LIQUEFACTION BEHAVIOUR OF TWELVE TROPICAL HARDWOOD SPECIES IN PHENOL

Liquefaction of ligno-cellulosic biomass is one of the chemical conversion technologies for developing new materials, adhesives and energy systems. The liquefaction process also provides an opportunity to utilize ligno–cellulosic wastes such as saw-dust, woody wastes, branches and twigs, agro-residues, etc. for the development of value added products. This paper presents the liquefaction behaviour of wood meal of twelve tropical hardwood species in phenol as liquefying media to produce chemically active liquid which has potential to be used as a raw material for developing different products. The liquefaction was carried out at 140°C temperatures for 120 minutes at different liquid ratios. In all the cases, a viscous and sticky black liquid was obtained after the stipulated reaction time. The liquefaction efficiency was found to vary with species and liquid ratio. The highest liquefaction efficiency of 93% was achieved in balsa wood at 1:3 (wood: phenol) liquid ratio. The liquefied wood and residues were characterized by FTIR spectroscopy. The liquefied wood was found to be highly acidic in nature. Viscosity of liquefied wood at higher liquid ratio was independent of wood species. Correlation analysis revealed a strong negative relationship between wood basic density and liquefaction efficiency.


INTRODUCTION
In recent years, the utilisation of lignocellulosic biomass has become appealing because of versatility, renewable nature and environmental concerns.One of the promising approaches of utilizing lignocellulosic wastes is its liquefaction and converting the liquefied wood into value added products (Worthy 1990).In liquefaction process, macromolecule compounds in biomass are degraded of hard-to-access cellulose.Thereafter several studies have been carried out on liquefaction of wood in different liquefaction systems and product development from liquefied wood (Alma et al. 1998, Alma and Acemioglu 2004, Yamazaki et al. 2006, Honglu and Tienju 2006, Lin et al. 2014, Petrič et al. 2015).Lin et al. (2001Lin et al. ( , 2004) ) have studied the mechanism of liquefaction of lignin model compound and cellulose in phenol to understand wood liquefaction mechanism.
Research on wood liquefaction has primarily focussed on effects of catalysts, solvents, temperature, and other factors on liquefaction efficiency.Studies on accelerating liquefaction using microwave and ultrasonication has also been attempted (Pan et al. 2011).In most of these studies, specific wood species has been used as the source of woody biomass.
Industrial woody wastes are generally mixtures of various wood species softwoods, hardwoods, composite wood, other ligno cellulosic material from forest, agricultural residues, weeds, shrubs, lops and tops of plantation species.Liquefaction behaviour of different species and, the physical and chemical properties of liquefied wood have not been studied much.Kurimoto et al. 1999, Kurimoto et al. 2001 reported strong influence of wood species on liquefaction process using Poly Ethylene Glycol as the liquefying agent at 3:1 liquid ratio and 120 minutes of reaction.The softwood species exhibited higher residual content as compared to hardwood species.Within hardwood species also significant differences were observed in liquefaction efficiency Ertas et al. (2014) compared liquefaction efficiency of Eucalyptus and pine wood in PEG-400 /glycerine system and found significant differences in characteristics of liquefied wood from two species.Hardwood species are highly variable in terms of their anatomical structure, density and extractive content which can influence the liquefaction process.There is hardly any study on liquefaction process of tropical hardwood species.Chauhan and Karmarkar (2009) studied the liquefaction efficiency of Hevea brasiliensis wood in phenol with hydrochloric acid as a catalyst and found liquefaction efficiency of 49% to 82% under different reaction conditions.
In this study, liquefaction efficiency of 12 tropical hardwood species and their mixture was determined using phenol as the liquefying agent and HCl as the catalyst.The liquefaction was carried out at different liquid ratio.The liquefied wood was characterized for properties like viscosity and pH.The wood characteristics namely wood density, ash content, extractive content, and lignin content were determined and related to liquefaction efficiency.The liquefied wood and residues were chemically characterized by FTIR spectroscopy.

Wood species and chemicals
Wood species were selected based on various density classes and these were

IV.
Mixture of all wood species Wood material of each of the selected species was collected from local sawmills and chipped into small pieces.The chips were air dried and pulverized using a laboratory scale pulveriser.The pulverized wood meal was sieved through 150 mesh sieve (BSS specification, ~80 µm particle size) and oven dried at 105°C for 24 h.Phenol (LR grade) was used as the liquefying media and hydrochloric acid (LR.Grade) was used as a catalyst.For characterizing the liquefied wood, laboratory grade reagents namely; 1-4 Dioxane, Liquefaction behaviour of twelve ..: Kumar et al. phthalic anhydride, pyridine, KOH and methanol were used.All the reagents and solvents were procured from Nice Chemicals, Kochi, India and used as received.

