Bioresource Technology 88 (2003) 17

Separation of cashew (Anacardium occidentale L.) nut shell liquid with supercritical carbon dioxide
R.L. Smith Jr.
a b

a,*

, R.M. Malaluan b, W.B. Setianto a, H. Inomata a, K. Arai

Department of Chemical Engineering, Research Center of Supercritical Fluid Technology, Tohoku University, Aoba-ku, Aramaki Aza, Aoba-04, Sendai 980-8579, Japan Department of Chemical Engineering Technology, Mindanao State UniversityIligan Institute of Technology (MSU-IIT), Andres Bonifacio Avenue, Iligan City 9200, Philippines Received 7 December 2001; received in revised form 24 October 2002; accepted 28 October 2002

Abstract Cashew nut shell liquid (CNSL) represents the largest readily available bioresource of alkenyl phenolic compounds. In this work, separation of CNSL from the pericarp of the cashew nut with supercritical carbon dioxide was studied. In the initial extractions with CO2 at 4060 C and at pressures from 14.7 to 29.4 MPa, low yields were obtained. However, when the extractions were performed with one or more intermediate depressurization steps, the yield of CNSL increased to as high as 94%. Most of the oil did not separate from the shell during the depressurization step, but was obtained during the subsequent repressurization. The CNSL extract had a clear light brownish pink color and exhibited no evidence of polymerization or degradation. The pressure prole extraction method proposed in this work increases the possible CNSL extraction yields and greatly reduces the amount of CO2 required for CNSL separation. 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Supercritical uid extraction; Cashew; Carbon dioxide; Pressure prole

1. Introduction The acajou or cashew nut tree (Anacardium occidentale L.) is a branchy evergreen tree that traces its origins to the Tupi Indians of northeastern Brazil (Ohler, 1979). The fruit of the tree consists of an outer shell (epicarp), a tight tting inner shell (endocarp), a strongly vesicant cashew nut shell liquid (CNSL), testa and a kernel that is highly regarded in edible nut markets. The CNSL is contained between the inner and outer shell (pericarp) in a honeycomb matrix. Besides the valuable cashew nut, the cashew nut tree is the source of many useful by-products including the peduncle of the nut or pseudofruit (cashew apple) that is used to make juices and wines (Da Silva et al., 2000), the pseudofruit waste that has been proposed as a feedstock for protein enriched animal feed or as a high-fructose source (Azevedo and Rodrigues, 2000), the tree gum

Corresponding author. Tel.: +81-22-217-7247; fax: +81-22-2177293. E-mail address: smith@scf.che.tohoku.ac.jp (R.L. Smith Jr.).

that has been proposed as an aqueous two phase extractant substitute for fractionated dextran (Sarrubo et al., 2000), the tree bark and leaves that have been used as an ancient remedy for toothache and malaria, and the CNSL that is used industrially in phenol/formaldehyde resins, brake uids, and coatings (Tyman, 1996). World production of cashew nut kernels was 907,000 metric tons/year in 1998 (FAO, 1998). On the average, the shell makes up about 50% of the weight of the nut-in-shell (NIS) while the CNSL makes up about 15 30% of the NIS weight (Behrens, 1996), which means that a potential 544,000 metric tons CNSL/year is available. According to available statistics, only a small amount of CNSL enters the world market most likely due to diculties in removing the oil from the shell and CNSL processing. In our research, separation of CNSL from the pericarp without harsh chemical or thermal treatment is of interest since this could allow for CNSL to be used in pharmaceutical and speciality polymer formulations. CNSL consists of 8090% anacardic acids and 10 20% cardols depending upon the source (Behrens, 1996). The anacardic acids have high pharmacological activity

0960-8524/03/$ - see front matter 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 9 6 0 - 8 5 2 4 ( 0 2 ) 0 0 2 7 1 - 7

