Sample Preparation: The State of the Art

A panel of experts discuss the latest trends in sample preparation.

This article was written as part of a special series to commemorate 30 years since the launch of LC Magazine, which evolved into LC GC Europe and LC GC North America.

This is certainly an accomplishment that we in the separations community have enjoyed and benefited from, and we wish many more happy years of success. This article is intended to be a follow-up of a sample preparation special report published in 1995 by Ron Majors, entitled "Trends in Sample Preparation and Automation — What the Experts Are Saying" (1). Now, we have asked a new panel of experts similar questions on sample preparation. This is a compilation of what these experts think about the current practice of sample preparation, and where they predict it will be 10 years from now. Just for fun, comparisons to the predictions of the expert panel in 1995 have been made where possible.

Sample preparation remains the single most challenging aspect of chemical analysis. Chromatographic instrumentation, including column technology, instrument reliability, durability, capabilities and most certainly detector technology, continues to advance and move the business of chromatography forward at a steady pace and frequently with an incremental leap. Sample preparation still relies on scientists' basic chemical knowledge of the analytes being measured; the techniques and methods they have to measure them; and the experience to couple this knowledge effectively. Fortunately, advances in chromatography and chromatographic detection have reduced some of the need for complex and tedious sample preparation. But, as our experts reveal, challenges still remain.

How valuable are automated sample preparation techniques, such as supercritical fluid extraction (SFE), microwave extraction, automated solvent extraction (ASE) and solid-phase extraction (SPE)?

In 1995, the overall general opinion of the earlier panel was that sample preparation was a manual technology, considered "low-tech" and assigned to the least trained staff. Now, that is changing. "Sample preparation stations are becoming more and more powerful, and aside from extraction and enrichment they offer features of standard addition, derivatization and so forth," says Pat Sandra from the University of Ghent. "The application of automated sample preparation units with injection (that is, on-line systems) is presently rather limited but will become more and more important in the future." Other experts concur that for automated sample preparation to realize its goals of being commercially advantageous, efficient and "green," samples need to be run in parallel batches, similar to how sample preparation is done using manual labour. If sequential sample handling is the automated mode of operation, it should be very rapid and subsequently integrated with parallel analysis of final extracts to expedite the whole process. All the panelists agree that simple solvent based extractions can be faster and easier for a few samples and sometimes using SPE or SFE just to concentrate, purify or desalt samples in preparation for further mass spectrometry (MS) analysis is the simplest procedure to follow.

Are preset sample preparation methodologies, in kit form, such as QuEChERS (quick, easy, cheap, effective, rugged and safe), decreasing method development time and making sample preparation a more routine, predictable laboratory operation?

These kits are recommended highly by our panelists as excellent tools for routine sample analysis at facilities that measure hundreds of analytes in large numbers of samples. Sergio Nanita, a residue analytical chemist from DuPont Crop Protection, states that the kits "allow laboratory personnel to become very familiar with the analytical steps, thus reducing analyst-to-analyst variability and errors that can occur when changing operations between multiple methods." Better expected precision is an added advantage of the kits.

Analysts should ensure that the kits for matrices and target analytes have passed validation studies. In addition, users should be mindful that the kits contain preweighed materials designed for a specific concentration range of the analytes themselves. Karyn Usher, an Associate Professor in the Department of Chemistry at West Chester University, reminds analysts that if the kits do not meet the exact requirements of sample concentration levels, custom kits can be requested from most suppliers or the content of the kit can be modified to meet specific sample requirements, for example, by adding more sorbent or more salt. Since the technology of QuEChERS was introduced in 2003 (2), additional kits for removal of proteins and phospholipids from serum have also been marketed (3). In 1995, these advanced sample preparation techniques were not available.

The Sample Preparation Expert Panel

Will advantages in detection limits in MS, when used with massive dilution of complex samples and single ion monitoring (SIM), eliminate the need for sample preparation altogether? Are there particular sample types where this just doesn't work?

According to Steve Lehotay, lead scientist at the US Department of Agriculture's Agricultural Research Service Eastern Regional Research Centre, dilute-and-shoot sample preparation methods for many analytes from both simple sample types and complex applications are available using mass spectrometers with exceptional detection limits. However, some analytes do not ionize well and cannot be diluted enough and still meet the detection limit needs of the desired analysis. With current instrumentation, the experts agree that there is a constant trade-off between the ionization capabilities of the molecules and how much dilution can actually be used without losing accurate and precise quantitation. This goes hand in hand with ion suppression and ion enhancement, which need to be evaluated on an analyte-toanalyte basis. Removal of high-molecular-weight material with simple sample preparation, using an automated liner-exchange system and backflush options on MS equipment were recommended by Sandra if sample preparation is reduced or omitted. Agreement among the experts was not obtained about the benefits vs. cost of mass spectrometers to achieve these sample preparation reductions. Generally, academics viewed the most sensitive mass spectrometers as having a "high capital cost" needing a significant up-front investment and a specially trained and expensive operator. Conversely, their industrial counterparts stated that even if a less expensive technology — for example high performance liquid chromatography (HPLC) with UV detection — may be the "logical" choice, using the highly specific mass spectrometer pays for the costly instrument over time because the overall procedures are faster, with shorter sample preparation times and chromatographic runs of only a few minutes. To achieve this investment in state-of-the-art technology that yields improved efficiency and can translate into savings in all laboratories, cost improvement trends in spectrometers with improved ionization capabilities and detector sensitivity must continue. Undoubtedly, they will.

Are there specific sample types where SPE disk and cartridge technologies coexist, or does each have specific applications where it excels?

