Emerging Applications Headspace Analysis and Stripping of Volatile Compounds from Apple and Orange Juices Using SIFT-MS Introduction Differences in fruit varieties, fruit ripeness and processing techniques are just some of the factors that can cause undesirable and unacceptable variations in fruit juice flavour and aroma. The ability of selected ion flow tube mass spectrometry (SIFT-MS) to rapidly detect and quantify aroma compounds creates new opportunities to monitor production of juices, with a view to optimizing consistency of juice aroma and flavour qualities. In this brief paper we apply the whole-air analysis capability of SIFT-MS to static headspace and dynamic stripping analyses of two types of apple and orange juices, as well as a reconstituted orange-flavoured powdered beverage. Experimental Squeezed and reconstituted (from concentrate) juices of apples and oranges were investigated, as well as a reconstituted powdered orange beverage. Target compounds for apples and oranges were identified from the monograph, Volatile Compounds in Foods and Beverages (edited by Henk Maarse, Marcel Dekker, New York, 99). A Syft Technologies Ltd Voice0 SIFT-MS was used to quantify the target compounds in selected ion mode (SIM). Samples for static headspace measurements were prepared by placing 30 ml of neat juice (or reconstituted powdered beverage made up as per packet instructions) into a one-litre Schott bottle capped with a pierceable septum. The sample was left to equilibrate for at least 5 minutes prior to analysis. Dynamic stripping of volatiles was undertaken in a 250-mL bubbler. Diluted juice (50 ml; usually one part juice to four parts water) was placed in the bubbler. The SIFT-MS was used to sample the stream of air in a flow-past mode. That is, the SIFT-MS sampled a portion of the 200 ml min - flow of zero grade nitrogen that was stripping volatiles from the solution. The purge flow through the solution was started about one minute after the SIFT-MS scan was begun, so that portion of the scan corresponds to levels of compound in laboratory air. For both types of analysis, water was used as the control. The levels of target compounds were very low in water. Results and Discussion. Static headspace analysis Figure summarises the static headspace concentrations of various volatile organic compounds (VOCs) associated with the aroma and flavour of oranges. Although the graph is somewhat truncated to show the minor components, it is obvious that the reconstituted concentrate is markedly different from the other juices in the level of the major flavour compound, limonene. It is likely that the concentration process drives off some of the volatile compounds and hence reductions in the concentration of ethanol and acetaldehyde are particularly evident. The EMA-002-0.0 Page of 5
powdered beverage has clearly been formulated to mimic the squeezed juice, except that the level of ethanol is much reduced. 2000 800 600 2700 6800 62000 6200 3600 Reconstituted concentrate Squeezed Reconstituted powder 400 200 800 600 400 200 0 Ethyl 2- methylbutyrate Figure : Headspace volatile organic compound (VOC) profiles of two types of orange juice and a reconstituted orange-flavoured powdered beverage. Profiles of volatile compounds in the headspace of two apple juices are shown in Figure 2. For these juices, there are substantially different profiles of volatile compounds. In the squeezed juice, the volatile compounds ethanol and acetaldehyde dominate, whereas several of the more identifiable apple-related compounds (E-2-hexenal, ethyl butyrate and ethyl 2-methylbutyrate) are reduced compared to the reconstituted juice. The higher level of methanol in the reconstituted juice suggests that the flavour added to the reconstituted juice is an apple essence containing E-2-hexenal, ethyl butyrate and ethyl 2-methylbutyrate in methanol solution. 2. Volatile stripping Dynamic headspace concentrations of volatile compounds are generated by purging the solution with a continuous flow of inert gas. Figures 3 to 5 show the results of such stripping experiments for diluted orange juices (one part juice to four parts water). Note that the vertical axis in these graphs is logarithmic, not linear. We do not analyse the rise and decay behaviours of the various volatiles here, but it is evident that the volatiles behave differently between juices. The rise is very rapid and decay much slower for squeezed juice, with its much higher level of flavour compounds. Dilution of the squeezed juice 25-fold (Figure 6) shows that the rise can be slowed and decay hastened in more dilute solutions. The alcohols and aldehydes have a slower decay than limonene (a hydrocarbon) due to their much higher affinity for the aqueous solution. The volatile compounds in apple juices displayed similar decay behaviours, but levels of volatiles were generally lower than expected from the headspace measurements. EMA-002-0.0 Page 2 of 5
2000 800 600 5200 8900 Reconstituted concentrate Squeezed 400 200 800 600 400 200 0 Methanol Propanol E-2-Hexenal Hexanal Ethyl 2- methylbutyrate Figure 2: Headspace volatile organic compound (VOC) profiles of two types of apple juice. 00 0 0 0 5 5 20 25 Figure 3: Stripping of selected VOCs from squeezed orange juice (diluted five-fold with water). EMA-002-0.0 Page 3 of 5
00 0 0 0 5 5 20 25 Figure 4: Stripping of selected VOCs from orange juice reconstituted from concentrate (diluted five-fold with water). At the time this experiment was performed, there was more ethanol in the ambient air than in the dynamic headspace of the diluted juice as evidenced by the dip when the purge began. 00 0 0 0 5 5 20 25 Figure 5: Stripping of selected VOCs from reconstituted powdered orange flavoured beverage (diluted five-fold with water). EMA-002-0.0 Page 4 of 5
00 0 0 0 5 5 20 25 Figure 6: Stripping of selected VOCs from squeezed orange juice (diluted 25-fold with water). Conclusion This paper has demonstrated the flexibility and suitability of the Syft Technologies Ltd Voice0 SIFT-MS for analysis of fruit juice headspaces. The Voice0 facilitates very rapid analysis of aroma and flavour compounds in static headspace concentrations, with subsequent applications in production quality assurance. It also provides a very simple technique for determining absolute concentrations of aroma and flavour volatiles in real time to low part-per-billion levels, which can be employed in experiments stripping VOCs from fruit juices and other media. EMA-002-0.0 Page 5 of 5