Analysis of Iodine-like (Chlorine) Flavor-causing Components in Brazilian Coffee

Transcription

1 Food Sci. Technol. Res., 17 (4), , 2011 Analysis of Iodine-like (Chlorine) Flavor-causing Components in Brazilian Coffee with Rio Flavor Hiroyuki Kato 1,2*, Kazunobu Sato 1 and Takeji Takui 1 1 Departments of Chemistry and Materials Science, Graduate School of Science, Osaka City University, Sumiyoshi-ku, Osaka , Japan 2 General Laboratories, Daiwa Can Company, Sagamihara, Kanagawa , Japan Received December 17, 2010; Accepted April 26, 2011 Green beans and roasted beans of coffee with Rio flavor were analyzed by gas chromatographyolfactometry to clarify the causal substances of Rio flavor in Brazilian coffee. In addition to 2,4,6-trichloroanisole (TCA) and 2,4,6-trichlorophenol (TCP), which are known as the main causal substances of Rio flavor, a new iodine-like flavor-causing compound, 2,6-dichlorophenol (DCP), was detected. The detected amounts of 2,6-DCP and 2,4,6-TCA were less than 0.1 and 0.1 ppb in Rio flavor-free coffee beans, 0.8 and 1.2 ppb in weak Rio flavored beans, and 3.0 and 2.6 ppb in strong Rio flavored beans, respectively. Musty and chlorine flavors were sensed more strongly in model coffee containing both 2,4,6-TCA and 2,6-DCP than in the coffee containing them individually. In the present study, 2,6-DCP was identified as a new causal substance of Rio flavor, and it was found that the presence of 2,4,6-TCA intensifies Rio flavor, as a synergy effect. Keywords: 2,6-dichlorophenol, 2,4,6-trichloroanisole, coffee, Rio, off-flavor, iodine-like, synergism, GC-O Introduction Together with black tea and green tea, coffee is one of the most popular drinks in the world. Coffee beans are produced worldwide, mainly in Africa, Southeast Asia and Central and South America. Among these, Brazil has the largest production area and exports a large amount of coffee beans. Coffee beans are used after roasting, but at the green and roasted bean stages the quality is checked by coffee flavor testers. In Brazil, depending on the flavor, coffee beans are ranked by sensory evaluation from the favorable soft to the non-favorable Rio rank. In addition, their appearance is examined, such as for the presence of dirt and injury, and the quality is graded, for example as type 2 or type 4/5 (i). In the coffee industry, extensive, even overzealous, efforts have been made to obtain high quality beans with a preferable flavor. Particularly in Brazilian coffee, close attention has been paid to an intensive off-flavor, called Rio flavor. There are many studies on the off-flavor of coffee (Blank and Grosh, 2002; Cantergiani et al., 2001; Czerny *To whom correspondence should be addressed. h-kato@mail.daiwa-can.co.jp et al., 2000; Lindinger et al., 2009; Spadone and Takeoka, 1990). The most non-preferable flavor among them is Rio flavor, which consists of medicinal, phenolic, iodine-like and chlorine-like flavors found in tap water in Japan. Causal substances of Rio flavor hitherto detected and discussed are 2,4,6-trichloroanisole (TCA) and its precursors, 2,4,6-trichlorophenol (TCP) and 2-methylisoborneol (MIB) (Spadone and Liardon, 1988; Spadone and Takeoka, 1990). 2,4,6-TCP has a high threshold and does not cause an off-flavor unless it exists in large amounts, but 2,4,6-TCA causes an off-flavor even at a low concentration. Spadons et al. analyzed Rio coffee from different areas such as Brazil and Puerto Rico, and identified 2,4,6-TCA and 2,4,6-TCP as off-flavor compounds (Spadone and Takeoka, 1990). Even in non-brazilian coffee, 2,4,6-TCA and musty flavor components such as geosmin have been detected (Cantergiani et al., 2001). Blank and Grosh reported that the off-flavor of robusta coffee was caused by 2-MIB (Blank and Grosh, 2002). In the present study, we analyzed the causal substances of the iodine-like flavor characteristic of the Rio flavor of Brazilian coffee.

