Select Page
Best practice for separation columns in ion chromatography (IC) – Part 3

Best practice for separation columns in ion chromatography (IC) – Part 3

This is the third and final post in our series providing you with tips and tricks on the proper use of ion chromatography columns. In the first part, we mostly discussed the standard operation conditions as well as operational limits for columns, while in the second post, we focused on application related topics and what elution parameters can be changed to modify the separation performance. In the conclusion of this series we will take a closer look at the ways to assess the column performance during its lifetime and offer some troubleshooting tips which can help in fixing issues that may appear. 

Essential column parameters

Let’s begin by looking at the main parameters that can be used to judge the column performance. Most of these reference values can be extracted from the certificate of analysis. The separation column certificates can be found in the Metrohm Certificate Finder. Reproducing the certificate chromatogram periodically to verify the column performance can be helpful to detect changes in performance early and to avoid irreversible damage of the column.

Below, we will consider the most important parameters one by one.
Retention time and column capacity

The retention time of the individual analytes is a good measure of the column selectivity and capacity. When the retention times of the analytes do not match the retention times of the certificate, there are many possible reasons for this behavior.

Possibly, the eluent composition is wrong, e.g., because the eluent components are not present in the correct concentrations. In case the eluent strength is too high, the retention times will be reduced. This phenomenon usually occurs for all ions simultaneously and shifts all the peaks closer together. Thereby, the multivalent ions are accelerated to a stronger extent than the monovalent ions. This issue can be fixed by preparing a fresh eluent with the correct composition.

The eluent may not be fresh or not sufficiently protected from the surrounding atmosphere, e.g., with a CO2 adsorber. Carbon dioxide from the ambient air can change the eluent composition (i.e., strength, pH, etc.) over time and this will affect the retention times of the analytes. Weak hydroxide eluents are particularly affected by this issue, since the elution strength of carbonate ions is much stronger than the one of hydroxide ions, leading to strong shifts in retention times. Therefore, multivalent ions are more affected by this than monovalent ones. This issue can be fixed by preparing a fresh eluent and by using a CO2 adsorber. When working with hydroxide eluents, it is also important to verify the status of the hydroxide stock solution as it can absorb CO2 from the air. Figure 1 shows the effect of CO2 uptake from the air on the retention times of common anions.

Effect of eluent absorption of carbon dioxide in the air on ion separation.
Figure 1. Example chromatograms showing the effect of eluent absorption of carbon dioxide in the air for seven standard anions measured on a Metrosep A Supp 5 – 250/4.0 column under standard conditions. Red: unprotected eluent. Black: eluent protected with a carbon dioxide adsorber.
Some samples may contain constituents that occupy the ion exchange groups of the column material. In that case, the number of available ion exchange groups on the column is reduced compared to the initial state, leading to an apparent smaller column capacity and shorter retention times. Depending on the type of contaminant(s), there are possibilities to reverse this effect and wash them from the column surface. These regeneration procedures are adapted individually to the different stationary phases and should be used as a last option to save a column.
The procedures as well as the operation limits are described in the respective leaflets of the column and can be found on our website. Please consult your Metrohm dealer before performing a regeneration procedure.
  • In case the contamination is from a multivalent ion, it is usually possible to wash them off by flushing the column with an eluent of a higher concentration. When doing so, do not forget the operation limits of the column, particularly regarding the eluent pH.
  • Sometimes, organic molecules can adsorb to the column stationary phase because of their strong affinity to it. This can lead to a blockage of the ion exchange groups and reduced retention times along with increased operating pressure. When this happens, it can help to wash the stationary phase with an eluent containing a fraction of an organic modifier. This will increase the affinity of the contaminant to the mobile phase and help to detach it from the stationary phase.

Usually, after such a regeneration procedure, the original column capacity can be restored.

An even better way to extend the lifetime of the separation column is by regularly exchanging the guard column. One of its purposes is to protect the separation column from contaminants by retaining them.

Make sure to exchange the guard column three to four times over the lifetime of the separation column, or even more frequently when dealing with very complex sample matrices such as dyes or food matrices. In these situations, we suggest using the Metrosep RP 2 Guard/3.5.

The chromatogram overlay in Figure 2 shows the effect of a defective guard column on the peak shape.
Effect of eluent absorption of carbon dioxide in the air on ion separation.
Figure 2. Example chromatograms showing the effect of a damaged guard column. Black: analysis of standard anions when using a defective guard column. Red: identical analysis without a guard column installed. In this case, the replacement of the guard column is strongly recommended.
In some cases, the column capacity is reduced due to chemical modification of the stationary phase. This happens when a column is used in conditions outside of the operation limits—e.g., extreme pH situations. Extreme pH conditions can alter the chemical bonds of the ion exchange groups to the base particle substrate. This irreversible process leads to permanent column capacity loss. In this case, no regeneration procedure can restore the original column capacity.

Please note that this loss of ion exchange groups also happens slowly during the regular use of ion exchange columns. The strength of the chemical bond of the ion exchange groups is thoroughly tested during the development of Metrohm columns to provide a long column lifetime.

