Reprinted from Chemistry in Britain, Volume 17, Number 5, May 1 981

Crysta0 growñng
Peter G Jones

In recent years structure determination by x-ray crystallography has crystals for method 1. The seed crystals
should be drawn off in a pipette together
become a commonly used technique in chemistry, largely due to the with some mother liquor—lf allowed
development of automatic diffractometers, powerful computers and to dry,
seed crystals lose
elegant programs. More and more chemists are thus encountering
the technique and need to familiarise thcmselves with some of its their
funda-mentals. Since most institutions have specialist effectiveness—and run into the warm
saturated solution . T his method has been
crystallographers to deal with the technical problems of data used to prepare very large crystals (se-
veral mm3) for neutron diffraction; as a
collection, structure solution and refinement, the essential róle of the general rule, the fewer the seeds, the
chemist is often restricted to the provision of suitable crystals. bigger the resulting crystals.

Crystal quality
Although many compounds emerge from The quality of crystals cannot be instantly
preparative procedures in 'crystalline' recognised in situ; exar•mnation with a
form, these crystals are rarely suitable ca. 5mm
for structure determination. The microscope is necessary. Remove some of
following special procedures are often the crystals from their mother liquor and
used to prepare good single crystals. place them on a slide. Do not remove all of
1. Slow cooling of saturated solu-tions. them, even if the compound is known to be
The flask and solvent to be used should be air stable, because crystals often
dust-free. The flask of warm solution is decompose by Ioss of solvent of
allowed to cool in a Dewar of warm water
over a period of several days. T his crystallisation once out of contact with the
classical method often gives results when mother liquor.
method 2 is inapplicable (eg when the A good single crystal Will normally have
compound is soluble in most liquids). weil-formed, Clear faces; a crazed or
2. Liquid diffusion. T his is normally • fibrous appearance IS an indication of
the most successful method of obtaining (v) poor quality. 'f crystals have grown an
single crystals (see Fig. I). A small (i)
amount of solution is placed in the tube clumps, single crystals may often be
and a sultable precipitant is layered care- Fig. l. The liquid diffusion method. (1) separated with a needle, or cut off with a
fu(ly down the side of the tube on to the Solution of compound to be crystallised. scalpel or razor blade. To prevent
solution. This second liquid should be crystals from flying off and becoming
less dense than, and miscible with, the (itHiiô Slowly layer on the precipitant. (i")
solvent. The tube is then corked and left Stopper and leave overnight: Do not Iost. they can be anchored to the slide
to stand undisturbed for 24 hours. As the disturbe ( v) Crystals form at the interface.
with a trace, of silicone grease.
precipitant diffuses into the solution,
crystals form at the interface. A suitable the same as the previous method, but the If a microscope with a polarising
precipitant is ailowed to diffuse into the
volume ratio solvent/precipitant is 1 : 4 attachment is available, the
solution from the vapour phase (see Fig. identification
or 1 2). Advantages of this method are that
: 5: the solution should be suf- of good crystats is made
several tubes may be set up in the same much easier.
ficiently concentrated to form an im-
mediate turbidity at the interface. If no container (a sealed desiccator When a good single crystal is viewed
crystals have formed in 24 hours, try ofien between crossed polars (je black back-
Used) and that very small quantities may ground) it Will probably show up bnghtly;
again
be used (on a spot plate)- Crystal form-
with ation is somewhat slower (several days). Fig. 2. The vapour diffusion method. (i)
a Solution of compound to be crystallised.
stronger 4. Convection methods. A simple ap-
(ii) Introduce precipitant and stopper
solution. paratus invented by Hope' is based on
Commonly continuous convection (Fig. 3). firmly. Crystals form after some
used 5. Diffusion of reacting solutions.
solvent/prec ipitant
pairs are water/acetone; chloroform or Compounds arislng as precipitates by
methylene chloride/petroleum ether or reaction of two solutions may often be 2
cyclohexane; obtained in crystalline form by layering one
and reacting solution on the other. If both
have
had reactants are in the same solvent, the
considerable layering may be facilitated by adding small
success amounts of other liquids to adjust the
with
formic densities. A 'buffer layer' of pure solvent or
acid/diisopropyl ether. T his method is a fine porosity frit may be used to prevent
also suitable when only small quantities excess immediate precipita-
of material are available (about 0.1 ml
solution is quite adequate). tion. This method is slower than 2
3. Vapour diffusion. The principle is because reaction must be complete
before the mixture is disturbed
6. Use of seed crystals. If Other meth-
P. G. Jones is at the Anorgan•sch- ods provide crystals which, although of
Chemrsches Institut der Universitát, reasonable quality•, are too small. then
Tammannstr. 4, 3400 Gôttingen, Federal these may profitably be used as seed
Republic of Germany.
222
 nature of the compound IS unknown, any
information (eg what elements are pre-
sent, whether the compound is optically
active) Will help the crystallographer.
2. To obtain accurate interatomic dist-
ances and angles. For normal organic
structures with no elements heavier than
chlorine, bond lengths and angles Will
probably have an accuracy of 0.005 A,
0.59 or slightly better. However, since x-
rays are diffracted by electrons, any
atoms appreciably lighter than the rest
Will be less accurately located —if at alll
Thus, hydrogen atom positions Will be
relatively Inaccurate in organic structures
(C —H bond lengths to accuracy 0.02 A
or less). In the presence of heavy ele-
ments (eg the heavier transition ele-
ments) first row elements Will be similarly
affected and locating hydrogen atoms is
usualty impossible.
3. To determine the absolute configur-
ation of a chira/ compound. T his
a less
common, but Still important requirement-
The determinauon of absolute configur-
ation depends on the measurement of
small intensity differences between re-
lated pairs of reflections; these differe nces
are greater (and the determination thus
A typical organic crystal structurc determinadon yields this picture of the natural
product loganin. Carbon and oxygen atoms are represented as thermal ellipsoids of 50 per facilitated) (1) when heavy elements are
cent probability (ie there is a 50 per cent chance of finding an atom within its ellipsoid present and (ii) when longer wavelength
during its thermal vibrations). Hydrogen atoms are plotted as spheres of constant radius. radiation is used. With copper radiation
(1.54 A) the absolute configuration may,
on rotation of the slide, all pans of the by extremes of shape such as plates or with luck, be determined when only first
crystal should extinguish the light needles, and these shapes are therefore row elements are present. Modern prac-
sharply and simultaneously. (Cubic and
best avoided (reflected beams Will be tice is however to use molybdenum radi-
some other high symmetr,f crystals are
rare exceptions to this behaviour.) If one attenuated much more when passing ation (0.71 A). which is absorbed less; in
pan of the crystal is bright and another along the long axis of a needle than when this case bromine or heavier elements Will
perpendicular to it). Modern programs

