While QuEChERS has been a standard sample preparation method for two decades, modern technology affords ever more efficient methods with an even broader scope, enabling more comprehensive food safety testing.
USDA Research Chemist Steven J. Lehotay discusses advancements in high-throughput residue screening. Principally, he describes how the QECh*All "mega-method" permits simultaneous analysis of a wide array of contaminants in the same sample, including complex matrices that are problematic in QuEChERS and other common methods.
Disclaimer: Mention of brand or firm name does not constitute an endorsement by the US Department of Agriculture (USDA) above others of a similar nature not mentioned. The opinions expressed in this discussion are the author's own and do not reflect the views of the USDA.
Where are standard QuEChERS methods struggling the most, and how are labs successfully modifying the approach to handle increasingly complex or high-fat matrices?
All analytical methods, including QuEChERS, have limitations in the analytes and matrices they can handle. Successful sample preparation requires the following conditions to be met:
The test portion must be representative of the original sample.
The analytes must be accessible to the extraction medium and partition fully (and/or consistently) into it.
Matrix components that hinder analysis must be excluded or removed via cleanup.
The analyte or its specific marker must remain stable throughout the process.
The analyte concentration in the final extract must be sufficient for detection that meets the purpose of the analysis.
The QuEChERS method requires thorough comminution to meet points 1 and 2 above. Conditions 2 and 3 are achieved by extraction with 1 mL acetonitrile per g watery sample (yielding ~44% water in the extract) and salting out with 0.5 g 4/1 MgSO4/NaCl per mL extract, which partitions the extract into an upper phase of ~16% water in acetonitrile. Cleanup is typically performed via dispersive solid-phase extraction (d-SPE) with 150 mg MgSO4 and 50 mg each of PSA and C18 per mL of extract, yielding a 1 g/mL sample equivalent in acetonitrile with ~6% water. This method is generally effective for any stable analyte with a logP of about -2 to 8 in various matrices, meeting points 4 and 5 for analysis of the same final extract by both LC-MS and GC-MS.
Working in analytical science?
Register for a FREE Separation Science account to subscribe to the Separation Science Newsletter.
Those conditions in QuEChERS generally exclude major food components, consisting of proteins, fats, and sugars. However, like all methods, it struggles to separate ultratrace analytes in matrices rich in components with similar physicochemical properties (such as caffeine, terpenes, cannabinoids). Labs often address this inherent problem by using sophisticated instrumentation to detect analytes in highly diluted, complex extracts, though this can still compromise analytical scope and necessitate increased instrument maintenance.
In fatty matrices, the effective upper logP range in QuEChERS decreases from ~8 to ~4, depending on the amounts and types of lipids. While consistently lower recoveries can be corrected with internal standards, sensitivity is reduced, and excessive co-extracted fats can harm instruments. Akin to other complex extracts, acceptable analysis relies on sophisticated instruments, dilution, flow-diverting techniques, and more frequent maintenance.
What key instrumental or sample-prep hurdles do commercial testing labs face when consolidating pesticides, veterinary drugs, and environmental contaminants into one workflow?
The QuEChERS method, developed 24 years ago, sacrifices better cleanup for ease of use, and its 1 g/mL equivalent sample concentration reduces extraction efficiency and exceeds the levels required for modern instruments to meet fit-for-purpose standards. Also, now that MS data acquisition is so much faster, the salting-out and cleanup steps in QuEChERS unnecessarily limit the scope in LC-MS analysis. In short, QuEChERS should be replaced with a general "mega-method" that integrates current technologies to enable wide-ranging regulatory monitoring while reducing costs and resource use.
The QECh*All (né QuEChERSER) mega-method overcomes the limitations of QuEChERS in several ways:
Test portion: Uses 1–2 g test portions, representative of 1 kg, achieved via appropriate 1-step comminution.
Extraction: Extracts with 5 mL of 20% water in acetonitrile per gram of watery sample. This upfront dilution improves the efficiency of both extraction and cleanup.
LC-MS analysis: A small volume (0.17 g/mL sample equivalent) of the initial extract is analyzed, thereby expanding the scope to include more polar analytes while eliminating the large amounts of salts that partition into QuEChERS extracts and are detrimental to LC-MS.
GC-MS analysis: The same salts as QuEChERS (0.25 g per mL extract) are used for analytes with logP > -1, yielding drier (<8% water) and cleaner upper layers. Final GC extracts are 0.25 g/mL in acetonitrile with ~3% water.
