Mechanism of Depolymerization of Cellulose in Low Sulfuric Acid Medium. Kinetic Investigation and Stochastic Simulation

TABLE OF CONTENTS

FIRST PAGES OF ORIGINAL THESIS

ABSTRACT IN FRENCH

EXPERIMENTAL STUDY I : CELLULOSE PRETREATMENTS
INTRODUCTION TO CELLULOSE
EXPERIMENTAL APPARATUS
COTTON PREATREATMENTS
COTTON CHARACTERIZATION
EXPERIMENTAL RESULTS AND DISCUSSION
Cotton structure and morphology
Effects of cotton extraction on cellulose acid hydrolysis
Rolling up process of cotton wax
Effects of cotton boiling on cellulose depolymerization
Effects of milling on cotton morphology
Effects of milling on cellulose depolymerization
Effects of milling on the rate of glucose formation
CONCLUSION
APPENDIX
NOMENC;LATURE
REFERENCES

EXPERIMENTAL STUDY II: KINETIC INVESTIGATION
INTRODUCTION TO CELLULOSE ACID HYDROLYSIS
EXPERIMENTAL RESULTS AND DISCUSSION
Glucose degradation
Cellobiose hydrolysis
Glucose build up during cellobiose hydrolysis
Cellulose sacharification
Effects of milling on cotton morphology
Effects of milling on cellulose depolymerization
Effects of milling on the yield of glucose
CONCLUSION
NOMENCLATURE
REFERENCES

STOCHASTIC SIMULATION OF THE YIELD OF GLUCOSE
INTRODUCTION TO STOCHASTIC SIMULATION
CELLULOSE ACID HYDROLYSIS PROCESS
STOCHASTIC SIMULATION OF A POLYMER DEPOLYMERIZATION PROCESS
STOCHASTIC SIMULATION OF CELLULSOE ACID HYDROLYSIS
RESULTS AND DISCUSSION
CONCLUSION
NOMENCLATURE
REFERENCES

PREFACE

This is an English rewritten version of the PhD thesis written originally in French. This version is based on the following publications:

1. Acid hydrolysis of glucosidic bonds in polysaccharides: Modelling and Stochastic simulation: Advances in thermochemical biomass conversion, edited by A.V. Bridgewater, vol. 2, pp1583-1597, 1992

2. Acid hydrolysis of cellulose. Part I: Experimental kinetic Analysis: The Canadian Journal of Chemical Engineering, vol. 71, December 1993

3. Effects of cotton pre-treatments on wax layers and cellulose acid hydrolysis: Cellulose Chemistry and Technology, vol. 27, pp 597-625, 1993

4. Acid hydrolysis of cellulose, Part II: stochastic simulation using a Monte Carlo technique: The Canadian Journal of Chemical Engineering, vol. 71, February 1994

Abstract for the thesis:

The final objective of this investigation is to model the kinetic behaviour of cellulose during hydrolysis by means of stochastic simulation. Part I of this study will thus report the experimental determination of kinetic parameters to be used in the simulation. These were established from kinetic experiments on cellobiose hydrolysis and glucose degradation. Furthermore, both cotton morphology and outer layer are analysed and the effects of cotton wax on cellulose depolymerization are studied. Finally, the effects of cotton milling on both cellulose depolymerization and glucose yield are investigated and presented in this first part.

Part II will deal more specifically with the stochastic modelling of these data. This simulation should be realistic enough to allow a representation of the effect of milling on the cellulose structure and its influence on acid hydrolysis kinetics.

Abstract for the kinetic investigation:

The main objective of this first part is to investigate the effects of milling on the rate of cellulose depolymerization during cotton acid hydrolysis. The data indicate that some glycosidic bonds of cellulose have very high accessibility to catalytic ions. It was also shown that a mechanical pretreatment increases the accessibility of some glycosidic bonds of cellulose and decreases the volume of the crystalline regions of cotton. From the glucose yield versus time data, it was found that the effects of milling on the rate of cellulose depolymerization depends on the reactivity and accessibility of the glycon rings of cellulose. Furthermore, conditions that eliminate the influence of cotton wax on the rate of cellulose depolymerization were found by investigations on the effects of cotton wax extraction and cotton boiling on the rate of cellulose depolymerization. It was shown that the shift in the controlling mechanism of cellulose depolymerization, from a mass transfer control to a kinetic control, is located in the melting temperature range of the fatty acids of cotton. It was concluded that, at temperatures higher than the wax melting point, the effects of cotton extraction on the rate of cellulose depolymerization become negligible. The reason for this result is that the wax rolling up process happens in original sample and solvent extracted sample in which wax is still present. This effect was confirmed by showing that cotton boiling affects significantly the rate of hydrolysis at temperatures below the wax melting point, due to the increased rate of cotton wetting associated with the observed process of cotton wax rolling up during boiling.