pan is dark it is probably not a sengle probably be needed.

are. however, capable of applying effec-
crystal. The crystal must of course be
tive absorption corrections in most cases.
transparent for this procedure. Finally, when you have obtained Choice of compound
Crystal size and shape are important.

Equidimensional reasonable-looking crystals, take them To some extent the chemist may have
crystals are best, several possibilities open to him in
(at least some Still in the mother liquor) to choosing a compound for x-ray analysis,
though plates or needles may often be
your crystallographer. eg different side chain substituentsa
He can then
used. The optimum size and shape are Fig. 3. Apparatus for crystal growth by
decide whether to mount them convection. The substance of which crys-

with tals are required is placed at the bottom

determined by the type of atoms present.
mother liquor in a capillary; this is not left. A and B are heating coils and C is a
If only light atoms are present, eg typical usually done unless necessary, as the cooling coil. Crystals form near C and
organic structures containing carbon. carried to the bend (bottom right) are
capillary scatters x-rays, lead ing to a high
nitrogen, oxygen and perhaps a few by
background level and less precise convection currents.
phosphorus, sulphur or chlorine atoms,
measurements. Since the waiting tist to
the crystal mav be of medium to large 3
use the diffractometer may be several
size, in crystallographic terms, about
months your compound, even if 'air
0.3 to 0.7 mm in the longest direction.
stable', may decompose over this period.
( Few diffractometers can provide a
uniform x-ray beam larger than 0.8 I have suffered severa' disappointments c
mm

due to sublimed or hydrolysed crystals? If

diameter. and it is essential that the
your crystals are at all likely to misbehave.
whole crystal always remains in the
beam.) If heavier atoms are presenta warn the crystallographer.
the

crystal Will absorb x-rays to an ap-

Why do you want the structure?

