A new simple and rapid technology finally unlocks the potential of stable isotopes to distinguish the authenticity and country of origin of natural food and beverage products.

The Need
The food and beverage supply chain can benefit from better real-time techniques to distinguish subtle adulteration in beverage and food products. Is that really honey, or is it diluted with cane sugar? Has corn syrup been added to that fruit juice? Is that pure olive oil or is it mixed with corn oil? Unfortunately, most natural products are incredibly complex in chemical terms, so an exhaustive compound specific analysis is often just not feasible as a high throughput, real-time screening tool. In addition, the new USDA food label law, COOL (an acronym for Country-Of-Origin-Labeling program), took effect March 16 2009. This law underscores the urgent need for a fast-screening tool that could determine not just the authenticity of beverages or other food products, but also provide information about its true geographic origin(s). Is that really 100% Florida orange juice or is it from a mixture of sources?  

The Power of Stable Isotope Ratios
The potential of using stable isotope ratios of the life elements for this purpose has long been discussed by researchers in our industry.[1-3]. Isotopes are two atoms of the same element that have a different atomic mass. The concentration of an isotope in a sample is usually measured as the ratio of the concentration of that rare isotope versus the most common isotope. Therefore, for the common life elements of organic compounds, carbon, hydrogen and oxygen, we are talking about measuring the ratio of carbon-13 to the more common carbon-12, deuterium (mass 2) versus hydrogen (mass 1), and oxygen-18 versus oxygen-16, respectively. These ratios are usually reported as deviations from the ratio found in internationally accepted reference standards and written as δ13C, δD and δ18O with units of ‰ or per ml (parts per thousand). It turns out that nearly every chemical and biological process fractionates stable isotopes leaving a characteristic signature of δ values. For example, in the case of carbon, C3-type plants photosynthetically fix carbon from atmospheric CO2 using the Calvin Cycle, whereas C4-type plants use the Hatch-Slack process. C3 products (e.g., orange, apple, beet) preferentially fix 12CO2 to a greater extent and therefore have much lower δ13C values (around minus 22-28) compared to their Hatch-Slack counterparts (e.g., corn and cane sugar), which are in the range minus 10-12.[4]

Furthermore, the exact conditions under which this process occurs will also influence the isotopic ratios, an effect that is particular pronounced in the deuterium/hydrogen ratio. For natural products, these factors include parameters unique to the growing location, such as local soil characteristics, latitude, groundwater supply, salinity, average temperature and so on. These ratios remain fixed when the product is harvested, until/unless mixing, dilution or other processes are carried out—processes that in turn will leave behind their own isotopic signatures. So, orange juice that has been concentrated and then re-diluted with added water will have very different δD and δ18O values than orange juice from the same location that is fresh or pasteurized.[5]

Limitations of Prior Art
Even though the use of stable isotopes to screen beverages and other food products for authenticity and/or origin has been long discussed, it has never been widely adopted. The main reason for this has been the significant practical limitations imposed by the available analysis equipment, based on isotope ratio mass spectrometry (IRMS). The IRMS is a vacuum-based instrument that is relatively complex to operate and requires a skilled and experienced technician. Moreover, it needs constant calibration and can cost several hundred thousand U.S. dollars when fully configured, not to mention the cost of consumables. It also requires significant and time-consuming sample prep. For instance, water, the ubiquitous major constituent of all beverages, can never be directly measured by an IRMS detector.  It has to undergo either a pyrolysis process to generate hydrogen and carbon monoxide, which are chromatographically separated and then detected to infer hydrogen and oxygen isotopic ratios, or it must undergo two off-line chemical treatments to extract oxygen and hydrogen which are analyzed independently. As a result, stable isotope ratios have not been used for routine high-throughput screening and have been mainly confined to a few specialty labs who’ve offered sample testing on a contract basis, with turnaround times as long as several weeks. To quote directly from the paper by Kelly and Rhodes,[3] “Although these methods are effective they have not been widely adopted because they tend to be labor-intensive, time consuming and do not readily lend themselves to high sample throughput on a routine basis.”  

WS-CRDS – Simple and Fast
This situation has now completely changed with the development of wavelength scanned-cavity ring down spectroscopy (WS-CRDS), a new type of compact, rugged, transportable instrument (Figure 1) that eliminates these drawbacks and provides simple, real-time access to high-precision isotope ratios. With similar or better precision than IRMS, these new analyzers require no special training, elaborate sample preparation or continuous re-calibration. Plus, they dramatically cut the cost per measurement—both capital cost and operating costs—versus IRMS. They can be used in quality control labs, in the field or at-line. And importantly for beverage applications, water samples can be directly injected into the instrument, which simultaneously provides both δD and δ18O in seconds or minutes, depending on the desired precision. 

Figure 1: The new family of isotope analyzers offers several revolutionary improvements over legacy vacuum-based IRMS technology, not least being their compact, rugged packaging.

Roughly the size of a briefcase, these new instruments are based on a high-precision optical technology that was originally developed at Stanford University and then further enhanced by engineers at Picarro. WS-CRDS is a type of gas-phase, high-resolution spectroscopy with incredible sensitivity and precision; commercial trace gas analyzers based on WS-CRDS are now widely proven in greenhouse gas applications both at leading research laboratories and at remote monitoring stations, routinely providing long-term parts per billion sensitivity. Recently this technology has been applied to stable isotopes. These instruments work by measuring the concentrations of isotopologues of carbon dioxide (13CO2 vs. 12CO2) or water vapor (DHO vs. H218O vs. H216O).

