Determination of deuterium content in components of water-organic solutions by a combination of the NMR 1Н and 2Н methods
The composition of any water-organic solutions (WOS), for example, alcoholic beverages, milk, wines, juices etc. containing H, C, N, O and S elements can be characterised at various hierarchical levels, such as component, fragment, atomic and isotopic ones. A comprehensive analysis of the obtained complementary data makes it possible to identify the components and origin of WOS with a high degree of accuracy. The elemental and component analysis methods were the most widely used for this purpose. The latter predominantly uses different types of chromatography, which, in combination with the mass spectrometric detection, is highly sensitive and makes it possible to identify and quantify the content of macro- and microcomponents in the presence of certified reference samples (CRS). In studying the WOS of an unknown composition, the quantitative nuclear magnetic resonance (NMR) 1Н and 13С spectroscopy, which does not require the use of CRS, offers some real advantages. However, even a combination of these methods does not provide a rigorous proof of the conformity of the WOS components to the declared varietal, agroclimatic (geographical) and/or technological origin, for example, wines- protected geographical indication (PGI) or the protected designation of origin (PDO), mineral water – source area, juices - NFC or reconstituted, milk – whole, adulterated or reconstituted milk. The WOS component composition, if economically feasible, can be simulated by preparing a solution of pure water as well as all necessary and widely available inorganic and organic components. However, a counterfeit, even if the formulation components are carefully manufactured and selected, will not correspond to the genuine product in terms of the content of minor stable isotopes of hydrogen for example (2Н), carbon (13С), nitrogen (15N), oxygen (17О & 18О) and sulphur (33S &34S). It is only isotopic analysis that can make it possible to discover a counterfeit.
The isotope ratio mass spectrometry (IRMS/SIRA) takes the lead in the isotopic analysis due to the highest sensitivity and precision. The initial difficulties in the application of this method were associated with the most difficult stage, the isolation of components of an object and their purification for subsequent authentication. The IRMS toolkit development, for example, the introduction of Gas Chromatography (GC) (for the evaluation of volatile compounds) and High Performance Liquid Chromatography (HPLC) systems (for the evaluation of non-volatile compounds) combined with the corresponding isotope interfaces made it possible to exclude the difficult and time consuming sample preparation associated with extraction and purification of analytes from the matrix. As a rule, the so-called integral ratios of the isotopes 2H/1H, 13C/12C, 15N/14N, 18O/16O, 34S/32S are determined by the IRMS/SIRA method. For example, for the ethyl alcohol molecule CH3CH2OH, only values of 2H/1H (δ2H(D)), 18O/16O (δ18O) or 13C/12C (δ13C) can be measured in the composition of wine, although these isotopes can be characterised by a non-equilibrium distribution in individual fragments of molecules which is typical for of agroclimatic conditions of the geographical region of grapes cultivation, i.e. three and two different values for hydrogen and carbon respectively. In addition, when studying, for example, alcohol-containing distillates, the presence of water, which influences the reliability of the results when measuring isotope ratios of hydrogen (2H(D)/1H) and oxygen (18O/16O) in alcohol, should be taken into account.
When studying the isotopy of light elements by the IRMS/SIRA method only, and in the case of coincidence of agroclimatic (geographical) conditions of plant growth, it is impossible to measure differences between ethanols from carbohydrates of any representative from the C3-pathway of photosynthesis (e.g. grapes, sugar beets, apples, wheat, potatoes, etc.) due to a single mechanism of isotope metabolism in this group of plants. At the same time, the latest developments IRMS/SIRA mass-spectrometry allow selective measurements of the total composition of hydrogen isotopes 2Н/1Н in the methyl and methylene groups of the ethanol molecule after its dehydration (EIM-IRMS/SIRA method).
METHODOLOGICAL BASIS FOR CUMULATIVE NMR SCREENING 2Н/1Н (CUMULATIVE SCREENING NMR /CS-NMR)
In the general case, the NMR spectroscopy of the minor isotopes 2H, 13C, 15N, 17O and 33S provides the measurement of their differential distribution in molecule fragments. In practice, this method can be used only for hydrogen which has a significant range of variation of the natural content (about 150 ppm). In natural waters, it is almost 100 ppm, and its differentiation even in individual fragments of organic molecules (for example, α-pinene) reaches 150 ppm. The non-equilibrium distribution of the hydrogen isotope deuterium (2Н(D)) in organic molecules is an isotope passport which is practically impossible to falsify. However, this method is difficult to use for the isotopes of other elements due to the low variability of their content in natural objects with respect to the precision of measurements of the integrated intensities of NMR signals.
