Ind. Eng. Chem. Res. 2009, 48, 6783–6788
6783
Ethanol Dehydration by Adsorption with Starchy and Cellulosic Materials Julian A. Quintero and Carlos A. Cardona* Departamento de Ingenierı´a Quı´mica, Plantas Piloto de Biotecnologı´a y Agroindustria, UniVersidad Nacional de Colombia Sede Manizales, Cra. 27 No. 64-60, Manizales, Colombia
Adsorption with starchy and cellulosic materials was evaluated as an energy efficient technology for ethanol dehydration. Corn (Zea mays), upright elephant ear (Alocasia macrorrhiza), cassava (Manihot esculenta), and sugar cane bagasse (Saccharum), were used as sources of starch and cellulosic fibers and further used as adsorbents for water removal. Enzymes (R-amylase and cellulase for starch and cellulose, respectively) were evaluated as modifying agents with the aim of increasing water adsorption capacity. Native and enzymetreated materials were examined with SEM, XRD, and BET analysis to determine the hydrolysis influence on structure and superficial area. Partial hydrolysis was obtained under operation conditions evaluated, as was shown by the SEM micrographs. Crystallinity increased with the enzymatic hydrolysis for all materials. For all materials, superficial area was not increased with enzymatic hydrolysis but water adsorption capacity was improved reaching an anhydrous product with enzyme-treated materials. Water adsorption capacity ranges from 4 to 19 g/ 100 g adsorbent were found for evaluated materials. Corn starch had the highest water adsorption capacity (19 g/ 100 g ads) while upright elephant ear starch presented the lowest (4.2 g/ 100 g ads). Tested materials showed affinity with water for both native and enzyme-treated cases. This allowed water to adsorb from the feed for obtaining anhydrous ethanol. It was suggested that water adsorption capacity was increased because of the more exposed hydroxyl groups from the polymeric structures evaluated (starch and cellulose). Applicability of these materials was oriented to small scale ethanol production plants in developing countries where rural areas are the objective and small productions are expected. 1. Introduction The removal of water from ethanol using a fixed-bed adsorber is a well-known process using mainly molecular sieves as adsorbents. Starch, starch-based materials, cellulose and hemicellulose have an affinity for water.1-3 The adsorption properties have been studied for corn starch, corn grits, and particles synthesized from a mixture of corn starch and either corn cobs or hemicellulose.3-11 Other materials less referenced as adsorbents for water-ethanol separation are cornmeal, wheat flour, and wheat straw.12 Starch is a mixture of amylose, a linear polymer of D-glucose units joined by R-1,4 bonds, and amylopectin, a polymer of linear, 1,4 D-glucose chains linked at branches points by R-1,6bonds. The polar attraction of water and the hydroxyl groups of the anhydrous glucose units are the main force for water adsorption on starch. In zeolites, the silica/alumina ratio can be adjusted to give a higher affinity for water and other polar molecules.6,13 The mechanism of adsorption of water is understood to involve hydrogen bonding with the hydroxyl groups on the starch chains.6,14,15 Both types of starch chains, amylose and amylopectin, interact with water molecules in this manner. However, the amylopectin structures also physically trap water molecules in the matrix of chain branches. When the water molecules are trapped this way, some of the nearby hydroxyl groups become unavailable for hydrogen bonding.15 Generally, it is accepted that water sorption isotherms on hydrophilic polymers have a sigmoidal shape and are type II isotherms, according to Brunauer’s classification.16 A similar relationship between polar interactions is evident in starch-based materials.6 * To whom correspondence should be addressed. Tel.: + 57 6 8879300, ext: 50417. Fax: + 57 6 8879300, ext: 50452. E-mail:
[email protected].
