HANDBOOK

OF LASER
SYNTHESIS
OF COLLOIDS

S. Barcikowski V. Amendola G. Marzun

C. Rehbock S. Reichenberger D. Zhang
and B. Gökce
Stephan Barcikowski1,2, Vincenzo Amendola3, Galina Marzun1,4,
Christoph Rehbock1, Sven Reichenberger1,4, Dongshi Zhang1,
and Bilal Gökce1,2

1 Institute of Technical Chemistry I, University

of Duisburg-Essen, Universitaetsstraße 7, 45141 Essen,
Germany

2 Center for Nanointegration Duisburg-Essen (CENIDE),

University of Duisburg-Essen, Carl-Benz-Straße 199,
47057 Duisburg, Germany
3 Department of Chemical Sciences, University of Padova,
Via Marzolo 1, I-35131, Padova, Italy.
4 NanoEnergieTechnikZentrum (NETZ), Carl-Benz-Straße 199,
47057 Duisburg, Germany

Cover & Internal Design: Sasha Ediger // designmodus.com

HANDBOOK
OF LASER
SYNTHESIS
OF COLLOIDS
CONTENT

Preface
Intro

Chapter I: Get started: Setup of Experiments and Starting Days 08

Step 1: Learn the basics

Step 2: Choose the target material

Step 3: Choose the liquid

Step 4: Choose the laser

Step 5: Choose the right optics

Step 6: Choose the ablation setup

Step 7: Find the focus

Step 8 Spice up your nanoparticles

Step 9: Now, just do it

Chapter II: Get more: Productivity 38

Step 1: Define nanoparticle productivity

Step 2: Cross the ablation threshold

Step 3: Maximize the penetration power

Step 4: Avoid fluid breakdown and nonlinear effects

Step 5: Dodge the cavitation bubble

Step 6: Keep it cool and thin

Step 7: Scaling it up – from batch wise towards continuous processing

Step 8: Lower the residence time using 3D-printing techniques

Step 9: Move between dimensions

Chapter III: Understand it: Size and Structural Characterization 62

Step 1: Prove that you actually have “nano” particles

Step 2: Measure their size

Step 3: Identify their morphologies

Step 4: Quantify them

Step 5: Analyze their surfaces

Step 6: Analyze their properties

Chapter IV: Keep it: Stability 110

Step 1: Define stability

Step 2: Increase colloidal stability

Step 3: Measure colloidal stability

Chapter V: Stay alive: Laser Safety 130

Step 1: Learn the risks

Step 2: Take precautions

Step 3: Turn on the laser

Step 4: Be organized

Step 5: Dispose of your waste

PREFACE

”We aimed at a new type of nanostructures, but we

really did not expect that the use of lasers could bring out
something revolutionary”, Anton Fojtik wrote in his abstract
for the 4th “Advanced Nanoparticle Generation and Excitation
by Lasers in Liquids – ANGEL” conference, 23 years after his and
Arnim Henglein´s pioneering work started this field in 1993.

Today, the field of laser synthesis and processing of colloids (LSPC)

has expanded to all over the world, so that whenever a researcher enters
in this field, she/he has to fumble her/his way forward amidst a mass of
scattered papers, focused reviews and Ph.D. theses all of which are not always
educational. This is why it seemed to us that one coherent presentation in
form of a handbook might help newbies (in the following referred to as “you”)
with this predicament.

The main purpose of this book is to introduce you to this interdisciplinary

topic, which contains elements from chemistry, physics, engineering, and
sometimes even biology. While LSPC is pretty simple to apply, it is not as easy to
understand, in part caused by its interdisciplinary nature. And to be honest even
among the experts in the fields there are many disagreements on how LSPC
really works. Although by now there is quite a collection of publications on LSPC,
none of them treats it in a way suitable for beginners unravelling the secrets of
experimental setups, reproducibility, colloidal stability, and yield.
Instead of aiming for a scientific and structural approach, which would
first establish the theory and properties of LSPC and then try to view the
consequences of these, we decided to arrange the book by phenomenological
“steps” that you will need to take to be successful in our field. In order to
visualize what really matters we tried to give instructional examples. Since
there is no need for you to make the same mistakes that we did (yes, we can
tell you!), we will also explain what you should avoid during your journey in the
world where ultra-small particles meet giant laser intensities. Of course, it would
be unreasonable to refrain from any theory if this really makes things simpler;
hence we limited the fundamentals to a minimum and refer to the scientific
reviews that are out there.

It was our intention to write all chapters of the book in a sufficiently leisurely
style, while only minor preliminary knowledge is required to read it. After
“getting started with setting up the experiments” in Chapter I and “getting more”
colloidal nanoparticles in Chapter II for the extensive “characterization” in
Chapter III, we explain how to “keep your colloids” stable in Chapter IV and how
to “keep you alive” during your contact with lasers and chemicals in Chapter V.
We hope that after reading the book you will not get lost in the zoo of materials,
liquids and instruments. Last but not least, we would like to thank all of the
25 scientists who contributed to this work by words and quotes. We hope you
enjoy reading this book as much as the authors enjoyed writing it!

Stephan Barcikowski and Bilal Gökce

INTRODUCTION

Two decades after the first colloidal nanoparticles were generated with a laser,
more than 350 institutes all over the world are actively involved in the synthesis
and processing of nanoparticles by laser ablation. If 3 students per institute
would work in this field we would have a total of 1000 students that are eager
to make their way into academia or industry using this laser-based method.
Different terminologies such as “pulsed laser ablation/fragmentation/melting
in liquids (PLAL/PLFL/PLML)” or “Laser Ablation Synthesis in Solution (LASiS)“
are used to describe the field that this book is about. If you want to know more
about LASIK (Laser-assisted in situ keratomileusis or simply laser eye surgery),
then you are wrong here and should consider reading another book. But if you
want to know more about the practical aspects of the exciting field, which we
call overall “laser synthesis of colloids” then you are at the right address.

The first chapter of this book starts as you would start your first days in the
laboratory, you would first read about the basics and then about the specific
experiment you plan to perform. Since the flexibility of this laser-based method
is very high you have a huge set of parameters that you need to choose from
and with each of these variations you will get a different result. If you don’t want
to waste expensive gold or platinum targets for your first “walking steps” you
might want to start with a cheaper and more abundant metal such as silver
or iron. What do you do next? You need to choose the liquid that you want to
use to collect your nanoparticles. However, for reactive metals such as iron
the liquid medium isn’t just a collecting medium it also defines the nature of
your synthesis product. If you use water you will get a mixture of iron oxide(s),
if you use an alcohol you will additionally obtain iron carbides. You see that the
choice is crucial. Let’s make a simple calculation to demonstrate the number
of possible outcomes: the period table tells us that there are 91 metals, you
can imagine that the number of liquids can’t even be numbered, estimates
range from 1018 to 10200, but the motto of this book is “keep it simple”, so let’s
limit our consideration to solvents. Solvents can be categorized into inorganic
solvents such as water or ammonia and organic solvents, which additionally
are divided into, oxygenated solvents (e.g. ethanol), hydrocarbon solvents (e.g.
hexane), and halogenated solvents (e.g. chloroform). If we now want to try five
liquids from each of these categories for each metal in the periodic table, we will
have almost 2000 possible combinations and possible outcomes. There is also
the laser, which will not just have an impact on the physical parameters of your
nanoparticles but might also change their chemistry. You see, you have to be
smart about choosing the right target, liquid and laser.
Let’s say you chose to synthesize iron oxide nanoparticles in water. If you use
a ns-laser and perform your ablation in a beaker for one hour you will get
micrograms of nanoparticles. This might be sufficient for UV-vis spectroscopy,
but if you want to analyze your nanoparticles properly, e.g. the crystal structure,
or even apply them in a real-word application you have to perform a lot better.
Chapter II will show you how you can boost your productivity by changing
your ablation chamber, laser parameters, fluidics or your target shape. You
don’t necessarily need to buy the most expensive laser in order to get more
nanoparticles, good news for your boss.

After you followed the steps in Chapter II and obtained, let’s say, 100 mg of
nanoparticles after one day of ablation you have to analyze your product(s).
Just by looking at your colloid you won’t be able to tell much, especially if you
ablate iron. How big are your particles? What is it? What is on its surface? These
are just some of the questions that immediately arise when you have a colloid of
unknown properties. Chapter III summarizes all major methods that are out there
to analyze nanoparticles or colloids. We will also mention practical aspects such
as how much nanoparticles you need or how you prepare your particles for the
analysis with each method.

You might not need to be a chemist in order to be successful in this field, but
you definitely won´t have an impact on the international community without
basic knowledge of chemistry. Especially colloidal chemistry is essential for
reasons such as to keep your colloids stable. You need stability for colloidal
characterization but also for further processing of your particles. If you have
aggregated particles on the bottom of your ablation chamber and the pure
liquid above, you won’t be happy with this outcome. So follow the steps in
Chapter IV and be happy!

We not only want you to be happy but also to be safe during your journey that
you’ve just started. The interdisciplinarity of laser synthesis of colloids is a big
advantage but from a safety point of view there are additional dangers, you have
to be concerned not only with chemistry but also with laser safety. If you don’t
want to perform unintended LASIK by your own – check out Chapter V. Many
things can happen during laser ablation in chemicals, you have to be extremely
careful. As you might imagine, flammable liquids don’t go well with powerful
lasers. We additionally included many “lab stories” from students, post-docs
and even professors that are evenly distributed over the whole book to help you
to realize what might get wrong during your experiments.
 CHAPTER I

GET STARTED

You probably know what it is like to start something completely new. You are

I highly motivated and at the same time know that all beginnings are hard.
Exactly here, this handbook comes into play. It will help you with tips and tricks
and will let you avoid typical beginners’ mistakes. You will not just learn how
to dodge the mighty cavitation bubble and tune your experiment but also to
study the exciting world of laser-generated nanoparticles. In order to start your
first experiments on the synthesis of nanoparticles by laser ablation in liquids,
you have to be well prepared. Thus not just experimental preparation is crucial
but also theoretical knowledge is of fundamental importance for a successful
work. At the beginning of this chapter the basics of lasers and laser ablation in
liquids are summarized. Then, the most important equipment and setup you can
choose will be described and you will learn how to manage your first steps of
the practical work when starting with your laser ablation experiments.

STEP 1: LEARN THE BASICS

It all started with a joke: “A laser is a solution seeking for a problem”, this is
how the first working laser was described in 1960 by sceptics who wanted
to make fun of Theodore Maiman’s invention. Maiman made the world’s first
laser operate in his laboratory at Hughes in Malibu. The laser was based on
an optical pumping of a ruby crystal using a flash lamp that generated pulsed
laser radiation at 694 nm. As you know every beginning is difficult, the same
happened to the father of optical lasers. Many researchers just did not realize
the significance and the dimension of his invention. After his first report on the
ruby laser was rejected by reviewers of the journal Physical Review Letters,
Maiman turned to Nature where his paper was published on August 6th 1960.
Even though the relevance of his discovery was not clear, Maiman believed in his
device’s potential. Nowadays lasers are everywhere and their applications have
become an indispensable part of our daily life. Just have a closer look at your
surroundings: applications can be found everywhere from research laboratories
to steel industry, from dentistry to medical clinics, from discotheques to
pointing devices. For example, lasers can be applied everywhere where an object
has to be quickly identified without direct contact. Cashiers only need to bring a
barcode close to the scanner without the need of typing the article number by
hand. Isn’t it great to have more time for shopping? At least for all women, this is
a great invention. Lasers are used for online shopping when you use a computer
mouse. In an optical computer mouse, a laser can identify the direction of the
movement. In addition, all optical disc drivers like DVD or Blue-ray use laser
diodes. Next to everyday products, pulsed lasers have gained huge importance
in medical applications such as eye surgery or photothermal laser resection of

08
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

organs. Huge industries have found multiple ways to make money from laser
technology including laser entertainment shows, cosmetics (laser tattoo and
hair removal) and measuring large distances by lasers, to name just a few
things. There is no car or airplane built without laser technology, either for steel
welding or for drilling the gasoline nozzles. In addition, pulsed laser structuring
is essential to build mobile phones and tablets. The skeptical colleagues of I
Maiman must have looked dumbfounded when they realized that indeed a lot of
problems have been solved by the laser.

Laser Ablation in Liquids (LAL)

Lasers are powerful tools that can lead to an astonishing outcome if the laser
beam is directed with high energy density onto a surface. You irradiate a piece
of gold in water and obtain a red-colored liquid. As a result of irradiating a
target, material detachment can take place, which can be exploited by scientists
for particle synthesis. What a brilliant idea, isn’t it? Particularly because the
technique was originally used to structure the surface of the target and the
particles were considered “just waste” for a long time. Focusing a laser beam
in a liquid environment, enabled to catch the ablated materials in the solvent
directly. This makes particle synthesis safer, because no fine dust or particulate
matter is released in the surrounding workplace, avoiding health risks, in
particular, respiratory diseases. But most important, colloids are easier and
safer to process into valuable products than nanopowders filtered off the gas
phase. The liquid is the key to quality and value of the synthesis. Fabrication
of nanoparticles by laser ablation in liquids has become an interesting and
important technique for many applications. Laser-generated particles are not
only very useful for fundamental research, but also for their widespread use in
medical devices, photonic materials, catalysts and more.
......................................

laser beam liquid

cavitation plasma
bubble plume

shockwave
bulk target
......................................
Figure 1: Sketch of what happens during LAL of a bulk target, just before the nanoparticles
disperse in the liquid

09
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

Considering the principles of laser ablation in liquids, nanoparticle formation

may take place by various mechanisms and is influenced by different laser
parameters (such as pulse duration, wavelength, energy, repetition rate) and
materials involved. When a laser beam irradiates a solid-state material in a
liquid medium, the energy provided by the laser pulse can be absorbed by the
I target and leads to the formation of an expanding plasma plume containing
the ablated material (Figure 1). This is accompanied by the emission of a
shockwave that releases energy in the surrounding liquid. When the plasma
cools down, it releases its heat to the liquid, which is transformed into hot vapor.
This leads to the formation of an oscillating cavitation bubble containing both
the ablated matter and the liquid vapor, where particle formation takes place.
After the collapse of the cavitation bubble, another shockwave is generated
and the particles are released into the solvent. In case of aqueous solutions
and materials such as gold or zinc, electrostatic stabilization of the particle is
then achieved due to partially or fully oxidized surfaces respectively, where ion
adsorption takes place and forms an electrostatic double layer. By this method
a variety of nanomaterials can be obtained purely in different solvents which
opens a wide range of applications.

Of course to fully understand the basics you would need to go into more detail
by reading one of the specific reviews that are out there.

Producing particles by LAL is quite simple. For the easiest setup, you just need
a laser, a vessel filled with the solvent and a target, which is immersed into the
liquid and appropriate focusing optics (Figure 2). The most expensive part of this
setup is the laser itself which costs at least 30.000 € if you want a reasonable
nanosecond laser. Maybe you share a laser with another lab, or look for used
ones. All the other costs are almost negligibly small, but also depend on the
target you use. Talking about laser ablation of gold in water the noble metal with
about 40 €/g (purity of 99.99 %) mostly influences the price. The remaining
equipment is quite cheap (a glass beaker, distilled water, lens). In total the
costs for consumable materials amount to less than a hundred Euros for many
weeks of fabricating colloidal nanoparticles. Search on the internet, you will
find that buying only some 100 ml colloid is often more expensive. Note that
for laser ablation not the whole target can be transformed to nanoparticles, but
can be recycled to a new target. By laser ablation of a target, you multiply the
value of your material. For example a liter of a colloid with 100 mg/L laser-
generated gold nanoparticles in water is available at least for 1,000 €, which
makes 10,000 €/g. In comparison, to obtain wet chemistry synthesized gold
nanoparticles (where a gold precursor is used and reduced to nanoparticles),
you will pay more and will get nanoparticles stabilized by citrate or other
ligands. Depending on the application, you need to remove the ligands on the
particles surface, which will again cost time and money.

10
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

......................................
laser
lens
LD
GO

target
I

beaker

. . . . . . . . . . . . . . . . . . . . . wat
. . er. . . . . . . . . . . . . . .
Figure 2: Basic elements needed for laser ablation in liquid

STEP 2: CHOOSE THE TARGET MATERIAL

Before starting with any experiments you should be well prepared. Take time to
think about the material you want to ablate and about the suitable solvent for
your research and application.

One of the main advantages of laser ablation in liquids is the possibility to

choose a variety of materials. So if you realize that for instance gold is not the
right material for your study, you just change the target and synthesize other
nanoparticles, maybe put the gold in an oven with silver and go for ablation of
alloys. Other chemists synthesizing nanoparticles would dream of it, because
simplicity and robustness of LSPC experimental setup makes the research much
more effective and more fun. Typically there are two different types of materials
that are mostly used to generate nanoparticles by laser ablation in liquids:
metals and oxides. But also other types of nanomaterials such as sulphides,
nitrides or carbides were successfully fabricated by LAL. We recommend to
go backward in material selection by the product you want, if its zinc oxide or
titania, it’s easier controllable to start with the oxide rather than going for metal
ablation and hope for controlled oxidation in water. In the latter case it is likely
that you will end up with a mixture of defect-rich crystal phases not always easy
to reproduce. Laser ablation in liquids does not just allow the synthesis of pure
materials such as metallic or oxide nanoparticles, but also alloy nanoparticles
possessing solid solution or core-shell structures are accessible depending on
the experimental parameter (target, solvent, solutes etc.) you use. For instance
in the case you want to have alloy nanoparticles you can just ablate an alloy
target. Try a series of AgAu, it gives beautiful colors, resembling the composition
of the alloy nanoparticles. The composition of the targets determines the
(starting) composition of your nanoparticles. However, you should carefully
chose your solvent, which strongly influences the formation and thus the mixture

11
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

and structure of the particles. Acetone is always a good start. Another fancy
approach for alloys is to use a self-made target. You can mix different powders
and press them to form mixed powder targets. Depending on the compressibility
of the materials and stability of the targets you can sinter the pellets by heating
them up subsequently (this also avoids breaking of the consolidated powder
I targets). In case the diameter of your laser beam is larger than the grain size
of your powder, you now will synthesize alloy nanoparticles. The same works
with other composites, such as mixing silver powder with an oxide. The key to
reproducibility is intense mixing of the powder in a mortar. Inspect color and
gloss of the pressed target, is it the same on both sides or has the particle
density and size difference caused inhomogeneities? This micropowder-based
preparation method is a simple approach to achieve alloys or doped oxides on
the nanoscale without much effort. It’s up to you which composition you would
like to take, because you prepare the educts as you want and it could be much
cheaper compared to bulk alloy targets, which are usually only available in
certain compositions.

Besides taking a solid bulk material for LAL, other shapes of a target can be
used. If only powder is available in your lab or you are either just too lazy to
press the powder or just don’t have a press, you can use a suspension of the
powder. This process is then called laser fragmentation in liquid (LFL). Main
disadvantage here is that after irradiation of the powder you need to separate
it from the colloidal nanoparticles. And also the determination of the particle
concentration is long-winded compared to bulk materials, where you just need
to weight the target before and after ablation.

~~~~~~~~~~~~~~~~~~~~~~~
Ina came desperat ely to me and said that the magnetic s tirrer isn´t working any more.
Af t er t es ting several others we found that the magnetic s tirring bar isn´t magnetic
and is changing it s appearance. Af t er smelling a mint y tas t e we realized that this is not
a magnetic s tirring bar but a tic tac. Imminently the ques tion raised who was willing
to sabotage our experiment s by put ting this tic tac into the box with s tirring bars.
Af t er detailed inves tigations and having several suspect s we found out that Galina los t
a tic tac several days ago and Ina finding it was wrongly identifying it as a s tirring bar ,
thus placing it in the box for the s tirring bars. The lesson is clear: tic tac´s should
get a new shape!

Marcus Lau and Ina Haxhiaj, Es s en

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

12
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

In case of powders you always need to separate them from the liquid and
it’s sometimes difficult to fully dry them before weighting. Sieve the powder
before use, to remove everything larger than 50μm, facilitating dispersion and
homogeneous ablation.

If you have already done some experiments by LAL with a solid target, you will
quickly realize that sooner or later you need to change the target because you
I
“burned” a hole into it. For a continuous ablation where you could synthesize
a huge amount of nanoparticles without stopping the irradiation of the laser,
a wire that is fed by a turning roll can be used. However handling wire ablation
is much trickier and needs some experience. You will find further information
on this in Step 6, where you choose your setup and Step 9 of Chapter II where
target geometry effects are explained.

But there´s one more thing to say. We always try to make our research
convenient, and as characterization of the nanoparticles after synthesis
consumes most of our time, we select the material always also by this criteria.
It’s a pity if you wait a long time for TEM or XRD data and it was the wrong
colloid that was analyzed: It’s easy to make 20 samples a day, but which to
select for in-depth analysis? It is very convenient if the material you ablate leads
to nanoparticles that have a color, since in this case you can immediately tell
if you had success or not. Examples include the metal nanoparticles (red gold,
bright yellow silver, brown copper) or the doped oxides, such as ruby. Colored,
doped oxide targets you may find in gemstone shops. Also crystal defects may
give colour, such as titania which turns nicely blue if titanium is ablated in water.

STEP 3: CHOOSE THE LIQUID

The liquid environment plays an important role in synthesizing nanomaterials
by LAL since it influences the nature of the materials you will get. Basically you
should have your application in mind. For instance if gold nanoparticles are
needed for biomedical applications, ultra-pure water (and maybe addition of
some albumin for stabilization in the saline biological media) is ideally qualified.
As you know, less noble metals such as copper may lead to oxidation, hence
alternatives are required if you want to obtain non-oxidized metallic particles.
Materials, which are oxidation-sensitive, can be ablated in the presence of
reducing agents or in solvents such as propanol, acetone, toluene etc. For
example, acetone works well with copper. However, you have to keep the
application in mind and you should be aware that some solvents (such as
toluene, and even sometimes acetone) lead to a formation of a thin carbon
layer on the particles’ surface. Working with solvents, which have a low flash
point and evaporate already at low temperatures, may cause flames if applying
a laser beam on the solvent vapor. Then a dust particle crossing the laser beam
in the vapor could be enough to ignite the mixture. In Chapter V, you’ll learn how
to deal with these powerful pulsed light sources. Mainly in case of non-noble
13
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

materials the pH value can influence the structure and the composition of your
material. If you look at the Pourbaix diagrams, which show the redox-potential and
pH-dependence of a certain material, you will note which oxides can be formed
at a certain pH. Furthermore, the pH strongly influences the colloidal stability (as
will be seen in Chapter IV). As a rule of thumb, stabilization of (noble) metals is
I achieved at high pH, and oxides often go well at neutral pH (not too low, to avoid
dissolution). Some oxides that tend to dissolve at low pH can also be stabilized
at very high pH, e.g. 11. Check the isoelectric points of the product in the web
(already the value for the bulk gives a good hint), and keep the pH away from
it. If you run into stability problems and notice aggregation of particles the next
day even though you have tried different pH, you may like to read the chapter
on particle stability. Just one word on that: add a trace of soap or protein. More
precise, add a pinhead piece of curd soap or add albumin (works with almost
everything). Of course, never use tab water, only distilled or deionized water.

As mentioned in the previous part, the liquid environment you use also influences
the formation of alloy nanoparticles or core shell structures. This can get
complicated. For instance laser ablation of an alloy target with an oxidation
sensitive material such as iron and noble metal gold (equimolar Au:Fe) leads
to a formation of an unoxidized elemental iron core with a gold shell in organic
solvents. In contrast to this, LAL in water form a gold core surrounded by an
oxidic iron shell. Just to name a few of other approaches for particle synthesis
with a core-shell structure: a quite less investigated method is the ablation of a
target which is already immersed in a colloidal solution with particles. But also
a combination of LAL and chemical reduction is possible. In that case you can
synthesize your particles and add a precursor solution or you directly ablate a
target in the presence of a precursor, this is then called reactive laser ablation
in liquid. Try ablation of a less noble element in a metal salt solution of a nobler
element, the ablated species will reduce the precursor and form a composite
nanoparticle. That’s advanced colloidal nanoredoxphotochemistry done in
1 minute (and maybe understood after 1 year). No worries, in this handbook w
e will keep it simple and most rules apply as well to the more complex material-
liquid combinations.

STEP 4: CHOOSE THE LASER

The literature is full of papers dealing with laser ablation in liquids and often many
different laser parameters are used, so that for beginners, it’s not really clear why
exactly that laser with that parameter was used. To be honest, mostly the laser,
which is often the only one available in the lab, is used while the parameters are
optimized for this given laser. But if you have an option to choose a certain laser
you could pay attention to some general details.

It is mostly established to use the fundamental wavelength of 800-1064 nm

for laser ablation in liquids, but also green light with 532 nm is often used
14
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

to ablate a material. However, if you want to prevent two different mechanisms

that take place simultaneously, the fundamental wavelength is more appropriate
to synthesize nanoparticles by laser ablation in liquids. For instance if you ablate
a gold target with 532 nm in a vial, where the concentration of the particles
increases and form an optically dense liquid, the particles will be re-irradiated.
At this wavelength e.g. gold nanoparticles can absorb a high amount of the laser I
energy because of their intense plasmonic resonance band at about 520 nm
(depending on particle size). Hence ablation of the target as well as re-irradiation
of the particles take place simultaneously, affecting reproducibility. In addition,
the colloidal solution efficiently swallows the laser energy and your productivity
will quickly decrease (as will be addressed later in Chapter II). To avoid this,
ablation synthesis at wavelengths where the particles show minimal absorption of
the laser light is advisable for better control and understanding of the mechanism
that takes place during ablation. In contrast, if you hunt for smaller particles,
short wavelengths and long irradiation times may shift the process balance from
ablation to colloid excitation, at the expense of productivity.

~~~~~~~~~~~~~~~~~~~~~~~
… everyone was quickly leaving the building , and heavily dressed fire workers seeking
the fire source. Smoke sensors point ed at my laser lab, and the surprised firefight ers
found my gues t scientis t (a renowned gues t professor ) smoking . He lat er said he was
blowing the cigaret t e smoke over in order to visualize the laser beam during it s
alignment on
the optical table.

S t ephan Barcikowski, Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Before talking about laser power, energy and intensity it is useful to understand
what these terms refer to. The power of a laser can be measured in Watts.
This refers to the mean power output of the laser. The definition of the mean
power output Pmean is simply calculated as the Energy Emean released during
the operation time top (Pmean=Emean /top). For instance as shown in Figure 3 if
a continuous wave-laser emits light with an energy of 3,000 J in 10 minutes,
the laser has a mean power output of 5 Watts (3,000 J / (10 · 60 s)). When
using pulsed lasers the situation is much different as there are two separate
definitions of the laser power possible. Remember that a pulsed laser releases
its energy in packages (laser pulses) during the operation time top with some
repetition rate (e.g. 20000 pulses per second) having a given pulse duration
tpulse. If the pulsed laser fires 20,000 (= 2 · 104) pulses per second having the
energy amount of e.g. 250 µJ (= 2.5 · 10-4 J) per pulse, its mean output
power simply equals to 5 W (2 · 104 [pulse/s] · 2.5 · 10-4 J [J/pulse]).

15
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

That´s the same average power output as the cw-laser before. But the cw one will
not even make the target hot whereas the pulsed one is already quite productive
in colloid synthesis. Accordingly, due the compression of the energy into pulses
of a specific pulse duration tpulse (e.g., 1 ns and 1 ps) the stream of single pulses
is by far more efficient in removing matter from a solid surface. You wonder
I why? Imagine two situations: you are the target getting irradiated with a 250 µJ
laser pulse of A: 1 ns pulse duration and B: 1 ps pulse duration. In case A you
absorb 250 µJ of energy within 1 ns which means that the laser is irradiating you
with a peak power of 250,000 W (Ppeak = Epulse/tpulse = 2.5 · 10-4 J / (1 · 10-9 s)
or in another unit 0.25 MW. This peak power Ppeak is comparable to the added
power of about 20 single-family houses. Now it’s better to stop imagining being
the target as we get to case B. In case of a 250 µJ pulse having a pulse duration
of only 1 ps the peak power (or pulse power) impacting the target equals
250 MW which basically is the amount of energy a small power plant produces.
Practically, nanosecond lasers often have similar peak powers than picosecond
lasers, as ns lasers provide mJ pulse energies and ps laser pulses are in the µJ
range. E.g., 200 mJ @ 10 ns or 200 µJ @ 10 ps will both shoot with 20 Megawatt
peak power. But a typical femto laser with 100 µJ @ 100 fs will allow you to fire
with 100 MW, allowing to disintegrate every solid (with the risk of vaporizing the
liquid before reaching the target).

Of course this peak power only affects the target for a very short time however
with a tremendous effect we call ablation. But do not underestimate the target’s
defense mechanisms, which mainly are light reflection and energy dissipation
(for more details refer to Chapter II). In order to crack the defense mechanisms
of the target you need to increase the penetration to use the full force of your
single pulse. To do so you need to decrease the lateral extension of the pulse,
which basically is the beam area. To achieve this, lenses are used to focus the
laser beam. As depicted in Step 5 the other option is to use a telescope of which
two telescope types namely the Kepler and the Galilei telescope are especially
famous. The advantage in using telescopes is that the light wave still propagates
in the same direction (parallel beam) while a focusing lens will result in a
diverging beam. The telescopes are limited by the destruction limit of the lens
material, which is why they can’t be used in cases where very small spot areas
and high pulse energy lasers are being used. Now, upon decreasing the beam
area two properties of light called “intensity” and “fluence” will increase.
Intensity is simply defined as the power (either mean power or, in our
community, the peak power in Watt) per beam area (in cm2). On the other hand,
laser fluence is defined as the beam energy (in J) per area (cm²) or speaking in
formulas:

16
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

Hereby F depicts the desired laser fluence, A the laser beam area and represents
the energy of a single laser pulse. The pulse energy EP can be obtained by dividing
the mean output laser power measured by the number of laser pulses per second.

pulsed laser cw-laser I

1s 1s
power

=0.15

0.5 s 1s
=> 50% on =>100% on

Figure 3: Comparison of laser power of a pulsed laser with a continuous wave laser beam.
The average power of both are the same so that the peak power of the pulsed laser is far higher.

The definition of the laser fluence only applies to pulsed lasers as it is based on
the pulse energy as shown above. The laser fluence is a key parameter for all
laser processing of materials in liquids, for laser ablation, laser fragmentation,
and laser melting. The fluence rules the productivity during laser ablation,
sets the smallest size you can achieve by laser fragmentation, and is the main
parameter to control particle size by pulsed laser melting in liquid. Hence, it is
of utmost importance for the sake of reproducibility to precisely measure this
parameter “laser fluence. And if you want to be kind to people that like to follow
your footsteps, you will need to write down how you measured it. This however
requires to measure the pulse energy and the spot area to be able to divide
both values. The pulse energy is acquired by a power meter at the beam outlet,
to be divided by the repetition rate. Better measurement position is a close to
the ablation chamber as possible (e.g. behind the last mirror or just before the
focusing optics), as energy losses accumulate at every optical element. Don´t
consider taking the value that is written on the laser or the last PhD thesis in
the lab. Pumping lamps or diodes naturally degenerate, so that laser power
decreases every year. Also, for most lasers laser power will be different in the
morning than one hour after laser operation, because it needs to “warm up”.
We recommend measuring laser power 30 minutes after the laser resonator is
on. Now you need to know the spot area on the target (for laser ablation) or the
liquid entrance (for laser fragmentation/melting). Don´t wonder if you will find
very different values in literature regarding the laser fluence for the same setup
(e.g. same laser, same focal length), because measuring the fluence is done
differently. In principle, there are 3 ways of doing it.
17
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

1) The “theoretical fluence” you get by simple calculation based on focusing law
(input parameters are raw beam diameter, wavelength, and focal length). We
recommend then to measure at least the raw beam diameter before the lens as
it often deviates from the value the system manufacturer provides. For this, you
take a beam detector card and hold it into the beam (read chapter on safety
I before). If you don´t have a beam card, (thermosensitive) telefax paper or thin
black paper does a good job, too. Wetting the telefax paper before is better
choice if you have high pulse energy to avoid non-representative big holes and
smoke. 2) The same you can do to acquire the “measured fluence at target
plane” at the distance from the lens where the target is going to be placed.
Maybe you put the paper on top of the target. Test how many pulses you need
to see a hole. But this will probably not work in liquid. 3) The most suitable way
to determine the fluence is to measure it at the target in liquid, that’s the closest
to what you want to know. In order to obtain the spot diameter based on this
setup, you will need to set a limited number pulses (e.g., start with 10 and 100)
and measure the average crater diameter in the microscope. Very often, a larger
heat affected zone surrounds the crater, but for calculating the fluence value,
we look for the ablated crater diameter.

In case of continuous laser beams the impact of a laser can only be measured
as a function of laser intensity, which was introduced above prior to the
fluence. If the laser intensity is high enough, it is possible to ablate a target
with continuous waves, however as already indicated while a 5 W pulsed laser
can create peak powers of, e.g. 250 MW (in case of 1 ps pulses with 250μJ.)
the continuous wave laser will still only operate at 5 W. That’s why in cw laser
manufacturing like laser welding or plate cutting, kilowatt laser output power
is standard today, whereas for pulsed laser machining tens of watt are often
enough. As hinted before the target has several “defense mechanisms” one
being the energy dissipation. Against continuous wave lasers this works perfectly
as the energy is continuously delivered it can be continuously dissipated
resulting in slow heating and eventually melting of the target. The reason for
this is that the different heat exchange mechanisms are influenced by the time
scales when irradiating the target and thus it is affected by the pulse duration
of the laser. The pulse duration mainly governs the heat affected zone (and with
it the ablation mechanism), which is higher for nanoseconds (ns) pulses than
for pulses much shorter than the heat diffusion time (fs to ps pulses). While
longer pulse widths such as in nanosecond scale give ability to heat transfer
into the liquid via the heated target surface, ultrashort pulses such as in pico
(ps) or femtosecond (fs) scales are so fast that heating of the environment
can be minimized. For instance, if you operate with ultrashort pulsed lasers,
the ablation rate is so fast that the solvent even does not realize if the target
is being hit by the laser (no heat transfer) and stays cool, which enables laser
ablation in solvents with low boiling point. Furthermore, because of energy

18
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

losses to the environment in case of ns pulses, the ablation threshold is higher

compared to ps pulses. The threshold fluence decreases with pulse duration
(< 100 ps, better < 10 ps) because of negligible energy dissipation by thermal
conduction, and thus leads to higher ablation efficiencies. That’s the theory,
but in practical application, also ps and fs laser ablation will heat up the target
slightly for longer ablation time, caused by accumulation of minimal residual I
heat that every pulse leaves at the target. It is not much, but noticeable for high
repetition rate lasers.

