Karl Fischer Titration: Volumetric, Coulometric and Performance Checks

Choose Karl Fischer titration from the expected amount of water introduced into the cell, the sample matrix and the required decision, rather than from a fixed percentage boundary. Volumetry measures delivery of a reagent with an established water equivalent; coulometry relates electrochemically generated iodine to electrical charge. Both need suitable chemistry, controlled sample introduction and demonstrated performance.

A low drift value or a passing water reference does not prove that a poorly soluble sample releases all its water. Equally, loss on drying is not automatically a water-specific result. This article connects instrument selection, two original comparison tables and a hypothetical case to a practical control strategy, without prescribing universal limits or a complete analytical procedure.

Define which water result the decision requires

Start with the measurand: water mass fraction in a specified material, on a stated reporting basis, determined by a defined procedure. Record whether the sample is hygroscopic, heterogeneous, volatile, hydrated or likely to change during handling. “Moisture” alone is too imprecise for a user requirement if different methods may report different quantities.

Loss on drying measures mass lost under specified conditions. Other volatile components may contribute, and some water may remain. Karl Fischer uses a chemical reaction to determine water, but interfering reactions and incomplete release can bias its result. NIST training explains this distinction. Agreement between the two methods in one sample is useful evidence, not proof of interchangeability for every formulation.

For compendial testing, identify the applicable monograph, chapter, edition and effective date before selecting the procedure. The public introduction to USP ⟨921⟩ makes the monograph central to the choice. Ph. Eur. 2.5.12 and 2.5.32 concern different water-determination approaches; their current full texts were not accessed for this article. Obtain the controlled applicable text before defining compliance tests.

Specify the decision the result supports: material release, investigation, stability or development. The required working range, selectivity, precision and reporting uncertainty follow that use. A research comparison on oils or cement can reveal a failure mechanism; it cannot by itself validate a pharmaceutical matrix or set its acceptance criterion.

Compare volumetric and coulometric operation at the intended water load

In volumetric Karl Fischer, a dosing system adds iodine-containing titrant. Its water equivalent, often called the titre or factor, connects the delivered reagent volume to water amount. Reagent standardisation establishes that factor for the relevant conditions. Dosing performance and endpoint detection remain separate contributors to the result.

In coulometric Karl Fischer, iodine is generated electrochemically from iodide. The measured charge supports calculation of the amount generated when the necessary reaction and current-efficiency conditions hold. There is no volumetric titrant factor to determine in the same way. This does not remove the need for electrical-system checks, appropriate reagents, reference measurements or method verification.

Coulometry is often considered for small absolute water amounts; volumetry is often convenient for larger loads and flexible sample introduction. The useful boundary depends on sample mass, water fraction, background, reagent capacity, dosing or charge performance and required uncertainty. A small portion of a wetter sample can introduce less water than a large portion of a drier sample.

Do not choose only from a catalogue concentration range. Ask whether a representative portion can be transferred reproducibly, whether its water is accessible and whether its matrix overloads or changes the cell medium. Establish an operating window that links sample amount to expected water load and acceptable performance, then challenge its edges during method development.

Make sample introduction part of the analytical method

Liquid transfer must deliver a known representative portion without uncontrolled water uptake or loss of volatile sample. Choose a compatible, sufficiently tight transfer device and justify whether mass or volume is used. If converting between them, establish the density and temperature information needed. Avoid assuming that a nominal syringe graduation proves the amount actually delivered.

For solids, distinguish visible dissolution from complete water extraction. A dispersed solid may release water effectively, while a visually clear solution does not exclude a chemical interference. Define particle preparation, mixing and extraction conditions through evidence. Any grinding, warming or prolonged exposure can alter the original water content and therefore needs evaluation.

An oven can separate released water from a matrix that is unsuitable for direct contact with the cell. A carrier gas transports the evolved vapour to the titration system. Evaluate sample-vial background, gas and transfer-path contributions, water-transfer efficiency and the release profile. Temperature and time must release the relevant water without generating additional water through decomposition.

Margolis and colleagues showed in their 2004 study that volatile products from certain refined oils could generate an apparent water response in the reagent. An oven therefore does not guarantee freedom from chemical interference. The study supports investigating this mechanism; its temperatures, reagents and conclusions are not a ready-made method for pharmaceutical samples.