Distillation is used to enrich the ethanol from about 10% (wt) in the broth to over 90% (wt). At atmospheric pressure, simple distillation can produce a maximum ethanol concentration of 95.6% (wt) (azeotropic composition). Compared to distillation, which requires about 2.8 × 106 J/L of ethanol to break the azeotrope using benzene, adsorption is energy efficient with estimates of 5.6 × 105 J/L to reach 99.6% (wt) ethanol from a 90% (wt) solution using starch,4 2.0, 24 > 106 J/L to go from 85% to 99.5% (wt) using cornmeal.17 Adsorption systems (using mostly molecular sieves, barium and calcium chloride, and others) are considered as energetically efficient technologies compared to conventional technologies using benzene and ethylene glycol.1 Main operational costs using starchy and cellulosic materials as adsorbents are caused for their regeneration (drying, compressing, and gas heating).10 In this work, water adsorption by starchy and cellulosic materials has been evaluated as a potential process based on non-conventional adsorbents for alcohol dehydration. Three different starches were evaluated, corn starch (recovered from Zea mays L plants), cassava starch (recovered from Manihot esculenta plants) and upright elephant ear starch (recovered from Alocasia macrorrhiza plants, commonly known as “bore” in Latin American countries). The cellulosic material used was sugar cane baggase (Saccharum). Ethanol dehydration capacity of evaluated materials was increased by their modification with enzymes (enzymatic hydrolysis). Adsorption capacity was determined by their operation on the ethanol dehydration process. 2. Materials and Methods 2.1. Raw Materials. Two of the raw materials, cassava and yellow dent corn, were acquired from a local market in the center of Colombia. Sugar cane bagasse was obtained from a small mill used for producing commercial sugar cane juice. Upright elephant ear roots were collected from a farm of the National
10.1021/ie8015736 CCC: $40.75 2009 American Chemical Society Published on Web 06/18/2009
6784
Ind. Eng. Chem. Res., Vol. 48, No. 14, 2009
Service of Learning (SENA) Manizales, Colombia. These raw materials were washed with enough water to eliminate impurities. Cassava and upright elephant ear were peeled manually and processed for starch recovery. Corn grains were separated from the corn cobs and processed for starch recovery. Sugar cane bagasse was milled in a mill blade (Wanvig, Ind. Colombiana Modelo CSM 20 L) for wet materials, the obtained fibers were dried for 48 h at 75 °C, and sieved (U.S.A. Standard Testing Sieve ASTME-11 Specification, USA). Each fiber fraction was milled in a mill blade (Thomas-WILEY, Laboratory Mill Model 4, Arthur H. Thomas, Company, Philadelphia, PA, USA) for dry materials. Each milled fraction was sieved again. 2.2. Starch Recovery. Cassava and upright elephant ear starch were recovered by the conventional method used at the industrial units dedicated to extraction and transformation of cassava starch. This method consists of root washing, peeling, milling and filtering.18 Starch solution obtained from filtering was decanted for 24 h. Residual liquid was extracted, and the solid starch was dried on an oven at 60 °C at a constant weight. Corn starch was recovered using the method exposed by Haros and Sua´rez,19 in which 100 g of corn was steeped in 500 mL of sufficient sodium bisulfite solution to give a dioxide concentration of 0.25% in distilled water and pH 3.5-4.0. The samples were held at 52 °C for 48 h and shaken periodically during the initial hours of steeping. After this time, the steep water was decanted and the excess of liquid water was removed from the corn by blotting. The purity on dry basis of upright elephant ear starch, corn starch, and cassava starch were 74.03 ( 0.52%, 73.92 ( 0.41%, and 74.03 ( 0.66, respectively, assuming that protein constitutes the rest part of the material. For sugar cane bagasse, the crust corresponded to 92.85 ( 0.002% of the dry matter while marrow was only 7.15 ( 0.002% of the dry matter. 2.3. Enzymatic Hydrolysis. The agents used for the modification were enzymes, which permit to accomplish a partial hydrolysis under special conditions. R-amylase (HT-340 L, Bacterial source: Bacillus licheniformis; Proenzimas, Cali, Colombia) was selected as one of the enzymatic modifying agent because of its action on starch polymers. Cellulase (CELULASE CE 2, Fungal source: Trichoderma longbrachiatum; Proenzimas, Cali, Colombia) was used for sugar cane bagasse modification. Before hydrolysis, all materials were dried in an oven for 48 h at 60 °C. Dried starches were milled in a ball mill (KM1, Germany) obtaining a dry powder. Enzymatic hydrolysis was carried out in a shaker (UNITRONIC OR, J.P. Selecta, S.A., Barcelona). Enzymatic modifications were based on procedures exposed by Beery et al.20 The modification of the starch with R-amylase occurred by preparing a 1/50 dilution of the R-amylase in phosphate buffer. The modification took place at pH 6.9 at 25 °C for 7 days. After 7 days, the starch was washed repeatedly and then dried in a convection oven for 24 h at 40 °C. Cellulose hydrolysis was carried out with a 1/2 and 1/6 cellulase solution in citrate buffer at a ratio of 10 mL of the enzyme solution to 1 g of sugar cane bagasse. The mixture was kept at 50 °C in a shaker for 6 days. After the enzymatic hydrolysis of the starch and sugar cane bagasse, the reducing sugar concentrations were measured by Nelson and Somogyi method.21 Two replicates were evaluated for determing the enzymatic effect and the reducing sugar concentrations. 2.4. Structural Analysis and Surface Area Measurement. All samples were dried in a vacuum oven at 60 °C for 3 days. SEM micrographs were obtained in a JEOL JSM-5910LV microscope. The X-ray diffraction plots were obtained in a
Figure 1. Diagram of experimental apparatus for fixed-bed adsorption: 1, adsorber; 2, adsorbent; 3, Q-ring support packing; 4, feed; 5, electrical heater; 6, water bath; 7, condenser; 8, product.