SPE, whether disk or cartridge, really shines for use with dirty matrices, such as urine (cartridges) and large sample volumes, such as in water analysis (disks). These are classical applications, which Nick Snow, Professor of Analytical chemistry and Separation Science as well as the Director of the Centre for Academic Industry Partnership at Seton Hall University, views as the "bread and butter" for SPE. He reasons that cartridges have high capacity so analytes will not easily be lost or hidden by interferences, while disks make it easy to pass large volumes of sample through quickly. Both disks and cartridges are helpful in enriching or concentrating solutes from a matrix while also eliminating some interferences. Matrix solid-phase adsorption techniques (MSPD), such as QuEChERS, provide the purification step, but not the enrichment step. In 1995, the question asked of the panel was if both SPE disks and cartridges would survive in the future, and the answer was that they would likely coexist. The answer is the same 12 years later: Disks are best suited for concentrating large volumes such as environmental water samples and cartridges are more suited to a wider variety of cleanup applications.

Are specific detectors, such as mass spectrometors, eliminating or greatly reducing the need or use of sample preparation in your laboratory today?

Amgen's Peter Grandsard, who leads the company's Discovery Analytical Sciences Group, comments that they are on a quest to perform "top-down" MS of large, heterogeneous molecules — that is, biologics and full antibodies. The "top-down" technique determines the intact molecular weight of each and every protein identified by peptide mass fingerprinting and peptide mapping. If Amgen's goal is achieved, the need for sample preparation currently associated with "middle-down" or "bottom-up" MS analysis will be greatly reduced. In "bottom-up" or "middle-down" MS, peptides are determined in given proteins, and online databases are used to identify the most likely structures (3). So, the quest to eliminate sample preparation and achieve a better answer remains in some specific fields. In general, as Usher representatively reports, using SIM mode in either GC–MS or LC–MS, most sample types still require some sample preparation, but this type of detection eliminates the need for more complex sample preparation methodologies that would otherwise be required. Laboratories need to prepare samples to the extent that they are clean enough to not foul the instrument and enable an accurate and reliable result to be obtained. W. Jeffrey Hurst, a member of the 1995 panel and a scientist at Hershey Foods Technical Centre in Hershey, Pennsylvania, USA, predicted the start of this trend with a statement that ideally analysts should perform no sample preparation, thus eliminating the use of solvents and reducing potential errors.

What do you consider the best tool in your current toolbox of sample preparation methodologies?

Snow states that solid-phase microextraction (SPME), using both direct immersion and static headspace extraction, along with classical liquid–liquid extraction techniques, is frequently the best place to start in his laboratory. Lehotay concurs; he relies on extraction by shaking with solvents, liquid–liquid portioning and dispersive-SPE cleanup with the options of a variety of sorbents. QuEChERS, cartridge-based SPE and gel permeation chromatography (GPC) are preferred for more extensive cleanup. Usher sums it up for all three laboratories, stating that the most useful for a number of different samples is SPE. "Method development time was often longer than for some of the other techniques, but we were typically able to get the desired result," she said. These views are consistent with the 1995 panel, which asserted that automated or on-line analysis would mimic traditional analysis and that until new analytical techniques were developed, sample preparation would continue to be performed as necessary.

If you had a crystal ball, where would you predict that sample preparation would be in 5–10 years?

Sample preparation, and even chromatography, will become less critical in MS- and nuclear magnetic resonance (NMR)-based methods in the next 10 years, predicts Grandsard. So it is dependent on the evolution and adoption of highend MS and NMR in the biologics arena. He does maintain that sample preparation techniques may be needed more in prototypical, biophysical methodologies such as cryogenic transmission electron microscopy (cryoTEM) and atomic force microscopy (AFM) that will likely have greater ongoing use in the analysis of biologics. For ultratrace-level analysis, DuPont's Nanita suggests that the survival of chromatography and MS for years to come will depend on rigorous methods using SPE, robotics, preconcentration steps and more-sophisticated instrumentation to achieve the most selective and sensitive chemical analysis. He feels that sensitivity improvements in MS will continue as will research on fast methods based solely on MS, making the techniques more rugged, placing them in laboratories around the world for routine sample analysis and replacing some of the classical extraction methodologies currently in chemists' toolboxes. Sandra simply states that application-dedicated sample preparation stations (off-line) will be increasingly developed, as he feels they are much more robust than on-line systems.

These were not the same futuristic viewpoints of the expert panel 17 years ago. However, upon review, most of their predictions have come true. Examine the list below and see which predictions you believe they hit the mark on.

• Greater integration of sample preparation and analysis with information feedback, automated record keeping and greater use of hyphenated techniques.

• Smaller samples and miniaturization.

• Increased standardization, automation and modularization.

• Techniques that use less organic solvent.

• On-line and at-location measurement.

• Less sample preparation because of improvements in selective spectroscopic measurements.

In summary, everything is expected to get smaller, labonachip is expected to be more widespread, kits are expected to be more broadly applied to a wider variety of sample applications, automation will increase and sample preparation will, in some cases, be eliminated because of more selective detectors. The panel predicted many of the same things with a few unexpected changes because of cautiously evolving technology.

Mary Ellen P. McNally is a Technical Fellow at the DuPont Crop Protection, Newark, Delaware, USA.

References

(1) R.E. Majors, LCGC North Am. 13(9), 742–748 (1995).

(2) M. Anastassiades, S.J. Lehotay, D. Stajnbaher and F.J. Schenck, J. AOAC Int. 86, 412–431 (2003).

(3) Wako Diagnostics, Clinical Diagnostic Reagents, Richmond, Virginia, Catalog No. 433-36201.

(4) I.S. Krull, A.S. Rathore and S. Kreimer, LCGC North Am. 6(29), 502–514 (2011).