Cleanup: Replaces d-SPE with more effective flow-through mini-SPE cartridges (20 mg anh. MgSO4, 12 mg each of PSA and C18) for streamlined matrix removal using robotic or centrifugal platforms.
QECh*All removes common food components (e.g. fatty acids, cholesterol) that plague QuEChERS, reducing matrix effects. Routine backflushing of GC and LC columns eliminates ghost peaks and minimizes system contamination (and their associated matrix effects and maintenance needs).
Matrix effects are a major issue. Is matrix-matched calibration the best approach? How can labs better use analyte protectants to improve quantification?
Matrix effects are the Achilles heel of otherwise powerful MS tools. However, injection of less matrix via dilution and better cleanup, coupled with the use of divert valves and backflushing, is an effective solution. I recently completed an interlaboratory study using QECh*All sample prep (including centrifugal mini-SPE cleanup), which showed insignificant matrix effects for nearly all analytes in GC-MS/MS, with analyte protectants for fatty and nonfatty foods. Internal standards compensated for the few affected analytes. One analyst, unfamiliar with the protocol, extracted five matrices and cleaned up 75 extracts in three hours, spending most of the time on labeling and sample entry to build the sequence, and it turned out that the extra time spent preparing 15 matrix-matched calibration standards was unnecessary. Analytical quality control (QC) yielded <5% RSD (n = 83).
In LC-MS/MS, no matrix effects (ion suppression) occurred within the retention time range of >200 pesticides in a 6 minute gradient using 1 µL injection of QECh*All extracts of similarly complex foods. The collaborating chemist, also new to the protocol, spent more time at the store to get the samples than it took to chop, weigh, pour, shake, centrifuge, and transfer sample/extract in the simple protocol. Use of a divert valve and alternating dual columns with backflushing not only maintains cleaner chromatographic and MS systems but also eliminates the re-equilibration time between injections, increasing sample throughput.
For data processing, my lab uses summation function integration (an often-ignored option in all chromatography software) for three ions per analyte to eliminate the time-consuming review of peak integration. Results are checked as a batch of signals and ion ratios (or high-resolution MS ions), with preset tolerances in identification criteria automatically applied to assess signal/concentration, retention time, and ion ratios. Unlike other integration approaches, the summation function also measures blanks the same way, providing the background mean and standard deviation to accurately calculate limits of quantification and identification. My lab recently used this approach to identify ultratrace veterinary drug residues in hundreds of QECh*All extracts of cattle kidneys and livers.
Elimination of matrix effects in both LC-MS and GC-MS is one aspect of a holistic process allowing one or two chemists to report accurate quantifications/identifications for a batch of 30 comminuted samples within a day, or 100 samples by the next morning.
To meet the demand for higher throughput and lower cost-per-sample, what are the key limitations in current commercial automated sample preparation solutions for residue labs?
All aspects of the QECh*All protocol, from start to finish, have been fully streamlined to the extent that current technology allows. Although sample processing is faster and better than ever, it remains the rate-limiting step in routine high-throughput monitoring, regardless of the approach. Labs spend millions of dollars on analytical instruments, and they could spend a fraction of that to buy several food processors, along with twice as many bowls and blades, to comminute a large batch of bulk samples.
When our lab needs to process a large batch of samples quickly, four of us work in parallel, almost like an assembly line. We handle comminution using two food choppers, four bowls and blades, two sinks for washing, and two balances for weighing test portions and extra samples, which are then placed in jars for freezer storage. After that, one or two analysts spend another one to two hours preparing the samples and setting up sequences for overnight analysis. For staff who usually spend the day at a computer, this time in the lab can be a welcome change of pace. It gets them moving and creates space for conversation, music, and a more social atmosphere. There is no reason contract labs could not apply the same model at a larger scale, with administrative or other staff spending a small portion of their time to help in the lab.
You often advocate the use of low-pressure (LP) GC-MS to achieve shorter runtimes. Why hasn't it been widely adopted by routine residue labs, and what's needed for this to change?
Actually, two companies sell a range of pre-connected LPGC columns, and all vendors offer custom LPGC column sets. Moreover, anyone can couple the restrictor/guard to the analytical column, as has been done routinely by chromatographers since the invention of capillary GC. Quite a few contract labs routinely use LPGC to gain a competitive advantage. LPGC provides higher sample throughput, larger batch sizes, better performance, and less downtime, especially when combined with QECh*All. LPGC does not require special hardware, operates like any GC-MS method, and, in practice, uses standard He flow rates of 1-2 mL/min.
