Technology Forum: Food and Beverage Analysis

Food and beverage analysis is a major concern in analytical chemistry.LCGC invited three experts from industry: Deepali Mohindra from Thermo Scientific, Paul Zavitsanos from Agilent and Norbert Helle, TeLA GmbH, to participate in this forum to discuss the hottest issues in this area of research.

Q. What is the importance of sample preparation in food and beverage analysis and what do you regard as the most important advances that help the working chromatographer in practice compared to, say, five years ago?

Helle: When LC–MS (liquid chromatography–mass spectrometry) was developed it was thought that it would enable food analysis without sample preparation because the instrument would be able to sensitively and selectively determine analytes — even in the presence of a complex matrix. Practice has shown, both for food, beverage and water analysis, that sample preparation is indeed required if one is to obtain reliable quantitation and the required limits of quantitation. This is especially true for electron spray ionization (ESI), the most widely used LC–MS ionization technique in food analysis. Matrix-induced signal suppression or enhancement are frequently experienced when sample extracts are injected without proper clean-up and such interferences lead to incorrect analysis results. Compared to five years ago, a major advantage is automated solid phase extraction (SPE), which reduces the manual workload significantly. In addition, considerable amounts of solvent can typically be saved, for example, in comparison with previously used methods for the determination of pesticide levels in meat and fish.

Mohindra: Sample preparation is key in food and beverage analysis as most of these matrices are complex and require removal of matrix interferences prior to sample injection. In recent years, the widespread adoption of QuEChERS as a sample preparation technique has offered significant process efficiencies and cost savings for the chromatographic laboratory. Automated and online sample preparation systems that require minimal or no manual intervention have also increased productivity. Another important advancement for chromatographers has been the ability to couple chromatography systems with high resolution accurate mass (HRAM) spectrometers which minimize interferences, especially in complex matrices. In GC–MS (gas chromatography–mass spectrometry), the availability of MS-MS techniques from triple quadrupole systems has provided lower limits of detection with greater precision in heavy matrices than was previously possible. Also, this ability has come at a significant decrease in cost to the laboratory due to advanced technologies. Finally, software for instrument control and data review/reporting has taken centre-stage. The ability of a laboratory to more easily balance personnel resource between LC and GC by using the same software eases one burden on the lab. The ability of a quality control department to easily review data from both techniques on the same software and then easily move that to a LIMS (laboratory information management system) relieves another burden.

Zavitsanos: It has become more important as instruments have become more selective and sensitive. While that may seem counter-intuitive it makes sense. The increased selectivity and sensitivity has allowed reduced sample prep for a given application that then generally leads to a higher throughput demand and that in turn is what drives the continual demand for faster cleaner sample prep.

Q. What special considerations do you place on sample preparation when dealing with trace quantification from different food matrices? Are there some overriding factors to consider?

Helle: Fat removal is one of the most difficult and labour-intensive clean-up processes.Preparation of oily fish samples for the determination of pesticides, fat-soluble vitamins or PAHs is one such example. The required steps are: (1) extraction; (2) gel permeation chromatography (GPC) or SPE; (3) a second fractionation on a clean-up column (SPE); (4) evaporation and finally (5) LC–MS determination. Even though two different SPE clean-up steps are needed, the process can be fully automated and it can be performed on a smaller scale using less solvent and leading to better detection limits. In general: Automation saves a lot of work, saves solvent and associated cost and also improves the laboratory work environment.

Mohindra: The actual sample preparation techniques are similar for most food matrices. Differences arise in sample preparation due to fat content, liquids, solids and semi-solids. For extreme low level analysis, like dioxins, additional sample clean-up can make a significant difference in method performance.

Zavitsanos: The biggest special consideration is whether the analysis is a one-off measurement or a recurring high volume application that will be used by many different groups. The one-off measurement can be labour-intensive and inefficient as long as it is correct in results. The high volume application must be more highly engineered.

Q. With the limits of detection (LOD) in LC– and GC–MS-MS dropping by an order of magnitude every decade, do you see other factors that may limit the acceptance of even lower LODs for food toxins and other undesirable compounds?

A: Helle: Laws and regulations generally require much lower maximum residue levels (MRLs) than five years ago, but automated sample preparation and more sensitive instruments certainly enable the modern food laboratory to reach the required MRLs. Sensitivity of ESI-QQQ LC–MS systems used for target compounds has reached levels that are highly satisfactory for food analysis. However, the past 3 to 4 years have not brought much change in terms of better sensitivity for non-target analytes, i.e. for unknowns. Techniques like ESI–TOF and Q-TOF still haven’t reached the sensitivity required. In food analysis, matrix interference is a great challenge for the analyst. Even if instrumentation is made ten times more sensitive, the matrix background signal will also increase by a factor of ten, leaving the signal-to-noise ratio — and thus the limits of quantitation (LOQ) — almost unchanged. First and foremost, we need efficient sample clean-up to improve both productivity and our limits of quantitation.

Mohindra: Based on historical experience, limits of detection will continue to drop as technologies are further developed. Instrument manufacturers will strive to update and create instrumentation that will exceed required reporting limits set by regulators. It is the job of the analytical chemist to tell the toxicologists and regulators that a compound is detectable at a certain level so that information can be taken into consideration when establishing LODs.

Zavitsanos: As long as selectivity increases at the same rate as sensitivity then there is little consequence. That is the instrument design challenge. If that relationship does not track then the burden is placed onto the sample prep. That will reach diminishing returns in one or two instrument generations.

Q. What unmet needs are there currently in food and beverage testing?

A. Helle: I would especially welcome improvements in the quantitation software for LC–MS. We need software that enables more intuitive operation with a simplified user interface. Another critical matter is the robustness of the ionization and transfer process. A positive development has been new ESI sources such as jet stream and curtain gas flow among others, but more needs to be done in this field. A clear improvement could also bring handsome rewards for instrument manufacturers in terms of increased sales.

Mohindra: The identification of unknowns, unknown compounds that you do not know you should be looking for, remains a major need as food, beverages and ingredients are imported/exported around the world. Unintentional contamination as well as intentional adulteration occurs and often is not identified through standard screening procedures. This type of contamination can pose serious public health risks.

Zavitsanos: The analysis of protein contaminants in protein-based food is a really big gap. The discovery of unknown contaminants in the general case is a constant challenge.

Norbert Helle, PhD, is the owner and general manager of TeLA GmbH, a contract laboratory in the field of food safety analysis. Norbert Helle has more than 15 years experience working in food safety and environmental analysis for various German federal and state agencies, mainly in the field of high performance liquid chromatography–mass spectrometry (HPLC–MS). He chairs two working groups under the German § 64 Foodstuffs and Commodities Act charged with developing food safety analysis methods for phycotoxins and biogenic amines.

Deepali Mohindra is a business development manager at Thermo Fisher Scientific. She joined the company in 2007. Previously, she held a number of positions in the food, nutraceutical and pharmaceutical industries. She holds a BS in Biological Sciences and masters in Business Development. She is active in AOAC.

Paul Zavitsanos is the global food industry manager at Agilent Technologies. In the past 25 years he has worked for the chemical industry at positions ranging from bench technician to senior method development scientist through to bio-analytical lab director in both the USA and Canada. He graduated from Concordia University in 1979 with a BSc specialization in Chemistry

Related content: Harnessing the Power of Multi-Hyphenation in Food Flavor and Odor Analysis