R.L. Smith Jr. et al. / Bioresource Technology 88 (2003) 17

including antitumor, inhibitory, properties, antifungal properties, hepatitis B inhibitory activity, and antivectorial activities (Toyomizu et al., 2000; Laurens et al., 1997). However, CNSL is dicult to remove from the shell with high yields due to the hard outer shell, the intricate honeycombed features of the pericarp and the thermally sensitive nature of the CNSL. Methods for removing CNSL from the shell include the artisanal method of roasting, hot-oil bath method, steam processing at 270 C, quick roasting at 300 C, cold methods, and solvent extraction (Behrens, 1996). In the hot-oil method used industrially, raw humidied cashew nuts are placed on a slow moving conveyor belt and submerged in a hot bath of CNSL at 180190 C (Tyman, 1996) for about 90 s. The heat causes the shell to become embrittled and the CNSL to seep out of the shell. According to Tyman (1996), the yield of CNSL with the hot-oil process is about 10% of the total NIS weight (40% maximum theoretical) but it invariably causes some of the CNSL to undergo polymerization due to the temperatures used. Thermal methods generally cause decarboxylation of the anacardic acids as shown in Fig. 1. Tyman et al. (1989) evaluated liquid extraction of the shell material with carbon tetrachloride, light petroleum, or diethyl ether and found that these solvents could give 1530% of the total NIS weight. When this is combined with catalytic decarboxylation to prevent polymerization, near theoretical yields of technical grade CNSL can be obtained. However, long extraction times (22336 h), large amounts of solvent, harsh mechanical pretreatment, and low yield of whole kernels make this process unattractive. Furthermore, solvent extraction of CNSL tends to extract undesirable colored compounds from the shell material (Tyman, 1996). Methods based on ammoniated solutions and mixed organic solvents have been proposed that can recover virtually all of the CNSL (Paramashivappa et al., 2001), but these can be expected to have substantial environmental considerations due to solvent recovery and recycle. Supercritical uids have been suggested as attractive alternative solvents for many organic solvents by Saito (1995), who reviewed research activity. For example, CO2 has been suggested as a replacement for n-hexane or toluene in desolventation of polymer solutions by Inomata et al. (1999). In their supercritical states, CO2

and water can be used in many extractions and reactions and have been recognized as earth compatible solvents (Arai and Adschiri, 1999). CO2 has been proposed for use in a process for improving the digestion properties of cellulosic material by Kim and Hong (2001). Furthermore, CO2 in its supercritical state has been used for fractionating thermally sensitive natural oils (Suzuki et al., 1997) and has the potential for rening the extracts reported in this study. Previous research exists on extraction of CNSL with supercritical carbon dioxide (SCCO2 ). Shobha and Ravindranath (1991) studied extraction of CNSL with SCCO2 and found that 18.7% yields (%75% maximum theoretical yield) could be obtained. Those authors suggested that raw undecarboxylated CNSL could be used. Arai et al. (1993) considered supercritical carbon dioxide with isopropanol cosolvent to concentrate cardanol in the critical phase and cardol in the residue. However, neither of these works examined the extraction yields in detail. Further, the results reported in those works showed that large quantities of CO2 solvent were required in the extractions. In this research, our objective was to evaluate supercritical carbon dioxide as a solvent for separating CNSL from the raw cashew nut shells and to examine methods to improve the extraction yields.

2. Methods Unshelled cashew nuts were obtained from Indonesia (Sumbawa Island) through courtesy of BPP Teknologi (Jakarta). Unshelled nuts from the Philippines (Cagayan de Oro) were used in some of the runs. Carbon dioxide (99.99%) was supplied from Nippon Sanso and used asreceived. Cashew nut shells were prepared by cutting in half with a micro-saw removing the kernel and testa. No crushing or other pretreatment was performed. The cashew nut shells (CNS) thus prepared were weighed and placed on a ne screen that was elevated within the extractor. The average shell-only weight was 64% of the total nut-in-shell (NIS) weight and within a range of 61 66%. 2.1. Experimental apparatus The apparatus used for contacting the CNS with SC CO2 is shown in Fig. 2. The extractor had a volume of 500 cm3 that was rated for working pressures of 40 MPa at 100 C. The CO2 supply pressure to the extractor was controlled with a backpressure regulator. Two methods were used for extraction: (i) typical extraction method and (ii) pressure prole extraction method. Only the weights of CNSL extracted and general CNSL characteristics are reported here. CNSL component analysis by HPLC will be reported in a future work.

Fig. 1. Heating CNSL results in decarboxylation of anacardic acids and formation of cardanol or polymerization products.