A proven LPGC configuration uses a 5 m, 0.18 mm i.d. restriction capillary connected to a 15 m, 0.53 mm i.d., 1 µm film analytical column, plus a 1 m integrated uncoated capillary at the MS transfer line (ITL). The guard/restrictor capillary fits snugly into the 0.53 mm i.d. analytical column, forming a zero-dead-volume connection. The ITL is important because QECh*All or QuEChERS final extracts contain 3 or 6% water, respectively, and air/water exposure at a hot transfer line ruins the stationary phase. Neglect of this critical precaution, along with connection leaks, is one reason LPGC hasn’t caught on more widely, but analysts need to take such precautions in all GC-MS systems.
LPGC typically speeds analyses by 3-fold without significant loss of separation efficiency. A major advantage is that the vacuum prevents solvent condensation at the analytical column head, allowing large-volume injection in standard hot-splitless inlets for lower detection limits and increased ruggedness. LPGC generally provides sharper peaks and more sensitivity than conventional GC using 30 m, 0.25 mm i.d. columns.
However, drawbacks include:
There is little benefit to analyzing volatiles using fast methods.
The vacuum prevents analytical column backflushing.
Thicker films generate more column bleed.
Trimming the analytical column requires leak-free reconnection to the restrictor.
0.53 mm i.d. capillaries are more brittle.
Wider transfer line ferrules can loosen more easily.
In my experience, analyte protectants obviate the need for column trimming, and my labmates and I have learned to avoid the last 3 concerns. Furthermore, newer low-bleed columns have been introduced to reduce the impact of the third point above.
How can high-volume testing labs balance the need for comprehensive extraction with the growing industry push to reduce toxic solvent consumption?
Working analytical chemists have always strived to develop and use cost-effective, safe, and high-performance methods that are fit-for-purpose. However, that last essential “fit-for-purpose” element is often ignored by those who develop methods solely for publication rather than real-world implementation. Nearly all of these type of authors ignore the first condition of sample preparation by using test portions that are too small to provide meaningful results, thereby wasting all the time, effort, and resources involved in doing the analysis at all. How is that green?
In terms of waste, QECh*All sample prep involves two polypropylene tubes, a mini-SPE cartridge, two autosampler vials, and 4 mL of acetonitrile per gram of test portion. Acetonitrile should be minimized for cost and safety reasons, and using sealed tubes and hoods reduces exposure, but its reduced lab use won't affect overall production because it is a byproduct from acrylonitrile manufacturing.
I don’t know the subjective greenness score of QECh*All, or which scoring system to use, but I know its greatest savings comes from its ability to cover the same analytical scope as multiple methods. Implementation of QECh*All can consolidate labs, reducing needs for instrumentation, space, energy, solvents, reagents, and plastics, among other overhead cost-saving measures.
What causes the lag between new method innovations and their adoption in commercial testing, and how can instrument vendors or regulators accelerate this process?
The rapid adoption of QuEChERS (introduced in 2002) was jump-started by the unfulfilled need at the time for a multi-class, multi-residue sample preparation method covering a wider range of pesticides analyzed by conventional GC-MS as well as newly commercialized LC-MS/MS instruments. Partnering with scientists from European and US regulatory labs also helped disseminate the new approach, plus extensive marketing from many companies and two interlaboratory collaborative studies led to its widespread implementation within a decade.
In contrast, the "mega-method" approach, first published by Hans Mol in 2008, initially failed to gain traction due to regulatory resistance against consolidating separate monitoring programs. From the beginning, I recognized that Mol’s optimization experiments led to nearly the same solvent conditions as Mastovska found to work in my lab for a wide range of veterinary drugs. However, it took me years to realize that it would be better to include pesticides and environmental contaminants in that method, and split the extract for GC and LC, rather than try to add veterinary drugs to the QuEChERS method.
Entrenched interests, inertia, and practical limitations still slow wider adoption of the mega-method concept, including QECh*All, but the USDA recently implemented QECh*All (called MEGA) for combined veterinary drug and pesticide residue analysis. It is also being evaluated in other labs, so it has a promising future.
Looking ahead at the next five years, what do you consider the next major paradigm shift or necessary evolution in routine, high-throughput residue screening?
The longstanding trend of “more more more” is very likely to continue, but improved efficiencies can allow more to be done with fewer resources. My hope is that more labs will switch from QuEChERS to QECh*All sample preparation for higher-throughput analysis to monitor more chemical residues in more samples of more types of foods to better ensure food safety and protect the ecosystem from improper agricultural practices.
