Section 03 · The Bench E-16
Recovery
How much of what went in came back out, and the correction applied.
When the method takes a cut
Every sorbent tube and diffusive badge captures something, but no sampler captures everything. Some analyte breaks through the bed, some stays bound too tightly to desorb cleanly, some degrades during storage. Recovery is the fraction that makes it through the entire analytical chain — from loading to final detector count — expressed as a percentage of what was originally present.
The measurement is deliberate. A known mass of target compound is spiked onto a sampler under controlled conditions, the sampler is then processed exactly as a field sample would be, and the result is compared against what was added. If 8 µg was spiked and 7.2 µg was quantified, recovery is 90 percent. That gap is not noise; it is the method taking a systematic cut.
That cut matters because field samples carry the same systematic loss. Without a recovery figure there is no basis for correcting reported concentrations upward — and an uncorrected result understates what was actually in the air. Most published methods and laboratory accreditation schemes specify an acceptable recovery range: 70–130 percent is common, though tighter windows apply when precision is critical. A result outside that range flags a problem with the spiking procedure, the sorbent, the desorption conditions, or sample degradation in transit.
Where recovery has been established for a given compound and method, the laboratory divides the raw measured concentration by the recovery fraction before reporting. A raw result of 45 µg/m³ at 90 percent recovery becomes a corrected result of 50 µg/m³. Whether that correction is applied — and whether it is reported explicitly or folded silently into the result — should always appear in the analytical notes accompanying a measurement report. A reader who cannot find it should ask.
Recovery is compound-specific and, importantly, matrix-specific. A sorbent that gives 95 percent recovery for toluene may give 60 percent for a polar compound on the same tube. Humidity during sampling, the co-loading of other compounds and the mass loaded all shift the figure. For this reason a recovery determination run on pure standard in a clean laboratory may not represent field performance precisely; some programmes therefore run spiked field duplicates alongside real samples. The duplicate and spike procedure is exactly this: a parallel check that recovery held on the day, in the conditions, with the actual sample matrix.
The number on the front of the report is already a corrected number, or it should be. Recovery is the evidence that the correction was grounded in something measured, not assumed.
Filed inSection 03 — The Bench