ELECTROCHEMICAL AUTHENTICATION METHOD

Abstract

The present invention relates to a method for product identification comprising subjecting a sample to cyclic voltammetry, wherein the sample is subjected to a plurality of voltammetric cycles to obtain a data set for each cycle and wherein the data sets comprise data points; and comparing the data set for each cycle with a data set for a corresponding cycle of at least one known product to determine whether the sample is the known product. The present invention further relates to a method for determining a profile for a known product, which may be used in determining the identity of a sample.

Claims

1. A method for product identification comprising: a) subjecting a sample to cyclic voltammetry, wherein the sample is subjected to a plurality of voltammetric cycles to obtain a data set for each cycle, wherein the data sets comprise data points; and b) comparing the data set for each cycle with a data set for a corresponding cycle of at least one known product to determine whether the sample is the known product.

2. The method according to claim 1, wherein the data set of a first cycle of the sample and the data set of a first cycle of the at least one known product are omitted from step b).

3. The method according to claim 1, wherein the sample is subjected to at least three voltammetric cycles.

4. The method according to claim 1, wherein the sample is subjected to five voltammetric cycles.

5. The method according to claim 1, wherein the or each voltammetric cycle is cycled through an oxidative voltage range and/or a reductive voltage range.

6. The method according to claim 1, wherein the or each voltammetric cycle is cycled between +1V and −1V.

7. The method according to claim 1, wherein step b) further comprises determining whether the data set for each cycle of the sample falls within a predetermined upper threshold and a predetermined lower threshold of the data set for the corresponding cycle of the at least one known product.

8. The method according to claim 1, wherein step a) further comprises subjecting the sample to a plurality of non-identical working electrodes for each voltammetric cycle to obtain a data set for each cycle for each electrode, and wherein step b) further comprises comparing the data set for each cycle for each electrode to a data set for a corresponding cycle for a corresponding electrode of at least one known product.

9. The method according to claim 8 wherein step b) further comprises determining whether the data set for each cycle for each electrode of the sample falls within a predetermined upper threshold and a predetermined lower threshold of the data set for the corresponding cycle for the corresponding electrode of the at least one known product.

10. The method according to claim 1 for determining whether the sample is an authentic product.

11. A method of determining a profile for a known product, said method comprising: a) subjecting the product to cyclic voltammetry wherein the product is subjected to a plurality of voltammetric cycles to obtain a data set for each cycle, wherein the data sets comprise data points; and b) recording the profile comprising the data set for each cycle in a reference file.

12. The method according to claim 11, wherein the data set for a first voltammetric cycle is not recorded in step b).

13. The method according to claim 11, wherein step a) further comprises subjecting the product to a plurality of substantially identical working electrodes for each voltammetric cycle to obtain data set for each cycle for each electrode.

14. The method according to claim 13, further comprising: (i) averaging corresponding data points for the plurality of substantially identical working electrodes for a first recorded voltammetric cycle to obtain the data set for the first recorded voltammetric cycle; (ii) determining a standard deviation value to obtain an upper threshold and a lower threshold for each data point of the data set for the first recorded voltammetric cycle; and (iii) repeating steps (i) and (ii) for each subsequent voltammetric cycle.

15. The method according to claim 14, wherein step a) further comprises subjecting the product to a plurality of non-identical working electrodes during each voltammetric cycle, wherein each non-identical working electrode has an associated plurality of substantially identical working electrodes, and wherein steps (i) to (iii) are repeated for each cycle for each non-identical working electrode.

16. The method according to claim 11, wherein the product is subjected to at least three voltammetric cycles.

17. The method according to claim 11, wherein the product is subjected to five voltammetric cycles.

18. The method according to claim 11, wherein the or each voltammetric cycle is cycled through an oxidative voltage range and/or a reductive voltage range.

19. The method according to claim 11, wherein the or each voltammetric cycle is cycled between +1V and −1V.

20. The method according to claim 11, further comprising entering the reference file into a database of profiles for known products.

