The Use of Gel Permeation-Chromatography for the Cleanup of Samples in the Analytical Laboratory

Gel permeation chromatography (GPC) is a subset of a very common technique called size exclusion chromatography (SEC), which involves the separation of molecules by their hydrodynamic volume in solution, commonly referred to as the molecular size.1 This should not be confused with the molecular weight. When the sample is passed through a porous stationary phase, the molecular size of the solvated molecules determines their ability to enter the pores of the resin. Thus GPC is the separation of molecules according to molecular size when using an organic solvent as the mobile phase (Figure 1). The vast majority of laboratories that use this technique use it as an analytical tool to measure the distribution within the sample. Commonly, the analytes of interest are a protein or a polymer.2  

Figure 1 - Schematic of GPC. Those compounds with smaller molecular size are retained by the pores of the resin while the larger compounds are excluded.

This article examines how this simple yet effective technique of separation by molecular size can also be used as a very thorough and robust means of sample cleanup prior to analysis. In this case, GPC is not used as an analytical tool, but rather as a preparative tool in the cleanup process.

In the areas of environmental testing and food and agriculture, when a laboratory must analyze for any variety of toxins or synthetic products such as pesticides, the task of separating the matrix from the analytes of interest becomes very challenging. In many cases, the analytes of interest are very small molecules hiding in the proverbial haystack of proteins, lipids, waxes,3,4 sugars, and any other matrix components in the sample that usually interfere with the chromatography. The good news is that typically these matrix compounds are much larger in size (hydrodynamic volume) than the analytes of interest, and thus are easily separated using GPC.

Figure 2 - UV chromatogram of standard mix used to determine forerun and main run. A laboratory looking for pesticides (Methoxychlor, an organochlorine pesticide, peak 3) will use the chromatogram as a reference. Example: A typical laboratory interested in both pesticides and polychlorinated biphenyls (PCBs) will set the forerun to 18 min, then the main run to collect between 18–43 min in order to ensure all of their analytes of interest are collected. The majority of matrix will wash out before 10 min.

Sample cleanup is achieved by eluting the sample from the column and collecting only the fraction that contains the analytes of interest. The fats and colored compounds are generally larger and thus excluded. They are eluted in the fraction known as the forerun and are directed to the waste stream. At the appropriate time, a switching valve directs the flow into a collection vessel; this fraction is known as the main run. This is the fraction that contains the analytes of interest. Once these have all been eluted from the column, the valve once again switches back to the waste stream and the column is then washed with solvent to clear any remaining matrix in preparation for the next sample. The timing of the forerun and main run is determined by calibrating the system using a UV detector and noting the retention times of the components of a mixture of target compounds (Figure 2). 

It is important to note that following GPC cleanup, the sample must be analyzed further to distinguish the concentration and identity of specific analytes. That is, the GPC will separate the group of pesticides from the fats in the sample, but an analytical instrument such as a gas chromatograph and detector are needed to determine which pesticides are present and in what concentration.5

Figure 3 - GPC Ultra automated sample cleanup and concentration instrument.

Following the collection of the main run, the sample must generally be concentrated, and depending on the analytical method to follow, the solvent has to be exchanged to one that suits the analytical method. Modern GPC instruments, such as the GPC Ultra (LCTech, GmbH, Dorfen, Germany; distributed by Pickering Laboratories, Inc., Mountain View, CA) can perform these two steps automatically (Figure 3).

A good example of this is the analysis of certain pesticides in meat using gas chromatography. To clean the sample prior to analysis, it is first dissolved in cyclopentane, and is then cleaned up with the GPC Ultra using ethylacetate/cyclopentane (70:30).6 The result is a clean, but dilute, sample solution. There are two issues with this: 1) The sample may be too dilute for the analytical instrumentation, and/or 2) the ethylacetate/cyclopentane mixture is not suitable for the analytical method.

Figure 4 - Recovery data graph of pesticides in meat. Recovery rates were between 79 and 110%. The cleaned extracts were then analyzed independently using a gas chromatograph-electron capture detector (GC-ECD) to determine recovery rates.

The GPC Ultra solves this by first evaporating and thus concentrating the sample by means of a vacuum chamber designed to ensure good recovery and reproducibility (Figure 4), and second, by exchanging the first solvent mixture for isooctane and concentrating this solution as well. This is all achieved through the use of a rotating, heated chamber under vacuum. The instrument then measures an aliquot of this solution and optionally places it directly into an injection vial, ready to go on the GC, or places it into another vial or container for further handling by the laboratory. For particularly dirty samples, the collected fraction can be cleaned up even further, such as by passing it through a solid-phase extraction (SPE) or silica column, as in the case with many fish samples. Figure 5 shows an example of the effectiveness of GPC cleanup. 

Figure 5 - Gas chromatograph-nitrogen phosphorus detector (GC-NPD) chromatograms of beef blank without GPC cleanup (top) and after GPC cleanup (bottom).