Theoretical plates and peak symmetry

The number of theoretical plates (TP) can be a useful tool to judge the column packing status. The higher the TP value, the better the packing of the column. For a meaningful judgement of the column packing bed, choose an analyte that does not elute too early in the chromatogram (these peaks can be affected by extra-column effects, in particular on 2 mm systems) and that is not affected by secondary interactions (i.e., avoid using nitrate). For cations, potassium is a good choice, while for anions, sulfate is a suitable choice.

Although the columns are packed at pressures higher than the normal operating pressure, the packing bed continues to further densify throughout the column lifetime with the continuous application of flowing mobile phase (eluent). While the packing bed of the stationary phase improves due to this effect, it may also lead to some dead volume at the entrance of the separation column. This dead volume can be responsible for peak broadening and reduced theoretical plates, particularly at the beginning of the chromatogram.

While this is a regular aging process of the column, it is possible to slow it down by taking good care of the column. Preventive actions include the slow initiation of the eluent flow rate and temperature at start-up as well as the proper and complete shutdown of the high-pressure pump before removing the column, as mechanical stress can be detrimental to the packing bed. Exchanging the guard column can often have a positive impact on the theoretical plates and peak shapes (see Figure 2).

Issues with the column packing bed often also become visible in the asymmetry factor values and the overall peak shapes. More pronounced fronting as well as peak broadening can be warning signs for channeling in the column or the guard column. Unfortunately, this kind of damage on the column is irreversible and requires an exchange of the separation column.

Before column replacement, it is important to check if the issues originated from the column or another connection in the IC system whenever peak broadening is observed in the chromatogram. Make sure that all capillaries in the high-pressure path have a diameter of ≤0.25 mm and that all capillaries have been installed and connected correctly without additional dead volume. Systems equipped with small inner diameter (2 mm, microbore) columns are more strongly affected by dead volume than those with 4 mm columns. This means that when using microbore columns, less dead volume is required for the peak broadening effect to become visible compared to when using 4 mm columns.

Column pressure

Another important parameter to regularly check throughout the column lifetime is the system pressure. High pressure is among the most frequent reasons leading to a column replacement. Whenever a pressure increase is noted, it is important to verify what part of the IC system is the cause.

If the sample contains particles and insufficient sample preparation is applied, the particles will accumulate at the entrance of the protective guard column and eventually lead to increased system pressure. In this case, the guard column performs its intended use of protecting the separation column and needs to be replaced. If the guard column is not replaced soon enough, then particulate contamination may break through and load onto the separation column. While there is no regeneration procedure for a guard column loaded with particles, separation columns contaminated with particles may be regenerated by rinsing the column at low flow rates in the reverse flow direction.

Please note that these regeneration procedures may not always be successful.

Avoiding particles from being injected into the high-pressure path of the IC system is a good way of protecting the guard and separation columns. Metrohm provides several automated sample preparation techniques which result in a positive impact on the column lifetime. The most common techniques for this purpose are Inline Ultrafiltration (Figure 3) and Inline Dialysis.

Effect of eluent absorption of carbon dioxide in the air on ion separation.
Figure 3. Inline Ultrafiltration is a useful automated sample preparation technique that protects the separation column from particle accumulation.
Learn more about Metrohm Inline Ultrafiltration in our related blog post.

Column end of life

All of the parameters described above should be considered to judge the column (and the IC system) performance. Many of these parameters can be monitored closely in the MagIC Net software so that a potential issue can be detected as early as possible.

Even though Metrohm IC columns are designed and manufactured to have very long lifetimes, at some point performance will decrease and even regeneration procedures may not be able to restore the required column performance to solve the application needs. This represents the end of the column life and makes the exchange of the separation column unavoidable.

Metrohm separation columns cannot be recycled and can be disposed with normal waste. However, depending on the samples measured as well as the used chemical types and their associated hazards, it may be necessary to consider a proper disposal option.

Irrespective of the sample nature, do not open the column at any time. 

Troubleshooting overview

Table 1 gives an overview about certain troubleshooting strategies when looking at your column performance behavior.

Table 1. Preventing and correcting performance loss in IC columns


Indicator Cause Preventive and corrective measures
Increasing counterpressure Particles on the guard column Replace the guard column.
Particles on the separation column

Rinse out the separation column in the reverse flow direction

  • Place the column inlet (i.e., the opening next to the intelligent chip) in a beaker, as this is now the outlet path for the contaminants.
  • Rinse out the separation column for approx. one hour.
  • Reinstall the separation column in the flow direction.
Particles in the sample Sample preparation, e.g., remove particles through lnline Ultrafltration
Shortened retention time Carbonate in the eluent

Carbon dioxide from the air affects the carbonate/hydrogen carbonate balance in the eluent. A carbonate/hydrogen carbonate eluent weakens over time; a hydroxide eluent strengthens.

  • Always tightly seal eluent bottles and bottles containing eluent concentrate.
  • Always use a CO2 adsorber.
Air bubbles in the eluent

Air bubbles make the eluent flow unstable. The counterpressure is an indicator of unstable flow. lt should remain stable
within a range of ± 0.1 MPa.


