The pyrolysis of reed black liquor (RBL) was studied in nitrogen atmosphere by thermogravimetric analysis at six different heating rates of 5, 10, 20, 30, 40, and 50 ˚C·min-1 from ambient temperature (25 ˚C) to 800 ˚C. Thermogravimetric (TG) and differential thermogravimetric (DTG) curves were obtained. The results show that there are three main weight-loss stages in the temperature ranges of 180 to 350, 350 to 560, and 560 to 800 ˚C, for which the error is about ± 10 ˚C. The kinetic parameters were determined by the Coats-Redfern method. A kinetic compensation effect (KCE) between activation energy (E) and pre-exponential (A) factor also was found.
Kinetics of Reed Black Liquor (RBL) Pyrolysis from Thermogravimetric Data
Xingfei Song, Rushan Bie,* Xiaoyu Ji, Pei Chen, Yun Zhang, and Jun Fan
The pyrolysis of reed black liquor (RBL) was studied in nitrogen atmosphere by thermogravimetric analysis at six different heating rates of 5, 10, 20, 30, 40, and 50 ˚C·min-1from ambient temperature (25 ˚C) to 800 ˚C. Thermogravimetric (TG) and differential thermogravimetric (DTG) curves were obtained. The results show that there are three main weight-loss stages in the temperature ranges of 180 to 350, 350 to 560, and 560 to 800 ˚C, for which the error is about ± 10 ˚C. The kinetic parameters were determined by the Coats-Redfern method. A kinetic compensation effect (KCE) between activation energy (E) and pre-exponential (A) factor also was found.
Keywords: Reed black liquor; Pyrolysis; Thermogravimetric analysis (TGA); Kinetic parameters; Kinetic compensation effect (KCE)
Contact information: School of Energy Science and Engineering, Harbin Institute of Technology, Harbin, 150001, China; *Corresponding author: firstname.lastname@example.org
Black liquor (BL), a by-product during the alkaline cooking process of fibrous plants, such as wood or non-wood (cereal straw, reed, elephant grass, Phalaris arundinacea, miscanthus, sorghum, kenaf, hemp, and sugarcane bagasse), is a water pollutant that contributes about 90% of the total wastewater in a paper mill. BL compositions are complex, containing lignin, hemicelluloses, other carbohydrates, and inorganic compounds. However, BL is also a special type of energy source because the chemicals and energy contained in its organics can be recovered and reused in pulp and paper production through special treatment technologies (Sánchez et al. 2005).
Currently, BL from wood is able to be burned in a conventional recovery boiler (Tomlinson recovery boiler) for generating energy and reusing chemicals in the papermaking process (Sjöström 1993). Compared with BL from wood, BL from straw contains more silicon and potassium. The high silicon content can increase the construction costs because it can reduce the concentration efficiency of evaporators. Additionally, the high potassium content can decrease the melting point of BL, which could cause a corrosion problem and decrease the boiler efficiency (Dickinson and Verrill 1988). Although the conventional recovery boiler has been used for many years, there are also some problems in safety and environment, such as low power to heat ratio, low thermal efficiency, high capital and maintenance cost, and risk of smelt-water explosion (Stigsson 1998). Therefore, an increasing amount of research has focused on alternative recovery technologies that are much safer, easier, and more energy efficient than the conventional technology (Whitty and Verrill 2004). Pyrolysis and gasification are two potential methods for gaseous production during BL capacity reduction. Some studies have been reported on the pyrolysis and gasification of wood-derived BL (Frederick et al. 1993; Alén et al. 1995). A series of studies on straw BL has been performed using thermogravimetric analyzers, different bench scale reactors, and a pilot scale reactor (Puértolas et al. 2001; Gea et al. 2002, 2003; Sánchez et al. 2005).