2 348 Materials and Methods Rio Coffee Samples Green and roasted Brazilian coffee beans without Rio flavor (control), and green and roasted beans with Rio flavor purchased from Unicafe Inc. (Tokyo, Japan) were used for the experiments. Details are shown in Table 2, in which A, B, C, D, E, F and G are Brazilian beans of Santos type 4/5 imported on different dates in These beans were frozen in liquid nitrogen and immediately powdered using an electric mill (Karita electric coffee mill C-120; Tokyo, Japan; roughness dial set at position 4). Chemicals 2,4,6-TCA, 2,4,6-TCP, 2,6-DCP, 2,4,6-tribromoanisole (TBA) and 2,6-dibromophenol (DBP) were purchased from Tokyo Chemical Industry (Tokyo, Japan). 2,4,6-TCAd5 and 2,4-DCPd3 were purchased from ISOTEC (Miamisburg, OH). Ethanol, n-hexane, 0.5 mol/l sodium hydroxide solution (NaOH), phosphoric acid and anhydrous sodium sulfate were purchased from Wako Pure Chemical Industries (Tokyo, Japan). Ultrapure water was obtained from a Milli-Q water purification system (Nihon Millipore, Tokyo, Japan). Extraction of Volatiles Figure 1 shows the distillerextractor used for coffee extraction (Iida et al., 1978; Oka et al., 1977). Twenty grams of the coffee sample supplemented with an internal standard (20 ng of 2,4-DCPd3 and 20 ng of 2,4,6-TCAd5) was added to 500 ml of pure water with 5 ml of phosphoric acid, and then 5 ml of n-hexane was added as extractive solvent and the resulting mixture was distilled for 2 h under atmospheric condition. The n-hexane extraction was put in a 10 ml test tube, and dried over with anhydrous sodium sulfate, then 2,4,6-TCA was analyzed by GC-MS. For the analysis of 2,6-DCP and 2,4,6-TCP, 2 ml of the n-hexane extraction was put in a 10 ml test tube and shaken together with 2 ml NaOH for 5 min, and the NaOH layer was collected by centrifugation (3,000 rpm at 20 for 10 min). The same procedure was repeated twice. The collected NaOH layer was extracted with 5 ml of n-hexane twice to remove neutral and alkaline organic matter. Then, the NaOH layer was adjusted to ph 2 by adding phosphoric acid, and chlorophenol was extracted with 2 ml of n-hexane. The extracted n-hexane layer was dried over with anhydrous sodium sulfate, and analyzed by GC-MS. Gas Chromatography-Olfactometry (GC-O) GC-O was performed using a 6890GC (Agilent Technologies) with a DB-5 column (30 m 0.53 mm 1.0 μm film thickness, Agilent Technologies). The GC conditions were as follows: column temperature, 100 (for 5 min) to 250 at a rate of 10 /min; injector temperature, 250 ; injector mode, splitless; injection amount, 2 μl. The gas obtained by branching immediately before entering the FID detector (4:1) was subjected to GC-O. The GC-O experiments were performed by H. Kato et al. five panelists, trained in-house (2 males and 3 females, years old), in order to detect the musty and iodine-like flavors characteristic of Rio flavor. The presence or absence of 2,4,6-TBA and 2,6-DBP, which give musty and iodine-like flavors, was also confirmed by GC-O. GC-MS For the GC and MS experiments the 7890A GC (Agilent Technologies) and the Jms-Q1000GC MKII (JEOL, Tokyo, Japan) systems, respectively were used. The column was DB-5MS (30 m 0.32 mm 0.5 μm film thickness, Agilent Technologies). The GC conditions were as follows: column temperature, 60 (for 5 min) to 250 at a rate of 10 /min; injector temperature, 250 ; injector mode, splitless; injection amount, 2 μl. The MS conditions were as follows: GC interface temperature 250 ; ion source temperature 230 ; ionization voltage 70eV (EI); SIM measurement ion M/Z, TCAd5 197, 199, 215; TCA 195, 197, 210; DCPd3 165, 167; DCP 162, 164; TCP 196, 198. A spiked surrogate reagent was used as an internal standard for quantification. The value of 2,4,6-TCA for 2,4,6-TCAd5 was 0.87 (CV=3.6%; n=5), that of 2,6-DCP for 2,4-DCPd3 was 1.03 (CV=4.1%; n=5), and that of 2,4,6-TCP for 2,4-DCPd3 was 1.55 (CV=3.0; n=5). Under these conditions, the detectable limit of the concentration of off-flavor-causing substance was 0.1 ppb, and the linear relationship was obtained in the range of 0.1 to 10 ppb. Sensory Perception Threshold of Off-flavor in Coffee by the Coffee-Tasting Panelists Four trained coffee-tasting panelists estimated the sensory perception threshold by tasting the coffee extract with 2,4,6-TCA and 2,6-DCP. The coffee extract was obtained from 20 g of coffee/400 ml, supplemented with the off-flavor-causing substance diluted Extractive solvent (n-hexane) Cold water (10 ) Fig. 1. Distiller-Extractor used in the coffee extraction.

