Carbonization of "algarrobo negro" (prosopis nigra,): a study of its microstructure and main volatile components

C.E López Pasquali. C. A. Wottitz. R.G. Martinez and H. A. Herrera

Instituto de Ciencias Químicas. Facultad de Agronomía y Agroindustrias. Universidad Nacional de Santiago del Estero. Avda. Belgrano (s) 1912. (4200) Santiago del Estero. Argentina

Abstract&— Modifications in the microstructure of wood from Algarrobo Negro (Prosopis Nigra) in function of the temperature were studied by scanning electronic microscopy (SEM). Some non condensable gases obtained by isothermal pyrolysis were investigated employing gas chromatography (GC). The results showed that the basic anatomic structure of wood remains almost unchanged in the working temperature range. It was also found that in the gas phase and with increasing temperature, the concentrations relative to carbon monoxide and water decrease whereas there is a pronounced increase in the concentration of methane from 300ºC.

I. INTRODUCTION

We have previously studied the thermal degradation process of wood sawdust of six hardwood species from the Chaco-Santiagueño region in Argentina under isothermal conditions. "Algarrobo Negro" (Prosopis Nigra) was one of the wood species analyzed there (Herrera et al., 1986), and is the species studied in this work. We have also reported on the pyrolisis of its lignins (López and Herrera, 1997). The thermal behavior of each of the main components of wood (lignin, a-cellulose and hemicelluloses A and B) gives information about the different processes that take place in each of them and that affect the overall thermal degradation of wood (Herrera et al., 1989 and Wottitz et al., 2000). In 1988 Bourgois and Guyonnet demonstrated that the hemicelluloses in pine wood are the most sensitive thermal components. Therefore, they are the first components to degrade and go into the gas phase. They also, start the decomposition of lignin that has a more stable structure.

As the size and distribution of the particles influence the reactions occurring during the wood degradation processes (Lahiri, 1980), we employed here wood cubes having a side of 2 mm and analyzed the degradation process under these new working conditions. The experimental information gathered was compared with the behavior predicted by the Avrami Erofeev and Arrhenius equations. Agreement was observed between the experimental results and the Avrami Erofeev equation. In contrast, no agreement was found with the Arrhenius equation in the whole temperature range studied but the model was found suitable when separate and definite temperature ranges were taken into account (López Pasquali and Suarez, 2000).

Scanning electron microscopy (SEM) has been employed by several authors to explain the changes that occur in wood as a result of temperature effects. Thus, Kollmann and Sach (1967) employed SEM to get an insight into the high temperature effects on certain wood cells from Fagus Sylvatica. Zicherman and Williamson (1981 and 1982) used SEM to observe the changes in wood microstructure as a result of fire effects and in 1982 and 1992 several authors have reported on the effect of different chemical substances applied to wood as fire retardants.

Connor and Daria (1993) found in wood from Eucaliptus Delegatensis that visual observation studies give quantitative information on the changes that occur in wood when it is carbonized. Rivera et al. (1994) used SEM to identify vegetal charcoal whereas the use of this technique allowed Donaldson (1995), to determine the fractures of the cells on the tangential surface of air dried woods.

Gas chromatography analysis of the gases generated during thermal degradation of woods has been used by Bourgois and Guyonnet (1988) in pine wood and Faix et al. (1991) in Picea Abies and Fagus Sylvatica. This technique allows identification of the gases as well as understanding the pyrolysis evolution during carbonization.

It is possibly that during this process the reactions are strongly influenced by the resistances to mass and heat transfer in wood particles. This resistance is dependent upon the inner structure of the wood and the physical and chemical changes that are produced during carbonization. As no information is available in the literature on wood from "Algarrobo Negro", we investigated the modifications of the microstructure in function of the temperature and of the analysis of some non condensable gases obtained from isothermal pyrolysis. Our aim was to obtain further information on the physical and chemical changes that take place during the isothermal pyrolysis of wood.

II . MATERIALS AND EQUIPMENT

A hardwood from the "chaqueña" region in the Argentine Republic, "Algarrobo Negro" (Prosopis Nigra) was selected. The samples preparation for their later analysis required close attention due to the hardness and other particular features of this wood. On this account, we considered the following steps: a) sectioning, b) softening and c) smoothening of the surface.