Liquefaction of wood powder
Oven dry sawdust of various wood species was used for the preparation of liquefied wood.Wood liquefaction was carried out in phenol as the dissolving agent and HCl as the catalyst.The liquid ratio i. e. wood to solvents ratio were 1:1, 1:2, 1:3 and catalyst content was 3 wt% of the wood content.Liquefaction reaction was performed for 120 minutes in a glass reactor equipped with a stirrer and a condenser.The reactor was immersed in an oil bath that was preheated to 140°C.For each liquefaction reaction, 10 g of oven dried meal was taken in the glass reactor.The pre-determined quantity of phenol and HCl was added to the wood meal.The mixture was kept as such without any stirring at 140°C for 30 minutes.Thereafter, stirring was started and continued for 90 minutes.After the liquefaction reaction process, the liquefied wood was diluted thoroughly with excess methanol and filtered through glass filter paper under reduced pressure on G1 filtration apparatus.Methanol from the filtrate was recovered by distillation and the liquefied wood was obtained.The residues were dried in an oven at 105°C until it reached to a constant weight.Liquefaction efficiency (LE, %) was calculated from the initial weight of wood meal and residual weight using Equation 1.
Where m 0 is oven-dry weight of the initial wood (g) and m 1 is oven-dry weight of the residue content (g).

Viscosity of the Liquefied wood
The viscosity of the liquefied wood was measured using a Ford viscosity cup N°.4 (B4 cup).The viscosity cup was fully filled with the liquid under the test, and the flow time through standard orifice was measured.Equation 2 was used to convert the time of flow in seconds, t to kinematic viscosity V4; ( ) Where; V4= kinematic viscosity using orifice N° 4 and t is flow time of liquid through orifice in seconds

FTIR analysis of the Liquefied wood and residues
The changes of components of the liquefied biomass based polyols during the liquefaction were analyzed using Bruker make Fourier transform infrared (FTIR) spectrometer (Tensor 27 model).Spectra were recorded for wood meal, liquefied wood and wood residue, with a co-addition of 64 scans at a spectral resolution of 4 cm -1 .Each spectrum was recorded in a frequency range of 400-4000 cm −1 using potassium bromide (KBr) disc.KBr was previously oven-dried at 105°C to reduce the interference of water.

Wood chemical analysis
Wood meal of all the studied species was analyzed for ash content, extractive content and lignin content.For ash content estimation, 10 g wood sample (oven dried and cooled) was weighed in a ceramic crucible and kept in a muffle furnace at 650°C for 8 hours.Ash content was determined from the initial and final weight of the sample.Extractive content was determined by extracting 10 g wood meal (oven dried and cooled) with a mixture of toluene, acetone and ethanol (4:1:1 wt ratio) using soxhelet extraction until colour less solution in the extraction unit was obtained.The extractive free wood was oven dried and weighed.Extractive content was estimated from the initial weight and final weight after extraction.Lignin estimation was carried out from the extractive free wood as per Tappi method (T222-om 15:1988).Air-dry density of wood was determined from the air-dried mass and volume of wood samples of each species.

RESULTS AND DISCUSSION
Wood characteristics namely air dry density, ash content, extractive content and lignin content of the species are given in Table 1.Wood density among the species was varying from 178 kg/m 3 to 831 kg/m 3 .Ash content was ranging from 0,84% to 5,27% extractive content was ranging from 1,92% to 16,81% and acid-insoluble lignin content was ranging from 18,84% to 31,86%.The liquefaction efficiency and characteristics of liquefied wood namely viscosity and pH at different liquid ratio are given in Table 2. Within a species, liquefaction efficiency increased with the increasing liquid ratio.The role of solvent is to dissolve the liquefied products and prevent them from re-polymerizing.At lower liquid ratio, during the initial stages of the process, the dissolving agent was wetting only a small part of wood meal and the stirring was difficult in the reactor vessel.However as the reaction proceeded with slow and careful stirring for longer time, it resulted in photolysis of lignin which initiated the liquefaction process resulting in satisfactory liquefaction.Liquefaction efficiency of 68,89% to 78,53% was achieved for different species at 1:1 liquid ratio.
At 1:2 liquid ratio, there was no concern with respect to stirring as the quantity of phenol was enough to wet wood surface and the liquefaction reaction proceeded smoothly.The liquefaction efficiency at this liquid ratio was nearly 10% higher than as observed at 1:1 liquid ratio.Increasing phenol content further at 1:3 liquid ratio did not result in any remarkable increase in the liquefaction efficiency.At 1:3 liquid ratio, the liquefaction efficiency was only about 3-4,5% higher than at 1:2 liquid ratio irrespective of species.Overall, liquefaction efficiency of wood varied from 69% to 93% under different reaction conditions and at different liquid ratio.Chauhan and Karmarkar (2009)

