Influence of the biomass components on the pore

Biomass and Bioenergy 97 (2017) 53e64

Contents lists available at ScienceDirect

Biomass and Bioenergy journal homepage: http://www.elsevier.com/locate/biombioe

Research paper

Influence of the biomass components on the pore formation of activated carbon Catalina Rodriguez Correa a, *, Thomas Otto b, Andrea Kruse a a

Institute of Agricultural Engineering, University of Hohenheim, Garbenstrasse 9, 70599, Stuttgart, Germany €ude 727, 76344 Institute for Catalysis Research and Technology, Karlsruhe Institute of Technology (KIT), Hermann-von-Helmholtz-Platz 1, Geba Eggenstein-Leopoldshafen, Germany

b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 13 June 2016 Received in revised form 16 December 2016 Accepted 18 December 2016

Uncontrolled management of agricultural wastes have strongly contributed to the increase of greenhouse emissions and pollution. On the other hand, these residues can be used as a sustainable source for the production of activated carbon. Currently, biomasses rich in lignin are the most widely used, due to the high yields and large surface areas attainable. The aim of this study is to understand the influence of each biomass component on activated carbon properties. Alpha-cellulose, xylan, kraft lignin, and mixtures with different ratios of the single components were used as model substances to represent biomass. These materials were pyrolyzed and subsequently activated with KOH to expand the surface area. TGA results showed no interaction between components during pyrolysis but there was a strong influence of the composition of the mixture on the activated carbon properties due to the different thermal stabilities of each char. The activated carbon with the largest apparent surface area was obtained from cellulose with 2220 m2 g1 and pure xylan showed the lowest with 1950 m2 g1. T-plot calculations showed that more than 90% of the surface area was composed by micropores. To understand the microporosity, CO2 isotherms were measured. The surface areas calculated were lower but followed the same trend as those obtained from the isotherms with N2. © 2016 Elsevier Ltd. All rights reserved.

Keywords: Biomass Activated carbon Surface area Microporosity

1. Introduction With attainably high surface areas and porosity volumes, lignin or lignin-rich biomass presents a highly interesting feedstockmaterial in producing activated. Additionally, woody biomass yields a considerably higher amount of product compared to grassy biomasses [1]. This leads to the assumption that not only the presence of lignin is important, but also the whole composition or the structure needs to be considered. However, whether lignin has the most appropriate structure for porosity development, especially microporosity, is a question which still remains. Additionally, the influence of the other biomass components (cellulose and hemicellulose) on porosity formation requires a more thorough research. The main scope of this work is to understand how biomass cell-wall components affect surface area and porosity formation. The first important criterion is concerned with steam activation

* Corresponding author. E-mail address: [email protected] (C. Rodriguez Correa). http://dx.doi.org/10.1016/j.biombioe.2016.12.017 0961-9534/© 2016 Elsevier Ltd. All rights reserved.

of carbon materials. Cagnon et al. [2] studied the contributions of biomass compounds to the porosity of steam activated carbons using model compounds and biomasses with different proportions of hemicellulose, cellulose and lignin. They identified that only the model compounds and those biomasses with high lignin contents (coconut shell, olive stones treated with sulfuric acid and soft wood) were able to form a fairly high micropore volume during the pyrolysis stage. Biomass with a high percentage of extractives showed no evidence of microporosity at all. Nonetheless, after steam activation, most biomasses had developed a moderate microporous structure, whereas those with high lignin content had the highest micropore volumes. The only exception was apple pulp, which showed the highest micropore volumes despite its low lignin content. The findings of Cagnon et al. [2] are consistent with several studies conducted with palm oil stones or coconut shells. Guo et al. [3,4] pyrolyzed palm oil stones and obtained relatively high surface areas (approximately 300 m2 g1). Additionally, the type I N2 isotherms of all the chars indicated the presence of a microporous structure, which corresponded to approximately 60% of the whole surface. The same group [5] activated these palm oil stones with

