Skip to main content
Nazmul Alam PhD
← Writing
SPME 6 min read

Why the matrix changes how fast SPME reaches equilibrium

Short answer

When a sample contains components that bind the analyte, such as proteins in plasma, only the free fraction partitions into the SPME coating. My 2015 modelling work showed three outcomes: a shorter equilibrium time, which comes from needing less extracted analyte to reach equilibrium and only happens when the binding component is above its dissociation constant; a slower uptake with about the same extracted amount, from local depletion near the coating; and fast-then-slow uptake with a longer equilibrium and less extracted, limited by how fast the analyte unbinds. In complex samples, binding also gives balanced coverage: it holds back hydrophobic analytes so they do not crowd out polar ones on the coating.


Put an SPME fibre into a vial of drug in water, and then into the same amount of drug in plasma. The plasma extraction gives less, which is expected. Less expected is that the time to reach equilibrium also changes, sometimes shorter, sometimes longer. When we started this work, it was not clear under what conditions each one happens.

Working this out was part of my Ph.D. in Janusz Pawliszyn’s lab at Waterloo. This piece explains what we found, in plain terms, because it changes how an SPME method in a complex sample should be set up.

Only the free analyte reaches the fibre

In plasma, blood, milk, or water with humic substances, some of the analyte is bound to matrix components. The SPME coating takes up the free fraction. As the fibre removes free analyte, some of the bound analyte dissociates to replace it, at a rate set by the binding chemistry.

So two processes run at once: diffusion of free analyte to the coating, and release of bound analyte from the matrix. Which one is slower decides what you see.

Three ways a binding matrix changes equilibrium time

In our 2015 Analytical Chemistry paper, we built a mechanistic model of SPME uptake from a sample containing one binding component, validated it against three published data sets, and used an asymptotic analysis to explain three behaviours that had been reported in the literature (Alam, Ricardez-Sandoval & Pawliszyn 2015).

1. Shorter equilibrium time. This looks like faster extraction, and it is not. The model showed that “the decrease in equilibration time is not due to increased rate of extraction but to the requirement of less extracted amount to reach equilibrium when binding matrix is present.” It also happens only under a condition: the concentration of the binding component has to be greater than the dissociation constant, K_D. Below that, the equilibrium time is barely affected by the matrix.

2. Slower uptake, about the same amount extracted. Here uptake is still controlled by diffusion in the sample, but the fibre depletes free analyte in the boundary layer next to the coating, so analyte has to diffuse from further away. The extraction is slower, but the amount at equilibrium is similar to a standard.

3. Fast, then slow, with less extracted. Uptake starts quickly, then slows, the equilibrium time gets longer, and the amount extracted at equilibrium is significantly lower. In this case the release of analyte from the matrix limits the process.

Which case applies depends on measurable things: how much analyte is free, how fast the complex dissociates compared with how fast the analyte diffuses (which also depends on the size of the container), and how much binding component there is. The paper’s point was that with these numbers, you can predict whether the matrix will shorten or lengthen equilibrium without running a long series of experiments.

Two limits, which we stated in the paper: the model handles one binding component, and real blood or tissue needs a multicomponent model. And the discussion was mostly for static samples. With stirring, mass transfer is controlled by the boundary layer at the coating rather than the whole vial, and agitation sets both the rate and the equilibrium value.

Friend or foe: binding and balanced coverage

In a 2018 follow-up, we asked how binding affects recovery across analytes with very different logP values, using matrix-compatible coatings in direct immersion (Alam & Pawliszyn 2018).

The answer was both. Binding can lower recovery of analytes with high affinity for the matrix, because their extraction depends on how fast they are released. But it also gives balanced coverage. Hydrophobic analytes bind strongly, so their free concentration is low and they extract slowly. Because of that, they do not displace the polar analytes from the coating. One device ends up extracting a balanced range of compounds.

Our 2018 review shows the same effect in simulation: analytes with log binding constants above 4.5 gave less than 2% recovery in a 5-minute extraction, and longer extractions raised them (Reyes-Garcés et al. 2018, Figure 2).

A real matrix: haematocrit in blood

Blood adds red cells to the picture. In a 2018 study on haematocrit, the effect on SPME recovery depended on the analyte and the extraction set-up: no effect on salbutamol, a polar drug with low protein affinity; a strong effect on the more hydrophobic propranolol with one device at 400 rpm; and negligible differences for several drugs with another device under high convection (Reyes-Garcés, Alam & Pawliszyn 2018). The analyte’s relative affinity for the matrix and the coating was the deciding factor.

