BIOTECHNOLOGY LETTERS Volume 17 No.8 (August 1995)pp.845-850 Received as revised 14th June
SUPERCRITICAL CARBON DIOXIDE EXPLOSION AS A PRETREATMENT FOR CELLULOSE HYDROLYSIS
Yizhou Zheng*, Ho-Mu Lin, Jingquan Wen, Ningjun Cao, Xuezhi Yu, and George T. Tsao Laboratory of Renewable Resources Engineering 1295 Potter Engineering Center, Purdue University West Lafayette, IN 47906, USA
SUMMARY Cellulosic material Avicel was treated with supercritical carbon dioxide to increase the reactivity of cellulose, thereby to enhance the rate and the extent of cellulose hydrolysis. Upon an explosive release of the carbon dioxide pressure, the disruption of the cellulosic structure increases the accessible surface area of the cellulosic substrate to enzymatic hydrolysis. This explosion pretreatment enhances the rate of the Avicel hydrolysis as well as increases glucose yield by as much as 50%.
INTRODUCTION A key problem in utilization of cellulosic materials for fuel and chemical production is the poor yields of glucose from cellulose by acids or enzymes (Tsao et al., 1978). Many techniques have been used to increase the hydrolysis of cellulose. These techniques can be characterized as either chemical or physical in nature. The power requirements for physical treatments include grinder/milling methods are so large as to make them quite cosily. Chemical treatments with strong acids or bases are generally quite corrosive, and cosily. A further problem is the recovery of these chemical agents, which increase the expense. Another treatment is the steam explosion process (Chahal et al., 1981). This process requires considerable expensive thermal energy in the form of
845
steam. T h e high temperature will cause sugar to decompose and thus reduce the available carbon source for fermentation. Besides the steam explosion, there has been an effort to develop an ammonia explosion (Dale and Moreira, 1982). This allows an explosion at a relatively low temperature to avoid sugar decomposition. However, it is difficult to recover all the feed ammonia as a reusable gas stream, since ammonia is expensive and a less than total recycle is detrimental to process economy. In addition, its corrosive and toxic nature will cause problems in design and operation of the process. As an alternative method, supercritical carbon dioxide explosion with a low temperature (compared with steam explosion) and reduced expense (compared with ammonia explosion) was developed. Carbon dioxide molecules should be comparable in size to those of water and ammonia and should be able to penetrate small pores accessible to water and ammonia molecules. Upon an explosive release of the carbon dioxide pressure, the disruption of the cellulosic structure should increase the accessible surface area of the substrate to enzymatic hydrolysis. The initial experimental results showed how the technique works below. This short article is, however, limited to reporting some interesting observations after a pure cellulose, Avicel, is supplied to an explosion pretreatment. A more comprehensive report will be prepared later after more experimental studies have been completed with substrates such as recycled paper and natural cellulosic materials.
MATERIALS AND METHODS Material Avicel (trade name of a eomnm~ial product of pure cellulose) is a microcrystalline cellulose which was bought from FMC Corporation (Philadelphia, PA). As a pure cellulose, Avicel has a relatively high crystallinity. It has been used to test the pretreatment method of carbon dioxide explosion. The cellulase enzyme used in the present studies was obtained from Iogen Corporation (Ottawa, Canada). The activity of the enzyme is 154 FPU/ml.
