Identification of Adulteration or origins of whisky and alcohol with the Electronic Nose Dr Vincent Schmitt, Alpha M.O.S AMERICA schmitt@alpha-mos.com www.alpha-mos.com Alpha M.O.S. Eastern Analytical Symposium, November 2004
Slide 2 Alcoholic beverage challenges Taste and smell in the beverage industry are important for the success of the product. Aroma and taste reflect the quality point of view, for example: - the vineyard - the type of distillation - the age - the contamination - the packaging
Slide 3 Alcoholic beverage challenges Taste and odor analysis are currently carried out by sensory panels. Actually, it is extremely problematic for production control. It is also very difficult to screen a large number of samples in one shot.
Slide 4 Needs of Industry Control objectively and rapidly from raw materials to finished products Monitor the Volatile Organic Compounds (VOC) quality of new production batches Detect out of specification odor and taste products Get a robust technique that correlates with human perception and provides fast answers
Slide 5 Human nose and Sensing system simulation Recognized as BRAZILIAN Odor Smell Brain Neurons Acquisition Data Processing Comparison of the odor by pattern recognition RESULTS Raw signals Processed signal Data treatment RECOGNIZED AS BRAZILIAN
Slide 6 Electronic Noses The advantages of this new technique: Volatile and odor measurement systems Both qualitative and quantitative results Solid, liquid or gaseous samples Correlation with human perception Fast (a few minutes per sample) Robust Very little sample preparation
Slide 7 αprometheus A development platform combining 2 technologies: αfox - Sensor array system Flexible sampling system, up to 18 sensors, QC or laboratory tool αkronos - Fingerprint MS High discrimination power, SCAN and SIM available, qualitative and quantitative
Slide 8 Fingerprint Mass Spectrometry (FMS) Principle
Slide 9 Multi organoleptic system: principle of one analysis Sampling Data Processing Headspace Sampler for volatile analysis Fingerprint measurement Multivariate analysis Printing the report SPDE (PDMS/AC) Crimping the sample in a vial and headspace generation Raw data are transferred in a Library Fingerprint identification with chemometrics Prediction Results
Slide 10 Multivariate Statistics Tools Data Processing Discrimination Quantification Quality control
Slide 11 Application 1: Whisky Study Aim: Cross polluted whisky discrimination. Detection of Orange, Mint, Pastis and Gin pollutions Reference Description of Whisky sample Characteristics A1, A2, A3, A4 Accepted quality - PO Polluted with orange Limonene (68 uma) Terpilonene (93 uma) PM polluted with mint Menthone (112 uma) PP polluted with pastis Anethole (148 uma) Estragole(149 uma) PG polluted with gin α and β pinène (93 uma)
Slide 12 System configuration αkronos - Fingerprint Mass Spectrometry Solid Phase Dynamic Extraction (SPDE) Option 2,5mL Headspace Syringe Steel needle Cross-section view New automated dynamic extraction technique 56 m m Internal coating Specific syringe coating Conical tip with side port
Slide 13 Analytical parameters Sample preparation Quantity of sample in the vial 0.5ml whisky / 0.5ml water Total volume of the vial 10 ml Headspace generation Headspace generation time 15 min Headspace generation temperature 70 C SPDE extraction / injection SPDE Syringe PDMS/AC syringe Stroke Number 80 strokes / 2500µl pumped / 500µl/sec Analysis time 15 min Predesorption time delay 15 sec, 0.6 ml N2 injected / 50 µl/sec
Application: results with the Fingerprint Mass Spectrometry Anis polluted sample MS The chemical compounds characteristic of Anis are clearly detected (anethole and estragole respectively mass 147 and 148). SPDE (PDMS/AC) Anethole estragole Slide 14 Calculation type RSD (%) 1-55 uma peak 1.7 2 - TIC average 3.0 In order to validate the analytical parameters the Residual Standard Deviation has been computed on one mass 55 and on the TIC (Total Ion Current). In this case, an excellent reproducibility is obtained. Clear detection of Anis characteristic masses Excellent repeatability of the measurements
Application: Principal Component Analysis results Acceptables Menthe A PCA model allows to represent the difference measured with the SPDE/MS system. Pastis Gin Orange The Accepted samples are grouped together and the samples with different defaults are spread around these reference samples. Slide 15 Excellent distinction between the acceptable samples and the polluted samples
Slide 16 Application 2: Wine packaging selection Aim: Study the shelf life of wine during eight months in two different packagings: Glass bottle (B) and Can (C). Sample designation Red wine in bottle 1 month Red wine in bottle 2 month Red wine in bottle 3 month Red wine in bottle 4 month Red wine in bottle 5 month Red wine in bottle 6 month Red wine in bottle 7 month Red wine in bottle 8 month Red wine in can 1 month Red wine in can 2 month Red wine in can 3 month Red wine in can 4 month Red wine in can 5 month Red wine in can 6 month Red wine in can 8 month Alpha-MOS Reference R_B_1_565 R_B_2_605 R_B_3_074 R_B_4_998 R_B_5_087 R_B_6_817 R_B_7_332 R_B_8_928 R_C_1_656 R_C_2_746 R_C_3_481 R_C_4_018 R_C_5_953 R_C_6_337 R_C_8_557
Slide 17 Analytical conditions System configuration: αkronos - Fingerprint Mass Spectrometry Sample preparation Quantity of product in vial 2 gr Volume of the vial 10 ml Head-space generation parameters Generation time 5 min. Generation temperature 65 C Head-space injection parameters Injected volume 5000 µl Injection speed 150 µl/s Syringe temperature 70 C Acquisition parameters KRONOS Mode Scan Scan range 50-110 Acquisition time 90 seconds Accuracy 25
Results Distance Slide 18 The distances between the sample of month 1 and each other month have been reported on the following graph: The First month is the 2,5 reference. 2 1,5 1 0,5 0 1 2 3 4 5 6 7 8 Reference Months Bottle After 5 months, the wines in can or in bottle reach the same target quality. Can At the beginning there is a big change in the quality of the wine compared to the reference month. The wine in can ages slightly more rapidly than the wine in bottle. After 5 months, the wine has changed but it reaches the same target whatever it is in Can or in Bottle.
Slide 19 Conclusion The results obtained demonstrate the ability for the Kronos Mass Spectrometry E-nose to identify several origins, process, packaging effect and to detect counterfeits on the market. Results show: Excellent discrimination between all the alcoholic beverage qualities Excellent repeatability of the measurements
Slide 20 Benefits of the method Fast, reliable and simple to use Objective identification of grade products Used for qualitative and quantitative analysis of complex odorous matrices
Slide 21 Raw material (rice, malt) selection and Quality control (Fox, Gemini) Ageing control (Prometheus, Kronos, Astree) Qualitative and quantitative control of yeast (Fox) Astringency determination (Astree) Control of barrel burning Blending control Wood chip toasting Packaging selection (Fox, Gemini) Counterfeit and fraud detection (Kronos, Astree) Trade mark protection
Slide 22 Electronic Nose and Tongue Products Research& Development 2nd generation of hybrid system α PROMETHEUS Research & Development Electronic tongue α ASTREE R&D Quality Control α FOX / α KRONOS Factory Quality Control Product conformity test α GEMINI
Slide 23 Please come and see us Booth # 411 Website: www.alpha-mos.com