R.L. Smith Jr. et al. / Bioresource Technology 88 (2003) 17

Fig. 2. Supercritical uid extraction apparatus.

2.2. Typical extraction method In the typical extraction method, CNS was loaded and the extractor was sealed and brought to extraction temperature. Then, CO2 was compressed to the desired pressure and allowed to ow into the extractor. After the pressurization, the metering valve was controlled to achieve a CO2 owrate of 0.55 L/min measured at standard temperature and pressure. Extract was collected in the trap periodically at intervals of about 1 h. Details of the apparatus are given in Fig. 2. 2.3. Pressure prole extraction method

depressurization rate. This generally required about 15 min for depressurizations from pressures of 10 MPa and 30 min for depressurizations from pressures of 20 or 30 MPa. Extraction periods for all runs reported in the tables do not include the pressurization and depressurization times. However, the mass of CO2 reported in the tables and gures includes the amount of CO2 required to pressurize the extraction vessel at the given conditions assuming a total extractor volume of 500 cm3 .

3. Results 3.1. Typical extraction method In the pressure prole extraction method, extractions were performed as a series of typical extraction methods but with one or more intermediate depressurization steps to 0.1 MPa. During the time at atmospheric conditions, the extractor and trap were isolated from the surroundings. 2.4. Pressurization and depressurization Times required for pressurization were approximately 5, 10, and 20 min for operating pressures of 10, 20 and 30 MPa, respectively. After the end of a given extraction period, the CO2 feed supply was closed o and the metering valve was opened further to achieve the desired Weights of CNSL extracted for the typical extraction method are shown in Table 1 for various pressures, temperatures, owrates, extraction time and loadings. From the data in Table 1, it was clear that supercritical carbon dioxide could extract CNSL from the cut shells. Color of the extracted CNSL was clear with a light brown/reddish tint. Cooling of the liquid to 4 C gave ne needle-like crystals indicating that the extract contained high purity components. According to the amount of CNS loaded, the theoretical amount of CNSL available for extraction can be estimated to be 16.5, 2, 6.7, 12.7 g, for runs 0, 1, 2, 3, respectively. However, the amounts of CNSL obtained

Table 1 Experimental conditions and data for extraction of cashew nut shells with the typical extraction method Run no.a P (MPa) T (C) Flowrate CO2 @STP (L/min) 3.0 3.0 0.5 5.0 Extraction time (h) 2 1 3 4 CO2 used (kg) 0.20 0.10 0.31 0.41 WCNS loaded (g) 50.0 6.001 20.123 37.974 WCNSL extracted (g) 2.00 0.07 0.50 1.30

0 1 2 3

30.0 14.7 29.4 29.4

60 40 60 60

CNS: cashew nut shells; CNSL: cashew nut shell liquid.

R.L. Smith Jr. et al. / Bioresource Technology 88 (2003) 17

with the typical extraction method were far below the theoretical amounts. Shobha and Ravindranath (1991) reported on a single experimental trial for extraction of CNSL with supercritical CO2 at 40 C and 25 MPa. In their work, approximately 80 kg CO2 was used over a 17.5 h extraction period to obtain 56.1 g CNSL from broken CNS from 300 g NIS. From their work, 0.7 g CNSL was obtained for every kilogram of CO2 . From the data in Table 1, it became clear that greater amounts of CNSL could be obtained for given amounts of CO2 solvent. This could be due to a number of factors such as CNS loading amount and placement in the extractor, mass transfer or the extraction conditions used. The extractor was opened at the end of each of these runs and it was found that liquid was on the outer portion of the shell. Just after removal of the shells from the extractor, a curious bubbling was observed, which indicated that the CO2 had high solubility in the CNSL and possibly caused the liquid phase to expand. Following up on this observation, Setianto et al. (2001) examined changing the pressure during the extraction. This led to our study of the pressure prole extraction method as described next.