Rigaku (MiniFlex II) unit. Superficial area was determined in a Micromeritics ASAP 2020 (Micromeritics Instrument Corporation) unit. 2.5. Water Adsorption. The device used for ethanol dehydration, using starches and sugar cane bagasse as desiccant, was a glass column. The experimental apparatus is shown in Figure 1. The adsorber was made of a glass shell with a water bath jacket. Temperature points represented in Figure 1 by T1-T4 were arranged from the feed to the end of the adsorbent. The ethanol-water mixture was stored in a balloon glass that was heated by an electrical heater. The temperature of the adsorber was well above the dew point of the vapor feed to avoid condensing. The composition of the samples was measured with a refractometer ABBE (Bausch & Lomb). Three replicates of the calibration curve were measured and three points of the curve were verified by gas chromatography. Predried adsorbent was packed in the adsorber and supported by Q-ring packing, and the temperature of the bed was stabilized for 2 h at 90 °C. The vapor feed went through the adsorber, and the ethanol-rich product was collected at the end of the condenser. Product samples were collected at intervals of 1 min for the first 60 and 5 min for the remainder and were analyzed to determine the breakthrough curve. Temperature profiles were determined with the recorded data from each temperature point (see Figure 1). Two replicates were obtained for each material evaluated on adsoption. In order to obtain the amount of water adsorbed in the solid phase, a mass balance was used between the phases, where alcohol was considered to be the nonadsorbable component for simplicity of the adsorption capacity calculation. This hypothesis agrees with that reported by Hong et al.2 and Rebar et al.,15 who found that ethanol adsorption was negligible after the adsorbent had been used for several cycles, and the chemical affinity of starch to water is much greater than that of alcohol. However, Hassaballah and Hills,17 Robertson et al.,22 and Crawshaw and Hills23 reported that some ethanol was adsorbed on starch materials under the conditions that resulted in high water loading. Chang et al.24 showed that a significant amount of ethanol is adsorbed on the adsorbent when equilibrium is reached. 3. Results and Discussion 3.1. Enzymatic Hydrolysis. Enzymatic hydrolysis conditions are shown in Table 1. The highest starch conversion was
Ind. Eng. Chem. Res., Vol. 48, No. 14, 2009
6785
Table 1. Enzymatic Hydrolysis Parameters raw material
dilutiona
activity
T (°C)
pH
% conversiond
upright elephant ear starch corn starch cassava starch sugar cane bagasse sugar cane bagasse
1:50 1:50 1:50 1:2 1:6
340 000 MWU/mLb 340 000 MWU/mLb 340 000 MWU/mLb 11 000 ECU/g ( 5%c 11 000 ECU/g ( 5%c
25 25 25 50 50
6.9 6.9 6.9 4.9 4.9
15.17 ( 0.28 11.41 ( 0.62 13.28 ( 0.54 12.09 ( 0.75 6.32 ( 0.10
a Defined as enzyme volume: solution volume. b MWU ) Modified Wohlhemuth ) one unit is the quantity of enzyme that will dextrinize one milligram of starch in 30 min at pH 6 and 95 °C. c ECU/g ( 5% ) one unit is the quantity of enzyme that will hydrolyze one gram of de hydroxiethilcellulose at 50 °C and pH 4.8. d Defined as (gram glucose formed/gram material initially present)*100.
