Acid and Enzymatic Hydrolysis of the Residue from Castor Bean ( Ricinus communis L . ) Oil Extraction for Ethanol Production : Detoxification and Biodiesel Process Integration

O processo de hidrólise do amido da torta de mamona por ácido e pela combinação de a-amilase, glicoamilase e pululanase foi realizado, seguindo de etapa de fermentação. A hidrólise química resultou em 27.3 g L de açúcares com 33.4% de eficiência de hidrólise. O hidrolisado obtido foi fermentado originando 11 g L de etanol (Y P/S =0.48 g g). A melhor condição de hidrólise enzimática foi (por grama de torta de mamona): 200 μL de a-amilase, a 90 oC; 200 μL de glicoamilase e 100 μL de pullulanase, ambas a 60 °C, que resultou em 75 g L de açúcares redutores totais, que corresponde a 91.4% de eficiência de hidrólise. Esses açúcares em seguida foram convertidos em 34.5 g L etanol. O processo de hidrólise ácida foi capaz de promover a destoxificação da torta de mamona.


Introduction
Castor Bean (Ricinus communis L.), also known as Castor oil Bean and wonder tree, is a member of Spurge family (Euforbiaceae) which is originated from tropical countries.The Castor Bean contains 40 % of oil e 1-5 % of a cytotoxic protein, named ricin, one of the most potent and deadly plant toxins known. 1 Castor Bean cultivar was always considered as a subsistence agricultural activity experienced by few groups of growers, mainly on the semi-arid region in Northeast of Brazil.On the other hand, efforts of the Brazilian government toward to develop this region, mainly by stimulating Castor Bean cultivation as potential source of development for biodiesel production, have changed this scenery.There are almost 4,000,000 ha in this region appropriated to be cultivated, which are capable to produce 1.5 tons of seed per ha. 2 The oil extraction process of Castor Bean seeds for producing biodiesel and lubricants generates the Castor Bean Cake (CBC) as waste.This waste is highly-concentrated in protein and is produced at the ratio of 1.2 tons per ton of extracted oil, corresponding roughly to 55% of weight of seeds. 3On the other hand, the presence of anti-nutritional factors (ricine and ricinine) does not allow the utilization in natura of this waste as animal food. 4,5dditionally, since it contains high concentrations of starch, CBC is a potential feedstock for the production of bioethanol. 3,6rojections carried out by Freitas & Fredo indicate that from 2006 onwards 20,000 tons of CBC will be available, due to the increasing biodiesel production. 2 According to recent studies on transesterification of Castor Bean Oil to Biodiesel production, if 160 L of ethanol can be generated from 1 ton of CBC (produced by pressed seeds), the ethanol demand for the transesterification process of extracted oil into ethylic ester (Biodiesel) would be fullfiled and the process self-sufficient regarding the transesterification agent. 7Table 1 presents the main differences between the methylic and ethylic routes.Using different combinations of physical variables, as well as stoichiometric amounts of catalyst, alcohol and oil, it is possible to attain similar characteristics to the biodiesel produced by methylic route, since the comparative values are not quite different.The possibility of the ethanol use in the biodiesel production process has a great potential in the Brazilian industrial context, because of the high produced volumes of ethanol.The cost of each process depends on its variables and operational conditions.Some environmental problems, like the greenhouse gas emissions and their consequences, show the importance of the biofuels on global warming attenuation. 8,91][12] Current policies stimulate the development of these technologies.European countries are beginning to use the bioethanol/diesel and bioethanol/gasoline blends, as started by the directive 2003/2030 years. 13onnecting the need of treating the residue generated from Castor Bean Oil extraction to the potential of its Cake as a fermentable source of sugars, the present study aims at the development of enzymatic and chemical hydrolysis processes to promote the detoxification of the leftover Cake.Additionally, it is objective of investigation the production of ethanol by fermentation of the CBC hydrolysate to be used as a transesterification agent in biodiesel production, and thus establishing a self-sufficient and integrated biodieselbioethanol process, as illustrated in the Figure 1.

