Home > Technical Articles > Assay vs Purity


With this article we would like to clarify our approach to these different tests and explain the difference to our clients. While the word "assay" generally means to measure a property or the concentration of a component, in the context of pharmaceutical quality control the word more specifically refers to the concentration of an ingredient, one which needs to be measured with a high degree of accuracy and precision (eg. 98-102% of label claim). The component may be a major ingredient (~100%) as in an active pharmaceutical ingredient (API) or a minor ingredient (0.1%) as part of a drug product. But the ingredient is expected to be present and the concentration should be

controlled with tight specifications. This requires that the ingredient be assayed using a test procedure with minimal error. An assay may not be highly specific for a particular ingredient. For example, all amino acids are typically assayed using a perchloric acid titration and most chloride salts are assayed using an argentometric (silver) titration. Many monographs require a combination of an assay which is not particularly specific with a related impurities test.

Assay (Meriam-Webster)

Assay (Wikipedia)

Titration Assay


Purity, on the other hand, is defined as the absence of impurity in a substance. The unwanted impurities are measured in the presence of the desired material. There is a big difference between the two types of measurement, analytically speaking.


Typically, trace impurities are present at low concentrations, sensitive and specific tests are required, and errors can be rather large (3-20%) as compared to an assay (2%). Specifications for purity can be >99% or even >99.999%. This would mean that the total impurities that one could detect would be <1% or even <10 ppm.

Purity (Wikipedia)

Trace Impurities Testing
Residual Metals by ICPMS


Let me explain the different approaches. In a trace analysis (which comprises >90% of our business) a contaminant may be present over a wide range of concentrations. Consequently, the test is performed using a calibration curve over a wide range, such as 1-1000 ppb. The lower the concentration, the closer the measurement is to the instrument background (or "noise") level, and the larger the measurement error will be. Typical errors in trace analysis are in the range of 5-25%. One method we often use to minimize error is called an isotope dilution technique, which can reduce errors to <5%.

In a pharmaceutical assay, one generally does not use a calibration curve over a wide range. Instead, a single calibration standard is analyzed at a high concentration near the upper working range of the instrument so that the noise level of the instrument is insignificant compared to the signal. The sample is analyzed at that same concentration. During validation, the linearity of the calibration is documented over a narrow range (e.g. 80-120%). This standard is analyzed multiple times and the relative standard deviation (RSD) of the results must be small (e.g. <2% over six replicates).

Then the unknown sample is tested, which we typically perform by preparing and analyzing the sample in duplicate. There are a few other QC parameters involved, but thatís the crux of the method.

There is also a difference in cost. Generally, trace analysis involves the testing of many samples with one calibration curve while assays often require a dedicated instrument run. Consequently, assays cost much more than a typical trace analysis, usually 6X more for a chromatographic assay.

A recent incident may help to clarify our dilemma. A client purchases cesium chloride with a specification of >99%. The client asked us to confirm the assay value of cesium chloride by testing for Cs by ICPMS and chloride by ion chromatography. The first problem is the specification. Neither of those methods is capable of attaining results with less than 1% error. A titration for chloride would be the best choice for an assay. However, the specification of >99% is probably a purity value, which means we should have tested his sample for impurities, with a specification of "not to exceed 1%"

See also Assay vs Identification Tests

For a quotation...

For a full listing of Exova services, please visit our new Exova web site,  Exova Pharmaceutical Solutions web site and our new Santa Fe Springs laboratory home page.


9240 Santa Fe Springs Rd
Santa Fe Springs, CA 90670
562.948.2225   Fax 562.948.5850


Formerly West Coast Analytical Service (WCAS) and Bodycote Testing Group