Figure 2. SEM micrograph of native (a) and enzyme-treated (b) cassava starch granules.
accomplished with upright elephant ear starch, followed by cassava starch, meaning that starch granules from roots are more exposed to enzymatic attack, as will be seen further. Sugar cane bagasse with a dilution of 1:2 presents greater conversion than when a dilution of 1:6 was used, as was expected due to the lowest enzyme quantity. Similar results were shown by Berry et al.,20 who obtained percentage conversions of 14.3% and 13% for corn grits and corn cobs, respectively, using amylase and cellulase at similar dilution conditions. In this step, R-amylase breaks only the R-1,4 linkages presented in amorphous zones producing glucose units. Partial hydrolysis is accomplished by the correct hydrolysis conditions. This procedure was necessary for holding the starch granule structure while exposing the hydroxyl groups of the amorphous starch chain. A similar process occurs with the cellulose from bagasse. Upon partial acid hydrolysis, cellulose was broken down into cellobiose (glucose dimer), cellotriose (glucose trimer), and cellotetrose (glucose tetramer), whereas upon complete acid hydrolysis, it is broken down into glucose. The β-1,4 glycoside linkages cause cellulose to be in a low surface area crystalline form. 3.2. Structural Analysis and Surface Area Measurement. Enzyme-treated upright elephant ear starch tends to get agglomerated, while native granules were scattered. Modified starch granules changed their diameters to 1-5 µm; the shape was altered, obtaining irregular shapes with a rough surface. Enzyme-treated corn starch granules kept their size but present some fractures by enzymatic attack. Enzyme-treated cassava starch granules, as well as upright elephant ear starch, tend to agglomerate but to a lesser extent. Cassava starch evidence the deterioration by enzymatic action, however, the granule sizes were kept. Enzymatic treatment generated deformations and breaking of the granules but not pores over the particle surface (see Figure 2). In general, for all starches, enzyme treatment did not create new visible pores, but did cause some fractures and shape alterations. The most appreciable enzymatic effect occurred with upright elephant ear starch granules. Hydrolyzed bagasse fibers with an enzyme/buffer ratio of 1:1 showed the layers breaking by enzymatic action exposing the inner channels of the fiber; the length was kept after hydrolysis. Bagasse fibers hydrolyzed with an enzyme/buffer ratio of 1:5, presented minor signs of enzymatic action.
XRD patterns are similar and characteristic of starches, the differences depend on the natural source of each starch. Corn starch is classified as type A, because it presents the highest peaks at around 2θ values of 14.92°, 17.18°, 17.8°, and 22.7°, characteristic angles of this type. Upright elephant ear and cassava starch present the maximum peak at 16.82° and 16.98°, respectively, moreover they have less intense peaks at 2θ values of 20.80°, 22.86°, and 24.20° for upright elephant ear starch and values of 19.70°, 22.8°, and 24.2° for cassava starch. This involves the fact that upright elephant ear and cassava starch are type B. Similar peaks were found by Jiping et al.,25 who has studied the morphological and crystalline properties of B-type (Fritillaria) and C-type (Rhizoma dioscorea and Radix puerariae) starches modified by acid hydrolysis. Several studies26-28 have demonstrated that the A-type starch was associated mainly with cereal starches, such as corn starch and wheat starch. The X-ray patterns of these kinds of starch gave the stronger diffraction peaks at around 15°, 17°, 18° and 23°. The B-type starch was usually obtained from tuber starches, such as potato starch and canna starch. The strongest diffraction peak of the X-ray diffraction pattern appeared at 17° 2θ. There were also a few small peaks at 2θ values of 20°, 22°, and 24°. The C-type starch was a mixture of both A and B types, such as smooth-seeded pea starch and various bean starch. Crust and marrow bagasse have different structures and crystallinity because the crust bagasse presents two diffraction peaks at 2θ values of 18.04° and 21.9°, while marrow bagasse presents only a peak at 21.86°, characteristic of the cellulose structures. Similar diffraction peaks have been reported for cellulose from sugar cane bagasse and methylcellulose.29,30 Nitrogen adsorption isotherms were obtained with the BET method. All starchy materials exhibited a type II (Brunauer’s classification) isotherm with hysteresis from the middle zone of the curve, indicating the presence of meso and macro pores. Sugar cane bagasse also showed a type II isotherm but with the hysteresis more pronounced, meaning that at high pressures, condensation in meso and macro pores is higher and is more difficult to evaporate the nitrogen, and it remains within the pores. Generally, it is accepted that the water sorption isotherms are hydrophilic polymers with a sigmoidal shape and are typeII isotherms, according to the results obtained by Lee et al.5 In the same way, Al-Muhtaseb et al.31 have evaluated water
6786
Ind. Eng. Chem. Res., Vol. 48, No. 14, 2009
Table 2. Structural and Adsorption Parameters of the Evaluated Materials
a
raw material
superficial area BET (m2/g)
Dp BJHa (Å)
Vp BJHb (cm3/g)
upright elephant ear starch enzyme-treated upright elephant ear starch corn starch enzyme-treated corn starch cassava starch enzyme-treated cassava starch sugar cane bagasse enzyme-treated sugar cane bagasse
2.1695 ( 0.0193 2.8686 ( 0.0219 0.9816 ( 0.0039 0.8267 ( 0.0064 0.4984 ( 0.0042 0.6625 ( 0.0163 1.0058 ( 0.0913 1.0945 ( 0.0436
181.967 96.186 157.670 168.129 143.669 208.422 136.012 172.946
0.008031 0.007675 0.003761 0.002807 0.001266 0.001125 0.003765 0.003266
BJH Desorption average pore diameter. b BJH Desorption cumulative volume of pores.