3 Taint Analysis of Coffee with Rio Flavor with ethanol and water, and the off-flavor of the extract was estimated at 50. The concentrations of 2,4,6-TCA were at the levels of , and 0.01 ppb, and those of 2,6- DCP at the levels of 0.01, 0.02 and 0.04 ppb. The sensory perception threshold was determined as the concentration when all the panelists correctly detected the flavor. Description of Off-flavor of Model Coffee by Coffee Tasting Panelists Model coffee extracts were supplemented with off-flavor-causing substances at different concentrations, and the musty or iodine-like flavor sensed by the panels was categorized into five ranks: The sign corresponds to undetected, + to slightly detected, ++ to detected, +++ to a little strongly detected, and ++++ to strongly detected, respectively. The concentrations of 2,4,6-TCA used were at the levels of and 0.02 ppb, and those of 2,6-DCP at the levels of 0.002, 0.04 and 0.08 ppb. Results and Discussion Table 1 shows the off-flavor components in different coffee beans detected by GC-O and GC-MS with retention 349 indices (RI) from 700 to Neither musty nor iodine-like flavor was detected in either the green beans or the roasted coffee control. On the other hand, in green beans and roasted coffee beans with Rio flavor, iodine-like and musty flavors, whose peaks appeared at RI 1203 and 1331, respectively, were detected by all five panelists. Thus, all the panelists confirmed the presence of 2,6-DCP and 2,4,6-TCA in the sample with Rio flavor. On the other hand, 2,4,6-TCP, known as an off-flavor-causing component in coffee and tap water, was detected only by GC-MS but not by GC-O. Neither 2,6-DBP nor 2,4,6-TBA was detected by GC-O and GC- MS. Thus, the off-flavor of the Rio sample in this study was caused by 2,6-DCP and 2,4,6-TCA. Table 2 shows the concentrations of 2,4,6-TCA, 2,6-DCP and 2,4,6-TCP as measured by GC-MS, together with the off-flavor intensity as estimated by the sensory test, of each hot water extract of the roasted coffee beans. Neither 2,4,6- TCA, 2,6-DCP nor 2,4,6-TCP was detected in the control, in which neither musty flavor nor iodine-like flavor was detected by the sensory test. Table 1. Off-flavor compounds detected in green beans and roasted coffee beans with or without (control) Rio flavor. Retention index Odor quality Odorant Green beans Roasted coffee beans A (control) C (Rio) D (control) F (Rio) 1203 Iodine-like 2,6-DCP a I b I 1331 Musty 2,4,6-TCA M c M 1355 Phenol-like 2,4,6-TCP p d p 1383 Iodine-like 2,6-DBP 1622 Musty 2,4,6-TBA a indicates undetected by any panelist by GC-O and peaks were not detected in GC-MS. b I (iodine-like) indicates detected by all panelists by GC-O and peaks were detected by GC-MS. c M (musty) indicates detected by all panelists by GC-O and peaks were detected by GC-MS. d p indicates that the peak was confirmed by GC-MS without odor detection. Table 2. Concentrations of off-flavor compounds in different samples of Rio coffee (n=3) in units of ppb. Sample Concentrations in Rio coffee (ppb) Intensity of off-flavor a 2,4,6-TCA 2,6-DCP 2,4,6-TCP Musty Iodine-like Green coffee, Brazilian A (control) nd nd nd Green coffee, Brazilian B (Rio) Green coffee, Brazilian C (Rio) Roasted coffee, Brazilian D (control) nd nd nd Roasted coffee, Brazilian E (Rio) Roasted coffee, Brazilian F (Rio) Roasted coffee, Brazilian G (Rio) Drip coffee a, Brazilian G (Rio) (0.01) b (0.04) (0.08) Values in the table were calculated based on bean weight. nd = not detected: < 0.1 ppb ; undetected, +; slightly detected, ++; detected, +++; a little strongly detected a Coffee sample was extracted at 20 g / 400 ml (90 hot water). b Equivalent value in coffee extract.

