Improvement of Tomato Paste Flavor via Addition of Tomato Essence By Rachel Gordon Advisor: Paul Sarnoski A thesis submitted in partial fulfillment of the requirements for the degree of Bachelors of Science with Honors in Food Science University of Florida Gainesville, FL April 6 th 2018
2 Abstract Fiv e tomato varietals (Roma, Garden Gem, Beaver Creek, Brandywine, Red Currant) were analyzed to observe the differences in flavor volatiles in the resulting tomato essence. Essence is an aqueous, clear liquid, filled with potent flavor volatiles. Essence was prod uced in fractions, each being based on ten percent of the total volume of the initial juice. The top five odor volatiles chosen for analysis were hexanal, (E) 2 hex enal, 3 hexenal, 1 hexanol, and 3 hexene 1 ol. The tomato varietals were juiced, then tomato essence was distilled using a rotary evaporator. To extract volatile compounds, a purge and trap (PT) was used along with Gas Chromatography/ Mass S pectro metry (GC MS) analysis. An internal standard ( (+) carvone) was added to each sample to quantify the compounds from the peak areas. GC/ MS data showed that the Roma followed by the Brandywine and Red Currant tomato varietals had the highest concentrations of the fiv e identified compounds Additionally, sensory data confirms that Roma and Red Currant were comparable for the highest values of tomato flavor, freshness, and aroma, with Red Currant derived essence receiving the highest scores for overall likeability and sweetness. These results, in addition to additional r esearch, could lead to the use of tomato essence in more tomato products.
3 Introduction production exceeding 70 million tons in 1993 (Thakur, Singh, & Nelson 1996). Consumers have complained of decreased flavor in fresh tomatoes, which has motivated horticultural research in improved tomato flavor. This may be due to tomatoes being harvested during the green stag e a nd later being exposed to ethylene gas, a plant ripening hormone, before going to market (Maul et al 1998). During tomato processing many undesirable changes occur leading to degradation of fresh tomato flavor (Buttery, Teranishi, Ling, & Turnbaugh 1990) One of the most significant changes is the loss of many of the green and fresh notes found in fresh tomato products. During processing, depending on the length and temperature of the heat treatment, the reducing sugar and amino acid content decreases. Ad components, including dimethyl sulphide and acetaldehyde form (Petro Turza, 2009). One possible way to prevent the loss of these flavor compounds is through the recombination of the proce ssed tomato product with a tomato essence. An essence is an aqueous clear liquid that is produced from the fresh product (Zhu et al 2017). Essences were first pioneered using oranges, collecting the essence from the saturated vapor during the juicing process (Attaway, Wolford, & Edwards, 1962). This same process can be used with tomato juice to capture tomato flavor volatiles. Additio nally, heirloom tomatoes are increasing ly popular in the fresh market. Although there is no legal definition of an heirloom cultivar, horticulturalists at the University of Florida consider it to mean an old cultivar that has a distinct diversity compared to other cultivars currently found on the market (Klee, 2015). H orticulture D epartment is currently breeding many different tomato variet als, four of which are evaluated in this study.
4 The aim of this study was to analyz e the es sences produced from four heirloom varietals (Garden Gem, Beaver Creek, Brandywine, and Red Currant) in addition to the common Roma variety for odor active volatile compounds that are associated with green and fresh flavors. The compounds evaluated were he xenal, (E) 2 hexenal 3 hexenal, 1 hexeanol, and 3 hexen 1 ol. These compounds were selected based on previous data collected by Zhu et al (2017) when characterizing the overall flavor volatiles found in Roma and Garden Gem tomatoes. Finally the essences that were produced were recombined with a tomato juice from concentrate and were analyzed in a consumer taste panel for preferences among the different cultivars. Methods and Materials Sample Preparation and Storage Roma tomatoes were purchased from a local Publix Super Markets, I nc store ( Gainesville FL, USA). Garden G em, Bea ver C reek, Brandywine, and Red C urrant tomato varietals were harvested from an on campus green house. All tomatoes were processed into juice within 24 hours. Tomatoes were cut i nto three pieces using toss and chop scissors (Comfify, Fernadale, WA, USA), and tomatoes were then juiced using a Model 2400 O.M.R.A. MTD Electric tomato milling machine (Brooklyn, NY, USA). The resulting puree was strained and mixed with food grade satur ated chloride (Modernist Pantry, Portsmouth, NH, USA) in a ratio of 3:1 to inactivate enzymes. The tomato juice was then stored in one gallon samples bags and stored at 20C. Essence Production Approximately three liters of each tomato juice were distill ed in a water bath at 40C at a pressure of 50 mbar using a rotary evaporator (Rotavapor R 220 SE, Buchi, Flawil, Switzerland), rotating at 30 RPM yielding 300 mL chilled fractions each. Five fractions of each distillate were collected.