Castor Bean Cake
Samples of Castor Bean Cake as a residue of biodiesel production process were harvested from the Biodiesel Pilot Plant of Petroleo Brasileiro SA (Petrobras).The solid residue was powdered and dried by 24 hours at 60 ºC to eliminate moisture.

Analysis of starch content from Castor Bean Cake
Percentage of starch as dry weight was determined according to the enzymatic-colorimetric method for starch analysis of seeds that contains mixtures of amylose and amylopectin. 14,15Analyses were performed in triplicate.

Chemical hydrolysis of the starch from Castor Bean Cake
Chemical hydrolysis was carried out in autoclave (model 103; 4000W Fabbe-Primar™, Brazil) utilizing conical flasks of 250 mL containing 10 g of CBC (dry base) suspended in diluted sulfuric acid.Screening experimental design was created considering solid-liquid ratio 1:6, following the Factorial Fractional Design (FFD) 3 k-1 .Independent variables were: time, concentration of sulfuric acid and temperature.Dependent variable was concentration of sugars (Table 2).The STATISTICA software 5.0 version (Statsoft Inc., Tulsa, UK) was utilized to generate and analyze the FFD.Model adjustment equation was confirmed by the correlation coefficient (R²) and statistical significance conditioned by the F-test.Significance for each parameter was depicted by Pareto diagram, with their correlations and curvatures.

Enzymatic hydrolysis of the starch from Castor Bean Cake
a-Amylase (Termamyl Type L, 144 KNU mL -1 ), glucoamylase (AMG, 300 AGU mL -1 ) and pullulanase (Promozyme, 400 PUN mL -1 ), purchased from Novozymes Lantin American Ltd., were combined to promote CBC saccharification according to a full screening experimental design 3 k (Table 3).Thirty independent runs were generated, three of them as center points.Hydrolytic assays were performed in a Dubnoff-type thermostatized bath, in 250 mL conical flasks containing 10 g of CBC.
Experiments were run with a solid-liquid ratio of 1:6.Due to the specificity of each enzyme and the structure of starch from the seeds, a-amylase (90 ºC) was utilized in the first stage of hydrolysis.Thereafter, the temperature was lowered to 60 ºC, for the action of glucoamylase and pullulanase, which were used combined. 14

Ethanol fermentation
Fermentation media were obtained by centrifugation of the remaining solid residue of CBC, after hydrolytic stage, followed by pH adjustment to 5.0 with calcium oxide of the liquid fraction.Fermentations were performed in an instrumented Bioreactor (Biostat B.Braun Biotech International) with a working volume of 2 L. The system was maintained at 30 °C under stirring of 100 rpm by eight hours.Saccharomyces cerevisiae was utilized as the fermentation agent with an initial concentration of 10 g L -1 .Samples were harvested at 1 hour-interval, for analytical measurements.

Analytical methods
Ethanol was analyzed by HPLC (PUMP model 510, Waters) equipped with ionic exchange column Aminex ® HPX-87P (Bio-Rad) and refraction index detector (model 2487, Waters).Standardized water MilliQ was used as mobile phase in a flow rate of 0.6 mL min -1 .Sample volume   was 5 µL.Oven and detector temperatures were 80 °C and 40 °C, respectively.Total reducing sugars were analyzed by DNS method and glucose was quantified by the enzymatic-colorimetric method GOD-POD (CELM ® ). 16iomass, after centrifugation and re-suspension in distilled water, was determined by absorbance measurement at 570 nm (Spectrumlab22PC spectrophotometer) which was further correlated to dry weight by a standard curve.

Detoxification of the Castor Bean Cake
CBC in natura, as well as the remaining solid residues from acid and enzymatic hydrolysis and the spend medium, resulting from the process of ethanol fermentation of both hydrolyzates, were evaluated for their possible cytotoxic effects.8][19] Results were expressed in terms of IC 50 (inhibitory concentration which inhibits 50% of cell viability), in mg L -1 .

Starch content from Castor Bean Cake
Experiments for starch hydrolysis of Castor Bean Cake (CBC) be it by acid or enzymatic procedures were proceeded by determination of starch concentration (percentage) in CBC, which resulted in 48±0.53 (%) in dry basis.This percentage, considering efficiency of 100% in the hydrolysis stage is equivalent to a potential amount of 53 g of glucose, which would result in 25.3 g of ethanol for each 100 g of Cake.These values were considered as reference for calculating the efficiency.