Abstract for the stochastic simulation:

A Monte Carlo procedure was developed to simulate cellulose acid hydrolysis at high temperatures. Both the kinetic information related to the model compound cellobiose and the morphological aspect of cellulose including crystalline, semi-amorphous and amorphous zones were estimated from experimental data and introduced in a FORTRAN program. In our model of cellulose acid hydrolysis, the cleavage of a glycosidic bond and the degradation of glucose are considered as two irreversible reactions in series. For all the temperatures, the overall glucose disappearance rate constant used in our model, was higher than the experimental constant obtained from the degradation of pure glucose. The changes related to the effects of milling on the cellulose acid hydrolysis were successfully considered in the procedure. Finally, the observed good agreement between the simulated and the experimental data of glucose yield versus time proved that Monte Carlo simulation associated with a Markov chain is a flexible connection between cellobiose (model compound) and cellulose conversion reactions.

EXPERIMENTAL STUDY I: CELLULOSE PRETREATMENTS

INTRODUCTION TO CELLULOSE

Cellulose is synthetized by all higher plants as well as by a wide variety of other organisms. The amount of synthesis is enormous, making cellulose the most abundant biopolymer on earth.¹ Cellulosic materials could then be used as a potential source of organic materials and fuels currently derived from petroleum fractions.²¯³ The objective of many research projects is to obtain the highest yield of glucose during acid hydrolysis of wood and cotton fibers.⁴⁻⁶ The glucose formed could for example be fermented to ethanol or hydrogenated into sorbitol. This last compound may be transformed into low molecular weight polyalcohol or used for the production ⁷ of vitamin C. In botanical science, cotton is defined as seed hair, but in common usage one refers to cotton as a fiber where the microscopy, each fiber appears as an irregularly twisted, collapsed, flattened cell. ⁸ It is also well known that the cotton fiber is covered with a thin layer of tightly moulded materials and, according to Peters, ⁹ waxes, resins and nitrogen-containing compounds are the main impurities detected in this outer layer of cotton cell. Cotton wax is described as a mixture of aliphatic monoalcohols in C₂₈ - C₃₄, fatty acids in C₂₄ - C₃₄, saturated and unsaturated hydrocarbons, resins and resin acids, sterol and sterol glucosides.¹⁰

Because of their low surface energy, the wax materials could represent a resistance to cotton wettability and therefore decrease considerably the rate of cellulose hydrolysis.¹¹ From the fact that wax is not an integral part of the cotton cell, many extraction techniques are used for the purification of fibers.¹¹ For example, for the hydrolysis of Egyptian cotton at 50° in 0.1N sulfuric acid solution, fibers were first purified by extraction with boiling alcohol and ether.¹² However, the ESCA analysis, reported by Ahmed el al.,¹³ indicate that cotton wax was not completely removed even after drastic treatments. On the other hand, Peters¹¹ reported that if cotton is immersed in hot water (97-99°) for 15 minutes and air-dried at 20°, it becomes wettable. This suggests that wax material melts when immersed in hot water, rolls up into droplets in the cotton surface and remains as such if the fiber is then dried at 20°. Moreover, the high temperature coefficient of detergency is partly associated with the melting of some wax materials.¹⁴ Finally, grinding was found to detach some wax from the cellulosic materials.¹⁵

Cellulose is a linear polysaccharide having the anhydro-cellobiose moiety as the repeating unit. This unit itself is formed by two anhydro-glucose linked by a β(1,4) glycosidic bond, ¹⁶ but the process by which the β (1,4) linked glucan chains are synthetized and assembled into the cellulose microfibrils of cotton remains one of the major mysteries of plant biology.¹⁷ Glucose and intermediate oligomers are produced when the catalytic agents disrupt these links. However, the highly ordered structure and, particularly, the crystallinity of cellulose constitutes a major obstacle for the hydrolysis of cellulosic materials.⁶ As a result, during cellulose hydrolysis in dilute acid solution, the production rate of glucose decreases drastically when the rupture of the β(1,4) bonds with very low accessibility is achieved. On the other hand, in concentrated acid solutions, the quantitative saccharification of cellulose is followed by a fast degradation of the glucose formed into furfural and levoglucosan.

Experimental investigations of cellulose hydrolysis demonstrate clearly that the decrease of the average degree of polymerization occurs in at least three major stages.⁹ According to the data presented by Sharples,¹² the first rapid stage of cellulose depolymerization represents the rupture of the bonds with high accessibility. Indeed, the postulated weak bonds, which were thought to be created by inductive effects, caused by the presence of some oxidized groups, were not found during the study of kinetics of cellulose depolymerization. ¹²̛ ²⁰ The following stages of the process have a smaller rate and are related to the breaking of the glycosidic bonds with lower accessibility, located in the noncrystalline regions.¹⁹ The third stage of the process starts very close to the so-called levelling-off degree of polymerization (LOPD) and its extremely low rate is due to the rupture of the β(1,4) bonds located in the crystallities.²¹

According to Harris,²² the accessibility of the glycosidic bonds is directly related to the rotational energy barrier encountered in their glycon rings flexure. Therefore, when the rate of cellulose depolymerization is controlled by the rupture of the glycosidic bonds with low accessibility, the low rotational energy of the glycon rings imposes a higher energy of activation. Indeed, the very slow degradation rate of the crystalline parts of cellulose is a consequence of the hydrogen bonds holding tightly the glycon rings.