  • Deaerate the high-pressure pump.
  • Use an eluent degasser.
Capacity loss in the column due to high- valency ions Regenerate the column as per the column leaflet to remove any inorganic deposits.
Resolution loss Eluent too old or produced incorrectly Eluents should be freshly prepared. Make sure that they are produced correctly and particularly that carbonate and hydrogen carbonate are not confused.
Adsorptive effect of the contamination deposited in the guard column Replace the guard column.
Adsorptive effect of the contamination deposited in the separation column Regenerate the column as per the column leaflet to remove any organic or inorganic deposits.
Loss of theoretical plates Guard column contaminated Replace the guard column.
Separation column contaminated Regenerate the column as per the column leaflet to remove any organic or inorganic deposits.
Separation column overloaded

The separation column can be overloaded by factors such as a high salt content in the sample matrix.

  • Dilute the sample.
  • Inject less sample.
Dead volume in the IC system
  • Check that all capillaries have a diameter ≤0.25 mm; if they don’t, replace the capillaries.
  • Check that all of the capillaries have been installed correctly. The installation process is described step by step in the «IC Maintenance» multimedia guide.
Asymmetry Dead volume or contamination on the guard column Replace the guard column.
Separation column contaminated Regenerate the column as per the column leaflet to remove any organic or inorganic deposits.



This article explained how IC column performance can be assessed and monitored throughout the column lifetime and what measures can be taken to ensure a long column lifetime. With that, we now conclude the series «Best practice for separation columns in ion chromatography». If you have more questions, do not hesitate to contact your local Metrohm IC salesperson.

Get free tips and tricks

for your IC columns 
Post written by Dr. Vincent Diederich (Product Manager IC Columns) and Dr. Anne Katharina Riess (Head of Column Division) at Metrohm International Headquarters, Herisau, Switzerland.
Ion-selective electrodes: General tips – Part 1

Ion-selective electrodes: General tips – Part 1

Ions – we encounter these tiny charge carriers constantly. Depending on the concentration of certain anions (negative ions) and cations (positive ions), they can have a significant impact on humans and the environment. Thanks to ongoing quality control in several industries such as food and beverage, the metallurgical industry, and water management, defined limits are neither exceeded nor undercut.
So, how can these small, ubiquitous ions be determined? Mistakenly, I thought at first that ion measurement is only possible by means of more costly analytical methods such as ion chromatography (IC), inductively coupled plasma optical emission spectrometry (ICP-OES), or atomic absorption spectroscopy (AAS). A promising cost-efficient alternative to these techniques is to use so-called ion-selective electrodes (ISE).

Types of ion-selective electrodes

If you want to determine, for example, the fluoride concentration in your toothpaste, the ammonium content of the water in your aquarium, or perhaps how much calcium is really in a fruit juice, then there are many suitable ion-selective electrodes for your application needs.

Membrane material

The very first ion-selective electrode was the pH electrode. However, this article will not discuss pH electrodes—you can find more information in our other posts pertaining to this special ISE.

Aside from the glass membrane used for pH electrodes, there are other membrane materials available for selective measurement of myriad ions. The most widely applied types are listed in Table 1.
Table 1. Various types of ion-selective electrodes.
Membrane material Description Ions Close-up view
Crystal membrane Crystal lattice containing defined gaps for the ion to be measured. Ag+, Cu2+, Pb2+, Br, Cl, CN, F, I, S2-
Polymer membrane Polymer membrane containing a molecule (ionophore) that only binds the ion to be measured. Ca2+, K+, Na+, surfactants, NO3
Glass membrane Framework of silicate glass with interstitial sites for H+ and Na+. Na+, H+
Gas permeable membrane Membrane acts as a permeable barrier through which only specific substances can move across. NH4+
The membrane material can limit the possible matrices in which the ISE can be used. For example, electrodes with a polymer membrane cannot be used to measure ions in organic solvents. For more information on the specific restrictions, check the user manual of your ISE.

Basic theory behind ISEs

Measuring range

Each electrode type has its own specific measuring range (see Table 2). Before beginning any ion measurement, first ensure that the ion-selective electrode is able to measure in the concentration range of the sample.

Table 2. Each ion-selective electrode has its specific measuring range. Note: The given measuring ranges only apply to ion-selective electrodes from Metrohm.
Cation of interest Measuring range
Ag+ 1×10–7 – 1 mol/L
Ca2+ 5×10–7 – 1 mol/L
Cd2+ 1×10–7 – 10–1 mol/L
Cu2+ 1×10–8 – 10–1 mol/L
H+ 1×10–14 – 1 mol/L
K+ 1×10–7 – 1 mol/L
Na+ (Polymer)

Na+ (Glass)

5×10–6 – 1 mol/L

1×10–5 – 1 mol/L

NH4+ 5×10–6 – 10–2 mol/L
Pb2+ 1×10–6 – 10-1 mol/L
Anion of interest Measuring range
Br 1×10–6 – 1 mol/L
Cl 1×10–5 – 1 mol/L
CN 8×10–6 – 10–2 mol/L
F 1×10–6 – sat. mol/L
I 5×10–8 – 1 mol/L
NO3 1×10–6 – 1 mol/L
S2- 1×10–7 – 1 mol/L
However, something more important than the measuring range is the linear range. Figure 1 illustrates a measuring range which also includes a linear range. Within this linear range the Nernst equation applies, and the signal is proportional to the analyte concentration. By performing an ion measurement within the linear range, the most precise and reproducible results will be obtained. Find out more about the Nernst equation in our previous blog post.
Figure 1. The measuring range of an electrode.
Outside of the linear range, the curve becomes flatter, and the potential difference becomes smaller, preventing a reliable measurement by standard addition. Even in this non-linear, flattened range, it is possible to determine the ion concentration by means of direct measurement – provided that your ion-selective electrode is also calibrated for this range.