All WOS have the main component in common, water. Therefore, let us first consider the determination of the content of deuterium 2H(D) in it by NMR spectroscopy. If the analysis object is natural water with an insignificant content of organic and inorganic impurities (up to 3.5% in mid-ocean water), a traditional IRMS/SIRA approach can be used, in which the water hydrogen for measurement is transferred in the gas phase to the molecular one when heated to 400° in the presence of zinc  or in another catalytic way . NMR spectroscopy is a direct method for measuring of the absolute content of deuterium in water by quantitative addition of a calibrated compound with the known deuterium content (e.g. tetramethylurea or dimethylsulfone) to the solution. Since the sensitivity of the 2Н NMR method is by six orders of magnitude lower than that for the main isotope of hydrogen - protium 1Н, during the measurement it is necessary to minimise the loss of volume of the water being analysed by reducing the proportion of calibration compounds. To this end, we propose to use special calibration compounds enriched of 2H by 1–2 orders of magnitude for NMR, which makes it possible to reduce their volume fraction in the analyte solution to 2–3%. In , sodium trimethylsilylpropionate-2,2-3,3-d4 (TSP) was used as the calibration compound, the deuterium signals of two methylene groups of which have chemical shifts of 0.60 and 2.00 ppm. In measuring the natural content of deuterium in water, such an internal calibrant is inconvenient because of the need to somehow reduce its concentration by several times in the sample, for example, by diluting by 104–105 times with acetic acid, the hydroxyl group signal of which should also be taken into account. In  another calibration method was proposed, with an external standard. It contains deuterium-enriched water and europium trifluoromethanesulfonate reagent (Eu[CF3SO3]3), which shifts the deuterium signal of the external standard water into a weak field by 2–3 ppm, which excludes its overlapping with the peak of the test sample. However, working with an external standard has a number of drawbacks. For 5 mm diameter NMR tubes, due to the loss of analyte volume by 15–20%, the time of each measurement increases by 30–40% depending on the water content in the WOS. The pairs of tubes (standard and capillary) carefully calibrated by volume are necessary; therefore, it is more difficult to conduct a series of experiments, especially with automatically changing samples.
To determine the water content in the WOS, it is necessary to know its volume quantity, which requires the identification and measurement of the remaining components based on their molar fraction established from NMR 1Н measurements. In the calculation it is necessary to take into account the contribution to the integral intensity of the protons of other components engaged in the rapid exchange with it of the 1Н signal. Without taking these factors into account, the signal area determination in the region of 4.8 ppm. always gives an understated value. Identification and quantitative determination from the NMR 1Н spectrum of the co-component in complex systems allow the calculation of the water content directly; and from the NMR 2Н spectrum, establish its isotopic composition important for research purposes or the authenticity characterisation.
INTEGRATED NMR SCREENING OF THE WATER COMPONENT OF WINES AND SIMILAR PRODUCTS
Let us consider the procedure for determining the content of deuterium water in the WOS by using the example of the expert task concerning quickly screening the authenticity of often adulterated wines. Screening means the operative quantitative analysis of the composition of the object by simple methods without any sample preparation with the possibility of immediately taking a decision .
The specificity of differences in the component and isotope composition of wines and other alcohol products (for example, cognacs, brandies or spirits) is due to the agroclimatic and technological peculiarities of the region of cultivation and processing of grapes including atmospheric precipitation, temperature, intensity of irrigation measures, surface and groundwater composition, soil etc. In particular, all these factors influence the content of minor components in wine and isotopic composition of grapes.
In the analysis of wine, the NMR 1Н spectrum is initially recorded, from which it is easy to calculate the content of water, ethanol (vol. %), glycerol, basic acids such as tartaric, malic, lactic, citric, succinic and acetic as well as identify residual carbohydrates, if any.
In measuring the content of certain minor components, it is convenient to use the high-field satellite of the СН3 ethanol group as a marker which is due to the presence of molecules containing the carbon 13С isotope (1.1%). The quantification of other minor components should be performed in the mode of satellite signal elimination, i.e. by recording the NMR 1Н spectra in the sample radiation at a frequency of nuclei 13С. Fig.1 shows as an example the NMR spectrum of the 1Н Cabernet Sauvignon wine of the 2015 crop (Anapa), in which the ethanol content (13.9 vol. %) is calculated from the integrated intensities of the signals of the ОН, СН3 and СН2 groups (they are partially cut off in Fig.1). Signals taken into account during the calculation of many minor components, which characterise the wine authenticity from the viewpoint of the component composition, can be seen.