Document the basis for blanks and corrections before routine use. A cell-background measurement, an empty oven vial and an extraction-solvent blank answer different questions. Use preparation and exposure conditions that represent the intended operation. Do not subtract a convenient blank to force agreement with a specification.

Evaluate the reagent system, closed cell and matrix chemistry

Compatibility concerns the complete combination of sample, solvent or extraction medium, titrant or electrolyte, electrodes and cell configuration. A reagent suited to one matrix may be unsuitable for another. Confirm the approved combinations and operating instructions; changing only a product name in the method is not sufficient evidence of equivalence.

Potential problems include reactions that consume iodine independently of water, reactions that generate water, slow extraction and matrix accumulation that changes the medium. Some carbonyl-containing matrices can react in certain media. Treat chemical structure as a risk signal to investigate, not as proof that every material in a broad class behaves identically.

Track reagent identity, lot, expiry, opening date, storage and in-use condition. Check accessible tubing, connections, closures and drying components against the permitted operator checks. A closed cell reduces exchange with ambient air but does not establish that every connection is tight or that every introduced component is dry.

Use the safety data and approved laboratory controls for the actual reagents. Account for solvent vapours, corrosive or harmful chemicals, sharps, glass breakage and waste. Oven and gas configurations add heat and pressure hazards. Do not improvise reagent mixtures, bypass protective devices or open electrical assemblies to investigate an analytical anomaly.

Separate titre, reference checks, qualification and method performance

A useful reference has an assigned water value appropriate to its intended use, with documented uncertainty or other relevant value information, traceability, handling requirements and stability. Select a form and water load compatible with the configuration. Read the certificate and instructions; a reference suitable for direct cell addition may not challenge the entire oven or extraction path.

For volumetry, establish and maintain the reagent titre using an approved standardisation procedure. Specify its units and how the software uses it. A satisfactory titre does not alone verify the dosing system, balance, endpoint settings or sample extraction. Define an independent performance check when justified by the control strategy and applicable procedure.

For coulometry, verify response with an appropriate water reference and retain relevant charge, endpoint and background information. Qualification and calibration of applicable instrument functions remain necessary according to the quality system. A theoretical relation between charge and iodine is not a substitute for evidence that the assembled system works correctly in use.

Blank is a defined comparison; drift is a background response evolving with time. Neither is interchangeable with the sample water content. Record the software correction mode, its basis and its application interval. A stable but large background can still compromise a small water determination, while an unstable background makes a simple correction difficult to defend.

EDQM clarifies that the amount mentioned in its 2.5.12 suitability exercise concerns that exercise, not a universal lower boundary for the technique. Its separate FAQ distinguishes the example for method A from method B. Consult the applicable full chapter for the actual test. Do not transfer an example’s additions, limits or calculation to every KF method.

Under the VIM, calibration, verification and adjustment have different meanings; titration is the analytical operation. Qualification documents fitness of equipment for intended use, while method validation or verification addresses the procedure in its matrix and setting. EU GMP Chapter 6, Annex 15 and ICH Q14 support a connected control strategy, not substitution of all these activities by a single passing standard.

Investigate drift and unexpected results sequentially

First preserve the original curve, raw response, result, blank, drift history, reagent state and settings. Describe whether the problem occurs before sample addition, only after a particular matrix, after a reagent change or during one transfer technique. Check masses, units, titre selection and correction settings before assuming a leak or instrument defect.

Follow an approved sequence that changes one relevant factor at a time. Establish the conditioned-system background, compare a suitable reference, then examine the introduction blank and a representative sample. If the system is abnormal before sample addition, inspect permitted external connections and consumable condition. If only the sample is abnormal, examine handling, release and chemical compatibility.

Hypothetical qualitative case — poorly soluble sample. A water reference introduced directly into the cell passes, but a solid gives variable responses that continue after the normal endpoint. This does not establish an instrument failure. Investigate release/extraction and endpoint suitability under a development or investigation plan. Compare justified preparation options; evaluate an oven only if release without decomposition and transfer can be demonstrated.