Even though the ablation threshold is smaller for ultrashort pulses, these may
cause an optical breakdown, namely a strong local ionization of a medium (the
liquid) due to energy absorption. For better insights look up in Chapter II, which
describes the effect of the pulse duration in more detail and nicely illustrates
these phenomena.

~~~~~~~~~~~~~~~~~~~~~~~
Recently at the T EM I could observe the presence of impurities in nearly all of my
colloids consis ting of an organic solvent. According to EDX measurement they were
made of silicon compounds and were present in a comparable concentration to the
nanoparticles. The impurities int eract ed with the electron beam of the microscope
and covered up the particles, no possibilit y to get nice images. Different sources like
air dus t, dirt y grids or production additives of the solvent s s tood in suspicion. Then a
mass spectroscopy analysis of different organic solvent s revealed high amount s of
polysiloxanes in a p.a. grade acetone I normally used for synthesis.

Friedrich Waag , Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
However ns lasers are usually available with a high pulse energy and high power
at low or moderate repetition rates and could be more appropriate concerning
high productivities because of another important point which influences the
ablation efficiency, namely the cavitation bubble. The bubble, which is caused
due to a laser pulse, can shield the following laser pulses and thus prevent
that energy reaches to the target. As a result less material can be ablated. It
should be noted that the lifetime and the size of the cavitation bubble depends
on the pulse energy, which is usually smaller for ps and fs lasers. The bubble
typically has a lifetime in the (hundreds of) microseconds to (single) milliseconds
regime, while its size is in the (hundreds of) micrometers to (single) millimeter
regime. Hence, the laser-cavitation bubble interaction takes place only at high
(> kHz) repetition rates, where the temporal pulse distance reaches values of
the bubble lifetime. To prevent that the cavitation bubble interferes with the
laser beam a spatial or temporal pulse separation is needed in order to get
higher productivities by bypassing the bubble. If you use a < 100 Hz laser, you
definitely will not have to worry about any cavitation bubble shielding.
19
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

However in case you think about ablation of a pressed powder target, mild
conditions are favorable to prevent a removing of the initial bulky particles.
High pulse energies (typical for low repetition rate lasers) are more likely causing
partial disruption of a pressed pellet. As you see, the laser parameters strongly
depend on your interests and the materials or liquids you use.
I Table 1: Pros and Cons of several laser pulses for pulsed laser ablation in liquids

Laser pulse Pros Cons

duration

• more and cheaper high power

lasers available (i.e. high • heat transfer from target to the
ns productivity) liquid
• high power at moderate • less efficient
repetition rates

• ‘gentle ablation’ • optical breakdown at high pulse

energies
• insignificant heat transfer to
liquid • high power at high repetition
ps rates (bubble shielding)
• compromise between efficiency
and productivity/costs • high power more costly than ns
• high power systems available

• ‘gentle ablation’ • high power at high repetition

rates (bubble shielding)
• very efficient ablation per pulse
fs
• optical breakdown
• insignificant heat transfer to
liquid • high power more costly

STEP 5: USE THE RIGHT OPTICS

Working with a laser requires at least basic knowledge about laser optics, since
they need to be chosen carefully. The main items are lenses and mirrors. The
area of optics is a perfect playground and you can tumble-around especially
with the variety of lenses. Starting from the simplest lens, common lenses to
focus or to spread the laser beam are shown in Figure 4.
......................................
plano-concave

bi-concave

plano-convex

......................................
Figure 4: Types of lenses: Plano-concave, bi-convex and plano-convex

20
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

Two main orientations should be kept in mind, namely convex and concave.
A convex lens is curved outwards and a concave lens inward. In order to
memorize the geometric structures you can help yourself by thinking on flat
surface, which is curved inward, to make a little cave (concave) or outward
to make humps like a T-Rex (convex). To focus the laser beam and thus
concentrate the entire laser power to a small spot a convex lens is used. You I
can decide whether you want to bundle the beam at long distances by using a
plano-convex lens or at shorter distances if working with a biconvex lens, which
is symmetrical on both sides (collecting lens). However focusing a laser beam
with a plano-convex lens leads to reduced spherical abberations (“sharper”
focusing) and is mostly used for laser ablation, whereby the curved surface of
the lens is faced opposite to the direction of the laser beam (Figure 4). Concave
lenses can be used to expand the laser beam or to increase the focal lens in
an optical system (diffusing lens). For instance if you want to change the laser
fluence (remember that this is energy per area and the entrance key to any
laser synthesis castle!) you can change the spot diameter by using such a lens.
Now you may ask, how can the beam be expanded to keep the beam parallel?
For this a beam expansion system can be assembled by combining two lenses.
You can build a beam telescope by displacing the concave and convex lens
against each other. There are two main assemblies. A Galilei telescope consists
of a collecting and a diffuse lens, whereas a Kepler system is build up with two
collecting lenses (Figure 5). How much the beam is expanded depends on the
focus of the lenses.

......................................
Galilei kepler

......................................
Figure 5: Systems for parallel beam expansion: Galilei (left) and Kepler (right)

21
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

As you can see in Figure 5 using a Kepler telescope a focal point is created
between the two lenses. If the laser is strong enough even the air can be
ionized and forms a plasma in that point, which can be clearly observed as
a bright white-purple spark of light. To avoid losses of laser power, Galilean
telescopes are more frequently used (Figure 5, left). Of course, you can use
I it in both directions, making a raw beam wider or smaller, keeping it parallel
before and after the telescope. Now think that you want to increase the laser
fluence. A plano-convenex lens will sharply focus the light to the highest fluence
(e.g., to increase productivity during laser ablation, see Chapter II), and the
fluence will be higher the larger your raw beam diameter is (better focusability),
and the shorter the focal distance (the supplier sells focal lenses with fixed
focal distances, very typical are 50, 60, 100, and 300 mm) the sharper the
spot. But sharpness also goes along with divergence around the spot, that is
high fluence gradient before and after the focal plane. This will increase the
required precision of target positioning at a defined distance, and would make
re-adjustments needed once the target gets thinner. Very often, we are quite
happy with 60 or 100 mm for laser ablation synthesis of colloids. When you read
papers on laser fragmentation or laser melting, often very long focal distances
(e.g. 300 mm) are used in order to minimize fluence deviation (sharpness) along
the beam path in the liquid. Hence, the thicker the liquid (e.g., 10 mm) you pass
during laser fragmentation or laser melting of suspended particles in liquids,
the longer the focal distance of the lens should be to allow similar condition
everywhere along the irradiated volume. For this situation, the Galilei telescope
is a good deal, as it allows to conveniently adjusting the fluence while keeping
the beam non-divergent, so that the fluence is always constant no matter where
you put the cuvette.

~~~~~~~~~~~~~~~~~~~~~~~
“A s trange smoky smell comes from laser lab!”, told me Fabrizio. Few seconds lat er
I realized for the firs t time that a lit tle aspiration chamber is not enough to avoid
the ignition of solvent s during laser ablation in alcohol! The holders of lenses and
mirrors (made of plas tic) were melt by a flame developed over the ablation flask .
From that time I always performed laser ablation in sealed cells, in cases of
flammable solvent s…

Lucio Lit ti, Padova

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Compared to a ‘standard’ lens as described above which focus a laser beam
on a spherical area an f-theta lens allows focussing on a larger planar surface.
This enables a steady focusing in one plane independent of the incoming
position of the beam on the lens. For instance, in case of LAL a focal point will

22
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

be always on the flat target and the same energy density reaches the target.
This is very useful if using a scanner system, where the beam is always scanned
and thus moves in several positions on the lens. This leads to a minimized
fluence deviation when scanning the beam by a galvanometric laser scanner, in
particular at short (< 100 mm) focal distances and large (> cm²) target areas.
...................................... I
f-theta lens s tandard lens

......................................
Figure 6: Comparison of the focal position using an f-theta (left) and a standard lens (right).

If you are working with lasers, you cannot avoid using mirrors. Mirrors are useful
to reflect the laser beam in a desired direction without changing the position of
the laser beam source. But, here you have to note a few things when working
with mirrors for instance in combination with high-energy lasers. In all cases,
you need to make sure that your lens and mirrors are suited for the working
conditions you want to apply. Important questions you need to answer before
using the optical device by checking the manufacturer specifications:

Is the wavelength you are planning to work at suitable for the

optical device?

What is the damage threshold of the optical device at the

wavelength and pulse duration you plan to work at?

Does the int ensit y you plan to work with exceed the damage
threshold?

Are the optics suitable for cw or pulsed lasers?

The story behind these questions is that optical devices are sealed with an
optical coating, which is optimized on a specific wavelength regime. Hence, it is
important to check whether the devices are suited for the given wavelength to
make sure they even perform properly. Secondly although the optical devices

23
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

are made of glass (e.g. BK7 or Quartz glass) a little amount of the laser energy
(<0,5%) will be absorbed, depending on the wavelength and the type of glass.
Therefore, there is a laser intensity (peak power per raw beam area irradiating
the mirror or lens) at which the glass will get damaged which is specified by the
manufacturer as the damage threshold. Making sure that the working conditions
I are in specification of the optical devices is essential for a successful ablation
(ask provider for information on: suitable wavelength, maximum raw beam
diameter, and threshold fluence).

You can differentiate between two main different types of coatings. Shiny metals
such as gold, copper or bronze have been used as mirrors since ancient history.
Nowadays glass is coated with a tiny layer of metals such as silver or alumina.
The reflectivity of metallic coated mirrors is determined by the physical
properties of the metal. Using mirrors with a metallic coating are quite cheap
but less resistant to high laser energies, thus dielectric coatings are mostly used
to synthesize particles by LAL, where strong laser beams are common.
By using dielectric coatings defined optical interference between the reflected
and transmitted light can be set and the reflectivity can be precisely adjusted by
the thickness of the coating layer. Each layer reflects the beam and weakens the
intensity that reaches the next lower layer. You already know that absorbing laser
energy lead to ablation of a substrate, the same can happen to glass.

~~~~~~~~~~~~~~~~~~~~~~~
Once we focused with a green Nd:YAG laser into a cylindric glass vial (one of the cheap
ones with the snap-on lid) from the side for laser pos t-irradiation of a platinum
colloid. Of course , we did not see the beam because we were wearing safet y glasses,
but we had a camera on and looked at the pict ures lat er . There was a met er-sized laser
light wing reflect ed into the half lab, on half body height. Luckily, the laser was not
on full power that day and no one was too close to the vial .

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Thus, the dielectric coating are used to prevent energy absorption due to many
reflective layers. A good deal is to buy mirrors with “HR” = high reflectance
coatings. Are you then on the safe side? Yes, but only when you avoid dust
particles to settle on the optics. A dust particle is an absorbance center and
will cause spikes at the optics that again act as even more effective collecting
centers. When you shut the laser down, put coverings (simple plastic bags) over
each mirror. When you notice that laser power is lost after a while, it’s very often
the optics that need to be cleaned. Imagine a single mirror “consumes” only 5%
and you have three of them aligned, what you get at the end is the cumulative
loss, mathematically 0.95 x 0.95 x 0.95 = 0.85 energy yield. Small thing, big

24
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

effect, and easy to avoid. The thickness of the layer is in nanometer range and
thus the coatings are sensitive to any scratches. You need to be careful not to
destroy the layers if you want to clean the mirrors. For this purpose specific
optical cleaning supplies are commercially available. These will be useful many
years for the whole laser lab. For instance, extremely soft tissues can be wetted
with very pure methanol and glided carefully over the lenses or mirrors without I
scrubbing it. If you just need to remove only the dust on your optics air duster
that blow the dust away are suitable.

STEP 6: CHOOSE YOUR ABLATION SETUP

Besides the simplest setup described in Step 1, several experimental
configurations are possible to perform particle synthesis by laser ablation in
liquids. To ensure uniform mixing of the dispersed particles in the solvent
a magnetic stirrer can be added. For a high utilization of the target, the vessel
or a target holder can be moved during laser ablation. Some examples of the
setups with different kind of sample holder are sketched in Figure 7.

Of course, it is also possible to penetrate a vessel with the laser beam and
for instance to work with closed setups with flammable solvents. In that case
you need to be careful not to damage the glass. You should carefully choose
the distance of the laser focus, which is not to close to the glass so that high
fluences are reached only in proximity of target and not on the glass window. For
a higher stability it is recommended to use quartz glass (like a cuvette) or
a self-designed chamber with a coated window that enables to minimize surface
reflections and thus any loss at the optical surfaces due to reflections.

motor
A B C D
focused
laser
beam

magnetic
solvent target
s tir bar
t urn
target table magnetic
nano- s tirrer
particles motor

Figure 7: Pulsed laser ablation in liquids using different setups in a batch operation mode.

25
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

Thinking about long-term particle production, where a large amount of particles

and solvent are required, a continuous ablation process can be quite useful.
Sounds complicated, but in fact it makes multi-milligram scale synthesis easier
I and more robust. Here, the solvent flows over the target and carries the just
formed particles (and the bubbles) away from the ablation zone and collects
the product in a storage vessel (Figure 8). More detailed setups are show in
Chapter II. As long as the target is not perforated you will have a continuous
ablation process, however if you need to change a thin target often this can
take longer than in a batch process. Most importantly, it enables a so-called
steady-state, where concentration does not change with time in a given volume
everywhere downstream. You may imagine that steady state synthesis conditions
favor steady state particle growth conditions, improving reproducibility. A gold
plate could feed a laser ablation process for many hours, and all you have to
do manually is the taking the vessel at the end, it is like tapping nanoparticle
colloids (see: http://youtube.com/nanofunction). Since time is money, and
being lazy is linked to creativity (in particular for sunshine-affine Italian guest
scientists), researchers also study how to optimize the process by wire ablation.
This technique enables a fully continuous synthesis of nanoparticles. For LAL the
wire may be either just dipped and fed into an ablation chamber or run without
a chamber being surrounded by a concurrent liquid jet. In the setup a motor
enables to move the wire continuously through a cannula (a tip broken from a
syringe) with a desired velocity (Figure 8). The setting of the feeding speed then
should be adapted for the entire wire thickness is ablated. By this method, a
higher productivity can be achieved due to different cavitation bubble dynamics
compared to the processes involved during LAL of a bulk target. The mass that
is ablated can be easily pre-calculated and pre-set by knowing the length of
the wire that is ablated. Note that many metals are available in wire shape at
no extra costs, and a roll of silver wire can run for a long time. A disadvantage
of the ablation of a wire is that process stability is crucial. For example, large
micro to millimeter nuggets can be removed from the wire if the feeding rate is
too high, and thus not the whole material can be converted into a nanoparticle
colloid. Hence using this setup is more convenient for experienced operators,
because you have to optimize your experimental setup very precisely.

A B motor [mm/s]

laser beam
wire

liquid jet

nanoparticles

solvent

Figure 8: Pulsed laser ablation in liquids using a continuous flow system (A) with a wire (B).

26
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

A B
laser
window beam

electroly t e
solvent

I
solution

time molecule
delay s tream

Figure 9: Continuous ablation with the possibility of particle property control by (A) delayed
conjugation and (B) in the presence of electrolytes in a flow-through chamber.

For many applications such as biology or catalysis small nanoparticles are

desired. In case you want to do laser ablation in a continuous operating mode
and tuning the particles properties by adding molecules or specific ions you
can construct a chamber with inlets. There you have a reservoir where you add,
e.g. surfactant, albumin, or salts. When pristine particles are synthesized by LAL
in liquid flow, they tend to grow on in the stream. In case a quencher molecule
is added, this growth stops immediately and the particles become smaller.
This can be utilized by using a flow-through reactor where quencher molecules
are added after certain time delays. The time delay within the reactor can be
controlled by the flow rate and by the distance between the point of particle
generation and the point of quencher addition (for more detailed information
please check Chapter II). Then the size can be controlled by the downstream
position of the delayed injection (Figure 9 A) or in situ by adding electrolytes or
stabilizers to the feed (Figure 9 B).

Overall, for experiments where you only need to do ablation for a couple of
minutes or need to change the target often, the batch chamber is in most cases
the best choice. It is always worth to add a mixing effect (magnetic stirrer, fast
rotating target, …). If you have to run laser ablation longer and have thicker
targets, add a pump to create a continuous flow. Depending on concentration
you want, the flow is easily switched to run recirculated or run in one path. One
path flow creates steady state for particle formation, increasing reproducibility.

STEP 7: FIND THE FOCUS

You may imagine that irradiating something with a laser beam e.g. such as a
laser pointer, does not automatically lead to ablation of a material. The criteria
for doing so include an adequate intensity of a beam. Using a high-energy laser
and focusing a beam on the target enables a high local fluence, due to the
bundled energy in a tiny spot area. In this manner, the ablation threshold – the
minimum fluence at which ablation takes place – can be overcome and the solid

27
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

can be vaporized. This threshold defines the minimal amount of energy per area
needed in order to heat up the target strong enough, such that material gets
vaporized and turned into plasma. For most materials this threshold typically
will be found in a range of 1-10 when using the very common nanosecond,
fundamental Nd:YAG laser wavelength of 1064nm. To focus the beam onto the
I target a convex lens is required. The focal position can be found by moving the
target or the lens, which has a certain focal length. But how does it work, to ‘find’
the focal position? In the field of laser materials processing many parameters
are used to describe the laser beam, focal position etc.. Since we do not want to
complicate the first steps, let’s keep it simple.

You knew that you have to use your hands and your eyes for pulsed laser
ablation in liquids, now you learn that you can use your ears as well! The
probably simplest and fastest possibility to position the focus onto the target is
by hearing the intensity of the sound, which a laser causes if ablating a material.
If the laser beam hits the target with an appropriate intensity, it starts to ablate
the material and makes noise. At the position where you hear the loudest noise,
the highest ablation rate can be affirmed due to highest effective fluence at
the focal position. However, for systematic research studies, this method is not
accurate and the sound intensity depends on the bubble cavitation (the sound
frequency you will hear in liquid is not the laser pulse frequency). So probably,
your boss won’t hear a sound, while you will hear more precisely high frequency
noise just because you are younger than he or she is. If you don’t believe it, just
test how old your years are by hearing to several frequencies in the internet.
You’ll find a lot of different self-tests at YouTube, simply by typing ‘how old are
my ears’.

Figure 10: Sketch of someone, who is listening to “laser music”

To find the focus in a way independent of your ears, a microphone can be used.
The magnitude of microphone signal linearly correlates with the laser fluence
hitting the target. You don’t need a special microphone system, if you want to

28
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

keep your approach simple. In case you want to make sure the productivity is
high, you can easily do that by using a basic ‘singstar’ microphone e.g. coupled
with a readout software. And look for high frequencies, maybe around several kHz.

I
~~~~~~~~~~~~~~~~~~~~~~~
I planned to do some UV-Vis measurement and I washed a cell . Af t er that, I want ed to
dry it as fas t as possible so I used a nitrogen blower . Since I didn’ t think of the gas
pressure and hence I underes timat ed it so that the cell blow up in my face and broke.
It was really high pressure!

Mizutaru Tsukasa, Tsukuba

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Back to other possibilities for an appropriate focal positioning, you can use
an indirect but very result-oriented method to find the properly focal distance
to the target. That is by mesuring the productivity. You can provide a series of
ablation experiments, where the distance between the lens and the target is
varied. The obtained colloids, which are synthesized at several positions of the
lens to the target, can then be analyzed by UV-Vis spectroscopy (see Chapter
III). Oftentimes (depending on the material that is ablated) the extinction of the
spectra at a certain wavelength correlates with the particle concentration. E.g.,
in case of gold the extinction at 380 nm increases with a higher production rate.
If there is no specific interband wavelength, use the intensity of the shortest
wavelength in the spectrum where the solvent does not absorb. By varying the
position of the target relative to the lens, you can find the maximum of the
extinction and by this way the focal distance. Alternatively just weighting the
target before and after ablation works as well, which is more appropriate for
several materials. An important thing to note is, before doing any laser ablation
tests, the laser spot should be located only on the target and not e.g. on the
ablation chamber itself. For this, first position the laser beam with a low energy
or use a positioning laser, which is directly integrated into some laser systems.

Once you know which materials you need and how to find the focal position,
you should be able to synthesize nanoparticles by pulsed laser ablation in
liquid. What do you think will happen if you perform laser ablation of the target
for a while - let’s say 5 minutes (note that the time depends on the repetition
rate, pulse duration, fluence and so on)? You’ll be able to efficiently drill a hole
into your target, however this is not what you call efficient use of material and
you will also ablate the material that is behind the target and get impurities in
the colloid. In order to avoid a fast hole-formation and thus contaminations,
two approaches are possible. On the one hand, the target can be mechanically
moved e.g. by circular movement of the whole chamber. By this way, you may

29
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

exploit a larger surface of the target that makes it possible to use it for longer
ablation time without the need of replacing the target.

A much more elegant but also much more expensive solution is the continuous
shift of the laser beam on the target, either by using a motor stage moving the
I chamber or by using a rapid scanner system moving the beam. By using a scanner
system (it’s always faster than most axis systems) the cavitation bubble can be
spatially bypassed and thus a higher ablation rate can be achieved, since the
laser beam is not shielded from the bubble. A scanner system can create different
ablation pattern. The simplest form of scanning patterns are stripes (Figure 11 B).
However if one strip is scanned to the end, the mirrors cause a certain delay time
by moving back to start the next strip. During this period, no pulse is coupled and
thus the productivity is influenced. For example in case of a 50 mm long strip and
a scan speed of 1 m/s, the scanner needs 0.05 s to move this distance back. By
working with a repetition rate of 5 kHz about 250 pulses cannot be coupled in this
time. Second, unless you have an advanced “on the fly” scanner, the software
will accelerate and decelerate the beam guidance mirrors at the beginning and
at the end of the line, respectively. This will cause deeper ablation there, and
again the target is not ablated homogeneously in the whole area, making early
target replacement necessary. Thus it is more efficient to move the laser beam
continuously. For example by moving the beam 50 mm to the right, 50 mm
downwards and then 49 mm to the left generates a rectangular spiral pattern
(11 C). However in that case only one mirror is moving in each direction, which
again cause a delay, before the other mirror reacts (and again, deeper ablation at
these turning points, and it’s the same with motor stages). Alternatively, scanning
the laser beam in an Archimedean spiral pattern with ‘rounded corner’ has proved
worth and preferable shape, which enables to exploit as much as possible of a
target per time and assures maximal target use (11 D). Also here at the end of
the spiral one delay occurs if the scanner had to move to its initial position. The
highest utilization rate then would be by using a Fermat’s spiral pattern, which is
unified within itself and allows a continuous movement of the laser beam (11E).
Anyway, for an optimal utilization of the target material, the target needs to have
the form of the scanning pattern, e.g. in case of stripes a rectangular target is the
best and a round target shape can be used with a Fermat’s spiral. A set of annex
spirals can mimic a rectangular shape.

A B C D E

Figure 11: Scan pattern on the target for laser ablation in liquids: (A) Without scanning,
(B) Stripes, (C) Rectangular spiral, (D) Archimedian spiral, (E) Fermat’s spiral.
30
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

STEP 8: SPICE UP YOUR NANOPARTICLES

A huge advantage of laser-generated nanoparticles is the exclusion of toxic
substances or by-products that adsorb onto the particles surface, since
chemical precursors are not needed. Depending on the laser parameter, pulsed
laser ablation in pure water very often leads to polydisperse particle sizes or I
even multimodal size distributions. To spice up your nanoparticles you can use
(either micromolar saline solution for noble metals, or) several post-treatment
methods. Centrifugation is such a method to play with the size distribution
(compare Chapter III). By using a combination of Svedbergs’ equation and
Stokes-Einstein relation a pre-defined cut-off can be calculated. The cut off is
the particle diameter value that defines the fractionation size level between
supernatant and sediment. Using this formula the centrifugation time can be
calculated if parameters such as particle size (intended cut-off), nanoparticle
material density (take the value of the bulk), viscosity of the solvent, rotation
speed of the centrifuge and dimension parameter of the centrifuge are known
(rotor type that holds the vials). Due to the faster sedimentation of larger
particles compared to smaller ones, different fractions can be separated to
desired particles sizes by taking advantage of centrifugal forces. In this manner,
after centrifugation the largest particles or agglomerates are located at the
bottom (sediment or pellet) and small particles (supernatant) at the top of the
centrifugation tube shown in Figure 12.

pellet supernatant
pellet supernatant

Figure 12: Particle size separation by centrifugation.

Even though size separation by centrifugation is a quite simple method, one

main disadvantage is the loss of mass if only one fraction is needed (the pellet
cannot always be redispersed). Typical loss of a bimodal colloid of laser-
generated noble metal nanoparticle colloid in water, if the pre-set cut-off is in
the range of 10-20 nm (to get rid of the large fraction by-product of LAL), is
about 30-50% of the mass. It will take only 3-5 minutes with typical benchtop

31
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

centrifuges (far longer centrifugation times should be avoided in any way, unless
you have a centrifuge or ultracentrifuge that cools down the colloid). Finally, you
will end up with a monomodal, often monodisperse colloid, free of any additive.

The size of nanoparticles can be also tuned by re-irradiation of a colloid with

I pulsed lasers. At high laser fluence, post-irradiation of nanoparticles leads to
fragmentation of nanoparticles and thus to smaller sizes and narrower size
distributions. This Laser Fragmentation in Liquids (LFL) generally benefits from
a wavelength at which the particles can absorb significant light, however laser
fluence should be still efficient. As a rule of thumb: the product of absorption
at the laser wavelength (measured in the UV-Vis spectrometer) and the pulse
energy at that wavelength should be maximal.

For very small nanoparticles, surfactants can be used during laser

fragmentation. But also re-irradiation of nanoparticles in totally ligand-free
environments or micromolar saline solution facilitate size reduction by high
laser fluence laser fragmentation. For noble metals in water, maybe adding
0.1-0.3 mM NaOH or carbonate buffer (there is CO2 in water, anyway) could
be a worth a try.

In contrary to laser fragmentation, which yields smaller nanoparticles, Laser

Melting in Liquid (LML) can be applied for the production of spherical particles
of several hundreds of nm in size by using low fluences applied to metal or
dielectric nanoparticle (aggregate) dispersions. At moderate fluences (try
10-300 mJ/cm²) nanoparticles are stimulated to aggregate and the aggregates
of nanoparticles then melt, fuse together and form submicron spheres. There
large spheres are easily sedimented by gentle centrifugation and are often well
redispersable.

Re-irradiation of a colloid can be followed in different setups. The simplest

approach is laser fragmentation in a batch chamber, where a colloid is irradiated
from above or from the side (Figure 13, left). Similar to this, post-irradiation can
proceed in a continuous operation mode, in which a colloid is operated in a
flow-through mode and is irradiated by a pulsed laser light (Figure 13, middle).
A significant disadvantage of these setups is that focusing a laser beam in the
colloid results in a deviance of fluence at several positions in the liquid, thus
leading to different mechanisms in particle re-formation. In sharply focused
situation with long beam path in liquid, you may cause LFL in the focal volume
but LML before and behind it, creating a mixture you may not want. Also, if the
total liquid volume is larger than then irradiated volume, educt-product mixing
appears, and the process renders to be less efficient. The most precise post-
irradiation can be performed in a liquid-jet reactor (Figure 13, right). Sounds
extravagant, but it is just a funnel, more detailed a flask with the educt on top
reservoir with a mm capillary outflow at the bottom, omitting the use of a pump.
The geometry of a reactor with a liquid jet allows a tight focusing of the laser
beam on a thinly volume, so that each particle is reached by the same laser

32
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

light fluence and losses of excitation light by scattering or absorption in front

of the focus can be reduced. Furthermore, even at high light intensity, there
is no risk to damage any optical element like windows or cuvettes, the beam
entrance is the liquid itself. If the reservoir is empty, you have done one passage
with very defined energy input (think of placing a laser powermeter behind the
liquid jet to quantify the energy intake) and can see if the colloid meets your I
demands. If not, just refill and go for the second passage. The outflow speed
slightly changes during the process because hydrostatic pressure changes when
the reservoir gets empty. A workaround is quite simple: use the concept of a
Marriotte´s bottle (see Wikipedia) to assure constant volume and particle mass
flow passing the beam.

colloid
low high
fluence
fluence

colloid

liquid jet

Figure 13: Set-ups for pulsed laser post-Irradiation of nanoparticles in liquids.

How to influence the particle size by using ions or other additives such as
biomolecules will be explained in Chapter III.

STEP 9: NOW, JUST DO IT

You are entering the lab and see this ‘magic box’, which is more powerful than
it may seem. You don’t need to be afraid of powerful lasers, but you should
treat them with respect. You need to take the security aspects into account,
which are described in detail in Chapter V (“Stay alive”). By this, you’ll have best
precondition to safely operate with lasers.

The day should start with all preparation steps for the set-up. Sometimes
experiments can fail, if only a screw is missing, thus it is worthy to make a list
with the required preferably on the afternoon before you actually intend to use
the laser (Figure 14). It may be satisfactory having everything waiting for you in
the lab and get very productive already right after your morning coffee.

33
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

I t io
n cha
mb
er

dr i v er ,
l- k
ey

bla , sr
ew
nd
-a ws pou
c r e o m
-s c a lc d
pti ar a ls
-o rc , vi
cto as k or
d et e , fl s ens
- t
o l ven t e r+
n
-s m e a
er t em
ow ip e
-p t t e, p e s
ip e l as
s
-p rg n
l as e k , pe
- bo o
l ab
- et
arg
-t
.
- ..

Figure 14: Check-list for preparing pulsed laser ablation in liquids in the afternoon before your
day in the laser synthesis lab.

When you have all the required things together, you can start to prepare the
setup. Before switching on the laser, don’t forget to put on the laser glasses and
let others know that laser is on!

You need to build up the optical components to align laser beam to a desired
position. All optic compounds should be free of any defects and dust, to ensure
a high yield of light reaching the target. To locate the laser beam a detector
card can be used (as your laser is invisible through safely glasses), where a
photosensitive region allows to view the light spot. Before use, such a card
requires “charging” with visible light, and for optimal brightness of the beam
one must move the card around because the emission from the card is not
persistent. If you use a laser scanner, align the laser beam with a low intensity
into a scanner optic and then onto a thermal laser power sensor, which is
connected to a powermeter. Note, that you should measure the power only in
the raw beam or out of the focal plane to avoid a destruction of the sensor.
Point out that you have to check the fluence of the raw beam directly at the
source and – most relevant for the synthesis - behind all mirrors (right before
the focusing lens) before you start in order to avoid potential fluctuations in the
laser fluence. Once you have directed the raw beam spot on the sensor, you can
slowly increase the laser intensity and measure the power that you intend to
use for laser ablation.

34
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

Wait a couple of seconds for thermal equilibrium of the sensor (and remember,
many lasers need a warm up period of 20-30 minutes to reach steady state of
output power).

Using the example of laser ablation in a flow-through chamber, the preparation

steps are briefly described. First, the mass of the target should be determined,
hence the ablated particle mass and with it the colloidal mass concentration can
I
be concluded by weighting the dry target before and dried target after ablation.
The ablation chamber should be prepared by embedding the target and screwing
all component parts. After placing the chamber in front of the scanner in a
distance of the focal length, very often, because of liquid refraction, the point
of maximal productivity is behind the focal plane in air (or the focal distance
the lens supplier has provided), e.g., at 1-2mm longer distance. Tubes should
be connected with the chamber and the pump. A certain volume of the solvent
should be filled in a flask and the tubes should be dipped into the solvent.
Before starting ablation, the liquid should be passed through the chamber and
recycled back to the flask until no bubbles can be observed to prevent ablation
in air or bubble scattering. Be sure, that no solvent is leaking during the flow.

The target can be ablated in a continuous process, where the solvent is pumped
through one flask to another collecting vessel. If you want a highly concentrated
colloid, quasi-continuous ablation (so-called semi-batch) can be carried out by
cycling the solvent during laser ablation in the same flask. It should be noted,
that by this mode of concentrating operation particles can be re-irradiated more
intensely by the laser beam. In most setups, you will have some post-irradiation,
and no flow setup is fast enough to cope with kHz repetition rates. But flow
brings everything in a controlled steady state.

Now you can align the laser beam through a scanner optic onto the target
(or directly on a moving target) by directing a laser beam with a lowest possible
energy. Some lasers have integrated an additional positioning laser, which helps
to adjust the beam on a target object. Oftentimes your laser is equipped with a
shutter, which can be controlled by a software or opened and closed manually.
After you have found the optimal set up parameter, and you are sure that the
beam hits only the target, you can start to pump the solvent and increase the
laser energy.

Since you have synthesized a colloid, you may notice that the concentration
of nanoparticles in the liquid is not as high or the particles even start to
agglomerate after a short time. Now the time has come when you can use your
knowledge on finding the focal position and continue to perform systematic
studies on laser synthesis and stabilization of colloids.

35
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

ablation
chamber
scanner

I target

laser

mirror

pump

Figure 15: Optimized experimental set-up for pulsed laser ablation in liquid using a flow-through
chamber with continuous operating mode.

At the beginning, many mistakes can be made and of course, you learn from all
of them. However, there are some typical mistakes that can be avoided. The list
of lessons learnt by us may help you to prevent unnecessary waste of time and
difficulties:

Forget ting to weight the target

Working without laser goggles

Leaving shut t er closed, when trying to find the beam

Aligning the beam with maximum laser energy

Leaving shut t er open, before pump was s tart ed

Adjus ting the beam out of focal position

Ablating till hole formation at the target causing ablation of

the chamber mat erial

Set ting optics incorrectly (maligned optics or covered with

dus t)

Sealing ablation set-up (chamber , pump) insufficient

Irradiating in open flammable solvent s or leaving combus tible

mat erial near laser set-up

36
 GET STARTED: SETUP OF EXPERIMENTS AND STARTING DAYS

You can find all the things mentioned above in the following that can help you to
synthesize your first particles by laser ablation in a liquid:

PREPARATION
Do your checklis t with all necessary object s for your set up
I
Bring all the things together the day before you s tart synthesis

Check your optics and clean it if necessary

Weight your target

Prepare the solvent

SET UP
Build up your set up (optics & chamber)

Remove all flammable it ems near the laser beam (solvent s,

paper , packaging , …)

Put your laser goggles on

Check laser alignment, use det ector card and low power

Allow laser warm-up and measure the laser power (raw

beam!) af t er all positions (behind the laser source, lenses and
mirrors)

Position the chamber and fill it with solvent

LASER ABLATION
Align the beam on the target

Find the focus

Start to produce your particles

Enjoy

Good luck and may the force be with you!

37
 CHAPTER II

GET MORE: PRODUCTIVITY

In the previous chapter, we used the power of light to produce very pure and
tiny nanoparticles by pulsed laser ablation of different materials in liquids. With
this fancy setup, you will be able to delve down into the amazing playground of
nano-sized objects. However, exploring this small universe doesn’t necessarily

II mean being content with tiny quantities. In case we want to characterize or even
apply our nanoparticles in some novel material combinations it is not enough
to end up with small amounts. Therefore, let’s grow the scales and make small
things huge – not just in size but also in quantity.

STEP 1: HOW TO DEFINE THE NANOPARTICLE

PRODUCTIVITY?
To start with our scale up procedure it is important to initially think about how
we want to measure productivity and consequently find a proper definition
for the expression “nanoparticle productivity”. Does “yield”, “efficiency” and
“productivity” have the same meaning? Yield refers to the product, e.g. 80%
after centrifugation or “100 ml with 1 mg monodisperse gold”. Efficiency
needs something to refer to, so that there is often the effort you put in for the
synthesis in the denominator, e.g. mass you obtained per laser energy (mg/J).
Productivity has the unit of time in the denominator, e.g. mg/h. The meaning of
productivity seems obvious, but once we look closer in this topic there will be
several possible ways to define the nanoparticle productivity.

Let’s begin with the most apparent choice and define the productivity as the
total amount of ablated target mass per unit time. It can be obtained by simply
weighting the target before and after an ablation interval. The precision of this
gravimetrical technique is only dependent on the chosen time of ablation and
the scale that is being used. Evidently, the significance of your results will be
enhanced, if an extended timespan is selected. In general the ablation time
should be chosen such that a single ablation experiment will last at least
5 minutes while the minimum amount of ablated target mass should not fall
below 3-5 mg. The latter limitation is necessary in order to keep the relative
weighting error lower than 5% (following a typical error of about 0,1mg for a
standard analytical balance). Additionally it should be kept in mind that most
lasers do need about 20-30 minutes pumping to reach a stationary power
output (new lasers maybe 2-5 minutes). Consequently, the laser beam should be
aligned into a beam trap or onto a power meter before starting laser synthesis,
until the stationary power output has been reached in order to maximize
reproducibility.

