Fail to Prepare, Prepare to Fail

Over the last two or three decades, sample preparation has advanced from being a required step ahead of instrumental analysis to become an integral part of the analytical process with a profound influence on both the total time required to complete the analysis and the quality of the results obtained. 

Is this the high point for sample preparation? Will ongoing improvements in liquid chromatography–mass spectrometry (LC-MS) and other techniques make sample preparation redundant or spur further innovation? 

What is current practice in your field?

Serge Rudaz
For complex matrices or at very low analyte concentrations, sample preparation is the most polluting step of the analytical process. Historically, the sample prep methods employed in biofluid analysis, namely conventional solid-phase extraction (see “SPE 101”) and liquid–liquid extraction (LLE), were tedious and labor-intensive. Automation solved these issues but the high sample numbers required from a cost-benefit perspective are not generated in a university research laboratory. 

The complete integration of SPE, achieved in the early 1990’s through the introduction of performant extraction supports, simplified the process and increased speed and automation. Using these, direct injection of biological fluids into LC-MS systems has been adopted by bioanalysts. LLE, which remains the technique of reference for producing clean extracts to be injected into LC-MS, has not lent itself so easily to automation. Just a a handful of approaches, such as supported liquid–liquid extraction (SLE), offer the opportunity.

Both LLE and SPE use large volumes of hazardous organic solvents that are harmful to both humans and the environment. We and others are developing new sample preparation techniques to replace the toxic organic solvents used and to reduce solvent consumption. This will be successful only if the extraction methodology reduces sample preparation time.

Lourdes Ramos
Although large-scale solvent-based techniques are still routinely used in many laboratories for certain analyses, SPE is widely used in routine and academic labs to preconcentrate and purify analytes from fluids and aqueous samples.

For (semi-)solid matrices, classical approaches, such as solvent-shaking and Soxhlet extraction, have been replaced by faster and more cost-effective versions, such as Soxtec.  Enhanced solvent extraction techniques have become markedly more popular in the food and environmental fields, particularly pressurised liquid extraction (PLE).

In PLE, the sample, typically dispersed in a drying or inert sorbent, is packed in a stainless-steel cell and extracted in a closed flow-through system, using solvents at high temperatures and pressures (up to 200 ºC and 20 MPa).  A range of extraction solvents and experimental conditions can be used, and the technique is straight-forward, with a limited number of variables to be optimized. Current trends include the packing of a co-sorbent on top of the sample column, providing in-cell (or on-line) purification of extracts. This approach, known as selective-PLE (s-PLE), yields analysis-ready extracts.

Conventional solvent-based techniques, such as LLE, SLE and Soxhlet extraction, have remained essentially unmodified for more than a century. They require large samples, consume much organic solvent, involve manual manipulation of the sample and extracts, and are difficult to automate. The extracts are often too diluted for direct instrumental analysis and, when used with complex samples, their non-selective nature make additional clean-up of the collected extracts mandatory. SPE and PLE permit faster, less manipulative, greener, more cost-effective, and automated sample treatment that in many instances result in ready-to-analyse extracts. They can be used in hyphenated procedures when the analytical approach is conveniently miniaturized.

What are the major challenges?

Rudaz
Gaining sensitivity is probably the key challenge. This can often be achieved through regular updating of our analytical tools, since improvement is continuous.

Our laboratory is accredited (ISO 17025), so new technologies must be carefully evaluated.  We need to see a significant improvement to justify the amount of laboratory documentation required to introduce a revised methodology.

Ramos
The increasing demand for food and environmental controls means that the diversity of compounds  and matrices that must be accurately determined is constantly expanding. In addition, public and legal pressure is driving more frequent analyses, for example, for trace pollutants in food and environmental samples.  This means that massive numbers of sample-analyte combinations need to be analysed quickly and with ever-more sensitivity.  The main current challenge is for higher throughput, cheaper, and increasingly sensitive analytical methodologies.