To make efficient use of the glucose potential of cellulosic materials, during acid hydrolysis, many physical and chemical pretreatments are reported in the literature. ⁵̛ ²³ It was found that milling enhances the degradation of cellulose by loosening its structure and disrupting the durable hydrogen bonds between cellulose molecules of the crystalline areas within the microfibril or between microfibrils. ⁵̛ ²⁴ The investigation by Millet et al.²⁵ of the effects of ball milling on dilute acid hydrolysis of cotton, shows that both the rate of hydrolysis of cotton and the maximum yield of glucose are increased by the mechanical pretreatment. Furthermore, the work of Saeman⁴ showed that the increase of the rate of hydrolysis of milled cellulosic material is partly related to the smaller particle size of the sample.

The main objective of the present work is to study the effects of cotton milling on the rate of cellulose depolymerization. Therefore, the effects of milling on the morphology of cotton, on the rate of cellulose depolymerization and on the net rate of glucose production are investigated. Furthermore, in order to study the influence of the cotton wax film on the rate of cellulose depolymerization, the effects of cotton extraction and cotton boiling on cotton wax and on the rate of cellulose depolymerization are also investigated.

EXPERIMENTAL APPARATUS

The apparatus used in our experiments was a 725 ml. steel autoclave (AMINCO), where the agitation was kept at 650 rpm by a magnetic stirrer. The heat input from an internal source (500-W cartridge) was regulated by a three mode temperature controller (Lindberg M 211). An external source (1200-W) was added during the warm-up period. During the reaction time, the temperature deviation was less than 0.6°. To avoid any cotton conversion during the warm-up period, a concentrated acid solution was placed in a vertical closed conical reservoir upstream the autoclave. When the desired temperature eas reached, the acid solution was pushed into the reactor by a gas supply. Liquid samples were collected from a micro-valve connected to a capillary tube immersed in the reacting solution. A quantity equivalent to the dead volume of the sampling system was discarded before collecting samples.

COTTON PRETREATMENTS

Milling

The milling pretreatment of the cotton fibers was done in a Thomas-Wiley intermediate mill where cotton samples were introduced through the hopper and swept around by the rotor until cut to sufficient fineness to pass through the sieve top of a delivery 20 mesh (0.85 mm) tube. The shearing action was produced by four cutting edges and two stationary blades.

Boiling

Cotton samples were boiled in de-ionized water (pH=6.7) at 100° for 20 minutes, then dried at room temperature for two days.

Extraction

Cotton samples were purified in a Soxhlet extractor using ethanol, for 8 hours. A second extraction by ethyl-ether was carried out in the same conditions. After this, the samples were washed with distilled water, and air-dried at room temperature.

COTTON CHARACTERIZATION

IR spectra

Each cotton sample (3 mg) was ground with 100 mg of IR grade KBr powder, and pressed into a 13 mm diameter disk. The infrared spectra were recorded using the Harrick diffuse reflectance Fourier transform (DRIFT) cell, in a Digilab FTS 60 spectrometer. The band at 1372 cm⁻¹ (C-H bending) was chosen by Nelson and coworkers²⁶ to monitor the crystallinity of the sample and the absorbance ratio of this band to the one at 2900 cm⁻¹ (C-H stretching) was used to measure the crystallinity index (CI).

Transmission Electron Microscopy (TEM)

Pieces of 1mm³, cut from cotton samples, were fixed with 3% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.2) for 2 hours, at 4°, and post-fixed with 1% osmium tetroxide in the same buffer. Samples were repeatedly dehydrated in ethanol and embedded in Epon 812. Ultrathin sections collected on nickel grids were processed for ultrastructural investigation.

These analyses were performed at the “Département de phytologic, Université Laval” with the assistance of Dr. N. Benhamou.

Scanning Electron Microscopy

A gold-Palladium coating was deposited on all cotton samples prior to examination in a JEOL-840A scanning electron microscope.

Degree of Polymerization by viscosity determination

Viscosities were measured in 0.5M cupriethyldiamine solution (CED) at 20° with a Cannon-Feske viscosimeter, size 100 (ASTM). From these measurements, the intrinsic viscosity was determined using the relation of Martin. Values for the average degree of polymerization were then obtained using the relation of Mark-Houwink (AFNOR NET 12-005). All samples were first washed with de-ionized water and dried at room temperature for two days. Cellulose solutions were prepared by dissolving dry cotton samples in the cupriethyldiamine solution at 20°, for two hours. These experiments were performed at the “Département des sciences du bois, Université Laval” with the assistance of Dr. J. Doucet.