Methods Explosion pretreatments of the cellulosic materials by supercritical carbon dioxide were performed in a static type of apparatus as shown in Figure 1. After a sample of Avicel with addition of a certain content of buffer solution (0.05 M potassium dihydrogen phosphate buffer solution, pH 4.7) was placed in the reactor at the beginning of an experiment, the reactor was then enclosed and immersed in the thermostated bath at a constant temperature of interest(35 - 80°C). When the temperature was equilibrated and all tubing connections were secured, carbon dioxide was injected from a Ruska pump into the reactor at the experimental pressure (1000 - 4000psi). Before entering the Ruska pump, carbon dioxide in a coil was placed in an ice bar which is helpful in easily obtaining a higher carbon dioxide pressure. The Avicel sample was agitated by a magnetic stirrer and subjected to the carbon dioxide pressure for a controlled length of time to let the carbon dioxide molecules penetrate the micropores in the cellulosic
846
structure. After the sample was exposed to carbon dioxide for a designated length of time, a quick pressure release was done by opening a valve attached to the reactor. The sample was then immediately taken out and placed in a standardized solution of cellulase enzymes to start the hydrolysis process. The solution contains 0.1% (w/v) of enzymes in buffer. The pH of the hydrolysis solution was 4.7 which was maintained by the buffer solution of 0.05 M potassium dihydrogen phosphate at the temperature of 46 °C. A Siemens powder X-ray diffractometer D500 was used to measure the structure changes before and after the pretreatment of supercritical carbon dioxide explosion. The change also was measured by a GE OMEGA-400 spectrometer showing solid-state NMR spectra. 13C and 1H relaxation data were obtained at 100.6 mHZ via cross polarization. A YSI 2700 select biochemistry analyzer (Yellow Springs Instrument Co., Inc., Yellow Spring, Ohio) was used to determine the concentration of glucose in the solution at different times during the course of hydrolysis. An BioRad HPX-87H IonExclusion column was used to cross check glucose concentration. The mobile phase, was 0.005M H2SO4 at a flow rate of 0.8 ml/min at 60 C. The detector was based on the measurement of the refraction index.
i
C3
T
1~
e
Pressure
2
Coil
.,-
Ice bath
_L
C ylinder
Pump
,
Water
,
bath
Figure 1. The supercritical carbon dioxide explosion apparatus.
RESULTS
AND DISCUSSION
The structure changes before and after the pretreatment of carbon dioxide explosion was determined by the X-ray diffractometer. An obvious change of cellulose structure in Avicel is shown in Figure 2. A peak around 20 = 35 ° is removed due to the explosion. In addition, the crystallinity of the Avicel is decreased by 50% as compared to that without the explosion. Temperature is an important factor in the cellulosic hydrolysis. In the experiments, supercritical (larger than 31.1 °C) and subcritical (less than 31.1 °C) temperatures of carbon dioxide were used to test the carbon dioxide explosion method. The effect of carbon dioxide temperatures on the Avicel hydrolysis is shown in Figure 3. Results indicated that subcritical carbon dioxide is less effective. A major cause for such 847
retardation in subcritical carbon dioxide is likely due to poorer diffusibility of liquid carbon dioxide. In comparison with supercritical temperatures, subcritical carbon dioxide molecules are relatively hard to penetrate the micropores in the cellulosic structures, and then disrupt them upon the carbon dioxide pressure releasing suddenly. Figure 4 shows that the higher pressure of carbon dioxide resulted in the higher glucose yield during the Avicel hydrolysis, which means that a higher pressure is desirable for faster penetration of the carbon dioxide molecules into the solid structures.
a,-a,m-~a,N~ #f |[
i" "s,,~,
(e)
"
(A')
°
•
ii
m
,b
4
:1
m
oJ
ill
i~
m
lb
s
~
•
m
Figure 2. X-ray powder diffraction patterns of Avicel. (A) Without supercritical CO2 explosion; (B) With supercrifical CO2 explosion (1000psi and 1 hour).
80
100i
25
(hours)
"nrm(t~n) Figure 4. The effect of CO2 pressure on Avicel hydrolysis.
Figure 3. The effect of CO2 subcritical and supercritical temperatureson Avicel hydrolysis.
When steam explosion is used as a pretreatment of cellulosic materials, the enhancement of the subsequent enzymatic hydrolysis of cellulose as measured by glucose formation is generally more with natural materials such as wood and agricultural residues than with processed materials such as paper and other cellulosic already subjected to a prior hydrolysis by a dilute acid (Tsao 1987). The explanation of this difference has been
848
that cellulose in native materials is much less accessible to enzymes than the processed materials where prior removal of lignin and hemicellutose by pulping and acid hydrolysis has occurred such that cellulose in these materials has already been made more accessible. A further steam explosion thus will result in less percent enhancement in glucose formation. Explosion of Avicel which is pure cellulose should result in much less enhancement of the subsequent enzymatic hydrolysis unless the agent of the explosion pretreatment can somehow penetrate the cellulose crystalline structures and decrease the crystallinity of cellulose. The current results from X-ray diffraction analysis showing a decrease in cellulose crystallinity by the carbon dioxide explosion pretreatment suggests a more intimate interaction between cellulose and hydrophobic molecules of carbon dioxide. The results suggest that the pretreatment might be able to achieve more than just shattering the higher level of structures involving hemicellulose and lignin, but perhaps can affect the cellulose structures. Generally, water molecules are considered unable to penetrate into the cellulose crystal lattices. An interesting question raised by the current results with pure cellulose is whether or not carbon dioxide molecules can.