Description

BRIEF DESCRIPTION OF FIGURES

[0045] The accompanying drawings illustrate presently exemplary embodiments of the disclosure, and together with the general description given above and the detailed description of the embodiments given below, serve to explain, by way of example, the principles of the disclosure.

[0046] FIG. 1 shows a typical cyclic voltammetry curve for a first branded whisky sample.

[0047] FIG. 2A shows an example of the changes observed in a set of cyclic voltammetry curves over the course of different cycles recorded for the first branded whisky sample.

[0048] FIG. 2B shows an expanded view in the high potential region of the set of cyclic voltammetry curves shown in FIG. 2A.

[0049] FIG. 2C shows an expanded view in the low potential region of the set of cyclic voltammetry curves shown in FIG. 2A.

[0050] FIG. 3 shows a graph of the average response of cobalt pthalocyanine sensors in branded whisky samples at +0.2V.

[0051] FIG. 4 shows a graph of the changes observed in the percentage similarity score recorded for three branded whisky samples compared to an unknown sample over the course of five voltammetric cycles.

[0052] FIG. 5 shows a graph of the percentage similarity score recorded for three branded whisky samples compared to an unknown sample by considering a) cycle 1 only; b) an average of the first three cycles; and c) five cycles independently.

[0053] FIG. 6 shows a graph of the percentage similarity score recorded for vodka brand 1 vs vodka brand 1 using six different electrode types.

[0054] FIG. 7 shows a graph of the percentage similarity recorded for vodka brand 1 vs vodka brand 2 using six different electrode types.

[0055] FIG. 8 shows a graph of the percentage similarity score recorded for Coca Cola® when tested against a database of other leading brand soft drinks.

[0056] FIG. 9 shows a graph of the percentage similarity recorded for dissolved paracetamol when tested against a database of leading brand headache medicaments.

DETAILED DESCRIPTION

[0057] FIG. 1 shows a typical voltammetry cycle based on a single cycle recorded for a first branded whisky sample, which may or may not be the first of several cyclic voltammetry (C.V.) cycles. To date, known methods have incorporated the use of multiple cycles into the identification of products by simply averaging out the data obtained from several cycles for the purpose of removing background noise or for the purpose of long term monitoring to identify fouling of the sensors.

[0058] However, both approaches neglect to appreciate the surprising fact that each cycle is unique and repeatable for any given product and sensor combination. An example of how the overall C.V. curve changes over the course of different cycles recorded for the first branded whisky sample can be seen in FIG. 2A. The expanded views shown in FIGS. 2B and 2C demonstrate the subtle differences observed at the extremes of the positive and negative ends of the cycles.

EXAMPLE PROCEDURE

[0059] In the following examples, Metrohm DropSens sensors were used, although other suitable sensors are known in the art. The voltammetric sensor is immersed in a liquid sample such as a beverage. Solid dissolvable products are dispersed in a suitable media prior to testing. Advantageously, no additional pre-treatment of the sample is required.

[0060] The sensor(s) may be attached to a handheld potentiostat to apply voltammetry to the sample. For example, the potentiostat may be configured to perform a plurality voltammetric cycles between −1V and +1V. The user initiates the voltammetry measurement and the current output is recorded for each voltage potential on each cycle.

[0061] The results are compared to a known product for authentication or cross-referenced against a database of known product fingerprints to identify the sample. For example, a user may upload the results to an online data storage platform, which can then be viewed remotely.

[0062] The following Examples were obtained using a potentiostat U400 by Metrohm DropSens. However, as would be appreciated by a person skilled in the art, any suitable device may be used.