ESCA spectra

C₁₅, O₁₅ spectra and both O/C and N/O ratio at the interface of cotton fibers were measured using an ESCA spectrometer. The equipment utilized was a VG ESCALAB MK 11 spectrometer fitted on the microlab system from Vaccum Generators. It was equipped with a dual Mg-Al anode X-ray source, non-monochromatized. Kinetic energies were measured using a hemispherical electrostatic analyzer with 150 mm radius working in the constant pass energy 20 eV mode. Vacuum was maintained in the range of 10⁻⁸ to 10⁻⁶ Torr.

Glucose HPLC analysis

Solutions taken from the reactor were first neutralized with calcium carbonate (CaCO₂), centrifuged and filtered through a 0.45 micron nylon filter unit (Cole-Parmer). High-performance liquid chromatography (HPLC) was used to analyze all samples. The mobile phase (filtered, degassed HPLC grade water from Fisher Scientific) was delivered through a pump (Perkin-Elmer 3B) at a flow-rate of 0.6 mL min⁻¹. The Aminex HPX-87P Pb form column (Bio-Rad, 300 х 7.8 mm) was equipped with a Carbohydrate Analysis ion-exclusion Micro-Guard pre-column (Bio-Rad). The column temperature was maintained at 85.0 ± 0.2°, by means of an insulated heating system connected to a temperature controller (model D921K; Omega). The injector (model 7125, Rheodyne) was equipped with a 20-µL sample loop. A differential refractometer (L.C25; Perkin-Elmer) was used as detector with a 3392A integrator (Hewlet-Packard) to record the signals. Calibration curves for glucose and cellobiose were constructed periodically using galactose as an internal standard in all samples.

Fatty acids identifications by GC/MS analysis

A HP 5890 gas chromatograph with splitless injector and helium carrier gas at about 1mL/min flow rate was used with HP fused silica column 12m х 0.2 mm coated with 0.25 µm film with crosslinked methyl silicone gum. The GC temperature was programmed from 50° to 290°, at a rate of 10°/min. The end of the column was directly introduced into the ion source of a HP 5970 series mass selective detector. Derivatization of the fatty acids was obtained using Diazald (N-methyl-N-nitroso-p-toluenesulfonamide).

EXPERIMENTAL RESULTS AND DISCUSSION

Cotton structure and morphology

Chinese white cotton fibers were obtained from the “Centre des technologies textiles, Conseil national de la recherché scientifique, Ste-Hyacinthe, Québec, CANADA”. A typical scanning electron photomicrograph shows that cotton fibers grow as twisted single cells (Fig, 1). A cross section of one fiber, obtained using a transmission electron microscope (Fig. 2), shows that cotton cell is a tube, having a width of 20µm, with a central canal or lumen running throughout its length. When the plant matures, the dry protoplasm becomes the lumen and the dried resides (minor amounts of protein and salts), either as solid deposits or as a thin layer on the lumen wall, give the characteristic dark areas observed in the TEM micrograph. Furthermore, a more detailed examination, by transmission electron microscopy (Fig. 3), reveals that the fiber is covered with a dark thin layer (0.2 µm) of tightly molded material, called cuticle.

Abbildung in dieser Leseprobe nicht enthalten

Abbildung in dieser Leseprobe nicht enthalten

Fig. 3 – TEM micrograph of cotton sample: cross section of a portion of a cotton cell.

Moreover, Peters⁹ reported that the major impurities are proteins (14%) and wax material (8%), located in the cuticle and primary wall. On the other hand, 99 wt% (on dry basis) of the secondary wall of cotton is pure cellulose. In agreement with this, the ESCA C₁₅ spectra, shown in Figure 4 and the N/O and O/C ratios, presented in Tables 1 and 2, indicate clearly that the surface of cotton fibers is covered with some non-cellulosic compounds. The high C₁ fraction (93.1%) is consistent with the large number of carbons in long chains monoalcohols and monoacids. Therefore, the small percentages of C₂ (5.7%) and C₃ (1.2%) are entirely different from the composition C₂ (83%), C₃ (17%) of cellulose. This drastic difference is also apparent in the O/C ratio of cotton (0.08), compared to the value of pure cellulose (0.83). Moreover, the GC/MS analysis of the other solution used for the cotton extraction, shows that the cotton wax contains fatty acids from C₁₅ to C₃₃ (Fig. 5).