So how are these instruments applied to beverages and their natural ingredients? The Picarro L1102-i is an instrument configured to accept liquid water samples, incorporating a simple evaporator between the input port and the sample chamber. It is available with an autosampler enabling around 100 samples to be automatically and quickly analyzed. Individual samples and/or calibration standards can also be manually injected through a septum. Plus, extensive testing proves that it is not even necessary to de-salinate so-called briny samples before injection. Any build-up of salts is very slow and confined entirely to the evaporator that can be periodically cleaned.  

For measuring organic carbon content, the iTOC-CRDS, available from Picarro, is an integrated, small and easy-to-use system that includes an automated front-end accessory that can either perform wet oxidation on the carbon content of juices and other aqueous solutions or combustion of solid/semi-solid organic materials (e.g., fruit, nuts) to generate CO2. The Picarro isotopic analyzer then measures the δ13C value of the liberated CO2. No drying or other purification of the CO2 is necessary, in stark contrast to legacy isotope analyzers based on IRMS. (The isotopic CO2 analyzer was recognized with a prestigious R&D 100 Award in 2008, and the integrated iTOC-CRDS instrument was awarded the Editors’ Choice Silver Medal at the PITTCON 2009 analytical conference and exposition.)

The following simple experiments in the Picarro applications laboratory serve to illustrate how stable isotopes can be used to distinguish even subtle differences between fruit/fruit juice of different origin, specifically, between U.S.-sourced apples of two different varieties, and between U.S.-sourced oranges from two different climates (Florida and California) and between chocolate processed in two different countries.

Apples/Apple Juice
Two apples were obtained from a supermarket; one carried a Pink Lady label and the other a Granny Smith label. To extract water for analysis, a portion of each apple was ground in a plastic cup. Juice samples from each apple were then pipetted into the caps of 5-ml vials. The samples were capped and inverted and then evaporated to dryness by distillation in a hot sand bath. The vials were then turned right side up and the condensed liquid water removed with a standard pipette, and added to a 2-ml vial with an insert and septum cap. These vials (three per apple) were added to the autosampler of the Picarro L1102-i and run using a series of three to five injections for each vial with an analysis time of 8 minutes per injection.  

Figure 2: Raw data for Granny Smith and Pink Lady apples. Each data point is an individual sample vial and is a superposition of several separate injections—note the excellent precision in each case.

As seen in Figure 2, the data show that each apple variety provides a distinct profile, significantly different from the other, especially in δD. This result is even more compelling when the huge simplicity of the method is taken into account. Also visible is the exceptional precision for each vial. The precision for δD was as low as 0.13 per ml and that for δ18O was a low as 0.008 per ml. Isotope standards were measured before, during and after the run and, although the values are calibrated against this VSMOW standard, at no point during the run did the standard data drift by more than 0.1 per ml for δ18O and 1.6 per ml for δD.

Oranges/Orange Juice
 The ability to screen oranges and orange juice is important since terms like “100% Pure Florida Orange Juice” and “Not From Concentrate” are descriptors used to distinguish premium product and premium brands and are governed by the new COOL law. In these tests, two California-grown oranges and two Florida-grown oranges were obtained from a local supermarket.  The juice from each was manually squeezed into vials that were then distilled as with the apple pulp above. The extracted water was then injected into the Picarro L1102-i.

As can be seen in Figure 3, there is a marked difference in the δD values for these differently-source oranges, moreover a difference that could easily be the basis for a fast-screening “pass/fail”-type test.

Figure 3: Water extracted from domestic oranges from two different growing locations—Florida and California—show a marked difference in their δD values.

Chocolate Samples and δ13C
A very different example is highlighted in Figure 4. This shows δ13C values obtained from the organic carbon component of supermarket-obtained solid chocolate samples—two sold as sourced from Mexico and two from Jamaica. These were processed automatically in the iTOC-CRDS, with two data points recorded for each chocolate. Of course, block chocolate and drinking chocolate are highly processed commodities with multiple different natural products, including dairy extracts and other ingredients in addition to the cocoa beans. These ratios therefore reflect a composite isotopic signature of the final chocolate. Any change in the way the beans or other ingredients are sourced, any unreported additives or any change in the processing can be expected to show up in a shift in these ratios and can be used to flag product for further testing. Unlike δD and δ18O values that are sensitive and most useful for indicating geographic origins, the main value of δ13C is in determining the plant type present in a beverage or food product, that is, C3- or C4-type plants. Both of these two samples indicate that they are in the normal range for C3-derived products and not heavily diluted with C4 products, such as from corn. 

Figure 4: δ13C values obtained from the organic carbon component of supermarket-obtained solid chocolate samples.

Summary
The concept of using stable isotopes to verify food source, authenticity and integrity has been around for many years, but cost, time and other drawbacks of the available measurement technology served to stunt the development of this field. However, the combination of simple, portable and cost-effective instruments that now provide real-time data and the need for COOL compliance can be expected to herald much more widespread use of stable isotopes in the food and beverage industry.

References
1. Krueger, H. W. and R. H. Reesmann 1982. Mass Spectrom. Rev. 1:205-236.
2. GuuÄek, M., J. Marsel, N. Ogrinc, and S. Lojen. 1998. Acta Chim. Slov. 45:217-228.
3. Kelly, S. D. and C. Rhodes. 2002. Grasas y Aceites 53:34-44.
4. Jahren, A. H., C. Saudek, E. H. Yeung, W. H. L. Kao, R. A. Kraft, and B. Caballero. 2006. Am. J. Clin. Nutr. 84:1380-4.
5. Doner, L. W., H. O. Ajie, L. S. L. Sternberg, J. M. Milburn, M. J. DeNiro, and K. B. Hicks. 1987. J. Agric. Food Chem. 35:610-612.