4 350 In Brazilian coffee beans F and G, which were judged to have a strong Rio flavor by the sensory test, 2,4,6-TCA, 2,6-DCP and 2,4,6-TCP were detected at , and ppb, respectively. In Brazilian coffee beans E, which was judged to have a weak Rio flavor, 2,4,6-TCA 2,6- DCP and 2,4,6-TCP were detected at 1.2, 0.8 and 2.4 ppb, respectively. Thus, in the roasted beans as in the green beans, 2,4,6- TCA, 2,6-DCP and 2,4,6-TCP were detected at a similar relative ratio of the concentration. In the sensory test, the beans with a concentration of > 1 ppb of 2,4,6-TCA had a musty flavor and those with > 2 ppb of 2,4,6-TCA had a strong musty flavor. On the other hand, the beans with a concentration of > 2 ppb of 2,6-DCP had a strong iodine-like flavor, but the iodine-like flavor tended to be overpowered by the strong musty flavor of 2,4,6-TCA at a high concentration level. In the drip coffee, the values of 2,6-DCP (after/before extraction 0.8/3.0) and 2,4,6-TCP (1.2/6.6) were greater than that of 2,4,6-TCA (0.2/2.6). This is possibly due to the fact that chlorophenol is more soluble than chloroanisole, or that 2,4,6-TCA is easily evaporated during the hot water (90 ) extraction. Sensory Perception Threshold of Off-flavor-causing Substances The sensory perception thresholds of 2,4,6-TCA and 2,6-DCP were estimated by triangle tests. All panelists detected the musty flavor caused by 2,4,6-TCA at ppb and one panelist could not detect it at ppb. Therefore, the threshold of 2,4,6-TCA detection in coffee was regarded as ppb (Table 3). The panelists who detected 2,4,6-TCA at ppb did not detect a musty flavor, but recognized that the coffee flavor was clearly weakened. Particular substances have such a smell masking effect, and Takeuchi and coworkers have studied in detail the mechanism of the effect of such biomolecules on the odorants (Takeuch et al., 2009). 2,6-DCP was detected by all the panelists at 0.02 ppb, but two panelists could not detect it at 0.01 ppb. Therefore, the threshold of 2,6-DCP detection was determined to be 0.02 ppb (Table 4). Previously, the threshold of 2,6-DCP detection in coffee Table 3. Sensory perception threshold of 2,4,6-TCA in coffee (n=4). 2,4,6-TCA ppb 0.001ppb 0.01ppb Number correct Table 4. Sensory perception threshold of 2,6-DCP in coffee (n=4). 2,6-DCP 0.01ppb 0.02ppb 0.04ppb Number correct H. Kato et al. was reported to be 0.1 ppb (Asakura et al., 1995), but in this study, the chlorine flavor was detected at 0.04 ppb. Description by Coffee Tasting Panelists of Off-flavor Caused by 2,4,6-TCA and 2,6-DCP as Added to Coffee Table 5 shows the results of a sensory test of Rio flavor-free coffee supplemented with 2,4,6-TCA and 2,6-DCP. In the coffee supplemented with 2,4,6-TCA alone, only a musty flavor was detected, and in the coffee supplemented with 2,6-DCP alone, only a weak iodine-like flavor was detected. However, in the coffee supplemented with 2,4,6-TCA and 2,6-DCP at a concentration level comparable to that detected in drip coffee, a strong musty and iodine-like flavor was detected. Thus, the Rio flavor of Brazilian coffee is caused not only by 2,4,6-TCA, the substance reported so far, but also by 2,6-DCP. Importantly, the