1. Sectioning: thick discs (1-1.5 cm) were cut with a saw from a trunk chosen at random. Then blocks with a side of approximately 1.5 cm were cut, taking into account that they should be displayed in one of the radial, tangential or transversal planes. A magnifying lens (10 x) was employed for observation of the plane.
2. Softening: the blocks were placed in a container with water which was connected to a refrigerant tube to avoid evaporation losses. The water boiled for around 1.40 h until the blocks became soft enough to facilitate sectioning. The softening time was established by trials and error tests, considering that the wood sample was soft enough when it was possible to cut 10-15 microns layers without fractures using a xylontrom.
3. Smoothening: the thickness of the blocks was reduced down to 2 mm with a xylontrom by smoothening the surface corresponding to the cross, tangential or radial sections. They were air dried and cubes of side 2 mm were cut. They were stored in sealed containers.

A Thermobalance Stanton Redcrof 750 TG was employed for thermal pyrolysis in a nitrogen atmosphere (50 cc/min) in the temperature range between 200 and 400 °C (when the pyrolysis process takes place, Herrera et al., 1986), using samples of around 20 mg of mass. The samples were placed in the oven, which was previously set to the working temperature and placed in the crucible of the thermobalance, with the smooth surface up. A different wood sample was used for each determination. Temperature was kept constant for 60 minutes, the solid residue (SR) obtained was removed and attached to the plate of the electronic microscope so as to keep the smooth surface up. The residues thus obtained were analyzed by scanning electron microscopy (SEM) using a Scanning Electron Microscope JEOL 35 CF.

The non condensable gases were simultaneously collected for chromatographic analysis using a solid-gas chromatograph Konik 3000 with a thermal conductivity detector, a 6" long column and 1/8" internal diameter, packed with Porapak Q 80/100 and nitrogen as gas carrier. The working conditions were: oven temperature 40ºC, detector 125 ºC, injector 110 ºC, filament current 135 mA, injection volume 5 ml, nitrogen flow 20 ml/min. The peaks corresponding to methane, carbon monoxide and dioxide and water were determined under these conditions using pure gases.

As on-line determinations could not be made, a glass ampule was fabricated for this purpose. It was (200 ml) had a teflon inlet and outlet and a device with an exchangeable silicon septa that allowed extraction of different volumes of gas. Other syringes with different volumes were also employed to injected gas into the chromatograph. used to carry the pyrolysis gases along a nitrogen current coming from the thermobalance. The ampule (200 ml) had a teflon inlet and outlet and a device with an exchangeable silicon septa that allowed extraction of different volumes of gas. Other syringes with different volumes were also employed to injected gas into the chromatograph.

III - RESULTS AND DISCUSSION

SEM analysis: we began analyzing the SR obtained by treating the samples at 200ºC and 15ºC intervals beginning at 200ºC. As no substantial differences were observed in samples treated in such a short temperature gap, the temperature increase from one sample to the next was set to 50ºC. Therefore, the samples obtained at 200, 250, 300, 350 and 400ºC were analyzed in detail and according to the different cutting.

Examination of the microphotographs revealed structural similarities and differences between wood and the carbonization product. Thus, the microphotographs of the SR according to the transversal and tangential planes (Fig. 1) show that the wood structure of "Algarrobo Negro" does not suffer significant qualitative changes because of the temperature effect, that is, the carbonized wood keeps its basic structure almost unaltered.

However, the mechanical tissue that strengthens the wood showed increasing changes with temperature rise until small fractures are noticed in the sample treated at 400°C. This alteration could be due to the effect of different contractions that are produced in its secondary and primary walls.

On the other hand, the analysis of microphotographs of the axial parenchyma from the cross surface, (Fig. 2) shows that although the cells from the axial parenchyma keep their shape and structure, the thickness of its walls decreases with increasing temperature. Thus and in order to look into the way in which such modification was produced, the thickness of the cell wall was determined from these photographs as the average of thirty measurements obtained at each temperature. The values obtained were plotted for wood without carbonization and for the different SR in function of the temperature (Fig. 3). An important diminution of the wall thickness was observed in the samples heated at 200 ºC, when compared with non-carbonized wood. The cell wall thickness keeps going down until the sample reaches 400 ºC. From 200° C to 300 ºC, the thickness decrease is more import than the decrease observed from there to 400ºC. These result coincide with the dehydration process (Herrera et al., 1986) which runs up to 200 ºC and the following stages that run consecutively, from 200 ºC to 300 ºC, and from 300 ºC to 400 ºC. On the other hand from the thermograms obtained under isothermal conditions, the experimental values of m_i = initial mass of the wood sample to heat, and m_t = mass corresponding to time t of the SR were determined for each temperature.