5 Flavor Isolation The tomato essence was placed in the vial for analysis without any pretreatment. Flavor extraction was performed using a Stratum purge and trap (Teledyne Tekmar, Mason, OH, USA). A 5 ml sample was taken by the autosampler and purged with a nitrogen flow at 200 mL/ min for 30 minutes. The vola tile compounds were adsorbed by a trap c ontaining Tenax, were desorbed and then sent to the gas chromatograph port via a transfer line. Id entification and quantification Identification was conducted on an Agilent (Sant a Clara, CA, USA) gas chromatograph, connected to a quadrupole mass spectrometer detector (5975). A ZB WAXplus (Phenomenex, Torrance, CA, USA) column (30 m x 250 m x 0.25 m) ) was used for separation. The GC oven temperature was held at 45C for seven minutes, and th en the temperature was increased gradually to 240C. The mass spectrometer scanned over a range of 45 300 m/z and the resulting peaks were identified using a compar ison with data from the 2011 NIST library. Volatiles were then quantified by comparing the peak area of the chromatogram to that of the internal standard ((+) carvone) (Figure 1) Sensory Analysis A sensory panel was conducted with five participants includ ing students and professors from the Food Science and Human Nutrition Department to rate the overall likeability and f ive sensory characteristics of each essence (tomato flavor, sweetness, freshness, mouthfeel and aroma ). Each attribute was rated on a 9 p oint hedonic scale from very weak (score of 1) to very strong (score of 9). Tomato essence fractions one and two were added to tomato paste and water with addition of a low salt level 295 mg salt per 50 mL sample ( same salt level as commercially available
6 Low Sodium Tomato juice). Panelists tasted a total of eleven samples: fraction 1 and fraction 2 of each varietal, and a control of only tomato paste, water, and salt. Panelists tasted 50 mL of each sample. Statistical Ana lysis Sensory data was analyzed using a two way analysis of variance (ANOVA) test using Microsoft Excel to determine if there were significant differences between the tomato varietals on the tested sensory characteristics: overall likeability, tomato flavo r, sweetness, freshness, mouthfeel, and aroma Results Volatile Compound Analysis Five tomato volatile compounds were quantified for each essence fraction (Tables 2 6 ) These five compounds were found by Zhu, et al (2017) The tomato varietal with the highest concentrations of these compounds was found to be the Roma tomatoes. Roma had especially high concentrations of hexenal and (E) 2 hexenal (Table 2) Brandywine an d then Red Currant followed Roma with the highest volatile concentrations (Tables 5 & 6) Garden Gem and Beaver Creek, the remaining two varietals, trailed with the lowest concentrations of all five volatiles (Tables 3 & 4) The first fraction contained th e highest concentration of each volatile, with the concentration in the thousands of parts per billion for both Roma and Red Currant essences. The second fraction had a significant drop, with concentrations in the hundreds of parts per billion. Concentrati ons continue to decrease through fractions three and four, with the concentrations in fraction five being m uch lower.
7 Sensory Analysis None of the differences in the taste panel results was found to be statistically significant. There Red Currant fraction 1 had the highest overall likeability with an average of 5.4, while Garden Gem was the low est with an average of 3.6. Roma fraction 1 and Red Currant fraction 1 both had the highest value of 5.4 for tomato flavor, while Red Currant fractions 1 and 2 both had the highest values (5.1 and 5.2) for sweetness). Roma fraction 1 had the highest value (5.4) for freshness, with Red Currant fraction 1 comparable (5.2). The control had the highest score for mouthfeel (5.4). Finally, Roma fraction 1 had the highest value (6.2) for aroma, with Red Currant fraction 1 also close (6.0). All sensory data is summ arized in Table 7. Discussion Over one hundred and fifty volatile compounds have been characterized in fresh tomatoes since the 1950s. Recently, research has turned to identifying the type of flavor volatile (fruity, grassy, cooked, etc) in both fresh and cooked tomato products. Zhu et al (2017) analyzed 38 compounds that had statistically different concentrations before and after thermal processing, which confirmed that thermal processing dramatically affects the flavor profile of tomato products. The mos t dra matic loss of green notes was in hexanal, 3 hexenal, 1 hexanol, 3 hexene 1 ol, 2 octenal, and 2 isobutylthiazole, which were reduced up to 80% after processing (Goodman et al. 2002). The produced essence contained approximately sixty volatile c ompoun ds as characterized by GC/MS. Five compounds generally accepted as influencing the green/ fresh flavor of the tomato were then quantified. Roma had the overall highest concentrations of any of the varietals in fractions one, two, and three. There was a gen eral trend for all the cultivars of hexenal and (E) 2 hexenal having the highest concentrations of the analyzed volatiles in the essence fraction one.