observed
Liquefaction behaviour of twelve ..: Kumar et al.
77% liquefaction efficiency of rubber wood in phenol with 1: 3 liquid ratio at 140°C which was lower than the observed liquefaction of the same species in this study.Liquefaction efficiencies for birch wood was reported to be of 90% and 100% in the presence of hydrochloric and sulphuric acid catalyst respectively at 150°C in two hours of reaction (Acemioglu andAlma 2002, Alma andAcemioglu 2004) which is in agreement with this study.Lee and Chen (2008) also reported 82% liquefaction of Japanese cedar in phenol at 1:2.5 liquid ratio with HCl as catalyst with reaction carried out at 130°C.The liquefaction with sulphuric acid catalyst was nearly 95% under the same conditions.Hrastnik et al.
(2013) obtained more than 90% liquefaction efficiency of black poplar wood at liquid ratio of 1:3 and more with ethylene glycol as the liquefying media and they did not find any significant improvement in liquefaction efficiency on increasing the liquid ratio beyond 1:3.Among the species, the highest liquefaction efficiency was achieved in Ochroma pyramidales and the lowest was in Terminalia tomentosa at all the liquid ratios.These species were also found to be having the extreme values in terms of wood density, extractive content and lignin content.It was observed that high density wood exhibited relatively low liquefaction efficiency as compared to low density wood.However, it is to be noted that the liquefaction efficiency at 1:3 liquid ratio was varying within a small range from 83% to 93% though the air dry density was varying from 178 kg/m 3 to 831 kg/m 3 , exhibiting nearly five-fold variation (Table 1).
The small difference in liquefaction efficiency of wood having extreme density differences may be attributed to the differences in anatomical structure and permeability which may influence the penetration of phenol in cell-walls and the mechanism of wood liquefaction.Once the depolymerisation of lignin takes place in the initial stages, the liquefaction process gets accelerated resulting in rapid liquefaction of other components of wood cell walls.This suggests that satisfactory liquefaction of wood can be achieved irrespective of wood species and density.This is further substantiated with the viscosity of the liquefied wood.At low liquid ratio i.e. 1:1 the viscosity of Ochroma pyramidales and Delonix regia liquefied wood was significantly lower than other species and this may be attributed to easy penetration of phenol into the cell wall resulting in rapid liquefaction of these two species.The viscosity at 1: 1 liquid ratio was ranging from 51,44 cp to 132,18 cp in different species.However at 1:3 liquid ratio, the viscosity of liquefied wood from different species was not varying much with its value ranging from 32,34 to 44,86 cp.
The pH value of liquefied wood is an indicator of hydrogen ions activity.pH value varied from 2,14 to 3,87 at 1:1 liquid ratio 1,84 to 2,86 at 1:2 liquid ratio and 1,53-2,16 at liquid ratio 1:3.Within a species, pH values declined with increasing liquid ratio indicating that acidity increases with increase in phenol content.This was expected as the liquefied wood ought to be more acidic with more phenol.

Table 2. Effects of species on liquefaction efficiency at different liquid ratio.
A correlation analysis was performed to understand the relationship between different parameters of the liquefied wood and wood characteristics.Table 3 represents the Pearson's correlation coefficients for 1:1, 1:2 and 1:3 liquid ratios with wood density, ash content, extractive content and lignin content.
Liquefaction efficiency exhibited a strong negative relation with wood density at all liquid ratios.Wood density exhibited significant positive correlation with viscosity, and negative correlation with pH of liquefied wood at 1:1 and 1:2 liquid ratio but these relationships were not apparent at 1:3 liquid ratio.The varying correlation with liquid ratio may be attributed to limited liquefaction at low phenol content during the stipulated reaction time.Ash content of wood was not found to be associated with liquefaction efficiency and other liquefied wood parameters.Extractive content and lignin content of species were negatively correlated to liquefaction efficiency at all liquid ratio.Higher extractive and lignin content may hinder the penetration of phenol into the cell-wall ultra-structure and consequently slow defragmentation of cell-wall constituents.Wood density was positively correlated to extractive content (r= 0,86) and lignin content (r = 0,72) which was on the expected lines.There was no consistent correlation of viscosity and pH of liquefied wood with extractive content.Therefore any relationship observed at a specific reaction condition may not be generalized.No significant correlation was observed between viscosity and pH of liquefied wood at all liquid ratio.
The results suggest that wood species irrespective of having different densities can be liquefied at higher liquefying agent concentration without any significant influence on characteristics of liquefied wood and therefore mixing of wood species may not any significant impact on characteristics of liquefied wood.This is evident from the fact that for mixed hardwood species the liquefaction efficiency of 88% was achieved at 1:3 liquid ratio and other characteristics of liquefied wood were close to the average of species.