3.2. Pressure prole extraction method At the completion of run 3, the apparatus was depressurized and left overnight. Then, the apparatus was repressurized and extraction was continued for 4 h at the same conditions. This is run 4 in Table 2. Unexpectedly, additional CNSL extract was obtained as shown in Table 2. Data obtained from several other

experimental runs made with a depressurization step are also shown. Pressure prole extraction methods for these runs and others are summarized in Fig. 3. As shown in Table 2, the amount of CNSL obtained with an intermediate depressurization step was very consistent for three dierent loadings (runs 46). In run 7, the rst extraction period was shortened to 1 h and the later extraction segment was lengthened to 5 h. A decrease in the weight of CNSL obtained was observed, which can be attributed to the equilibration times as explained below. It should be pointed out that the amount of CO2 required to extract a given amount of CNSL in Table 2 was much less than that reported by Shobha and Ravindranath (1991). Those authors used more than 80 kg CO2 to obtain comparable amounts of CNSL. Table 3 and Fig. 3 summarize the additional studies performed with various pressure proles. A large amount of CNSL could be obtained with the various pressure proles. It should be emphasized that the pressurization steps, although as rapid as possible for the given equipment, were relatively slow and gentle towards the natural materials. Slower depressurization rates (%1 L CO2 /min at STP) were performed but without signicant dierences in the extraction yields. Rapid depressurization rates could not be considered with the present experimental setup. From the longest run made (run 12), the shells were clearly dry. Examination of shells cut in sections with a low power microscope did not reveal any free oils. However, some dried CNSL-like solids were observed. Some of the shells exhibited cracks along the internal wall of the endocarp but no common structural defects

Table 2 Data for extraction of cashew nut shells with a single intermediate depressurization step to 0.1 MPa Run no. 4a 5 6 7 Extraction time (h) 44 44 44 15 CO2 used (kg) 4.5 4.5 4.5 3.5 WNIS (g) 58.4 56.2 63.2 52.4 WCNS loaded (g) 37.974 36.986 41.309 33.239 WCNSL extracted (g) 5.989 6.212 6.134 4.961

CNS: cashew nut shells; CNSL: cashew nut shell liquid; NIS: nut-in-shell. a Run 4 NIS calculated from CNS/0.65. Extraction conditions: 60 C, 29.4 MPa, CO2 owrate of 5 L/min@STP. Symbol: : reduction of pressure to 0.1 MPa, then repressurization to extraction conditions.

Fig. 3. Pressure prole extraction methods for runs 412; run 4 includes run 3; run 12 includes run 11.

R.L. Smith Jr. et al. / Bioresource Technology 88 (2003) 17 Table 3 Data for extraction of cashew nut shells with single and multiple intermediate depressurization and pressurization steps Run no. 8 9 10 11 12a Extraction time (h) 5 min 5 min 4 5 min 1=5 3:5 5 min 14 14 1 4 4=4 Pressure (MPa) 29:4 29:4 29:4 9:8 9:8=29:4 29:4 29:4 9:8 29:4 9:8 19:6 9:8 19:6 19:6=29:4 CO2 used (kg) 3.69 5.67 2.90 2.73 6.72 WNIS (g) 53.7 56.5 53.1 56.3 56.3 WCNS loaded (g) 33.629 36.184 32.418 37.326 37.326

WCNSL extracted (g) 4.660 6.566 4.657 4.201 7.951a

Extraction conditions: 60 C; CO2 owrate of 5 L/min at STP. CNS: cashew nut shells; CNSL: cashew nut shell liquid; NIS: nut-in-shell. a Run 12 is the continuation of run 11. Mass of CO2 used and CNSL extracted include respective masses from run 11. Symbols: : depressurization to 0.1 MPa. /: intermediate pressurization to given pressure.

were observed. Therefore, we concluded that most of the CNSL had been extracted from the NIS and that the ultimate amount of CNSL available was close to 15% of the NIS weight based on the lower of the average values summarized by Behrens (1996). It should be noted that the precise value of the ultimate amount of CNSL used in Eq. (1) does not change the general conclusions of this work. In the analysis of the results, it is convenient to dene a yield of the CNSL extracted, dened as: Yield % CNSL extracted g 100% NIS g 0:15 1

Fig. 4 shows results for runs 68 and 10, which used simple pressure proles (Fig. 3). At a owrate of 5 L/min at STP, roughly 0.5 kg/h of CO2 were consumed. Values in the gure include the CO2 needed to pressurize the extractor as explained previously. The eect of the depressurization on the extraction yield for each run can be clearly seen. As explained in detail next, the eect of the depressurization on each run could be clearly observed. In runs 68, extraction pressures were held constant at 29.4 MPa. In run 6, yields began to greatly increase after the rst (4 h) extraction period (%2 kg CO2 used) and depressurization. In run 7, the extraction yield began to increase earlier due to the shorter initial (1 h) extraction period and depressurization. In run 8, two