Figure 1: Microphotographs of the SR according to the transversal and tangential planes.

The mass loss was determined from these parameters and is defined as follows:

Dm = m_i - m_t                               (1)

The gas analysis was done simultaneously with the SR obtained.

As the relative concentrations of each gas for a certain temperature varies according to the heating time (Oren et al., 1987), the samples were taken at 1.5 and 3 min. after placing the sample in the oven. Fig. 4 (a, b and c) shows the chromatograms performed at 3 minutes at 200, 300 and 400 ºC. The carbon monoxide and water peaks are observed in all these temperatures, except for the carbon dioxide whereas the methane peak appears only after 300 ºC.

The area for the peaks of carbon monoxide, water and methane was calculated from the chromatographed gases obtained at 1.5 and 3 min after placing the sample in the oven, The relative concentration (X) of each gas is represented by the equation.

X = area / Dm (2)

The values obtained for each temperature are listed in Table 1. These values are plotted in function of the temperature and are shown in Figure 5 (5a for 1.5 min and 5b for 3 min).

Figure 2: Microphotographs of the axial parenchyma.

It can be clearly seen in this figure the variations of the relative concentration for each gas according to time and temperature. Both plots show that the temperature increase occurs with a diminution in the relative concentrations for carbon monoxide and water, an increase in methane relative concentration takes place after 300ºC, more evident in samples taken at 3 minutes.

The kinetic study of wood from "Algarrobo Negro" under isothermal conditions (López Pasquali and Suarez, 2000) indicated the presence of two well defined kinetic processes, one between 200 and 300°C and the other between 300 and 400°C, with activation energy values of 6.55 and 69.45 Kj mol^-1 respectively, for reactions in the order of n = 0.5.

Figure 3: The thickness of the cell wall for wood without carbonization and for the different SR in function of the temperature.

In a previous study we showed the effect of the degradation processes of the main components of wood from "Algarrobo Negro" on the overall degradation process (Herrera et al., 1989, Wottitz et al., 2000). It was observed that the components do not degrade simultaneously but follow a certain sequence. It was also found that almost complete degradation of hemicellulose A and B takes place at 300ºC whereas between this temperature and 400ºC, the degradation of a-cellulose and lignin occurs.

On account of the foregoing results it can be suggested that the degradation process of "Algarrobo Negro" takes proceeds in two steps, the first between 200°C and 300°C, with a pronounced diminution of the cell wall thickness, an important contraction of the tissues, the appearance of the maximum peaks of carbon monoxide and water.tissues, the appearance of the maximum peaks of carbon monoxide and water.

Figure 4: Chromatograms performed at 3 minutes: (a) 200 ºC, (b) 300 ºC and (c) 400 ºC.

Table 1: Values of mass loss Dm and of area / Dm in function of temperature.

Tempera-ture ºC	Dm g	areas/ Dm at 1.5 min	areas/ Dm at 3 min
CO	H2O	CH4	CO	H2O	CH4
200	0.95	335..9	218.3		256.5
230	0.75	298.0	203.1		243.1	86.2
250	1.63	202.1	46.2		240.7	5.3
280	1.55	183.2	33.1		111.3	14.1
300	4.25	153.4	35.2	44.8	102.3	2.1	6.4
310	3.27	135.0	30.5	47.5	71.2	9.8	18.1
325	5.80	122.5	15.6	51.2	67.8	8.6	12.7
350	6.30	117.8	10.4	137.0	65.1	5.2	27.3
400	12.5	105.4	5.8	220.3	31.3	3.6	44.2

 

Figure 5: area / Dm in function of temperature: (a) for 1.5 min and (b) for 3 min.

It all agree with the first kinetic process whose activation energy is 6.55 Kj mol^-1 for n=0.5 and coincide with the temperature range where the degradation of hemicelluloses A and B is produced. The second step occurs between 300 and 400ºC in accordance with the second kinetic process and whose activation energy is 69.45 Kj mol^-1 for n=0.5. This step shows a lesser variation of the cell wall thickness, a larger structural alteration, appearance of cracks and

larger alteration of the mechanical tissue along with appearance of the methane peak. This step also coincides with the temperature range where the degradation of a-cellulose and lignin is produced.