If the concentration is too low or the sensor is saturated, this situation is considered outside of the measuring range. Potential changes can no longer be determined.

Interfering ions

Compared to a pH electrode with a linear range over 14 decades, the sensitivity of ISEs is limited since interfering ions reduce the linear and measuring ranges (see Figure 1).

There are two different types of interfering ions, both outlined in Table 3.

Table 3. Types of interfering ions and their impact on measurement with ISEs.
Interfering ion Description Impact
  • Binds to the membrane material and reacts with it or
  • Forms complexes and precipitates with the measuring ion
Destruction of the ISE since the irreversible interfering ion reacts with the membrane and is unavailable for further analysis.
  • Exhibits cross-sensitivities
Erroneous results since the reversible interfering ion binds to the membrane material and contribute to the signal.
Nowadays, the most important interfering ions for an ISE are known, and information about them is provided by the sensor manufacturer. For the measurement, the impact of interfering ions is considered in a so-called selectivity coefficient, which in turn is used in the Nikolsky equation—an expanded version of the Nernst equation.
Table 4. Each ion to be measured has at least one interfering ion that must be considered when setting up an experiment.
Measuring cation Interfering ion(s)
Ag+ Hg2+, proteins
Ca2+ Na+, Pb2+, Fe2+, Zn2+, Cu2+, Mg2+
Cu2+ Ag+, Hg2+, S2-, Cl, Br, I, Fe3+, Cd2+
K+ Na+, NH4+, Cs+, Li+, H+
Na+ (Polymer)

Na+ (Glass)

SCN, acetate

H+, Li+, K+, Ag+

Pb2+ Ag+, Hg2+, Cu2+, Fe3+, Cd2+
Measuring anion Interfering ion(s)
Br Hg2+, I, S2-, CN, NH4+, S2O32-
Cl Hg2+, Br, I, S2-, CN, NH4+, S2O32-
CN S2-, Ag+ complexing substances, I, Cl, Br
I Hg2+, S2-, CN, Cl, Br, S2O32-
NO3 Br, NO2, Cl, acetate
S2- Hg2+, proteins
Some examples for the most important interfering ions of ISEs are listed in Table 4. For more information on the theoretical background of pH as well as ion-selective electrodes, download our free monograph below.

Ionic strength adjustment

The measurement depends on the activity of the measuring ion in solution, which in turn depends on the ionic strength. For this reason, ion-selective measurements are always carried out in solutions with approximately the same ionic strength. By addition of ionic strength adjuster (ISA) or total ionic strength adjustment buffers (TISAB), a constant ionic background can be achieved.

ISA and TISAB are chemically inert respective to the measurement, and they contain a relatively high concentration of salt so that the ionic strength of the sample solution can be neglected. Some examples can be found in Table 5. Check the user manual of your ISE to find its ideal ISA or TISAB solution.

Table 5. Recommended ISA and TISAB solutions for ion-selective electrodes and further information about each procedure.
Ion to be measured ISA / TISAB Further information about procedure
Fluoride (F) NaCl / glacial acetic acid / CDTA Application Bulletin AB-082
Potassium (K+) c(NaCl) = 0.1–1 mol/L Application Bulletin AB-134
Sodium (Na+), glass membrane c(Tris(hydroxymethyl)aminoethane) = 1 mol/L Application Bulletin AB-083
Sodium (Na+), polymer membrane c(CaCl2) = 1 mol/L
Ammonium (NH4+) c(NaOH) = 10 mol/L Application Bulletin AB-133

Electrode maintenance and lifetime

Cleaning the ISE
  • After each measurement or titration, the ISE has to be rinsed thoroughly with distilled water.
  • Never use organic solvents for cleaning. They may attack or irreversibly destroy the polymer membrane ISE or reduce the lifetime of your crystal membrane ISE.
Conditioning the ISE

Conditioning steps must be performed before the first use as well as in between measurements. This step activates the measuring membrane and provides a stable equilibrium of the measuring ion in the membrane. By doing so, an accurate ion measurement is possible. An ion standard solution with a concentration of c(ion) = 0.01 mol/L is recommended as the conditioning solution.

Storing the ISE

An overview of proper storage instructions for your ion-selective electrode is shown in Table 6. For more detailed information, check the ISE manual.

Table 6. Recommended storage conditions for different ISE types.
Membrane material Short storage period Long storage period
Crystal membrane In c(ion) = 0.1 mol/L Dry, with protective cap
Polymer membrane Dry Dry
Polymer membrane, combined In c(ion) = 0.01–0.1 mol/L Dry, with some residual moisture
Glass membrane In c(ion) = 0.1 mol/L In deionized water
Lifetime of an ISE

The lifetime of an ion-selective electrode depends on several influencing parameters including membrane type, sample matrix, and electrode maintenance. Don’t forget to regularly exchange the electrolyte of your combined ISE or – in case of a separate ISE – your reference electrode. Furthermore, never touch the membrane with bare fingers.

