Kinetic Study of the Catalytic Pyrolysis of Elephant Grass Using Ti-MCM-41

Programa de Pós-Graduação em Ciência e Engenharia de Materiais – PPGCEM, Centro de Tecnologia – CT, Universidade Federal do Rio Grande do Norte – UFRN, Natal, RN, Brazil Departamento de Química, Universidade Federal do Rio Grande do Norte – UFRN, Natal, RN, Brazil Departamento de Química, Universidade Federal de Campina Grande – UFCG, Cuité, PB, Brazil Departamento de Engenharia de Energias Renováveis – DEER, Centro de Energias Alternativas e Renováveis – CEAR, Universidade Federal da Paraíba – UFPB, João Pessoa, PB, Brazil Programa de Pós Graduação em Engenharia Química – PPgEQ, Centro de Tecnologia – CT, Universidade Federal do Rio Grande do Norte – UFRN, Natal, RN, Brasil


Introduction
In recent years, intensive research has been carried out for the development of fuels from biomass, either for environmental (conversion of biomass into energy with lower environmental impacts) or economic reasons (alternatives to fossil fuels), requiring greater knowledge on the kinetic parameters involved in thermal conversion reactions for better control of the products obtained.
Among the techniques used for biomass thermal conversion, pyrolysis is considered the most promising due to the greater process flexibility, which allows obtaining liquid, solid or gas 1 .Thermogravimetric analysis (TG / DTG) has been an essential tool for modeling the pyrolysis kinetics, based on the principle that the reaction velocity depends on temperature (T), conversion (α) and time (t).Several mathematical models have been proposed for the identification of kinetic parameters such as activation energy, reaction order and conversion 1 .
The use of lignocellulosic materials in thermal processes (pyrolysis and / or gasification) has proven to be advantageous, given that these materials are renewable energy sources that can be converted into bio-oil and various products of industrial interest 2,3 , and among these materials, elephant grass stands out, which is lignocellulosic biomass easy to grow and abundant in Brazil.It consists mainly of cellulose, hemicellulose, lignin and small amounts of extractives and minerals 4 .The proportions of EG components play an important role in the distribution of pyrolysis products.Elephant grass has significant potential as a renewable energy source for fuel conversion of greater calorific power 5 .In this context, this study aimed at assessing the apparent activation energy involved in the holocellulose decomposition (cellulose + hemicellulose) of elephant grass (Pennisetum purpureum) through catalytic and conventional pyrolysis using Ti-MCM-41 as catalyst.

Catalyst synthesis
Ti-MCM-41 molecular sieve was hydrothermally synthesized using experimental procedure adapted from literature syntheses described in Braga 6 to obtain gel with the following molar composition:1.00CTMABr: 4.00 SiO 2 : xTiO 2 : 1 + x Na 2 O: 200.00 H 2 O.The "x" value was set so that the Si / Ti molar ratio remains equal to 50.After gel preparation, the material was transferred to a Teflon container with stainless steel autoclave and heated in an oven at 100°C for 120 hours (5 days).The content was washed with distilled water and then with a 2% HCl solution in ethanol for the partial removal of surfactant.Finally, the solid resulting from the crystallization process was ovendried for 12 hours and calcined at 550°C for 2 hours.

Biomass
The biomass selected for this study was elephant grass (EG) (Pennisetum purpureum Schum) from the Laboratory of Biomass and Biofuels at UFRN.The biomass was submitted to drying in the sun, ground and crushed in a Wille-type mill and sifted in sieve with mesh of 0.8-1.0mm.About 20 mg of biomass was used, which was submitted to thermal pyrolysis and for the catalytic pyrolysis, a mechanical mixture containing 70% by weight of elephant grass plus 30% of catalyst was used.

Ti-MCM-41 characterization
Ti-MCM-41 was submitted to different characterization methods to evaluate its properties for application in the catalytic pyrolysis of EG.X-ray diffractogram of the catalyst synthesized was obtained using a Shimadzu X-ray diffractometer, model XRD-7000, with CuKα monochromatic radiation (λ = 1.5406Å) obtained by 30 kV, in 30 mA filment current.The morphological characterization of the material synthesized was determined using N 2 adsorption / desorption isotherms, in QuantaChrome equipment, Model NOVA-2000.To determine the specific area, the BET method was used and for pore size distribution 7 , the Barret-Joyner-Halenda (BJH) model was used.The micrographs performed in this work were obtained in JEOL scanning electron microscope model JSM-6360 with the aim of observing the morphology, particle size and homogeneity of samples.

Determination of the kinetic parameters
Many mathematical methods have been proposed to obtain kinetic parameters from TGA experiments.In this work, two models were selected, Flynn and Wall and Vyazovkin in order to compare the results obtained in the determination of the kinetic parameters.
ASTM E 1641 norm is based on the method proposed by Flynn and Wall 8 using the Least Squares method to determine the estimated activation energy (E a ) by means of Equation 1.
Where β is the heating rate, T is the absolute temperature (K), b is an interaction variable, R is the universal gas constant and Ea is the activation energy (J mol -1 ).The Model-free Kinetics method is based on the theory of Vyazovkin 9 , which applies isoconversional techniques, in which for each conversion α, lnβ/T α 2 was graphically represented as a function 1/Tα, also providing a series of lines with slope -E a /R, so that this theory is based on Equation (2).
Where α is the conversion degree, K 0 is the preexponential factor (s -1 ), E a is the activation energy (kJ mol -1 ), R is the gas constant (8.314Jmol -1 K) and T is the temperature.Thermogravimetric analyses simulating the thermal pyrolysis were performed in a thermobalance, model DP-QA500 from Union Instruments, using nitrogen atmosphere at flow rate of 50 ml.min - in the temperature range from 30 to 900°C and heating ratesof 5, 10 and 20°C min -1 .