Reed black liquor (RBL), a type of good quality straw BL, plays an increasingly important role in the Chinese paper industry. The extractability of hemicellulosic components from RBL was investigated (Hao and Shigetoshi 2014). The combined effect of ultrasound and polyferric sulphate on the RBL has been researched (Shen et al. 2005). The reuse from RBL has also been researched (Sun et al. 2006, 2007; Xu et al. 2012). Moreover, CO2 gasification characteristics of reed kraft black liquor have been studied (Yang et al. 2012, 2013). However, little work has been published on the pyrolysis of RBL. Therefore, it is necessary to study the pyrolysis characteristics and kinetics of RBL.
The objective of this work was to investigate the thermal degradation behavior of RBL in N2atmosphere at different heating rates. Moreover, the kinetic parameters (E and A) of RBL pyrolysis were determined by the Coats-Redfern method. The results could accelerate the development of RBL pyrolysis and provide some useful experimental data for practical application. What’s more, E and A could be used in numerical simulation models to investigate the RBL reaction behavior under different operating conditions.
The RBL used in this work was received from a Chinese paper mill in Daqing of Heilongjiang province. The RBL has a solids content of 52.60%. Proximate analyses of RBL were obtained by drying oven (CS101-E, China) and muffle furnace (YX-HF, China) using ASTM standards. Ultimate analyses of the dried basic for C, H, O, N, and S were obtained by an organic elementary analyzer (Vario Micro cube, Germany), and other elements were tested by X-Ray fluorescence (AXIOS-PW4400, Netherlands). The results are shown in Table 1, and the net heating value was determined as 13.35 MJ·kg-1 (YX-ZR/Q 9704, China). Similar to other black liquor from straw , high ash content (20.63%) and a larger quantity of Na (17.64%), K (1.92%), Si (1.78%) and Cl (1.57%) were found. Before being used in the TGA experiment, the RBL was completely dried in a drying oven (105 ± 3 oC). After dying, the RBL solids were smashed. Two sieves of 75 and 105 μm were used to filter RBL solid powder. The residue powder retained in the 75 μm size sieve was used in the following experimental procedure.
Table 1. Ultimate and Proximate Analyses (Dry Basis) of RBL
A TGA ZRY-2P Series TG analyzer (Shanghai Precision and Scientific Instrument Co., Ltd., Shanghai, China) was used for all TGA experiments. During all the experiments, nitrogen was used as the flow gas (60 mL·min-1) to ensure the occurrence of the pyrolysis reaction. The initial mass of the dry RBL powders, placed in an alumina crucible (5 mm internal diameter and 4 mm height), was approximately 7 mg. Samples were heated from ambient temperature (25 oC) to 800 oC at different heating rates (β=5, 10, 20, 30, 40, and 50 oC·min-1).
Kinetics Analytical Methods
The Coats-Redfern equation (Eq. 1) was used for analysis.
During the pyrolysis of a specified substance, the reaction mechanisms should be the same. There were 30 main reaction mechanisms shown in the following reference (Zhao et al. 2010). Therefore, from the linear relationship between ln[G(α)/T2] and 1/T, the parameters E, and Acould be obtained from the slope and intercept, respectively. Among the 30 results, the best reaction mechanism can be chosen to be called the most probable mechanism function, according to the coefficient of determination (R2) that is closest to one.
RESULTS AND DISCUSSION
The TGA and DTG curves are shown in Figs. 1 and 2, respectively. Different liquors can behave very differently in the same reactor under the same conditions (Whitty et al. 1997). Thermal decomposition of straw black liquor seemed to happen in two different steps which temperature ranges were 200 to 550 oC and 550 to 900 oC (Gea et al. 2002). The pyrolysis of black liquor from the mixture of bamboo and hardwood included 3 stages (Wu et al. 2007). But there was only one weight loss stage during wheat straw pyrolysis (Yang and Wu 2009). However, in this work, three weight loss stages had been found during the pyrolysis process. The three temperature ranges were 180 to 350, 350 to 560, and 560 to 800 oC ( ± 10 oC) which were similar to the result of the thermal degradation of neutral sulfite semi-chemical pulp black liquor (Zhao et al. 2010).
