Summary

A COA reports “Moisture: 0.4% by Loss on Drying,” while the purchasing specification requires “Water: NMT 0.5% by Karl Fischer.” Can the material be approved because both values are below 0.5%?

No. The percentages are not comparable until the measured attribute, method conditions, reporting basis, and sample history are aligned.

Karl Fischer titration is intended to determine water under a defined procedure. Loss on drying measures the mass removed under specified drying conditions. That lost mass may include water, residual solvents, absorbed volatile compounds, or material affected by heating.

The quality decision should focus on one question:

Does the reported method measure the attribute that the specification and intended application need to control?

A method name on a COA does not answer this by itself. The reviewer must also examine sample preparation, matrix suitability, drying conditions, exposure during sampling, calculation basis, and whether the tested sample represents commercial supply.

What Each Test Result Can Support

ASTM E203-24 covers volumetric Karl Fischer titration for water determination. A COA entry stating only “KF” still leaves important details undefined. The procedure may involve volumetric or coulometric titration, direct sample addition, solvent extraction, or thermal transfer.

The selected approach must release the relevant water from the material without introducing unacceptable interference.

ASTM E1868-26 covers loss on drying by thermogravimetry. It determines the mass lost under defined test conditions, not the chemical identity of that mass. It is an example of thermogravimetric LOD methodology rather than a universal procedure for every oven-based drying test.

Neither method is automatically better. The appropriate method depends on whether the specification is intended to control water specifically or total condition-dependent mass loss.

Data Distinctions That Prevent False Approval

Several analytical concepts are often compressed into a single percentage on specifications and COAs. They must remain separate during review.

The distinction between HPLC area purity and assay remains relevant because water correction does not convert area purity into quantitative content. High HPLC area purity combined with a water result does not automatically establish the as-is assay of the delivered material.

Hygroscopic Samples Can Change Before Testing

For hygroscopic raw materials, sample handling may alter the measured result before the analytical procedure begins.

Water uptake can occur during:

• opening of the original package;
• sampling from a bag, drum, or liner;
• grinding or homogenization;
• transfer into a smaller container;
• transport to the laboratory;
• weighing and sample preparation;
• repeated opening of a retained sample.

A supplier’s production-line sample, a buyer’s incoming sample, and a repacked evaluation sample may therefore produce different results without either laboratory performing the analysis incorrectly.

The sampling record should establish whether the test represents:

• material before final packaging;
• an unopened commercial unit;
• a laboratory retention sample;
• material after repacking;
• material after a defined exposure period.

This distinction is particularly important for organic chemical intermediates used in moisture-sensitive processes. An as-is water result may affect molar charging, hydrolysis risk, catalyst performance, or batch reproducibility.

Karl Fischer Still Requires Matrix-Specific Suitability

Selecting Karl Fischer does not prove that all relevant water was measured.

The reviewer should determine:

• whether the procedure is volumetric or coulometric;
• whether the sample is added directly, extracted, or thermally transferred;
• whether it dissolves or disperses adequately;
• whether entrapped or less accessible water is released;
• whether the matrix reacts with the reagent system;
• whether blank contribution is controlled;
• whether sample size and repeatability are appropriate near the specification limit.

A low KF result may reflect genuinely low water. It may also result from incomplete water release or unsuitable sample preparation.

A more useful supplier question is not simply whether the method has been “validated.” The reviewer should ask what evidence shows that the selected preparation releases the relevant water without material interference.

Loss on Drying Is Defined by Its Conditions

An LOD result cannot be separated from the procedure used to generate it.

Relevant conditions may include:

• sample quantity and physical form;
• sample-layer thickness;
• drying temperature;
• drying time or endpoint;
• atmospheric or reduced-pressure conditions;
• cooling and weighing procedure;
• protection from moisture reabsorption;
• observations of softening, oxidation, sublimation, or decomposition.

Two COAs may both report “LOD: 0.5%” while using materially different procedures. Comparing only the percentages may amount to comparing different quality attributes.

Thermal instability creates an additional risk. A mass decrease may include decomposition or sublimation rather than removable moisture. A mass increase caused by oxidation may also distort the apparent drying result.

How to Investigate a KF–LOD Difference

A difference between KF and LOD should be treated as an investigation signal, not a diagnosis.