Although NMR 1Н spectra directly characterise the content of individual components, they are insufficient to recognize the wine authenticity or its compliance with certain agroclimatic conditions of grapes cultivation (geographic origin). A spectrum identical to the 1Н NMR shown in Fig. 1 can be simulated by an artificial mixture of water, ethanol and minor components. The water signal in the NMR 1Н & 2Н spectra has a characteristic position (4.8 ppm), nor it overlaps with peaks of other components. Its integral intensity in the NMR 2Н spectrum should be used for wine authentication when taking into account the hydrogen contribution of the hydroxyl and carboxyl groups of the remaining ingredients, primarily ethanol. For a strict expert opinion, it is desirable to know the deuterium content in water of the authentic wines of this region and the agroclimatic conditions of the declared harvest year. A more accessible criterion is a comparison of deuterium content in wine water and its averaged value in precipitation and soil waters of the region. It is known that the amount of deuterium in the wine water is 8-15 ppm higher than in sediments because of the peculiarities of isotope fractionation in the molecules of carbohydrates (sugars) of grapes when they are converted to ethanol during fermentation. On the territory of the Russian wine-growing regions, all surface and groundwater contain less than 150 ppm deuterium, so in the water fraction of the wine it should be more than 156–158 ppm. Therefore, the presented NMR method is able to detect the dilution of wine with a mixture of ethanol/water or water as well as a complete falsification of its composition with exogenous components. At the same time, it should be that by studying the isotopy of hydrogen in the water fraction of the product on the basis of the NMR method, it is not possible to detect the addition to musts of exogenous sugars of other plants of the C3 and/or C4 photosynthesis pathways (e.g. sugar beet, sugar cane, maize, sorghum and others) before fermentation (“shaptalization”) or direct introduction into wines of alcohol of the non-grape origin. To do this, it is necessary to measure the content of deuterium in the methyl ((D/H)I) and methylene ((D/H)II) groups of the ethanol molecule isolated from the wine. Therefore, the measuring the content of deuterium in wine water is the first rapid stage of its detailed analysis to ensuring that the most gross counterfeits can be detected.
The algorithm including an integrated analysis of the NMR spectra of protium 1H and deuterium 2H of wine, was first proposed by us for differentiating the wines of Georgia, Russia (the Kuban region) and Western European countries 15 years ago . The main results of the research are shown in Fig.2, where the value 0 is the deuterium content in the international VSMOW standard corresponding to the equatorial water of the world ocean (155.8 ppm). The amount of deuterium in the water fraction of wines in Georgia, Spain, Italy and France is close, but higher than in the wines of the Krasnodar and the Rostov Regions. This is partly due to agroclimatic reasons and seems to be chosen as a criterion in the differentiation of the wines of Georgia and Russia. However, in the latter, the content of deuterium in the water fraction varies widely. Possible causes include the influence of agroclimatic conditions of grape growing and the intensity of irrigation measures as well as the addition of some water or a water/ethanol mixture to the level typical for wines. Manipulations with the composition of some wines were revealed from NMR 1Н spectra by the ratio of the content of ethanol, glycerin and acids in them. Moreover, in three samples, the content of deuterium in the water fraction of the wine turned out to be lower than the values for the surface waters of this region (145–150 ppm), which is unrealistic. Such an approach to the identification of counterfeit of natural wines in the following years was not applied, as follows from the review of the article of 2015 .
The analysis algorithm developed by us was tested in 2017, when wine samples were selected at the VINORUS trade exhibition held in Krasnodar along with the 1st International Scientific and Technical Conference Innovative World of Modern Viticulture & and Wine-Making: Russia (InnoWine: Russia-2017). There were samples of dry wine of Cabernet Sauvignon and Merlot grapes (vintage 2015) from different regions - Krasnodar region (No 2–11), Crimean peninsula - Bakhchisaray (No 12), Volgograd region (No 13) and Spain (No 14). The contents of ethanol and glycerol were measured by the NMR 1Н spectra analysis of samples (Table 1). The contribution to the signals 1Н and 2Н of the water of the hydroxyl and carboxyl groups of glycerin and other minor wine components was not taken into account. Table 1 shows the relative amounts of deuterium in the water fraction of various wines measured by the NMR method. For clarity, they are presented in Fig.3, together with data by the content of 2Н in the tap water of Krasnodar dated 28 April 2017 (No 1). In the water fraction all of tested wine, the deuterium 2Н content varies in the range from 159 to 167 ppm. Two samples of Cabernet Sauvignon wine from the Taman area of the Krasnodar region (No 2 and 6) were found as an exception because of the reduced deuterium content in water which can be explained by the composition of the source water used for irrigation of corresponding grapes.