In the first stage, the weight loss ranged from 19.83% to 21.51%, which can be attributed to the decomposition of various organic matters, such as hemicelluloses (200 to 260 oC), low molecular weight organic acids (200 to 300 oC), and some lignin (280 to 500 oC) (Oren et al. 1984; Gu et al.1992). In the second stage, the weight loss ranged from 9.13% to 15.37%, which was mainly due to the decomposition of lignin (280 to 500 oC) (Alén et al. 1995). In the third stage, the weight loss was about 11.5% to 22.11%, due to the decomposition of organic compounds and inorganic salt sodium in RBL, in addition to the reactions among salt sodium, in which carbon and a minor portion of the organic matter remained (Wäg 1996).
From Figs. 1 and 2, it can be seen that the peak temperature of each stage increased as the heating rate increased. As the heating rate increased, reaching the same temperature required shorter time, and the reaction became much stronger. In other words, the RBL was heated inhomogeneously, and some reactions did not occur while the temperature rose. Therefore, the temperature lag effect caused the peak temperatures of curves to move to the high-temperature side.
Fig. 1. TGA curves of RBL on dry basis
Fig. 2. DTG curves of RBL on dry basis
Table 2 shows the pyrolysis analysis data of the RBL solid powder at different heating rates. During the wood BL pyrolysis process, final solid conversion increased as the heating rate increased, due to a higher production of volatiles, as has been found (Frederick et al. 1994). But Gea et al. (2002) found that the trend of the final solid conversion was not clear during the thermogravimetric study of alkaline straw BL. They found that these deviations may be attributed to differences of composition of BL solid powder.
In the present work, the final solid conversions of RBL used also did not show the same trend, which is evident by the last row of Table 3. However, we consider this to be a result of the deviations of the intensity of the reaction. When β is greater than 10 oC·min-1, as β increases, pyrolysis time decreases, reactions strengthen, and more matter is released, so the solid conversion rises. When β is smaller than 10 oC·min-1, the reactions are closer to an ideal process and each reaction has sufficient time to react isothermally and completely. Therefore, the final solid conversion is greater than 0.5, and the solid conversion decreases when β increases. But when β is very small, some reactions do not occur. Therefore, the final solid conversions of 5 oC·min-1 (53.75%) and 10 oC·min-1 (50.70%) were smaller than those of 40 oC·min-1 (52.48%) and 50 oC·min-1 (55.56%).
It also was found that the highest weight loss occurred in the first stage and third stage at a low heating rate (5 and 10 oC·min-1) and a highest heating rate (β = 50 oC·min-1). The first stage weight loss value was similar, having a value between 19.83% and 21.51%. However, as the heating rate increased, the second stage weight loss increased at first and then decreased, and the third stage weight loss decreased at first and then increased. This may be an indication of the temperature lag effect.
Figure 2 and Table 2 also show that the devolatilization rates of RBL when β = 50 oC·min-1 were greater than for other conditions. When β was 5 oC·min-1, the devolatilization rate of the second and third stages were similar, and the first stage value was the highest of the three.
Not including the 5 oC·min-1 result, it can be seen that the devolatilization rates of the three stages were ranked as the first stage > the second stage > the third stage. Generally speaking, organic matter fraction devolatilization (T < 560 oC) was higher than that of the inorganic matter fraction (T > 560 oC).
Table 2. Pyrolysis Analysis Data of RBL on Dry Basis at Different Heating Rates
Comparing the 30 results, the R2 value of the 13th reaction mechanism is closest to one, which suggests that it is the most probable mechanism function to describe the kinetics for the RBL pyrolysis in nitrogen atmosphere during all three stages. The integral form is G(α)=[-ln(1-α)]4, and the function of this reaction mechanism, the Avrami-Erofeev equation, is f(α)=(1-α)[-ln(1-α)-3]. Therefore, the RBL pyrolysis reaction mechanism may be random nucleation and subsequent growth. This model is based on the assumption that in the course of the pyrolysis of RBL, active centers are generated at random local points, and some of the active centers produced decomposition products or become deactivated, but the others continue to produce some active centers. The reaction order was four.