Table 1. The effect of COz explosion pretreatment on 1H and 13C relaxation times of Avicel by solid-state NMR.
C & H positions C1H C4H C6H
C2H, C3H, CsH
Avicel, Control ~3C, Relaxation ~H, Relaxation time (sec.) time (sec.) 119 4.8 180 4.7 79 5. l 89,102 4.9, 5.0
Avicel, C02 Pretreated ~3C, Relaxation time (sec.) 109 114 57 79, 82
1H, Relaxation time (sec.) 1.8 1.8 1.9 1.8, 1.8
Solid-state NMR allows the measurement of the mobility by measuring relaxation times of carbon and hydrogen atoms. The solid-state NMR analysis for a pretreated Avicel sample is shown in Table 1. After the carbon dioxide explosion pretreatment, the relaxation rates of all carbon atoms in the Avicel sample are increased by about 1.4 folds, particularly those at 4- and 6-positions, indicating the enhancement of the mobility. While the relaxation rate of hydroxymethyl carbon at 6-position is determined not only by its internal rotation and other motions, but also by the environment where it is
849
located, and the relaxation rate of carbon at 4-position is directly related to the glycosidic linkage along the chain and fibrillar dimensions. In addition, the increase in relaxation rate is about 3-folds for hydrogen atoms. The increase in relaxation rate or the mobiliy of almost all carbon and hydrogen atoms could mean that the hydrogen bonds are broken and the three-dimensional structure of the cellulose is disrupted to some extent during the explosion pretreatment. Thus, the cellulose crystalline structure is no longer so rigid and orderly and carbon dioxide molecules penetrate into the cellulose crystal lattices.
CONCLUSION As a pretreatment for cellulose hydrolysis, the supercritical carbon dioxide explosion method increases the accessiblc surface area of thc cellulosic substratc to cnzymatic hydrolysis. Bccausc of its characteristicsof a "gas-like" mass transfer with a "liquid-like"solvating power of (Larson and King, 1986), supcrcriticalcarbon dioxide is cffcctivc. An increase in pressurc facilitatestbc faster penetration of carbon dioxide molecules into the crystalline structures, thus more glucose is produced from the cellulosic material after the explosion as compared to those without the prctrcatment. The results suggest the possibilityof penetration of carbon dioxide molecules into the cellulose crystal lattices. As an alternative method, this supercritical carbon dioxide explosion can reduce expense compared with ammonia explosion and since it is operated at the low temperature it will not cause degradation of sugars such as those treated with steam explosion due to the high temperature involved.
ACKNOWLEDGMENTS This research was supported by a grant from the US EPA CR822932-01-0. The X-ray diffraction assistance from the Depa,iment of Material Science Engineering, Purdue University is appreciated.
REFERENCES (I)
(2) (3) (4) (5)
Chahal, D. S., Mcguirc, S., Pikor, H., and Noble, G. (1981). Proceedings of the 2nd World Congress of Chemical Engineering, Montreal, Canada, October, 4-9. Dale, B. E. and Morcira, M. J. (1982). Biotechnology and Bioengineering Symp., 12, 31- 43. Larson, K. A., and King, M. L. (1986). Biotechnology Progress, 2(2), 73-82. Tsao, G. T. (1987). In: Anaerobic Digestion of Biomass, Chap. 6, London: Elsevier Publishers. Tsao, G. T., Ladisch, M., Ladisch, C., Hsu, T. A., Dalc, B. E., and Chou, T. (1978). In AnnualReports on Fermentation Processes, Vol. 2, Chap. 1, New York: Academic Press.
850