38
 GET MORE: PRODUCTIVITY

~~~~~~~~~~~~~~~~~~~~~~~
Working with int ense laser radiation is a task , which requires concentration due to
the risks of burning ones skin and eyes. S till , there are easier ways to get hurt. When
checking lenses and mirrors one should make sure not to bend over in a way that leads

II
to hit ting the edge of a safet y screen and suffering of a bleeding ear af t erwards.

Alexander Letzel , Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

As easy and convenient as the gravimetrical approach may seem, it always

includes interrupting the ablation process making time resolved productivity
studies impossible. However, in case some knowledge about the temporal
alteration of the productivity is required, a second definition of productivity may
be introduced. In contrast to the “gravimetrical productivity”, let’s additionally
define a “visible productivity” considering the amount of colloid being detectable
e.g. by optical measuring techniques. You may be thinking – shouldn’t those
two definitions be the same? Bear in mind that nanoparticles may not be stable
and precipitate or even dissolve in the liquid. It may also be required to exclude
generated micrometer sized particles from the productivity definition as they
often aren’t the desired product. Therefore, it is obvious that both definitions
may differ significantly depending on the chosen metal-liquid system and the
required nanoparticle specifications.

In order to quantify the “visible productivity” the nanoparticle concentration

may be monitored by UV-VIS absorbance spectroscopy over time. With this
method the concentration of a colloid can usually be calculated considering
the extinction of incident light over a given path length. More about this type of
detection will be given in Chapter III. As the ablated mass always gets converted
into particles with 100% yield, both the “gravimetrical” and the “detectable”
definition are equal as long as the particles are sufficiently stable over some
period of time. For the ablation of most noble metals this statement holds,
however when ablating metal oxides the “detectable productivity” and the
“gravimetrical productivity” approaches may significantly deviate as nano-/
microparticle precipitation is often observed in those systems.

The final question about which of the two definitions is to be chosen, may
arise. To answer this question: it depends on what you’re interested in! The
gravimetrical approach is easy to measure (and does not need material-specific
calibration curves) but lacks temporal resolution and ignores precipitation
and dissolution whereas the “visible productivity” definition gives rise to a
well-defined time resolved productivity profile while requiring more effort

39
 GET MORE: PRODUCTIVITY

and knowledge about the used nanoparticle system. The latter will however
become a very powerful tool when thinking about long term continuous ablation
methods. However, it requires a calibration curve, to be done by simply diluting
a colloid and find out which UV-Vis wavelength gives best correlation coefficient
in the linear fit (don´t force it to go through zero) of the diagram concentration
vs intensity. By knowing the different ways to measure productivity we are now fit
to generate our first nanoparticles.

II
STEP 2: CROSS THE ABLATION THRESHOLD
Well, maybe this chapter should be called crossing the information desert
without some refreshments in you back-pack but unfortunately, our sleeves will
still remain unwrapped until we haven’t thought about the basic idea behind
the laser synthesis of colloids. In fact, to generate some of our desired midgets
it is required to reach the so called ablation threshold already introduced in
Chapter I Step 7. This threshold depends on a manifold of different properties.
Next to material properties, it is defined by laser parameters such as laser
wavelength, laser fluence and the pulse duration. Keep in mind - Laser synthesis
of colloids is like performing martial arts – so let’s consider the aforementioned
parameters from the angle of e.g. Muhammad Ali. Let’s say the laser wavelength
represents your striking precision, the laser fluence creates your jap virtue and
the pulse duration is your quickness. Analyzing your opponent’s weaknesses
first, you want to precisely strike where it hurts most. For an ablation target,
you’ll find its weaknesses in its wavelength dependent light absorption. Choosing
the right laser wavelength will significantly increase your impact. But maybe you
can’t choose a weak spot (i.e. fixed laser wavelength of your laser) on your own,
it is obvious that you can always strike harder (increase laser fluence) to make
up for that. Of course your striking power is limited so you have to make use of
your quickness (pulse duration) as well. A fast strike with a laser pulse of very
short duration will surprise the target making it “bleed” and loose nanoparticles.

~~~~~~~~~~~~~~~~~~~~~~~
I ent ered the laser lab and found the surface of the power met er significantly
damaged, with a dis tinct ablation spiral . Obviously one of the s t udent s tried to measure
the energy of the laser close to the focus during operating and scanning the ablation
spiral .

Sebas tian Kohsakowski, Duis burg

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

40
 GET MORE: PRODUCTIVITY

The reason for the higher impact of a laser pulse with short pulse duration is
that it strongly influence’s the energy loss due to energy transport processes in
the following way:

Ftheshold: femtosecond pulse ≤ picosecond pulse < picosecond pulse

resulting in different time scales for the successively occurring heat exchange
mechanisms. Basically, the target material will just be struck faster than it is able
to react. This situation is roughly summarized in Figure 16. As you can see, there
are mainly two different extreme cases: one when using a nanosecond pulse, II
and the other with a femtosecond pulse. The reason for the two extreme cases
is a property of materials, which is the so-called electron-phonon-relaxation
time. When the laser beam starts being absorbed the electrons get excited.
It’s the same as if Muhammed Ali struck his opponent: in the first second after
striking his opponent, he doesn’t feel any pain. However once the brain of his
maltreated victim realizes there was a strike, his quarry collapses due to KO. The
same happens in the material when the excited hot electrons start to interact
with the cold lattice, heating up the material. The delay for this heating process
is the aforementioned electron-phonon-interaction time. The usual timescale
of these is in the picosecond regime for metals. Subsequent on this interaction
delay the target temperature increases while the electron temperature
decreases until both, lattice and electron temperature are roughly the same.
Now if you consider the first extreme of, e.g., a 10 nanosecond laser pulse,
the pulse duration is far longer than the time required for the electron-phonon
relaxation Figure 16 (middle). Therefore electron- phonon relaxation (several
picoseconds) is already occurring, thus heating the target in the irradiated
area, while the laser pulse is still continuing. Due to the increasing temperature
gradient between target and liquid as well as inside the target itself an additional
heat transfer kicks in after a few nanoseconds, leading to heat conduction into
the target and heat transfer (via the plasma plume) into the liquid while the
laser pulse is still ongoing (Figure 16, middle). Now, from this argument you may
already imagine that this condition results in deep heating, finally resulting in
vaporization and plasma formation, which is why this case is called hot ablation.
The laser is still on when the plasma arises into the liquid, hence the incoming
light is shielded and consumed by the plasma, lowering ablation efficiency
(ablated mass per pulse or ablated mass per joule pulse energy) compared to
ultrashort pulsed (often called “ultrafast”) picosecond and femtosecond lasers.
Going back to boxing, this nanosecond technique is similar to Nikolai Walujew
standing in the ring. Due to his slow punches he needs more force (energy) to
knock his opponent out which is why we may get the impression nanosecond
pulses are rather less desirable. Therefore, let’s think about a very short laser
pulse in the femtosecond regime. In this extreme case the laser beam duration
is already finished before the electron-phonon-interaction is over with the result
that the whole energy gets stored by the electrons and suddenly released into
the lattice, creating a sudden expansion and explosion of the material.

41
 GET MORE: PRODUCTIVITY

As no heating occurs this ablation is often called cold ablation (in practice,
there is a small portion of heat deposit in the target that is often not noticed
when working in liquid unless you go for very high repetition rates). It is obvious
that this technique creates much less heat loss which is why a higher fraction of
energy is used for the real ablation process.

II
ns-laser f s-laser pulse Response Hot Cold
Ablation Ablation
pulse Response
pulse duration

pulse absorbtion
+
heating /evaporation
pulse absorption
+
energy storage

Figure 16: The effect of pulse duration on the ablation mechanism.

Hence, when using nanosecond pulses, a higher fluence is needed to cross

the ablation threshold. Of course, from this behavior you may infer that
femtosecond or picosecond pulses should be highly favored especially because
the ablation is more energy efficient. Yet we’re only at the beginning of our
journey and we will see that high efficiency also comes with a cost, creating
additional efforts and therefore rendering inefficient brute force is the easier
option in some situations…

STEP 3: MAXIMIZE THE ABLATION EFFICIENCY

With the gained knowledge about the ablation threshold, especially about its
dependence on the pulse duration, it is time to discuss how we can increase
the ablation efficiency without varying the pulse duration of the laser beam.
That will become your routine work, because pulse duration is fixed at most
laser systems. You may already smell the rat – in order to increase the fluence

42
 GET MORE: PRODUCTIVITY

we have two options: either increase the pulse energy or decrease the beam
area. Before working on practical approaches on how to achieve either of these
alternatives, we may split our ideas into cost-intensive and non-cost intensive
measures. It is obvious that you will always endorse yourself by using a non-cost
intensive approach saving lots of money. Did you already figured out which of
the two given approaches may embody the trail towards success and popularity
amongst your boss? If your choice fell upon maximizing the pulse energy you
probably chose the sledgehammer to crack a nut. It is almost certain that
initially following this approach you will sooner or later end up buying a new II
expensive laser because the old one seems too feeble. Of course, the laser
manufacturers will be pleased by this approach however your budget may not
be amused. Also, very high pulse energies don´t make you happy once you´ve
seen how huge pulse energies damage your optical components or ablation
chamber windows. To save some of your money, initially the non-cost intensive
approach of simply changing the beam spot area should be chosen. In order
to do so the laser beam can be focused using a collecting lens. This is one of
the easiest actions available to significantly increase your ablation efficiency
and should always be your first priority when thinking about scale up measures.
With the collecting lens introduced into our setup your first scale up action is
to optimize the distance between lens and target as shown in Figure 17. When
ablating in gaseous media, the highest ablation rate will usually be found when
the beam focus is aligned with the target surface as the fluence reaches a
maximal value within the focal spot. Having a target immersed in a liquid – say
water - interactions of the laser beam with the liquid have to be considered.
One of those interactions is the refraction of the laser beam at the gas-liquid
phase boundary, shifting the focal spot depending on the chosen liquid height.
This shift can easily be calculated using linear optics for the known liquid
height. Depending on setup, laser and liquid applied, you may however want to
discard the calculation of the refraction (considering the manifold of possible
disturbances to be addressed soon) and simply choose an experimental
approach following Figure 17. Here, the displacement between lens and target is
being varied until the highest productivity has been found. As a general rule of
thumb, the initial distance between target and collecting lens should equal the
working distance of the collecting lens given by the manufacturer of the lens
(Figure 17, point A), after which you increase the distance.

43
 GET MORE: PRODUCTIVITY

nanoparticle productivity
x1

B
x2
II focal position
in air
C
x3
relativ focal shift

Figure 17: Optimization of productivity by changing the distance between collecting lens and
target beginning at point A where the equals the focal distance of the collecting lens.
The productivity optimum is reached at point B where the beam waist is located on the target
surface. Shifting further towards point C the productivity is again decreased.

The productivities shown in Figure 17 may be determined by the gravimetrical or

the visible approach (or sound amplitude) discussed earlier. Once approaching
point B, the productivity reaches a desired maximal value as the beam area at the
target surface becomes minimal (fluence is maximized) with the focus located on
the target surface. A further increase of the distances should be avoided. Above
the optimal distance your laser focus will be located in the liquid, typically
forming a plasma within the liquid leading to severe absorption, local boiling and
bubble formation, significantly decreasing the LAL´s nanoparticle productivity. In case
the ablation is done in an ablation chamber where the fluid is confined by a quartz
window (e.g. dashed line in Figure 17) a focus placed on the used window should
be avoided. Otherwise damaging of the window and termination of the experiment
(maybe even inflammation of the liquid, e.g. when using acetone or ethanol) due
to a bursting window may be happening. Therefore, it is highly recommended to
only conduct this optimization in between the range of A and B while starting at
point A until B is reached. This routine is one of the main basics when trying to
maximize productivity. It should always be repeated if the setup (e.g. liquid height,
target thickness, different lens) or the liquids (refractive index) were changed in the
meanwhile. You do not need to do this with the precious target material reserved
for your flagship experiments series. Instead, you may take another material with
same thickness. Avoid using high reflective materials for finding the focus (such as
silver) if you have no idea where the focus is roughly, as it may back-reflect the partly
focused laser light (from position A in Figure 17) and inscribe or break the window
from the inner side.

However, it was already insinuated that there are numerous biasing effects that may
still decrease the productivity or even avoid finding a proper productivity maximum.
The next section will address those interferences and prevent you from spending too
much of your precious lifetime while waiting for your nanoparticles being generated.

44
 GET MORE: PRODUCTIVITY

STEP 4: AVOID FLUID BREAKDOWN AND

NONLINEAR EFFECTS
The previously described basic method will always be a starting point of your
optimization. But as soon as the productivity starts rising, the previous simplicity
becomes a rare luxury. When applying high laser fluencies and/or very short
pulse durations a whole series of stumbling blocks may be evoked on your way
towards higher productivities. Let’s shed some light on how to evade those
obstacles and make sure we boost our productivity to the maximum. II
The first “banana peel” may possibly occur when focusing your laser beam into
an open beaker filled with the target immersed in e.g. water. Let’s consider the
worst case scenario of high laser intensities, lenses with a long focal distance
and a low liquid layer thicknesses being applied. Once exceeding a given fluence
threshold in the liquid, you risk causing a significant fluid vaporization followed
by severe bubble generation as shown in Figure 18 – Case A. This effect will
drastically reduce your nanoparticle productivity, due to severe scattering
of light at the bubble interface, rendering all effort of finding the best focal
position inefficient and pointless. In order to avoid liquid vaporization when
using lenses with long focal distance (small focusing angle a1 see Case A and
B) you will be forced to work at rather high liquid layer thicknesses (or smaller
pulse energies). However increasing the liquid layer thickness (or decreasing
the pulse energy) will again cause a substantial depletion of our desired
productivity, as laser light might be scattered by possible gas bubbles (from
solute gas) or absorbed of already produced nanoparticles.

high focal distance high focal distance low focal distance

A small liquid layer height B big liquid layer height C small liquid layer height

1 2
focal distance

focal distance

focal distance

Figure 18: Influence of focal distance and liquid layer height on the generation of an optical
breakdown at high pulse energies (high peak power). Case A: High focal distance and small
liquid layer resulting in a small beam area (high fluence) at the phase boundary causing in
optical breakdown. Case B and Case C: Fluence at phase boundary doesn’t excess breakdown
threshold  no optical breakdown occurs.

45
 GET MORE: PRODUCTIVITY

To avoid the aforementioned disadvantages it is feasible to refrain from using

collecting lenses with a very long focal length. The application of shorter focal
distances should be favored as those lenses possess higher focusing angles
(short Rayleigh length) and therefore a bigger beam area at the liquid surface
even at small layer height (Figure 18 – Case C). A smaller focal length will
additionally result in a smaller beam waist at the focal spot leading to higher
fluencies or, generally speaking, higher productivities. A big downside of lenses

II with short focal distance (f << 100 cm) is the limited action window. In this
context, the action window refers to the spacing of the plateau at Point B
in Figure 17 shown in Step 3. The shorter the focal distance the smaller this
plateau, resulting in a stronger sensitivity of productivity on changes in the focal
distance when using this type of lenses. One effect causing constant variation
of the focal distance is caused by the ablation itself. Let’s consider ablating a
target, which implies removing material. It is obvious that a crater of increasing
depth will be forming with ongoing ablation, continuously increasing the distance
between the target surface and the focal spot position. Considering a lens with
short focal length, you’ll find a fast drop in productivity with increasing crater
depth (due to the small action window / short focal distance) in case you’re not
constantly adjusting the distance between target and focusing lens. Therefore
working with lenses of short focal distances increases short term productivity
which on the contrary quickly drops if the distance is not constantly adjusted
(e.g., every 15 minutes) increasing the working effort. On the other hand, a long
focal distance (f > 100 cm) will increase the operating window but strongly
limits the productivity due to the higher beam waist area. Hence, a suitable
compromise between these opposing tendencies is necessary. A proper trade-
off was usually found when using f-theta lenses with a focal distance of around
60 mm (and liquid layer thickness of around 4-6 mm).

~~~~~~~~~~~~~~~~~~~~~~~
In my laser lab there is also a micro Raman spectromet er , that of t en needs long
measurement sessions by my colleagues. For this reason, I ’m sure to obtain some
t errible curses by them when I tried laser ablation in iodo-benzene. Perhaps this
solvent has not a scaring datasheet, but it has a persis t ent and really unlovable smell!

Of course I realized that a sort of “ t echnological limit ” s tays in front of me: lasers
cannot be positioned under laboratory extractor fan, and the leas t cannot be moved
over the entire optical bench. So the solution was a sort of lit tle portable extractor
fan, linked with a flexible t ube and equipped with some metal shields that allow to
fit on my laser ablation set up. The next Raman users sincerely appreciat ed my set-up
improvement!

Lucio Lit ti, Padova

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
46
 GET MORE: PRODUCTIVITY

Now, since we discussed what may happen when working with high laser
fluencies, there is another extreme we didn’t address until now – short pulse
durations. As already discussed in Step 2, using short pulse durations in the
femtosecond regime will result in nearly no heat losses and therefore high
ablation efficiencies (based on energy input). In contrast to the mean power
being released by the laser while switched on - the peak power is much closer
to the apparent ablation process. From the example given in Chapter I Step 4 it
is obvious that an ultrashort pulse duration will cause a very high peak power
during the pulse, which can be similar to the power a power plant produces, II
causing a more efficient ablation (as discussed in Step 2). However, this peak
power will also have a tremendous effect on the liquid in which the target is
immersed. As soon as the peak power exceeds a threshold that depends on the
liquid nature, events like self-focusing and optical breakdown will be the result.

In this sense self-focusing comprises a nonlinear effect usually occurring at peak

powers in the order of GW. Mostly these high peak powers can only be reached
with femtosecond pulses. The basis of self-focusing is the so-called Kerr effect
which describes a dependence of the refractive index on the laser intensity. In
normal applications, the refractive index of a liquid is only dependent on the
wavelength of light. However, in case of such an intense electromagnetic field
strength like in a laser pulse with a peak power of several GW, the positive Kerr
effect results in an increase of the refractive index with increasing laser intensity.
This results in a lensing effect where the medium acts like an additional
collecting lens significantly shortening your focal distance. Therefore when
working with femtosecond pulses in liquid, the focal position maybe found even
above the geometrical focal plane. You will notice self-focusing in liquid by a
white-light filament, like a straw of light. It causes some energy loss, but benefits
from tight focusing.

The other effect called optical breakdown mainly happens when a laser beam
of sufficient peak intensity enters a liquid out of air. And it will strongly diminish
productivity. The intensity threshold for optical breakdown to occur was found
to be in case of water (800 nm and 120 fs) while the thresholds
of organic liquids are often even lower. To get an idea about the scale of this
value, let’s just consider an example. In typical femtosecond ablations setup
100 fs and a beam waist diameter of 100 μm are common. If you consider
the fluence threshold given above, optical breakdown will already occur when
a pulse energy of about 170 μJ has been exceeded. However, this actually is
a typical pulse energy when working with ultrafast lasers. Therefore, optical
breakdown and reduced productivity are very likely when femtosecond lasers
are being used at high pulse energies. Unfortunately, your only option to avoid
optical breakdown of the liquid is to increase pulse duration or lower the
fluence. Moreover, in many cases even this measure does not let you escape,
because colloidal nanoparticles, in particular plasmonic nanoparticles are very
effective “sensitizers” for filamentation. In practice, unless you have created

47
 GET MORE: PRODUCTIVITY

severe optical breakdown and splashing of the liquid, filamentation can still be
handled. Many groups work with fs and ps lasers in liquids since decades and
are quite happy with that. Maybe productivity is not always optimal for laser
ablation synthesis, but there are no better laser tools for efficient fragmentation
(downsizing) of colloids than ultrashort pulsed lasers.

Also keep in mind that during the “cold ablation” where the pulse duration
is shorter than the electron-phonon-relaxation time (Figure 16), the fluence
II threshold for laser ablation and therefore the productivity (assuming the same
laser fluence and wavelength) is less dependent on the pulse duration, rendering
100 fs - 10 ps laser ablation with similar efficiency. The shorter pulse duration is
more efficient at the ground, but disturbed to a higher extend before reaching it.
Increasing the beam area may be a proper alternative in these cases. However,
when applying pulse durations significantly longer then the electron-phonon-
relaxation time (that is, nanosecond lasers) the ablation efficiency forces you to
bring out the earlier called sledge hammer to crack the productivity nut. In this
context, the focusing conditions are crucial for nanosecond lasers.

Wrapping up the take home message of Step 4, if you try cutting target
chunks using a scalpel (femtosecond laser) make sure to be gentle (limit pulse
energy) and easygoing (prefer ps pulses) or you may spook the solvent (liquid
break down) while carving out the desired target areas. In case you prefer
using the machete (nanosecond laser) avoid wielding it like a claymore (using
inappropriate lenses with f > 300 cm) while also not begin fencing it (going
too short for f < 50 cm). Nanoseconds are the working horses of LAL, and with
ultrafast lasers you are on the safer side when working in organic liquids. This in
mind let’s move on to the next step trying to outrun some bubbles.

STEP 5: DODGE THE CAVITATION BUBBLE

In Chapter I, the concept of cavitation bubbles occurring during laser ablation
was already introduced and discussed. These cavitation bubbles are a significant
issue for productivity as they operate as a lateral limited protection shield of the
target deflecting the strikes of subsequent laser pulses during bubble lifetime
scattering. Consequently, we need to think how to handle this shielding wall in
order to stop it from limiting our desired productivity.

We actually suggest two different ways – the temporal and the lateral
evasion of the cavitation bubble. The idea of temporal evasion technique is to
successively reduce the repetition rate until the subsequent laser pulse hits
the same ablation spot only after the cavitation bubble is gone. Hereby the
ablation efficiency (mass per laser pulse) can be significantly increased once the
temporal spacing between the pulses is longer than the lifetime of the cavitation
bubble. A typical cavitation bubble lifetime is 200 μs which would equal a
maximal “undisturbed” repetition rate of 5 kHz. Hence, this technique is only at

48
 GET MORE: PRODUCTIVITY

all feasible for lasers operating in the lower kHz regime. Not a big issue if you
work with a laser from the 90ies that delivers only 10 Hz. But, if you are lucky
to run a kHz laser with more power, you may already sense another flaw in this
idea. When the repetition rate (number of pulses per second) is being decreased
the mean laser power output may decrease as well depending on the type of
laser and its characteristics. Even if the amount of ablated mass per pulse
increases by the temporal bypassing, decreasing the number of pulses (due
to a lower repetition rate) may still cause the overall productivity to stagnate
or only slightly increase. Therefore, the temporal evasion is a robust strategy II
for increase of productivity, but has physical limits. Again, the bubble size and
lifetime scales with the pulse energy. If you are allowed to select a laser from
scratch, 5-10 kHz is enough at high (>> 10 mJ) pulse energies, maybe
100 kHz at low pulse energies (<< 1 mJ) may still get to the ground undisturbed
by cavitation bubbles.

~~~~~~~~~~~~~~~~~~~~~~~
Af t er conducting a series of measurement s for different calibration curves at the UV/
Vis spectrophotomet er , nothing but noise could really be extract ed from the data.
Af t er reassuring that the samples were prepared correctly, we expect ed a major defect
of the ins trument. However , upon opening the lid, it became apparent that only the
cuvet t e changer was not calibrat ed correctly so that the light beam could neither
reach the sample nor the det ector . Lesson learned: check the ins trument set up and
ques tion your result s directly before continuing with 100 more samples and was ting
your time.

Carmen S treich, Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Still following the idea of a peaceful solution, the second and probably most
generally applicable idea is the lateral evasion of the cavitation bubble. In order
to achieve this goal it is mandatory to use a scanner or displace the target during
the ablation (e.g. by spinning the target). For simplicity, let´s only consider using a
scanner at this point. One of the most efficient patterns applied during ablation
is the use of a Fermatian spiral (closed spiral) as described in Chapter I. In this
case, the laser follows this spiral, homogeneously ablating the target. The curved
nature of the path makes sure the mirrors in the scanner are constantly moving,
whereas the adoption of a path with corners would result in sudden operation of
one mirror and consequent introduction of mapping errors. Back to the cavitation
bubble, in case you plan working at high repetition rates, e.g. >> 5 kHz there may
be an interaction of subsequent laser pulse and cavitation bubble even when

49
 GET MORE: PRODUCTIVITY

the target displacement is applied. Yet if you manage to guide the second laser
pulse towards a space next to the cavitation bubble fast enough, no interaction
between pulse and bubble occurs. Consequently, the ablation efficiency and,
as the repetition rate remains unchanged, the overall productivity will increase
significantly. Note that this also makes sure that every shot gets the same energy
to the ground, increasing reproducibility. To manage the previously said you
simply need to increase the lateral interpulse distance by increasing the scanning

II speed. Measuring the productivity after stepwise increase of the scanning speed
(interpulse distance) an optimal interpulse distance can be found. A good starting
point for this optimization is to start at a rather low interpulse distance of e.g.
mm
5 μm (which equals a scanning speed of 50 s at 10kHz). The idea of the
Pulse
technique is shown in Figure 19.

x < xopt x = xopt x > xopt

xlow
cavitation
bubble xopt
boundary
xhigh

target temperature

Figure 19: Variation of the interpulse distance Δ x in order to avoid scattering of a subsequent
laser pulse at the boundary of a cavitation bubble generated by the preceeding laser pulse.
Therein Δ xopt represents the optimal interpulse distance. On the bottom of the picture, a
scheme with dashed lines (close lines depict higher temperatures) depicts the temperature
evolution in the target considering residual heat accumulation.

In case of too small interpulse distances, the laser pulse is likely to be scattered
at the phase boundary of the cavitation bubble. In this case, your precious
productivity will be low unless you increase the interpulse distance. Reaching the
optimal interpulse distance Δ xopt you will just bypass the prior cavitation bubble
increasing your overall productivity. When exceeding this optmal interpulse
distance heat diffusion and preheating of the subsequently ablated area needs
to be considered. Consequently, it is feasible to ablate this already preheated
spot and limit the interpulse distance to an upper value. In case of e.g. 10 kHz
the time between two pulses is 100 μs which is more than enough to have heat
conduction into several hundreds of μm around the previously ablated spot (in
particular for ns laser). A thorough examination of the optimal ablation distance
should be made in this case. Typical galvanometric scanners easily provide

50
 GET MORE: PRODUCTIVITY

lateral scan speeds of 1 m/s, allowing to separate 10 kHz pulses laterally by

100 μm, maybe already enough if you work at moderate pulse energies (causing
moderate bubble sizes).

~~~~~~~~~~~~~~~~~~~~~~~
S t udent: The device isn’ t working , i’ve tried every thing . II
Assis tant: Is it plugged in?

S t udent: Oh

Alex Heinemann, Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
However, in case the scanner being used has high scanning speed of up to
100 m/s (e.g. a polygon scanner), the previously mentioned upper interpulse
distance limit will be rather irrelevant when working in the MHz laser pulse
repetition rate regime. The reason for this statement is the short interpulse
delay of only several tens of nanoseconds being shorter than the heat diffusion
timescale, rendering it impossible to use residual heat from a previous laser
pulse. Simply speaking, the timescale for heat diffusion to take place between
two subsequent laser pulses in MHz regime is just too short. Therefore, when
you’re working with MHz-repetition rates (as modern high-power pulsed lasers
often provide), while having a sufficient scanner at hand, you don’t need to
worry about working at too high interpulse distances and rather run your
scanner at maximum speed. However, in case you have a galvanometric scanner
with vscan << 10 m/s, even your highest scanning speed may not be sufficient
to fully bypass the cavitation bubble in case you work with repetition rates
of several hundreds kHz. This again is strongly depending on the cavitation
bubble size and lifespan which again is based on applied pulse energy, pulse
duration etc. As you may realize, due to the huge number of different possible
parameters depending on the setup chosen, it is hard to find a general rule
directly giving the optimal scanning speed (interpulse distance) for obtaining
maximal productivity. Therefore, following the previously explained technique
the optimal working conditions need to be determined exclusively for the used
setup, target material and liquid you are working in. To keep it simple, when
you work at 100 Hz or below, you do not have to worry about the bubble.
Approaching the kHz regime may render faster movement of beam relative
to the target a good choice to improve both productivity and reproducibility.
Approaching the MHz regime, polygon scanners of latest generation are
demanded to bring every pulse to an undisturbed site at the ground.

51
 GET MORE: PRODUCTIVITY

STEP 6: KEEP IT THIN

During the remarks made in Step 4 it was already mentioned that scattering and
absorption of laser light by nanoparticles is a possible reason for the reduction
of productivity. Actually, this is another big issue when working with a batch
chamber. There are three ways to compensate this effect – either decrease the
ablation time such that the nanoparticle concentration remains low or work
with non-circulation liquid flow or decrease the liquid layer height so that the
II laser beam has a very short path through the liquid containing nanoparticles.
All approaches aim at minimizing absorption and scattering losses due to
nanoparticles or bubbles. Those vapor bubbles should not be underestimated.
In order to maximize productivity and avoid disturbances of the laser beam
during its path through the liquid it is therefore feasible to minimize the final
concentration and additionally decrease the liquid layer height. When working
in an open environment with the laser beam passing a liquid-air boundary it was
already stated in Step 4 that there is a lower boundary for the liquid layer height
to avoid liquid vaporization. Using the setup shown in Figure 20 the case is a
little different. In this setup, the laser passes a glass-liquid phase boundary
(like in a cuvette irradiated from the side or a glass-covered chamber).
Due to both the similar refractive index liquid and the (quartz) glass window and
the high damage threshold of glasses (higher than liquid vaporization threshold)
no direct ablation of the window is expected at its entrance side. However,
window damage is often observed if the liquid layer is too thin. With the setup
using a quartz-glass window the lower limit of liquid layer thickness is mainly
determined by the dimensions of the cavitation bubble and the range of its
mechanical ablation by outward jets ejected from the bubble during collapse.
Usually lowering the distance between target and glass window below 2-3 mm
will result in the abrasion and consequently in the destruction of the window.

~~~~~~~~~~~~~~~~~~~~~~~
We once designed an innovative set up for wire ablation in a flow-through chamber .
However , it had one downside. For evaluating the wire position you had to put your
head directly on the table and look through the optics. I had a s t udent who worked
with this set up for months. One day he felt a slight s tinging on the back of his head
while put ting his head in. Apparently the laser was s till on. From this we can learn 1)
Accident s happen particularly when every thing is routine 2) Do not put your head into
the path of the beam, EVER , but put safet y firs t during design of optical set ups.

Chris toph Rehbock , Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

52
 GET MORE: PRODUCTIVITY

Therefore, it is recommended to not reduce the liquid layer height below 3mm
when using glass windows (air/glass/water boundary) instead of an air/water
phase boundary for the laser to enter the liquid. For maximal LAL productivity,
of course, the liquid layer should be as thin as possible (4 – 6 mm). And liquid
flow always helps.

cavitation bubble shrinKage liguid jet abrasion during collapse possible damage profile
glass window glass window
II
liquid liquid

bubble
abrasion
of glas

target target

Figure 20: Typical glass damage scenario during ablation with a too low liquid layer height. To
the right a typical picture of a glass cover is shown after abrasion occurred.

STEP 7: SCALING IT UP – FROM BATCH WISE TOWARDS

CONTINUOUS PROCESSING
Since we already learned how to find the best operation conditions and
geometry for an ablation setup it is now time to think about optimizing the
whole setup concept. Of course, the widely used batch process is good for
a “quickie”, but will not be very efficient. Due to the previously mentioned
accumulation of nanoparticles during ablation, productivity and nanoparticle
yield per pulse will decrease during the process. Additionally, when thinking by
economical means, emptying the batch chamber after a specific concentration
is reached, re-adjustment and setting it up for another run will consume crucial
and “precious” working hours. In order to circumvent the problem about the
nanoparticle accumulation and the high workload of a scientist, it is time to
think about changing to a more compartible continuous nanoparticle production
using a flow chamber. A simple and very effective setup to achieve a temporally
continuous ablation process is given in Figure 21.

53
 GET MORE: PRODUCTIVITY

ablation chamber sealing gasket

quarz glas window

II
target

back plate front plate

pump

Figure 21: Scheme of a continuous flow processing including a flow chamber cross-section view.

The design is held simple and chambers can be built with basic workshop
equipment. It is recommended to use aluminum plates as a basis material as it
is light and easy to handle when mechanically working it. As basic shape a cube
with 40mm x 40mm x 5mm dimension for each chamber part is convenient.
The chamber is built of 3 basic parts:

• back plat e (holding the target)

• middle plat e (ablation channel for liquid perfusion and particle

transport)

• front plat e (fixing the quartz window)

Additionally the middle piece should be equipped with suiting hose connection
in order to connect tubes to your chamber.

In order to assemble the 3 chamber parts, four drilling holes in the corner of
each part and slid bolds may be used as depicted in Figure 22. The different
plates should be separated by some gaskets preventing leaks. In this context
0.5mm Teflon or Nylon slabs are recommended. In case of persistent leaks,
Teflon paste applied onto the gaskets proofed to be a handy tool to prevent
leaks occurring after assembly. The chamber may be fixed on a breadboard
attached to a 1 axis or 2 axis translation stage using a suitable chamber holder,
facilitating chamber positioning towards the focal plane. With this chamber, you
are now able to easily perform a continuous nanoparticle production.
54
 GET MORE: PRODUCTIVITY

target

window

breadboard

II

Figure 22: Assembly of a simple, but fine working flow chamber to achieve continuous ablation
conditions. Liquid flow is directed upwards. Use Teflon sealings with additional Teflon paste
between the parts.

Of course, the target needs to be changed once it is completely ablated. Yet,

before starting your first experiments, it is highly recommended to anodize all
chamber parts to make sure the surface is covered by an aluminum oxide layer.
Without passivation, aluminum may dissolve into e.g. the water pumped through the
chamber. The dissolved aluminum ions will significantly destabilize your produced
colloids leading to agglomeration (yes, we suffered from this experience). More
details on the stability of nanoparticles will be given in Chapter IV; however keep in
mind that a healthy aluminum oxide layer is very important. Now, before we start
looking into more detail about the performance of the new ablation process let’s
address some issues you should be considering in the first place:

• avoid pH < 4 and pH > 9 in order to keep the aluminum oxide layer intact.
For corrosive liquids, use Telfon or polyamide

• do not use tubes and seals made of silicone and other materials that have
considerable porosity, to minimize nanoparticle losses and contamination
carryover due to deposition inside the pores

• check solvent compatibility table for all seals, tubes or plastics in contact
with the solvent you plan to ablate your nanoparticles in. Recommended
tube materials when using solvents like acetone are Polyamide (PA12) or
Teflon or PVC.

• consider leaching of plasticizers from your tube (i.e. from cheap PVC) and
seals material especially when ablating in organic liquids like acetone, in
order to avoid impurities (one example siloxanes leached from silicones
and seals immersed in acetone – siloxane impurities will adsorb onto
nanoparticles, even after several cleaning steps!).
55
 GET MORE: PRODUCTIVITY

• make sure the pump is clean and specified for the liquids you plan to work
with (e.g. solvent). Pump oil is transparent, not dark-brown.

According to the previously listed points, silicone should be avoided and Teflon
based tubes and seals should be preferred. Those however may provide less
efficient sealing properties and some tinkering to attach the tubes to the hose
connector resulting in leaks during liquid perfusion. To avoid leaks consider
using Teflon paste additionally to the already used gaskets and tubes. Teflon is
II thermoplastic, so it tightens best only during the first mounting.