In general, the following can be said:

  • Polymer membrane electrodes: Limited lifetime of about half a year since the membrane ages, resulting in a loss of performance.
  • Crystal membrane electrodes: Lifetime of several years – the membrane can be regenerated by polishing using an appropriate polishing material. Watch the video below for more details.


  • If you decide to perform an ion measurement using an ion-selective electrode, you must consider the measuring range and any interfering ions that may be present in advance.
  • In addition to the membrane type and the sample matrix, the cleaning, storage, and conditioning all have an influence on the lifetime of your ion-selective electrode.
Hungry for more information regarding direct measurement and standard addition? Check out Part 2 (coming soon!) where we will discuss the different determination methods.

Can’t wait? Download our free White Paper: «Overcoming difficulties in ion measurement: Tips for standard addition and direct measurement».

Free White Paper:

Overcoming difficulties in ion measurement: Tips for standard addition and direct measurement
Post written by Doris Hoffmann, Product Manager Titration at Metrohm International Headquarters, Herisau, Switzerland.
USP  – simple automated analysis of ultrapure water

USP <645> – simple automated analysis of ultrapure water

H2O – two simple elements, oxygen (O) and hydrogen (H), fuse together to form one of the most important molecules in the world: water. Water is everywhere on Earth and it is vital for our health and survival. It often contains other ions like calcium, magnesium, and chloride which are essential for the human body to function. However, in specific situations, ultrapure water (UPW) is needed to prepare e.g., injections or other solutions used in hospitals. How is the quality of UPW ensured so that it is always suitable for such medical purposes? The answer to this comes from USP <645>. This standard explains how the water quality can be determined and how this analysis must be performed.


For this analysis, a measuring device capable of measuring the conductivity and the pH value is required. If a combined device is not available, then using two separate ones is also fine. Then, a pH electrode that is especially suitable for the determination of the pH value of water and a conductivity cell for measurement of low conductivities is necessary. In this case, the Aquatrode plus and the stainless steel conductivity cell are recommended.

The Aquatrode plus responds very quickly in ion-deficient matrices (such as ultrapure water) and thanks to its double junction system, the bridge electrolyte can be chosen freely.
The stainless steel conductivity cell has been specially designed for measurements in samples with low conductivity. With a cell constant of c = 0.1 cm-1, it is ideal for conductivities ranging from 0–300 µS/cm. The built-in temperature sensor makes handling even easier as no additional sensor is needed for the temperature measurement.

USP <645> procedure

Now that the necessary instrumentation has been introduced, it’s time to take a look at the standard procedure itself. At first glance this looks a bit difficult as it is a three-step analysis, or actually a four-step analysis if you count the calibration as well.

Step 1: First, calibrate the pH electrode and the stainless steel conductivity cell (sensor). The pH electrode is calibrated with pH 4 and 7 solutions, whereas the stainless steel sensor is calibrated with a 100 µS/cm standard.

Find the standard solutions you need here:

Step 2: After calibrating the sensors, both the temperature of the water and the conductivity are measured without temperature compensation. If the measured conductivity is lower than the value mentioned in the table of USP <645>, then the requirement for the conductivity is met and the water can be used for medical purposes. If this is not the case, then step 3 must be performed.

Step 3: 100 mL water is transferred to an external titration vessel where its temperature is adjusted to 25 ± 1 °C. The water is stirred vigorously to incorporate carbon dioxide present in the air. If the conductivity does not change by more than 0.1 µS/cm per 5 minutes, the value is noted for further evaluation. If this value is below 2.1 µS/cm, then the water is usable for medical purposes. If not, then proceed with step 4.

Step 4: The solution is tempered to 25 ± 1 °C. Once the temperature is stable, 0.3 mL of saturated KCl solution is added and the pH of the water is measured. The pH value must lie between pH 5 and 7. If this is not the case, the water does not meet the requirements and must be discarded. If the pH is measured between 5 and 7, then the conductivity must additionally be lower as mentioned in the USP <645> table. If this is the case, the analysis passes and the water can be used for medical purposes.

Automation of USP <645>

The analysis can be quite time-consuming and therefore Metrohm has provided a solution to make this process much easier. Our system combines all of these steps into one method, allowing you to perform walk-away automation and focus on more important tasks.

To expand the capabilities of the measuring system, the 856/867 modules can be exchanged with a modular OMNIS Titrator which can also be used for other standard potentiometric titrations.

Take a closer look at our automated analysis solution

Proper electrode immersion depth
  • Aquatrode plus for accurate pH measurement in ion-deficient matrices
  • Stainless steel conductivity cell for low conducting samples
  • Special holder for performing step 2 of the standard procedure
  • Thermostated vessel (for step 2)
  • DIS-Cover lid to prevent the sample from CO2 uptake (before step 3)
  • Rod stirrer to saturate the solution with carbon dioxide (for step 3)


With this complete system, standard analysis of UPW quality according to USP <645> is performed in a fully automatic and reliable manner. At the end of each analysis a clear message is received if the deionized water (UPW) has passed the test or not. Handling is very easy and allows users to check if an analysis passes or not with just a glance.