Abbildung in dieser Leseprobe nicht enthalten

Abbildung in dieser Leseprobe nicht enthalten

TABLE 1: Surface analysis of C₁₅ spectra by ESCA technique of (1) original, (2) boiled, (3) extracted and (4) milled cotton samples

Abbildung in dieser Leseprobe nicht enthalten

TABLE 2: Surface analysis of O₁₅ spectra by ESCA technique of (1) original, (2) boiled, (3) extracted and (4) milled cotton samples

Abbildung in dieser Leseprobe nicht enthalten

Furthermore, inside the cuticle, there come respectively the primary and secondary walls where cellulose is concentrated.²⁷ The high crystallinity index of cotton sample (79%), obtained by infrared spectra (Fig. 6), indicates that cellulose microfibrils, located in the inner walls of the cotton fiber, are highly organized. We may conclude that the accessibility of the glycosidic bonds of cellulose may be affected by both the wax outer film of the cotton fiber and the organization of cellulose microfibrils detected by a high crystallinity index.

Abbildung in dieser Leseprobe nicht enthalten

Effects of cotton extraction on cellulose acid hydrolysis

It is well known that, during cotton hydrolysis, the accessibility of the β (1,4) glycosidic bonds by the reactive ions depends on the wettability of cellulose. The hydrophilic surface character of cotton is confirmed by many authors but, as they are covered with wax materials, cotton fibers behave like a “low-energy surface”. ¹¹̛ ²⁸ Therefore, water forms droplets on the surface of fibers and the rate of cotton wetting is considerable lowered.¹⁰̛ ²⁸ In order to investigate the effects of cotton wax on cellulose depolymerization, the interface of the extracted cotton sample was first analyzed by ESCA technique (Fig. 4) and compared to the original cotton sample in Tables 1 and 2. The increase of the C/O ratio from 0.08 to 0.47 indicates that some of the cotton wax has been removed from the cotton outer layer. However, this value remains small compared to the theoretical value of pure cellulose (C/O=0.83). The relative small C₂ and C₃ values of the extracted cotton sample, compared to the values of pure cellulose (C₂ = 83% and C₃ = 17%) also indicate a poor extraction of the cotton wax. Moreover, in agreement with the ESCA analysis, Figure 3 shows clearly that the catalytic ions should first diffuse through the wax layer before reacting with the glycosidic bonds of cellulose located in the inner walls of cotton. Under these circumstances, the variation in time and space of the acid ions concentration in the cotton wax layer could be described by Fick’s second law of molecular diffusion:

Abbildung in dieser Leseprobe nicht enthalten

Considering diffusion as an activated process, the temperature dependence of the molecular diffusivity could then be estimated by a relationship similar to Arrhenius equation:²⁹

Abbildung in dieser Leseprobe nicht enthalten

Where the activated energy of diffusion ( is defined as a measure of the energy expanded against the cohesive forces of the polymer in forming the gaps through which diffusion will occur. Figure 7 shows that the effects of temperature on the volume of a condensed film of stearic acid [CH₃(CH₂)₁₆COOH] are very small.³⁰ Therefore, we assume that the of a temperature increase on the diffusivity ( of the catalytic ions remain small, below the melting point of the fatty acids. In agreement with this, the experimental data of cellulose depolymerization, during the hydrolysis of both original and extracted cotton samples in 0.16 N sulfuric acid for 15 minutes (Fig. 8), show that the decrease of the average DP of cellulose remains most constant when the temperature is raised from 20° to 60°. These results indicate that the rate of cellulose depolymerization is mass transfer controlled when the temperature is below the wax melting region. On the other hand, Figure 8 shows that, at temperatures higher than 70°, the decrease of the average DP of cellulose is significantly affected by an increase in temperature. The experimental data suggest that a resistance to the mass transfer of the catalytic ions, caused by the fatty acids with higher melting points, is still present, but the rate of cellulose depolymerization seems to be controlled by the chemical process in the high temperature region. From the fact that the transition between these two controlling regimes is located in the melting temperature range (52.3° – 84.15°) of most fatty acids of the cotton sample (C₁₅, C₂₄) (Fig. 9), the shift in controlling regime is assumed to be mainly caused by the melting process of the cotton wax.

Abbildung in dieser Leseprobe nicht enthalten

Rolling up process of cotton wax

The rolling up process of cotton wax consists in the formation of wax droplets upon wax melting. It is a thermodynamically driven process associated with the change in free energy of the cotton-wax-water system upon wax melting.

Abbildung in dieser Leseprobe nicht enthalten

Fig. 9 – Melting point of some cotton fatty acids as a function of carbon number.³⁸

ESCA analysis as performed on cotton samples boiled in water, as described above. From the data in Figure 4, the small value of O/C ratio (0.17) and the small percentages of O₂ (10.7%) and O₃ (3.5%) (Table 1) show that boiling does not remove wax material from the cotton outer layer. However, by comparison with the original cotton sample, the rolling up phenomenon of cotton wax is clearly observed in scanning electron photomicrographs of the boiled sample (Fig. 10). It must be underlines that, before the rolling up process of cotton wax takes place, water molecules should first reach the cotton surface. As shown in Figure 7, the large increase of the wax volume in the crystal melt transition is caused by a high thermal expansion ( αm) of the melted material, compared to the small thermal expansion αc of the crystalline material. The resulting decrease of the wax surface tension could be described by the following equations: 3,1