Table 3. Operation Parameters of the Adsorption with Starchy and Cellulosic Materials raw material (RM)
ads weight (g)
packing density (g/cm3)a
upright elephant ear starch enzyme-treated upright elephant ear starch corn starch enzyme-treated corn starch cassava starch enzyme-treated cassava starch sugar cane bagasse enzyme-treated sugar cane bagasse
6.409 6.478 10.258 10.121 9.900 2.967 2.059 2.124
0.2590 0.2618 0.3869 0.3579 0.4001 0.3358 0.0832 0.0858
a Adsorbent packing density. adsorbent.
b
Ethanol concentration of the feed.
c
adsorption by potato starch and it was concluded that starchy materials show a type II isotherms. Upright elephant ear starch was the material with the highest superficial area of both native and enzyme-treated cases (2.17 and 2.87 m2/g, respectively). Enzymatic hydrolysis did not significantly change the superficial area for all tested materials. Obtained superficial areas were slightly higher than that reported by Beery and Ladisch11 and Beery et al.20 for native and modified corn grits. The specific surface areas obtained are low compared with other commercial materials,6 a fact which generates concern about viability for using these materials as adsorbents for alcohol dehydration; however, its evaluation in the dehydration process showed that adsorption capacity was increased by enzymatic hydrolysis, reaching anhydrous ethanol composition. Last, we suggest that enzymatic treatment may have exposed more functional groups (hydroxyl groups) or changed their orientation to become more accessible to the water molecules, and they are not expressed by the BET method in the surface area, this result agrees with the work of Beery and Gulati32 for corn grits. Pore diameters obtained by BJH method (procedure developed by Barrett, Joyner, and Halenda)16 confirmed the existence of macro pores in the studied materials (see Table 2). 3.3. Water Adsorption. All materials were tested in the adsorption operation in both native and enzyme-treated cases. Water was adsorbed in all materials evaluated but in different extend. Only enzyme-treated materials, except enzyme-treated upright elephant ear starch, reached anhydrous composition. Operation parameters and results of the adsorption process are listed in Table 3. The packing density of each material was used as the fixed parameter jointly with bed height for determining the mass of adsorbent. Packing density values over 0.38 g/cm3 caused system pressurization because the adsorbent at these values became an obstruction for feed vapors, and they cannot cross the bed. But a lower packing density could leave free spaces for feed vapors avoiding their contact with the adsorbent. In the studied regime of operation conditions, the possibility of having feed vapor temperatures greater than those of the bed temperatures was avoided. This phenomenon could appear only if the system is autopressurized depending on some problems
CETOH Feedb (wt. %)
CETOH Productc (wt. %)
CETOH Equilibriumd (wt. %)
water adsorbed (g/ 100 g ads)
93.05 92.53 91.81 92.01 91.87 91.01 91.81 92.92
97.16 ( 0.00 98.72 ( 0.12 95.45 ( 0.40 99.44 ( 0.00 97.99 ( 0.14 99.49 ( 0.47 97.41 ( 0.54 99.40 ( 0.18
93.70 ( 0.20 93.86 ( 0.11 92.16 ( 0.15 92.04 ( 0.00 91.46 ( 0.00 92.06 ( 0.15 91.74 ( 0.16 93.09 ( 0.40
4.1999 6.2565 9.8178 19.0218 6.0906 11.4765 5.8962 7.8912
Maximum product concentration.
d
Product concentration at equilibrium Ads,
Figure 3. Breakthrough curve of water adsorption by cassava starch.