off-flavor caused by these two compounds was intensified by their coexistence, as a synergy effect. Such synergism was observed in the effects of 2,6-DCP and 2,6-DBP, which are causal substances of iodine-like flavor of orange juice contaminated with Alicyclobacillus. The off-flavors caused by these compounds were intensified by the coexistence of guaiacol (Gocmen et al., 2005). Wise and coworkers prepared a mixture of two aliphatic carbonic acids with different lengths of the carbon chain (C2, C4, C6 or C8), and found an interaction between the two components such as C2+C4 (Wise et al., 2007). Miyazawa and coworkers reported that the flavor of coffee-flavor components such as maple and lactone was intensified by adding, at a concentration level less than the threshold value, butyric acid, which exists in coffee (Miyazawa et al., 2008; Miyazawa et al., 2009). 2,4,6-TCA and 2-MIB are known as substances that cause important off-flavor claims in the drink and food industries. In particular, 2,4,6-TCA is decomposed or synthesized by pentachlorophenol (PCP), which is used as a fungicide on wood and is a cause of many off-flavors. In addition, 2,4,6-TCA in drinks causes an off-flavor even at the level of ppt, and in the container industry, the use of wood pallets is prohibited in certain cases. Similarly, 2,4,6-TBA is known as a causal substance of musty-flavor (Whitfield et al., 1997). Among the dichlorophenol compounds, 2,6-DCP, which has a very low detection threshold, caused chlorine or iodine-like off-flavors in foods, drinks, tap water and mineral water, even at extremely low levels, and this also gives rise to a crucial problem in coffee, as illustrated above. 2,6-DCP is produced by the reaction of minute phenolic substances in water with chlorine or by decomposition of agricultural pesticides, and is known as a main component of the iodine-like flavor of tap water, vegetables and fruits (Acero et al., 2005; Asakura et al., 1995; Fukaya et al., 1994; Goldenberg and

5 Taint Analysis of Coffee with Rio Flavor 351 Table 5. Sensoly evaluation of Model coffee. Sample Add concentrations (ppb) Intensity of off-flavor 2,4,6-TCA 2,6-DCP Musty Iodine-like Description of off-flavor in coffee by tasting panel (1) 2,4,6-TCA Musty (2) 2,6-DCP Weak iodine-like (3) TCA + DCP Strong musty, iodine-like (4) TCA + DCP Very strong musty, little weak iodine-like (5) TCA + DCP Strong musty, phenolic strong iodine-like Coffee was extracted at 20 g / 400 ml (90 hot water). Sensory evaluation was performed by four panelists. ; undetected, +; slightly detected, ++; detected, +++; a little strongly detected, ++++; strongly detected Matheson, 1975; Montiel et al., 1999; Ogiwara et al., 2003; Saxby and Wagg, 1985; Tanaka et al., 1997). Both 2,4,6-TCA and 2,6-DCP have a low threshold and cause off-flavors in drinks and foods even at very low concentrations. It should be noted that almost anyone can sense mustiness at high concentrations (Buser et al., 1982; Curtis et al., 1974; Griffith, 1977; Tindale et al., 1989; Whitfield et al., 1985). 