Acid hydrolysis from Castor Bean Cake
Assays in duplicate of the acid hydrolysis of the CBC through experimental planning 3 3-1 (3 variables and 3 levels) are showed in Table 4.According to the variances analysis, the values of R 2 (0.985), pure error (2.457) and lack of fit (0.265), shows that the model is in accordance with the experimental datas.Analysis of the response-factor (concentration of reducing sugars) gives emphasis to experimental condition number 6, with the lowest acid concentration and the highest temperature, as the most efficient.Under these conditions 30 g L -1 of reducing sugars were obtained, corresponding an to acid hydrolysis efficiency (Ef HA ) of 33.5%.Statistical treatment of the results, by Statistica software, generated predicted values with low residues (Table 4).Pareto diagram analysis (Figure 2) showed that, variables, 1 (time) and 2 (Temperature) presented significant positive linear effects on CBC hydrolysis.On the other hand, variable 3 (H 2 SO 4 concentration), displayed negative effect (-7.64%), suggesting that the acid hydrolysis prevails at lower concentrations of acid.This statistical results does not corroborate to expected effect of acid concentration in the hydrolytic process.The hydrolysis efficiency is reduced when the acid concentration is higher than 0.25 mol L -1 .However, the explanation of this antagonic effect may be due to the possible decomposition of sugars to furfural and hydroxy-methyl-furfural under extreme conditions of temperature and pH, that turn the system more oxidizing. 20This hypothesis also finds support by the negative effect evidenced in the interactions of variables 1 (time) and 2 (Temperature).Recent researches relate that increased yields of chemical hydrolysis of biomass are achieved when conventional hydrolysis accompanies thermal processing.However, ultra-high-pressure has the highest cost. 21

Enzymatic hydrolysis from Castor Bean Cake
The enzymatic hydrolysis of starch from CBC was initiated with the use of a-amylase enzyme (-1,4-glucan-4glucanohydrolase) at 90 °C to liquefy starch.On the sequence, glucoamylase (amyloglucosidase or 1,4-glucanohydrolase) acting on -1,4 bounds from the reducing polysaccharide extremities releasing glucose.Glucoamylase was combined with pullulanase, which acts with high specificity on the glycoside -1,6 bounds in amylopectin, reducing the remaining oligosaccharides, and therefore contributing for the release of fermentable sugars. 22,23 sults of the screening experimental design combining three kinds of amylases had as the response-factor the concentration of reducing sugars.These experiments were carried out utilizing 1:6 solid-liquid ratio in order to improve the enzyme action, resulting in a better mass transfer in this heterogeneous media.Results are shown in Table 5.
Experiment 27 (Table 5), performed with the highest concentrations of enzymes, presented the best response among the evaluated concentration (74.5 g of reducing sugar), leading to a starch hydrolysis efficiency of 91.4 %.This result indicates that the process is efficient, becoming unnecessary the evaluation of others levels of enzymatic concentration.Combined use of pullulanase, a-amylase and glucoamylase (experiment 27) when compared to the use of only a-amylase and glucoamylase (experiment 21) resulted in an increase of 26.6 % on the concentration of reducing sugars.This is surely due to the specific action of pullulanase on the residual -1,6 bound, increasing the rate of free glucose production.Pareto chart (Figure 3) evidenced the statistical significance of linear effects of the three amylases evaluated in the hydrolysis process and the combined effect of a-amylase and glucoamylase action. 14he best conditions for starch saccharification of CBC indicated by the experimental design analysis of the acid hydrolysis (experiment 6, Table 4) and enzymatic  hydrolysis (experiment 27, Table 5) were used to prepare the fermentation media for ethanol production.