Example 1: Two Leading Brands of Scotch Whisky

[0063] To illustrate that the change in curve shape observed between cycles is repeatable, two leading brands of Scotch whisky were tested with a series of identical sensors comprising an electrode coated with a carbon-based ink doped with cobalt phthalocyanine. Three identical sensors were immersed in three separate samples of whisky brand 1 and the samples submitted to C.V. cycling between −1V and +1V at 50 mV/s for 5 cycles. The procedure was then repeated for whisky brand 2. The current recorded at specific voltages for each sensor was compared for each of the five cycles to ensure the current recorded was repeatable when using the same sample type. An example of the average result per cycle for the sample of whisky brand 1 and whisky brand 2 at +0.2V is shown in FIG. 3 and Tables 1 and 2 below.

TABLE-US-00001 TABLE 1 Results obtained from whisky brand 1 +0.2 V Sensor Sensor Sensor 1 2 3 Average STDEV c1 2.50 2.85 2.85 2.73 0.20 c2 9.65 9.70 10.10 9.82 0.25 c3 7.37 7.89 8.10 7.79 0.38 c4 7.45 7.40 7.79 7.55 0.21 c5 6.90 7.11 7.36 7.12 0.23

TABLE-US-00002 TABLE 2 Results obtained from whisky brand 2 +0.2 V Sensor Sensor Sensor 1 2 3 Average STDEV c1 3.78 3.84 3.92 3.85 0.07 c2 9.82 9.85 9.54 9.74 0.17 c3 6.99 6.10 6.40 6.50 0.46 c4 6.79667 6.93 6.69 6.81 0.12 c5 6.24 6.29 6.24 6.26 0.03

[0064] The results demonstrate that when using the same sample and sensor combination the amperometric response is repeatable between corresponding samples. Furthermore, it is clear that apart from the results obtained from cycle 2 there is a significant difference in the response between whisky brand 1 and 2, as exemplified in FIG. 3. If, as could be the case in known methods, only the second cycle was used as the basis for the comparison, it would not be possible to determine whether the sample was whisky brand 1 or whisky brand 2.

[0065] Furthermore, when this strategy is expanded for all of the data points obtained from a C.V. scan (i.e. all the observed currents for all of the voltage points measured) it is possible to generate a complete fingerprint for a product. The fingerprint is generated by testing multiple identical sensors in the same sample of a known product and recording the results. An average current is then determined for each voltage point on the C.V. scan by averaging the current recorded at the corresponding voltage points on each sensor on a given cycle. The standard deviation for each value is then calculated. The standard deviation is used to calculate the upper threshold of the fingerprint by adding the standard deviation value to the average value. The lower threshold of the fingerprint is calculated by the subtracting the standard deviation value from the average value. This is repeated for each cycle independently to produce a fingerprint consisting of multiple cycles with an amperometric range at any given voltage point on the C.V. scan for a specific product and sensor combination. If the current recorded for an unknown tested sample falls within amperometric range, this is indicative that the sample it is the product to which it is being compared.

Example 2: Comparison of Three Whisky Brands

[0066] To test a sample against the fingerprint, the same sensor type and testing methodology is used to generate the C.V. data set as was used for the known product. The recorded current for each voltage is then compared to the fingerprint to determine if it falls within the expected ranges. This may be expressed as a percentage similarity score between the unknown sample and the product fingerprint in the database to which it is being compared.

[0067] To illustrate this, a database consisting of three whisky brands was created using sensors coated in carbon doped with Prussian Blue. A blind test was performed to determine if whisky brand 2 could be accurately identified. Whisky brand 2 (unknown to the operator) was tested using the same Prussian blue doped carbon electrode type and methodology and the results were compared with the fingerprints from whisky brands 1, 2 and 3.

TABLE-US-00003 TABLE 3 % similarity of whisky brand 2 to finger prints 1, 2 and 3 Whisky Whisky Whisky brand 1 brand 2 brand 3 % similarity 21 100 47

[0068] As shown in Table 3, the sample correctly matches the fingerprint from whisky brand 2. In contrast, it has only a 21% similarity to whisky brand 1 and a 46% similarity to whisky brand 3.