Abbildung in dieser Leseprobe nicht enthalten

On the other hand, in the crystalline region, the effects of temperature on the wax surface tension are negligible and could be described by:

Abbildung in dieser Leseprobe nicht enthalten

Where ϒ₀ and T₀ are the wax surface tension at T₀ (K) and the wax critical temperature (K), respectively. The data reported by Singleton ³⁰ show that all the saturated fatty acids, when spread on the surface of distilled water at 20°, exhibit essentially similar behavior with respect to the compressibility of their films. This behavior, illustrated by the force-area curve shows that, when the area occupied per molecule is lower than 21 10⁻¹⁶ cm², the compressing force increases practically linearly with decreasing area up to the point at which the film collapses, due to piling up of the molecules of the film. From the fact that 21 10⁻¹⁶ cm² is the area of the cross section of the CH₂ group, determined by other methods, the author concluded that the molecules of fatty acids are oriented. It was concluded that the film attracts water molecules and could be subject to mechanical compression. In agreement with these results, Peters¹¹ reported that the spreading coefficient of water on an oily film increases by decreasing the surface tension of the wax materials. Moreover, during the cotton hydrolysis experiments, the negative charge, arising from the (COOH) groups of the fatty acids, could have caused an unequal distribution of the catalytic ions, between the film and the solution, ¹⁸̛ ³² which could cause the increase of pH in the bulk solution detected during our hydrolysis experiments (Fig. 11).

Abbildung in dieser Leseprobe nicht enthalten

Fig. 10 – Scanning electron photomicrograph of (1) original and (2) boiled in cotton samples

Therefore, we may conclude that the decrease of the film-water interface free energy (ϒm), caused by wax melting, results in attracting both fatty acid chains and water molecules at this interface. Moreover, as shown in Figure 7, the volume occupied by a fatty acid increases drastically in the crystal-melt region. As a result, the smaller remaining solid islands of the expanded film are separated by an increased number of water molecules.³³ The resulting film pressure (π), caused by the difference between the surface tensions of water and the film will tend to disrupt the film.³⁴ As a consequence, water molecules will reach the cotton surface. According to the cotton-wax interface free energy balance (appendix), water molecules could easily displace the wax materials from the cotton fiber surface because the large cotton-film interface free energy (ϒƒ) is replaced by a smaller cotton-water interface free energy (ϒm).¹⁴

Abbildung in dieser Leseprobe nicht enthalten

Abbildung in dieser Leseprobe nicht enthalten

Fig. 11 – pH evolution during cotton hydrolysis in 0.03N sulfuric acid solution.

Effects of cotton boiling on cellulose depolymerization

In order to investigate the effects of cotton boiling on cellulose depolymerization, both boiled and original cotton samples were hydrolysed, during fifteen (15) minutes, in 0.16N sulfuric acid solution at temperatures ranging from 20° to 90°. The experimental data show that the decrease of the average DP of cellulose is lower after the hydrolysis of the boiled cotton sample (Fig. 12). These results indicate that the mass transfer resistance of the wax film was considerably decreased by boiling. On the other hand, the ESCA C₁₅ spectra (Fig. 4) and the small value of the O/C ratio (0.17), shown in Table 1, indicate that, after boiling, almost no wax was detached from the cotton surface.

Abbildung in dieser Leseprobe nicht enthalten

Fig. 12 – Effects of cotton boiling on cellulose depolymerization as a function of temperature in 0.16N sulfuric acid solution.

Moreover, the effects of boiling are still important at the high temperature region and the observed shift from the mass transfer control to chemical control, during the hydrolysis of the boiled sample, may be explained by the fact that not all the wax materials were affected by the rolling up process during the boiling pretreatment. Therefore, the rolling up process of the fatty acids continues during the hydrolysis of the boiled cotton sample, when the temperature becomes higher than the wax melting point.

Furthermore, the effects of wax extraction on the rate of cellulose deloymerization, during hydrolysis in 0.16N sulfuric acid solution at 100, were investigated (Fig. 13). One of the samples was originally extracted following the procedure described in the apparatus and methods section and the two samples were preliminary boiled at 100°, for one hour, before hydrolysis. The time evolution of the average DP of both samples shows clearly that the degradation process is characterized by an initial rapid stage related to the hydrolysis of the completely amorphous areas and a final stage of persistent DP corresponding to the hydrolysis of the crystalline zones.³⁵

Moreover, Figure 13 shows also that the decrease of the average DP of cellulose was not affected by the extraction pretreatment. These results suggest that an increase of the rate of cotton wetting created by a rolling up process of the wax, during the one hour warm up period, eliminated the effects of the decrease of wax film thickness caused by cotton extraction. Finally, Figure 13 shows that, after 15 minutes, the depolymerization rate of cellulose is still controlled by the fast rupture of the glycosidic bonds located in the amorphous regions. As a consequence, the shift in the controlling mechanism of cellulose depolymerization could be discernable in Figures 8 and 12. On the other hand, this shift will not be observed in the experiments of hydrolysis of the most crystalline regions of cotton.