Ethanol fermentation of the hydrolysates from Castor Bean Cake
The CBC acid hydrolyzate was inoculated with a cell concentration of 9.5 g (d.w.) L -1 , and the fermentation was operated batchwise (Figure 4).The process lasted 8 h, when the ethanol concentration reached 11 g L -1 , corresponding to a yield of product on substrate consumed of 0.480 g g -1 and a volumetric productivity of 1.38 g L -1 h -1 .Considering the acid hydrolysis efficiency (33.5%) and the response variables of ethanol fermentation, we can estimate a ratio of 108 liters of anhydride ethanol per ton of CBC (dry base), which would supply 83% of the ethanol demand for biodiesel production from Castor Bean oil.
Similarly, the CBC enzymatic hydrolyzate was fermented batchwise for 10 h, and the main process variables (ethanol, sugar and cell concentrations, as well as pH) were monitored throughout the fermentation (Figure 5).After 5 h of fermentation, the ethanol concentration reached its maximum value (34.5 g L -1 ), corresponding to a yield of product on substrate consumed of 0.465 g g -1 and a volumetric productivity of 6.98 g L -1 h -1 .
Then again, considering the enzymatic hydrolysis efficiency of 91.4 %, it is possible to estimate that 270 liters of ethanol could be produced from 1 ton of CBC.A mass balance for the transesterification reaction allows to figure out that the process has the potential to supply a double of the ethanol demanded for biodiesel production from Castor Bean oil.

Analysis from detoxification of the Castor Bean Cake
These results point out for the starch enzymatic  hydrolysis of CBC.Furthermore, waste residues of hydrolytic and fermentative were also analyzed as far as their effectiveness on detoxification (Table 6).The results show the inhibitory concentration of 50% for cell population to different viability markers (mitochondria, lisossomes and nucleic acids) obtained in assays with hepatocites from male Wistar mice.Table 7 present results of cytotoxicity (IC 50 in mg L -1 ) obtained in assays with V79 cells.

Figure 1 .
Figure 1.Summaries of the integration of the present proposal with biodiesel production within the context of residue/waste valorization.

Figure 2 .
Figure 2. Pareto chart of starch acid hydrolysis of the CBC, considering minimal confidence range of 95%.

Figure 4 .
Figure 4. Ethanol fermentation of the acid hydrolyzate of Castor Bean Cake.

Table 2 .
Variables and levels of Statistical Factorial Fractional Design (FFD) to evaluate chemical hydrolysis of Castor Bean Cake

Table 3 .
Full Factorial Design for the action of amylolytic enzymes and their concentrations in the hydrolysis process a KNU is defined as the amount of enzyme capable to promote rupture of 5.26 grams of starch per hour, in the conditions specified by the Novozymes stand method; b AGU is defined as the amount of enzyme that hydrolyses 1 µmol of maltose per minute under standard conditions; c PUN is defined as the amount of enzyme that hydrolyses pullulan, releasing 1 µmol glucose per minute under standard conditions (Novozymes Product Sheet, 2001-00746-03.pdf,2001-12270-02.pdf e 2001-00464-40.pdf).

Table 4 .
Chemical hydrolysis of Castor Bean Cake with a solid-liquid ratio of 1:6, according to the experimental design shown in Table2Acid and Enzymatic Hydrolysis of the Residue from Castor Bean (Ricinus communis L.)

Table 5 .
Action of amylases and their concentrations following full factorial design (Table3) *Center Points.Figure3.Pareto chart amylases effect at the hydrolysis of Castor Bean Cake.

Table 6 .
Cytotoxic effect (IC 50 ) in hepatocites utilizing residues from hydrolysis and fermentation of Castor Bean Cake Solid residue from acid hydrolysis, FA: waste of fermented acid hydrolysate, HE: residue from enzymatic hydrolysis, FE: waste of fermented enzymatic hydrolyzate.IC 50 : Inhibitory concentration which inhibits 50% of cell viability.

Table 7 .
Cytotoxic effect (IC 50 ) in V79 cells utilizing residues from hydrolysis and fermentation of Castor Bean Cake HA: Residue from acid hydrolysis, FA: waste of fermented acid hydrolysate, HE: residue from enzymatic hydrolysis, FE: waste of fermented enzymatic hydrolysate.IC 50 : Inhibitory concentration which inhibits 50% of cell viability.