KF-Competence_Guide_LR
KF-Competence_Guide_LR
Titration Excellence
KFTitration
Compact Titrators
Excellence
KF Compact Titrators
products more difficult, or lead to quality problems with plastic products. The determination of the water content
in different materials is therefore crucial in production processes and quality control.
Chemical methods for the water content determination include techniques where a chemical reaction takes place
between water and a suitable reagent. The amount of water converted during the reaction is determined from the
amount of reagent consumed. The most well-known method based on this principle is the Karl Fischer titration.
The main advantages of this quantitative analytical technique are its
METTLER TOLEDO offers you a complete portfolio of solutions for successful water content determination by Karl
Fischer titration, including autosamplers and dedicated accessories for the unattended analysis of a high
number of samples. Detailed information can be found at Volumetric and Coulometric Karl Fischer Titrators – For
Accurate and Precise Water Determinations (mt.com).
This Competence Guide covers the most relevant points to be considered when performing a Karl Fischer titration
- without starting from basics. Beside a brief overview, this Guide will focus on three parts:
We are convinced that this document is a valuable support for your work on water content determination.
Titration is a straightforward, easy and affordable technique for the content determination of a specific analyte
in sample solutions. If the analyte is water, then the titration according to Karl Fischer is the method of choice.
Briefly, the water content is determined by addition of an iodine I2 solution of known concentration to the sample
dissolved in an alcoholic solution e.g. in methanol. Iodine reacts with water in a 1:1 ratio in a two-step reaction
involving methanol, an organic base and sulfur dioxide (SO2). This chemical reaction is designated according to
the German Chemist Karl Fischer who developed this analytical technique for the determination of the water
content in petroleum products i.e. the so-called Karl Fischer (KF) reaction.
During titrant addition, the potential difference is monitored with a polarized double pin platinum potentiometric
sensor (a redox electrode), such that the process can be visualized in a graph showing a titration curve. The
iodine titrant solution is added until a predefined potential endpoint is reached i.e. until a potential of 100 mV is
reached. The titrant consumption in mL up to this point allows for calculation of the water content in the sample
(Volumetric KF titration). Alternatively, the iodine needed for the KF reaction can be generated directly in the
titration vessel by electrochemical processes (Coulometric KF titration).
Figure 1: Titration curves showing the determination of water in cellulose by volumetric KF titration. 0.15 g cellulose was added directly
in the vessel of a volumetric KF titrator, stirred during 120 s to achieve complete dissolution, and subsequently titrated. The potential
(blue curve) and the titrant volume (red curve) are plotted versus time.
Two determination techniques are used in KF titration depending on the water content of the sample:
The techniques differ in the way they provide iodine for the Karl Fischer reaction with water. Based on the
expected water content, the most suitable instrument i.e. a coulometric or a volumetric KF titrator is selected:
Figure 2A: Coulometric (left, Coulometric KF Titration - Tutorial - YouTube) and volumetric (right, 13. Run titration - Karl Fischer Volumetric
Titrator - YouTube) KF titrators. Typical samples are indicated below each instrument.
When measuring several aliquots of the same sample in a sample series, the repeatability – expressed as
relative standard deviation Srel in % - is changing depending on the water content, as shown in figure 2 B. In
particular, decreasing water contents lead to poorer repeatability i.e. the relative standard deviation is increasing:
Figure 2B: Measurement range for coulometric and volumetric KF titration. The repeatability (rel. std. deviation Srel, %) is also given as a
function of the water content.
Select the suitable sample size for accurate and precise measurement i.e. small deviation from expected value
and low relative standard deviation:
Competence Guide
Coulometric KF titration
Volumetric KF titration
Obtaining accurate results requires extracting all water from the sample. Many liquid samples dissolve easily in
the classical KF solvent methanol, and thus the titration can be performed after direct injection of the liquid
Competence Guide
sample in the vessel. On the other hand, viscous oils, pastes or solid samples requires additional preparation
steps to release water from the sample matrix. These steps can include the use of auxiliary solvents, increased
dispersion e.g. longer (stirring times) or even higher temperatures to extract water completely:
Mixing
Many liquid or solid samples dissolve in methanol or in methanol
mixtures using a normal magnetic stirring bar:
• Methanol: liquid foods, cosmetics, organic solids
• Chloroform: petroleum products, fats, organic solids
• Long-chain alcohols (e.g. octanol): edible oils, fuels, creams
• Toluene/xylene: tars, waxes, crude oil
• Formamide: sugars, starches, solid foods, pharmaceuticals
Dispersion
Thicker, viscous liquids, pastes or powders may require more intense
dispersion with a high-speed homogenizer to fully dissolve them.