Many techniques developed during the last two decades have tried to fulfil these demands. Their success, particuarly their final acceptance by commercial laboratories involved in food and environmental control, have varied widely. On-line SPE, commercialized more than twenty years ago,  is well-established and accepted. However, complete hyphenated systems involving PLE await miniaturization of the system and the development of selective-PLE-based processes. 

Other miniaturized techniques have been incorporated into automatic systems, allowing unattended preparation-plus-instrumental analysis of samples. These include solid-phase microextraction (SPME) and stir-bar sorptive extraction (SBSE), which share advantages and shortcomings. On the plus side is their miniaturized nature, virtually solvent-free operation, simplicity, limited – if any – sample manipulation requirement, adaptability to different liquid samples and extracts, and capacity for automation and hyphenation with conventional separation-plus-detection techniques, namely LC and GC. On the minus side, the limited number of sorbent coatings commercially available restricts potential and legislation sets stringent performance criteria that are difficult to meet. Similar considerations have limited the impact of solvent-microextraction (SME) techniques in routine labs. 

Where are things going in the near-to-medium term?

Ramos
I expect no dramatic changes in the short-term (that is, 1-5 years). Both SPE and PLE will benefit from progress in the synthesis of new materials, including very pure small-size sorbent particles, highly sortive sorbents, coatings based on nanomaterials, new molecularly imprinted polymers (MIPs) and solvents with improved solubilization properties, such as ionic liquids (ILs).

SPME and SBSE have probably reached their maximum development from a technical point of view, so advances will be associated with new coating materials. These will expand the application field for certain promising SME techniques. 

Dispersive-solid-phase extraction (d-SPE), for which QuEChERS is the most popular application method, will enjoy increasing acceptance and use in routine laboratories. A constantly increasing number of applications, the simplicity of the procedures, and the commercialization of ready-to-use tubes will push this trend, while difficulties in automation may represent a real limitation.

Rudaz
Sample pre-treatment, including dilution and protein precipitation, will be improved and faster, even where the quality of the sample is low. But I do not forsee a revolution as the basic principles of extraction remain the same. 

New sample storage formats, such as dried blood spots (DBS), will gain in importance. 

Miniaturization, which reduces the sample volume, analytical time, operating costs and loss of compounds will continue to make inroads as detection sensitivity improves. Both LLE and SPE can be miniaturized, the former by solid-phase microextraction (SPME) or disposable pipette extraction (DPX), the latter using liquid-phase microextraction (LPME) techniques, such as single drop microextraction (SDME), or dispersive liquid-liquid microextraction (DLLME).

From a personal point-of-view, I would like to see more electro-assisted sample preparation.

SPE 101

An appropriate sorbent is packed in a syringe-like cartridge (glass or plastic, depending on the application). The liquid (or dissolved) sample is percolated through the SPE column followed by one of two analytical strategies. In the most common approach, target analytes are selectively retained in the sorbent while other interfering and matrix components are eluted out of the column. The preconcentrated analytes are then eluted with a relatively small volume of an appropriate solvent as a purified and relatively concentrated extract. In the alternative strategy, matrix components are selectively retained on the sorbent while the test analytes are eluted as a clean, but diluted, extract for concentration before instrumental determination. The wide variety of sorbents and formats commercially available allows determination of analytes with divergent chemical structures and polarities for a wide variety of application studies. 

And in the long term?

Rudaz
I see greater focus on the use of chemical processes in an environmentally- and human-friendly way to suit green approaches. The latter goal consists of designing chemical processes to either reduce or eliminate hazardous substances. New analytical procedures have been developed to protect people’s health and to eliminate, or at least reduce, the negative impact of chemical products (e.g., organic solvents) on the environment. This was the case for separation techniques, with the advent of ultra-high performance liquid chromatography (UHPLC), capillary electrophoresis (CE) and supercritical fluid chromatography (SFC), which use significantly reduced quantities of organic solvents. Sustainable and solventless approaches are likely part of the future for sample preparation. 