preciable extent; large crystals cannot
Before requesting an x-ray structure it is
then be used because few x-rays Will get as well to ask yourself exactly What you
through! By way of compensation, how-

want to know. There are three main

ever, such materials usually diffract x- reasons for wanting a structure.
rays more strongly. so smaller crystals are 1. To identify an unknown compound.

quite usable. A suitable size would be 0.1 Although crystallography is becoming a

to 0.3 mm maximum, depending on the

routine analytical technaque, it is Still

'heaviness' of the atoms present—a crys- tame- and resource-consumlng. It is
tal as small as 0.03 mm may be therefore best to exhaust other. more
adequate if it is very strongly diffracting, conventional, approaches (ir, nmr. ele-
The effects of absorption are enhanced

mental analysis) first. Even if the exact

224
counter-ions or solvates. A sensible CH2CIa which form no strong secondary
choice may make both his and the interactions to other molecules. It is usu-
crystallographer's work much easier. ally best to get crystals first and worry
The following guidelines should about this later.
therefore be observed if at all possible. 4. Fifteen years ago the solution of an
1. Try to reduce the possibilities of dis-order, organic crystal structure normally de-
X-ray analysis assumes that the contents of pended on the incorporation of a heavy
each unit cell are identical. If groups are free to atom (often iodine or bromine) into the
rotate, or can occupy different positions in structure. Vastly improved structure
different cells, this assumption is only true in solv-ing methods now render this unnec-
essary, and It can in fact often be a
an average sense and structure refinement disadvantage; the precision of light atom
location
can is
become difficult. This reduced,
phenomenon is and
organic
known as disorder. Relatively small, bromides and iodides often decompose
highly symmetrical ions (eg BF,-, CIO- when irradiated.
S03F-) are unlikely to partake in strong 5. Make sure the compound is the r•ght
secondary interactions such as one! Simple checks may be possible. eg
hydrogen bonds and are thus often if the compound is meant to contain
dis-ordered; The problems engendered by the use of highly phos-phorus
bulky symmetrical, small ions. T his is an
spectra. and fluorine, check its nmr
ions (eg
SbFe- ion, which shows a virtually
Your friendly local crystallograp-her Will
Ph4As not be amused if time is wasted
spherical distribution of its F atoms. It has on the wrong structure, and this happens
(Pha been refined as if it were occupying two
BPhÃ, picrate) are less likely different sites, but the large size of the F more frequently than one might imagine.
to rotate freely and their use as counter- ellipsoids shows clearly that the dis-order 6. Check that the structure has not
ions is strongly recommended. Similarly, must bc more severe. (I am grateful to Dr been done before! The Cambridge
torsionally symmetric groups such as Crystallographic Data Centre reviews
Gabriela Weber for this example). yearly all structures containing at least
3. CFa, P03 (especially when bonded one C—H bond (Molecular structures and
to spa atoms and therefore experiencing practical difficulties and always requires dimensions); inorganic structu res are
no torsional energy barrier) and long frequent attention. Low melting point presented in the BIDICS series of publi-
side chains (as in sterols cations
and many other natural products) are solids, if not cooled, often exhibit some from
often disordered. The most recommen- form of disorder due to their high McMaster
ded substituent is phenyl; this is rarely thermal motion . University,
disordered and has the additional ad- 3. Thechoice of solvent is often dictated Canada. Both should be in most un-
vantage that phenyl derivatives often iversity chemical libraries.
crystallise more easily. by What Will give the best crystals.
2. Avoid low melting point solids or Solvent of crystallisation may, however,
often be disordered, especially for sol- Reference
liquids. Low temperature x-ray work, 1. Hope, J App'. Cryst, 1971. 4. 333.
although possible, is often beset with vents such as benzene, CHCla, and