Table 1    -    Cleanup efficiency rates of GPC and QuEChERS; “Matrix Burden” corrected for 1 g/mL Dried Orange Powder*

The GPC Ultra; its counterpart, GPC Vario; and the GPC technique are not limited to a single matrix or a single type of analyte. The technique is very versatile. It is efficient in the cleanup of fish tissue, vegetation, feeds, and edible oils.7 Spices with high pigment content and aroma substances (e.g., chili, mace) can be processed without any problem, although further cleanup is quite common as the colored and other matrix compounds in these types of samples, while tolerable to a MS detector, can interface with an electron capture detector (ECD).3 Even environmental samples such as soil and sludge8 are routinely cleaned up using GPC. The range of analytes includes pesticides, PCBs, polycyclic aromatic hydrocarbons (PAHs), azo dyes, and phthalates.7 The GPC Vario can even be used for the cleanup of edible oils for the detection of 31 mycotoxins, as shown by some very promising research done by the Technische Universität Muenchen (Munich, Germany).9,10 

Today, there are many other methods of cleaning up a sample prior to analysis, including the effective and very useful QuEChERS (quick, easy, cheap, effective, rugged, and safe) method,11 which is a simple extraction protocol. Several other techniques include microwave-assisted extraction (MAE), matrix solid-phase dispersion (MSPD), SPE, and pressurized liquid extraction (PLE).4,5,8 While very simple and overall very effective, it seems the extraction efficiency for some of these methods varies greatly for fatty or very highly colored matrices. The QuEChERS method, for example, can bring coextractives such as fatty acids, sugars, and carotinoids12 into the solution along with the analytes of interest (Table 1), which can often interfere with the chromatography. Thus, GPC remains one of the most effective cleanup methods for complex samples because it removes the matrix entirely rather than try to work around it. Indeed GPC sample cleanup can be particularly effective when used in combination with one of many extraction protocols.5 The issues of coextractives and baseline noise are less of an issue when using tandem-MS detection because of the very selective nature of these detectors. However, this instrumentation is not available to all laboratories, and there remain some challenges with high lipid content in which lipophilic pesticides may remain in the fatty layer even after the extraction.5 Additionally, there remains the challenge of matrix suppression (generally addressed by using an internal standard [usually isotopic] or matrix-match calibration). In the case of GC-MS analysis, a high level of coextractives will foul the liner which must be replaced.

While there are other methods available to a laboratory for cleaning up samples, GPC remains a powerful and effective means of doing so on a daily basis. This is especially true for highly colored samples or those with high fat content. With the modern, automated, and reproducible GPC Ultra, a laboratory can easily clean up samples and rest assured that its recoveries will be high and reproducible, and its chromatography will be virtually free from matrix interferences, thereby yielding accurate results.


  1. Skoog, D.A. Principles of Instrumental Analysis, 6th ed.; Thompson Brooks/Cole: Belmont, CA, 2006; Chapter 28.
  2. Jordi, M.A.; De Mesa, M. Gel permeation chromatography: conventional versus multiple detection; Chromatography Techniques;
  3. Pardo, O.; Yusa, V.; Leon, N.; Pastor, A. Development of a method for the analysis of seven banned azo-dyes in chili and hot chili food samples by pressurized liquid extraction and liquid chromatography with electrospray ionization-tandem mass spectrometry. Talanta2009, 78, 178–86.
  4. Lambropoulou, D.A.; Albanis, T.A. Methods of sample preparation for determination of pesticide residues in food matrices by chromatography-mass spectrometrybased techniques: a review. Anal. Bioanal. Chem.2007, 389, 1663–83.
  5. Gilbert-Lopez, B.; Garcia-Reyes, J.F.; Molina-Diaz, A. Sample treatment and determination of pesticide residues in fatty vegetable matrices: a review. Talanta2009, 79, 109–28.
  6. Mazawa, D.; Rasmussen, W.; Torma, L.; Aulwurm, A.; GPC Clean-up for Pesticide Residue in Fatty Food. Poster Presentation, AOAC International Meeting, Philadelphia, PA, 2008;
  7. Cavaliere, B.; Macchione, B.; Sindona, G.; Tagarelli, A. Tandem mass spectrometry in food safety assessment: the determination of phthalates in olive oil. J. Chromatogr. A2008, 1205, 137–43.
  8. Smalling, K.; Kuivila, K. Multi-residue method for the analysis of 85 current-use and legacy pesticides in bed and suspended sediments. J. Chromatogr. A2008, 1210, 8–18.
  9. Gottschalk, C.; Barthel, J.; Aulwurm, U.; Englehardt, G.; Bauer, J.; Meyer, K. Application of a GPC-LC-MS/MS Method for the Determination of Various Mycotoxins in Edible Oils. Poster Presentation, Mycotoxin Workshop, 2008; Workshop1. pdf.
  10. Gottschalk, C.; Barthel, J.; Aulwurm, U.; Englehardt, G.; Bauer, J.; Meyer, K. Application of a GPC-LC-MS/MS method for the determination of various mycotoxins in edible oils. LC·GC Feb 1, 2009;
  11. Majors, R. QuEChERS—a new technique for multiresidue analysis of pesticides in foods and agricultural samples. Chromatography Online Feb 1, 2008;
  12. Steinbach, P.; Sinderhauf, K.; Voegler, P.; Schwack, W. Development of Clean-up Procedures to Minimize Interfering Matrix Effects in Residue Analysis. Poster Presentation, European Pesticide Residue Workshop, Berlin, Germany, 2008;

Ms. Rasmussen is Technical Sales Manager, Pickering Laboratories, Inc., 1280 Space Park Way, Mountain View, CA 94043, U.S.A.; tel.: 650-694-6700; fax: 650-968-0749; e-mail: