Food Safety: Analysis at the Limits of Detection

April 2013

Hardly a month goes by without the media reporting the discovery of trace amounts of some unwanted chemical in the things we eat and drink. Some enter the food chains by accident, such as pesticides, others are deliberately added to deceive, while some arrive through natural processes. In this article I will be looking at the work of those who investigate them, the analytical chemists.

The amounts detected are at levels of parts per million (ppm) and parts per billion (ppb), which, in the words of the analysts, are nanograms per gram or picograms per gram, respectively. Sometimes levels are even measured as teragrams per gram, which is one part per trillion (ppt).

When explaining these units to the general public, I make a more easily understood analogy,  with time as a scale: One ppm is one minute in two years, 1 ppb is one minute in 2,000 years, and 1 ppt is one minute in 2 million years. With this as a comparison, the response is the one I hope for: admiration of the analytical chemists who measure them. Previous generations would have been unable even to detect at these levels.

There are tens of thousands of papers reporting chemical analyses. In deciding which to review in this article, I have chosen those published in the past five years, and which have accumulated at least 10 citations. These are listed in the accompanying table, by number citations.

Selected Papers on Analysis of Food Residues,
2008 – 2012

(Listed by citations)

Rank Paper Citations
1 S.J. Lehotay, et al., “Identification and confirmation of chemical residues in food by chromatography-mass spectrometry and other techniques,” Trends Analytical Chem., 27(11): 1070-90, 2008. [US Dept. of Agriculture, Wyndmoor, PA] 49
2 J.F. Garcia-Reyes, et al., “Desorption electrospray ionization mass spectrometry of trance analysis of agrochemicals in food,” Analytical Chem., 81(2): 820-9, 2009. [Purdue University, IN] 40
3 V. Gianotti, et al., “A new hydrophilic interactions liquid chromatography tandem mass spectrometry method for the simultaneous determination of seven biogenic amines in cheese,” J. Chromatography A, 1185(2): 296-300, 2008. [U. Piemonte Orientale, Alessandria, Italy] 37
4 V.A. Lemos, et al., “Development of a new sequential injection in-line cloud point extraction system for flame absorption spectrometric determination of manganese in food samples,” Talanta,77(1): 388-393, 2008. [U. Estadual do Sudoeste da Bahia, Brazil.] 33
5 E.G. Amvrazi, N.G. Tsiropoulos, “Application of single-drop microextraction coupled with gas chromatography for the determination of multiclass pesticides in vegetables with nitrogen phosphorus and electron capture detection,” J. Chromatography A, 1216(14); 2789-97, 2009. [U. Thessaly, Volos, Greece.] 30
6 J. Yuan, et al., “Surface plasmon resonance biosensor for the detection of ochratoxin A in cereals and beverages,” Analytica Chimica Acta, 656(1-2); 63-71, 2009. [New Zealand Institute for Plant and Food Research Ltd, Hamilton, New Zealand.] 29
7 M. Kirchner, et al., “Fast gas chromatography for pesticide residues analysis using analyte protectants,” J. Chromatography A, 1186(1-2); 271-280, 2008. [Institute of Analytical Chemistry, Slovak U. of Technology, Bratislava, Slovak Republic.] 26
8 Y.C. Fan, et al., “Ionic liquid extraction of Para Red and Sudan dyes from chilli powder, chilli oil and food additive combined with high performance liquid chromatography,” Analytica Chimica Acta, 650(1); 65-69, 2009. [Zhejiang U., Hangzhou, China.] 24
9 K. Buonasera, et al., “Separation of organophosphorus pesticides by using nano-liquid chromatography,”  J. Chromatography A, 1216(18); 3970-6, 2009. [National Council for Research, Rome, Italy.] 23
10 R. Huskova, et al., “Analysis of pesticide residues by fast gas chromatography with negative chemical ionization mass spectrometry,” J. Chromatography A, 1216(35); 6326-34, 2009. [Institute of Analytical Chemistry, Slovak. of Technology, Bratislava, Slovak Republic.] 22
11 S. Vichi, et al., “Determination of volatile phenols in virgin olive oils and their sensory significance,” J. Chromatography A, 1211(1-2): 1-7, 2008. [U. Barcelona, Spain] 22
12 M.J. Ramalhosa, et al., “Analysis of polycyclic aromatic hydrocarbons in fish: evaluation of a quick, easy, cheap, effective, rugged and safe extraction method”,  J. Separation Sci., 32(20); 3529-3538, 2009. [U. Porto, Portugal.] 20
13 M. Lombardo-Agui, et al., “Laser induced fluorescence coupled to capillary electrophoresis for the determination of fluoroquinolones in foods of animal origin using molecularly imprinted polymers,” J. Chromatography A, 1217(15), 2237-2242, 2010. [U. Granada, Spain.] 13
14 R.J.B. Peters, “Identification of anabolic steroids an derivatives using bioassay-guided fractionation, UHPLC/TOFMS analysis and accurate mass database searching,” Analytica Chimica Acta, 664(1), 77-88, 2010. [Wageningen U., Netherlands.] 10
SOURCE: Thomson Reuters Web of Science