Warming
A thermostatable titration vessel allows warming up to 50°C with an
external silicon oil or water bath to improve dissolution and extraction
of water from the sample.
KF Oven
For solids or liquids causing side reactions or releasing water very
slowly, a drying oven heats samples up to 280°C to drive water out of
the sample and into the titration beaker.
Easily soluble samples can be titrated by injection (addition) directly into the titration cell:
Competence Guide
Ketones (e.g. acetone) and aldehydes react with methanol to form water. To suppress this undesired side
reaction, dedicated KF reagents are needed when dealing with these samples.
Diesel oil, petroleum, edible oil, ethereal oils, ointments, butter, fats, etc.
Solvents for one-component KF volumetric reagents Solvents for two component KF volumetric reagents
Honeywell HYDRANAL Methanol Dry + max. 50% v/v decanol HYDRANAL Solvent Oil (with chloroform)
or hexanol
HYDRANAL LipoSolver CM (with chloroform) Solvent CM (with chloroform)
LipoSolver MH (with hexanol) Solver (Crude) Oil (xylene, chloroform)
Merck Aquastar CombiSolvent Fats (butyl acetate, Aquastar Solvent Oils & Fats (hexanol, 2-propanol)
decanol)
CombiSolvent Oils (with toluene)
Mineral oils are very difficult to be completely dissolved. Auxiliary solvents (chloroform, toluene, xylene and
mixtures thereof) added to methanol or dedicated KF solvents can cope with these challenging samples:
Solvents for one-component KF volumetric reagents Solvents for two component KF volumetric reagents
Honeywell HYDRANAL Methanol Dry + up to 70% v/v HYDRANAL Solvent + up to 70% v/v chloroform
chloroform
Solver (Crude) Oil (xylene, chloroform) Solvent CM (with chloroform)
Solver (Crude) Oil (xylene, chloroform)
Merck Aquastar CombiMethanol + up to 70% v/v Aquastar Solvent + up to 70% v/v chloroform
chloroform
CombiSolvent Oils (with toluene) CombiSolvent Oils (with toluene)
Products containing sugar, starch or gelatin also need the use of formamide as auxiliary solvent to be
completely dissolved in the titration cell:
Solvents for one-component KF volumetric reagents Solvents for two component KF volumetric reagents
Honeywell HYDRANAL Methanol Dry + up to max. 50% v/v HYDRANAL Solvent + up to max. 50% v/v formamide
formamide
Merck Aquastar CombiMethanol + up to max. 50% v/v Aquastar Solvent + up to max. 50% v/v formamide
formamide
Water determination in liquid samples is generally straightforward and does not involve any challenges other
than improving the sample solubility to achieve a complete release of water into the KF solvent. The situation is
different when the water content has to be determined in solid samples.
In fact, water molecules present in solid samples are bound to the sample matrix in different ways i.e. water
molecules can be
Hence, water must be first completely released from solid samples to be available for the Karl Fischer titration:
The Karl Fischer reaction consists of two steps: first, sulfur dioxide SO2 reacts with a base RN and methanol
CH3OH to a methyl sulfite, (RNH)SO3CH3. Subsequently, methyl sulfite reacts with water H2O, iodine I2 and a
base giving methyl sulfate as a product.
The methods to efficiently release water from the sample matrix have been summarized in the following list:
Based on these methods, different sample preparation techniques have been developed:
• Direct titration
• Internal extraction
• External extraction
• External dissolution
• Gas-phase extraction
The choice of the suitable technique depends on the sample to be titrated. The different tehcniques will be
presented in the next chapters.
Diesel fuel, petroleum edible oils and fats, ethereal oils, ointments:
These samples are difficult to dissolve in the KF solvent. The solubility is improved by mixing decanol or hexanol
up to 50% v/v with methanol. Both alcohols have long non-polar chains, thus improving the dissolution.
pH value adjustment:
When the samples are acidic or basic, thee optimum pH for the Karl Fischer Titration i.e. between pH 5 and 7 is
achieved by adding buffers (imidazole, or benzoic acid) to the KF solvent before starting the titration.