With your finished sealed setup you’re ready to perfuse liquid through the
chamber. Prior to pumping make sure the inlet is attached to the lower hose
connector following the sketch in Figure 21. In doing so you make sure that all gas
bubbles either sucked in by the pump or generated during ablation are leaving the
ablation chamber at the top without accumulating. In case you decide to connect
the inlet to the top you may end up with a chamber full of gas bubbles.

Finally, you are ready to start your new continuous production line tapping
nanoparticles. This setup helps in minimizing reassembling times as the
interruption of the process is only necessary when the target is consumed.
Considering the reassembly time as a timespan where no nanoparticles are
produced the productivity of your process will be drastically increased when
using this setup as a guide. And you will benefit from reproducibility. In order to
examine the continuously produced nanoparticle, the setup shown in Figure 21
may be upgraded downstream with a fiber UV-VIS set up probing a flow cuvette.
As every LAL process has a tune-in period of a few minutes, we recommend to
waste the first 3-5 min of the colloid stream and then relax while your setup is
doing the work. A batch process also suffers from this tune-in period, but here
extraction of that fraction is impossible

STEP 8: LOWER THE RESIDENCE TIME USING

3D-PRINTING TECHNIQUES
In the previous step, a very flexible chamber and setup design was introduced.
The chamber is easy to access and can be manufactured with a simple tool
kit. However, this design is still not optimal when it comes to consider the
flow profile inside the cell. Now before thinking about better chamber designs
exhibiting a better flow profile, we may first think about the impact of liquid flow
on the productivity of nanoparticles.

In previous sections, the issue of nanoparticle accumulation was already

addressed several times, stating that enrichment of nanoparticles within the
beam path way needs to be avoided in order to minimize absorption and
scattering losses. Within the batch process, we found that our only degrees
of freedom are the liquid layer height and nanoparticle concentration. Using
continuous flow setup we created a new degree of freedom to handle “laser

56
 GET MORE: PRODUCTIVITY

shielding” minimizing absorption and scattering by already produced particles

(and some drifting bubbles). But chamber volume geometry has a big effect,
and needs not much effort to be optimized. Figure 23 displays unwanted
geometries on the left and a laminar flow pattern on the right with minimized
back-flow. As you can imagine, if some particles are moving into the flowing
direction away from the ablation spot, while others are surfing on eddies into
the opposite direction, right back into the ablation spot, large channel width
(and a steep change in channel opening diameter) may not be the best solution.
In case of the chamber design presented on the left in Figure 23 strong levels II
of re-mixing are to be expected due to the small liquid entrance / outlet and
the quick widening of the channel. On the other hand, a more tube like shape
with smooth geometry changes results in a more homogeneous flow pattern. In
order to get an idea of the flow pattern inside the chamber you can simply use
software tools like “open foam” or “ANSYS fluent” in order to create a rough
CFD (computational fluid dynamics) simulation similar to the sketch in Figure
23. A good is to choose a design where the liquid is directly guided towards the
drain avoiding obstacles like corners. A proper design to assure a laminar flow
pattern would be that of a standard pipe having no surface roughness. However,
a tube is far from feasible to act as an ablation chamber holding a planar target.
Yet if we cut off two sides of the cylinder such that one side can be covered with
the quartz glass and the other with the target, a laminar flow pattern without
too much re-mixing is likely to appear. Note that laminar flows have the highest
flow rate in its center, exactly the place where the laser ablation happens and
nanoparticles need to be effectively drained off.

calm down boys!

I’m here since the first laser
pulse bacause the chamber
is not optimized!
B i’ll be
first!

no
I am!
laser

I will
make it!
puh this laser beam
is so hot, i think
i am melting again

Figure 23 (A): Flow pattern inside a steep channel geometry change in the chamber volume
(left) and the optimized chamber design (right). (B): Sketch of fresh particles arriving, while
previously produced old ones are still present.
57
 GET MORE: PRODUCTIVITY

With this new concept in mind you should however avoid attending your
workshop manager in case you want to leave the workshop without bruises.
Actually, the optimized chamber concept, especially the gradual diameter
reduction shown in Figure 23 (left) is not that simple to manufacture. Yet once
you are able to consider yourself to be a proud owner of a 3D printer you’re
able to simply print this chamber within a couple of minutes. Obviously, the 3D
printer won’t complain whatever design you have in mind. In case you decide

II to use a 3D printer it is however crucial to choose the right material especially

when working with organic solvents. A good choice for a printable polymer
material with good acidic, basic and solvent stability would in this case be nylon.
By following all previously given steps (focussing, scanning, liquid level) and
using an ablation flow chamber with optimized flow properties like the one in
Figure 23 (right) included into the flow concept presented in Figure 21 you will be
able to maximize the nanoparticle productivity towards hundred mg nanoparticle
per hour with tens of Watt laser power.

STEP 9: MOVE BETWEEN DIMENSIONS

Since you already became an expert regarding optimizing the laser ablation
of a two dimensional target, it is time to finish this chapter on productivity
maximization with some additional novel concepts regarding the laser ablation
technique. Until now, the presented ablation concept concluded in a continuous
process which significantly decreased assembly times and outages due to target
or liquid change, compared to the initially presented batch process. However, to
think big scale a fully continuous automatic setup is required. Therefore, the last
obstacle to overcome is to get rid of time-consuming target replacement steps.
Now how can we achieve something like that?

The answer to this question may be counter intuitive to our goal but as we are
already in the predicament of going big scale with the small scale one additional
bafflement shouldn’t be too hard to swallow. So prepare yourself! It is time to cut
off unnecessary target dimensions.

Let’s begin with the 2nd dimension and get rid of this large area of metal but
instead consider a very long metal strip of rather small thickness (e.g. several
tens or hundreds of microns, as thin as the focused laser beam). Therefore
let’s decrease the dimensions move to a 1D shape which would be a simple
target wire. Wire ablation has been studied in literature for a while and can be
considered as one of the ablation techniques with the highest potential to be
applied in fully automated, continuously driven ablation processes.

58
 GET MORE: PRODUCTIVITY

wire
feeder

cannula colloid
wire

stock

window II

laser

water

Figure 24: Sketch of a continuous wire ablation setup.

As depicted in Figure 24 wire ablation is done in the same ablation chamber

presented in Figure 22 only with the exception that the back plate holding the
target is exchanged by a second window and a second front plate such that
the laser can pass through the chamber. The wire is to be fed into the ablation
chamber using a cannula inserted at the top. Using a wire conveyor the wire
may be pushed through the cannula while the laser spot position needs to
be adjusted such that the wire tip is being ablated. By controlling the feeding
speed and choosing a wire diameter that is slightly thinner than the focal spot
diameter the wire can be constantly ablated over a long period of time. The
ablation time is only limited by the chosen feeding speed and wire length which
can be chosen to be a couple of 100 m if needed. When properly adjusted the
wire is ablated with 100% material yield without any non-ablated metal waste.
Using the feeding speed set for the wire conveyor, the productivity can directly
be set. And isn’t a continuously fed target a perfect match to continuous flow
and continuously provided laser pulses?

59
 GET MORE: PRODUCTIVITY

Obviously wire ablation already is a very suitable concept and we’ve seen that
with decreasing target dimensions more and more issues disappear. Maybe it is
time to think about cutting off the last dimension as well and move towards a
zero dimensional target. It may sound paradox – in the mathematical definition
a zero dimensional target would be represented by a point in space with
infinitely small diameter taken in absolute units. Naturally, this is not possible in
reality but we may converge when considering decreasing target dimensions up

II to the point where the target particle diameter is considerably small compared
to the laser spot diameter in the working area. This assumption is valid for most
commercially available micro-powders. Using the same setup suggested for the
wire setup, probably leaving the cannula, it is possible to form nanoparticles
from irradiating micro-powders dispersed into a liquid (e.g. water) under
continuous flow conditions (Figure 24).

Figure 25: Comparison of different ablation geometries including an overview of the major
up- and downsides.

Applied to micro- or nanoparticles this technique is commonly discussed as

laser fragmentation in liquid (LFL) as already mentioned in Chapter I and mainly
used to produce particles with downsized particle diameter depending on the
applied laser fluence. Obviously by changing the target dimensions down to
1D or even 0D (Figure 25) we gracefully managed to pass the last obstacles
and introduced a final concept for a 100% continuous process with 100%
nanoparticle yield 0% waste and high productivity potential.

60
 GET MORE: PRODUCTIVITY

Concluding, much effort is necessary during the adventure of climbing the

productivity mountain. The path may be cloudy and every herein discussed step
needs to be taken with diligence and caution in order to make sure not to fall off
the steep cliffs appearing on your way up to the top. However, by understanding
and avoiding the obstacles and barriers presented you may soon be sitting on
the peak of the mountain enjoying the view over the nanoparticle ocean.

II

~~~~~~~~~~~~~~~~~~~~~~~
One day I went into the lab to produce some gold nanoparticles immersed in acetone. I
was a bit worried as acetone is flammable it was by far not my int ention to burn down
the lab. I decided to use the flow chamber – which is a good choice as it is sealed and
ones filled with acetone , it s int erior is free of oxygen. I remembered a short lesson
about safet y issues s tating that a burnable liquid can only burn if both oxygen and
an ignition source are present. Now as the oxygen was ruled out in the sealed flow
chamber I considered this experiment quit e save. While I was ablating using a high
power laser that I know likes to kill the glass cover of the ablation chamber I s till
was a lit tle cautious about the whole process. But thinking - no risk no colloid! I
s tart ed the acetone flow and the laser – surprisingly without any accident s … well at
leas t in the firs t place… I was nearly finished when I heard some s trange deviations in
the ablation sound which hasn’ t been there before. Now I thought – what ever - eyes
shut and go for it … jus t a lit tle more … but this sound … where does it come from? So
I looked a lit tle closer not being aware of that leaky lit tle hole in the glas cover .
Suddenly the laser ignit es this eject ed jet of acetone forming a pret t y big flame
making my heart jump. Lucky for me – as soon as the flame appeared it was gone again
as the acetone jet burned off ins tantly and couldn’ t continue burning as the acetone
flow through the leak was too slow. To avoid further fire spectacles I directly t urned
off the pump followed by the laser . Except of a lit tle shock not much had happened
however nowadays my fingers do not leave the t urn-off-but ton of pump and laser
control in order to t urn every thing of directly when I hear something suspicious.

Sven Reichenberger , Duis burg

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

61
 CHAPTER III

UNDERSTAND IT:
OPTICAL AND SOLID MATTER CHARACTERISATION
Imagine you synthesized one-of-a-kind nanoparticles and spiced them up. What
would you do next? Show them to the world by putting it on your website? Make
a video of them and put it on YouTube (e.g., http://youtube.com/nanofunction)?
Or maybe you would write a manuscript about your particles and submit it
to “Nature”? You should first hold your horses and imagine the following:
someone gave you a shiny diamond ring and disappeared. What would you do
with this ring? Run to the love of your life and propose to her (sorry this story

III only works if “I” am a man and “you” are a woman)? You know that this might
backfire if the ring is fake. The obvious step would be characterizing it or having
it characterized. Is it a gold ring? Is the diamond real? What is its clarity, color
grade, shape, Carat weight? All these questions need to be answered for a
proper estimation of its value and if it’s suitable for the love of your life. Same
thing goes for your nanoparticles. You need an exhaustive characterization to
fully identify and confirm the nature of your products.

The questions that usually arise in front of a sample of nanoparticles with

unknown nature (no matter if obtained by laser ablation or by other methods)
can be summarized as follows (Figure 26):

......................................
What 's on the surface? What is it?

> XPS, F TIR , NMR , TGA > T EM, SEM, ICP-MS,

raman , Ftir , XRD, XAS

how big is it?

> T EM, SEM, DLS,

DC , saxs, OAS

how much do i have?

microbalance
> iCP-ms, icp-oes, oas,

......................................
Figure 26: Illustration of nanoparticles and questions that arise while trying to understand
them. Once this is done, function (e.g. optics, catalysis, biomedicine) can be attributed to
structure. This structure-function-relationship is exactly what everyone is hunting for.

62
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

To answer each of these questions, you can use the different characterization
methods in a step-by-step approach, as described in the following paragraphs
and summarized in Figure 26. Of course, depending on the material and
synthesis objective, the sequence of these questions can change from case to
case. In fact, characterization of nanomaterials is often non trivial and Ancients
would have said that we must proceed “cum grano salis” (latin for “with a grain
of salt”, i.e. use your brain before acting!) to avoid misinterpretation of results
and exponential growth of the research work. In particular, for each technique, it
is important to have very clear in mind the answer to the following questions:

a. What do I see?

b. How does it work? III

c. Which amount of sample is needed?

d. How do I prepare the sample for analysis?

e. Are there any tips/ warnings/hazards?

f. Where to look if you want to learn more

Let’s see how the answer to the above questions can help us to achieve a
perfect knowledge of our laser-generated nanoparticles.

STEP 1: PROVE THAT YOU ACTUALLY HAVE

“ “NANO” PARTICLES
Try to ask any Italian who was abroad: he/she will confirm that not enough to
have pasta, tomatoes, salt and water to make a genuine Italian style pasta...
The same goes for the synthesis of nanoparticles by laser ablation: it is not
necessarily true that shooting the laser at an object immersed in a liquid is
enough to collect a colloidal nanoparticle solution (or make a paper on such
“shoot and run” experiment and ruin your reputation). Therefore, first of all
we need direct evidence on the existence of nanoparticles in our sample.
The “Queen” of characterization techniques for nanoparticles is, without any
doubt, transmission electron microscopy (TEM), because of its unmatched
spatial resolution and completeness of achievable data. Though with lower
spatial resolution, scanning electron microscopy (SEM) permits in most cases
the direct visualization of nanoparticles with size larger than ~ 5 - 10 nm (and
larger particles often may be a mass-dominant side product). A great advantage
of these electron microscopies is that they can analyse a single nanoparticle
at time. A relevant consequence of this peculiar ability is that electron
microscopies are the only methods allowing real and direct “number weighted”
analysis (i.e. by counting nanoparticles one by one), contrary to all the other
techniques that only give volume-weighted results (i.e. measuring signals which
are proportional to the x-th power of the volume of each nanoparticle, with x ≥ 1).

63
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

However, TEM and SEM are not easily or immediately available, and analysis of
one particle by one may be time consuming without the support of appropriate
software (e.g. the freeware ImageJ). Also, EM techniques provide bad statistics
(< 1000 particles of millions in a drop) and tend to over-estimate small
particles. In this case, the quickest and simplest way to check the existence of
nanoparticles in a liquid medium is the optical absorption spectroscopy (OAS).
In fact, visible (vis) and ultraviolet (UV) light interact in a very peculiar way with
objects of nanometric size homogeneously dispersed in a transparent matrix.
Besides, when dealing with nanoparticles dispersed in a liquid matrix, OAS does
not need any sample preparation. A great advantage of OAS is that it can be

III performed also in real time during laser ablation or irradiation experiments,
especially if the absorbance is monitored at a single wavelength, by coupling
a photodiode with a laser beam crossing the sample chamber.

In a limited number of cases, depending on the properties of nanoparticles

under investigation, fluorescence spectroscopy (FS) can be a viable technique
for immediate characterization of nanoparticles which emit light. Indeed,
emission bands are usually located in a smaller spectral range than absorption
bands for the same compound (i.e. fluorescence bands are more specific and
easier to recognize). Besides, position and intensity of fluorescence bands in
semiconductors are distinctive of their size. Let’s look at the features of these
techniques and what we can learn from each of them.

~~~~~~~~~~~~~~~~~~~~~~~
During my research int ernship I was working with centrifugation of colloids to
separat e big particles from smaller one. To avoid unbalances I filled my t ubes all
equally to a certain volume and insert ed them into the t ube holder . For saving the
time I chose the highes t rotation number , which was 5000 RPM. Once I s tart ed the
centrifugation, every thing went so fas t. During the acceleration the sound was slightly
different than commonly and suddenly the centrifuge it self s tart ed to dance on the
table. A moment of shock came into me and my heart beat s s trong and fas t. Immediat ely
af t er a short second of rigidit y I pushed the but ton to s top the centrifuge. Fort unat ely
the speed was jus t at about 100 RPM and the centrifugation s topped before something
worse happened. I opened it and realized that one of the t ube holder was int erlocked
and thus didn’ t have the possibilit y to float freely which caused the unbalances. The
machine seemed so mons trous to me already only at very low speeds, so that af t er this
happening I always checked every thing at leas t twice with care. In the meantime we
purchased a new centrifuge , where the holder are already in the right position and no
int erlocking is feasible. However this experience always makes me wait until the full
speed is reached before dedicating to other work .

Galina Marzun, Duis burg

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
64
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

1.1. Optical absorption spectroscopy (OAS).

a. What do I see?

By OAS, we see if nanoparticles are present in the liquid medium. In the large
majority of cases, OAS is performed with bench-top spectrometers, giving an
averaged information on all the nanoparticles dispersed in the solution, clearly
meaning that it is not a single particle technique.

In case of nanoparticles with well defined absorption bands, OAS gives the
absolute certainty of their presence. If the absorption band is related in some
way to nanoparticles size and shape, this information is also achievable, such as
in case of plasmonic (Figure 27, left) and semiconductor nanoparticles, although III
the precise interpretation of the results requires accurate elaboration of the
experimental spectrum with complex theoretical models.

Some polymer, oxide or wide band gap semiconductor nanoparticles may not
have well defined absorption bands in the UV-visible range. In that case, OAS
can still detect the presence of nanoparticles due to light scattering effect,
whose intensity scales with the -4th power of wavelength ( λ-4, the so-called
Rayleigh scattering curve). Once you’ve seen this characteristic curve (Figure 27,
right), you can be sure to have submicro or nano inside.

b. How does it work?

The principle of OAS is simple: a light beam with well defined wavelength (λ)
is transmitted through the sample, typically contained in a cuvette or on a
transparent substrate. A light intensity detector is placed behind the sample.
The procedure is repeated in a range of wavelengths predetermined by the
operator. At each wavelength, the ratio between transmitted (I ) and incident
light (I0 ), called transmittance (T ), is measured. More useful is the absorbance
(Abs) defined as the

Abs = –log10(T)

because it is proportional to the concentration (C) of absorbing objects

according to the Lambert-Beer law:

Abs = εbC

where b is the length of the path travelled by light in the sample (“optical path”,
it depends on the sample geometry), and ε is the coefficient of molar extinction
(it depends on sample nature, being higher for highly absorbing objects).

65
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

It is worth to stress that two distinct phenomena contribute to the decrease

of I/I0 light “absorption” and “scattering”. Absorption means that photons
-4 are annihilated and their energy transferred to the absorbing object, whereas
scattering means that photons just change their travel direction, and will not
reach the light intensity detector behind the sample, without releasing energy
to the scattering object. In case of “real” absorption, it is often possible to
recognize distinct peaks in the spectrum (Figure 27, left) In case of scattering,
the spectrum shows a typical ~λ-4 trend, without a clearly identifiable absorption
peak (Figure 27, right).
......................................
III absorbance scat t ering
a g nps
a u nps -4
~

C u nps
bs

bs
a

wavelength ( ) wavelength ( )

......................................
Figure 27: Typical OAS spectra of purely absorbing (left) or scattering (right) nanoparticles.

c. Amount of sample needed?

A great advantage of OAS is that, with the use of appropriate cuvettes, 100μL
of solution may be enough, although typically 0.5 – 3 mL of liquid is required.
However, the detection of samples depends on the combination of three factors
in the Lambert-Beer law (b, ε and C). When nanoparticles are present at very
low concentration in the liquid solution, the simplest way to improve the signal
intensity is to increase the optical path (b) of the light beam passing through
the sample. This can be done using cuvettes with 1 cm or even longer optical
path. In ordinary spectrometers, the optimum interval of absorbance is between
0.5 and 2.5. Obviously, ε is a material property and cannot be modified. Most
spectrometers are not linear anymore above absorbance of 2, dilution is
a simple workaround.

d. How do I prepare the sample for analysis?

In the majority of cases, you just need to fill the cuvette with the solution.
However, if nanoparticles are not stable in the liquid phase for the time required
from the analysis (< 5 minutes), they can be included in a polymeric matrix (e.g.
add PVP) and/or deposed on a transparent substrate (i.e. a microscope slide).

66
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

e. Tips / warnings / hazards?

Simplest way to stabilize an aqueous colloid is: add a pinhead-sized peace of

soap (the cheap, solid one, not the coloured and perfumed). Of course, any
organic additive will screen the UV part of the spectrum, same with plastic
cuvettes compared to quartz cuvettes. In water, sometimes sodium hydroxide
helps for stabilisation (adjust pH around 10 or 11), keeping the UV range free for
analysis.

You must consider that absorption and scattering are phenomena strongly
correlated to nanoparticles volume. In particular, absorption scales linearly with
the volume of the nanoparticle, whereas scattering scales with the 2nd power of
the volume. As a consequence, the information obtained by OAS are the volume III
(or volume2) weighted average of the nanoparticles in the sample.

Store and handle it cold (fridge is enough, don´t freeze) to minimize

sedimentation.

f. To know more?

- Fundamentals of modern UV-visible Spectroscopy – Primer; Owen, Tony;

2000; Agilent Technologyes; free at http://www.agilent.com/cs/library/
primers/public/59801397_020660.pdf

- UV-VIS Spectroscopy and Its Applications; Perkampus, Heinz-Helmut;

Springer-Verlag; 1992.

1.2. Transmission electron microscopy (TEM) imaging.

a. What do I see?

TEM allows the direct visualization of a single nanoparticle with nanometric

or subnanometric spatial resolution. In the high-resolution modality (HR-TEM),
atomic planes and rows within the single nanoparticle are visible. This gives
straightforward access to a series of information such as nanoparticle size,
size distribution, structure homogeneity or, alternatively, phase segregation,
formation of shells, crescents and heterostructures, crystalline order, lattice
defects.

b. How does it work?

The principle is the same of an optical microscope, with the remarkable

difference that a monochromatic electron beam is transmitted through the
sample, instead of a photon beam. Electrons are scattered (sometimes also
absorbed) by nanoparticles, with an efficiency which depends on the square
of the atomic number (z) of elements composing the nanoparticle, and on their
crystalline order and orientation. This generates a contrast between electron
dense and not dense regions (i.e. an inorganic nanoparticle and a thin carbon

67
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

film acting as substrate), which is at the basis of image generation. As in the

optical microscopy, spatial resolution depends on the wavelength of electrons,
which is determined by their kinetic energy according to the De Broglie equation
(corrected for relativistic effects due to the high speed of electrons in a TEM):

where h is Planck constant, m0 electron mass, c light speed and E the energy
of the accelerated electron. The consequence of the De Broglie equation is that,
usually, nanometric spatial resolution is possible with TEM operating at 100 kV
III of accelerating voltage, whereas subnanometric resolution is possible only with
acceleration voltages of 200 kV or higher.

c. Amount of sample needed?

Being a technique with single nanoparticle capability, virtually 1 nanoparticle

is enough. Obviously a statistic analysis of each sample is possible only by
considering more than N~500 nanoparticles each time. In practice, 1 – 10 μL of
solution are always enough.

d. How do I prepare the sample for analysis?

Nanoparticles must be deposed on TEM grids, which host a thin film with high
transparency to the electron beam, such as nanometric thick carbon films or
Si3N4 membranes. A drop of solution can be casted on the grid and left drying in
air (TEM operates in vacuum to avoid interference of atmospheric gas with the
electron beam, therefore samples must be dried prior to analysis). If there are
non-volatile organic residues in the sample, prior plasma cleaning (standard TEM
sample preparation equipment) may be required to improve the contrast.

e. Tips / warnings / hazards?

The electron beam depositeda considerable amount of energy on the

nanoparticles, and this energy increases with increasing magnification (the
energy density per unit area increases). Therefore, care must be used when
observing materials which undergo modifications under electron beam
illumination (especially organic nanoparticles and semiconductors). Even noble
metals or refractory materials can undergo modification, melting, segregation,
regrowth and etching when exposed to an electron beam. This is especially true
for laser-generated nanoparticles, whose surface is clean, uncoated and very
reactive, and whose composition can be metastable. Nanoparticle modification
can usually be observed in real time during TEM imaging and can be reduced
only by lowering the magnification (i.e. decreasing the energy density for unit
area), the exposure time and the TEM operating voltage. Another workaround is
dilution. If you see peanut-shaped gold particle twins, they often are artefacts
from fusion caused by the electron beam.

68
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

Note that today´s SEMs provide transmission mode as well, it´s called “STEM” (ask
operator if the instrument is equipped with this option, its also not too expensive
to buy later for modern instruments, because it´s essentially just a sample
holder), and before said for TEM is true as well for this STEM technique.

f. To know more?

- Transmission Electron Microscopy: A Textbook for Materials Science; Williams,

David B., Carter, C. Barry; 2009, Springer-Verlag.

~~~~~~~~~~~~~~~~~~~~~~~ III
He checked out his set up and examined all the optical part s on the table. He s tart ed
the laser and fright ened because of a machine gun-like regenerat ed noise. Af t er he
switched off the laser he examined what happened. It took him several minut es to find
out the reason but he realized that his laser had drilled some holes in the coating of
the laboratory door . Obviously a loose mirror had twis t ed independently.

Sebas tian Kohsakowski, Duis burg

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1.3. Scanning electron microscopy (SEM) imaging.
a. What do I see?

It gives the size and morphology down to the single nanoparticle level, for
objects larger than about 5 – 10 nm. Due to the spatial resolution of ~1-10 nm,
morphological information is reliably extracted only for larger objects (>10 nm).

b. How does it work?

A focused electron beam is scanned over the sample area, and a detector
collects the electrons emitted from the sample. There are two types of electrons
emitted from the sample under electron beam illumination, and both can be
used for image formation in a SEM: backscattered and secondary electrons.
Backscattered electrons are high-energy electrons of the incident electron beam,
reflected from the specimen by elastic scattering interactions with sample atoms.
Secondary electrons are extracted from atoms in the specimen by an inelastic
scattering process promoted by the interaction with the electron beam. Each type
of electrons is collected by a distinct detector, which works pretty much like a
photocamera that collects the light reflected by a shiny three-dimensional object.
Therefore, SEM images contain information on surface topography of the sampled
area and show the typical “light and shade” effect due to the directional imaging.

Backscattering is more intense with high – z elements, providing complementary

information to secondary electron imaging in samples where high – z compounds
are deposed on low – z substrates. In this imaging modality, heavier atoms
appear brighter.
69
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

c. Amount of sample needed?

Same as TEM, few μL of solution is enough.

d. How do I prepare the sample for analysis?

Same as TEM, a drop of solution can be casted on a substrate and left drying
before analysis (also SEM operates in vacuum). The substrate should be as flat
as possible (in this way, topographical information will enhance nanoparticles
over the flat surface), conductive, and should allow z-contrast with your
nanoparticles. For instance, carbon tape for electron microscopy is suitable
for oxides, and doped Si is very well suited for metals and almost all types of

III nanoparticles with elements heavier than silicon.

e. Tips / warnings / hazards?

Specimen topography is greatly affected by any impurity or solute (i.e. salts,

surfactants, chemicals, synthesis by-products) eventually present in the solution
and forming a matrix embedding nanoparticles in the dried sample. The smaller
the nanoparticles are, the higher must be the purity of the liquid.

You get nice pictures from your particles if you tilt the sample to the maximum
possible: in this way you will get nice 3D appearance. Note that this modus it is
not accurate to measure height, it´s just to make it beautiful.

f. To know more?

- Scanning Electron Microscopy and X-ray Microanalysis; Goldstein, J., Newbury,

D.E., Joy, D.C., Lyman, C.E., Echlin, P., Lifshin, E., Sawyer, L., Michael, J.R.; 2003,
Springer.

1.4. Fluorescence spectroscopy (FS).

a. What do I see?

Only a limited number of materials shows appreciable fluorescence: in these

lucky cases, FS easily allows the detection of these compounds in your sample.
In case that fluorescence bands are size-dependent, direct proof of the
presence of nanometric particles is achievable at the same time!

b. How does it work?

In FS, photons are emitted by the sample as a consequence of a light absorption

process. Therefore, monochromatic light is extracted from a lamp and conveyed
to the sample, where it is absorbed by the fluorescent compound (fluorophore),
whose electronic structure passes from the ground state to an excited state.
Fluorophores have the ability to release an appreciable fraction of the absorbed
energy by emission of other photons, which have longer wavelength (lower
energy) compared to the absorbed ones. The photons emitted by the sample

70
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

are dispersed by a diffracting optic and counted by a photodetector. The

fluorescence spectrum reports fluorescence intensity (in arbitrary units) versus
photon wavelengths. Absolute fluorescence efficiency (defined as the ratio of
emitted to absorbed photons and called “quantum yield”) is achievable only
by comparison with a standard with known quantum yield in the same spectral
range of the sample.

c. Amount of sample needed?

Typically 2 – 3 mL of liquid solution with absorbance <0.15 in the excitation and

emission spectral regions. The solution is placed in a 1 cm x 1 cm cuvette.

d. How do I prepare the sample for analysis? III

As in OAS, you just need to fill the cuvette with the solution. However, if
nanoparticles are not stable in the liquid phase for the time required from the
analysis (< 5 minutes), they can be included in a polymeric matrix and deposed
on a transparent substrate (i.e. a microscope slide). The slide can be analyzed
with a ~30° or ~60° geometry to the direction of excitation and of collection
(which are at an angle of 90° one each other). Avoid placing the film at 45°
to minimize reflection of excitation beam toward the detector.

e. Tips / warnings / hazards?

In principle, fluorescence intensity is proportional to the concentration

of fluorophores. However, self-absorption takes place already at very low
absorbance. Therefore care must be used in the choice of the absorbance of
the sample for analysis. Absorbance higher than ca. 0.15 in the absorption and
emission spectral regions usually results in artefacts due to self-absorption.
Run calibration curves to find a robust concentration range of linearity.

f. To know more?

- Principles of Fluorescence Spectroscopy; Lakowicz, Joseph R.; 2006, Springer.

}
~ · /=\{

71
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

~~~~~~~~~~~~~~~~~~~~~~~
The person prepared his sample and measurement device for the automatic pH dependent
zeta potential measurement, to find out the isoelectric point of his sample. He lef t the
device to go to lunch. Af ter one hour he came back with his colleague, examines the
measurement and immediately began to grumble on the device. The zeta potential was
only measured for one pH-value. His colleague examined the device and started laughing.
His college forgot to switch the pH electrode from buffer solution into the device.
III Sebas tian Kohsakowski, Duis burg

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

STEP 2: MEASURE THEIR SIZE

Nanomaterials take their name from their size, hence sizing the nanoparticles
of our sample is an obvious step. Therefore, an obvious question arising about
your colloid is “What is the size of your nanoparticles?”. Pulsed laser ablation
generally yields not one defined particle size but a whole “zoo” of particles
composed of different sizes. So a more meaningful question in this context would
be: “How many particles of which size can be found within your colloid ?”. This
correlation between the frequency of particles in a colloid and the individual
particle sizes is called a particle size distribution. The main points raised in this
paragraph will be how to interpret particle size distributions, how to measure
them and finally how to alter them by simple methods during laser ablation
in liquids.

Expressing particle size distributions

Particle size distributions can give lot of interesting information about your
colloid. In this context, however, you need to keep in mind that there are at
least three meaningful ways to express particle size distributions in colloidal
science, which require weighting based on number, volume and surface.
In a number weighted size distribution you basically count the number of
particles in your sample associated to a certain size class, e.g. all particles with
a diameter of 20 nm and plot this number against the respective particle size.
If you repeat this procedure for all particle sizes in your sample, you will end up
with your number weighted size distribution as illustrated in Figure 28 A.
Number weighted size distributions are generally highly relevant as they are the
direct result of size distributions obtained from electron microscopy. And this is
the standard in literature. Unfortunately, for most “nano-functions”, this statistic

72
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

is less meaningful (and sometimes without any meaning), as most functions are
proportional to either the volume or mass (e.g., ion release, optical effects), or
the surface (catalysis, bio-response, …) of the particles. In a volume weighted
size distribution, on the other hand, you basically plot which volume is filled by all
particles of a certain size class, e.g. which is the combined volume of all particles
with a diameter of 20 nm, and plot this against the particle size (Figure 28 C).
Volume weighted size distributions are highly relevant as they are generally the
data obtained from light scattering measurements. In a surface weighted particle
size distribution (Figure 28 B), the particle surface area found in a certain size
class is displayed. In case you know the geometry of your particles, conversion
of the different size distributions is pretty straightforward. During laser ablation
you are generally lucky as most nanoparticles can be considered spherical and III
Vsphere=4/3 π r3 and Asphere= 4 π r2 are the only equations you will need. When you
compare the particle size distributions depicted in Figure 28 you will notice that
volume weighted size distributions generally over represent larger particles, while
number weighted size distributions are shifted to smaller particles and surface
weighted size distributions are located in the middle. This is a universal rule
and it is important to consider this correlation when further interpreting particle
size distributions.

......................................

C
A

B
VOLUME
NUMBER

SURFACE

SIZE
SIZE
SIZE

......................................
Figure 28: Illustration of particle size distributions for a typical example of laser-generated
particles with 2 different particle sizes (bimodal): (A) Number weighted, (B) Surface weighted,
(C) Volume weighted particle size distribution. It’s the same sample, and all results are “true”.

73
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

As you have now learned about the different types of particle size distributions
you may now pose the question: “Which of these distributions do I use and
when?”. In general, all particle size distributions carry the same information
and are equivalent. However, in some cases one or the other may be physically
more meaningful. For example, when you want to support metal nanoparticles
on metal oxide carriers (e.g., for heterogeneous catalysis), this is usually
expressed as mass load and hence volume distributions should be used as
volume is directly proportional to mass. On the other hand, in many relevant
areas of colloidal science e.g. toxicity, binding of surface ligands and catalytic
reaction rate, the surface area is the driving factor and surface weighted size

III distributions should be used in this case. So there is no simple answer to this
question as it highly depends on the addressed application, this is something
you can choose. The next thing you may want to learn is how to interpret
particle size distributions.

Interpreting particle size distributions

In all cases, nanoparticles will not be identical to each other, and the average
87 size and the size distribution can be used to describe our sample. By measuring
x452 the size of a reasonable large number of particles (N), generally larger than
300, a size histogram can be generated. When particle size distributions is
fitted with an appropriate mathematical model, its interpretation is much
more straightforward. Nanoparticles obtained by laser ablation usually show
lognormal size distribution fLN(d):

where dmax is the point of maximum and w is the parameter determining

the width of the distribution. Conversely, a Gaussian size distribution fG(d)
is often observed in nanoparticles obtained by laser irradiation (e.g. after
laserfragmentation):

where is the mean size and σ its standard deviation.

In general, you have multiple options to present the most abundant (average)
particle size of a size distribution fitted by a gaussian or Log-normal curve.
The first and most obvious way would be to look at the peak maximum (mode)
and to pick the value with the highest abundance and to blank out all other
particles. Other common ways of presenting average values in particle size
distributions are the mean and the median value. While the mean represents the
sum of all particle sizes, divided by the number of all particles, the median or

74
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

D50 represents the value where exactly 50% of the particles are larger and 50%
are smaller. Generally mean > median > mode is a rule of thumb. The difference
between these values is illustrated in Figure 29.