[0069] FIG. 4 and Table 4 show how the innovation of analysing multiple cycles independently of one another to more accurately dismiss a sample as not matching the product fingerprint can be demonstrated when the similarity of the sample to product fingerprint is compared using fewer cycles. Specifically, when only a single cycle is used, the percentage similarities of whisky brands 1 and 3 to the test sample increases to 56% and 74%, respectively. Further, as can be seen, a general decrease in the percentage similarity of whisky brands 1 and 3 is observed as the number of cycles increases. In contrast, it can be seen that the high percentage similarity score is observed for whisky brand 2 is maintained.

TABLE-US-00004 TABLE 4 Analysis of multiple cycles Whisky Whisky Whisky brand 1 brand 2 brand 3 Cycle 1 56 100 74 Cycle 1, 2 40 100 80 Cycle 1, 2, 3 30 100 65 Cycle 1, 2, 3, 4 25 100 54 Cycle 1, 2, 3, 4, 5 21 100 47

[0070] FIG. 5 and Table 5 show a comparison of the results obtained using the method of the present invention with those obtained by considering the first cycle only and those obtained by averaging the first three cycles. The results clearly demonstrate that by analysing each corresponding data set of multiple cycles independently of one another, it is possible to more accurately distinguish an authentic candidate sample from those that are not.

TABLE-US-00005 TABLE 5 Comparison of results Whisky Whisky Whisky brand 1 brand 2 brand 3 Considering one cycle only (cycle 2) 24 100 86 Averaging the first three cycles 30 100 65 Considering each cycle 21 100 47 independently (five cycles)

Example 3: Comparison of Different Sensor Types

[0071] Furthermore, as each CV data set is generated from the combination of sensor surface chemistry and the sample being tested, the specificity of the data set to the fingerprint of a given product can be increased by using a plurality of sensors, each having different surface chemistries. The use of a plurality of sensors having different surface chemistries enables the detection of a broader range of sample components.

[0072] To illustrate this, a database using six different sensor types, the details of which are provided in Table 6, was generated for two leading vodka brands.

TABLE-US-00006 TABLE 6 Composition of working surface for sensors used FIGS. 6 and 7. All sensor types were obtained from Metrohm DropSens. Abbreviation Sensor Type 110 Carbon 110Bi Carbon doped with Bismuth 110Ni Carbon doped with nickel 410 Carbon doped with cobalt phthalocyanine 610 Carbon doped with Meldola's Blue F10 Carbon doped with Ferrocene

[0073] As can be seen in FIG. 6, a comparison of Brand 1 against Brand 1 found 100% similarity for all six electrode types. However, as shown in FIG. 7, a comparison of Brand 1 against Brand 2 showed that while most of the sensor types were significantly different some were very similar. In particular, the results obtained by sensor 110 indicated a 100% similarity between brand 1 and brand 2. However, while identical results were obtained for sensor 110 for brand 1 and brand 2, it will be appreciated that for another two brands, sensor 110 would be able to determine between the two as the response is highly dependent on the make-up of the product, and even the product batch, being tested. Therefore, using multiple sensors creates fingerprints that on at least some of the surface types will show a significant difference between even similar sample types. This provides the advantage of a more accurate fingerprint.

Example: Comparison of Soft Drinks

[0074] FIG. 8 shows a comparison of Coca Cola® against a database of other leading brand soft drinks. Using the method described herein, it was possible to clearly distinguish between the authentic product and eliminate those that were not, in order to allow for sample identification.

Example 5: Comparison of Pharmaceuticals

[0075] The method is also applicable to dissolvable solid products, which can be dispersed in a suitable media prior to testing. For example, FIG. 9 shows a comparison of an aqueous solution of paracetamol against a database of leading brand headache medicaments. In this case, the dissolved sample was again correctly identified as paracetamol.

[0076] It will be appreciated by persons skilled in the art that the above embodiment has been described by way of example only and not in any limitative sense, and that various alterations and modifications are possible without departing from the scope of the invention as defined by the appended claims.