Heading the table of papers is a report from Steven Lehotay of the Agricultural Research Service, US Department of Agriculture, Wyndmoor. Pennsylvania (#1). This paper, written with co-authors in Israel, Canada, the Netherlands, and Spain, is a timely warning of the pitfalls that analysts can stumble into, even when using that most essential of analytical tools: the mass spectrometer (MS). This is generally assumed to be the most reliable, so much so that its results are often unquestioned. But, says Lehotay, unless care is taken, even MS may mislead. Thankfully most analysts are aware of this.


One of the leading chemists in the development of MS has been Graham Cooks of Purdue University, Indiana. Several new types of MS devices have been developed, including hybrid sector/quadrupole and advanced ion trap instruments. His group has made significant contributions to the development of desorption ionization and tandem mass spectrometry as methods of analyzing complex mixtures. One particularly highly cited paper from Purdue reports on the use of desorption electrospray ionization (DESI) MS for detecting the presence of trace amounts of 16 pesticides in food (#2).

Cooks proved that DESI was a viable method, using samples of food that were purchased from supermarkets. In one case, the pesticide ametryn was detected at picogram levels. Such sensitivity allows analysts to meet even the most stringent of regulations that legislators seek to impose.

Organophosphorus pesticides have been, and still are, widely used, and can result in alarming news stories because these chemicals owe their origins to the nerve gases developed by the Nazis during World War II. (Organophosphates per se are not necessarily threatening; DNA is an organophosphate.) Identifying the pesticide ones, individually, poses a problem, but this was the aim of analytical chemists at the Consiglio Nazionale delle Ricerche in Rome. Under the direction of Salvatore Fanali, the team have developed a method of separating these compounds using nano-liquid chromatography (#9). They demonstrated how eight such pesticides could be detected and measured at picogram levels and showed that their method could be used to analyze products like baby foods.

Pesticide residues of 25 different kinds were the subject of research under the direction of Eva Matisová of the Slovak University of Technology, at Bratislava, Slovakia (#10). She showed that narrow-bore column gas chromatography and bench-top quadrupole MS could be used to investigate pesticides in a range of fruits and in lettuce. Her group measured concentrations in the ppb range and showed their results were in line with those limits now being imposed by the EU.

This paper was a follow-up to an earlier one (#7) in which the team focussed on how overestimation of some of these chemicals might occur. They concentrated on ones which they refer to as “troublesome” when it comes to analysis, and for which they noted that overestimations of up to 80% might be recorded. Matisová’s alternative method shows how more repeatable and reliable results can be obtained.

Single-drop micro-extraction and gas chromatography have been combined by Nikolaos Tsiropoulos and Elpiniki Amvrazi of the University of Thessaly, Volos, Greece, in order to detect and measure pesticides in vegetables. The researchers tested their method on tomatoes and courgettes (zucchini). Levels less than 1 ppm were measured with high precision (#5).


Not all food contaminants are man-made. Fungi are capable of producing some of their own, such as mycotoxins, and these may be even more of a threat because they are known to be powerful endocrine disruptors as well as carcinogenic. Ochratoxin A, a particular threat to human and animal health, is the subject of the paper by Yinqiu Wu and colleagues at the New Zealand Institute for Plant and Food Research at Hamilton, New Zealand (#6). Their research came up with a rapid and highly sensitive method based on surface plasmon resonance. Using gold nanoparticles to enhance signals, they achieved a 35-fold increase in sensitivity. The method simply involved extraction with 50% methanol followed by 5% poly(vinylpyrrolidone). Their findings revealed ochratoxin levels of around 0.5 ppb in oats and corn, and around 0.4 ppb in wines.

We generally rely on our sense of smell to detect food that contains the by-products of microbial activity and we avoid eating such food if it smells wrong (although we may seek out foods with such smells — see below). Sometimes contamination is present but cannot easily be detected this way, and a case of this is the formation of phenols in virgin olive oils. Thanks to researchers at the University of Barcelona, Spain, led by Stefanis Vichi, it is now possible to detect nine phenol types even when these were below the odor threshold and before they reached levels that would make the olive oil repulsive to users. Vichi developed a solid-phase microextraction-gas chromatography/MS method for doing this (#11).