Indeed, the most general parameters for identification of nanoparticle size

(derived from Gaussian distribution) is the average size expressed as the
arithmetic mean and its standard deviation σ, expressed as

The advantage of and σ are twofold: 1) they are of immediate interpretation

III
for the comparison of different samples with different size distributions, and
2) they can be always extracted from any dataset. The disadvantage is that
only the size histogram tells you all the truth on the polydispersity and mono
or multimodality of a sample. Therefore, additional useful information are the
size range (i.e. minimum and maximum size observed) and the width of the size
distribution, obtained by fitting the size histogram with lognormal or Gaussian
curves.
......................................

A B
FREQUENCY

FREQUENCY

MODE = MEAN = MEDIAN SIZE MODE < MEDIAN < MEAN SIZE

......................................
Figure 29: Illustration of mode, median and mean values during interpretation of (A) Gauss and
(B) log-normal fitted particle size distributions.

Based on what you have learnt so far, you can definitely say that the
presentation of an average particle size is far from trivial. You need to
consider the type of particle size distribution as well as the mathematical
model. Consequently you will have to be extremely careful when presenting
and particularly when comparing average values from different studies. And if
someone tells you about the size he made: ask how it was weighted and how the
maximum was found. Meaningless: “size of 3 nm”. Precise: “Number-weighted,
lognormal mode particle diameter measured by TEM” or “Surface-weighted
Gaussian mean diameter of 3 nm +/- 1 nm, as measured by XY”.

75
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

The second point which needs to be addressed about particle size distributions
is that you need to differentiate between the modality and the dispersity.
Modality indicates whether there are two or more distinct populations of
particles within a particle size distribution, which can be clearly differentiated.
The clearest indication is a disruptive size distribution with multiple peaks.
Based on this, a colloid can be characterized as monomodal (having one peak/
most abundant size) or bi-, tri-, multimodal (having more than one peak/
most abundant size). Interpretation of multimodal size distributions usually
necessitates fitting of every mode with an individual fit function, an endeavor
which may be tricky, particularly in case there is a certain overlap between the

III different modes.

On the other hand, dispersity indicates how broad your particle size distribution
is. Dispersity is generally used to characterize monomodal size distributions
or individual modes within a multimodal distribution. The difference between
dispersity and modality is illustrated in Figure 30. So within a multimodal size
distribution, each mode can be characterized concerning its dispersity. Suitable
ways to characterize this value are the standard deviation and the variance.
Both can be taken directly from your fit function (mostly log-normal) and give
you a rough idea about to what extent the different particle sizes scatter around
the average value, generally the higher the variance and the standard deviation,
the broader the size distribution. Based on these values one can define the
polydispersity index (PDI), calculated via: PDI = variance/mean2. In this formula
the Gaussian “mean” may be substituted by the Lognormal mode (xc). Note that
variance is usually displayedin units of nm².

~~~~~~~~~~~~~~~~~~~~~~~
My s t udent and I were performing a fragmentation experiment. I was jus t checking in
on him when I scent ed a funny smell . Af t er some time I realized the bright spot on my
sleeve originating from some laser reflection. It was a mirror that some other s t udent
had muddle headedly put in the wrong box.

Sven Reichenberger , Duis burg

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

E.g. a diameter of “5 +/- 2 nm” tells: mean of 5 nm and standard deviation is

2. But 2 nm is not the variance! Depending on the software you use, you might
not get the variance or standard deviation. For instance, OriginPro gives you the
width (w) of your Gaussian fit, which equals to two times the standard deviation.

76
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

The variance (that you need for a PDI calculation) in turn is obtained by
squaring the standard deviation (in case of Gaussian distribution). Hence, 5 nm
+/- 2 nm is monodisperse (PDI = 0.16), but 5 nm +/- 3 nm is not (PDI = 0.36).

Based on their PDI, particle size distributions can be characterized as

monodisperse (having a narrow size distribution and a PDI<0.3) or polydisperse
(having a broad size distribution PDI>0.3). However, you need to keep in
mind that these definitions are by no means fixed within the nanoscience
community and can greatly differ based on what PDI values are achievable by
the corresponding standard technique. For example, you may proudly call your
laser-generated colloid monodisperse based on the above mentioned definition,
however a chemist synthesizing oligoatomic clusters with atomic precision or a III
physicist requiring very narrow size distribution e.g. for optical application may
tend to strongly disagree with your definition even laugh at you or even worse,
reject your beloved manuscript during peer review.

A MODE 2
FREQUENCY

MODE 1

MODALITY
SIZE

B
FREQUENCY

DISPERSITY
SIZE

Figure 30: Modality and dispersity: (A) Illustration of a bimodal particle size distributions. (B)
Illustration of a monomodal and polydisperse particle size distributions.

Note that it is easier to achieve monidisperity with number-weighted histograms

than with volume-weighted histograms. If you hunt for very high quality, rate
your own colloid by volume-weighted criteria. The other statistics will always be
better, so that you´re on the safe side.

So up to this point it can be concluded that asking for “nanoparticle size” is a

huge simplification and needs to be expanded to the question of the particle
size distributions within the colloid. Furthermore, you have learned how to
express and term a particle size distribution weighted by number, surface
or volume and how to analytically interpret them based on average value,
modality and dispersity. Now you know about the interpretation of particle size
distributions, a suitable follow-up should be how to measure them.

77
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

Altering particle size distributions

There are multiple strategies suitable for controlling particle size distributions
in laser-fabricated colloids, while some of those were already mentioned in
Chapter I. These strategies are basically: I) post processing methods like e.g.
centrifugation or laser post irradiation where size distributions are changed
after the laser process and II) In situ methods where particle size distributions
are changed during the process. While the focus in Chapter I is on ex situ
methods, here emphasis will be on in situ methods which are easy to implement
during basic laser ablation synthesis.

III I) Influence of ablation time: Particle size may be changed by altering

the ablation time. Even though the effects are not very strong, usually
changes of a few nm can be achieved, however this strategy is very
easy to use. As a rule of thumb you can memorize that the particles
become larger with increasing ablation time. The reason for this is
pretty obvious. The final particle size is driven by ripening and growth
processes in the solution. In case the concentration of particles are
higher due to long ablation times, these growth processes are more
pronounced, yielding larger particles (unless you use high repetition or
high-intensity ultrasoft pulsed lasers that cause downsizing by laser
fragmentation during laser ablation). If you want to avoid concentration
supersaturation, simply use liquid flow and drain the colloid off the
chamber.

II) Influence of the solvent: laser ablation experiments can be conducted

in multiple solvents. The critical parameter controlling particle size, in
this context, is the in situ stabilization of initial nanoparticles by solvent
molecules. The stronger the interaction, the smaller the particles you
obtain. For example, the ablation of gold in acetone yields smaller gold
nanoparticles than in water, probably due to an efficient adsorption of
solvent molecules on the particle surface.

III) Influence of pH, ionic strength and stabilizing ligands: as a general rule
of thumb you can remember that whatever stabilizes your nanoparticles
in solution usually also induces a size quenching effect. As described
previously, stabilization can be efficiently done by adding ligands and
low salinity electrolytes with highly polarizable anions. Furthermore, it
may be concluded that a pH far from the isoelectric point is good for
stability. So you can simply use the same approach to efficiently reduce
the average size of your particle size distributions as well as the PDI.
Typical additive concentrations are in the sub-millimeter range for size
quenching.

78
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

Measuring particle size distribution.

Here you can basically differentiate between four types of methods: electron
microscopy, dynamic light scattering, analytical disc centrifugation and small
angle x-ray scattering.

According to what was stated in the previous paragraph, TEM (and SEM in case
of particles larger than 5 – 10 nm) allows direct imaging of nanoparticles and
consequent identification of their shape and size distribution. However, laser
synthesis frequently gives a sol, i.e. a colloidal dispersion of solid particles in
a liquid matrix. Although the sol can be stabilized by the presence of surface
charges on particles surface, the formation of agglomerates of nanoparticles
with final size larger than that of its single constituting “monomers” is very
III
frequent in laser generated nanomaterials (as will be discussed in Chapter IV).
Due to spontaneous soft agglomeration or hard aggregation of nanoparticles on
a substrate during solvent evaporation, the size of these assemblies in the liquid
phase is not accessible by SEM. It can be accessed by TEM only using cryo-
microscopy (a type of microscopy carried out below the freezing temperature
of the solvent, which is very time consuming and expensive!), or by embedding
the nanoparticles in a polymeric matrix which can “artificially” freeze the
aggregation state in solution (you will find literature reports about how to do it
in detail).

~~~~~~~~~~~~~~~~~~~~~~~
S t udent: I think the laser doesn’ t work anymore. For three samples I didn’ t ablat e
any thing even af t er 15 minut es.

Assis t ent: Did you remove the prot ection cap of the scanner?

S t udent: Yes.

Assis t ent: Did you press „s tart “ at the laser? S t udent: Yes.

Assis t ent: Did you t urn on the scanner?

The s t udent made a thunders truck face and moved back to the laser without any further
comment s.

Nina Million, Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The size (d) measured by electron microscopy is the geometric size, also called
Feret diameter (Figure 31), and it is defined as the distance between the two
parallel lines tangential to the object, as if you use a caliper to measure particle
diameter.

79
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

In comparison to all other techniques, electron microscopy has a couple of

striking advantages. The main point is that it will give you a direct insight
on the shape of your nanoparticles and tell you the exact size of the metal
core (Feret diameter). However, electron microscopy also suffers from some
drawbacks when it comes to particle size distributions. First, unless you use
cryomicroscopy or a trick such as inclusion in a polymeric matrix, it works on
dried samples, which means you can obtain no information of your particles
in the colloidal state. For example, information whether your particles are
agglomerated in the colloid are very difficult to verify with electron microscopy
as particle agglomerates you see in the images could also form during sample

III drying. The next drawback you have to face with electron microscopy is limited
statistics. When you look at typical TEM images of nanoparticles, you usually
find, depending on the resolution, up to a few hundred particles per image.
Consequently, you usually characterize less than 103 particles per sample.
Just for comparison, 1 mL of a gold colloid with a mass concentration of
50 mg/L (that’s already reddish in color) and a particle size of 10 nm will
contain about 5·1012 particles, which means you will have characterized only
~10-8 % of your sample with TEM. As a result of this, proper TEM image
interpretation necessitates that your sample is very homogeneous and thorough
mixing prior to loading your sample on the carrier is absolutely vital. In addition,
even this characterization is very labor intensive because you usually have to
measure all particles by hand. Even though there are algorithms for automated
analysis of TEM images, they usually fail for non-uniform particles or overlapping
particles, which are unfortunately quite common in laser-generated colloids. So
your will have to get used to spending hours staring at your computer screen.

Other techniques exist that can measure nanoparticles size directly in the
colloidal solution, without the need for solvent evaporation. One of these is
small angle X-ray scattering (SAXS), which is able to simultaneously extract the
size of aggregates (secondary particles) and of monomers (secondary particles)
contained therein, although at the price of complex data analysis.

~~~~~~~~~~~~~~~~~~~~~~~
S t udent: It takes ages to transfer always 1.5 mL with the 1 mL pipet t e. I have always to
change between 0.5 and 1.

Assis tant: Jus t put it on 0.75 mL .

Alex Heinemann, Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

80
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

In addition to the Feret diameter, also the hydrodynamic size is important for
nanoparticles in a liquid solution. The hydrodynamic size, also called Stokes-Einstein
size, is defined as the size of a hard sphere that diffuses in the liquid medium at the
same rate as that of the object (i.e. the size of the sphere with the same diffusion
coefficient D of the object, Figure 31). There are two techniques which are sensitive
to the hydrodynamic size: dynamic light scattering (DLS) and disc centrifuge
(DC). Remarkably, DLS and DC also give information on the hydrodynamic size of
nanoparticles coated by a layer of stabilizing molecules (the typical example is that
of organic ligands grafted on the surface of noble metal nanoparticles). The shell of
organic ligands has lower electronic density than the inorganic core, which makes its
detection with electron microscopy or by SAXS uncertain. Besides, the ligand shell
often has a “soft” structure, meaning that the size in the liquid phase is generally III
different than after drying of the sample for TEM/SEM imaging. It is worth to observe
that also in case of charged ligand free nanoparticles, the hydrodynamic size may
differ slightly from the Feret one, due to the presence of an electrochemical double
layer slowing Brownian motion or sedimentation in centrifuge (mimicking smaller
primary particles as would be expected).

Therefore, in comparison to electron microscopy, DLS and DC can characterize

nanoparticles in their colloidal state. This has a few nice advantages: DLS and DC are
generally much faster, do not need sophisticated sample preparation and possess
better statistics as they are prone to characterize a huge collective of particles. Also,
most colloids are tested for functionality in liquid state, to DLS and DC assess a
size that is often more close to the colloid´s in operando state. However, you need
to be careful when interpreting the results because size distributions from electron
microscopy and light scattering are fundamentally different:

I) Light scattering characterizes colloidal particles consisting of the metal/

hard matter core as well as a shell of counter ions or potential organic
ligands. This sum of core diameter and shell is called hydrodynamic diameter,
which of course is larger than the Feret diameter obtained from electron
microscopy. So as a rule of thumb you can memorize that particle size
distributions from light scttering are always larger than those from electron
microscopy. This difference is usually negligible when completely ligand-free
particles are evaluated, however, differences by a factor of two or three can
be found in case bulky ligands like proteins or polymers are used.

II) Light scattering can usually not differentiate between single particles and
agglomerates/aggregates dispersed in solution. This can lead to huge
differences between particle size distributions from electron microscopy
and light scattering. Therefore, whenever you encounter discrepancies
between electron microscopy and light scattering techniques, which are too
big to be explained by a hydrodynamic shell (in most cases the hydration
shell is even less than a nanometer, and ligands or sufactants have typical
length of only a few nanometers), agglomeration/aggregation in the
dispersion is the most probable cause.
81
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

While electron microscopy gives direct access to the particle size, light
scattering yields indirect results which originate from an approximation by a
mathematical model. In case you do not properly understand these underlying
models, data can be easily misinterpreted. Hence it is mandatory that you
critically evaluate all data obtained by light scattering.

It is now clear that the definition of nanoparticle size may be tricky, unless you
don’t have in mind that it is important to discriminate between:

i) the Feret diameter of the objects physically surrounded by the liquid

environment in the sample (which often is an aggregate of “monomers”),

III ii) the Feret diameter of constitutive units (“monomers”) in the aggregates,

iii) the hydrodynamic size of the aggregates (or of monomers if no aggregation

occurs), which also contains the contribution of the ligand shell if present.

Depending of which size we are interested in, we must use the appropriate
investigation technique, and none of the techniques is generally better or worse
for colloid analysis, but they bring complementary information. So it is very
useful to verify your particle size distribution always (!) with at least two different
methods. It is very unlikely that 2 methods give the same value. Instead, by the
difference in diameters, you will learn a lot about your sample and may give
impressively precise statements in your thesis or manuscript.

monomer (primary particle)

feret size
dobject = dsphere

feret size hydrodynamic equivalent

of the aggregate sphere size
(secondary particle) of the aggregate

Figure 31: Understanding nanoparticle “size” coming out from different techniques.

2.1. Dynamic light scattering (DLS).

a. What do I see?

The total (core + shell) average hydrodynamic size of all the objects dispersed
in the sol, and an estimation of their size distribution. Typically, size distribution
can be weighted on the volume or on the number of nanoparticles. However DLS
is not applicable to very small nanoparticles (i.e. below 5 nm).

82
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

b. How does it work?

The physical principle exploits here the fact that the diffusion coefficient of a
particle is size dependent, i.e. small particles diffuse faster than large particles.
DLS exploits a coherent monochromatic polarized laser beam (typically in
the red) to collect the light scattered by the nanoparticles in the sample. The
scattered beam forms a speckle pattern on the detector area, i.e. a pattern
with inhomogeneous density of photons due to the occurrence of constructive
or destructive interference in different parts of the detector area. In fact, the
scattering objects have different position in the sample, and light scattered from
each object will have a different phase on the detector surface. By repeating the
measurement at different time intervals, time evolution of scattered patterns
can be correlated to the Brownian motion of the scattering objects (i.e. to the
III
random walk of particles suspended in a liquid, resulting from their collision
with the liquid molecules or with other particles). It is intuitive that scattering
signals coming from smaller particles change/fluctuate much more quickly than
those stemming from larger particles. Based on this, the instrument acquires an
autocorrelation function, indicating how quickly signal intensity changes and
based on this calculates diffusion coefficients and particle sizes.

The use of complex mathematical functions can be applied to the time

dependent scattered pattern to extract the diffusion coefficients D of the
scattering objects, and their relative distribution. Finally, the size of the objects
is extracted from D by using the Stokes – Einstein equation:

with T the temperature, kB the Boltzmann constant, ƞ solvent viscosity and R

S
the equvalent sphere radius.

However, no worries: the mathematical analysis is done automatically by the

instrument, and the operator just needs to set parameters, place sample and
launch analysis with the mouse. The underlying software for data interpretation
is designed to work even for unskilled personal. The main advantage of DLS is
that measurements are very fast and results can usually be obtained after a
few minutes. In fact, DLS is the standard technique for the characterization of
size distribution in colloids, and this type of instrument is frequently found in
research labs.

However, the technique has one striking disadvantage which can be particularly
critical when analyzing colloids from laser ablation experiments, where you
may be faced with polydisperse or even polymodal size distributions: DLS
measurements, the scattering intensity is generally proportional to the sixth
power of the particle diameter! Even though the underlying algorithm considers
this correlation, the light emitted from small particles in the presence of larger
ones may be too weak to be detected (the detector is dazzled). Just to illustrate

83
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

this point, let us assume you intend to analyze a sample collective containing
5 nm and a 50 nm particle with DLS. In this case the light intensity from the 5
nm particles is a million times lower. Based on this it can be concluded that
DLS often fails in characterizing small particles in the presence of large ones.
Note that the widely used Malvern DLS instrument always puts out artificially
symmetric Gaussian distribution around a calculated peak value, independent
on how asymmetric your sample size distribution looks like.

c. Amount of sample needed?

Approximately 1 – 2 mL of solution are enough, at a concentration which allows

scattered light to cross the sample without considerable re-absorption. Actually,
III the use of high concentration and high absorbance solutions is not compatible
with DLS: in these cases dilution of the sample until convergence of DLS results
is suggested. Obviously, excessive dilution will result in a weak scattering signal
and noisy/not reproducible measurements: the presence of few dust grains in
solution can completely alter the results in this case!

d. How do I prepare the sample for analysis?

You just need to fill the DLS cuvette with the colloidal solution, as in OAS.

If you don´t see the small particles, try again after dilution. Typical concentration
threshold (e.g. for gold) is 20 mg/L, better 50 mg/L. Use a simple syringe dead-
end-filter (e.g. with 400 nm pores, often called virus filter) to quickly remove the
dust. Repeat measurement after 1-2 days.

e. Tips / warnings / hazards?

In case of samples with a wide distribution of size, the largest size will dominate
the scattering response, and the only reliable size distribution is the volume
weighted one. In this case, DC can give more reliable information because the
sample is sorted in size during the analysis.

For the same reason, the presence of large objects in the liquid, such as dust
grains, must be avoided.

f. To know more?

- Dynamic Light Scattering: Applications of Photon Correlation Spectroscopy;

Pecora, Robert; 1985, Springer-Verlag.

- Dynamic Light Scattering: An Introduction in 30 Minutes; Malvern Inc.; free at

http://www.malvern.com/en/support/resource-center/technical-notes/
TN101104DynamicLightScatteringIntroduction.aspx

84
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

2.2. Disc Centrifuge (DC)

a. What do I see?

DC gives the full size distribution of the sample and is compatible to

polydisperse and multimodal colloids. In case of high density nanoparticles, DC
is not applicable for sizes below 5 nm, and this limit increases with decreasing
particle density.

b. How does it work?

In a DC, particles are separated by size using centrifugal sedimentation in a

liquid medium. The basic principle is the separation of particles of different sizes
in a gravitational field, generated by a rotating disk. Naturally, larger particles
III
will sediment more quickly than small ones. The sedimentation is promoted
by the centrifugal force generated by the rotation of a discoid cell containing
the sample (in the inner part) and a liquid phase with increasing viscosity from
the center to the extremity of the disc (Figure 32). A sucrose initiated density
gradient within the disk ensures that individual size fractions of particles
uniformly move to the edge of the disk. The cell and the liquid phase are
optically clear so that, when particles approach the outside edge of the rotating
disc, where an optical detector is located, they are detected by absorption
or scattering of a light beam passing through the disc at a fixed position. The
change in light intensity recorded at a fixed wavelength over time is converted
by the software into a particle size distribution, because the time required for
nanoparticles to cross the light beam is a function of the hydrodynamic particle
size and medium viscosity. In order to correlate sedimentation time and particle
size, a calibration standard needs to be added prior to each measurement.
Based on the known speed of the disk and an external calibration standard, the
instrument “knows” which particle size fraction is meant to pass the detector
and which time. The scattering intensities for each size fractions can then be
recorded transformed into a particle size distribution. Due to this measuring
principle, analytical disk centrifugation is particularly well suited to characterize
polydisperse and polymodal particle size distributions, predominantly formed
during laser ablation experiments. Main disadvantages, however, are that
measurements can be highly time consuming, particularly for very small
particles or particles with low density (measurment time ~ (diameter)-2).

85
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

reinforcing ring
sedimentation
gradient liquid

detector

light beam

disc closure
time2

III
diameter
injection point

Figure 32: Sketch of the DC disc, sedimentation principle and chromatogram.

c. Amount of sample needed?

Only 0.1 mL of solution is required, at a concentration that allows detection by

the light beam, corresponding to 10-8-10-6 g of sample in the
best cases.

d. How do I prepare the sample for analysis?

The sample just needs to be injected in the center of the DC discoid cell,
after appropriate build up of the viscosity/density gradient.

e. Tips / warnings / hazards?

DC works under two main assumptions:

i) particles will have the same hydrodynamic size in the viscous medium as in
the original solution.

ii) absorbance of nanoparticles with different size is well reproduced by the

intrument software (the software uses a correlation function to consider
absorbance of different sizes, that is why optical parameters of measured
particle species need to be added).

Obviously, these assumptions should be checked before use with independent

techniques. Concerning assumption i), you must verify particles aggregation
or disaggregation for instance by mixing the nanoparticles solution with the
viscous solution and performing DLS, or by crossing DC analysis with TEM
imaging, at least the first times the analysis is performed.

Assumption ii) is more complicated to be verified (think for instance to

aggregates of plasmonic nanoparticles, whose absorption is strongly related
to size and shape), and actually requires a solid a priori knowledge of the
sample. But even for unknown absorption property, not the peak intensity but its
hydrodynamic diameter will be analyzed correctly.

86
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

f. To know more?

- Centrifugal Separations in Biotechnology; Leung, Wallace Woon-Fong; 2007,

Academic Press.

- Disc Centrifuge Particle Size Analyzers; CPS Instruments Europe; free at

http://www.cpsinstruments.eu/library.html

~~~~~~~~~~~~~~~~~~~~~~~ III
We had a s t udent who has been working on the laser for a couple of days. One morning
he was working within the laser lab with a colleague being present. Af t er several
ablations and breaks he put on his laser safet y goggles and s tart ed the laser for the
next ablation. However , he didn’ t make sure that everyone else in the lab was wearing
their laser safet y goggles. The colleague noticed the sound of the ablation and was
able to react right away. When this break in safet y procedure was addressed the s t udent
tried to make excuses s tating that the colleague was wearing his goggles some time
before. However , it is the responsibilit y of the laser operator to ensure that everyone
within the lab wears the personal equipment prior to t urning on the laser every single
time.

Elisabeth Maurer , Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
General comment on DLS and DC:
The software of both methods, DLS and DC often prints out the “z-Average”
or “intensity distribution”. Both are useful to see if you have strong scatterers
in the sample that may have screened the small ones, but are scientifically
meaningless and need to be converted into volume or surface or number
weighted statistics.

We have made round-robin tests and compared many instruments under

different conditions. Although instrument suppliers may claim different, you
cannot reproducibly measure particle diameters < 4 nm. We always measure
down to ≥ 3 nm diameter and take everything serious until 3-4 nm. For the
smaller fraction, you will need the help of TEM. Also, you will need to put the
density or an optical parameter into the software before measurement. A big
gold particle property will not be influenced largely by a monolayer of ligand
in density or extinction, so you take the value of gold. But for lighter particles
or bulky ligands this may have an effect. Choose the material parameter of the
core and disclose this in the experimental part of your thesis or manuscript.
Now you know why a hydrodynamic diameter measured by DC may be counter-
intuitively even smaller than a diameter measured by TEM, and larger by DLS
87
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

(even if it does not aggregate). A ligand has two effects. It may reduce the
particle´s density, to make it appear smaller. But most importantly, it adds a drag
force to the particle´s mobility. Thereby it is diffusing slower mimicking a bigger
particle in DLS, and sedimenting far slower, mimicking a smaller particle in DC.
That’s great to know, since it provides evidence for successful grafting a ligand
on a colloidal particle.

2.3. Small angle X-ray scattering (SAXS)

a. What do I see?
III Size and structure of particles in the 3 – 150 nm range can be measured.

b. How does it work?

SAXS is based on the elastic scattering of monochromatic X-rays by atoms in a

sample. Collection of scattered X-ray intensity takes place in a very small angle
interval (0.1° - 10°, Figure 33) from the beam propagation direction (forward
scattering), where the scattering events are more probable and diffractive
effects are nearly absent (see III.3.3). Although theory behind SAXS is not simple,
the main concept is that X-ray scattering depends on the discontinuity of
electron density in the sample, when it occurs on a length scale of 1 – 100 nm.
Consequently, the scattering pattern at small angles contains information on
the typical size and shape of the scattering objects with high electronic density
immersed in the matrix with low electronic density. It can measure polydisperse
samples, but intensities are dominated by high mass fraction of the sample
(the large particles). And size differentiation of overlapping modes is almost
impossible (e.g. differentiation of 6nm and 8nm particles).

The typical SAXS plot reports X-ray intensity versus angle, and it is fitted with
mathematical models which includes the size and shape of the scattering object
as fitting parameters. Contrary to DLS, these mathematical models must be
applied by the operator, making SAXS data analysis a non trivial task. Besides,

saxs
SAXS instrumentation is not available in most labs.

On the other hand, the advantage of SAXS compared to DLS is that it can
be used to characterize polydisperse samples and it may provide additional
information on the particle shape. In addition, SAXS is the only method suitable
for monitoring processes in optically opaque media e.g. inside the laser-induced
cavitation bubble.

88
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

ing
t t er
sca

x-ray beam

sample
in solution x-ray scattering curve
detector

III
Figure 33: Sketch of the working principle of SAXS.

c. Amount of sample needed?

The technique requires ~1 mL of sample with concentration of the order of

0.1 – 0.5 mg/mL (depending on the z of the elements in the scattering object).

d. How do I prepare the sample for analysis?

Stable liquid dispersions of nanoparticles are fine.

e. Tips/ warnings/hazards?

If the colloid is not stable, viscosity of solution can be increased to grant

stability for the duration of the analysis, which may require up to several days to
achieve acceptable signal-to-noise. To this purpose, you can use glycerol or high
concentration polymer solutions.

f. To know more?

- Structure Analysis by Small-Angle X-Ray and Neutron Scattering; Feigin,

L.A. and Svergun, D.I.; 1987; free at http://www.saxier.org/forum/viewtopic.
php?t=282

89
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

Table 1: Summary of properties, advantages and limitations of the principal

methods for the characterization of particle size distribution.

Electron Dynamic Light Small Angle Disk

Microscopy Scattering X-ray Centrifugation
(DLS) Scattering (DC)
(SAXS)

Type of particle
Feret (core) hydrodynamic hydrodynamic hydrodynamic
size

Acquisition of
III particle size
direct modelling modelling modelling

Sample
very slow fast very slow average
processing

Skills required
for data low low very high low
interpretation

Raw data output

number volume volume volume
(size distribution)

Statistics
(number of
~10-8 % up to 100% up to 100% up to 100%
analyzed
particles)

Information on generally no
colloidal state (unless using yes yes yes
specific tricks)

generally no
Information
(sometimes from
on ligand yes no yes
interparticle
conjugation
distance)

Characterization
yes no no no
of single particles

generally no
Characterization
(unless using yes yes yes
of aggregates
specific tricks)

Polymodal and
polydisperse yes limited yes yes
particles

Particle
yes no yes no
morphology

90
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

STEP 3: IDENTIFY THEIR MORPHOLOGIES

Imagine you’re in Italy, on a sunny day, but well arranged under a vine-covered
pergola in the porch of a restaurant from which you can admire green hills
planted with dozens of different types of fruit and vegetables. You’re tasting a
wonderful pasta with a red flavorful tomato sauce as you did not know that it
existed. The question you put to the waiter at the first opportunity will be “what
are the ingredients of this dish?” (or maybe it will be your second question, just
after asking him “Can I have some more?”). With our beloved nanoparticles it is
the same: the time comes when we have to exactly know about their ingredients.
Therefore, once you are sure that nanoparticles are present in our sample,
and you know their size distribution, the next step is to clearly identify their III
elemental composition and phase.

The quickest way to qualitatively identify elemental composition is the energy

dispersive X-ray spectroscopy (EDS). Most TEM and all SEM are equipped
with EDS, with the obvious difference that TEM allows elemental analysis with
nanometric resolution, while SEM only allows resolution of the order of 10 nm
or higher in ordinary conditions. Quantitative information on the elemental
composition by EDS is also possible, but it suffers of matrix effects and becomes
reliable only on isolated clusters of nanoparticles.

Alternative techniques such as inductively coupled plasma assisted (ICP)

mass spectrometry (MS) or optical emission spectroscopy (OES) gives reliable
quantitative information on particles ensemble, with sensitivity as low as part
per trillions (ppt) in best cases.

Elemental composition still does not tell us the phase of the nanoparticles,
as solid matter can be either amorphous or ordered, and every combination of
elements can be achieved in a multitude of phases. Just to give an example,
iron and oxygen can form magnetite, maghemite, hematite, wustite, goethite and
amorphous iron hydroxide, and all of them are achievable by laser synthesis.
Ordered phases are accessible by diffraction techniques, which for nanoparticles
are X-ray diffraction (XRD) or electron diffraction (ED). XRD is accurate and
easier to apply, and it can also give reliable quantitative information on the
weight fraction of all phases constituting the sample, but these are averaged
information over all particles in the specimen. ED is possible with TEM and can
give information on a single nanoparticle at time, but its quantitative application
is often not possible.

For some classes of materials such as oxides and semiconductors, qualitative

identification of ordered phases is possible by vibrational spectroscopy, such as
Raman and Fourier-transformed infra red (FTIR).

In the most complicated cases, where complex, disordered or unknown phases

are obtained, the above investigation methods may be not enough to precisely
identify the nature of the sample. In these “desperate” cases, you can refer to

91
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

a group of “magic” techniques which is applicable to any material: the group

of X-ray absorption spectroscopies (XAS), including X-ray absorption near
edge structure (XANES) and extended X-ray absorption fine structure (EXAFS).
Although the XAS group gives average information on the sample, these
information are complete and cover the composition, the chemical state of
each element and the composition and arrangement of nearest neighbour (NN)
and next nearest neighbour (NNN) atoms. This is very useful for multielement
materials, such as alloys or doped oxides. Unfortunately, the payback is high
because XAS analysis requires a synchrotron radiation facility and at least
one dedicated expert for data collection and analysis (which is absolutely not

III trivial!).

3.1. Energy dispersive spectroscopy (EDS)

a. What do I see?

EDS allows the detection of elements with z larger than Be, although sensitivity
increases with z and depends on matrix composition. Quantitative information
can be obtained easily when matrix effects are negligible.

b. How does it work?

EDS is based on the fact that i) X-rays are produced by the interaction of an
high energy electron beam with matter, and ii) each element has a unique
atomic structure allowing unique set of peaks on its X-ray emission spectrum.
In particular, the collision of the electron beam with sample atoms promotes
the kick off of electrons from inner atomic shells, leaving an empty energy level
which is filled by electrons of the outer shells. The energy difference is released
as a photon with frequency typical of X-rays. Since the electronic structure is
specific of each element, the spectroscopy of X-rays produced by the sample
allows the identification of elements contained therein. X-rays are measured by
an energy dispersive spectrometer exposed on the sample, and compared to
a database for element identification. The typical EDS spectrum reports X-ray
counts (in arbitrary units) versus their energy (in keV, Figure 34). Integration of
peaks and comparison with a standard or a database allows the extraction of
quantitative information.

92
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

c (K-line)

EDS spectrum of A F u e
alloy nps coated with
COUNTS (A.U.)

polyethyeneglycol

O (K-line)
u III
A (M-line)
e
F (K-line) A (L-line) u
x-ray energy (keV)
Figure 34: Typical EDS spectrum of AuFe alloy nanoparticles coated with polyethylenglycol.

EDS is based on the fact that i) X-rays are produced by the interaction of an
high energy electron beam with matter, and ii) each element has a unique
atomic structure allowing unique set of peaks on its X-ray emission spectrum.
In particular, the collision of the electron beam with sample atoms promotes
the kick off of electrons from inner atomic shells, leaving an empty energy level
which is filled by electrons of the outer shells. The energy difference is released
as a photon with frequency typical of X-rays. Since the electronic structure is
specific of each element, the spectroscopy of X-rays produced by the sample
allows the identification of elements contained therein. X-rays are measured by
an energy dispersive spectrometer exposed on the sample, and compared to
a database for element identification. The typical EDS spectrum reports X-ray
counts (in arbitrary units) versus their energy (in keV, Figure 34). Integration of
peaks and comparison with a standard or a database allows the extraction of
quantitative information.

c. Amount of sample needed?

See in III.1.2.c (TEM) and III.1.3.c (SEM).

d. How do I prepare the sample for analysis?

See in III.1.2.d (TEM) and III.1.3.d (SEM).

93
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

e. Tips / warnings / hazards?

Avoid thick or crowded agglomerates of nanoparticles to minimize matrix effects

and to increase the reliability of quantitative results, which is best in isolated
nanoparticles. Don´t worry about carbon, it´s always there, and sometimes
you will find silicon (or silicates), which is an additive in seals and are found
everywhere in the lab and sample holders, in particular where organic solvents
were used for cleaning by you or your colleagues. Of course, you will not use
any equipment made of metal (spatula etc.) during synthesis or sample handling.

f. To know more?

III See III.1.2.f (TEM) and III.1.3.f (SEM)

3.2. Inductively coupled plasma assisted (ICP) mass

spectrometry (MS) and optical emission spectroscopy
(OES).
a. What do I see?

All the elements in the sample can be detected and precisely quantified,
excluded hydrogen. It will also tell you the amount of dissolved elements (e.g.
analyzing the supernatant after centrifugation)

b. How does it work?

Measurement takes place in two stages: atomization of the sample and analysis
of atomic species. The inductively coupled plasma serves for the atomization
of the sample. The ICP is alimented by an electromagnetic coil which has the
function of inductively heating the plasma, which is composed of a highly
ionized and electrically conductive argon gas. The liquid sample is introduced
in the plasma with a nebulizer, immediately reaching the same temperature and
ionization conditions.

In ICP-MS, detection takes place by conveying the atomized sample to a mass

spectrometer, where charged particles are separated by a mass selector which
exploits the Lorentz force, and counted by a charged particles counter, to obtain
signal intensity versus mass for unit charge.

In ICP-OES, the plasma is coupled to a diffraction grating and a photodetector

(typically a photomultiplier or a CCD), to measure the intensity of light emitted
at the wavelength characteristic of highly ionised atomic species. Atomic species
emit a set of very sharp bands, which clearly are related to the electronic
structure of the emitting element.