Abbildung in dieser Leseprobe nicht enthalten

Fig. 13 – Effects of cotton extraction on cellulose depolymerization at 100° in 0.16N sulfuric acid solution (both samples were preliminarily boiled at 100° for one hour).

Effects of milling on cotton morphology

The direct effects of milling on cotton sample show that pretreatment reduced the sample particle size from large pieces (3 cm) to a powder (< 0.85 mm), but that the average degree of polymerization of the cotton sample (DP=1868) was not affected. Moreover, the TEM micrographs show that both the outer later and inside walls of the milled cotton microfibril are less dense and more disordered (Fig. 14). The changes in the ESCA C₁₅ of the milled sample, presented in Figure 4 and Table 1, are related to the structural changes of the cotton outer layer caused by milling. Therefore, as shown in the TEM micrographs, the increase of the C₂ component is associated to the unravelling of the cellulose surface. ¹³ In agreement with Figure 14, the low value of the O/C ratio (0.19) of the milled sample, compared to the value of cellulose (0.83), indicates that the wax material is not very much broken off by milling. Furthermore, the ESCA O₁₅ spectra of both original and milled cotton samples are also presented in Table 2 and Figure 4. The large changes observed in the O₀ and O₂ value suggest that these peaks are respectively related to wax material and cellulose. Moreover, the fact that the N/O ratio was not affected by milling, indicates that the nitrogen compounds are equally distributed in the primary wall of the cotton cell. Finally, as already shown in our TEM micrographs and according to the IP spectra (Fig. 6), milling had also decreased the crystallinity index (CI) of the cotton fibers from 70% to 64%.

We may conclude that, compared to cotton extraction, less wax materials are detached from the cotton fiber. However, the TEM micrographs and the IR spectra indicate that milling affects the wax layer morphology as well as cellulose crystallinity.

Abbildung in dieser Leseprobe nicht enthalten

Fig. 11 – TEM micrographs of original (1) and milled (2) cotton samples

Effects of milling on cellulose depolymerization

Changes in the average degree of polymerization of cellulose, during the hydrolysis of both original and milled cotton samples, were used to investigate the influence of milling on the acid hydrolysis of cotton. The two samples were first boiled for one hour, and the experiments, conducted at 100° in 0.16N sulfuric acid solution, are reported in Figure 15

Abbildung in dieser Leseprobe nicht enthalten

Fig. 15 – Effects of cotton milling on cellulose depolymerization during cotton hydrolysis at 100° in 0.16N sulfuric acid solution.

Unlike cotton extraction, these data show clearly that, for the same reaction conditions, cotton milling affects considerably the rate of cellulose depolymerization. Furthermore, from the fact that these experiments were conducted at temperatures higher that the cotton wax melting point, we assume that the influence of cotton wax is negligible and the effects of milling on cellulose depolymerization are principally caused by the changes in cotton morphology observed before.

TABLE 3 : Effects of milling on initial and final degradation rates of cotton fibers (from data in Fig. 15)

Abbildung in dieser Leseprobe nicht enthalten

*these slopes are obtained by averaging over four experimental points

These experimental data indicate clearly that the initial stages of cotton hydrolysis are enhanced by milling. Furthermore, in accordance with the IR spectra, the lower LODP value of the milled samples demonstrates that the volume of the crystalline regions of cotton is also decreased by milling. On the other hand, from the fact that the ratio of the two cellulose depolymerization rates is very close to ours (1.07) in the last degradation stages (Table 3), it may be concluded that the rate of degradation of the more resistant crystalline parts of the cotton sample is not increased by the mechanical pretreatment. Therefore, as in these experimental conditions, the rate of rupture of the β(1,4) bonds located in the amorphous regions is increased by milling, some bonds, originally located in the crystalline regions must have been disrupted, which is in line with the decreased cotton crystallinity.

Effects of milling on the rate of glucose formation

To eliminate any resistance related to the transport of catalytic ions through the wax film, the experiments related to the effects of milling on the yield of glucose were conducted at temperatures higher than 100°. Moreover, in order to investigate the effects of milling on the rate of rupture of the glycosidic bonds with different accessibilities, these experiments were performed in a sulfuric acid solution with low acidity (0.03), compared to the experiments on cellulose depolymerization (0.16N). Furthermore, from the fact that the probability of rupture of the glycosidic bonds depends on both the reactivity and accessibility of the glycon rings. The effects of milling were investigated at different temperatures (120°C to 160°C). The yield of glucose, reported in Figures 16-18, shows that the rise in temperature increases both the initial rate of formation and the maximum yield of glucose during hydrolysis of the original sample. On the other hand, the experiments for the milled cotton sample show that rising temperature from 140° to 160° did not increase the maximum yield of glucose. Because of the higher rate of cellulose depolymerization, this maximum value appears after a shorted time at 160°C.