Download our Application Bulletin

Automatic conductometry in water samples with low electrical conductivity in accordance with USP<645>
Post written by Iris Kalkman (Product Specialist Titration) and Heike Risse (Product Manager Titration – Automation), Metrohm International Headquarters, Herisau, Switzerland.
Side reactions in Karl Fischer titration

Side reactions in Karl Fischer titration

Many chemists that utilize Karl Fischer titration are nervous about the presence of side reactions because they know that the water determination in their samples can only be specific without any side reactions. Other KFT users do not know what the possible side reactions are and therefore may obtain incorrect results.

What are side reactions?

These are reactions with substances in the sample that:

  • interfere with the stoichiometry of the KF reaction
  • change the pH value of the KF reagent
  • either produce or use up water themselves
  • oxidize on the anode of the generator electrode
  • reduce on the cathode of the generator electrode
  • react with the ingredients of the KF reagent

Recognizing side reactions

One of the worst things that can happen with KFT is not knowing that a side reaction is falsifying your results. Below are some characteristic signs of side reactions.

Titration time and titration curve

Some indications of side reactions include longer titration times compared to the titration of a water standard, slow endpoint detection, and a higher drift value after the titration finishes than at the titration start. Comparing the titration curves of the sample and a water standard with a similar water quantity makes it easier to evaluate the situation. Just plot a graph of the volume against time (or µg water against time, in the case of coulometry). If the graph exhibits a curve that increases steadily as illustrated in Figure 1 (in orange), this can indicate a side reaction.

Figure 1. Side reactions can often be identified from checking the titration time and the titration curve, as shown in this graph.

If you notice that the water content depends on the sample weight or the titrant consumption (µg water for coulometry), then you can check the slope of a regression line after plotting water content against titrant consumption (µg water).

Ideally, the slope (b) should be 0. Significantly positive or negative values can indicate a side reaction, as shown in Figure 2.

Figure 2. If the slope of the regression line for the water content / titrant consumption value pairs deviates significantly from 0, this indicates a side reaction.

If the water recovery value found after spiking the samples is not within 100 ± 3%, this can indicate a side reaction. Depending on the type and speed of the side reaction, the recovery may be too high or too low. For example, samples which contain DMSO (dimethyl sulfoxide) change the stoichiometry of the Karl Fischer reaction and therefore result in false low readings.

Please note that a recovery rate of almost 100% does not guarantee the absence of a side reaction. Side reactions that take place very rapidly will not be detected, since the side reaction is already complete when the spiking process begins. A spiking procedure is described in detail in chapter 2.5.12 of the European Pharmacopoeia.

Preliminary tests

The oxidation of iodide or reduction of iodine leads to incorrect results.

How can you check whether your sample is undergoing a side reaction with iodine or iodide? A simple preliminary test can clarify the situation. Dissolve the sample in a weakly acidic (alcoholic) solution and then add some drops of iodine or potassium iodide solution. Based on the coloring (a discoloration of iodine or the formation of brown iodine), a side reaction can be detected.

Evaluating redox potentials

Comparing the redox potentials of the redox pairs of sample substances with the redox potential of iodine/iodide can be helpful to assess whether an undesired redox reaction may occur.

If the standard potential is higher than that of iodine/iodide, as in the case of e.g., chlorine, the oxidation of the iodide may result in false low readings.

If it is lower (e.g., lead), the reduction of the iodine may result in values that are too high.

Element changing oxidation state oxidized form + x e → reduced form Standard electrode potential E°
Cl Cl2 + 2e ⇌ 2 Cl +1.36 V
I I2 + 2e ⇌ 2 I +0.54 V
Pb Pb2+ + 2e ⇌ Pb -0.13 V

Avoiding side reactions

Most side reactions can be suppressed by taking suitable measures, such as those listed here.

  • For ketones and substances that react with the methanol present in the KF reagent: Use methanol-free reagents.
  • For samples that lower the pH range of the KF reagent: Add buffer solution for acids or a stoichiometric excess of imidazole.
  • For samples that increase the pH value (e.g., aminic bases): Add buffer solution for bases or a stoichiometric excess of salicylic acid / benzoic acid.
  • High drift after titration: Postdrift correction may help. This is done by stopping the titration at a defined time and recording the additional consumption over several minutes. This allows the calculation of the drift after the titration. This postdrift is then used to correct the water quantity found.
  • Samples that reduce iodine: Subtract the iodine consumption of the reductant in the sample from the overall iodine consumption of the sample.
  • Samples that oxidize iodide: Reduce the oxidant, e.g., Cl2, in advance with an excess of SO2, for example, by treating the sample with the solvent of a two-component reagent.
  • General: Carry out the titration in a thermostatically controlled cell connected to a circulation thermostat at, e.g., -20 °C in order to slow down the side reaction. Note that the titration parameters should be adjusted to the low temperatures.
  • General: Extract the water with the KF oven method if the interfering components are thermally stable at oven temperature.
  • General: Mask or eliminate the interfering component, e.g., by adding N-ethylmaleimide in the case of thiols.
Find out more about the Karl Fischer oven method in our blog article.