Quantitative analysis requires calibration with a standard in both MS and OES

methods.

94
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

c. Amount of sample needed?

Detection limits of ICP-MS are impressively low (usually below parts per billions,
ppb), thus requiring few μL of solution. ICP-OES has higher detection limits, but
fully compatible with a typical sample containing laser generated nanoparticles.

d. How do I prepare the sample for analysis?

For quantitative analysis, sample preparation requires the complete dissolution

of the solid phase into ions (sample digestion), which typically requires the use
of acids or bases at high temperature. This means that a colloidal solution is
not suitable for analysis without appropriate sample preparation, if reliable
quantitative results are desired.
III

~~~~~~~~~~~~~~~~~~~~~~~
Several days af t er my firs t experience with high int ensit y lasers, I began to feel more
confident in handling these machines. However , once af t er a short break I went to the
laser lab and continued my experiment s, when I observed increasing smoke formation I
realized that i forgot to remove the cap of the focus lens.

Sandra Jendrzej, Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

e. Tips / warnings / hazards?

Sample digestion is critical for quantitative measurements: check literature

for the most appropriate procedure for your sample (for instance aqua regia,
which is a 1:3 mixture of HNO3 and HCl, works very well in dissolving gold
nanoparticles, but cannot be used with Ag nanoparticles because it can form the
insoluble AgCl precipitate). For contamination with C, Si, metals, see (III.3.1.e).

f. To know more?

- AAS, XRF, and MS Methods in Chemical Biology of Metal Complexes; Ott, Ingo;
Biot, Christophe; Hartinger, Christian; in Inorganic Chemical Biology: Principles,
Techniques and Applications; Gasser, Gilles; 2014, John Wiley & Sons.

95
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

3.3. X-ray diffraction (XRD)

a. What do I see?

The ordered phases can be identified, and the volume-weighted average size
of ordered domains in the sample can be evaluated. By applying the Rietveld
analysis, quantitative information on the mass fraction of each crystalline phase
can be achieved. Hence, crystal diameter can never be smaller than the particle
diameter. But particles (e.g. larger metal particles) may consist of several
crystals (e.g. twins) even if they are spherical, so that combination of XRD
with TEM of DC/DLS may already tell a lot about the sample. XRD is “blind”

III for amorphous materials, whereas SAXS sees both, crystalline and amorphous
scattering objects.

b. How does it work?

X-rays are scattered by atoms. When atoms form ordered arrays with well
defined translation periodicity on a length scale comparable to X-ray wavelength
(~0.1 nm), the X-rays are scattered only at well defined angles (Ɵ) typical of the
crystalline structure. In this way, a diffraction pattern is obtained by the measure
of the intensity of scattered X-rays versus the scattering angle. The XRD pattern
(intensity and diffraction angle of all peaks) is related to the atomic position and
interplanar distances in the crystal, as described by the Bragg’s law (Figure 35):

where n is any integer, λ is X-ray wavelength, l is the distance between

reflection sites (typically interplanar distances d), and Ɵ is the angle between
the reflected beam and the plane containing the sample.

The comparison of the XRD pattern with a database allows the identification of
the compound. The volume-weighted average size (τ) of ordered domains in the
sample can be measured by using the Debye-Scherrer formula

where K is a constant (typically ~ 0.9), λ is the X-ray wavelength and β the full
width at half maximum (FWHM) of the diffraction peak located at the diffraction
angle Ɵ. Hence, XRD tells you the type of crystal and its size. Sharp, high peaks
indicate large particles.

In case of multiple compounds, the Rietveld analysis is used to extract

information on the relative mass abundance of all species involved. By this,
lattice strains from doping, defects or alloying can be attributed. It’s good to
have Rietveld fits for a phospors or alloy nanoparticle series.

96
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

incident
beam
reflected
0 0 beam
i

i
d
dsin 0
i

B III
XRD spectrum of A u Fe
intensity (A.U.)

alloy nps

Rietveld fit

40 60 80 100 120 140

20
i

Figure 35: (A) The principle of Bragg’s law and (B) a typical XRD spectrum of AuFe alloy
nanoparticles.

c. Amount of sample needed?

At least 2 – 10 mg of nanoparticles are required. Yes, you might need upscaling

(Chapter II)

d. How do I prepare the sample for analysis?

Nanoparticles are deposed on XRD substrates (for instance quartz plates).

e. Tips / warnings / hazards?

It is important to avoid heating of sample during deposition, in all those cases

where oxidation, segregation or phase transition may occur, such as with
iron oxides or metastable alloy phases. Because of the volume-sensitivity
of the method, minute amount of larger particles will screen the information
(crystal structure and size) of the small fraction. First calculate crystal size by
Scherrer equation, if it´s far larger than expected (e.g. sharp, high peaks are not
nanocrystals) maybe filter or centrifuge the sample and repeat.

97
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

f. To know more?

- Basic Concepts of X-Ray Diffraction; Zolotoyabko, Emil; 2014, John Wiley &
Sons.

~~~~~~~~~~~~~~~~~~~~~~~
Every body knows that one should wear laser prot ection glasses when the laser is
running . So what to do, when you notice that your colleague does not wear glasses? For

III sure , do not slap him into the face while trying to prot ect his eyes by hand. It ’ s bet t er
to t ell him: Please wear your prot ection glasses.

Lisa Gamrad, Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

3.4. Electron diffraction (ED)

a. What do I see?

Ordered phases can be detected and identified by ED, with the possibility to
perform the analysis at the single nanoparticle level.

b. How does it work?

The principle is the same of XRD (III.3.3.b), with the sole difference that here the
diffraction pattern is produced by scattering of a monochromatic high energy
electron beam instead of an X-ray beam.

c. Amount of sample needed?

See III.1.2.c (TEM).

d. How do I prepare the sample for analysis?

See III.1.2.d (TEM).

e. Tips / warnings / hazards?

Measurement of scattering intensity is not simple with ED, because of its

dependence on nanoparticle orientation and on the limited number of
nanoparticles probed. Therefore, ED is more suited for qualitative identification
of nanoparticle structure, while quantitative information on the ordered phases
are preferably obtained by XRD.

98
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

f. To know more?

See III.1.2.f (TEM).

3.5. Vibrational spectroscopy (Raman, FTIR).

a. What do I see?

Several compounds (mostly oxides and semiconductors) can be identified based

on the presence of peaks ascribable to well defined chemical bonds.

b. How does it work?

In both cases, the energy of vibrational levels is probed. In Raman spectroscopy,

III
the sample is irradiated with a monochromatic visible laser beam, and the
portion of backscattered light is collected. The backscattered light has a main
component due to elastic scattering, which is discarded, and a low intensity
inelastic scattering portion originated by Raman scattering. The Raman
scattered light is dispersed on a detector to measure intensity versus photon
energy. The energy difference between inelastic scattered photons and incident
photons (the Raman shift) is the energy of vibrational levels in the sample.
Raman spectra are reported as intensity (counts) versus wavenumber or Raman
shift (expressed in cm-1, Figure 36). The position of each peak is distinctive of
specific functional groups in a compound. However, a certain level of crystalline
order is required to observe appreciable Raman bands from crystalline solids,
which typically happens only for sizes above ~5 nm.

In FTIR, vibrational levels are probed directly by measuring the transmittance

of the sample as a function of infrared radiation energy. The term Fourier
transform infrared spectroscopy originates from the fact that a Fourier
transform of the transmitted beam is used to obtain the final spectrum (don’t
worry the instrument software will do it for you). FTIR spectra are reported in
transmittance (T) versus wavenumber (in cm.-1, Figure 36).

99
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

raman spectrum of hematite

e
(F 2O3) nanoparticles wavenumber (cm-1)
2000 1500 1000 500
100
COUNTS (A.U.)

t (%)
80

III
raman shift (cm-1) 60 FTIR spectrum of iron oxide
nanoparticles

Figure 36: Typical Raman and FTIR spectrum.

c. Amount of sample needed?

Raman spectroscopy only requires few μg of material, when coupled to a

microscope (micro-Raman). FTIR requires more than 0.5 – 1 mg of material.

d. How do I prepare the sample for analysis?

Solvent must be removed for the analysis, and nanoparticles must be deposited
on a substrate with negligible background. For Raman spectroscopy, this can be
a metal foil or a glass slide. For FTIR, specific crystals which are transparent to
infrared radiation must be used (such as CaF2 or KBr windows).

e. Tips / warnings / hazards?

Often Raman and FTIR gives complementary information about inorganic

materials, hence the two techniques give their best when used together.

f. To know more?

- Raman Spectroscopy for Nanomaterials Characterization; Kumar, Challa S. S. R.;

2012, Springer.

- Infrared and Raman Spectroscopy Principles and Spectral Interpretation;

Larkin, Peter; 2011, Elsevier

100
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

3.6. X-ray absorption spectroscopy (XAS).

a. What do I see?

Composition, chemical state, type of NN and NNN and their location in space.

b. How does it work?

XAS principle is pretty much like OAS, being based on the measurement of X-ray
absorption bands from the sample. However, the physics behind absorption
of X-rays is rich and strongly related to the chemical nature of the sample.
Contrary to OAS, that typically shows absorption peaks, in XAS you find
absorption edges. These edges can be divided in three portions: the pre-edge,
the rising edge and the extended edge structure (Figure 37). The pre-edge
III
contains the information on the element, being determined by the electronic
structure of the atom absorbing X-rays. Absorption intensity is related to the
abundance of the absorbing element by the same Lambert-Beer law exploited
for OAS (see III.1.1.b).

X-ray absorption is followed by emission of a photoelectron (photoelectric

NNN
effect). The great point of XAS is that photoelectrons undergo a scattering
process with surrounding atoms, and this generates a modulation of the X-ray
absorption intensity in the region of the rising edge and the extended edge. In NNN
particular, the rising edge reflects the interaction of the photoelectron with the
NNN
NN, and it’s the subject of XANES. The extended edge reflects the interactions
with NNN and other atoms, and it’s the subject of EXAFS. The XANES and EXAFS
portions of the X-ray absorption edge can be fitted with complex mathematical
N
models which account for the relative position and composition of NN, NNN and NN
so on. In this way, you can find the average composition and atomic disposition
around the absorbing element. In particular, XANES is more sensitive to the
oxidation state and to the atomic site symmetry, while EXAFS can give the full
atomic coordination number, the chemical nature of the surrounding atoms and
the interatomic distances.

nexafs exafs
(=xanes)
absorption

0 100 200

e - Ek (eV)

Figure 38: Sketch of a XAS edge and its three parts.

101
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

c. Amount of sample needed?

XAS is very sensitive, thanks to the high performance of synchrotron light: the
amount of material needed is the same necessary to cover a small area of few
mm2 on the sample holder.

d. How do I prepare the sample for analysis?

Simply by drop casting on the XAS substrate, followed by solvent evaporation. It

can also run wet samples

e. Tips / warnings / hazards?

III Anyone working with nanoparticles sooner or later needs a friend expert of XAS.
The tip is...be nice to any XAS expert!

f. To know more?

- X-Ray Absorption and X-Ray Emission Spectroscopy: Theory and Applications;

van Bokhoven, Jeroen A. and Lamberti, Carlo; 2016; Wiley.

STEP 4: QUANTIFY THEM

In Chapter II you already obtained good information about how to evaluate the
productivity of your set up. In addition, it is good to know that the best way for
the quantification of matter in a sample is ICP-MS or ICP–OES, as discussed in
section III.3. Sometimes, you may be tempted to use a microbalance to quantify
the mass of a sample, but this procedure can be error prone for amounts lower
than ~5 mg, and you must always account for solvent residuals and solutes
when drying the colloid. On the other hand, the weight of the target can be
altered by the formation of micrometric fragments which are not part of the
colloid.

In case of absorbing nanoparticles, OAS can be used for quantification thanks

to the Lambert-Beer law, which simply relates absorbance to nanoparticles
concentration (see III.1.1.b). This method works well with metals and oxides
whose size and optical properties are already known, and especially benefits of
start-up calibration with an independent and more reliable technique such as,
for instance, ICP-MS.

It is worth to stress that the most common unit for the quantification of a
sample of nanoparticles is mass (e.g., mg) or mass per volume (e.g., mg/
mL). However, the comparison of laser ablation yield among nanoparticles
with different composition is possible only by considering the number of
atoms in your sample, expressed as mol or mol/L (or M). The reason can be
easily understood considering that a sample with concentration 0.001 mol/L
correspond to 0.197 mg/mL if composed of Au, and to 0.063 mg/mL
if composed of Cu (Figure 38).

102
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

mass of 1 mol of A u ~ mass of 3 mol of C u

C u
A u C u u C
III

Figure 38: Comparison between the mass of 1 mol of gold and 1 mol of copper.

STEP 5: ANALYZE THEIR SURFACES

How many times you said “You can’t judge a book by its cover”? Sorry, but
this is not the case of nanoparticles: since the proportion of surface atoms to
volume atoms increases while decreasing the size of condensed matter, the
surface has a special importance in nanomaterials. For instance, approximately
1 atom every 10 is a surface atom in a 10 nm Au nanoparticle.

Surface atoms have different reactivity than bulk ones, due to incomplete
saturation of their atomic bonding capability. In case of laser generated
nanoparticles, which are often obtained in the absence of any stabilizer or
coating agent with an highly reactive surface, this is even more true. For this
reason, surface stoichiometry and oxidation state are typically different than
in the nanoparticle core. The consequences for electronic, optical, catalytic or
biomedical properties are huge, and these justify the need to obtain a thorough
knowledge of nanoparticle surface, with the same dignity of the “body” of
nanoparticles.

In addition, the surface of nanoparticles has a crucial importance when it is

coated with specific functional molecules, such as in noble metal nanoparticles
conjugated with biopolymers, dyes, DNA or antibodies.

The “queen” technique for surface studies is X-ray photoelectron spectroscopy

(XPS), which gives complete information on the chemical nature of the sample.
However, in case of nanoparticles coated with organic molecules, other
techniques can be more informative such as FTIR, nuclear magnetic resonance
(NMR) or thermogravimetric analysis (TGA).

103
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

5.1. X-ray photoelectron spectroscopy (XPS).

a. What do I see?

It tells you about the composition of the nanoparticle surface and its oxidation.
Surface elements with Z ≥ 3 (Li) are detected, and their oxidation state identified.
Although in principle quantitative information on the surface abundance of each
element is achievable, in case of nanoparticles these information are averaged
between the surface and the inner layers, due to the complex matrix effects
and geometry of the sample. Hence it is good for seeing oxidation trends and
roughly knowing what composition you have created.

III b. How does it work?

Monochromatic X-rays are used to extract electrons from the sample

(photoelectrons). The energy difference between the kinetic energy of electrons
and the energy of X-rays is related to the binding energy of the electrons in the
atoms of provenience, being therefore characteristic of each element. Besides,
the energy resolution of XPS is high enough to be sensitive to minimal energy
modifications which take place when an atom establishes chemical bonds with
other species, thus giving information on the oxidation state and chemical
environment of each element.

Since photoelectrons of the nanoparticle surfacevery low mean free path in

condensed matter, only the first 1 – 10 nm are probed (depending on the
electronic density of the sample and the used angle between sample and
detector).

O 1s
s 2 p C 2s Fe 2 p
A g 3d
COUNTS (A.U.)

intensity (A.U.)

c-c s p 2

c-c s p 3

c-oh

binding energy (eV) binding energy (eV)

Figure 39: Typical XPS spectrum (left) and magnification of one of the peaks (C peak for
example, right), showing that it can be deconvoluted in different components corresponding to
different oxidation states and/or chemical environment.

104
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

A typical XPS spectrum reports the number of electrons detected (as counts)
versus their binding energy (in eV, Figure 39), and each element produces
a characteristic set of XPS peaks at characteristic binding energy values.

c. Amount of sample needed?

XPS is only sensitive to the first nm of a sample, therefore a monolayer of

nanoparticles spread over an area of ~1 cm2 is enough.

d. How do I prepare the sample for analysis?

Sample must be dried before use (XPS is performed in ultrahigh vacuum), and
can be prepared by drop casting on a clean and flat substrate which has no
interfering peaks with those of our sample. III
e. Tips / warnings / hazards?

Due to the high surface sensitivity of XPS, contaminants like salts or synthesis
byproducts can shield the signals from the nanoparticles. In principle, the
topmost layer can be removed by ion sputtering in the XPS chamber, but
sometimes this can alter the stoichiometry of the surface. Sputtering also
significantly changes the oxidation state of the remaining species making it
hard to trace the oxidation state before sputtering.

In case the material of interest being shielded e.g. by a carbon layer –

the oxidation state can be accessed if the sputtering time is adjusted so that the
covering carbon layer is not fully removed. However some uncertainty about the
oxidation state remains when sputtering is used. Note that the holder itself has an
effect on the peak shifts, e.g. carbon needs different calibration than silicon.

Peak deconvolution (e.g., to extract the degree of oxidation) is very tricky, please
ask a trained person to help you at least when you do it the first time, you
will see that it makes a big difference how you deconvolute (deconvolution is
“dissectioning” a shouldered peak into two or more Gaussian peaks, Figure 42).

f. To know more?

- Auger- and X-Ray Photoelectron Spectroscopy in Materials Science: A User-

Oriented Guide; Hofmann, Siegfried; 2013, Springer.

105
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

5.3. Nuclear magnetic resonance (NMR).

a. What do I see?

The typical NMR spectroscopy is performed on protons (1H nuclei), and it is

therefore sensitive to their presence and chemical environment. The NMR
pattern tells us about which organic molecule is present in a sample, and its
relative abundance.

b. How does it work?

Some nuclei (such as 1H, 13C or 14N) have a non-zero nuclear spin. In the

III presence of an external magnetic field, the nuclear spin acts much like a small
magnet, and can assume a collinear (lowest energy in 1H) or anticollinear
(highest energy in 1H) orientation with respect to the direction of the external
field. Electromagnetic radiation can promote the transition from one state to the
other. Therefore, in NMR spectroscopy, an external magnetic field is exploited
to produce a splitting of spin energy levels in 1H nuclei, and the absorption
of electromagnetic radiation by the sample is measured as a function of its
wavelength (pretty much like in absorbance spectroscopy). Since the energy gap
in 1H nuclei is very small, radiofrequency radiation (102 MHz) is used instead of
visible light. The energy gap in 1H nuclei is sensitive to the chemical environment,
because electron density exerts a diamagnetic shielding effect on the external
magnetic field, which depends on the distance and number of electrons in
proximity of the hydrogen nuclei. Besides, the spin of 1H nuclei interacts with
each other (when sufficiently close), generating a splitting of the energy levels.
For these reasons, NMR spectroscopy gives information on the structure of the
molecule, and thus the NMR spectrum is a fingerprint characteristic of each
molecule.

c. Amount of sample needed?

NMR typically requires 0.1 – 1 mg of organic matter. However, signal intensity is

sensibly damped when short molecules are bound to nanoparticles, and in these
cases up to 10 mg of organic matter can be needed.

d. How do I prepare the sample for analysis?

Samples for NMR are in liquid phase. The only requisite is matching the
concentration required for appreciable results.

e. Tips/ warnings/hazards?

Since the signal of molecules bound to nanoparticles are greatly damped,

sometimes it is simpler and quicker to remove the ligands from nanoparticles
surface and analysing only the organic component. With noble metal
cores, this is performed by addition of strongly interacting thiols (such as
dithiothreitol), which replace the original ligands on a timescale of 1-2 days.

106
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

Then nanoparticles can be removed by centrifugation or the surnatant can be

collected by dialysis for NMR analysis.

f. To know more?

- The basics of NMR; Hornak, Joseph P.; 2014, free at https://www.cis.rit.edu/

htbooks/nmr/

5.4. Thermogravimetric analysis (TGA).

a. What do I see?

In case of organic ligands which undergo complete oxidation into volatile III
compounds (CO2, H2O, NOx etc.), the mass fraction of the organic component
can be easily and precisely quantified. This is possible only when the
nanoparticle (inorganic) core is chemically inert at the combustion temperature
of the organic component. TGA can also be run with inert gas. So you may find
out the amount of ligand that is covering the nanoparticle.

b. How does it work?

TGA basically consists of a microbalance and of a chamber with controlled

temperature and atmosphere. While temperature is increased, the
microbalance register any weight increase or decrease. In case of organic
compounds, combustion in presence of oxygen is detected by weight decrease.
The temperature ramp can be continuous or step by step, and weight can be
reported as a function of temperature or time.

c. Amount of sample needed?

In micro-TGA, few milligrams of oxygen-reactive matter are enough.

d. How do I prepare the sample for analysis?

Solvent must be removed and the dried sample deposed on the TGA holder.

e. Tips / warnings / hazards?

In the first run, take a slower heating program. To exclude chemical modification
of the inorganic core, run TGA also on uncoated nanoparticles as a reference
measurement to exclude background contributions, and on a known amount of
ligands without nanoparticles. Plot the first derivative of the signal. Do you see a
shift of the ligand’s temperature? This indicates ligand binding (immobilization),
in particular for polymers grafted on nanoparticles. TGA is very powerful if
coupled with calorimetry (so-called, TGA-DSC), then you also know if mass loss
is endotherm (e.g. degassing or drying) or exotherm (oxidation).

107
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

f. To know more?

Thermogravimetric Analysis (TGA) A Beginner’s Guide; PerkinElmer, Inc.; free at

http://www.perkinelmer.com/CMSResources/Images/44-74556GDE_
TGABeginnersGuide.pdf

STEP 6: ANALYZE THEIR PROPERTIES

The reason for generating nanoparticles is that they have some properties of
interest for scientific or technological applications. Obviously this opens up a
range of different possibilities for completing the investigation of our product,
III and each nanomaterial requires a specific technique or a group of techniques.
However, no matter what is the property of interest, it will be always related
to the structure of the nanomaterial, and therefore all the previous steps are
starting pillars for the functional study or the direct application of our laser-
generated nanoparticles.

Conclusions
In summary, the range of techniques available for the characterization of our
laser-generated colloids is wide, but they should be used “cum grano salis”
to save resources and obtain reliable results. As a last step to guide the
characterization of our samples, it is useful to conclude the chapter with a
sketch where all the techniques previously described are classified on the
basis of:

i) the amount of matter required for the investigation, starting from the single
nanoparticle techniques up to the most demanding ;

ii) the smallest size detectable.

The power of colloid characterization comes from combining the methods.

2 methods may already be enough to extract nice insights. Examples :

• XRD gives smaller sizes than other methods : you have multi-crystalline
particles

• DC gives smaller particles than TEM or XRD : you have successfully

conjugated the particle

• DLS gives volume-weigthed diameter of 50 nm, but TEM tells 6 nm.

It´s aggregated, you may want to add surface charge by anoins or pH,
or add a surfactant or polymer for stabilisation

• You have ablated elemental Pt but find 40% of its surface is oxidized in
XPS. Nice to know for catalysis. Maybe try Zeta potential measurement
(see Cahpter IV) to see this surface charge in liquid as well?

108
 UNDERSTAND IT: OPTICAL AND SOLID MATTER CHARACTERISATION

1 np ”0.1-0.5 mL 
“0.1 mM•
” 1 mL 
“ 1-10 mM
” 50 mL 
“ 1-10 mM
”100 mL 
“10 mM
• TEM (HRTEM,
ED, EDS) • OAS • XPS • XRD • NMR
• SEM (EDS) • FS • DC • FTIR
• DLS • XAS • TGA
• RAMAN
• ICP-MS, ICP-OES
• SAXS

> 0 nm > 1 nm > 5 nm > 10 nm

• OAS • TEM (HRTEM, ED) • SAXS • SEM

III
• DLS
• FS • XRD
• DC
• XPS
• Raman
• XAS

• NMR

• FTIR

• EDS (TEM, SEM)

• ICP-MS

• ICP-OES

• TGA

Sketch 1: Described techniques classified by amount needed and smallest detecteble size.

• You ablated an alloy but the XPS or EDX composition gives a different
molar ratio compared to the educt. Try ICP to see what has been dissolved
in the liquid

• XRD crystal structure does not fit to elemental composition. Try SAXS or
HR-TEM to learn about amorphous fraction.

• You know the above table but have no time and your supervisor needs
results by tomorrow: start with UV-Vis and DLS/DC. Collect arguments
(e.g., present different histogram statistics) to ask for more. Tell that always
2 independent analytical methods are needed to validate results ready for
publication with her/his highly reputed name on it.

• You tried everything but are lost. Send an email to the authors of this book.

109
 CHAPTER IV

KEEP IT: STABILITY

After your first days in the laser lab you will probably have produced your first
samples of laser-generated colloids and you will be able to proudly present
them to your boss with a full set of structural characterization data. At this
stage, two questions will inevitably come to mind, if not to yours then definitely
to your boss:

How stable are the nanoparticles? And how can stability be influenced during
or after laser ablation?

These will be basic questions sticking with you up to the end of your life in
the world of laser-generated colloids and nanoscience. Guided by these

IV fundamental questions, this chapter will give you a basic understanding of how
to define “stability” in the field of colloidal science, how to measure its value,
how to interpret the results and how to alter the corresponding parameter using
simple modifications of the laser ablation process.

How stable are the nanoparticles ?

You will certainly have to experience the hard way that particle stability is a
critical issue during synthesis and handling of laser-generated colloids. After
you have spent an entire day in the laser lab generating beautiful colloidal
samples, the next day you may be in for quite an unpleasant surprise. You
could find them either containing large soft flocks or situated as an ugly, brown
or black precipitate on the bottom of your vessel or almost irreversible fused
with your container´s wall. Another phenomenon you may encounter (and we
often did) could be that some of your samples are perfectly stable while others
precipitated, even though you are sure you treated them exactly the same.
At this point you will certainly, depending on your level of stress and your
personality, either think or cry out: “Stability Sucks”. In order to avoid or at least
minimize such pitfalls this paragraph will give you a step-by-step overview on
how to define stability, how to control it during laser ablation in liquids and how
to measure it.

STEP 1: DEFINE STABILITY

The definition of this term in colloidal science is not always universal, however
a straightforward answer could be that stability actually defines to what extent
your sample changes with time. When specifying the word stability you have
to differentiate between chemical stability and colloidal stability. Chemical
stability defines whether your colloid is subject to chemically-induced changes

110
 KEEP IT: STABILITY

caused by the environment, in most cases by the solvent. The most relevant
processes are oxidation and dissolution of your solid material in the respective
solvent. While chemical stability is a very specialized phenomenon strongly
depending on the used material, colloidal stability is universally applicable to
all colloidal systems. It is, furthermore, particularly critical in laser-generated
colloids which are generally completely ligand free and hence only “meta-stable”
without any external stabilizing agents. Consequently, this paragraph will focus
on the aspect of colloidal stability. Colloidal stability indicates to what extent
your nanoparticles are resistant to return from the dispersed (finely distributed
within the solvent) state to the bulk state. From a thermodynamic point of view,
all colloids are unstable since a reduction of their surface area is energetically
favored. However, this process can be kinetically inhibited, which could lead to
colloids stable in solution for centuries (as the gold colloid made by Faraday
150 years ago). So in order to understand the concept of colloidal stability you
need to basically understand the nature of attractive and repulsive forces in a IV
colloidal system. This knowledge can be provided by the DLVO theory named
after its inventors Boris Derjaguin, Lev Landau, Evert Verwey and Theodoor
Overbeek. Based on this theory, the interactions between colloidal particles are
driven by 2 main forces 1) the van der Waals attraction and 2) the electrostatic
O
repulsion. Van der Waals forces are caused by the formation of induced HO
dipoles between uncharged but polarizable materials and are much stronger oh
on a short distance. Electrostatic forces on the other hand are caused by the O
so-called electrical double layer around a charged particle, and are usually oh
O
more pronounced at longer distances. Based on this, a potential curve can be
constructed (Figure 40 A), which gives you the sum of all the potential energy
in correlation with the distance between two particles in the colloids. Stability
means keeping a large distance. Looking at this graph you can observe that
there are two minima of potential energy.

~~~~~~~~~~~~~~~~~~~~~~~
Our laser lab has sheet s covering the window. On a Monday I ent ered the room in the
morning and ins tantly got the feeling that something was different. Af t er some time
I realized the sunlight coming in through holes in the sheet s covering the windows.
I also saw some s traight lines in some covering boards leading towards the holes.
Act ually the las t s t udent that had worked with the laser on a Friday was dithery
adjus ting the laser beam while the laser was running on high power set tings thereby
unint endedly optimizing the sunlight incidence into the laser lab…

Sven Reichenberger , Duis burg

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

111
 KEEP IT: STABILITY

The first, the so-called primary minimum, is located at lower distance and
indicates a state where two nanoparticles are irreversibly fused, this process is
called aggregation and the resulting materials are hence named aggregates.
In addition, there is a second minimum found at greater distance between
the particles. Here the particles are more loosely bound, predominantly in
a reversible fashion. This process is called agglomeration and the resulting
materials are termed agglomerates. For the stability of the laser-generated
colloids you can conclude that there are two main processes, which could
interfere with the colloidal stability of the particles, being reversible “loose”
agglomeration and irreversible “tight“ aggregation. While agglomerates
may generally be re-dispersed easily e.g. by intensive shaking or short
ultrasonication, aggregates are generally permanent and you may need to adapt
your synthesis parameters in order to avoid it. In the potential curve diagram

IV (Figure 40 A), a stable colloid needs particles kept at a distance of Min2 or

better even farer away from each other.

To enhance the general understanding of colloidal stability you now need to

take a closer look at the main stabilizing force (that is, the repulsive force) in
ligand-free colloids, which are surface charge and the consequent formation
of the electrical double layer. The first fundamental question relevant in this
context should be why laser-generated colloids, e.g. composed of metal,
possess a surface charge at all. A pure metal surface in a solvent should not be
charged and hence the particles should be subject to immediate aggregation.
E.g. elemental gold is hydrophopic, insoluble in water and loves to binds to
other particles by van der Waals force, and once the solids are in contact,
inter-particle atom diffusion diminishes particle boundaries causing irreversible
fusion. However, the laser process generally yields a partial surface oxidation of
the colloidal metal nanoparticles, and partial reduction of oxide nanoparticles.
Hence, the surface of laser-generated nanoparticles carries defects (which are
very valuable for both application and colloidal stability), and these defects
attract surface adsorbates. These surface adsorbates are of course counter-
charged. Coming back to an elemental metal nanoparticles with electronic
defects at its surface (e.g. a few Au+ and Au3+ sites at the surface of a sphere
dominated by Au0 atoms): These oxidized surface atoms can then attract
hydroxyl from water and form a pH-dependent equilibrium between M-OH
and M-O- (M=metal) groups, which generally yield a partially negative surface
charge in laser-generated colloids. Hence, the particle became an acid, ready
to exchange protons with water. Yes, it even has a defined value where the
charge flips towards positive charge (M-OH2+) if more protons are taken up at
low pH. E.g., for laser-made gold nanoparticles, the flipping point, the so-called
isoelectric point is reached at a pH of around 2, which shows that ligand-free
gold nanoparticles are very acidic.

In case these charged nanoparticles are dispersed in an electrolyte like water

they attract counterions and form an electrical double layer. As the term

112
 KEEP IT: STABILITY

double layer suggests it consists of two components (Figure 40 B): i) An inner

layer termed Stern or Helmholtz layer where ions are densely bound to the
nanoparticle surface (e.g., M-OH-) and ii) a diffuse layer of loosely bound,
more or less mobile counter ions (the beneath negative Stern layer attracts
cations, which in turn bring anions, and so forth). The most critical criterion for
colloidal stability is, whether the electrical double layer extends far enough to
keep the nanoparticles at an appropriate distance and prevent aggregation or
agglomeration. This can be described by the Debye screening length (K-1), which
indicates the distance at which the electrostatic potential has dropped by 1/e.
This means that the Epot at the Debye´s distance has dropped by about 67% of
its original value.

For oxides, the same rules apply, just the charge sign is turned, with oxygen
defects resulting in positive Stern layer, attracting negative counterions, etc.
E.g. ZnO made by laser ablation in water is a base with oxygen defects (oxygen IV
vacancies) and (slightly) positive surface charge at neutral pH.

-1
stern
(epot )

A B
repulsion
potential energy (epot )

min 2
distance
min 1
agglomerate
aggregate
distance
attraction

Figure 40: Stay away: DLVO theory and electrical double layer. (A) Potential energy-distance
curve in a colloidal system indicating the primary minimum (Min1), secondary minimum (Min2)
and activation energy (ΔE). (B) Illustration of the electrical double layer around a spherical
particle. Indicated is the potential energy curve, the Stern layer and the diffused layer. κ-1
illustrates the extension of the Debye screening length.

Now you know how to define the term “colloidal stability” and you have learnt
that electrostatic repulsion and van der Waals attraction are the main forces
at work. Based on this you have learnt the physical origin of surface charge as
well as the electrical double layer in a completely ligand-free laser-generated
colloidal system. As a follow up it would be interesting to know how you can
affect colloidal stability.

113
 KEEP IT: STABILITY

STEP 2: INCREASE COLLOIDAL STABILITY…

In this context, it is essential to know practically how to affect colloidal stability
by external parameters like temperature, particle concentration, pH, ionic
strength, impurities and surface ligands.

…by t emperature:

The impact of temperature on colloidal stability is straightforward when you look

at the potential energy curves derived from DLVO theory. From this graph, you
can clearly see that agglomeration and aggregation processes are generally
inhibited by a barrier of potential energy (ΔE in Figure 40). The only way for a
certain particle to cross this barrier is when it collides with another particle: only
when the kinetic energy of the two surmounts this energy barrier, the barrier
will be surpassed. And it´s our intention in a stable system to avoid that this
IV happens, so that particles keep their distance.

It is well known that the mean kinetic energy which corresponds to the
movement of particles (kBT, with kB being the Boltzmann constant and T the
absolute temperature) increases with temperature. And this movement is the
cause of collisions between two particles. This means that at higher temperature

T the number of particles exceeding the necessary energy barrier for agglomeration
and aggregation is elevated. As a result, aggregation and agglomeration are more
pronounced at higher temperatures. To clarify this point let´s assume you have a
collective of 6 particles with 1 particle possessing the necessary kinetic energy
to induce agglomeration or aggregation, each of them exposed to one collision
event within the observed timeframe (Figure 41 A). Let us further assume that
elevated temperature doubles the number of particles with the necessary kinetic
energy to 2. Consequently, the probability of agglomeration or aggregation due
to a collision with a particle of sufficient energy would increase (Figure 41 B).
Based on this, you should consider keeping the temperature of your colloids
as low as possible to enhance stability. Concluding this thought, storage in the
fridge is recommended, 8 degrees Celsius is fine. However, you should keep in
mind that freezing a ligand-free system should be avoided.

~~~~~~~~~~~~~~~~~~~~~~~
S t udent: the contact angle measurement is not working .

Assis tant: the cover is s till on the camera.

Sven Koenen, Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

114
 KEEP IT: STABILITY

There is another negative effect of high temperature: the liquid´s viscosity

decreases exponentially with the temperature. That is a pity, because the
diffusion constant will in turn increase by the same factor. Knowing that
diffusion is also proportional to kinetic energy (Tk, see above), means that
the speed of a particle that approaches collision with another particle is
proportional to T · eT. Hence, every degree of temperature is your foe in the
lab. Note that your finger tips are permanent 36°C heaters, and a small 1 ml vial
taken freshly from the 8°C fridge has nothing to defend itself while you carry it
to the DLS machine measuring aggregation.

Cooling is the most simple, effective, and contaminant-free way to preserve a

colloid´s stability. Everybody in a cool lab is chilled, so are the samples.