These data allows us to suggest that wood from "Algarrobo Negro" behaves in a similar way to the species studies by Connor and Daria (1993), when we analyze the influence of particle size in the carbonization process, and the resistance to mass and heat transfer.

IV CONCLUSIONS

The results obtained suggest that the basic anatomic structure of wood remains almost unchanged during the pyrolysis. This in turn would indicate that mass and energy transfer in bigger particles would be the critical step of the carbonization process.

Even though the temperature does not substantially modify the basic structure, there is a general contraction of tissues which is seen in the diminution of the cell wall thickness of the axial parenchyma and the contraction of the mechanical tissue. The fact that the original basic structure is kept in the solid residues can be attributed mainly to residual lignin, which is the structural component of wood that degrades more slowly.

V. REFERENCES

Bourgois J. and Guyonnet - Characterization and analysis of torrefiel wood - Wood Sci. - 22, 143-155, 1988        [ Links ]

Connor M.A. and Daria V. - Changes in wood structure during the course of carbonisation - Advances in Thermochemical Biomasa conversion- Editorial Elsevier, 1993.        [ Links ]

Donaldson L.A. - Cell wall fracture properties in relation to lignin distribution and cell dimensions among three genetic groups of radiata pine - Wood Sci. 29-1, 50-63, 1995        [ Links ]

Faix O., Fortman I, Bremer J. and Meier D. - Thermal Degradation Products of Wood Holz als Roh-und Werkstoof, 49, 213-219 , 1991        [ Links ]

Fengel D. and Wegener G., Wood chemistry ultraestructure reactions , Edic. De Gruyter , 1984.        [ Links ]

Herrera H., Araya C. E., Wottitz A., Velez S. and Carranza M. - Descomposición térmica de celulosa, lignina y hemicelulosa A y B, First Brazilian Symposium on the Chemistry of Lignins and Other Wood Component, vol II, 356-359, 1989.        [ Links ]

Herrera H., Soria T. and Araya C. E., A Kinetic Study on the Thermal Decomposition of Six Hardwood Species, Holz als Roh-und Werkstoof, 44, 357-360, 1986.        [ Links ]

Kollmann FFP and Sach I.B - The Effect of Elevated Temperature on Certain Wood Cells - Wood Sci. 1, 14-25, 1967.        [ Links ]

Lahiri A. K., The effect of Particle size distribution on TG, Thermochimica Acta 46, 289-295 (1980).        [ Links ]

López Pasquali C. E. and Herrera H. - Pyrolysis of lignin and IR analysis of residues, Thermochimica Acta 293, 39-461 (1997).        [ Links ]

López Pasquali C.E. and Suarez E. - Parámetros Cinéticos de la Pirólisis Isotérmica de Prosopis Nigra. Análisis Estadístico de los Resultados - Información Tecnológica 11-2, 105-108, 2000        [ Links ]

Oren M.J., Karunakaran K.P., Pegg M.J. and MacKay G.D.M. - Mechanism of wood combustion 1: Flue gas analysis - Fuel 66, 9-11, 1987.        [ Links ]

Rivera S.M., Magnin.S. and Cavatorta M. - Aporte del microscopio electronico de barrido en la identificacion del carbon vegetal - Quebracho 2, 21-27, 1994.         [ Links ]

Wottitz C. A., López Pasquali C. E. and Herrera H. A.- Thermal Decomposition of Algarrobo Negro (Prosopis nigra). Wood and its main components: lignin, alpha cellulose and hemicelluloses "A" and B" - Latin American Applied Reseach, 31, 37-40 2001        [ Links ]

Zicherman J.B. and Williamson R.B. - Microstructure of Wood Char - Part 1- Whole Wood - Wood Sci. 15, 237-249, 1981.        [ Links ]

Zicherman J.B. and Williamson R.B. - Microstructure of Wood Char - Part 2: Fire Retardant Treated Wood - Wood Sci. 16, 19-34, 1982.        [ Links ]

Zicherman J.B. and Allard D.L. - Fire Performance of Fire-retard Wood Fiberboard Ceiling Tile Fire and Materials 16, 187-196, 1992.         [ Links ]

Received: December 7, 2000.
Accepted for pubblication: April 19, 2001.