NMR SPECTROSCOPY OF DEUTERIUM IN ETHANOL OF WINES AND SIMILAR PRODUCTS
The 2Н NMR method was most widely known and applied in conducting evaluations of the deuterium content in the methyl (D/H)I and methylene (D/H)II groups of the ethanol molecule in order to determine its botanical origin.
The characterisation of WOS by the content of deuterium in fragments of ethanol molecules of wine requires its quantitative isolation. For this purpose, a procedure has been developed in detail, since in different countries it is permitted to add exogenous sugars to the grape must, also from the plants of other botanical species. In 1990, the International Organisation of Vine and Wine (OIV) adopted for the first time the SNIF-NMR method  described in  as official in the European Union. To date, a significant base of scientific knowledge on the intervals of the content of deuterium in the methyl (D/H)I and methylene (D/H)II groups of the ethanol molecule of grape origin has been accumulated according to the results of foreign researches. Taking into account agroclimatic (geographical) and technological factors, the minimum and maximum levels for the (D/H)I in grape ethanol are 98.0 and 106 ppm respectively, and for the (D/H)II – 121.0 and 136.0 ppm respectively. These ranges also include ethanol extracted from certain other fruits but not sugar beet. Sugar beet belongs like grape to the group of plants of the photosynthesis C3 pathway) and plants of the photosynthesis C4 pathway (e.g. from corn, sugarcane, sorghum and others) [10–11].
The SNIF-NMR method requires the use of NMR spectrometers, with an operating frequency of at least 400 MHz for 1Н nuclei and equipped with a 10 mm tube sensor and a NMR signal stabilisation system 19F.
The Research and Educational Centre of the Peoples’ Friendship University of Russia uses a modified SNIF-NMR method based on the author’s development  for the standard NMR spectrometers to study the botanical origin of ethanol contained in wines and similar products.
One of the features of the developed method is that the stabilisation of the resonance conditions with the 19F signal is not used. Therefore, the measurement of the deuterium content in СН2- (D/H)II и СН3- (D/H)I fragments of ethanol molecules will have significant errors in height of the corresponding peaks due to a small drift of the signals. However, this practically does not affect the areas which are measured in the quantitative NMR spectroscopy by the integral but not by the height intensity. A distinct advantage of this approach over the method  is that the deuterium content in the hydroxyl group (OH) of extracted ethanol or residual water can be measured by the intensity of the signals which, due to exchange processes, always have a larger width than deuterium (2Н) of the methyl СН3 (D/H)I and methylene СН2 (D/H)II groups. Since the deuterium content in the hydroxyl group is exclusively influenced by the water fraction of the wine, there is no need for a detailed analysis of the 1H NMR spectrum. It is sufficient to measure the density of the analyte solution to estimate the content of residual water in ethanol containing distillate. Another feature of the developed method is the use of dehydrated dimethylsulfoxide (DMSO) as a calibration additive with a deuterium content increased by two orders of magnitude. This reduces in 10 times (from 30 vol. % in the case of the method  to 3 vol. %) reduces the volume fraction of the calibration substance and thereby halves the time of recording the NMR spectrum. The total error of this method ensures the precision of the measured values ± 3%, which is quite enough to estimate the botanical origin of ethanol.
The method was developed and tested on a JEOL JNM-ECA 600 (Japan) spectrometer with an operating frequency for 1Н nuclei 600 MHz and 2Н nuclei 92.102 MHz. The device is equipped with an autosampler and has 5 mm and 10 mm sensors (Fig.4). Calibrated tubes with a diameter of (4.97 ± 0.013) mm and a length of 178 mm were used in the research. Sodium trimethylsilyl propionate-d4 (TSP) was used as an internal standard for the quantitative determination of components to record NMR 1Н spectra, the signal of methyl protons of which does not overlap with the peaks of the analytes. When NMR 2Н spectra are recorded as an internal standard (3 vol. %), DMSO was used with an artificially increased deuterium content by two orders of magnitude. A typical spectrum is shown in Fig.5. The following experiment conditions were selected for recording the spectra on 2H nuclei: pulse 45°, gain 72, shift 5 ppm, scan 11 ppm, 16K points per spectrum, observation time of free-induction decay 2 s, scan number 800, the delay time between pulses is 5 T1 (2D2O) ≥ 5 s. The automatic correction of the baseline, manual phase adjustment and exponential multiplication by the value 2.0 were used to process the spectrum. The table 1 presents the averaged data of five measurements of the deuterium content 2H in each sample, for which the relative standard deviation is 2%.