… by concentration: IV
While the impact of temperature on colloidal stability is pretty obvious, the
effect of particle concentration is less plain. You learnt that a lower mean kinetic
energy of the particles is beneficial in order to not exceed the activation barrier
of aggregation and agglomeration. In addition to this, not only the energy of the
corresponding particles can enhance the probability of aggregation but also the
frequency of collisions, which, of course, is elevated at higher particle number
concentration. Lets take the example from the previous paragraph: a collective
of 6 particles, 1 exceeding the activation barrier, one collision each. Let`s now
assume that an increase in particle concentration would double the number
of collisions (Figure 41 C). Consequently, the probability of agglomeration or
aggregation would also increase. Based on this it can be assumed that colloidal
stability is generally favored in less concentrated colloids. So to make it simple:
dilute your colloid in order to enhance colloidal stability. You will often need to
dilute it for analysis, anyway. We have had good experience keeping gold at 100
mg/L stable for months stored in the fridge. But ligand-free 300 mg/L or even
500 mg/L gold stored in at room temperature is a lucky or bad shot, maybe
depending on the degree sunshine on the day of synthesis.

500 mg
100 mg /L
/L

/L
mg
300

115
 KEEP IT: STABILITY

A B

(number of particles)
T1

T T2

n
T2
Ekin (particles) E
C > T > E D C2

(number of particles)
C1
C

n
C1 Ekin (particles) E

Figure 41: Impact of temperature and concentration on energy distribution of colloidal

IV particles: (A) Cartoon illustrating an increasing number of particles with a kinetic energy (Ekin)
> activation energy for aggregation (ΔE) and (B) energy distribution of the corresponding
particles showing a higher number of particles with Ekin > ΔE at higher temperature. (C)
Cartoon illustrating an increasing number of particles with increasing particle concentration
as well as an elevated collision probability (indicated by a higher number of arrows). (D) Energy
distribution of the corresponding particles showing an increasing number of particles with Ekin
> ΔE at increasing particle concentration, due to the higher number of particles and due to
elevated collision probability.

… by pH value:

Laser-generated metal colloids in aqueous medium are at least partially

oxidized and carry a portion of M-OH2+/ M-OH/ M-O- (M=metal) groups on
their surface. Again, this is true for most oxides as well, resulting in cationic
defects (oxygen vacancies), e.g., positively charged ZnO(1-x). Hence, noble metals
are often acidic nanoparticles, and oxides are often alkaline.

It is obvious that alterations in the pH-value could lead to a protonation and

deprotonation of these surface groups, which can critically influence surface
charge and colloidal stability. Consequently, there is a certain pH-value where all
surface groups, in sum, are in an uncharged state. This value where the charge
flips sign is termed point of zero charge or isoelectric point (IEP). Naturally, this
point highlights the regime of lowest colloidal stability and should be avoided
during fabrication and storage of laser-generated colloids. A general rule for
achieving maximum stability should be to adjust your medium pH to be as far
as possible from the isoelectric point (but still avoiding dissolution), and with
the medium being about one pH unit away from it you are often already on the
safe side. This concept is illustrated in Figure 42 A, with the pH of the medium
increasing from the left to the right, crossing the IEP in the middle picture.
Knowledge on the isoelectric point can be either obtained from literature
(already Wikipedia has a short list at the entry “isoelectric point”), in case
you are working with a well established material, or determined by titration
experiments. In the latter case you titrate your colloid by adding acid or base

116
 KEEP IT: STABILITY

and the point of zero charge is reached when your particles aggregate and
precipitate. This procedure is ideally probed by zeta potential measurements
(the concept of zeta potential is described later). But the aggregation point
is often so easy to see that IEP can be found simply with the naked eye and a
pipette of acid or base.

… by ionic s trength:

The impact of ions on the colloidal stability of laser-generated colloids is rather

ambiguous. But once you understood it you will never want to miss this mighty
tool, in particular for laser synthesis in water. In our lab, the salinity-trick is used
on a daily basis and basically entails adding low salt concentrations to increase
stability. But let’s first understand the basics. The first and foremost thing you
need to be aware of in this context is that high ionic strengths critically reduce IV
the Debye length (K-1) and a low Debye length is actually a stability death
sentence. The underlying physical phenomenon is a screening of repulsive
surface charges by an overabundance of counterions (Figure 42 B). To put it
simply: With a lot of counterions around, the particles can no longer “feel” the
repulsion from neighboring particles and hence they aggregate due to attractive
forces. This correlation between the Debye screening length and the ionic
strength (I) in water at 25 °C is given by the following equation:

Based on this you could calculate a Debye screening length (κ-1) of 30 nm in the

>
presence of a 100 µM electrolyte, however this value is reduced to only 0.7 nm
in a 200 mM electrolyte, completely annihilating colloidal stability and initiating

T
aggregation under these circumstances. As a rule of thumb you can conclude
that high ionic strengths > 2 mM should be avoided when synthesizing and >
handling laser-fabricated colloids. In addition, you need to keep in mind that the E
destabilizing effect is intensified in case multivalent ions MZ+/- like Ca2+ or SO42-
are present. So the utilization of these ions should be totally avoided during
laser ablation in water. Whereas the concentration screens stability by c0.5, the
ions valence contribute additionally by cZ·0.5 to kill stability.

In contrast to the above mentioned well known fact of oversalting (or “salting-
out-effect”), recent studies have concluded that the page is turned at a minute
amount of salt! This is an effect that is very specific and can be very beneficial
for ligand-free particles, as those are the once we make by laser ablation or
laser fragmentation in water.

The in situ presence of certain anions at micromolar salinity during the particle
synthesis process can be beneficial for colloidal stability. In situ means that the
salt has to be dissolved in the liquid before you switch on the laser. In this way, it
is most effective. This effect has been reported by the authors of this handbook
to work best with highly polarizable anions (e.g. Cl-, Br-, I-) and only on noble
117
 KEEP IT: STABILITY

metal nanoparticles with low oxidation tendencies (Au, Pt, Pd).

Using this method, the colloids are far more stable and size distribution is
narrowed. That’s two flies with one clap. But how does it work? These effects
were attributed to a specific anion adsorption to the nanoparticle´s surface
during the particle formation process. Sodium chloride works well, also
carbonate or phosphate works in our lab. Best friends are sodium bromide and
sodium hydroxide at 100 – 300 mg/L. So, during your noble metal particle
synthesis you may add low salinity electrolytes to enhance colloidal stability.
Make your first try with dissolving 200 micromol table salt. Note that this is only
0.1 mg per liter, a bit less than a grain of salt.

A
HO O
HO OH2 OH
HO

IV OH HO - H oh
+ H OH O
H 2O oh
OH OH
OH2 O
HO

(0) (-)
B
Epot Epot Epot Epot

-1 -1 -1 -1

Figure 42: Neutral is worth nothing. Stay in charge. Add a grain of salt. Effect of pH and ionic
strength on colloidal stability: (A) Cartoon illustrating the effects of stable protonation at acidic
pH (left) and stable deprotonation at alkaline pH (right) of a particle at the colloidal instable
isoelectric point. (B) Cartoon illustrating the potential energy (Epot) of two particles in the
presence of low counterion concentrations (left) and high counterion concentrations (right) as
well as the extension of the Debye length (κ-1) as a measure of repulsive stabilisation. Arrows
indicate repulsive forces between the particles (left) and attractive forces (right).

…by impurities:

During your laser ablation experiments, sometimes colloids aggregate, while

others remain perfectly stable, even though your synthesis conditions seem
completely the same. This lack of reproducibility can be highly frustrating and is
in most cases because you have worked dirty, so in other words it is attributed to
impurities in the sample. These can be simply dust particles from the surrounding
atmosphere (often partly soluble in liquid), residual particles remaining in your
chamber from previous ablation experiments as well as multivalent ions leaking
from your glass vessel or metal chamber. Hence, if you want to precipitate a
colloid (of your nasty colleague in order to keep him/her in good mood):

118
 KEEP IT: STABILITY

add calcium chloride, or make sure to deposit some salty fingerprints. Also
metal containers, stainless steel spatula and chambers deliver multivalent ions
even after using them for years. If you need to work with metal in contact with
the liquid, use anodized metals (mechanical workshops do that easily, or use
so-called Piranha liquid), or passivated metals such as aluminum chambers, or
just Teflon.

Other impurities may come from the target material, in case it is not of highest
purity. A noble metal even at 99.9% purity naturally contains silver and, even
worse, copper, easily dissolving from the target or the particles after ablation
and destabilizing the colloid via charge screening. So in conclusion, etch before
use, and buy highest grade purity.

IV
~~~~~~~~~~~~~~~~~~~~~~~
I once asked a s t udent to perform laser ablation experiment s in saline solutions,
while I gave her a concentration regime of 5 – 500 millimolar . She discovered
some int eres ting effect s on size quenching , however , we lat er discovered that she
made a mis take in her dilution series and she act ually examined 5- 500 micromlar
concentrations. This is how we discovered that salinit y quenching occurs even at very
low ionic s trengths.

Chris toph Rehbock , Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

In addition, the liquid itself is a well known source of impurities. Organic solvents
always contain unknown impurities, whose concentrations can be way higher
than your colloid mass concentration. Let´s assume a typical colloid mass
concentration of 50 mg/L and technical grade acetone with 99.9 % purity. In
this case the amount of undefined dirt (0.1 %) would still sum up to 800 mg. So
in total you have a 16 times higher mass of dirt in your sample compared to the
mass of your particles. As a consequence, the only way to avoid interference
from impurities is to work in an environment as clean as possible. But still very
high commercial solvent purities still will have enough contaminants inside to
cover the surface of all particles you´ve fabricated. So you can´t be lazy, you
need to clean your solvents on your own by distillation and checking purity with
gas chromatography is the only way out if 100% purity is what you want. But
very often, in our lab, simply the solvents used for cleaning have been detected
as source of contamination. Once the solvent bottle is unsealed, the solvent
takes up softeners and additives from the seal, the tubings, your protective
wear, etc. And this is retarded in the equipment and found as traces even during
the aqueous synthesis post cleaning. Practical advice you should adhere to is:

119
 KEEP IT: STABILITY

I) Thoroughly clean all your glassware (or use disposables), your chamber as
well as your target. Use the solvent you work with later for cleaning as well.

II) Get target materials at maximum purity

III) Purify your solvents e.g. by distillation

IV) Preferably do not use the same chamber for different materials, never use
non-anodized metal chambers or spatula

V) Use a chamber composed of a material that can be easily cleaned by the

strongest possible oxidants like “aqua regia”. We like Teflon.

VI) In case of aqueous solutions: use plastic containers instead of glass

vessels for long time storage, however in case of organic solvents,
interactions between solvent and container may render glass more
IV suitable.

VII) For the tubings, silicon is nicely flexible but the worst you can do because
it retards everything in its pores. Better go for Telfon or something similar
that is resistant to solvents and has a smooth surface.

… by ligands:

The most important characteristic of laser-generated colloids is their purity due

to a completely ligand-free synthesis process. However, some applications e.g.
utilization in a salty biological environment may necessitate the presence of
stabilizing or biofunctional surface ligands. Once you have effectively grafted
stabilizers to your surface, in particular bulky ones, your colloid will survive
everything, even freezing and repeated freeze-drying. It´s wonderful to see a
freeze-dried, golden pellet to eject a perfectly rubin-coloured liquid once you
put a drop of water on it.

These stabilizers can either just be serum proteins, present in a biological

environment anyway, chemical polymer ligands, as well as functional ligands
used in bioactive nanobioconjugates. Basically, conjugation processes can
either be conducted in situ or ex situ. During in situ conjugation, the ligands
are present during the laser ablation or laser fragmentation process. The main
advantages in this case are generally a more narrow size distribution as well
as a very high surface coverage due to an instantaneous reaction (surface
adsorption) between the just formed nanoparticle surface and the ligand. The
downside of this procedure, however, is also pretty obvious. Firstly, organic
ligands are present during the laser ablation process. This may go along with
a significant degradation of conjugates by post-irradiation events by the laser
beam, a problem which necessitates careful adjustment of the applied laser
fluence during ablation (Chapter II). Another impairment originates from the
fact that the nanoparticle-to-ligand-ratio cannot be properly controlled in situ.

120
 KEEP IT: STABILITY

While the ligand concentration in the ablation medium is fixed, the particle
concentration increases during the laser synthesis process. As a consequence,
the ligand-to-nanoparticle-ratio decreases with time and uniform conjugation is
unlikely. Ex situ conjugation, on the other side, entails a two step process where
the ligands are added to the previously generated colloids. This process allows
for a precise adjustment of the ligand-to-nanoparticle-ratio and a well controlled
conjugation process. This approach, however, requires alternate strategies for
size control, e.g. salinity-induced size quenching (see above). As a rule of thumb,
single-digit micromolar ligand concentrations already work (if ligand has high
affinity).

The presence of surface ligands has a significant impact on the stability

of colloids as surface charge density in this case will be dominated by the
properties of the ligand as well as by those of the metal core. Based on the
nature of the ligand, stabilizing effects can either be electrostatic, steric, or IV
electrosteric (illustrated in Figure 43 A), and the steric is always the most robust.

~~~~~~~~~~~~~~~~~~~~~~~
Working with subs tances in the micromolar concentration range of t en includes the
necessit y of using volumes below 10 μL which is in fact not more than a droplet. In this
case , never trus t a physicis t doing bioconjugation and t elling you to dip the tip with
your biomolecule into your colloid. You will never get nearly reproducible result s, in
particluar concerning the surface coverage.

Lisa Gamrad, Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Electrostatic stabilization usually occurs when small highly charged ligands

(e.g. citrate or short peptides) are used. When these ligands covalently or
electrostatically attach to the particle surface, their net charge is added to the
particle´s surface charge. In case the charge of the particle and the charge of
the ligand carry the same sign, e.g. negatively-charged citrate on negatively
charged gold nanoparticles, this may lead to a significant stabilization of the
colloid. However, in case the charge of the particle and the charge of the ligand
carry opposite signs, an effect commonly referred to as charge balancing occurs
(illustrated in Figure 43 B). As a consequence, colloidal stability in oppositely
charged systems is generally reduced. On the other hand, steric stabilization
is basically found when nanoparticles are coupled to bulky uncharged ligands
e.g. hydrophobic polymers (bound by van der Waals force). The main stabilizing
force at work here is Born repulsion, that is, the ligands prevent the particles

121
 KEEP IT: STABILITY

from reaching the necessary distance required for aggregation and hence
keep the particles dispersed in the solvent. Amphiphilic polymers are the best
for metal nanoparticles, since they may bind to the particle’s hydrophobic
surface, and at the same time provide hydrophilic groups keeping the particle
dispersed in water. Such a polymer widely used in laser synthesis in literature
is polyvinylpyrrolidine (PVP). If you want to go Bio, a typical serum protein like
albumin is another amphiphilic macromolecule that provides steric stabilization,
and is present in most biofluids and cell culture media, anyway.

A hybrid form of stabilization, termed electrosterical stabilization, is found in

case particles are conjugated with bulky, charged surface ligands e.g. proteins
or oligonucleotides or classical surfactants. How big shall the ligand be? As
a rule of thumb, 12 carbon bond lengths are required to add enough steric
character to the electrostatic stabilization. Maybe you heard about sodium
IV dodecylsulfate? Same works in non-polar solvents, e.g. dodecane-thiol is a
gold standard of capping agent in hexane. But if the ligand gets too bulky, its
diffusion coefficient is lowered, decreasing the effectiveness of reaching the
particle’s surface during its growth (in situ) or before aggregation starts (ex situ).

The final thing you want to optimize is the tightness of binding to the
nanoparticle surface. The stronger it binds the more is the equilibrium
shifted towards the nanoparticle, and you will need less excess of ligand for
grafting a monolayer. In ideal cases, such as thiolated peptides pr thilated
oligonucleotides, almost 100% conjugation efficiency is achieved. Decorating
the surface of laser-generated nanoparticles with surface ligands usually
requires a strong interaction between the ligand and a metal core, which is
usually achieved by electron donor moieties like amines or thiols, which strongly
interact with the metal surface atoms. It is an old rule of conjugation chemistry:
the soft Lewis acid (noble metal nanoparticle) has highest affinity to soft Lewis
base (disulfide > thiols > thioethers > carboxy, amines). For noble metals,
thiols work very efficiently, so that this is the gold standard for fabrication
gold nanoparticle bioconjugates. For the peptides it is most simple, just add
a terminal cysteine amino acid. Same works with polymers, gold conjugated
to mercapto-polyethyleneglycol (m-PEG) has survived the highest salinities
among several stabilizing agents we have tested. And for metals that oxidize
or oxides? Carboxy groups have been reported to work well to be anchored
on oxides, and you will need a carbon chain pointing outwards to contribute
to steric stabilisation. Without going into detail, you will need to have the right
pH adapted to both the molecule (deprotonation) and particle (IEP). Maybe the
molecule MEEAA (check the internet for that) or other complexing agents are a
good start.

122
 KEEP IT: STABILITY

A B

surface charge
0

ligand concentration

IV

Figure 43: Decorate. Stability delivered by ligands: (A) Cartoon illustrating electrostatic (top),
steric (middle), and electrosteric (bottom) stabilization of nanoparticles. The electrosteric
stabilization by surfactants is often a double-layer with the charges pointing both inwards and
outwards. (B) Scheme illustrating charge compensation while adding positively charged ligands
to a negatively charged colloid.

In the previous part of this chapter you learned how to define the term colloidal
stability and how it can be correlated with surface charge as well as properties
of the electrical double layer. In addition, you learned how colloidal stability
changes when temperature, pH, ionic strength are altered and when impurities
as well as different surface ligands are present. The final step to a unified
understanding of colloidal stability is how to measure it.

STEP 3: MEASURE COLLOIDAL STABILITY…

In order to measure colloidal stability you need to monitor to what extent the
properties of your colloid change with time. In this context you need to keep in
mind that visible aggregation, like discoloration of your colloid and formation
of precipitates are only the tip of the iceberg, more subtle effects are far more
frequent. Popular and very powerful tool to monitor colloidal stability are UV-
Vis spectroscopy and zeta potential measurements, which are explained in the
following subsections.

123
 KEEP IT: STABILITY

… by UV-Vis absorption spectroscopy

UV-Vis spectroscopy is a fast and easy to use standard method available

in most laboratories. The basic principle entails shining light through your
sample while the attenuation (extinction) of your light intensity is observed
in comparison to a standard, e.g. the pure solvent. When this is done for
all wavelengths you acquire a spectrum were extinction is plotted against
wavelength. More information on how UV-Vis spectroscopy works and what else
it can be used for is included in Chapter III where all relevant characterization
methods are listed. Almost all colloidal nanoparticles show a distinctive
extinction spectrum, the most obvious example are plasmon resonant metal
nanoparticles like Au, Ag or Cu. While the details of spectra interpretation are
explained in Chapter III this paragraph focuses on how this technique can be
used to evaluate colloidal stability. The simplest way to utilize this method is
IV to measure whether or to what extent your spectrum changes with time. In this
case the simplest rule of thumb is: I) The spectrum does not change with time
= stable colloid II) The spectrum changes = unstable colloid. This approach is
admittedly rather crude but gives you general information about your colloid
and it is generally applicable for all colloids that possess any kind of extinction
in the UV-Vis. A proper follow-up could be to use either the extinction at a
certain wavelength or even better the sum (integral) of all extinction values in
a certain wavelength regime to quantify your stability e.g. the loss with time
(Figure 44 A). The decrease of extinction with time is the most straightforward
phenomenon, as precipitation is meant to occur in case colloidal stability
A deteriorates. Maybe a good definition is that the integral of extinction does not
change by more than 5% in 4 weeks.

Unfortunately, it is not always that easy. Sometimes you may encounter

B
that your extinction even increases with time, would this mean that particles
get more stable? Most certainly not! This observation would be a clear
indication that there is more going on in your sample (it is still “aging”) like
chemical changes e.g. the formation of reaction products with different optical
parameters or change of particle concentration due to particle growth from
C
precursors with time. In this case the aforementioned simple stability criterion
fails and more detailed investigations e.g. of particle size need to be added
(compare Chapter III). To give one example, after laser fragmentation (post-
irradiation) of a platinum colloid is terminated, the extinction still drastically
increases as there are still atom clusters that undergo coalescence and ripening
to particles, increasing the colloid´s extinction. This is happening in real time, you
can follow it in the UV-Vis directly after laser fragmentation. So better wait until
it is calmed down, in order to probe and report reproducible properties.

However, UV-Vis spectroscopy can do more than just measure total change
in extinction. In many materials it can be used to monitor the aggregation or
agglomeration process as larger particles generally scatter light at a longer
wavelength. As a consequence, an increase in infrared extinction can be an
124
 KEEP IT: STABILITY

indicator to monitor particle aggregation. A commonly used parameter is the

ratio of primary particles to aggregates; the Primary Particle Index (PPI). For gold
nanoparticles, this value is already well established and is defined by dividing
the extinction at the gold interband extinction (@λ=380 nm) by the extinction
of the aggregates or agglomerates in the NIR regime (@λ=800 nm). A high
PPI value goes along with a low number of agglomerates and consequently
with a good colloidal stability (Figure 44 B). A similar PPI value can also be
used for semiconductor nanoparticles e.g. ZnO where the extinction at the
bandgap energy in the UV can be divided by the NIR aggregate/agglomerate
scattering. So concerning the PPI you can memorize that it is a suitable way to
quantitatively evaluate aggregation and colloidal stability. As a rule of thumb
you should keep in mind that for ligand-free gold colloids a PPI < 4 should be
avoided in order to prevent precipitation of your colloid. A PPI of > 10 sounds
stable.
IV
A
Extintion

Extintion

time
wavelength
1
2
B
Ext ( = 380nm)
PPI = Ext ( = 800nm)
Extintion
Extintion

PPI1 > PPI2

380
800

wavelength

Figure 44: Stay with the color. UV-Vis spectroscopy to evaluate particle stability. (A) Illustration
of how to use time-resolved changes in the UV-Vis extinction to evaluate the loss of colloidal
stability. (B) Illustration of the primary particle index (PPI) in gold colloids evaluating colloid 1,
with a high PPI, to be more stable than colloid 2 with a lower PPI.

… by Zeta Pot ential:

Many works use zeta potential measurements to evaluate colloidal stability, but
information obtained from this parameter can tell you much more. Therefore,
the first question which needs to be address is: “What is the zeta potential?”.
The zeta potential (ζ) or electrokinetic potential is closely linked to the above
mentioned surface charge and it is a powerful technique for the characterization

125
 KEEP IT: STABILITY

of the electrical double layer. In case charged colloidal particles are exposed
to an external electric field these particles will move within the field; e.g. in case
they have a negative surface charge, they will drift to the positively charged
electrode. The tightly bound ions in the Stern layer will follow the movement of
the particle, those situated in the diffuse layer cannot and will slip off when a
certain potential is reached (because they cannot cope with the speed of the
particle core and are peeled off by the drag force). This potential at the slipping
plane is called the zeta potential and can be considered an indicator of the
surface potential caused by surface charge. Naturally, it requires more energy
(a higher potential) to remove ions from a highly charged particle in comparison
to a particle with lower charge. However, you need to be really careful when
interpreting zeta potential measurements. On the particle you will basically
have 3 values of relevance: 1) The surface potential caused by charged surface

IV atoms, 2) The Stern potential caused by tightly bound counter ions and 3) the
zeta potential at the slipping plane (Figure 45 A). The zeta potential is the only
one of these potentials which is experimentally accessible in liquid but during all
your interpretations you always need to keep in mind that it is not necessarily
equivalent to surface potential/ surface charge. The surface charge/oxidation
state of the inorganic core can be measured by XPS in the dry state.

~~~~~~~~~~~~~~~~~~~~~~~
A weakness when working with PLAL is the glass of the chamber . Once when working
with ethanol in a flow chamber the glass cracked and ethanol was pumped out of the
chamber . The organic solvent was pumped with 250 ml/min right into the laser beam.
And so the inevitable happened: the organic solvent was ignit ed before the laser and
the pump could be s topped. Af t er the ignition source was eliminat ed the flames were
easily extinguished, but only due to a timely response.

Elisabeth Maurer , Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The next question which needs to be addressed is: “How can you measure the
zeta potential?”. The zeta potential cannot be measured directly but it is usually
calculated from the drift velocity of the charged particles in an AC electric field.
Measurement of the particle´s drift velocity (v) is usually conducted via light
scattering techniques. The drift velocity itself, however, is not a suitable value
as it is linearly dependent on the external electric field (E). In order to obtain a
measured value which is independent of the external field, the electrophoretic
mobility (μ) is usually used, which can be obtained via:

v= μ·E

126
 KEEP IT: STABILITY

The electrophoretic mobility (μ) is linearly correlated with the zeta potential and
can be calculated by approximations made by Smoluchowski and Hückel. The
most important rules of thumb you have to be aware of are:

I) Smoluchowski: Valid for high ionic strengths and larger

(i.e. > 300 nm @ 1 mM) particles

II) Hückel: Valid for low ionic strengths and smaller

(i.e.< 20 nm @ 1 mM) particles

III) Transformation: ζHückel = 1.5 ·ζSmoluchowski

A Epot
B
IV
Smoluchowski

m
particle diameter

100 m
tr
an
sit
io
n
re
gi
m e
huckel

surface
potential
stern zeta
potential potential
ionic strength m
300 m

Figure 45: Measure the charge. (A) Illustration of surface potential, Stern potential and zeta
potential in the electrical double layer of a ligand-free nanoparticle. Dashed lines represent
(from left to right): the surface charge, the Stern layer and the slipping plane. Ions located
outside the slipping plane are stripped off during electrophoretic mobility measurements. (B)
Representation of the validity of the Hückel and the Smoluchowski regime for different ionic
strengths and particle diameter.

Based on what you have learnt so far, you basically know what the zeta
potential is and how to measure it. As a logical follow-up it has to be addressed
what practical uses the zeta potential has for evaluating the properties of your
colloid, or in other words: “What can you learn from the zeta potential ?”. The
most straightforward answer is that the zeta potential can give you a general
idea on the surface charge of your colloid. The sign of your zeta potential is
a good representation on whether you find a positive or a negative surface
charges. In addition, the value of the zeta potential gives you a rough indication
on the electrostatic stability of your colloidal system, as a high value usually
goes along with a high surface charge and hence a high stability.

127
 KEEP IT: STABILITY

This correlation may, in some cases, even be used to calculate the surface
charge density. A general rule of thumb often found in literature is that colloidal
stability can be found in case ζ > ± 30 mV. However, you need to keep in mind
that this value is not universal. It was calculated based on the potential energy
required to surmount a kinetic energy kBT found at room temperature (25°C).
So obviously, there are some important points you need to consider when
correlating the zeta potential with surface charge and colloidal stability:

I) Effects of temperature and concentration: As previously mentioned,

high temperatures and high particle concentrations can impair colloidal
stability. This is usually not represented by appropriate changes in zeta
potential. Consequently you need to figure out individual stability criteria
in case these values dramatically deviate from standard conditions. And
only measure the zeta potential at the temperature that is relevant for the
IV stability information you want. Most instruments have thermostats, but
you need to wait at least 10 minutes for your sample to equilibrate to the
temperature set in the instrument´s software.

II) Effects of particle size distributions: The effect of particle size is generally
considered while evaluating whether the Smoluchowski or the Hückel
regime are used. Some difficulties may arise in case you try to determine
zeta potentials of particles with a very broad size distributions (see chapter
on Light scattering DLS). These problems, however, do not originate
from the modelling but are rooted in drawbacks of the corresponding
measurement of the electrophoretic mobility via light scattering. As the
scattering signal from bigger particles is significantly larger than that
of small particles, the signals from small particles may not be properly
detected. As a consequence you only determine the zeta potential of your
large particles while the smaller ones remain invisible (at least significantly
underestimated) to your method.

III) Effects from specific ion adsorption: When laser ablation is carried out in
the presence of micromolar salinity electrolytes, the particles are prone
to specific anion adsorption. This can yield a significant electrostatic
stabilizing effect, which does not necessarily show in the zeta potential
as the adsorption may occur in the Stern layer behind the slipping
plane. Hence these effects would not be detectable by zeta potential
measurements (but by the PPI in the UV-VIS, and in the XPS).

IV) Effects from ligands: In the presence of ligands, particularly when they are
bulky, zeta potential can no longer be considered a reliable indicator of
surface charge and stability. This drawback basically originates from the
measuring principle. During measurements of the electrophoretic mobility
you determine the movement of charged particles in an electric field.

128
 KEEP IT: STABILITY

If these particles are now covered with surface ligands, these ligands will
induce a drag force which slows down the particle. Consequently, the
correlation between electrophoretic mobility and zeta potential based on
simple models like Smoluchowski and Hückel are no longer valid and the
displayed zeta potentials are too low. This reduction of the zeta potential
was frequently observed in case nanoparticles were surrounded by a
protein corona when exposed to a biological medium and has nothing to
do with changes in surface charge. In addition, the correlation between
colloidal stability and zeta potential also fails in ligand-capped systems
as the underlying theory is based on electrostatic effects and does not
consider steric stabilization effects. A polymer-stabilized particle often has
a zeta potential around zero, although it is very stable.

In this chapter you have obtained basic insight into colloidal stability, you
know how to define it, how to measure it and how to affect it via different IV
physical and chemical parameters.

~~~~~~~~~~~~~~~~~~~~~~~
During fragmentation experiment s of an oxide in ethanol smoke emerged from the flow
chamber used. Firs t it was thought that the organic solvent was a problem. Af t er a
while it was realized that the chamber it self was the reason. The chamber was cus tom-
made by 3D-printing of nylon. Due to scat t ered radiation the chamber was heat ed up,
melt ed and charred. Therefore it is important to choose the right chamber and
to monitor

Elisabeth Maurer , Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

129
CHAPTER V

STAY ALIVE: LASER SAFETY

After reading the former chapters, you must be very confident and excited
to do some experiments on your own, maybe now you are planning which
materials and liquids to use. Also, you must be astonished by the convenience
and simplicity of this technique, just one-step for the nanoproducts. And laser
synthesis of colloids is an emerging field, growing since decades. How amazing
is that!

For experimental operation, in fact, you do not need a master’s or doctor’s

degree. But is it indeed this simple? It depends on the precautions you take and
the hazards you prevent and how smooth your experiments will go. You can
compare laser ablation of a target to a “nanomine” you dig, where the laser is
your mining tool and nanoparticles are the minerals you extract (Figure 46). In
the newspapers, television news, broadcasts, etc. lots of reports cover mining
accidents, some disastrous ones even causing peoples’ death. The average
V lifetime of the miners who worked tens of years’ ago was less than 40 years.
From this fact, you should get a glance at the danger hiding behind traditional
mining. Since the technique of laser ablation in liquid is an interdisciplinary field of
chemistry, physics, material science, optics and engineering, the dangers during
mining of nanomaterials by laser digging in liquids are summarized from different
subject perspectives (e.g., chemistry, material and optics), in order to show you
the appropriate precautions to prevent the dangers of being poisoned or injured.
In detail, the contents of this chapter follows the sequence of Laser risk →
Precaution → Experimental operation → Sample Labeling → Nanowaste disposal
→ Nanocolloids Storage. Note that the main purpose for this chapter is to provide
some introductive information about experimental workplace safety specific
to laser synthesis and processing of colloids. It is not intended to replace the
mandatory chemical and laser safety training from institution. Perhaps you read
this chapter to prepare yourself with questions for the safety training day?

liquid

gold

Figure 46: Nanomining in liquid using laser gun

130
 STAY ALIVE: LASER SAFETY

STEP 1: LEARN THE RISKS

Maybe, till now you have never had a chance to use a laser but most of you are
familiar with the scenes of lasers. They have become an important element in
science fiction movies, laser guns in Terminator 2 and Star Trek or lightsabers
in the Star Wars universe. Although your interaction with lasers won’t be as
lethal as in those movies, the consequences can still be so severe. In the
following part, four sections related to laser risks are introduced, including laser
classification (section 1), laser damage on human eyes (section 2) and skin
(section 3) and some typical accidents that have already happened before.

1.1 Laser Classification

Lasers are generally classified into four groups (Class 1-4) according to U.S.
Food and Drug Administration (FDA) laser regulations and international standard
IEC 60825, as shown in Table 2. The laser beam of a Class 1 laser is safe under
normal operation. Even direct exposure to eyes will not induce any damage.
But the laser power threshold is very low and varies with laser wavelength.
V
For example, the often mentioned Nd:YAG laser system at a fundamental
wavelength of 1064 nm has a laser security threshold that is only 0.6 mW, at the
second harmonic wavelength (532 nm), this threshold even drops down to 0.39
μW, and at fourth harmonic (266 nm), the threshold will be further lowered to 0.8
nW. You should be aware that a potential hazard still exists when the laser beam
is enhanced or refocused by a telescope or microscope with large aperture.
When a laser is in the range of Class 2-4, a label should be put on the door
of your laser lab at the eye level to warn the potential risk of “laser exposure”
as shown in Figure 47. Class 2 lasers are also considered to be safe thanks to
the self-defense mechanism of human eyes, the so called “blink or aversion
response” which limits the continuous exposure to less than 0.25 s when a bright
light enters your eyes. Jet pilots have that kind of fast reflex. If you intend to
suppress your eye blink under this condition, eye injury will still occur.

~~~~~~~~~~~~~~~~~~~~~~~
The s t udent switched on the high power laser sys t em to perform the proper adjus tment
of the laser . During the adjus tment he noticed a burning smell and switched off the
laser again. Af t er examining the set up and lab, he realized some black spot s in the
windows curtain. An inadvert ently movement of one of his mirror with his elbow was
the trigger for this incident.

Sebas tian Kohsakowski, Duis burg

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
131
 STAY ALIVE: LASER SAFETY

Most laser pointers at 1 mW fall under this class. But note that these are only
eyesafe for healthy, awake people. After 6 hours working in the lab or listening
to 8 hours of chemistry classes, your reflex will be far slower and even these
“supermarket lasers” are not eye-safe at all for you and your colleagues,
anymore.

When the laser scales up to Class 3, especially 3B with medium laser power from

1
5-499 mW, eye hazard appears whereas skin hazard is not so apparent. The
degree of eye damage depends on the laser power. The higher the power is, the
less time needed to cause irreversible retinal damage. Therefore, eye protection
by eyewear is indispensable and a mark of eye risk should be posted close to

2
the label of “laser exposure” warning. The most dangerous lasers are Class 4
lasers because they can cause severe skin burn or permanent eye damage. Also,
in this case, the laser power is high enough to ignite flammable/combustible
materials/gases, such as paper, ethanol and hydrogen, as a result it can trigger
an explosion when the flammable gases are in high density. If you want to make
a grill outside the lab and you are short of lighter for the fire, lasers at class 4
V are good candidates as fire-starters. Note that practically all lasers used for
laser synthesis and processing of colloids – your lasers – are in that high class.

3
It’s the same class as lasers used to weld a car roof. You think the power of
your laser is lower? No, it’s only the average power that’s lower, pulsed lasers
often create pulse powers and intensities that are even higher than those used
in steel industry. And a single nanosecond pulse may be enough to cause

4
severe, irreversible damage to the eye. Note that by using special laser safety
encapsulation you can also reduce the class of your laser system from class
4 to class 1. This has to be done by certified personnel and you know this by
getting a certificate and a plate installed at the laser. But note that its only safe
when the enclosing is shut, once you open the door to install a chamber or
sample, it’s a class 4 laser again, and in this situation it is enough that electrical
power is on, no matter if the beam shutter is closed since these sometimes fail.
Hence, before you remove the casing on a class 1 laser like this, everyone in the
lab has to wear laser glasses and lab entrance has to signal laser operation.