Various kinetic parameters calculated with the Avrami-Erofeev equation (n=4) reaction mechanism using Coats-Redfern method are shown in Table 3. The R2 value in the first stage was the closest to one, meaning the linear fitting relevance between this reaction mechanism and the experimental result is satisfied. This reaction mechanism can accurately describe RBL pyrolysis in the first stage. But in the second and third stages, the R2 values are less than the first stage and far less than one, meaning that using this reaction mechanism to describe the second and third stages of RBL pyrolysis is not satisfactory, especially for the third stage. The results could also mean that pyrolysis of RBL is a complex reaction, especially in the third stage (560 to 800 oC).
Table 3. Kinetic Parameters of the RBL Pyrolysis Using Coats-Redfern Method
Kinetic compensation effect
From Table 3, it can also be seen that the variation of E gave rise to a variation in A, which can be explained by a predictable relationship between E and A (ln(A)=aE+b). This relationship is referred to as kinetic compensation effect (KCE). It reflects a compensation effect between the exponential and pre-exponential factors in the Arrhenius equation (Andreasen et al. 2005). It can be used to calculate E or A at other experimental conditions according existing experimental data to reduce the number of experiments. And it can also be used to test the experimental data. The plots of lnA–E of different methods are shown in Fig. 3. The KCE relationship between E and Ahas been found in the thermal decomposition of RBL.
Fig. 3. Linear fitting between Y=E and X=ln(A) from Table 4 and 5
- Three main weight loss stages were observed during RBL pyrolysis: 180 to 350 oC, 350 to 560 oC, and 560 to 800 oC. The three stages obtained at different heating rates were similar, and the errors were within ± 10 oC. Solid conversions were greatly influenced by the heating rates.
- The most probable mechanism function of the RBL pyrolysis in nitrogen was determined to be the Avrami-Erofeev equation (n=4), which can be expressed in integral form as G(α)=[-ln(1-α)]4. It showed that during RBL pyrolysis, active centers become generated at random local points. And then, some of the active centers produced decomposition products or are deactivated, but the others continue to produce some new active centers.
- E and A obtained by the Coats-Redfern method at different heating rates satisfied a kinetic compensation effect.
The authors are grateful for the support of the Scientific and Technological Research Projects of Harbin (2012DB2CP024).
Alén, R., Rytkönen S., and McKeough P. (1995). “Thermogravimetric behaviour of black liquors and their organic constituents,” J. Anal. Appl. Pyrolysis 31, 1-13.
Andreasen, A., Vegge T., and Pedersen A. S. (2005). “Compensation effect in the hydrogenation/dehydrogenation kinetics of metal hydrides,” J. Phys. Chem. B 109, 3340-3344.
Dickinson, J. A., and Verrill, C. L. (1998). “Development and evaluation of a low-temperature gasiﬁcation process for chemical recovery from kraft black liquor,” Proceedings of the International Chemical Recovery Conference, Florida.
Frederick, W. J., and Hupa, M. (1994). “The effects of temperature and gas composition on swelling of black liquor droplets during devolatilization,” J. Pulp Paper Sci. 20(10), 274-280.
Frederick, W. J., Wag K., and Hupa M. (1993). “Rate and mechanism of black liquor char gasification with CO2 at elevated pressures,” Ind. Eng. Chem. Res. 32(8), 1747-1753.
Gea, G., Murillo, M. B., and Arauzo, J. (2002). “Thermal degradation of alkaline black liquor from straw – Thermogravimetric study,” Ind. Eng. Chem. Res. 41(9), 4714-1721.
Gea, G., Murillo, M. B., Sánchez, J. L., and Arauzo, J. (2003). “Thermal degradation of alkaline black liquor from wheat straw. 2. Fixed-bed reactor studies,” Ind. Eng. Chem. Res. 42(23), 5782-5790.
Gu, P., Hessley, R. K., and Pan, W. P. (1992). “Thermal characterization analysis of milkweed floss,” J. Anal. Appl. Pyrolysis 24, 147-164.
