SPME for Bioanalysis

By: Craig Aurand, Bob Shirey, Katherine Stenerson, Reporter US Volume 26.4

Craig Aurand, Bob Shirey, Katherine Stenerson


SPME for analysis of non-volatile analytes in biological fluids has been greatly restricted due to limitations of the GC type fiber coatings. These coatings tend to swell when desorbed in organic solvents and eventually strip from the fiber when retracted into the needle. New fibers being investigated contain a solvent-stable binder. In addition, this binder is biocompatible such that proteins and other large molecules do not stick to the coating. This enables small molecules to interact with the phase coating on the silica particle and produce a very clean sample. The SPME fibers used in the analysis discussed here were functionalized with a polar-embedded liquid chromatography stationary phase enabling the extraction of polar analytes from a high aqueous environment.

back to top


The utility of biocompatible SPME fibers is demonstrated for the extraction of the beta-blocking agent propranolol and 4-hydroxypropranolol (4-HP) metabolite from a rat plasma matrix. Extraction efficiency calibration curves for the SPME fibers were generated from buffer solutions and rat plasma across a range from 5 to 100 ng/mL. Typical dosage levels of propranolol have been found to be in the range of 10-100 ng/mL, with toxicity occurring at levels > 2000 ng/mL. Samples were spiked with propranolol and the 4-HP metabolite and extracted under static conditions. Prior to extraction, fibers were conditioned in methanol followed by equilibration in water. Phosphate buffer contained 0.8% sodium chloride to mimic concentrations consistent with blood plasma. Desorption was carried out by placing the SPME fiber into 100 μL of desorption solvent for 1 hour prior to LC-MSMS analysis.

back to top

Reducing Sample Usage

Because SPME is concentration dependent, reduction in sample volume should not affect analyte extraction efficiency. Therefore, reducing the sample size from 500 μL to 100 μL was not expected to result in a 5X reduction in analyte response.

To confirm this, a reproducibility experiment was conducted from rat plasma, using both 100 μL and 500 μL sample sizes. The spiking level of the analytes was 50 ng/mL.

Table 1 shows little statistical difference in the amount of extracted propranolol or the 4-HP metabolite between the 500 μL and 100 μL sample volume, confirming that a smaller sample volume will not proportionately reduce response.

Table 1. Extraction from Rat Plasma, Comparison of 500 μL and 100 μL Sample Sizes

back to top

Linearity Using 100 μL Sample Size

To evaluate the affect of reduced sample volume across a range of sample concentrations, a linearity experiment was conducted for both buffer solution and rat plasma using a 100 μL sample volume and a 100 μL desorption volume. The spiked concentrations ranged from 5 ng/mL to 100 ng/mL for both propranolol and 4-HP metabolite.

Overall recovery for both propranolol and 4-HP metabolite analytes was lower from the plasma when compared to buffer samples (Figure 1). The calibration curve for propranolol in rat plasma was not as linear as the buffer samples with a slightly decreased response at the 100 ng/ mL spiked concentration. The difference in recovery between buffer and plasma supported an observation from previous reproducibility experiments (not shown here).

Figure 1. Linearity of Extractions from 100 μL Samples; Comparison of Buffer and Rat Plasma

back to top

Matrix Extraction Efficiency Comparison

To compare the amount of matrix (phospholipids) extracted during the sample preparation process, an SPME extracted rat plasma sample was compared to a rat plasma sample subjected to protein precipitation. Protein precipitation was performed by adding 100 μL rat plasma sample to 300 μL of acetonitrile, and then centrifuged to remove proteins. Both SPME extracted and protein precipitation samples were analyzed for propranolol and 4-HP metabolite concentration along with multiple reaction monitoring (MRM) transitions to monitor for phospholipids.

Interestingly, the amount of propranolol and 4HP-metabolite were nearly equivalent between the SPME extracted sample and the protein precipitation sample (Figures 2 & 3). In comparison, the amount of phospholipids extracted in the SPME sample was less than 5% of the amount extracted in the protein precipitation sample. The use of SPME offered similar recovery as the protein precipitation method but with much less matrix extraction.

Figure 2. Protein Precipitation, Extracted Phospholipids from Spiked Plasma Samples

Figure 3. SPME, Extracted Phospholipids from Spiked Plasma Samples

back to top


  • Biocompatible SPME fibers with polar embedded particles demonstrate the ability to quantitatively extract propranolol and the 4-HP metabolite from both phosphate buffer and rat plasma at therapeutic levels.
  • Recovery of the propranolol and 4-HP analytes in plasma using SPME was sufficient for calibration down to 5 ng/mL.
  • Reducing sample size from 500 μl to 100 μl did not affect recovery of propranolol and the 4-HP metabolite.
  • The SPME extracted sample contained significantly less phospholipid matrix when compared to protein precipitation samples.
  • Further experiments have demonstrated desorption times can be decreased to 30 min. with no affect on analyte recovery.

back to top

Related Links