~~~~~~~~~~~~~~~~~~~~~~~
Assis tant I: You can’ t walk around here with your coat.

Assis tant II: Sure - as if this coat ever saw chemicals

Alex Heinemann, Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

132
 STAY ALIVE: LASER SAFETY

To help researchers circumvent the potential risks from the class 4 laser lab,
a brief summary titled as “workplace safety standards” about the parameters
of the laser systems (e.g., maximal power, laser wavelength, pulse duration,
repetition rate, etc.) as well as the correct operation procedures should be also
posted near the laser warning label, as shown in Figure 47 For the sake of the
researchers’ security, a lamp to signify the laser current status (on or off) should
be also assembled in a prominent place outside the laser lab. A lamp for the
laser status is indispensable in order to rule out the possibility that people get
injured without notification.

LAB workpalce
safet y s tandards

Figure 47: Warning label for laser in Class 2 and higher.

Table 2: We´re in a high class. In laser synthesis, you will have class 4 laser
setups according to official Laser Classification Table.

Class US:FDA/CDRH IEC60825 Example

(Amendment 2)

• No known hazards during to eye or skin during

• laser printers
normal operation
Class 1 • CD players
• Note: Service Operation may require access to
• DVD devices
• hazardous embedded lasers
• geological survey
• No known hazards • equipment
to eye or skin,
Class 1M N/A • laboratory
unless collecting
optics are used

133
 STAY ALIVE: LASER SAFETY

• Visible lasers not

intended for viewing

Class 2a • No known hazards N/A

to maximum
exposure time of • visible continuous
1000 seconds wave Helium-Neon
lasers
• Visible lasers
Class 2 • laser pointers
• No known hazards with 0.25 seconds (aversion
response) • laser scanners

• No known hazard
with 0.25 seconds
Class 2M N/A
(aversion response)
unless

• Similar to Class 2
with the exception
that collecting
Class 3a optics cannot be N/A

V used to directly view

the beam • HeNe lasers above
• Visible only 1 milliwatt but
not exceeding 5
• Replaces Class 3a milliwatts radiant
(with different limits) power

Class 3R • 5× Class 2 limit for

N/A
visible

• 5× Class 1 limit for

some invisible

• Spectrometry
• Medium-powered (visible or invisible)
• stereolithography
Class 3B • Intrabeam and specular eye hazard
• entertainment light
• Generally not a diffuse or scatter hazard
shows

• High powered lasers (visible or invisible)

• (all!) LAL lasers
• Acute eye and skin hazard intrabeam, specular
Class 4 and scatter conditions
such as Nd:YAG or
Ti:Sapphire
• Non-beam hazard (fire, toxic fumes, etc.)

134
 STAY ALIVE: LASER SAFETY

1.2 Eye Damage

simplified cross sectin of the human eye

vitreous
Ciliary Body humour
(focus muscle)
selera
lrns
retina
iris
fovia
laser (macular
lut ea)

optic nerve

laser beam are almos t parallel thus the eye’ s lens will focus
them down to a small spot causing retinal burns

Figure 48: Laser focus when it enters the human eyes

V
The interior of an eye is shown in Figure 48. It mainly contains the cornea on
the transparent front part being responsible for protecting eyes, the iris serving
as light modulator for controlling the light volume inside the eyes, and the lens
which focus the light on the retina. The retina consisting of many photoreceptors
detect the light information and convert it to electrical signals which are then
carried to brain through optic nerve. Your friends think one watt is nothing, less
than a light bulb? To better understand the laser’s effect, you can do a small
experiment by yourself using sun light as a reference. When you directly view the
sun (please do it only shortly when you are in Australia), the power that arrives
at the retina is about 10 W/cm2, in which case you can not bear long-period
of exposure and may become dizzy afterwards. In comparison, 1 watt laser
beam can be focused to already 100,000 W/cm2 (!) at the retina, four orders of
higher than sun exposure, therefore damage may immediately occur to the eye
structures and cause permanent loss of vision. The pathological effects (e.g.
cataract, corneal burn and photokeratitis) towards eyes (e.g., retina, eye lens
and corneal) are laser wavelength dependent due to resulting different photon
energies, as summarized in Table 3. The lasers with wavelengths from 400 –
1400 nm are much more dangerous than other wavelengths so that this zone is
specially defined as a “Retinal hazard region”. When the retina is damaged, the
vision will be dramatically reduced. Such vision loss is normally permanent with
little chance to be repaired, while laser with wavelength of 315 ~ 400 nm cause
injury toward eye lens and trigger the photochemical cataract. Near infrared
(near-IR) and infrared (IR) laser (i.e. those with the wavelength in the region of
0.78 μm – 1.4 μm and 1.4 μm – 1 mm, respectively); both lead to corneal burn.

135
 STAY ALIVE: LASER SAFETY

~~~~~~~~~~~~~~~~~~~~~~~
Af t er her accident, Anne H. said she would not have expect ed that she did not notice
immediat ely that a part of her retina has got irreversibly damaged. A part of her vision
angle was los t, and she only noticed it lat er the day when she drove a car and got
headache. The doctor did the right diagnosis, but couldn´t do any thing anymore. She
can s till see with both eyes and drive car again (af t er some months her brain needed to
compensat e the vision angle on her upper right vision hemisphere that was los t), but
always has to keep in mind having a “dead angle”.

She was trying to remember how that could have happened. She was wearing a clock on
her arm while adjus ting an invisible 1 wat t femtosesond laser with a beam viewer card,
and she thinks the glass or housing of the arm clock might have reflect ed the beam
towards her eye.

V ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

If this just happens to the corneal epidermis, the vision can recover in two or
three days, but if it occurs to deeper part of the corneal and causes corneal
scar, you will become incurably blind. Therefore, put your security first, you and
all other persons in the lab, including your professor, have to wear laser goggles
to lower the risk for your eyes to be exposed by laser beam.

Are you safe with the certified laser goggle? Yes, if you use it as protection
and within its specification. This means that it has to be suitable for your laser
source, if you have more than one laser source the goggles might not be
certified for both lasers. They are made to stand some seconds of direct laser
radiation, but are not build to allow directly looking into the beam for more than
a minute. Unfortunately, this laser break though happens suddenly, due to
So-called incubation effects.

In our lab, we only have fully closed safety goggles. Unfortunately, you will not
look like Keanu Reeves or Lara Croft with that decoration, but is has several
advantages: None will borrow or steel it for a party. You can share with friends
wearing eye glasses. And dangerous side reflections (from chamber holders,
misalignment, back reflection, walls, etc.) are captured.

136
 STAY ALIVE: LASER SAFETY

Table 3: Pathological effect towards eyes and skin of laser exposure at

different wavelength (as summarized by some institutes and universities for
safety guidance http://www.oseh.umich.edu/pdf/LaserEyeSkinHazards.pdf).

Wavelength range Pathological effect towards Pathological effects towards

skin

photokeratitis (inflammation Erythema (sunburn), Skin cancer,

180–315 nm
of the cornea, equivalent to Hyperpigmentation, accelerated
(UV-B, UV-C)
sunburn) skin aging

315–400 nm photochemical cataract

Hyperpigmentation, Erythema
(UV-A) (clouding of the eye lens)

400–780 nm photochemical damage to the Photosensitive reactions,

(visible) retina, retinal burn Erythema V
780–1400 nm
cataract, retinal burn Erythema
(near-IR)

aqueous flare (protein in the

1.4–3.0μm (IR) aqueous humour), cataract, Erythema
corneal burn

1.3 Skin Damage

Besides being harmful to eyes, laser beams are also detrimental to the skin after
long time exposure originating from either photochemical reaction (mainly from
UV laser) or thermal reaction (mainly from IR laser). The laser power threshold
for skin injury is almost the same as eye damage except in the retinal hazard
region (400–1400 nm). Due to large surface area of skin compared to eyes, it
is highly vulnerable when the operators are not dressed properly. Similar to
eye damage, skin damage is also laser wavelength dependent due to different
accessible depth of skin tissues by lasers at different wavelengths.

Human skin is composed of epidermis, dermis and subcutaneous tissue. The

epidermis layer with a thickness of 50-150 µm is the outermost living layer of
the skin laying beneath the stratum corneum. A layer of dermis consisting of
collagen, elastic tissue, and reticular fibers locate underneath the epidermis with
varying thickness of 0.3-3.0 mm, providing the elasticity and supportive strength
of skin. The subcutaneous tissue is a layer of fat and connective tissue to
help regulate the skin temperature. In general, the harmful effect is associated

137
 STAY ALIVE: LASER SAFETY

with the laser exposure period. When skins are exposed to laser for several
nanoseconds, only superficial evaporation happens, while in the range from 100 μs
to several seconds, thermal effects are predominant, longer exposure than 100 s
can induce photochemical effect. The symptom of erythema (e.g., burn) normally
originates from immediate effects and thermal effect. Similarly to the blink self-
defense mechanism of eyes for visible radiation (not for the invisible radiation),
skin also has a warning signal when it feels the danger of laser exposure from
almost all lasers except for some high-power far-infrared lasers.

~~~~~~~~~~~~~~~~~~~~~~~
Also a professor is not impeccable: every body who works with anti-reflex coat ed
windows knows: „never touch the coating”. But sometimes, when a hardworking s t udent
present s proudly his new chamber to his professor he of course touches the window and

V damages it s coating .

René S treubel , Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

After skin absorbs laser energy, the skin temperature increases and a sensation
of warmth comes and then human subconsciously responds to get rid of further
laser exposure. Any laser with power larger than 0.1 W/cm2 can give a warmth
warning. A temporary skin pain form short period of laser exposure will heal very
soon and will not cause severe injury, but it may become more serious when skin
are exposed to high power laser irradiation for a long period.

There are three degrees of skin burn according to the injury degree. The
first degree only affects the epidermis like the sunburn and causes pain
and erythema, but erythema often heals within a few days without a risk of
leaving behind a skin scar. If all layers of the epidermis and part of the dermis
are affected by laser irradiation, erythema scales up to the second degree.
Furthermore, when the full thickness of both epidermis and dermis and part
of subcutaneous tissue are affected, third-degree erythema forms and usually
leads to a permanent scar. Lasers with wavelengths less than 200 nm or larger
than 1 μm are mainly absorbed by the epidermis layer, UV-A can reach dermis
layer, while visible and near IR (700-1200 nm) can reach subcutaneous layer.
Compared to other laser wavelengths, UV lasers, particular UV-B lasers
(280 nm - 315 nm) can only induce the second degree of erythema, however
they are much more harmful because of their induced long-term carcinogenic
effects leading to skin cancer. Besides carcinogenic effect, long-term

138
 STAY ALIVE: LASER SAFETY

exposure to UV lasers are also able to accelerate skin aging and give rise
to hyperpigmentation (pigment-darkening effect), erythema and blistering.
Although large areas of skin exposure is unlikely to happen in normal laser
work, precautions are still needed to be taken to prevent these adverse effects,
especially in summer when people dress less, and do not work with chemicals
(colloids) that are to be handled with a labcoat and gloves. Hence in a laser lab,
even when working with gold and water only, protect your skin.

Ultrashort pulses may not even hurt in a classical sense, and photochemical
damage may feel “just itchy” with red skin. As skin damage by lasers is quite
complex to classify how severe it is, and it may even promote cancer or be
underneath the skin: always seek the doctor´s advice.

1.4 Accidents: Little Mistakes with Big Consequences

The danger of laser exposure towards both eyes and skins has been discussed
above. In the following, we will introduce some accidents to show you what
can happen if you make similar mistakes, giving you also a warning to be V
very careful to circumvent similar risks from the static “laser weapon” in you
lab (Figure 49). It is worth pointing out that if you even have many years of
laser experience, you still need to be very careful because you are the most
endangered group who might neglect the rules and take your experiments for
granted, even though nothing bad has happened in your lab.

In March 2004, a Berkeley graduate student was subject to a serious injury to

his left eye from an Nd: YAG laser. At the beginning, he carefully wore eyewear
during his experiment. But when he returned back after dinner he thought there
was no danger for laser exposure any more since the system was aligned. Then
an accident happened to him from a reflected laser when he leaned over the
optical table and adjusted the power meter. This accident affected his professor
as well. Since then, his funding was significantly cut. Although the student’s
vision was impaired a little several months later, he must have regretted about
his misconduct and the corresponding bad effect on his professor (a typical
laser accident for students and professors, http://www.photonics.com/EDU/
Handbook.aspx?AID=25167). Wearing the eyewear would have completely
avoided this accident. But similar accidents also occurred in other laser labs,
even though some of the personnel had many years of laser experience.
For example, in October 2001, an Argonne scientist with over 15 years laser
experience suffered a serious eye injury from beam while manipulating the
800 nm beam from a Ti: sapphire laser (invisible, femtosecond) with a mirror.
The consequence was disastrous, reducing his vision to near-blindness. A
Chinese scientist burned his retina and lost some sight in his left eye because
he removed his goggles to check the crystal he used. Secondary-reflection of
laser light from a cylinder of liquid caused a central foveal lesion with vitreous
hemorrhage of a Dutch scientist.
139
 STAY ALIVE: LASER SAFETY

Above are some painful lessons from experiment users, from which the laser
safety training is strongly reinforced in every country nowadays. Under the
current condition that laser power is increased almost year by year to a new
level, any mistake may be disastrous. But also 1 watt is too much for our eye it
is a pulsed laser. Generally, in human’s unconscious mind, they fear the things
unknown or dangerous so that they will strictly follow the rules whatever they
are taught, especially the young students. When they become experienced and
see nothing bad has happened, they may relax and decline his/her vigilance
level and deliberately challenge the rules, hence the danger comes and may
be triggered by a small mistake. This scenario is very familiar to the most often
reported drown swimmers who have good swimming skills. But they normally
overestimated their strength and underestimated the potential danger of the
torrent or river they challenged, they eventually went to heaven quietly without
telling their parents and friends.

V
danger
do not
look into
laser with
remaining
eye

Figure 49: A scenario of eye damage by laser when you do not wear goggles.

Other non-beam potential hazards should also be considered. Laser systems

are typical high voltage devices with several kV at high laser power. Most of
them are equipped with a high pressure water cooling system which may also
create a danger if there is something wrong with pressure control. “Laser
accident database (http://www.rli.com/resources/accident.aspx)” lists many
kinds of accidents on account of the researchers’ misconducts or the laser
malfunction. For example, the fiber of a fiber laser suddenly broke in half and
shot the operator who was calibrating laser and the shutter failure made the
laser leak out of the laser system and burned the labia of the operator. A faulty
shutter accidently caused exposure from ruby laser to a French scientist and led
to central scotoma and pigmented scar. These accidents should arouse your
attention of regular maintenance of your laser system and the optical elements.
Furthermore, during laser alignment, pay attention to applicable range of the

140
 STAY ALIVE: LASER SAFETY

optics, which may also cause big consequences. For example, a laser hit a
researcher’s eye and resulted in a retinal burn when the beam exceeded prism’s
critical angle (USA) or burned a hand and the base of a thumb while doing
adjustments to the laser (Germany). In some cases, the optical elements are not
fixed well, it may deviate of desirable direction and hit an unprotected people
working into lab or just passed by.

The layout of the laser lab is another important issue from the safety concern,
especially the room with glass windows. Chemical labs often have a window for
safety reasons, but laser labs shall not. In 1980, in a laser demonstration class
in an American University, the beam went out of the class window and struck a
student passing by, resulting in his shadow interference and blurred vision. So
put a blackened alumina plate on the window, and place a warning sign outside.
There are sometimes people who cannot read, or bosses who mainly follow on
what they have written themselves. Best and cheapest way is a safety curtain
behind the door (it’s a heavy plastic curtain that can easily fixated and cut in
any length needed), anyone who is entering is protected against direct laser
light, and can communicate with the people in the lab. That’s also the place V
where we the laser glasses are hung (outside installment is often not wanted,
and inside they do not provide optimal use to entering people in case the laser
is already on).

STEP 2: TAKE PRECAUTIONS

Since your nanoparticles are synthesized in liquids, both optical and chemical
precautions are indispensable. Here are some precautions tips (Table 4)
leading you to the shining way that at least can keep yourself healthy and
alive if no natural disaster happens. Let’s start how you should dress. The
principle of clothing is to reduce the chance for you to be exposed by both
chemicals and laser exposure. Your shoes should be resistant to the chemical
solution permeation. No flip flops! Your pants should be as long as possible
to cover most of your skin, in which case those with long-sleeve and no cuffs
are preferred. Another choice is to use professionally protective clothing, one
is for chemical protection containing garments, gloves, boots and coveralls.
Experimental clothes cannot be repeatedly used in different experiments and
cross used by different persons. Each individual should have his/her own
protective clothing and one closet for clothing storage. Gloves, especially the
chemical resistance gloves are generally selected according to the solutions/
solvents you want to use. Take DMSO for example, when it gets in touch
with human skin, it may induce allergic reactions or headaches, even worse
symptoms if it brings about the aggregation of contaminants and toxins on
human skin. To prevent the penetration of DMSO to human skin, butyl rubber,
fluoroelastomer gloves should be adopted rather than nitrile ones because
they may be degraded by DMSO. Ventilation system is also suggested to be

141
 STAY ALIVE: LASER SAFETY

established in the working places to reduce the concentration of volatile

flammable gases (e.g. ethanol or acetone) or to remove poisonous substances
generated during laser ablation.

Table 4: Different protective equipment during nanoparticle synthesis by laser

ablation or processing.

Type General precautions

• closed-toed shoes made of a low permeability material

• long pants without cuffs, a long-sleeved shirt

Dressing • gauntlet-type gloves or nitrile gloves with extended sleeves

• wear splash goggles

• watches and jewelry (e.g., bracelet, ring, etc.) are not allowed
V • Put on eyewear fitting your laser system

• Do not wear the gloves outside the lab

• Cover your hair under the clothing or bundle it if they are long
Laser
• Check the wavelength and power fitting between optics/
scanner and laser beam

• Check the stability of every optical elements along the laser

path

• Check the cracks and deformities of tubes and bottles before

use

Tools • Ensure the containers are dry and clean.

• Never overfill the tubes (threshold of 3/4 volume).

• Remove nearby flammable/combustible materials to limit fuel

Environment
in the case of fire

Others • No food or drink inside

142
 STAY ALIVE: LASER SAFETY

Both men and women dress nicely to give a good impression to others. What
are favorite things for them? In my opinion, most men like cars/watches to
show their wealth and taste, while women like all kinds of jewelries as more
and shining as possible to put on their head, ears, wrist, etc. Most of these
decorations are shining under the sunlight, making you feel comfortable when
you touch them. Sometimes, they are the gifts from your friends and are
very meaningful for you as a period of good memory is behind them. If you
get married, you also have to wear rings to demonstrate your marital status.
Do you know that some of your favorite watch/jewelry (e.g., bracelet, ring,
etc.) may endanger you during your laser experiments, especially during the
laser alignment? The surface of watches/jewelries are always smooth enough
to reflect laser light, which may impose risks to your skin or eyes or other
researchers or devices nearby causing skin/eye damage or device destroying.
For instance, watches have shiny metallic housing and glass window, and you
put your own hands on the optical bench: in principle, no laser beam is on while
operating, but in reality it can easily happen that the laser is on and your clock
reflects its beam in random directions. The necklaces are more subtle because
you are used to suppose they are well adherent to the body, and instead they
V
are ready to hang in unknown areas of the optical bench when you lean on it
to manipulate optics. Although these can be bad news for girls, for boys it may
actually be a good one: they are legitimated to adopt an elegant pocket watch
because the wrist is not recommended to those who operate on optical tables!

Therefore, for your own security, take them off before you enter the laser lab.
Additional attention should be paid to your long hair, which may catch fire by
laser expose. Put your long hair under the lab clothing or fasten it on your head.
For laser eyewear, goggle, it is often functional in specific wavelength and has
a maximal threshold of laser power, check the applicability of your goggles,
otherwise it will be useless to protect your eyes.

Another rule to minimize risks with laser light is NEVER to sit at the same height
of the optical beam, especially when someone else is operating on it. If the optical
table is equipped with a PC, move it (in particular the screen) higher than the
optical bench, so that it has to be used while standing, making sure the operators
head is higher than the optical bench (for example while sitting on a stool).

In last section, the danger from laser exposure has been discussed. Precautions
should also be taken to other optical elements/scanners other than the laser
system itself. First check all the optical elements and fasten them during the
light alignment and make sure that the laser is stable all the time and has no
chance to deviate its direction. Another suggestion for the optical elements
is to make them arrange in a good order, not so crowded and crossed in
another experimental set up. When two laser experiments are implemented
simultaneously, the risk is squared. Meanwhile, these optical elements are better
to be kept away from the edge of the laser stage to give your arms some free
space to move and for your notebooks to be put on top to record the data.
143
 STAY ALIVE: LASER SAFETY

Attention should also be paid to the wavelength/power fitting range of the

optical elements/scanners. If they are not fitted well, the optics and scanners
will be damaged, which means you will lose thousands of euros and have to
postpone your experiments to wait for their substitutes to come back.

In order to make your experiments reproducible, all the tools you use must be
very clean, including the ablation chamber, glassware, etc. This is because you
are not sure whether your colloids will interact with the pollutants and whether
the pollutants or the reaction products are poisonous. Meanwhile, as already
addressed in Chapter IV, from the viewpoint of nanoparticles reproducibility, it is
also strongly recommended that all of the tools you use are pure, if possible as
pure as laser-generated NPs.

Great concern should be drawn to the environmental factors, including

inflammable solutions, windows blocking and spilled colloids collection. Do not
store large volume (only the volume for 11 day consumption) of inflammable
solvent in the lab, such as ethanol, methanol, etc. because once there is a fire,

V these solutions are good fuels for fire expansion. In some cases, it may also lead
to explosions. The experiments using these inflammable solvents need also to
be very careful, in particular when the laser energy is high.

~~~~~~~~~~~~~~~~~~~~~~~
Once upon a time , a s t udent (not named here) used UV laser to do some UV laser
fragmentation using ethanol flowing out of a funnel . A fire s tart ed at the beginning
of int eraction between laser pulses and the falling down ethanol , it immediat ely
expanded to the beaker which was used to collect the ethanol . Some of the splashed
ethanol even dropped out and burned the laser table. Fort unat ely, the s t udent
responded very quickly. He s topped the laser sys t em and moved his not ebook away. Af t er
about one minut e , the fire extinguished nat urally without causing serious consequences.
Jus t image if there is some flammable solution nearby, the fire may expand very quickly
without control , and the whole lab may be des troyed, in particular the laser beam
source which is very sensitive to aerosol contaminations.

Dongshi Zhang , Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Most of you have heard of the “laser tag” game, some of you may have already
played it before. It is a funny game where you can use laser guns to your
opponents to collect points and become the hero of your team. In this game,
once you are hit, you are deactivated and after tens of seconds you revive and
can go on playing. As mentioned above, real lasers can shoot people passing
by like in this game. But the big difference is that the one who is shot by laser in
reality might not have a chance, and you will regret to have doing this to him/

144
 STAY ALIVE: LASER SAFETY

her. Then how to prevent such “laser tag” game happen in your lab? One way
is to block all the possible spaces for laser beam to go out of your lab. You can
either cover the windows using high ablation threshold materials or install a
metallic fence outside the window to absorb the laser energy. Since the laser
arrived at these materials is not focused, the materials will not be ablated
dramatically. To ensure the security of your colleagues in you lab, you can use
some metallic resisting shields to block the laser from all possible reflected or
transmitting directions (we installed simple metal sheets and tubes guiding the
beam wherever often justage is not required. Of course the metal in anodized
and blackened to minimize reflection, just in case).

Before experiments, get to know something about your materials and predict the
products you could get. Take the corresponding precautions according to the
potential risks you may face.

STEP 3: TURN ON THE LASER

When all the materials and the tools are ready for your experiments and you V
can take all precautions described in the previous section. The time has come
for you to do experiments in liquid. When you step in front of the laser lab,
even though the laser status lamp is off, still do not enter the lab without laser
goggles. Wear them and check the condition in the lab yourself. When you go
inside the lab, do not forget to close the curtain shield near the door to give
your colleagues a safe space to dress goggles.

In this section, some hints that are easily neglected are summarized. First,
before you start the laser system, check whether the lens cap is still on. If
so, take them off. Sometimes, you start the laser but you can not find the
beam. You may get confused or become worried whether the laser system is
malfunctioning during your experiments. This small mistake often happens. The
ablated cap fogging will condense at the lenses, and our laser experiments will
have less yield. Lens cleaning is very time consuming.

Second, put all your things in a good order, do not put them randomly on the
table, especially fragile glassware. A messy layout of your samples may distract
you and might result in the spill of the chemicals when you behave improperly ,
in particular when you are in hurry. And scientists always are.

When you switch on the laser system, you have responsibility to inform other
colleagues who are still working in the lab and give them enough time to put on
the goggles. For those students who are curious about the goggles, sometimes
they have a compulsion to check the ability of goggles by holding it bin the
beam. Do never do this, trust their quality, otherwise you and your colleague
have the risk to be exposed by high power laser beam (as the goggle stands the
beam only for limited time, once, at one spot. It incubates the exposure but you
don´t see it). You should also know how you can turn off the laser in the fastest
145
 STAY ALIVE: LASER SAFETY

possible way in case of an emergency, so check for emergency-buttons. When

you finish your experiments, first remember to check whether your colleague
have finished their experiments. Under this condition, saying is better than
seeing, ask them: is laser on? If they went for lunch, shut it off and ask them
to pay next lunch. Only when you are sure that other laser system is off, you
can take off your goggles. During your experiments, maybe you are short of
materials or liquid or just want to go to the restroom, during you absence,
please remember to turn the laser power to 0% and close the shutter (not
the laser system), otherwise, someone may enter the laser lab and get injured
without be notified (I remember my Professor was demonstrating to a guest the
laser lab while my student was out for a break and he almost burned his hand
by the UV laser). During break, of course, leave gloves and protective clothing
in the lab.

3.1 Focus finding

V In Chapter I and II you have learned how to find the laser focus. For this purpose,
some people like to use a piece of paper, in which case due to high energy near
the laser focus, paper is easily burned, releasing some unpleasant smoke. Trying
to stop the fire will distract your attention to forget the existence of the laser.
And as a result, you can easily be burned yourself. Therefore, paper should be
totally restricted for searching for the laser focus. The proper way to follow
the beam is by using low laser power and a detector which is suitable for the
wavelength you use, e.g. an IR detector for 1064nm (so-called beam viewer).
Alternatively, a watered telefax paper can be used at low power. Precise finding
of the focus is to weight the target before and after several minutes’ ablation
at different positions along the beam transmission direction and then choose
the position where you achieve the biggest difference of the target weight. Note
that targets reflect laser light. This method is a little time-consuming because
the sample must be dried and weighed twice for each focus position. But it’s the
only method that is correlated to the real nanoparticle yield, as highest yield is
often achieved mm away from the geometrical focal plane in air (see literature
or email the authors abut that issue).

3.2 Long-period Ablation

In Chapter II you have learned how to obtain a large amount of nanoparticles.
For this purpose, sometimes, long-term laser ablation is needed, which is always
implemented by continuously flowing liquid with the aid of a pump. In this case,
liquid may be spilled out due to the improperly sealed pipe and will arouse a
disaster for the electrical lines or devices embedded underneath or positioned
on the floor. Meanwhile, it will also rust your optical table, which will not be
repairable any more.

146
 STAY ALIVE: LASER SAFETY

~~~~~~~~~~~~~~~~~~~~~~~
When I firs t arrived at my current lab and did some experiment s using flowing liquid,
due to inappropriat e connection of the chamber , some wat er flowed out and then fell
down on the ground. I did not recognize this was a big problem, but af t erwards when
it became a pool , even enough for a frog to swim. I suddenly saw some electrical wires
and a socket were floating on the wat er , “oh, how s t upid I am and how dangerous it
is!” Then I s topped the liquid and then s tart cleaning the floor , and luckily nothing
happened at time. But the optical table now has rus t spores.

Dongshi Zhang , Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
When you plan to do long-period laser ablation, liquid evaporation is another
issue to be taken into account. Organic solutions are volatile, as the ablation
proceeds, the liquid thickness gradually decrease. Once they are unable to
immerse the target underneath, you are performing laser ablation in air, in
which case the ejected particles will no longer be confined in liquid, but haunt
V
in the air. As a result, it is easy for you to inhale them and receive toxic effect
after inhalation. Another concern is the flammability of solvent vapor. Once
the environment is rich of this flammable gas, the laser will then become the
matchstick to ignite it and an explosion may also happen. By law, it is not
allowed to combine an ignition source (a pulsed laser) with an inflammable
liquid. So how to do laser ablation in ethanol or acetone, as widely reported
in literature? To answer this questions expands the scope of this handbook. In
general, you need to write a risk assessment scenario together with your safety
officer, taking into account what may happen and which safety measures can be
appropriate. In our lab, we focus a 500 watt picosecond laser in a flow chamber
pumping large amounts of inflammable liquid. For that, we have built an active
safety chamber around it with gas sensors and valves that automatically open
and flood the chamber with nitrogen, and shuts the beam and the liquid pump.
This is the upper end. For laser ablation of gold in a 20 ml ethanol batch the
safety measure will be less extensive, but still you will have to take into account
what may happen in the worst case. Consult your safety staff before working
with organic liquids and lasers.

STEP 4: BE ORGANIZED
When you collect colloids, what you have to do is labelling your sample. Sample
labelling is a good habit for you to know the information after many days. In
China, there is famous saying “good memory is inferior to sodden ability
to write”. Even if you have memory like a computer and can search for the
information whenever you want, it is still better for you to categorize the

147
 STAY ALIVE: LASER SAFETY

samples. Sample labelling should be done labelling at least to their composition,

production date, and name (e.g., “gold-silver-alloy in acetone, 28.4.2014,
S. Yusuf”). In the lab book, you write down processing parameters (e.g., laser
power, repetition rate, scanning speed, wavelength, pulse duration, processing
period; liquid volume, liquid thickness, liquid composition), etc. Do not forget to
put the foreseen disposal date as well.

One of the most important things from the security point of view is to check
whether your colloidal solutions have one or some of the following hazards
(Figure 50), e.g., exploding bomb, corrosion, flame over circle, gas cylinder,
environment, skull & crossbones, exclamation mark, health hazard and flame.
Besides the exploding bomb and gas cylinder are very rare for the nanoproducts
from laser ablation, all other marks are closely associated with your desirable
products.

V
flammable corrosive explosive

compressed oxidizing toxic

gas

health harmful/ dangerousfor

hazard irritant the environment

Figure 50: Labels for different kinds of chemicals.

The danger of corrosive substances (e.g., acids, alkalis, organic halides) towards
other substances (e.g., metals, organic, living tissue) upon contact should
be kept in mind during your pH adjustment of the liquid for laser ablation
or container washing. If there is little chance to form corrosive substances
after laser ablation, even if you are not sure, please label them using this tag.
Oxidizing materials, like hydrogen peroxides (e.g., H2O2), that can supply oxygen
or other oxidizing substances, or react chemically to oxidize combustible
materials to increase the risk of a fire or explosion should be used very carefully.
Attention to acute toxicity during synthesis of arsenic-based compounds should
be paid since rapid occurrence of adverse effects often takes place within
14 days. “Flammable chemicals” means that they are very easy to catch fire.
The most frequent flammable chemicals are organic solvents you use for

148
 STAY ALIVE: LASER SAFETY

~~~~~~~~~~~~~~~~~~~~~~~
Since I am an engineer I only had a limit ed chemical background. Before I suffered
from an accident I hardly knew any thing about chemical and nanoparticle toxicit y.
One day, with the aim to do some experiments with cobalt oxide powders, I had to first
mechanically press them into a pellet. But unfortunately, before I started experiments,
I did not look up the toxicity information so that I forgot to wear gloves (That is a bad
habit for me! May be also for you!). I touched the powders with my hand. Even though I
washed my hand af terwards, I still had a severe skin allergy at that night (see Figure 51),
I did not know the reason. The firs t thing that came to mind is that there should be
something wrong with the food. But that day I jus t at e a döner which I have at e t ens
of times before. One hour lat er , it came to my mind that I had done some experiment s
with cobalt oxide. Then I quickly searched from the int ernet and found skin touching
would cause skin allergy. From this event, I learned that some mat erials are really
dangerous. Something happened lat er taught me that even laser ablation in liquid is
one of the pures t synthesis method but not totally safe for your health. Af t er pressing
these cobalt oxide palat es, I did some experiment s with gloves. But at that night, I
s till had a skin allergy and much severe than las t time which should be relat ed to
V
the wat er evaporation during long period of laser ablation, containing cobalt oxide
or cobalt ions inside. They were airborne and were inhaled inside my body during my
breath. When I went to the hospital , even the doctor did not know the toxicit y of cobalt
oxide. You see , I became a knowledgeable person from my accident. If you want to try
something else and t es t the abilit y of the doctors in your country, you can do your own
experiment s carelessly like me. But I can not promise you are so lucky enough to jus t
have the symptom of skin allergy.

Dongshi Zhang , Essen

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Figure 51: Skin allergy induced by exposure of

cobalt oxide colloids and powders.

particle generation by laser ablation, including ethanol, methanol, isopropanol,

etc. Carcinogenic colloids (cobalt-, nickel- and copper-based nanoparticles) or
solutions that can directly cause cancer are the most dangerous substances
towards human. Reproductive toxicity is related to your offsprings because it
influences your sexual function and fertility. Since oocyte cell of women can be
endangered by DMF’s exposure leading to birth defects, women should be very
careful with DMF or replace it with other solvents.

149
 STAY ALIVE: LASER SAFETY

STEP 5: DISPOSE OF YOUR WASTE

was t ewat er
treatment
plant

effluent

solids/
plant
biosolids
fish
surface animal
run-off

Nanoparticle contamination in soil , wat er and are

retaken by human through eating .

V Figure 52: Nanoparticle contamination in soil, water and are retaken by human through eating.

Not every laser shot is successful in generating nanoparticles. Hence, some of the
colloids become trash during your focus search or rechecking the reproducibility.
Then, how to dispose of these colloids? The easiest way that might come to your
mind is to dump them in the sewage, like you do with water after you boiled eggs
in it. Its only mg of nanoparticles in standard liquid, isn´t it? Yes, this is the most
convenient way, but it is not a wise choice. Some solids are biopersistent and are
taken up during plant growth, and then eaten by the animals. Once these animals
or fishes or vegetables are eaten by human (Figure 52), you can image that what
you have done to the environment has gone back to you. So do not throw the
colloidal trash away in the drain! Store them in the specific waste containers
for professional staff/company to collect. Organic and aqueous solutions are
better separately stored as well as the glass and plastic trashes. Meanwhile, the
colloidal solutions you want to keep for further characterization should best be
stored in a fridge away from heat and light at 2-25 °C.

FINAL STEP
After reading this book you hopefully realized that
the force is strong with you. Now, it’s time for you
to take your laser, spread your wings, dive into the
liquid and start synthesizing nanoparticles that
will make a difference in science and technology.

150
Photography and illustrations credits: Angel and gold crystal illustrations by Sasha Ediger,
water background image by BravissimoS/Shutterstock.com, internal doodle graphics by
Sasha Ediger and Nikolaeva/Shutterstock.com
Download:

DOI: 10.17185/duepublico/41087

S. Barcikowski, V. Amendola, G. Marzun, C. Rehbock, S. Reichenberger, D. Zhang,

and B. Gökce. Handbook of Laser Synthesis of Colloids, DuEPublico, 2016.