Results and Discussion
The X-ray diffractogram of calcined MCM-41 and Ti-MCM-41, assessing the structural properties of these samples, is shown in Figure 1.The presence of three typical peaks was observed, one of high intensity attributed to the reflection line of the plane (100) and other two, of less intensity, attributed to reflections of planes ( 110) and ( 200), characteristic of the hexagonal mesoporous structure.The introduction of titanium into theMCM-41 structure caused a decrease in the X-ray intensity of peaks, indicating a change in thestructural arrangement in relation to that found in the MCM-41structure 10 .
The morphology of the mesoporous material is formed by the agglomeration of particles with irregular and rounded shapes, as can be seen in Figure 2 through scanning electron microscopy.The thickness of the silica wall that forms the hexagonal structure of MCM-41 was obtained by the difference of the lattice parameter a 0 , obtained by X-ray diffraction, and the pore diameter d p that was obtained by BJH.Table 1 shows that the presence of titanium in the MCM-41 structure caused anincrease in the pore diameter and consequently a slight reduction in the wall thickness when compared to pure MCM-41; According to literature, increasing titanium content in the synthesis mixture results in an increase in the lattice parameter (a 0 ) and this change is probably due to the greater length of the Ti -O bond (1.80 Å) when compared to the Si -O bond (1.61 Å) 11 .Other factors that influence the structural parameters are differences in ionic radius of Ti +4 (0.68 A°) and Si +4 (0.41 A°) as can be seen in Table 2, the length of the Ti-O-Si bond differs from that of O-Si bond, certainly leading to deformation of its structure.In addition, the replacement of Si +4 by Ti +4 can block the action of the structure template, changing its ionic strength, preventing the formation of the tubular mesoporous structure.This could result in the formation of pores partially broken as can be seen by the decreased specific areas 12 .
Figures 3a and b illustrate, respectively, TG and DTG curves obtained from the catalytic and thermal pyrolysis of Ti-MCM-41/Pennisetum purpureum and pure Pennisetum purpureum, at heating rates of 5, 10 and 20°C min -1 .According to the DTG curves, Figures 4a  and b, the initial and final temperatures of thermal events involved in the decomposition of the EG biomass could be determined.Three mass losses were observed, the first at temperatures below 200°C, relative to moisture loss corresponding to approximately 8% of mass loss.In the second temperature range, from 200 to 395°C, mass loss of about 77% was represented by the most volatile matter of biomass attributed mainly to cellulose, hemicellulose and part of lignin decomposition, and the third mass loss is continuously developed overlapping the other events, being more pronounced at temperatures above 400°C.The last event is related to lignin decomposition, which has polyaromatic structure thermally more resistant than hemicellulose and cellulose 13 .Comparing the curves, it was observed that both samples exhibited pronounced mass loss at temperatures between 187 and 380°C, which represents the highest percentage of biomass volatile compounds; thus, this temperature range has been selected for kinetic studies.
Figures 5a and b shows the conversion curves versus temperature for the thermal and catalytic pyrolysis of the biomass.The kinetic study determines the decomposed fraction (α) as a function of the reaction time (t) in processes in which the temperature is kept constant or in processes in which the temperature varies linearly with time.In TG, the decomposition reaction (α) was defined by Equation 3 14 .
Where: α is the conversion, m t is the sample mass that varies with time (t), m 0 is the initial sample mass and m f is the final sample mass.
In the Model-free Kinetics, each ln β/Tα 2 conversion was graphically represented as a function of 1/Tα, providing a series of straight lines with slope -Ea/R (R = 8.314 J mol -1 K -1 ).The Flynn and Wall method graphically represented the logarithm of the heating rate (log β) versus inverse of temperature corresponding to conversion (1/Tα) from the three curves.Table 3 shows the data of the apparent activation energy at conversions of 5 to 80%, obtained by Model-free Kinetics and Flynn and Wall models, observing a decrease in the apparent activation energy for biomass pyrolysis in the presence of the catalyst, evidencing that the mesoporous Ti-MCM-41 acted as catalyst for the pyrolysis of elephant grass.

Conclusion
It was observed through the catalyst characterization that the addition of titanium did not change the MCM-41structure and morphology.Thermal analysis is a good tool to investigate the kinetic behavior of the biomass during thermal conversion processes.The apparent activation energy values for the methods applied were very close, indicating that the Model-free kinetics and ASTM methods are both adequate to determine the kinetic parameters.
The Authors acknowledge the financial support and scholarship granted by CAPES and Graduate Program in Materials Science and Engineering (PPgCEM).

Table 3 .
Apparent activation energy of biomass decomposition.

Table 1 .
Crystallographic parameters obtained by X-ray diffraction and specific area by the BET method for calcined MCM-41 and Ti-MCM-41.

Table 2 .
Si-O and Ti-O bond energy.