2,6-DCP is detected by most persons as a chlorine or disinfection flavor even in minute quantities, and is supposed to be a causal substance of the chlorine odor in Rio flavor. Like 2,6-DCP, 2,6-DBP is also a causal substance of the iodine-like flavor, giving a long-standing issue in tap water, fruits and seafood industries (Acero et al., 2005; Chung et al., 2003; Whitfield et al., 1988). Including the Rio flavor of Brazilian coffee, chlorophenol, bromophenol, chloroanisole and bromoanisole are responsible for claims of off-flavors in drinks and foods, and can cause economic losses. We cannot rule out any possibility that Rio flavor of Brazilian coffee is caused by the decomposition of 2,4,6-TCP and PCP, which has been reported previously, and by disinfection using chlorine (Buser et al., 1982; Chatonnet et al., 2004; Curtis et al., 1974; Fukaya et al., 1994; Goldenberg and Matheson, 1975; Griffith, 1977; Montiel et al., 1999; Ogiwara et al., 2003; Saxby and Wagg, 1985; Tindale et al., 1989; Whitfield et al., 1985; Whitfield et al., 1997). Conclusion Rio flavor of Brazilian coffee was characterized by an iodine-like flavor, and its causal substance was identified as 2,6-DCP by quantitative and qualitative analyses and by a sensory test, as performed by trained coffee-tasting panelists. However, since the iodine-like flavor of 2,6-DCP was intensified by the presence of 2,4,6-TCA, 2,4,6-TCA was considered to be the main causal substance of Rio flavor. The coexistence of 2,4,6-TCA at about 0.01 ppb and 2,6-DCP at > 0.04 ppb in drip coffee was found to cause an iodine-like flavor. Because the presence of 2,4,6-TCA intensifies the off-flavor, which is caused by other substances, a problematic off-flavor is suggested to occur even when the off-flavor compounds are at a level below the threshold value. Acknowledgements We thank Mr. Tetsuo Takano and Mr. Yosuke Nakamura, Unicafe Co., for performing the sensory test of coffee samples, and Professors Yasuyoshi Hayata and Fumiyuki Kobayashi, Faculty of Agriculture, Meiji University, for their valuable advice. References Acero, J. L., Von Gunten, U. and Piriou, P. (2005). Kinetics and mechanisms of formation of bromophenols during drinking water chlorination: Assessment of taste and odor development. Water Res., 39, Asakura, M., Tsutsumi, F. and Hayashi, T. (1995). Off - flavor caused by chlorophenols in canned beverages. Kanzume Jiho (The Canners Journal), 74, Blank, I. and Grosh, W. (2002). On the role of (-)-2-methylisoborneol for the aroma of Robusta coffee. J. Agric. Food Chem., 50, Buser, H. R., Zanier, C. and Tanner, H. (1982). Identification of 2,4,6-trichloroanisole as a potent compound causing taint in wine. J. Agric. Food Chem., 30, Cantergiani, E., Brevard, H., Krebs, Y., Feria-Morales, A., Amado, R. and Yeretzian, C. (2001). Characterisation of the aroma of green Mexican coffee and identification of mouldy/earthy defect. Eur. Food Res. Tech., 212, Chatonnet, P., Bonnet, S., Boutou, S. and Labadie, M. D. (2004). Identification and responsibility of 2,4,6-tribromoanisole in musty, corked odors in wine, J. Agric. Food Chem., 52, Chung, H. Y., Ma, W. C. J. and Kim, J. (2003). Seasonal distribution of bromophenols in selected Hong Kong seafood. Agric. Food Chem., 51, Curtis, R. F., Land, D. G., Griffiths, N. M., Gee, M., Robinson, D.