Generator
electrode
Measuring electrode
(double platinium pin)
Diaphragm Anolyte
+
1 2 3 4
Figure 5: Back-weighing procedure
Summary :
Direct titration
Sample Soluble If required: Adjust pH value
No side reactions Improve solubility
- Auxiliary solvents
- Warming/heating
No specific sample preparation
The internal extraction technique -more precisely: the direct titration with internal extraction- is applied when the
water is extracted directly in the titration vessel by means of e.g.
Competence Guide
The use of a high-speed homogenizer allows for an efficient and quick dispersion of the sample. Consequently,
the water is released completely and fast from the sample matrix:
Figure. 6B/6C: Homogenizer before (left, B) and during action (right, C) in the titration vessel of the KF volumetric instrument.
Summary :
Internal extraction
Sample Insoluble e.g. almonds Extraction by Long stir time = 5-10 min
No side reactions e.g. instant coffee High–speed mixer
Fast water release e.g. tablets Improve solubility - Aux. solvents
This sample preparation technique is needed when the water is extracted from the sample in an external vessel
containing a solvent of known water content during a defined time e.g. 1-2 hours. Subsequently, an aliquot of the
Competence Guide
1 2 3 4
Blank Determination Weigh solvent and sample Extract water Settle and titration
Summary :
External extraction
Sample Not soluble e.g. tobacco Extraction by Suitable solvent or water content << sample
solvent mixture water content
Slow water release e.g. cheese Long extraction time 1-2 h and more
Very low water content Large sample size Large sample size for the
e.g. surface water of sugar extraction to get a large
sample water amount with
respect to the solvent water
content
No side reactions e.g. wool
Solid sample – Direct coulometric NEVER enter solid sample Generator cell will be
titration not possible e.g. plastic pellets in a coulometric KF titration damaged!
vessel
Measuring electrode
(double platinium pin)
Diaphragm Anolyte
+
-
Competence Guide
where
x Water content (ppm) of titrated sample/aliquot ppm
f1 Water content of extraction solvent (blank) ppm
f2 Weight of extraction solvent used grams
f3 Total weight of sample used for extraction grams
Similarly, the calculation for the water content in % of the extracted sample is defined in the following formula:
where
x Water content of titrated sample/aliquot %
f1 Water content of extraction solvent (blank) %
f2 Weight of extraction solvent used grams
f3 Total weight of sample used for extraction grams
Lyophilized substances (or freeze-dried substances) have a very low water content in the order of e.g.
10-50 ppm. For this reason, they are very hygroscopic when exposed to air, i.e. they easily absorb moisture
Competence Guide
To avoid this, the extraction of water from the lyophilized sample is performed directly into the sealed vial by
adding dried coulometric anolyte during pretitration:
↓
↓ ↓
Generator
electrode
Measuring electrode
(double platinium pin)
Diaphragm Anolyte
+
In the case when the water content is extremely low, a much too large sample mass (e.g. 30-40 g) has to be
used to achieve accurate and precise results. This will quickly fill up the titration vessel, and not more than
Competence Guide
1-2 samples could be titrated. Thus, the KF solvent must be often replaced. In the worst case, it has to be
replaced even after one single KF titration since the vessel is too full to accommodate other samples.
A large amount of the sample is dissolved in a flask containing a suitable solvent. Subsequently, an aliquot of
the extraction solvent is titrated by injection in the KF titration vessel. The whole procedure is similar to the
external extraction, however the calculations are slightly different:
)–(
where
x Water content of titrated sample/aliquot %, or ppm
f1 Water content of extraction solvent (blank) %, or ppm
f2 Weight of extraction solvent used grams
f3 Total weight of sample used for dissolution grams
As in the case for the external extraction, to achieve accurate and precise results it is necessary that
Summary :
External dissolution
Sample Soluble e.g. phenol Dissolution Suitable solvent or water content << sample
solvent mixture water content
Non-homogeneous water distribution Long extraction time 1-2 h and more
e.g. sucrose
Very low water content Large sample size large water amount
e.g. petroleum products
No side reactions e.g. phenol
Solid sample e.g. sugar Direct coulometric titration not possible
The specific calculations for both sample preparations are built-in in the high-end KF titrators to improve
easy-of-use and reduce errors when performing these techniques prior to the KF titration.