derived from the packing density. If the feed temperature is greater than the bed, then the feed vapor will condense on the adsorbent, causing its wetting and a bad performance in the adsorption operation. The operating time was kept long enough for the concentration of the product to approach that found at the inlet, to make the adsorbents saturated. Equilibrium was reached at this condition (see columns 4 and 6 from Table 3). Water adsorption capacity was calculated from a mass balance at maximum product concentration and results are shown in the last column from Table 3. In general, water adsorption capacity was increased by enzymatic treatment, suggesting that enzymatic hydrolysis acted on evaluated materials despite the fact that the surface area was not increased. More hydroxyl groups were exposed by enzyme-treatment, allowing more water adsorption for all tested materials. As a typical example, the breakthrough curves for native and enzyme-treated cassava starch are shown in Figures 3 and 4, respectively. All tested materials presented similar behavior to that shown in these figures. Main differences are in the maximum product concentration and the time required for saturating the adsorbent. Temperature profiles for adsorption with enzyme-treated cassava starch are shown in Figure 5. The highest water adsorption occurs just after 10 min of adsorption operation (see Figures 3 and 4) because at the start some water (presented in the column or other sections of the adsorption
Ind. Eng. Chem. Res., Vol. 48, No. 14, 2009
Figure 4. Breakthrough curve of water adsorption by enzyme-treated cassava starch.
Figure 5. Temperature profiles for adsorption with enzyme-treated cassava starch.
systems like joints) is first evaporated with some of the feed ethanol and the product consists of a mixture of both components. Maximum ethanol concentration was kept near 50 min throughout, until the adsorbent saturation was reached, and each material presented a different time for reaching the saturations because of the different adsorption capacities. Similar results were reported by Chang et al.,24 who found that the cornmeal bed adsorbent got saturated at 50 min of operation. However, the temperature decreases at first because the water-ethanol feed vapors come with a lower temperature provided by the column and they take more energy while crossing the adsorber (see Figure 5). It is known that chemical affinity to water is much greater than to ethanol, but some alcohol can be adsorbed, mainly due to its great affinity to the layers of adsorbed water. However, in this work, it was assumed that only water was adsorbed, for simplicity of the adsorption capacity calculation. Moreover, most of the literature shows that the ethanol is adsorbed to a lesser extent than water, as shown by Westgate and Ladisch3 and Vareli et al.33 Enzyme-treated materials give better results than nontreated materials reaching the anhydrous composition. However, this treatment involves an extra cost due to the use of enzymes and additional equipment for materials hydrolysis. This fact suggests that additional research is required for improving adsorption capacity of this type of materials. 4. Conclusions Tested materials showed affinity with water for both native and enzyme-treated cases. This allowed water to adsorb from the feed for obtaining anhydrous ethanol. Their hygroscopic nature allows trap water into the material by forming hydrogen
6787
bounds with hydroxyl groups. Enzymatic hydrolysis increased water adsorption capacity for all materials, however, it was not due to increasing superficial area or creating pores over the material surface, as occurred for starches hydrolyzed with glucoamylase.34,35 It was suggested that the water adsorption capacity was increased because of the more exposed hydroxyl groups from the polymeric structures evaluated (starch and cellulose). Partial hydrolysis broke polymer chains but keeping the granules and fiber structures, except for upright elephant ear starch, and conserving the size and shape. The size of the upright elephant ear starch granules was decreased by enzymatic attack, meaning that they were more exposed to enzymes, because of their shape and structure, than other starches with the same hydrolysis conditions. Enzymatic hydrolysis promoted crystallinity by attacking first the amorphous zones and keeping the structure and size. Adsorption operation was influenced by packing density because of the high drop pressure that can be generated at high packing densities or the poor vapor contact with the adsorbent at low packing densities. Biomass adsorbents were presented as a promising alternative for water removal from a different process than air dehumidification and alcohol dehydrations. Moreover, their high water adsorption capacity, low