8 This is true for all the cultivars excluding Garden Gem, which had a very low concentration of (E) 2 hexe nal (Table 3). This result matches those of Zhu et al. (2017), which showed that Roma tomatoes had higher levels of green notes and were generally lower in fruity notes. This matches the results found in Table 2 of the relative concentrations of hexenal an d (E) 2 hexenal (considered green and leafy green notes) compared to 1 hexanol (fruity green) in Roma essence. Although the taste panel results were not statistically significantly different, some generalizations can be made for future and further study of tomato volatiles in the essence. Roma and Red Currant tomatoes scored the highest on all sensory characteristics excluding mouthfeel, as seen in Table 7. Some panelists commented that the Roma fractions had an aftertaste. This may be due to the higher con centrations of all five green related compounds, leading to an overwhelming of the senses (Table 2) Interestingly, in an independent taste test of 38 heirloom varieties, Garden Gem scored second (below Maglia Rosa another heirloom varietal not studied here ) on likeability (Tieman et al 2012). That result places it far above Brandywine, another varietal analyzed in this study, which had at least twice the concentration of each volatile in the first fraction. Even though it scored w ell in that taste pane l, Garden Gem scored second to last on overall likeability in this taste panel, and Garden Gem had the second to lowest volatile concentration s of the five varietals. Brandywine received the lowest overall likeability score of all the cultivars observed, c orroborating the results found by Tieman et al. (2012). All though Bra n dywine received the lowest likeability score, fraction one had the second to highest volatile concentrations, after the Roma varietal (Table 5) It is possible there was an addition al, unanalyzed compound that le d to the decrease in likeability for Brandywine essence. Further analysis of the overall flavor profile should be performed.
9 Beaver Creek had the lowest volatile levels of any of the cultivars, though it did well in the taste pan el Beaver Creek matched Roma on overall likeability in the taste panel and had average scores for tomato flavor, sweetness, and freshness (Table 7). Conclusion Overall, this data will help with mapping the volatiles found in tomato essences and support further research in this area. Continuing analysis of the sensory aspect of this data, additional and more extensive taste panels should be conducted, specifically with trained panelists. Some of the data obtained was more in relation to the consumers pers onal op inion on tomato juice instead of analyzing the differences among flavors of the essence variations. From this preliminary data, it appears that Red Currant tomatoes are the most preferred when added back to a concentrate product al though the Roma variety also scored well The volatile levels of the Red Currant tomatoes were not as high as those found in the Roma tomatoes, but panelists found that the Red Currant were able to give a balanced and pleasant flavor to the products.
10 Cited Works Buttery R. G. Teranishi, R., Ling, L. C., & Turnbaugh, J. G. (1990). Quantitative and sensor y studies on tomato paste volatiles. Journal of Agricultural and Food Chemistry, 38(1), 336 340.. Buttery, R. G., Teranishi, R., Ling, L. C., & Turnbaugh, J. G. ( 1991). Cooked tomato flavor composition. Washington, DC: U.S. Patent and Trademark Office U.S. Patent No 5,064,673. Goodman, C., Fawcett, S., & Barringer, S. (2002). Flavor, viscosity, and color analyses of hot and cold break tomato juices. Journal of Food Science, 67(1), 404 408 Klee, H. J. (2015). Heirloom variety trials. 2015. http://hos.ufl.edu/kleeweb/newcultivars.html Accessed 5.9.2017 Maul, F., Sargent, S. A., Balaban, M. O., Baldwin, E. A. Huber, D. J., & Sims, C. A. (1998), Aroma volatile profiles from ripe tomatoes are influenced by physiological maturity at harvest: An application for electronic nose technology. Journal of the American Society for Horticultural Science, 123(6), 1094 110 1. Petro Turza, Martha (2009) Flavor of tomato and tomato products, Food Reviews International, 2:3, 309 351, DOI: 10.1080/87559128609540802 Tieman, D., Bliss, P., McIntyre, L. M., Blandon Ubeda A., Bies, B. Odabasi, A., Rodriguez, G., van der Knaap, E., Taylor, M., Goulet, C., Mageroy, M., Snyder, D., Colquhoun, T., Moskowitz, H., Clark, D., Sims, C., Bartoshuk, L., Klee, H. (2012) The chemical interactions underlying tomato flavor preferencce s, Current Biology, 22 :11, 1035 1039.