Competence Guide
In fact, these KF instruments have preprogrammed standard method templates as well as tested titration
methods to perform automatically the whole procedure including the calculations. Depending on the analysis to
be performed, the corresponding template is selected:
From the list of templates the method KFC002 is selected to develop a KF titration with external extraction (or
dissolution)
In the method function “Calculation” you can select the corresponding procedure sample preparation procedure,
as well as and the desired units i.e. “ppm” or “%”
2.
1. Under the function “Sample”, select “Blank” to first
determine the water content (blank) of the 1.
extraction (dissolution) solvent.
2. Subsequently, the sample is determined
(select “sample”)
Briefly, the operator is guided through all steps of the analysis. In the case of an external extraction followed by
coulometric Karl Fischer Titration, the detailed workflow is given in the next page.
↓
2. Start sample titration
When the sample undergoes side reactions with the KF reagents, water has first to be released by heating the
sample in an oven with subsequent transfer into the titration vessel. The KF oven is connected to the KF titrator
Competence Guide
with a transfer tube with an inlet in the titration cell, and the evaporated moisture is transferred using a stream of
dry and inert gas such as nitrogen.
Figure 10: A continuous flow of dry gas enter the sample vial in the oven and transport the evaporated water into the KF titration vessel.
Summary :
Gas-phase extraction
Sample Side reactions with the KF reagents e.g. Heating and No contact with reagents
motor oil collecting the
Insoluble e.g. plastic chips vapor
Slow water release e.g. polymers Heating time Recommendation:
• At least 10 min
• max. 30 min
Thermally stable – no decomposition Heating temperature • Select heating
e.g. polylactic acid (sutures) temperature 10-20°C
below melting point
• Avoid oxidation by using
nitrogen as inert transfer
gas
To determine the water content of gases, the gas is purged through the titration vessel for a defined time period.
During this time, the KF solvent absorbs the gas humidity. First, the gas is passing through a flow meter to
Competence Guide
ensure a constant flow rate to achieve an accurate determination of the total gas volume. Subsequently, the gas
transfer is stopped, and the titration is started:.
Figure 11: The gas humidity is collected in the titration vessel during a specific time e.g. 5 min before titration.
The amount of water mwater determined by KF titration allows for the calculation of the water content:
1. The total gas volume V is determined by multiplying the gas flow rate (mL/min) with the time (min):
2. The water content is determined according to:
( )
Summary :
The air humidity is the big enemy of a Karl Fischer titration! Measures are implemented to reduce the influence of
air moisture on the KF reaction. More specifically, the penetration of air humidity in the titration vessel has to be
limited by checking the following points:
• Connections to openings of the KF titration vessel have to be greased i.e. tube inserts, electrode, adapters,
septum and generator cell (KF coulometer), as well as stoppers.
• Use dry and fresh molecular sieve and silica gel in the drying tubes to ensure an efficient absorption of
moisture from the air.
• KF coulometer: the ceramic frit of a generator cell with diaphragm has to be dry.
Explanation:
During pretitration, the titrator add (generates) iodine to react with water present in the KF solvent. When the KF sol-
vent is dry, the titrator switches from pretitration in the stand-by mode. At this point, the online drift is still too high to
start a sample analysis. In general, it takes additionally 20-30 min to get a low and stable drift value of 10 μg/min.
To decrease the drift value it is also important to use fresh drying agents in the drying tubes. We recommend
exchanging molecular sieves every 6 weeks. The molecular sieves is regenerated by drying it in the oven
overnight at e.g. 250-300°C (check product data sheet)
Furthermore, if you use a cell with diaphragm, wash the diaphragm with dry methanol and store it for at least one
night in an oven at max. 80°C, or at a well-ventilated place to evaporate water trapped inside the ceramic diaphragm.
Solution:
1. Work in suitable working spaces (low humidity, room temperature, no strong ventilation)
2. Check tightness of the system
3. Use fresh KF reagents, solvents and molecular sieve
4. Regularly renew and dry exhausted molecular sieve
5. Gently shake the titration vessel to collect traces of water from the glass walls
6. Coulometric generator cell: Dry it in the oven if the ceramic frit is too wet (max. 80°C)
What to do If after 20-30 min the online drift is still too high e.g. 20-30 μg/min
Check again points 1.-6. (1.-7.)