Hypothetical qualitative case — moisture introduced during handling. A similar sample workflow gives a larger response after longer open handling, while protected transfers and the direct reference remain consistent. Compare matched handling blanks and sample exposures without changing reagent and method simultaneously. This pattern supports a handling contribution, but does not prove that the sample’s original water content is acceptable.

Neither example contains experimental data, acceptance limits or a prescribed number of repeats. The first branch needs method evidence; the second needs a controlled transfer technique and an impact assessment. Where a required control fails or the result is suspect, apply the site’s deviation or OOS process. Do not keep repeating measurements until one passes or discard inconvenient results.

Escalate when the checks exceed operator authority, a safety concern appears or the evidence does not distinguish causes. Provide the complete sequence, blank and drift records, reference certificate, lots, maintenance history and the exact sample-introduction configuration. After an intervention, justify the checks needed to return to use and assess potentially affected earlier results.

Record a selection decision that survives routine use

Build the user requirements around representative samples and decisions. Evaluate the entire workload: achievable reporting performance, sample preparation, solvent and waste handling, operator exposure, consumables, maintenance, software records and service continuity. An attractive analysis time has little value if the preparation is unstable or the raw data cannot be reviewed.

The following GuideGxP checklist is a planning aid, not an additional regulatory standard. Tie each item to evidence and an owner before approving routine use.

1. Define materials, water-result basis, intended decisions and applicable controlled methods.
2. Justify the operating window through sample amount, absolute water load and matrix compatibility.
3. Demonstrate introduction, extraction or oven transfer with representative samples and appropriate blanks.
4. Assign responsibilities for qualification, relevant calibrations, titre determination and performance checks.
5. Set justified criteria and frequencies from applicable requirements, risk, stability and observed performance.
6. Retain raw curves, masses, reagent factors, corrections, reference results, user actions and reviewed conclusions.
7. Define change triggers, investigation routes, maintenance boundaries and return-to-use evidence.

Changes in reagent formulation, cell configuration, transfer accessories, sample preparation, software calculation or matrix scope can invalidate an earlier assumption. Assess their impact before use and decide whether verification, partial revalidation or broader work is needed. Do not let a stored method name conceal a materially changed procedure.

The selection is defensible when the laboratory can explain why its complete measurement process is suitable for the actual samples and decisions. Keep residual limitations explicit: inaccessible bound water, suspected chemistry, background sensitivity or untested matrices require defined restrictions or further evidence, even when routine references are satisfactory.

Sources and applicability

Sources checked on 29 September 2026. Compendial public previews and FAQs do not replace the current controlled chapters. NIST training and the cited studies support scientific principles; their example conditions are not universal QC requirements. The decision tables, diagnostic branches and checklist are GuideGxP recommendations. No manufacturer or distributor links are used.

1. USP (2025). General Chapter ⟨921⟩ Water Determination. Public introduction; full current chapter not accessed.
2. EDQM. FAQ on the water amount in the suitability exercise of chapter 2.5.12. Updated 15 September 2021.
3. EDQM. FAQ on suitability for methods A and B of chapter 2.5.12. Updated 15 September 2021.
4. NIST SP 1209 (2017). SIM Chemical Metrology Working Group training, July 2016. Quantitative Water Determination, PDF pp. 56–60. Scientific training material.
5. Margolis SA, Vaishnav K, Sieber JR (2004). Measurement of water by oven evaporation using a novel oven design. 1. Water in water-saturated 1-octanol, coal, cement, and refined oils. Original research; NIST-hosted full text.
6. Suiter C, Widegren JA (2021). Hygroscopic Tendencies of Substances Used as Calibrants for Quantitative Nuclear Magnetic Resonance Spectroscopy. Analytical Chemistry 93(51). DOI 10.1021/acs.analchem.1c04268. Abstract consulted.
7. European Commission. EudraLex Volume 4, EU GMP Chapter 6: Quality Control. Effective 1 October 2014.
8. European Commission. EudraLex Volume 4, Annex 15: Qualification and Validation. Effective 1 October 2015.
9. ICH Q14, Analytical Procedure Development. FDA final guidance, March 2024.
10. JCGM. International Vocabulary of Metrology, VIM 2.39: calibration.

Related Pharma Lab articles

• Analytical instrument qualification and USP ⟨1058⟩
• Qualification, method validation and system suitability