Fig. 4. Extraction yields versus CO2 used for runs 6, 7, 8, and 10. Basis for yields: 15% of NIS being CNSL.

short (5 min) extraction periods were used, however, the extraction yield did not greatly increase after the initial (5 min) extraction as expected. From these results, it appeared that some contact time was necessary before depressurization was initiated. In run 10, we examined the eect of reducing the initial extraction pressure compared with that used in run 7. As shown in Table 3 and Fig. 3, the initial extraction pressure for run 10 was 9.8 MPa for a period of 1 h and that for run 7 was 29.4 MPa. Even for the different initial extraction pressures, the trends of the extraction yields were similar for the two runs as shown in Fig. 4. These results may be understood as follows. Generally, only small changes in the CNSL can be expected according to the pressurevolumetemperature behavior of pure oils unless there is a phase change of some of the solid components (Acosta et al., 1996). It would seem, therefore, that it is unlikely that solidliquid phase transitions occurred as a result of the hydrostatic pressure. Such a phase transition would cause the extraction yield curves for runs 7 and 10 to be signicantly dierent. Further, if the available CNSL changed according to the hydrostatic pressure on the shell, then, one would expect the two curves (runs 7 and 10) to be signicantly dierent. Since the two curves were similar, the eects of the hydrostatic pressure on the shell can be said to be similar. On the other hand, if the CO2 penetrated into the interstices of the shell matrix and dissolved into the CNSL, then for both cases there would be considerable swelling of the liquid phase, which would result in a viscosity reduction of the CNSL and a tendency for the liquid to ow. A swelling phenomenon has been reported by Smith et al. (1998), who found that the volume expansion of oils in contact with supercritical CO2 could be as high as 200% for n-decane or close to 50% for sh oil methyl esters. The possibility of CNSL swelling was conrmed to some extent by examination of the CNS after a single depressurization. The CNSL showed a frothy appearance in which slow evolution of CO2 was apparent. The CNSL swelling would increase the available surface area of the CNSL to the bulk CO2 phase and would account for some of the higher extraction yields observed and for the similarity in extraction yield curves for runs 7 and 10.

R.L. Smith Jr. et al. / Bioresource Technology 88 (2003) 17

solid-like samples showed the presence of needle-like crystals that were well dened. 3.4. Further processing of extracted CNSL and shells The pressure prole extraction methods presented in this work gave high extraction yields and CNSL that appeared to be of high quality. Analysis of the time evolution of the CNSL compositions as a future work will provide information on whether fractionation of CNSL with supercritical CO2 solvent is feasible using the inverse temperature gradient column developed by Suzuki et al. (1997) or whether optimization of the separations can be approached according to the methodology of Yonei et al. (1995). Regarding the remaining shell material, supercritical water has been found to be a useful solvent for converting cellulose and biomass materials into valuable pharmaceuticals and sugar compounds (Adschiri et al., 1993; Malaluan, 1995; Sasaki et al., 1998). In some preliminary experiments (Smith et al., 2001), we have found that the cashew nut shell can be dissolved completely into supercritical water using techniques described in Fang et al. (1999). This means that it is highly probable that a comprehensive process for the bioresource cashew tree can be developed, which uses only carbon dioxide and water as solvents. 4. Conclusions In this work, we have examined the separation of CNSL with supercritical carbon dioxide. It was found that extractions performed with intermediate depressurization steps increased the yields of CNSL obtained and decreased the amount of CO2 required. Contact period seemed to be an important factor in the initial pressurizations. Further, staged pressurizations seemed to provide increased extraction yields. A number of pressure prole extraction methods were used and yields as high as 94% (15% NIS basis) were obtained. In the initial extraction period, dissolution of CO2 into the CNSL phase probably reduced the CNSL viscosity and caused the liquid phase to expand. Then, perturbations made in system conditions probably led to increased available surface area of CNSL to the bulk CO2 . Experiments are needed regarding the phase behavior of CNSL with supercritical CO2 to allow understanding of the fundamental phenomena that are occurring during the extraction. However, use of supercritical CO2 with the pressure prole extraction method is an eective technique for separating CNSL from bulk cashew nut shells. Acknowledgements The authors wish to thank Mr. Priyo Atmaji (BPP Teknologi, Jakarta) for supplying the raw Indonesian

Fig. 5. Extraction yields versus CO2 used for runs 9 and 12 (includes run 11). Basis for yields: 15% of NIS being CNSL. Basis for Ref. FAO, 1998: 25% of NIS being CNSL.