FTIR Characterization
FTIR spectra of wood and residuals shown in Figure 1 have been evaluated into two regions namely 3800-2700 cm -1 which is assigned to the O-H and C-H stretching vibrations and the 1800-900 cm -1 fingerprint region assigned to stretching vibrations of wood components.A strong broad band at 3000-3600 cm -1 is related to O-H stretching and a band at 2800-3000 cm -1 is related to asymmetric methyl and methylene stretching groups present in the wood (Poletto 2017).In the fingerprint region, band at 1740 cm -1 is assigned to non-conjugated carbonyl group in hemicellulose and the band at 1510 cm -1 and 1610 cm -1 have their origin in the aromatic ring of lignin.The bands at 1375 cm -1 , 1160 cm -1 and 898 cm -1 are mainly due to carbohydrates, while the band at 1460 cm -1 is attributed to syringyl aromatic skeletal vibration and 1220 cm -1 is assigned to guaiacyl ring breathing with carbonyl stretching (Pandey andVuorinen 2008, Poletto et al. 2012).In case of residues, the carbonyl peak intensity at 1740 cm -1 has almost disappeared indicating the dissolution of hemicellulose component of wood.The characteristic peaks associated with lignin (1510 cm -1 , 1610 cm -1 , 1460 cm -1 and 1220 cm -1 ) have also shown marked reduction in the wood residues particularly obtained with 1:2 and 1:3 liquid ratios.Pan et al. (2007) has also reported disappearance of peak at cm -1 , attributed to the ester carbonyl group in xylan, in the FTIR spectra of the residues from liquefied wood.They observed reduction in the Klason lignin content and increase holocellulose and α-cellulose contents as the phenol to wood ratio (P/W) increased.At 1:1 liquid ratio, peak corresponding to aromatic vibration stretching are visible and this may be attributed to incomplete liquefaction of lignin components and also presence of phenol in the wood residues.
The spectra of wood polyols at different liquid ratio (Figure 2) differ significantly from the spectra of wood residues and wood (Figure 1) and shows that almost all peaks gradually become stronger with increase in liquid ratio i.e. from 1:1 to 1:2 and 1:3.The spectra at 1:1 and 1:2 were nearly overlapping in the fingerprint region.The liquefied wood had peaks associated with both carbohydrate (1740 cm -1 , 1375 cm -1 ) and lignin components (1610 cm - 1 , 1510 cm-1 , 1220 cm -1 ) and therefore is an indicative of liquefaction of all the components of wood.The spectral analysis clearly indicates that almost all the component of wood converted into liquefied wood polyols.

CONCLUSIONS
The study revealed that wood meal from different hardwood species can effectively be liquefied in phenol using HCl as a catalyst with liquefaction efficiency ranging from 69% to 93% under similar reaction.With increasing phenol content, liquefaction efficiency increased but the liquefied wood was more acidic in nature.Extractive content had an adverse impact on liquefaction efficiency.Viscosity and pH value will help in deciding further processing for developing products from liquefied wood.Although wood density was correlated negatively with liquefaction efficiency but the range is relatively narrow.FTIR spectra suggest complete dissolution of lignin and hemicelluloses and the residuals are mainly the crystalline cellulose.It is desirable to have high liquefaction efficiency at a low liquid ratio to make liquefied wood economical and 1:2 liquid ratios appears to be effective in achieving satisfactory liquefaction.The liquefied wood polyols having huge potential as a raw material for development of value added products adhesive, energy, preservative and materials.

Figure 2 .
Figure 2. FTIR spectra of the liquefied wood at different liquid ratio.

Table 1 .
Air-dry density, ash content, extractive content and lignin content in wood.