Side reactions can negatively influence and falsify your results. Recognizing and avoiding side reactions in KF titration is therefore crucial for the most accurate determinations.

For more information, check out our blog series about frequently asked questions in Karl Fischer titration.

Download our free monograph:

Water determination by Karl Fischer Titration
Post written by Michael Margreth, Sr. Product Specialist Titration (Karl Fischer Titration) at Metrohm International Headquarters, Herisau, Switzerland.
Ph. Eur. 2.2.48 Raman Spectroscopy: How Raman instruments from Metrohm comply with the 2022 update

Ph. Eur. 2.2.48 Raman Spectroscopy: How Raman instruments from Metrohm comply with the 2022 update

The European Pharmacopoeia (Ph. Eur.) is a single reference work for the quality control of medicines. Ph. Eur. contains norms, suggests analytical methods, and lists many properties that define quality control (QC) during the production of medicines, the raw materials used, and the instruments required to perform such tests. These official standards are legally binding in several countries – not only in Europe, but worldwide.

Raman spectrometers—especially handheld and portable instruments—are increasingly used for QC of medicines and raw materials (RMID). Instrument interfaces are user-friendly, requiring little technical expertise, and they provide flexible sampling options for most sample types with rapid, non-destructive measurements.

Metrohm’s Raman systems exhibit great flexibility—from see-through to standoff to immersion sampling.
An excerpt from the Ph. Eur. 2.2.48 Raman Spectroscopy chapter says:
«Raman spectroscopy is commonly used for qualitative and quantitative applications and can be applied to solid, liquid, and gaseous samples. Raman spectroscopy is a rapid and non-invasive analytical method and can be performed off-line, at-line, on-line, or in-line[…] Raman spectrometers can be situated far from the point of measurement using long-distance optical fibres to collect the Raman signal.»

Technological developments and their increasing adoption in the pharmaceutical industry prompted a revision of Ph. Eur. 2.2.48 which ensures the reliability of Raman results. The updated chapter 2.2.48 was published in the Ph. Eur. Supplement 10.7 (October 2021) and will ultimately take effect in April 2022.

While much of the Ph. Eur. 2.2.48 chapter has remained the same, the latest revision features:

  • new requirements for spectral resolution for qualitative Raman analysis using a suitable reference material
  • updated requirements for the Raman response-intensity scale
  • detailed procedures for the comparison of spectra

We will address these new requirements across our Raman spectroscopy product lines in the rest of this article.

Spectral Resolution

«Spectral resolution is the ability of a spectroscopic system to separate adjacent bands, which makes it possible to characterise complex samples (e.g., brand analysis, crystallinity, polymorphism).

[…] For identity tests, unless otherwise prescribed in a monograph, the spectral resolution must be less than or equal to 15cm-1 (measured in the wavenumber range between 1000cm-1 and 1100cm-1).

The spectral resolution is verified using a suitable reference material. The instrument parameters used for the test, such as laser, slit-width and grating for dispersive instruments and circular aperture […] for FT-instruments, must be the same as those applied for sample measurements. For example record the Raman spectrum of calcium carbonate for equipment qualification CRS, and determine the full width at half height (W1085) of the band located at 1085 cm-1. The spectral resolution (R) using calcium carbonate is then given by the following relation:»

Handheld Raman instruments: MIRA P and NanoRam

All MIRA P and NanoRam devices (including both NanoRam and NanoRam-1064) for the pharmaceutical industry are designed and tested to meet stringent resolution requirements. During QC, the resolution of each instrument is tested to be less than 15 cm-1 against a secondary USP (US Pharmacopeia) reference standard of calcium carbonate according to ASTM E2529, which is the same procedure recommended in this newly released Ph. Eur. chapter.

The measured spectral resolution value for each instrument, along with its identifying serial number, is included in the instrument final test report. A certificate or final test report is packaged with the device and sent to the customer (starting in April 2022 for MIRA P). This resolution is fixed by the optical design of the instrument and is stable over time.

Portable Raman instruments: i-Raman series, QTRam, STRam, and PTRam

The instrument resolution for all of Metrohm’s portable Raman instruments from B&W Tek are factory-tested with calcium carbonate and displayed on final instrument test reports. The spectral resolution is dependent on the instrument design and defined for each specific instrument configuration. Depending on the instrument model, the spectral resolution is between 3.5–11 cm-1. Additionally, the instrument control softwares Vision and BWAnalyst have the performance test function that verifies spectral resolution using the 1001.4 cm-1 peak of polystyrene.

Handheld and portable Raman instruments from B&W Tek.

Response-Intensity Scale

«The verification of the response-intensity scale is principally performed for quantitative methods.

Appropriate acceptance criteria will vary with the application. A maximum variation of ± 10 per cent in band intensities compared to the previous instrument qualification is achievable in most cases. Response calibration may involve the use of white-light standards or luminescent glass (e.g., NIST SRM 2241).»

Handheld Raman instruments: MIRA P, MIRA M-3, and NanoRam series

MIRA P, MIRA M-3, and NanoRam systems are designed for qualitative analysis, not for quantitative purposes. Therefore, this criterion is not a strict requirement for handheld Raman products.