A KF Compact Volumetric Titrator can determine the smallest amount (LOD) of 62.5 μg water in a 5 g sample,
using a 5 mL burette with a 5 mg/mL KF titrant. This corresponds to a water content of 12.5 ppm under optimal
conditions.
A KF Compact Coulometer can determine the smallest amount of 10 μg water. This corresponds to a water
content of 1 ppm for a 10 g sample – under optimal working conditions.
Explanation:
The highest dosing accuracy is reached when the titrant consumption is within 10-90% of the absolute burette volume.
Hence, the most accurate results will be obtained using half of the burette volume for one titration. For instance,
best results for a sample with an expected water content of 5% are achieved if you use a sample size of 0.25 g
in combination with a 5 mL burette and a 5 mg/mL KF one-component titrant.
If the water content is very low e.g. 0.1%, a much bigger sample size is required to obtain best results, or the
sample has to be titrated by a KF Coulometer.
Indeed, it is always necessary to determine the concentration of the Karl Fischer titrant since the value printed on
the label is a nominal value, and not the current concentration. Thus, if one calculate the water content using the
Competence Guide
nominal titrant concentration after a sample determination, the result will not be accurate. The determination
procedure is shown in a short film under Determine titrant concentration - Karl Fischer Titrator - YouTube.
Explanation:
The concentration of e.g. the one-component 5 KF titrant is always slightly higher than 5 mg/mL to allow for a
longer storage time of the bottle. Moreover, the titer is always changing with time even in a closed bottle due to
air moisture still penetrating into the bottle.
The concentration value indicated in the Certificate of Analysis is determined just after production of the reagent.
As soon as you open the bottle, air humidity reacts with it, thus decreasing its value. Hence, it is necessary to
perform a concentration determination value.
Note that the concentration determination using reference materials is a straightforward way to verify the correct
working of the whole titration instrument.
Solution:
Once the new KF titrant bottle has been connected to the titrator, a concentration determination is performed
using
To achieve optimum titration time, it is recommended to define tolerances for the concentration determination:
The use of liquid water standards is strongly recommended since they allow for seamless determination of the
KF titrant concentration.
Competence Guide
Two liquid water standards for the volumetric KF titration are commercially available with a concentration
10 H2O mg/g and 1 mg H2O/g. Ready-to-use titration methods developed for these standards are stored in the
KF titrator, hence no additional steps are needed.
Explanation:
Beside liquid water standards, also certified solid standards are available for the concentration determination of
the KF titrants. The reference material of choice is sodium tartrate dihydrate, an organic compound with an exact
water content of 15.66%. Briefly, water is tightly bound in the crystal structure of sodium tartrate. The dissolution
in the KF solvent allows for release of water from the structure, and thus water can be subsequently titrated.
However, sodium tartrate dihydrate is only partially soluble in the KF solvent, hence the solvent must be replaced
as soon as it becomes turbid when titrating several samples:
• Max. 0.15 g sodium tartrate in 40 mL KF methanolic solvent i.e. max. three samples of each 0.05 g can be
titrated in the same solvent.
• Max. 0.05 g sodium tartrate in 40 mL KF methanolic i.e. after one sample of 0.05 g sodium tartrate the
solvent should be replaced.
Hence, the best results and most convenient use will be obtained with the liquid water standards. In fact, liquid
standards are soluble in most solvents without any problem when compared to sodium tartrate. When using
liquid standards, you can even perform six determinations in the same solvent. Moreover, you get a lower
standard deviation and the analysis time is much faster since no long stirring time is needed as for a solid water
standard, which requires a stir time of 120-180 s prior titration.
It is also possible to perform the concentration determination using pure water. However, the accuracy is strongly
affected due to very small sample size used i.e. 10-30 μL. Thus, the influence of sample handling will be highly
increased. As an example, if you lose e.g. 5 μL of your 10 μL pure water sample, then there will be a huge error
of 50%. On the other hand, the sample size with liquid standard is in the order of 1000 μL, and therefore a loss
of 5 μL is not so relevant as in the case of pure water.