Subtracting the KF result from the LOD result does not quantify residual solvent. The difference represents unexplained mass loss until a suitable analytical method identifies and quantifies its components.

The Reporting Basis Must Match the Decision

A material may be reported with:

• HPLC area purity;
• assay on dried basis;
• water by KF;
• loss on drying.

All four results may be technically valid while answering different questions.

HPLC area purity does not establish weight percentage. Dried-basis assay does not describe the exact composition of the delivered material. KF does not measure every volatile component. LOD does not identify what was lost.

The purchasing specification should define:

1. the attribute being controlled;
2. the method used to generate the result;
3. whether the value is reported as-is or on a corrected basis;
4. how the result will be used in release, dosing, or process calculations.

A specification stating only “Moisture ≤0.5%” leaves unresolved whether the limit applies to actual water, total drying loss, or another operational result.

Approval Risk Changes by Decision Stage

The same moisture evidence should not be used unchanged for exploratory work, trial production, and bulk procurement.

Exploratory Sample

A preliminary sample may still be used for reaction screening, formulation direction, or method development.

It should not receive formal approval when:

• the test item is stated only as “moisture”;
• LOD is reported against a KF requirement;
• the sample was repacked without a controlled exposure history;
• the value is typical rather than batch-specific;
• the reporting basis is undefined.

Trial Production

Trial production introduces exposure that may not exist during small-sample testing. Material may be opened in a warehouse, staged before charging, sampled repeatedly, or transferred from a larger package.

A small sealed bottle can therefore pass while the production-handling condition remains unverified.

Method alignment is particularly important when water affects:

• reaction charging;
• hydrolysis-sensitive steps;
• catalyst activity;
• mixing or dissolution behavior;
• storage stability;
• downstream drying load.

The measured result must be connected to the process effect under review. A numerical limit copied from another material does not establish that connection.

Bulk Procurement

A compliant evaluation sample does not demonstrate continued commercial control.

Bulk approval should be held when:

• commercial batches use a different release method;
• results are reported on a different basis;
• the sample received additional drying or special packaging;
• production-scale packaging has not been assessed;
• supplier and incoming methods show unresolved systematic differences;
• process changes may alter water or other volatile components;
• representative commercial-batch evidence is unavailable.

The decisive question is whether the same attribute will remain controlled through production, packaging, transport, storage, and incoming testing.

Questions That Can Change the Approval Decision

Supplier communication should focus on evidence that changes the approval status.

The supplier does not need to disclose every internal laboratory detail. It should provide enough information for the buyer to understand what was measured, how the sample was handled, and whether the result supports the intended decision.

Why the Greater Risk Is Controlling the Wrong Attribute

Quality reviews sometimes reduce the issue to choosing between a more selective method and a simpler operational test. That comparison is incomplete.

A precise KF result has limited value when the process requirement concerns total volatile loss. A reproducible LOD result is insufficient when water specifically drives hydrolysis, stoichiometric error, catalyst sensitivity, or storage instability.

Using both tests does not automatically strengthen the evidence. Each result needs a defined purpose and an expected relationship to the material risk. Without that framework, the buyer receives two percentages but still lacks a defensible approval rule.

The overlooked problem is often specification design. One party may use “moisture” to mean water while another uses it to mean drying loss. The apparent laboratory disagreement may begin with an undefined quality attribute.

For R&D, quality, production, and procurement teams, method alignment should occur before the sample result is interpreted. Waiting until an incoming commercial batch fails creates a more difficult investigation because method differences, sample exposure, packaging, and process variation must then be separated at the same time.

A complete COA cannot compensate for a method that measures the wrong attribute. The document may be present and the number may be precise, yet the available evidence may still be insufficient for approval.

Final Approval Principle

Karl Fischer and loss on drying should not share a specification limit merely because their results are expressed as percentages.

Approval requires confirmation that:

• the method measures the intended quality attribute;
• the matrix and sample preparation support the result;
• the reporting basis matches the purchasing specification;
• sampling preserves the condition being evaluated;
• commercial packaging represents the qualified sample;
• the same control continues through trial production and bulk supply.

When a hygroscopic raw material requires alignment of the water method, reporting basis, sampling procedure, packaging condition, and commercial-batch controls, buyers can submit a specification-focused RFQ to ChemicalCell before formal qualification.