However, the relative intensity response of MIRA P and NanoRam series instruments is calibrated with a NIST standard SRM calibration material (SRM 2241, SRM 2242) or NIST SRM 2241-traceable calibration standard to achieve better uniformity from instrument-to-instrument.

The NanoRam series instruments have an acceptance criterion for relative intensity response in the instrument performance validation in alignment with Ph. Eur. 2.2.48 and USP<858>. To pass the performance validation, <10% relative intensity error is required using the factory-supplied polystyrene cap.

Portable Raman instruments: i-Raman series, QTRam, STRam, and PTRam

The relative intensity response of these portable Raman instruments is calibrated using a proper NIST standard SRM calibration material to achieve better uniformity from instrument-to-instrument. Additionally, the Vision instrument control software includes the performance test function that verifies the intensities of several Raman peaks of polystyrene relative to its 1001.4 cm-1 peak, to a maximum variation of ±10% compared to the previous instrument qualification.

Comparison Procedures

For qualitative methods, additional information for identification has been defined.

«Several comparison procedures may be used, and the analyst must document and justify the method used and the specific acceptance criteria that allow a conclusion for identification. The spectra can be compared by either overlaying the spectra (in the whole spectral range or in the region of interest specified in the monograph) or by using mathematical calculations of the software. It is possible for example to perform:

  • visual comparison based on band positions and relative intensities unless otherwise specified[…]
  • a statistical determination of the similarity between the spectra of the material to be examined and the reference standard[…]
  • evaluation by chemometric methods[…]»

While an experienced Raman spectroscopist can certainly compare spectra visually and assess sample validity based on peak location, fluorescence, saturation, and signal-to-noise ratio, the widespread implementation of Raman in the real world means that complex analysis must be done by the device and not the user. Statistical comparison methods are used primarily for identification of unknowns through correlation of a sample spectrum with library spectra. The software performs library searches and returns a Hit Quality Index (HQI) value indicating the level of correlation as defined by a user-defined threshold.

Chemometric methods rely on dimensionality-reduction methods that are performed by the software, such as Principal Component Analysis (PCA), where new sample data is compared within a multivariate model created from representative samples. This permits highly accurate verification of known materials according to how well a spectrum fits into the model limits, which are determined by a confidence interval. In the analysis of medicines and raw materials, chemometric methods are used to distinguish the quality and consistency of a material. MIRA P (and its dedicated software, MIRA Cal P) and NanoRam instruments use both statistical and chemometric methods for sample identification and verification based on the needs of the end-user.

For more information, download our free technical and application notes as well as a White Paper below.

Wavenumber accuracy requirements of Ph. Eur. 2.2.48

«Verify the wavenumber scale for Raman shifts using a suitable standard that has characteristic maxima at the wavenumbers under investigation, for example an organic substance such as polystyrene, paracetamol or cyclohexane[..]

A minimum of 3 wavenumbers covering the working range of the instrument intended for measurements should be selected.[…]»

This chapter maintains the same requirements for Raman wavenumber accuracy and is consistent with the USP <858> and JP 2.26. All of Metrohm’s handheld Raman instruments meet these requirements. Users are recommended to run performance validation tests at regular intervals using polystyrene or another ASTM Raman shift calibration material.

Download our free White Paper below to learn more about instrument calibration, system verification, and performance validation.

The Calibrate/Verify Attachment (CVA) shown here is a dual-ended accessory containing a toluene/acetonitrile ASTM standard for calibration/verification of the wavenumber axis and polystyrene for a second wavenumber verification according to Ph. Eur. 2.2.48.
Performance tests for Raman wavenumber accuracy are included the Vision and BWAnalyst softwares for the i-Raman series and other portable Raman products (STRam, QTRam, PTRam), with acceptance criteria in accordance with the pharmacopeial requirements.

Metrohm’s unique way of compliance with Ph. Eur. 2.2.48

Better representation of the material

«When using Raman spectroscopy[…] care must be taken to ensure that the measurement is representative. This can be achieved by, for example rotation of the sample, performing multiple measurements on different preparations of the sample, using orbital raster scanning (ORS), increasing the area of illumination by reducing the magnification, by demagnification of the laser beam or by changing the focal length between measurements to scan at different depths.»

ORS™ is Metrohm Raman’s proprietary method for moving the excitation laser in a pattern over a sample in order to collect more representative data from a larger area of the sample, especially on heterogeneous samples. All MIRA and MISA instruments are equipped with ORS.
Learn more about ORS by downloading the related Application Note.
For more details about how we comply, please check the U.S. Pharmacopeia Raman Chapters Updates page on the B&W Tek website. For more general information, download this General Compliance Statement for MIRA handheld Raman systems
For a more comprehensive look at raw material identification and verification in the pharmaceutical industry, there is a significant amount of information on this topic in our related blog post.

Post written by Dr. Melissa Gelwicks (Technical Writer at Metrohm Raman, Laramie, Wyoming), and Dr. Xiangyu (Max) Ma (Handheld Raman Product Manager) and Dr. Jun Zhao (R&D Director) at B&W Tek, Plainsboro, New Jersey.