Solution:
1. Flush the whole system i.e. the burette and tubes three times, and discard the dispensed reagent using the
manual operations of the titrator. This will remove any bubbles in the tubes and also guarantees that the titer
in the dosing unit and also in the tubes will be the same.
2. Use the 10 and/or 1 mg/g liquid water standard for the concentration determination.
3. If the ambient conditions are changing during the day, e.g. due to a humidity increase and/or a temperature
increase by more than 3-5°C, the concentration determination must be repeated.
Certified liquid water standards consist of an exact amount of water dissolved in a solvent mixture. The composi-
tion of the water standards has been optimized to reduce the absorption of air humidity. In any case, Good Titration
Competence Guide
Practice (GTP) strongly recommends not to leave the ampoules open with liquid standard inside. The liquid stan-
dard should be completely aspired and transferred from the ampoule in a gas-tight syringe (e.g. a glass syringe).
Explanation:
Every solvent that is exposed to air may absorb humidity until saturation is reached i.e. until chemical
equilibrium between the liquid and the gaseous phase is achieved. The amount of water absorbed from the air in
the liquid standard mainly depends on the
Solvents such as
easily absorb water from the air due to the presence of oxygen at-
oms in their chemical structure leading to electrical polarity. The
water molecule H2O is strongly polar, where the oxygen atom O
is negatively charged and the two hydrogen atoms H are positively
polarized. Thus, water is easily absorbed in a polar solvent because
polar substances tend to mix up.
Solution:
The Karl Fischer titrant is very sensitive to air humidity; thus, its concentration must be determined every time
before determining the water content of a sample by volumetric Karl Fischer titration. The procedure of the
Competence Guide
concentration determination for KF titrants is shown in a short film under Determine titrant concentration -
Karl Fischer Titrator - YouTube.
Explanation:
The determination of the titrant concentration is a crucial step in titration workflow. Without knowing the exact
concentration it is not possible to achieve correct results. Thus, the determination must be performed at least
every time before starting a series of measurements. Moreover, KF titrants are sensitive to air humidity and their
concentration slowly decreases with time – even if the titrant bottle is tightly closed. It is estimated that the
decrease in concentration of the KF titrant is in the order of -0.1 mg/mL per week.
Therefore, depending on the ambient conditions, the concentration of the KF titration have to be determined more
frequently. For instance, if the temperature and humidity are changing a lot during the day, it is necessary to
repeat the determination even during the daily routine.
It is not necessary to determine the concentration of the coulometric KF reagent since coulometry is an absolute
method because the amount of current used to generate iodine by electrochemical processes is directly
Competence Guide
measured by the electronics in the titrator. Thus, the generated current does not have to be compared to any
reference material.
Explanation:
The amount of electrical charges (Coulomb, C) needed for the iodine generation is used in the calculation of the
water content through the Faraday constant (96485 C/mol). It is required that the generator and electronics in
the instrument are correctly working to achieve accurate results.
Anode +
Generator
electrode
2H+ H2
Catholyte
Diaphragm Anolyte
I I I I I2 I2
• 96485 C are required to produce 1mol of a chemical compound with 1 mol of electrons.
• Iodine I2 is generated from two iodide ions I- by oxidation at the anode of the generator cell.
• 1 mol I2 reacts with 1mol H2O, i.e. 1 mole H2O (18 g H2O) is equivalent to 2 x 96485 C.
This corresponds to 10.712 C/ mg H2O (Assumption: 100% electrochemical conversion).
It is recommended to periodically verify the performance of the KF coulometer. This is achieved by checking the
recovery (%) of certified liquid water standards 1 and 0.1 mg H2O/g.
In general, if the results are outside the tolerances stated in the certificate of analysis (CoA) of the water
standards or they differ too much from the certified value, there is nothing left but to systematically check the
Competence Guide
whole titration system step-by-step to limit or suppress the influence of all possible sources of uncertainty.
Additionally, most errors come from handling problems, especially when working with the lowest water standard
with 0.1 mg/g (100 ppm).
Explanation:
The measurement uncertainty indicated on the CoA of the water standard e.g. 0.99% ± 0.02% only indicates the
uncertainty of the specific water standard measured just after production at the production site of the provider of
liquid standards. Thus, it indicates the calibration uncertainty of the water standard obtained at production plant
of the provider.