As briefly discussed, miniaturized sample preparation that drastically reduces solvent consumption is the most obvious route. Toxic solvents should then be replaced with alternative, nontoxic extraction agents. This goal will also be a direction requested in regulations like REACH (the European Community Regulation on chemicals and their safe use).

Ramos
Miniaturisation is a clear (and necessary) trend in food and environmental analysis. Advances in instrumentation made over the past two or three decades has allowed the development and commercialization of a number of separation-plus-detection techniques. These have improved selectivity and sensitivity and are used in many laboratories for routine analysis. In most cases, these techniques involve mass spectrometric detection for final unambiguous determination, allowing a significant reduction in initial sample size without affecting performance. Large volume injections without affecting chromatographic performance is an additional tool that can achieve similar detectability levels. These are key components of the development of miniaturized methods. Another is the coupling of analytical treatments for the preparation of complex matrices for automated and/or unattended sample analysis. Here, selective mass spectrometry-based techniques will  reduce or eliminate tedious multistep purification protocols, simplifying and speeding up the analytical process. This is already being done for liquid samples and fluids while the development of equivalent methods for (semi)solid samples awaits the commercialization of appropriate miniaturized systems.

Although I believe that miniaturization is the way ahead, the representativeness of the subsample used for the analytical determination can become an issue when analysing complex heterogeneous matrices. Careful homogenization of the sample before analysis is essential.

Do you see a bright future for automation?

Rudaz
That is my strong expectation. I remember the development of selective extraction supports, allowing the direct and multiple injections of biological matrices, as an attractive means to reduce sample preparation time. Now, with ten years experience behind us, my conclusion is that the full automation of sample preparation is possible for a limited number of applications, which need to be very well characterized.

Ramos
Full automation of food and environmental samples is absolutely required to meet the increasing demand, as described earlier. 

Full automation has already been achieved for certain techniques and applications, particularly for those related to the analysis of volatiles and liquid samples. For many others (including semi-solid and solid matrices), today’s sample preparation and instrumental techniques could yield novel hyphenated instruments that allow equivalent approaches for many other application studies. To achieve this, more work is required of academia and, especially, of manufacturers in developing new analytical instrumentation.

Is the future sample prep-free?

Rudaz
Yes and no. Resolution has always depended on three steps: sample preparation, separation and detection. The tremendous increase in the power of detection in recent years has modified how we view resolution, arguing against the need for separation.

However, new problems have arisen, most notably the matrix effect, which is the suppression or enhancement of the MS signal. These effects are generally not reproducible and the desired deuterated internal standard is not always available. This compromises quantitative analysis. I therefore believe that sample preparation, even simple, should always be considered in most cases.  (Where qualitative estimation  or low-level quantitation is needed, direct MS injection may be considered.)

This sample preparation may be rather generic since the subsequent steps (separation and detection) can compensate for some loss of selectivity. Rapid, simple, generic, and automated sample preparation methods are on the way.  For example, simple protein precipitation or dilution could easily be automated, which would be particularly useful for samples containing a high concentration of analytes. 

Note also that good sample preparation maintains the analytical platform. The dirtier your sample, the more often you have to clean your device. This is an important aspect of productivity.

Ramos
The selectivity provided by  modern MS-based (and multidimensional chromatographic) techniques has already reduced the requirement for sample purification. Strategies like in-line purification can be incorporated into the instrumental procedure to preserve column integrity. 

I have assisted in the development of a number of analytical procedures that allow direct determination of selected target compounds without (or with minimum) sample preparation. The feasibility of the approach, for example, for lipids or fibre in foodstuffs, or for impurities in drugs, has already been demonstrated. However, for the determination of trace components in complex matrices, such as food and environmental samples, a sample-preparation-free approach is far from being a real analytical alternative, despite being an attractive concept.