cost, and possibility for use later as fermentation material, without being regenerated, lead to further reductions of energy consumption and operating costs. Other adsorption materials like zeolites have more efficiency than the materials evaluated in this research. Moreover, if zeolites are used in a pressure swing adsorption system, then they will endure long life. However, for small scale ethanol production plants (i.e., developing countries that are implementing fuel programs where the rural areas are the objective and small productions are expected) more economic adsorption materials are required, and the starchy and lignocellulosic materials presented in these regions are seen as a potential option for these applications. Acknowledgment ´ scar H. Giraldo O., Nano-Structured The authors thank Dr. O and Functional Materials Laboratory (National University of Colombia at Manizales) for technical assistance in XRD and BET analyses, Dr. Medardo Pe´rez, Advanced Microscopy Laboratory (National University of Colombia at Medellin) for technical assistance in SEM analyses. Supporting Information Available: This information is available free of charge via the Internet at http://pubs.acs.org. Literature Cited (1) Ladisch, M. R.; Dyck, K. K. Dehydration of Ethanol: New Approach Gives Positive Energy Balance. Science. 1979, 205, 898. (2) Hong, J.; Voloch, M.; Ladisch, M. R.; Tsao, G. T. Adsorption of Ethanol-Water Mixtures by Biomass Materials. Biotechnol. Bioeng. 1982, 24, 725. (3) Westgate, P. J.; Ladisch, M. R. Sorption of Organics and Water on Starch. Ind. Eng. Chem. Res. 1993, 32, 1676. (4) Ladisch, M. R.; Voloch, M.; Hong, J.; Bienkowskl, P.; Tsao, G. T. Cornmeal Adsorber for Dehydrating Ethanol Vapors. Ind. Eng. Chem. Process Des. DeV. 1984, 23, 437. (5) Neuman, R.; Voloch, M.; Bienkowski, P.; Ladisch, M. R. Water Sorption Properties of a Polysaccharide Sorbent. I&EC Fundam. 1986, 25, 422. (6) Lee, J. Y.; Westgate, P. J.; Ladisch, M. R. Water and Ethanol Sorption Phenomena on Starch. AIChE J. 1991, 37, 1187.
6788
Ind. Eng. Chem. Res., Vol. 48, No. 14, 2009
(7) Westgate, P.; Lee, J. Y.; Ladisch, M. R. Modeling of Equilibrium Sorption of Water Vapor on Starch Materials. Trans. ASAE 1992, 35 213. (8) Anderson, L. E.; Gulati, M.; Westgate, P. J.; Kvam, E. P.; Bowman, K.; Ladisch, M. R. Drying of Gases Using a New Polysaccharide Based Adsorbent. Ind. Eng. Chem. Res. 1996, 35, 1180. (9) Carmo, M. J.; Gubulin, J. C. Kinetic and Thermodynamic Study on Adsorption by Starchy Materials in the Ethanol-Water System. Braz. J. Chem. Eng. 1997, 14, 3. (10) Vareli, G.; Demertzis, P. G.; Akrida-Demertzi, K. Effect of Regeneration Thermal Treatment of Cellulosic and Starchy Materials on their Capacity to Separate Water and Ethanol. J. Cereal Sci. 2000, 31, 147. (11) Beery, K. E.; Ladisch, M. R. Adsorption of Water from LiquidPhase Ethanol-Water Mixtures at Room Temperature Using Starch-Based Adsorbents. Ind. Eng. Chem. Res. 2001, 40, 2112. (12) Vareli, G. D.; Demertzis, P. G.; Akrida-Demertzi, K. Effect of Adsorbent Particle Size and Temperature on Water-Ethanol Separation by Starchy and Cellulosic Substrates. Z. Lebensm-Unters Forsch A. 1998, 207, 122. (13) Ruthven, D. M. Principles of Adsorption and Adsorption Processes; Wiley: New York, 1984. (14) Kulik, A. S.; Chris de Costa, J. R.; Haverkamp, J. Water Organization and Molecular Mobility in Maize Starch Investigated by TwoDimensional Solid-State NMR. J. Agric. Food Chem. 1994, 42, 2803. (15) Rebar, V.; Fischbach, E. R.; Apostolopoulos, D.; Kokini, J. L. Thermodynamics of Water and Ethanol Adsorption on Four Starches as Model Biomass Separation Systems. Biotechnol. Bioeng. 1984, 26, 513. (16) Webb, P. A.; Orr, C. Analytical Methods in Fine Particle Technology; Micromeritics Instrument Corporation: Norcross, U.S.A., 1997. (17) Hassaballah, A. A.; Hills, J. H. Drying of Ethanol Vapors by Adsorption on Corn Meal. Biotechnol. Bioeng. 1990, 35, 598. (18) Alarco´n, M. F.; Dufour, D. CassaVa Starch in Colombia: Production and Recommendations (Almido´n agrio de yuca en Colombia: Produccio´n y recomendaciones, In Spanish); Centro Internacional de Agricultura Tropical: Cali, Colombia; Centre de Coope´ration Internationale en Recherche Agronomique pour le De´Veloppement; De´partement d’Ame´lioration des Me´thodes pour l’Innovation Scientifique: Montpellier, Francia, 1998. (19) Haros, M.; Sua´rez, C. Effect of Drying, Initial Moisture and Variety in Corn Wet Milling. J. Food Eng. 1997, 34, 473. (20) Beery, K.; Gulati, M.; Kvam, E.; Ladisch, M. Effect of Enzyme Modification of Corn Grits on Their Properties as an Adsorbent in a Skarstrom Pressure Swing Cycle Dryer. Adsorption. 1998, 4, 321. (21) Nelson, N. A Photometric Adaptation of the Somogyi Method for the Determination of Glucose. J. Biol. Chem. 1944, 153, 375.