11 Thakur, B. R., S ingh, R. K., and Nelson, P. E. (1996). Quality attributes of processed tomato products: a review, Food Reviews Int., 1996; 12: 375 401. Wolford, R. W., Alberding, G. E., & Attaway, J. A. (1962). Analysis of recovered natural orange essence by gas chromatography. Journal of Agricultural and Food Chemistry, 10(4), 297 301. Zhu Y, Klee H, Sarnoski P. (2017). Development and characterization of a high quality plum tomato essence. Food Chemistry. DOI:1 0.1016/j.foodchem.2017.07.160
12 Tables and Figures Table 1. Sensory characteristics of odor active volatiles found in tomato essence. Compound Sensory characteristic h exenal g reen (E) 2 hexenal leaf g reen 3 hexenal grassy g reen 1 hexanol fruity g reen 3 hexen 1 ol fresh grassy T able 2 Concentrations (in parts per billion) of five key tomato volatile concentrations in Roma tomatoes separated based on essence fraction. ROMA Concentration (ppb) Compound Fraction 1 Fraction 2 Fraction 3 Fraction 4 Fraction 5 Hexanal 5558.69 658.95 388.90 59.40 12.28 (E) 2 hexenal 5826.93 526.61 481.11 48.51 0.00 3 hexenal 1101.07 193.75 132.57 34.64 0.00 1 hexanol 2140.57 399.74 221.28 24.65 25.42 3 hexen 1 ol 1975.46 640.24 410.09 95.66 30.22 Table 3 Concentrations (in parts per billion) of five key tomato volatile concentrations in Garden Gem tomatoes separated based on essence fraction. GARDEN GEM Concentration (ppb) Compound Fraction 1 Fraction 2 Fraction 3 Fraction 4 Fraction 5 Hexanal 1475.95 576.04 118.49 42.16 13.05 (E) 2 hexenal 208.73 244.63 114.00 54.21 14.01 3 hexenal 674.43 273.85 105.10 42.56 18.56 1 hexanol 123.58 69.84 0.00 0.00 0.00 3 hexen 1 ol 384.00 302.02 83.89 34.69 13.89
13 Table 4 Concentrations (in parts per billion) of five key tomato volatile concentrations in Beaver Creek tomatoes separated based on essence fraction. BEA VER CREEK Concentration (ppb) Compound Fraction 1 Fraction 2 Fraction 3 Fraction 4 Fraction 5 Hexanal 1019.55 142.91 16.13 23.03 0.00 (E) 2 hexenal 1784.72 82.47 17.38 15.37 0.00 3 hexenal 413.47 118.05 16.80 21.32 0.00 1 hexanol 354.54 64.15 0.00 0.00 0.00 3 hexen 1 ol 816.54 273.89 61.24 53.94 11.39 Table 5 Concentrations (in parts per billion) of five key tomato volatile concentrations in Brandywine tomatoes separated based on essence fraction. BRANDYWINE Concentration (ppb) Compound Fraction 1 Fraction 2 Fraction 3 Fraction 4 Fraction 5 Hexanal 2802.86 971.38 181.16 40.27 12.01 (E) 2 hexenal 2693.36 254.75 63.81 16.32 0.00 3 hexenal 920.03 394.18 118.87 39.44 12.51 1 hexanol 509.93 181.60 22.03 0.00 0.00 3 hexen 1 ol 1145.85 542.19 135.18 46.19 19.14
14 Table 6 Concentrations (in parts per billion) of five key tomato volatile concentrations in Red Currant tomatoes separated based on essence fraction. RED CURRANT Concentration (ppb) Compound Fraction 1 Fraction 2 Fraction 3 Fraction 4 Fraction 5 Hexanal 2337.10 295.83 81.84 65.68 11.27 (E) 2 hexenal 1954.22 190.37 58.51 41.95 0.00 3 hexenal 606.91 136.20 69.24 54.34 10.42 1 hexanol 301.98 45.37 9.95 0.00 0.00 3 hexen 1 ol 856.55 239.35 83.14 51.65 9.98 Table 7 Attribute averages for Fractions 1 (F1) and 2 (F2) of the five tomato varietals from the sensory analysis panel Higher values indicate a more favorable attribute. Characteristic Control Roma F1 Roma F2 GG F1 GG F2 BC F1 BC F2 BW F1 BW F2 RC F1 RC F2 L ikeability 4.4 4.8 4.4 4.8 3.6 4.8 5 4.4 4.4 5.4 5.2 Tomato flavor 5 5.4 5 4.6 4 5.2 5.2 4.6 4.6 5.4 5 Sweetness 3 4.8 3.6 4.4 3 4 4.6 3.6 3.8 5.1 5.2 Freshness 4.4 5.4 4.4 5 4.2 4.8 4.6 4 3.6 5.2 4.8 Mouthfeel 5.4 4.6 4.6 4.6 4.6 4.8 4.8 4.6 4 4.8 5 Aroma 3.6 6.2 3.8 5.4 4.4 4.6 4.4 4.2 3.4 6 3.8
15 Figure 1. Chromatogram of the Red Currant fraction one essence.