Next, we considered more complicated pressure proles as shown in Fig. 5. In run 9, a short pressurization period at 9.8 MPa was used followed by a two-step pressurization. Extraction yields increased rapidly during the pressurizations. After approximately 3 kg of CO2 was consumed, slow depressurization followed by pressurization to 29.4 MPa gave a new clear pattern in the yield as shown in Fig. 5. A nal depressurization and repressurization resulted in a slight increase as shown by the last three points of run 9. In run 12, a 1 h extraction period at 9.8 MPa was used before depressurization. Upon repressurization, the extraction yield increased as expected and seemed to reach a plateau. In the next pressurization step at approximately 4 kg CO2 used (%6 h), a two-stage pressurization gave a considerably high extraction yield. The eect of the pressure proles can be clearly seen for each step. Fig. 5 also shows the initial experimental data from Shobha and Ravindranath (1991), who performed extractions at 40 C and 25 MPa. Although the basis for the extraction yields and conditions are dierent, the slopes of the extraction yield curves from the data in this work were similar to those of the literature during the initial extraction period before the depressurization step. These results conrm that a large quantity of CO2 would be required to obtain the desired extraction yields with the typical extraction method. 3.3. CNSL characteristics and quality The CNSL obtained in all runs was clear with a light brownish red tint at room temperature. No signs of polymerization or degradation of the CNSL product were apparent from visual analysis. In some cases, the extracted CNSL was brighter in color. Most samples became solid or exhibited a solid-like appearance when placed in a refrigerator at 4 C. Examination of the

R.L. Smith Jr. et al. / Bioresource Technology 88 (2003) 17

cashew nuts used in this study. The authors also wish to thank Mikio Kikuchi for assistance with the experimental apparatus and the many sample preparations. References
Acosta, G.M., Smith Jr., R.L., Arai, K., 1996. High-pressure PVT behavior of natural fats and oils, trilaurin, triolein, and n-tridecane from 303 to 353 K from atmospheric pressure to 150 MPa. J. Chem. Eng. Data 41, 961969. Adschiri, T., Hirose, S., Malaluan, R., Arai, K., 1993. Noncatalytic conversion of cellulose in supercritical and subcritical water. J. Chem. Eng. Jpn. 26, 676680. Arai, K., Adschiri, T, 1999. Importance of phase equilibria for understanding supercritical uid environments. Fluid Phase Equilibria 158160, 673684. Arai, K., Ajiri, T., Suzuki, S., Nishimura, M., 1993. Method for separating and purifying cashew nutshell juice. (Japanese Patent) JP 05-000979, August 1. Azevedo, D.C.S., Rodrigues, A., 2000. Obtainment of high-fructose solutions from cashew (Anacardium occidentale) apple juice by simulated moving-bed chromatography. Sep. Sci. Technol. 35 (16), 25612581. Behrens, R., 1996. Cashew as an Agroforestry Crop, Tropical Agriculture [9]. Margraf Verlag, Weikersheim, Germany. Da Silva, K.D.P., Collares, F.P., Finzer, J.R.D., 2000. A simple and rapid method for estimating the content of solids in industrialized cashew juice. Food Chem. 70, 247250. Fang, Z., Smith Jr., R.L., Inomata, H., Arai, K., 1999. Phase behavior and reaction of polyethylene terephthalatewater systems at pressures up to 173 MPa and temperatures up to 490 C. J. Supercrit. Fluids 15, 229243. Food and Agriculture Organisation (FAO) Yearbook, 1998. Food Statistics Series no. 148, 52, p. 167. Inomata, H., Honma, Y., Imahori, M., Arai, K., 1999. Fundamental study of de-solventing polymer solutions with supercritical CO2 . Fluid Phase Equilibria 158160, 857867. Kim, K.H., Hong, J., 2001. Supercritical CO2 pretreatment of lignocellulose enhances enzymatic cellulose hydrolysis. Biores. Technol. 77, 139144. Laurens, A., Fourneau, C., Hocquemiller, R., Cave, A., Bories, C., Loiseau, P.M., 1997. Antivectorial activities of cashew nut shell extracts from Anacardium occidentale L.. Phytother. Res. 11, 145 146. Malaluan, R.M., 1995. Study on cellulose decomposition in subcritical and supercritical water. Ph.D. Thesis, Tohoku University (in English). Ohler, J.G. Cashew, 1979. Communication 71, Koninklijk Institut voor de Tropen, Amsterdam.