(22) Robertson, G. H.; Doyle, L. R.; Pavlath, A. E. Intensive Use of Biomass Feedstock in Ethanol Conversion: The Alcohol-Water Vapor Phase Separation. Biotechnol. Bioeng. 1983, 25, 3133. (23) Crawshaw, J. R.; Hills, J. H. Sorption of Ethanol and Water by Starchy Materials. Ind. Eng. Chem. Res. 1990, 29, 307. (24) Chang, H.; Yuan, X. G.; Tian, H.; Zeng, A. W. Experimental Study on the Adsorption of Water and Ethanol by Cornmeal for Ethanol Dehydration. Ind. Eng. Chem. Res. 2006, 45, 3916. (25) Jiping, P.; Shujun, W.; Jinglin, Y.; Hongyan, L.; Jiugao, Y.; Wenyuan, G. A Comparative Studies on Morphological and Crystalline Properties of B-type and C-type Starches by Acid Hydrolysis. Food Chem. 2007, 105, 989. (26) Cheetham, N. W. H.; Tao, L. P. Variation in Crystalline Type with Amylose Content in Maize Starch Granules: An X-ray Powder Diffraction Study. Carbohydr. Polym. 1998, 36, 277. (27) Elsenhaber, F.; Schulz, W. Monte Carlo Simulation of the Hydration Shell of Double-Helical Amylase: A Left-Handed Antiparallel Double Helix Fits Best Into Liquid Water Structure. Biopolymers. 1992, 32, 1643. (28) Zobel, H. F. Starch Crystal Transmissions and Their Industrial Importance. Starch 1988, 40, 1. (29) Zhao, H.; Kwak, J. H.; Zhang, Z. C.; Heather, M.; Brown, H. M.; Arey, B. W.; Holladay, J. E. Studying Cellulose Fiber Structure by SEM, XRD, NMR and Acid Hydrolysis. Carbohydr. Polym. 2007, 68, 235. (30) Rodrigues, G.; de Assunc¸a˜o, R.; Vieira, J. G.; Meireles, C. S.; Cerqueira, D. A.; Barud, H. S.; Ribeiro, S.; Messaddeq, Y. Characterization of Methylcellulose Produced from Sugar Cane Bagasse Cellulose: Crystallinity and Thermal Properties. Polym. Degrad. Stab. 2007, 92, 205. (31) Al-Muhtaseb, A. H.; McMinn, W. A. M.; Magee, T. R. A. Water Sorption Isotherms of Starch Powders Part 1: Mathematical Description of Experimental Data. J. Food Eng. 2004, 61, 297. (32) Beery, K.; Gulati, M. Effect of Enzyme Modification of Corn Grits on their Properties as an Adsorbent in a Skarstrom Pressure Swing Cycle Dryer. Adsorption 1998, 4, 321. (33) Vareli, G. D.; Demertzis, P. G.; Akrida-Demertzi, K. Water and Ethanol Adsorption on Starchy and Cellulosic Substrates as Biomass Separation Systems. Z. Lebensm-Unters Forsch A. 1998, 205, 204. (34) Aggarwal, P.; Dollimore, D. A Thermal Analysis Investigation of Partially Hydrolyzed Starch. Thermochim. Acta 1998, 319, 17. (35) Aggarwal, P.; Dollimore, D. Degradation of Starchy Food Material by Thermal Analysis. Thermochim. Acta 2000, 357-358, 57.
ReceiVed for reView October 18, 2008 ReVised manuscript receiVed April 16, 2009 Accepted June 3, 2009 IE8015736