6 352 and Peel, J. L. (1974). Chloroanisoles as a cause of musty taint in chicken and their microbiological formation from chlorophenols. J. Sci. Food Agric., 25, Czerny, M. and Grosch, W. (2000). Potent odorants of raw Arabica coffee. Their changes during roasting. J. Agric. Food Chem., 48, De Moraes, M. H. P. and Luchese, R. H. (2003). Ochratoxin A on green coffee: Influence of harvest and drying processing procedures. J. Agric. Food Chem., 51, Fukaya, T., Muraoka, A. and Ishiguro, Y. (1994). Off flavor derived from 2,6-dichlorophenol in canned foods. Kanzume Jiho (The Canners Journal), 73, Gocmen, D., Elston, A., Williams T., Parish, M. and Rouseff, R. L. (2005). Identification of medicinal off-flavours generated by Alicyclobacillus species in orange juice using GC olfactometry and GC MS. Letters in Applied Microbiology, 40, Goldenberg, N. and Matheson, H. R. (1975). Off-flavours in food, a summary of experience : Chem. Ind. (London), July 5, Griffiths, N. M. (1977). Odour properties of chloroanisoles - effects of replacing chloro - by methyl groups. Chem. Senses Flavour, 2, Iida, K., Watanabe, S. and Ikeda, H. (1978). Analysis for residue of Chlorophenols in Vegetables. Shokuhin Eiseigaku Zasshi (J. Food Hyg. Soc. Jpn.), 19, Lindinger, C., Polline, P., de Vos, C. H. R., Tikunov, Y., Hageman, J. A., Lambot, C., Fumeaux, R., Voirol-Baliguet, E., and Blank, I. (2009). Identification of ethyl formate as a quality marker of the fermented off-note in coffee by a nontargeted chemometric approach. J. Agric. Food Chem., 57, Miyazawa, T., Gallagher, M., Preti, G. and Wise, P. M. (2009). Odor detection of mixtures of homologous carboxylic acids and coffee aroma compounds by humans. J. Agric. Food Chem., 57, Miyazawa, T., Gallagher, M., Preti, G. and Wise, P. M. (2008). Synergistic mixture interactions in detection of perithreshold odors by humans. Chem. Senses, 33, Montiel, A., Rigal, S. and Welte, B. (1999). Study of the origin of musty taste in the drinking water supply, Wat. Sci. Tec., 40, Ogiwara, T., Kondo, H., Miyakoda, M., Yamada, Y., Aoyagi, Y., Sato, H., Amakawa, E., Matsumoto, H., Kasuya, Y., Nakazato, M. and Yasuda, K. (2003). A case study on complaint about glazed H. Kato et al. adzuki beans contaminated by chlorophenol compounds, Ann. Rep. Tokyo Metr. P.H., 54, Oka, H., Nagai, I. and Tasaka, M. (1977). Analytical method of pentachlorophenol in wooden chips. Shokuhin Eiseigaku Zasshi (J. Food Hyg. Soc. Jpn.), 18, Saxby, M. J. and Wagg, S. (1985). Contamination of cocoa liquor by chlorophenols. Chem. Ind. (London), May 6, Spadone, J. C. and Liardon, R. (1987). Identification of specific volatile components in Rio coffee beans. In 12th International Conference on coffee, Montreaux, ASIC: Paris, France, pp Spadone, J. C., Takeoka, G. and Liardon, R. (1990). Analytical investigation of Rio off-flavor in green coffee. J. Agric. Food Chem., 38, Takeuchi, H., Ishida, H., Hikichi, S. and Kurahashi, T. (2009). Mechanism of olfactory masking in the sensory cilia. J. Gen. Physiol., 133, Tanaka, Y., Hosoi, S. and Shimizu, R. (1997). Detection of tolclofos-methyl and its decomposition product (2,6-dichloro-p-cresol) in phenolic-smelling tomatoes. Shokuhin Eiseigaku Zasshi (J. Food Hyg. Soc. Jpn.)., 38, Tindale, C. R., Whitfield, F. B., Levingston, S. D. and Nguyen, T. H. L. (1989). Fungi isolated from packaging materials: their role in the production of 2,4,6-trichloroanisole. J. Sci. Food Agric., 49, Whitfield, F. B., Hill J. L. and Shaw, K. J. (1997). 2,4,6-Tribromoanisole: a potential cause of mustiness in packaged food. J. Agric. Food Chem., 45, Whitfield, F. B., Last, J. H., Shaw, K. J. and Tindale, C. R. (1988). 2,6-Dibromophenol: the cause of an iodoform-like off-flavor in some Australian crustacea. J. Sci. Food Agric., 24, Whitfield, F. B., Nguyen, T. H. L., Shaw, K. J., Last, J. H., Tindale, C. R. and Stanley, G. (1985). Contamination of dried fruit by 2,4,6-trichloroanisole and 2,3,4,6-tetrachloroanisole adsorbed from packaging materials. Chem. Ind. (London), October 7, Wise, P. M., Miyazawa, T., Gallagher, M. and Preti, G. (2007). Human odor detection of homologous carboxylic acids and their binary mixtures. Chem. Senses, 32, URL Cited i) (Apr. 21, 2011)