54

C. Rodriguez Correa et al. / Biomass and Bioenergy 97 (2017) 53e64

CO2 after a previous impregnation with KOH and H2SO4. They again found a microporous structure by means of the N2 isotherm, and in both cases the microporous volume had increased proportionally to the impregnation ratio. In a different work [6], palm shell and coconut shell with lignin contents of 53.4% and 30.1%, respectively, were pyrolyzed at 850 C and attained surface areas of 260 m2 g1 and 183 m2 g1. By subsequently activating these chars at 850 C with CO2, coconut shell had exhibited a faster activation rate in comparison to palm shell, due to its higher cellulose and hemicellulose content. It is well known that coconut shell and wood are the most widespread natural precursors for the production of activated carbon, however successful trials have been conducted with wheat straw [7,8], rice husks and rice straw [9e11], sugar beet bagasse and pulp [12,13], as well as with sugarcane bagasse [14,15]. These agricultural residues are similar in that their lignin content is considerably lower than that of nut shells, fruit stones, or wood. Another characteristic of these biomasses is the low surface areas and porosity that develop during the pyrolysis process. Nanda et al. [16] studied Timothy grass residue and wheat straw residue for the production of biochar for soil amending. They pyrolyzed the substrates at 600 C for 4 h to obtain surface areas and total pore volumes smaller than 5 m2 g1 and 7 mm3 g1, respectively. Rice straw biochar was carbonized by Hammes et al. [17] in a staged pyrolysis process with a maximum temperature of 450 C and residence time of 5 h to obtain a surface area of 5.9 m2 g1. Bornemann et al. [18] pyrolyzed Phalaris grass at different temperatures and determined that surface area increased with temperature. Independent of the carbonization temperature, the charcoals did not develop a measurable microporosity. The low surface areas shown for grassy biomasses like rice or wheat straw are usually related to their inherently high ash content [19]. Nevertheless, high surface areas can be achieved by leaching out the minerals as shown by Alvarez et al. [20]. They obtained a rice husk char with a surface area of 487 m2 g1 and activated carbons with surface areas larger than 800 m2 g1. Guo et al. [21] also leached the minerals from rice husk and used the organic residues to chemically produce activated carbon with KOH, obtaining surface areas larger than 1000 m2 g1. Their finding supports the idea that minerals in biomass may hinder the surface area development. Even so, the presence of an effect from high hemicellulose and cellulose contents in grassy biomasses should not be overlooked. The question remains as to what extent do the components of biomass play a role during activation? The reaction steps during the activation of biomass with H3PO4 have been studied to some extent. Jagtoyen and Derbyshire [22] proposed that H3PO4 stabilizes cellulose and inhibits the formation of levoglucosan through a mechanism during which phosphate esters form by the phosphorylation of cellulose. Guo et al. [23] examined the influence of temperature, impregnation ratio, and of the precursor on the formation and type of surface groups that can be formed during the activation of biomass with H3PO4 by means of Boehm titration. This technique helps to quantify acid surface groups by exposing the activated carbon to three different basic solutions: NaOH is the strongest base and neutralizes all Brønsted acids (phenols, lactonic groups and carboxylic acids), Na2CO3 neutralizes lactonic and carboxylic groups, and NaHCO3 reacts only with carboxylic acids [24,25]. The surface group concentrations are quantified from the difference between the base consumption. In the study by Guo et al. [23], cellulose, xylan and kraft lignin were used as model compounds for biomass. They observed that xylan and cellulose were significantly more reactive towards H3PO4 than lignin, therefore they presented a higher volume of mesopores. They determined as well that the surface groups are either temperature-sensitive, formed due to the low temperature hydrolysis of the raw

precursors in acidic surroundings, or temperature-insensitive, which are phosphorous containing compounds formed after the reaction of the raw material and H3PO4. Furthermore, they also concluded that all the carbons contained the same type of functional groups, however the abundance of each group type depended on the precursor. Acidic groups, especially strong and weak acidic groups, were present in higher concentrations in cellulose activated carbon than in kraft lignin. Due to the formation of highly developed porous structures, many studies concentrate on activation with alkali compounds, especially with KOH. The process begins as a solid-solid reaction, followed by a liquid-solid reaction at higher temperatures. During the heating stage, potassium penetrates the carbon layers creating a lamellar structure of carbon-potassium-carbon layers (intercaladenas et al. [27e29] proposed a mechanism for tion) [26]. Lillo-Ro the interaction between KOH and carbon (Equation (1)), however this reaction cannot be thermodynamically proven at low temperatures. On the other hand, at temperatures higher than 570 C, the Gibbs energy becomes negative. Especially during activation processes under nitrogen, hydrogen is diluted dropping its partial pressure to less than 0.1 MPa, therefore the Gibbs energy can become negative at even lower temperatures [30]. Nevertheless, it is clear that the alkali salts react directly with the carbon that results from the carbonization process and not with the substrate [31]. This explains why it is more effective to conduct a two-step (carbonization followed by the activation) rather than a one-step activation.