What this means when you set up a method

  1. Do not assume equilibrium time from a standard in water. Measure the uptake curve in the real matrix.
  2. A shorter equilibrium in the matrix is not a faster extraction. It usually means less is being extracted.
  3. Decide whether you need free or total concentration. The fibre sees the free fraction. Total needs matrix-matched calibration, or a calibrant-loaded approach with the distribution constant measured in matrix, as in how SPME is calibrated.
  4. Keep agitation identical between samples and standards, because it changes the boundary layer and the rate.
  5. Use an internal standard that binds like the analyte, ideally isotope-labelled, added before extraction. See SPME without an isotope-labelled internal standard is a guess.

A related point for LC-MS users: binding and balanced coverage are about what the fibre extracts. Whether SPME reduces ion suppression in the mass spectrometer is a separate question. One 2025 study did measure it, for PFAS in meat with a matrix-compatible coating, and found absolute matrix effects between −13.7 and +11.1% by post-extraction spiking (Li et al. 2025). That is one matrix and one coating, and the comparison it gives with SPE comes from a different study, so I would still measure it in your own matrix. I wrote about measuring matrix effect in LC-MS matrix effect: how to measure it.

Sources

  • Alam MN, Ricardez-Sandoval L, Pawliszyn J. Numerical modeling of solid-phase microextraction: binding matrix effect on equilibrium time. Anal Chem 2015; 87:9846–9854.
  • Alam MN, Pawliszyn J. Effect of binding components in complex sample matrices on recovery in direct immersion solid-phase microextraction: friends or foe? Anal Chem 2018; 90:2430–2433.
  • Reyes-Garcés N, Gionfriddo E, Gómez-Ríos GA, Alam MN, et al., Pawliszyn J. Advances in solid phase microextraction and perspective on future directions. Anal Chem 2018; 90:302–360.
  • Reyes-Garcés N, Alam MN, Pawliszyn J. The effect of hematocrit on solid-phase microextraction. Anal Chim Acta 2018; 1001:40–50.
  • Li Y, Zhou W, Jiang RW, Pawliszyn J. Solid-phase microextraction with recessed matrix compatible coating for in situ sampling of per- and polyfluoroalkyl substances in meat. Food Chem 2025; 480:143891.
  • Alam MN, Pawliszyn J. Numerical simulation and experimental validation of calibrant-loaded extraction phase standardization approach. Anal Chem 2016; 88:8632–8639.

If you are extracting from plasma, blood or another binding matrix and the numbers do not make sense, tell me the analyte and the matrix on LinkedIn and I will tell you which of the three cases I would suspect.

Common questions

Does a binding matrix make SPME extraction faster?
It can make equilibrium arrive sooner, but not because extraction is faster. In our 2015 Analytical Chemistry paper, the shorter equilibration time came from needing less extracted analyte to reach equilibrium when part of the analyte is bound, and only occurred when the binding component's concentration was above the dissociation constant.
Why does SPME take longer to reach equilibrium in some matrices?
Two reasons in our model: local depletion of free analyte in the boundary layer next to the coating, so analyte must diffuse from further away, or slow dissociation of the analyte from the matrix, which gives fast initial uptake followed by a slower phase and less extracted at equilibrium.
Does SPME extract the free or the total concentration?
The coating takes up the free analyte. The total can be obtained by calibrating in a matrix-matched sample, and in our 2016 work on the calibrant-loaded extraction phase, free or total concentration came from one experiment depending on how the distribution constant was measured.
What is balanced coverage in SPME?
In complex samples, hydrophobic analytes bind strongly to matrix components, which lowers their free concentration. They therefore extract more slowly and do not displace polar analytes from the coating, so one device can extract a balanced range of analytes with different logP values.
Does haematocrit affect SPME recovery from blood?
It depends on the analyte and the conditions. In a 2018 study, haematocrit had no effect on salbutamol, a polar compound with low protein affinity, but strongly affected propranolol under one set of conditions, while high convection with another device gave negligible differences for several drugs.

Stay current

New essays on analytical chemistry, methods, and industry careers. No noise.

Related reading