Common Causes of Underestimation in Quantitative Analysis and How to Prevent Them
Author: Chromatography Mound (expert chromatographer, Welch's contracted writer)
Introduction
When it comes to deviations in quantitative results, many analysts have encountered difficulties in proficiency testing, struggling to pinpoint the cause and determine how to correct it. In fact, the key question is: how do we know if our results are biased?
A common criterion is recovery rates. But sometimes, even if the recovery rate appears normal, the results can still be incorrect—and analysts may even be unable to ascertain it.
In this article, we will discuss how to recognize when errors have occurred, potential causes of deviations in quantitative results, so as to prevent them in the future.
Why Are Quantitative Results More Often Biased Low Rather Than High?
While deviations in quantitative results can go in either direction, common laboratory practices tend to cause underestimation rather than overestimation. The following reasons illustrate why this happens:
1. Low Recovery Rate
The formula of recovery rate calculation essentially reflects two factors: analyte loss and matrix effects.
Analyte loss is related to sample preparation procedures, whereas matrix effects depend on whether matrix-matched calibration curves were used or whether the matrix solvent used for curve preparation originated from the actual sample. Since recovery rate provides an intuitive reflection of these factors, a significantly low recovery rate often signals an immediate problem.
2. Increased Standard Solution Concentration
Preparing a standard calibration curve is a meticulous process, especially for inexperienced analysts, who may take considerable time to calculate the dilution factors for each concentration level. This process becomes even more time-consuming if mixed standard solutions are involved.
A common but problematic practice during handling is leaving the stock solution bottle open throughout the preparation process. Since stock solutions are typically prepared in organic solvents such as acetonitrile or methanol, these solvents evaporate quickly in a fume hood, leading to an increase in stock solution concentration. This, in turn, results in lower measured concentrations. The impact is particularly difficult to notice for those unfamiliar with the method.
3. Relying Solely on Recovery Rate for Assessment
As mentioned above, when the stock solution concentration is incorrect, the recovery rate loses its reference value. In many experiments, spiked recovery tests are conducted using blank samples. If analyte loss during sample preparation is minimal, a high recovery rate can still be observed even when the stock solution concentration is inaccurate. In such cases, even if a blank sample is used to prepare the calibration curve, the quantitative results may still be incorrect.
For example, when the stock solution concentration is correct, the baseline peak area of the sample contributes to the curve intercept, leading to a quantitative result of X1 (see diagram below). However, if the stock solution concentration increases while the nominal calibration concentrations remain unchanged, both response values and the slope of the curve will increase. Meanwhile, the baseline peak area of the sample remains unchanged, meaning the curve intercept stays the same. This leads to a new quantitative result, X2, which is evidently biased. The extent of this bias depends on the slope change, which is directly influenced by the deviation in stock solution concentration.
4. Incorrect Internal Standard Addition Procedure
The purpose of adding an internal standard is to ensure that it undergoes the same sample preparation processes as the analyte (external standard), including extraction, cleanup, heating, and concentration. Then both the analyte and the internal standard will be subjected to chromatographic separation and ionization. Therefore, the internal standard should be added immediately after weighing the sample.
If the internal standard is added at the final stage of sample preparation (e.g., after nitrogen evaporation but before reconstitution), it fails to experience the same treatment as the analyte. Although both the internal standard and the analyte are injected from the same vial, the internal standard in this case only corrects for matrix effects. The quantitative result, however, is still influenced by both analyte loss and matrix effects, leading to potential errors.
Conclusion and Further Insights
Besides the above reasons, there are also other potential causes for underestimation in quantitative analysis, such as natural isotopes, structural isomers, chromatographic resolution, and the selection of characteristic ions. This article shares only a few scenarios based on the author’s own experiences, aiming to help readers refine their analytical approach and achieve more accurate and reliable experimental results.