Paramashivappa, R., Kumar, P.P., Vithayathil, P.J., Srinivasa Rao, A., 2001. Novel method for isolation of major phenolic constituents from cashew (Anacardium occidentale L). nut shell liquid. J. Agric. Food. Chem. 49, 25482551. Saito, S., 1995. Research activities on supercritical uid science and technology in Japana review. J. Supercrit. Fluids 8, 177 204. Sarrubo, L.A., Oliveira, L.A., Porto, A.L.F., Duarte, H.S., CarneiroLeao, A.M.A., Limo-Filho, J.L., Campos-Takaki, G.M., Tambourgi, E.B., 2000. New aqueous two-phase system based on cashew-nut tree gum and poly(ethylene glycol). J. Chromatogr. B 743, 7984. Sasaki, M., Kabyemela, B., Malaluan, R.M., Hirose, S., Takeda, N., Adschiri, T., Arai, K., 1998. Cellulose hydrolysis in subcritical and supercritical water. J. Supercrit. Fluids 13, 261268. Setianto, W.B., Smith Jr., R.L., Inomata, H., 2001. Application of supercritical CO2 to the extraction of cashew nut shell oil. Published proceedings of the Japan Oil Chemists Society, 40th annual meeting, Sendai, October 45, paper 1E-05 (in Japanese). Smith Jr., R.L., Yamaguchi, T., Sato, T., Suzuki, H., Arai, K., 1998. Volumetric behavior of ethyl acetate, ethyl octanoate, ethyl laurate, ethyl linoleate, and sh oil ethyl esters in the presence of supercritical CO2 . J. Supercrit. Fluids 13, 2936. Smith Jr., R.L., Malaluan, R.M., Setianto, W.B., Inomata, H., Arai, K., 2001. Biobased manufacture of alkenyl phenolics and polysaccharide feedstocks from cashew nut with supercritical water and carbon dioxide processing. Proceedings of the 5th Biomass Conference of the Americas (cancelled), September 1721, Orlando, Florida, session 26. Shobha, S.V., Ravindranath, B., 1991. Supercritical carbon dioxide and solvent extraction of the phenolic lipids of cashew nut (Anacardium occidentale) shells. J. Agric. Food Chem. 39, 2214 2217. Suzuki, H., Chen, J., Adschiri, T., Inomata, H., Arai, K., 1997. Purication of EPA ethyl ester from sh oil with a supercritical CO2 temperature gradient rectication column. JOCS 46, 331 340. Tyman, J.H.P., Johnson, R.A., Muir, M., Rokhgar, R., 1989. The extraction of natural cashew nut-shell liquid from the cashew nut (Anacardium occidentale). J. Am. Oil Chem. Soc. 66, 553557. Tyman, J.H.P., 1996. In: Synthetic and Natural Phenols Studies in Organic Chemistry, vol. 52. Elsevier, Amsterdam, pp. 518521. Toyomizu, M., Okamoto, K., Ishibashi, T., Chen, Z., Nakatsu, T., 2000. Uncoupling eect of anacardic acids from cashew nut shell oil on oxidative phosphorylation of rat liver mitochondria. Life Sci. 66 (3), 229234. Yonei, Y., Kobari, M., Yokoyama, C., Arai, K., 1995. A search for optimum extraction conditions of avor from tabacco leaves by high pressure CO2 . Kagaku Kogaku Rombunshu 21, 674682 (in Japanese).