This means that the uncertainty of the CoA refers only to the conditions of content determination at the provider’s
site, i.e. to the analytical methods, instruments procedures, standard operating procedures SOPs, ambient
conditions etc. used at provider’s site. The uncertainty value is only valid under the same working conditions and
instruments at the provider’s site. However, this is never the case, and thus it is not related to the result
uncertainties of measurements. Therefore, it is not correct to state that a value is wrong if it outside the
tolerances of the CoA.
To assess if a measurement is acceptable or not, note that the providers of reference materials give
recommendations, as given below:
Figure 18: Certificates of analysis (CoAs) indicating both a) the calibration uncertainty of the water standard determined after production at
provider’s site, as well as b) recommendations for the assessment of tolerances for the determination of water by KF titration.
Most errors come from handling problems, especially with the low water standards that have only 100 ppm. It
is recommended to use airtight glass syringes with long needles since glass is not absorbing air moisture as
plastic:
Suggestion:
Use the limits recommended by the providers of the water standards as highlighted in yellow at the end of the
certificates in figure 18, and also of additional documents related to the standards.
Figure 19B: Warning message on the titrator display if the result is outside the limits.
METTLER TOLEDO Competence Guide 29
What do you do if the result is not correct?
It is expected that the titrator has been adjusted and its performance verified with water standards. Thus, the
most important point to be checked is related with the applied sample preparation i.e. if the sample preparation
Competence Guide
has achieved a complete release of water from the sample matrix. In other words: Has all water been released
from the sample? Is it completely dissolved in the KF solvent?
Explanation:
For a successful water content determination by the Karl Fischer titration, the sample needs to be fully dissolved
or dispersed to completely release water. Moreover, no side reactions should take place, and the correct pH for
the KF reaction should be between 5.5 and 8.
For example, the determination of water content in lubricants requires the use of a KF oven to evaporate water,
since the additives present in these samples lead to side reactions that affect the result. Thus, a direct titration is
not the method of choice. In fact, the direct injection of oil samples containing additives such as thiols R-SH lead
to higher results because the thiol group R-SH reacts with iodine giving an erroneous higher consumption.
Note that there are also some oils and lubricants without additives, therefore they can be titrated by direct
injection in a mixture consisting of the KF solvent and auxiliary solvents such as e.g. chloroform facilitating the
dissolution of the oil samples.
Figure 20: Water is bound to the sample matrix and it has first to be released before starting the KF titration.
Solution:
In general, deviations from the expected results are due to incomplete release of water from the sample.
Depending on the sample nature, there are different preparation techniques for the complete release of water as
well as for the handling of the sample:
1. Direct titration
2. Internal extraction (inside titration cell)
3. External extraction
4. External dissolution
5. Gas-phase extraction
6. Water content in gases
In general, the stand-by (or conditioning) of the KF titration cell should take no longer than 20- 30 minutes to
reach low online drift values. This can be achieved if the KF coulometric electrolyte or the KF volumetric titrant
Competence Guide
and solvent used are fresh and the titrator is well tight, i.e. all apertures and connections of the cell are slightly
greased, correctly assembled and tighten.
To decrease the online drift value it is also important to use fresh drying agents in the drying tubes. We
recommend exchanging molecular sieves every 6 weeks. Furthermore, if you use a cell with diaphragm, wash
the porous ceramic diaphragm with dry methanol and store it at least for one night in an oven at max. 80°C, or
at a well-ventilated place to evaporate water trapped inside the ceramic diaphragm.
Explanation:
After assembling and filling the titration cell, the pretitration is started to eliminate the water present in the KF
solvent or the coulometric electrolyte. When the solvent is dried, the titrator switches from pretitration in the
stand-by mode. However, at this point the online drift is generally still too high to start a sample analysis.
During stand-by, the online-drift is slowly but continuously decreased since the titrator automatically reacts to
any potential change caused by the presence of humidity e.g. from air. This step may require 20-30 minutes in
general.
To speed up the process, it is beneficial to gently shake the titration vessel in order to efficiently collect traces of
water adsorbed on the glass wall of the titration cell.
