spike recovery
/ spyk rih-KUV-er-ee /
Suppose you suspect a parcel-delivery service of losing packages. A clever test is to send through a few parcels you know about and count how many arrive. Spike recovery is that test for an analytical method: you deliberately add a known amount of analyte to a real sample, run it through the whole method, and check how much of your addition comes out the other end.
In a spike-recovery experiment you take a portion of the actual sample, add (spike) a known mass or concentration of the analyte, and analyse both the spiked and unspiked portions. The recovery is the extra amount you measured divided by the amount you added, as a percentage. Because the spike rides through the very same matrix as the real analyte, it tests the method under true working conditions.
It matters because it is the most direct check of whether the sample's matrix is sabotaging your result through losses or interference — far more honest than testing a clean standard. The caveat is that a spike sits on the surface and may not behave exactly like analyte naturally bound inside the sample, so good spike recovery is strong reassurance but not absolute proof that real, native analyte is measured correctly.
Soil already containing 12 mg/kg of copper is split; one portion is spiked with a further 20 mg/kg. The spiked portion measures 31 mg/kg, so the recovered spike is 31 − 12 = 19 mg/kg, a spike recovery of 19/20 = 95%.
Adding a known amount to a real sample to see how much the method recovers.
Spike recovery is the standard way to measure recovery, and it underlies the method of standard addition; the key difference is that standard addition uses the spikes to actually quantify the analyte, whereas a spike-recovery check uses them to test the method's trustworthiness.