The slow increase of glucose concentration observed at 120° (Fig. 16) demonstrates that the rate of depolymerization of the cellulose chain molecules was low. The fact that the two curves are identical indicates that, during the reaction time, the rate of cellulose depolymerization was independent on the sample morphology. On the other hand, the experimental data obtained at 140° (Fig. 17) show that both the initial production rate and the maximum yield of glucose are increased by milling.

Abbildung in dieser Leseprobe nicht enthalten

Fig. 16 - Glucose build up during cotton hydrolysis at 120° C in 0.03N sulfuric acid solution

Abbildung in dieser Leseprobe nicht enthalten

Fig. 17 – Glucose build up during cotton hydrolysis at 140° in 0.03N sulfuric acid solution

These results reflect the fact that the milled cotton sample is characterized by a higher accessibility of the glycosidic bonds located in the amorphous regions and a lower crystallinity. However, the same initial fast increase of glucose concentration observed at 160° (Fig. 18) during the hydrolysis of both the original and milled samples, shows that the initial rate of cellulose depolymerization was not affected by the changes in the morphology of cotton sample. The slightly higher maximum glucose yield could be related to the rupture of glycosidic bonds, originally located in the crystalline regions.

Abbildung in dieser Leseprobe nicht enthalten

Fig. 18 – Glucose build up during cotton hydrolysis at 160° in 0.03N sulfuric acid solution

These experimental observations demonstrate that the effects of milling on the rupture of the glycosidic bonds depend on the temperature of reaction. At 120°, the reactivity of the glycosidic bonds might be low and the acid ions could then disrupt only the glycon rings in the amorphous regions which have no increased rotational energy barrier, due to interfibril hydrogen bonds and are therefore not affected by milling. As a result, the mechanical pretreatment did not affect the probability of rupture of the bonds with high accessibility. At 140°, we observe that the rise in temperature increased the probability of rupture of the glycosidic bonds of both original and milled samples. The relatively large difference in the maximum yield of glucose shows the effects of milling were significant. These results could be related to the fact that the depolymerization rate of cellulose was controlled by the accessibility of the β(1,4) bonds and the rupture of the links having some rotational energy barrier was an important step of the process. However, at 160°, we observe that milling did not affect the initial rate of glycosidic bonds rupture. This could be explained by the fact that the rupture of these bonds was increased not only by a higher reactivity of their glycon rings but also by the increase of their accessibility caused by a faster rupture of bonds with a lower rotational energy barrier. The former effect was certainly more important that the unravelling caused by milling. This behavior was also observed ²⁵ during cotton hydrolysis in 0.1N sulfuric acid solution and 180°.

CONCLUSION

In order to eliminate the influence of the transport of the catalytic ions through the wax film on the rate of cellulose depolymerization, a study was undertaken to determine the effects of cotton extraction and boiling on the fatty acids located in the outer layer of the cotton fiber. The data show clearly that the rate of cellulose depolymerization is controlled by two different mechanisms steps, depending on the temperature of the reaction. From the fact that the shift between these two controlling regimes is located in the melting temperature range of the fatty acids of cotton, this phenomena was related to a rolling up process of the fatty acid. Therefore, wettability of the cellulosic materials plays an important role in acid hydrolysis.

It was shown that cotton extraction has no observable effects on the rate of cellulose depolymerization at temperatures higher than the melting point of the fatty acids located in the cotton outer layer. On the other hand, diffusion of the acid ions through the cotton wax layer seemed to control the cellulose depolymerization rate when the temperature is lower than the melting point of all fatty acids. Our investigation demonstrated that cotton boiling is more effective than cotton extraction in the low temperature region. Furthermore, the use of strong chemicals (ethanol, ether,...) is expensive and could also give rise to problems associated with toxic waste when the cotton materials are washed after extraction.

The experimental results showing a decrease in the average degree of polymerization of cellulose allows us to conclude that milling increases the accessibility of the glycosidic bonds located in the amorphous regions and decreases the crystalline of the cotton sample. On the other hand, data on the yield of glucose, during the hydrolysis of cellulose, show that the effects of milling on cellulose saccharification will depend on both the accessibility and the reactivity of the glycosidic bonds. At high temperatures, only the maximum yield is affected by the pretreatment, which could be more related to the decrease of the sample crystallinity. Therefore, a more severe milling is needed to increase the maximum glucose concentration.²⁵ On the other hand, at lower temperatures, the production of glucose is mostly affected by the loosening of the cotton sample structure. The maximum effect, observed at 140°, could be related to both the loosening of the structure of the same and to a smaller volume of the crystalline regions. Moreover, the data observed at 120° are in agreement with the fact that cellulose contains glycosidic bonds with very high accessibility.¹²

[...]