A new method of a cumulative NMR screening analysis (CS-NMR) was proposed to determine the deuterium content in water of various water-organic solutions including wine and similar products, juices etc. The first stage is the recording of quantitative NMR 1Н spectra to determine the water content in the sample by taking into account other identified components of the solution. Then, the NMR 2Н spectrum is calculated from which the isotope 1Н of water as an important characteristic of origin or authenticity with a control sample. A method that does not require sample preparation is convenient for rapid screening of many water-organic objects. Detailed determination of the isotopic composition of hydrogen of other components of WOS after their isolation can be performed with a standard equipment by a faster original technique.
A modified NMR method is used at the Research and Educational Centre of the Peoples’ Friendship University of Russia, to study the botanical origin of ethanol in wine and other alcohol-containing products which is based on the determination of the deuterium content in the methyl ((D/H)I) and methylene ((D/H)II) groups of an alcohol molecule. The quantitative data on the content of deuterium in the methyl group (D/H)I) of alcohol make it possible to draw a conclusion not only about the botanical form of the plant, but also about the geographic region of its origin, as well as about the nature of the water used in the fermentation of carbohydrates and formation of ethanol. The data on deuterium content in the methylene group (D/H)II alcohol indicate the climatic conditions of the growth of the plant (e.g. grapes).
* The publication has been prepared with the support of the RUDN University Program 5–100.
1. Hachey D.I., Wong W.W., Boutton T.W., Klein P.D. Isotope ratio measurements in nutrition and biomedical research // Mass Spectrometry Reviews. 1987. V. 6. P. 289–328. DOI: 10.1002/mas.1280060203
2. Scrimgeour C.M., Rollo M.M., Mudambo S.M., Handley L.I., Prosser S.J. A simplified method for deuterium/hydrogen isotope ratio measurements on water samples of biological origin // Biological Mass Spectrometry. 1993. 22. P. 383–387. DOI: 10.1002/bms.1200220704
3. Ma X., Deng P., Wang X., Zhang C., Hou X. Direct determination of deuterium of wide concentration range in water by Nuclear Magnetic Resonance // Talanta. 2012. 97. P. 450–455.
4. Dzhimak S.S., Basov A.A., Fedulova L.V., Bykov I.M., Ivlev V.A., Melkonyan K.I., Timakov A.A. Determination of the deuterium concentration in biological fluids by NMR spectroscopy // Aerospace and Environmental Medicine. 2016. 50 (3). P. 42-47.
5. Varcarcel M., Caerdenas S., Gallego M. Sample screening systems in analytical chemistry // Trends in Analytical Chemistry. 1999. 18(11). P. 685–694.
6. Kalabin G.A., Kozlov Yu.P., Rykov R.S., Kulagina N.V., Kushnarev D.F., Rokhin A.V. A new algorithm for identifying natural Georgia wines by NMR spectroscopy // Bulletin of the Peoples’ Friendship University of Russia. Ecology and Life Safety Series. 2003. 9. C. 77–82.
7. Christoph N., Hermann A., Wachter H. 25 Years authentication of wine with stable isotope analysis in the European Union – Review and outlook // BIO Web of Conferences. 2015. 5. 02020. 8 p.
8. Martin G.J., Martin M.L., Mabon F., Michon M.J. Identification of the origin of natural alcohols by natural abundance hydrogen-2 nuclear magnetic resonance // Analytical Chemistry. 1982. 54. P. 2380–2382.
9. EC-Regulation 2676/1990 // Official Journal of the European Community. 1990. 64. L272.
10. Bauer-Christoph C., Wachter H., Christoph N., Rossmann A., Adam L. Assignment of raw material and authentication of spirits by gas chromatography, hydrogen- and carbon-isotope ratio measurements. I. Analytical methods and results of a study of commercial products // Zeitschrift fuer Lebensmitteluntersuchung und-forschung A. 1997. 204. P. 445–452.
11. Perini M., Guzzon R., Simoni M., Malacarne M., Larcher R., Camin F. The effect of stopping alcoholic fermentation on the variability of H, C and O stable isotope ratios of ethanol // Food Control. 2014. 40. P. 368–373.
12. Kalabin G.A., Kulagina N.V., Rykov R.S., Kozlov Yu.P., Kushnarev D.F., Rokhin A.V. Identification of the raw material nature of ethanol by the NMR spectroscopy method // Bulletin of the Peoples’ Friendship University of Russia. Ecology and Life Safety Series. 2003. 7. P. 87–98.