Hao, R., and Shigetoshi, O. (2014). “Comparison of hemicelluloses isolated from soda cooking black liquor with commercial and bacterial xylan,” Cellul. Chem. Technol. 48(7-8), 675-681.
Oren, M. J., Nassar, M. M., and Mackay, G. D. M. (1984). “Infrared study of inert carbonization of spruce wood lignin under helium atmosphere,” J. Spectrosc. 29(1), 10-12.
Sánchez, J. L., Gonzalo, A., and Gea, G. (2005). “Straw black liquor steam reforming in a fluidized bed reactor. Effect of temperature and bed substitution at pilot scale,” Energy Fuels19(5), 2140-2147.
Shen, Z. S., Lan, C. Q., and Shen, J. Z. (2005). “Combined effect of US/PFS on the black liquor of making paper,” J. Environ. Sci. 17(1), 110-112.
Sjöström, E. (1993). Wood Chemistry: Fundamentals and Applications, 2nd Ed., Academic Press, San Diego, California.
Stigsson, L. (1998). “ChemrecTM black Liquor gasification,” Proceedings of 1998 International Chemical Recovery Conference, Florida.
Sun, Y., Zhang J. P., Yang, G., and Li, Z. H. (2006). “Removal of pollutants with activated carbon produced from K2CO3 activation of lignin from reed black liquors,” Chem. Biochem. Eng. Q.20(4), 429-435.
Sun, Y., Zhang, J. P., Yang, G., and Li, Z. H. (2007). “Preparation of activated carbon with large specific surface area from reed black liquor,” Environ. Technol. 28(5), 491-497.
Wäg, K. (1996). Characterization and Modelling of Black Liquors Char Combustion Processes, Doctoral thesis, Oregon State University.
Whitty, K., Backman, R., Forssén, M., Hupa, M., Rainio, J., and Sorvari, V. (1997). “Liquor-to-liquor differences in combustion and gasification processes-pyrolysis behavior and char reactivity,” J. Pulp Pap. Sci. 23(3), 119-128.
Whitty, K., and Verrill, C. L. (2004). “A historical look at the development of alternative black liquor recovery technologies and the evolution of black liquor gasifier designs,” Proceeding of 2004 International Chemical Recovery Conference, South Carolina.
Wu, S. B., Tang, Y., Guo, Y. L., and Liu, J. Y. (2007). “Thermogravimetric properties of black liquor and corresponding kinetic analysis,” J. S. China, University, Technol. (Na. Sci. Ed.) 35(6), 59-63. (in Chinese)
Xu, K. M., Liu, H. Y., Dong, L., Tong, S., and Liu, Y. J. (2012). “Lignin recovery from the black liquor of reed pulping,” Adv. Mater. Res. 512-515, 2376-2380.
Yang, Q., and Wu, S. B. (2009). “Thermogravimetric properties and kinetic analysis of wheat straw,” Trans. Chin. Soc. Agric. Machinery 25(3), 193-197. (in Chinese)
Yang, Q., Yin, X. L., Wu, C. Z., Wu, S. B., and Guo, D. L. (2012). “Thermogravimetric-Fourier transform infrared spectrometric analysis of CO2 gasification of reed (Phragmites australis) kraft black liquor,” Bioresour. Technol. 107, 516-516.
Yang, Q., Yin, X. L., Wu, C. Z., Wu, S. B., and Guo, D. L. (2013). “The influence of sodium cation on characteristics of CO2 gasification of lignin from reed kraft black liquor,” Trans. Chin. Soc. Agric. Machinery 44(3), 137-141. (in Chinese)
Zhao, Y., Bie, R. S., Lu, J., and Xiu, T. C. (2010). “Kinetic study on pyrolysis of NSSC black liquor in a nitrogen atmosphere,” Chem. Eng. Commun. 197(7), 1033-1047.
Article submitted: September 8, 2014; Peer review completed: October 15, 2014; Revised version received and accepted: November 3, 2014; Published: November 11, 2014.