Low drift values are necessary to achieve accurate and precise results. The following table gives indicative
values that can be reached with a KF titrator:
Solution:
1. Work in suitable working spaces (low humidity, room temperature, no strong ventilation)
2. Check tightness of the system
3. Use fresh KF reagents, solvents and molecular sieve
4. Regularly renew and/or dry exhausted molecular sieve
5. Gently shake the titration vessel to collect traces of water from the glass walls
6. Coulometric generator cell: Dry it again in the oven
What to do If after 20-30 min the online drift is still too high e.g. 20-30 μg/min
Check again points 1.-6. (1.-7.)
A too high drift value e.g. 60 μg/min after sample titration can be attributed either to
Competence Guide
Thus, either special KF reagents have to be used, or the sample must be heated with the KF oven, or the sample
preparation has to be improved to achieve the complete release of water.
Explanation:
Side reactions are caused by interfering chemical reactions between the sample and the KF reagents. Most of
these side reactions can be almost completely suppressed by using special KF reagents. For instance, in the
case of samples containing ketone groups e.g. R1-CO-R2, the use of so-called K-reagents allows to achieve
correct results by suppression of the side reactions.
Figure 21: KF titration of 5 mL acetone using two different solvent: blue line - classical KF solvent, green line - solvent for ketones and
aldehydes. The titration in methanol is not stopping even after 600 due to side reactions.
Another possibility to avoid side reactions is to heat up the sample in a KF oven and thus to evaporate water to
be titrated in the KF vessel.
In many cases, a longer stir time, a higher stir speed and the use of selected auxiliary solvents have proven to
be very beneficial to improve dissolution or extraction allow for complete release of water.
Solution:
The KF instrument always determines the total amount of water present in the titration vessel, i.e. the water
present in the sample, the air humidity entering the vessel with time (drift) as well as adherent water on the
Competence Guide
Drift and blank values are subtracted from the total amount of water. If the sum of both contributions are bigger
than the total amount of water, then the result becomes negative.
Explanation:
To calculate the water content of the sample, the contributions due to the air humidity (i.e. the drift) and – where
needed - of the adherent water on inner surfaces of the vials (i.e. blank value) have to be subtracted from the
determined water amount:
Volumetric KF titration:
Coulometric KF titration:
Solution:
1. Drift:
a. Start the KF titration only when the online drift is very low e.g. below 5 μg/min. It may be necessary to wait
at least 15-20 min in stand-by to achieve low drift values.
b. Take into account or the drift by determining the blank using the same heating time as for the sample. In
this case, the drift value is set to 0 μg/min.
2. Blank value:
a. Use dry glass vials – previously dried in an oven,
b. Perform the blank determination for at least three vials, and calculate the average
c. Be sure that the inert gas used is correctly dried
3. Heating time:
Do not use too long heating time for the sample determination: a too long time increases the influence of the
drift value. Recommendation: 10-30 min.
“METTLER TOLEDO Methods for Water Content Determination”, Titration Applications Brochure No. 38,
ME-51725075B, August 2012.
“Karl Fischer Titration with a Homogenizer – Practical Examples”, Titration Applications Brochure No. 27,
ME-51725053B, April 2013.
“Validation of Titration Methods - A Guideline for Customers”, Titration Applications Brochure No. 16,
ME-51724912C, March 2017.
“Selected Applications with Karl Fischer Oven Autosampler”, Titration Applications Brochure No. 49, 30452052,
November 2017.
Links
InMotion Karl Fischer Titration Oven Autosampler - METTLER TOLEDO – YouTube
This application bulletin represents selected, possible application example. These have been tested with all
possible carein our lab with the analytical instrument mentioned in the bulletin. The experiments wre conducted
and the resulting data evaluated based on our current state of knowledge.
However, the application bulletin does not absolve you from personally testing its suitability for your intended
methods, instruments and purpose. As the use and transfer of an application example are beyond our control,
we can not accept responsibility therefore.
When chemicals and solvents are used, the general safety rules and the directions of the producer must be
observed.
After a short introduction on Karl Fischer titration, the main points affecting result
accuracy such as low drift determination, tightness of the instrument, fresh KF reagents
and specific sample preparation will be presented in detail. In addition, frequently asked
troubleshooting questions will be presented and answered.
Several KF applications have been developed by the Market Support Group Analytical
Chemistry. They are available for download on the METTLER TOLEDO
website www.mt.com.
www.mt.com
For more information