Food can be analyzed not only to detect adulteration or the presence of unwanted chemicals, but to check the amounts of key ingredients. The detection of the various biogenic amines that give cheeses their characteristic flavors can be done by a new method developed by Valentina Gianotti and co-workers at the University of East Piedmont, at Alessandria, Italy (#3). These graphically named amines include cadaverine, putrescine, spermidine, and spermine, as well as the more usual suspects, histamine, tyrptamine and tyramine. The levels of these amines provide information about the freshness and maturity of the cheese, but analyzing them is not without its difficulties, chiefly because these amines can vary considerably with cheese and with time.

What Gianotti has done is to adapt further their own HPLC-MS/MS set-up (high performance liquid chromatography and tandem mass spectrometer) with HILIC (hydrophilic interaction liquid chromatography) and thereby analyze samples of a Piedmont cheese known as Castelmagno, which is noted for its unique flavor, aroma, and crumbly texture.


Polycyclic hydrocarbons (PAHs) are known to be carcinogenic and are present in fish products that have been smoked. They are also present in fresh fish, and that is what Maria Oliveira and others have been investigating at the University of Porto, Portugal. They targeted 16 PAHs in mackerel, sardine, and farmed seabass and found levels in the ranges 2.5 to 15 ppb, with the highest levels in the seabass (#12). In a more recent paper — Food and Chemical Toxicology, 50(2); 162-7, 2012 — they have extended their investigation to Atlantic fish and looked at seasonal differences in the level of PAHs and related this to the fat content of the fish and the season of the year. Sardines were found to have more PAHs in winter, whereas for mackerel it was highest in the summer months. However, they say that consuming fish with these ppb levels poses no threat to health.

Ana Garcia-Campaña and colleagues at the University of Granada, Spain, have concentrated on measuring traces of antibiotics of the fluoroquinolone kind, which are used by farmers to keep livestock healthy, and which can enter the human food chain. Those given to cows and pigs are detectable in their milk and meat. The Granada group used capillary electrophoresis in conjunction with laser-induced fluorescence, thereby making use of a natural property of these kinds of drugs. The method has detection capabilities of 0.2 to 10 ppm for milk and 1 to 10 ppm for pig kidney (#13).

Anabolic steroids can sometimes be found in herbal remedies. Ruud Peters of Wageningen University in the Netherlands has published in this field [#14], and his group have used ultra-high performance liquid chromatography, combined with time-of-flight MS, to detect them. Two of the herbal mixtures they tested showed the presence of testosterone itself and four other steroids.

The intense red dyes known as Para red and Sudan can be, and have been, used to adulterate chilli powder in India, and this created international concern in the past decade. It generated research into ways of detecting these dyes and a recent paper on this subject comes from Yan Zhu and co-workers at Zhejiang University, Hangzhou, China (#8). This reports a simple analytical method based on ionic liquid extraction followed by high performance liquid chromatography. The ionic liquid method gave reproducible results and was able to measure amounts of the dyes in the region of 10 ppm.

The paper from Valfredo Lemos and colleagues at the Southwestern University of Bahia, Brazil, concerns the analysis of manganese in foods, and the team has developed a new and fast method based on flame atomic absorption spectrometry (#4). The method was proved to be accurate based on certified reference samples of rice flour and tomato leaves, and then used to determine the manganese levels in a range of foods. These researchers have concentrated on the detection of other metals in foods, and a recent paper of theirs — Food Additives & Contaminants. Part A, 29(11); 1689-95, 2012 — is devoted to the analysis of arsenic in chicken feed, which is known to boost the weight of these birds and was once widely used as such, although now banned.

There is no doubt that analytical chemists are doing a good job, and it is hardly their fault that the media sometimes sensationalizes their findings, no doubt from press releases originating from those who blame “chemicals” for the ills that beset us. However, an ill-informed public clamoring for changes based on the vested interests of these alarmists is not what we need. Of course, we have cause to worry when antibiotics used by farmers to protect farm animals enter the human food chain. However, for many molecules detected in food and drink, we should put the information in context and precede the units in which they are reported with words like “minute,” “miniscule,” and even “infinitesimally small.” Let the scaremongers remove these adjectives, if they must.

Dr. John Emsley is based at the Department of Chemistry, Cambridge University, U.K.

The data and citation records included in this report are from Thomson Reuters Web of ScienceTM. Web of ScienceTM is a registered trademark of Thomson Reuters. All rights reserved.