6KOH þ 2C42K þ 3H2 þ 2K2 CO3

(1)

Khezami et al. [32] studied the feasibility of producing activated carbon from biomass model compounds and obtained adsorbents that could remove pollutants from aqueous as well as from gaseous sources. They measured high surface areas, especially for carbonized xylan, followed by wood, cellulose and finally lignin. It must be highlighted that biomasses with high lignin content are the most promising precursors for the production of activated carbon (e.g. coconut shell) and the reason is not only the high yields, but also the large surface area that is attainable. The results presented by the group of Khezami show the complete opposite: in this case lignin had developed the poorest surface area compared to the other biomass main components. The authors of the work presented here believe that the reason is the performance of the carbonization (first) step previous to the actual chemical activation with KOH. The precursor materials were carbonized at 300 C, which is very low temperature to allow the formation of a fully carbonized structure. Xylan and cellulose decompose in a narrow temperature range of 220e315 C and 315e400 C, respectively. On the other hand, lignin decomposes slowly and over a wide temperature range between 100 and 900 C. At 300 C, only xylan undergoes a relatively complete decomposition and therefore has a stable and carbon-rich structure [33]. The second step of the activation conducted by Khezami et al. [32] occurred at 700 C in the presence of KOH, however this activation wasn't solely an activation but also a further carbonization of undecomposed cellulose and lignin. This “mixed” set of reactions may have affected the pore development of the activated carbons produced by this group. By taking this into account, the aim of this present research was to study the biomass components with complete carbonization (T ¼ 600 C) prior to activation. This temperature was not high enough to allow a full decomposition of the kraft lignin, however this did not prevent the formation of a wide surface area as well as a highly microporous structure. With regards to climate change, increasing greenhouse emissions, and the pollution caused by inadequate treatment of


agricultural waste, it is necessary to find an alternative way to produce sustainable activated carbon. By understanding the role that each biomass component has on the determination of the activated carbon properties, there is an opportunity to make use of more precursor materials, even if their lignin content is poor. 2. Materials and methods 2.1. Raw materials Alpha cellulose (Sigma Aldrich; CAS: 9004-34-6), xylan from beech wood (Sigma Aldrich; CAS: 9014-63-5), and alkali kraft lignin (Sigma Aldrich; CAS: 8068-05-1) were used as model substances for biomass, and were studied separately as well as mixed with each other in different ratios: 1:0:0, 0.5:0.5:0, and 0.33:0.33:0.33. Alkali kraft lignin was chosen due to its low ash content, since it has been shown that mineral matter can have a negative effect on the adsorption properties and porosity development of activated carbon [19,34]. 2.2. Pyrolysis Each sample was pyrolyzed individually in a batch reactor placed inside a muffle furnace and constantly purged with gaseous nitrogen (13 L min1 at STP (298 K, 101.3 kPa)) to ensure an inert atmosphere. The temperature was increased to 600 C with a constant heating rate of 10 K min1 and maintained for 120 min at 600 C. After the reaction, the carbonized substances were rapidly quenched using gaseous nitrogen to abruptly stop the pyrolysis reaction and cool the sample down. After each experiment, the mass loss was measured and recorded for comparison, and the products were analyzed for their elemental composition. 2.3. Activation The activation reactions were carried out chemically in nickel crucibles and under an inert atmosphere provided by constantly purging with gaseous nitrogen (13 L min1 at STP (298 K, 101.3 kPa)). The activating agent was KOH (Sigma Aldrich; CAS: 1310-58-3) and the weight ratio char:KOH was 1:4. The mixtures were constantly heated up to 600 C with a heating rate of 10 K min1 and the reaction time was 2 h. Afterwards, the reactor was rapidly quenched with gaseous nitrogen to abruptly stop the reaction and to avoid spontaneous combustion. Subsequently, the product was mixed with HCl (2 mol L1) to ensure that all unreacted KOH or any other possible potassium compounds were removed. The samples were then washed with deionized water until the electrical conductivity was lower than 30 mS cm1. The activated carbons were dried at 105 C for at least 16 h and stored for further analysis. 2.4. Characterization of the feedstock, chars and activated carbons 2.4.1. Thermogravimetric analysis Thermogravimetric analyses were conducted in a Netzsch STA Jupiter 449 F5. The sample size was approximately 15 mg and it was heated up to 800 C with a constant heating rate of 10 K min1 under a constant nitrogen flow of 70 ml min1. Cellulose, xylan and kraft lignin were available as powder, therefore the particle size was left unmodified. The pyrolysis carbons were ground in a ball mill before the activation to a particle size smaller than 500 mm. 2.4.2. Adsorption capacity The adsorption capacity of the activated carbons was determined by means of methylene blue adsorption (MB) and iodine

55

number (I2 number). The first was conducted by mixing 0.1 g of each activated carbon with 100 cm3 of a 750 mg l1 methylene blue solution. The mixture was stirred for 48 h to ensure that equilibrium was reached. After this time, the solution was filtered to remove the carbon and the liquid phase concentration was analyzed with a Hach-Lange photometer set to a wavelength l ¼ 664 nm. The maximum adsorption of methylene blue was calculated using Equation (2), where Co is the initial concentration of the methylene blue solution, Ce is the solution concentration after equilibrium is reached, V is the solution volume, and W is the amount of activated carbon used for the analysis:

qe ¼

ðCo Ce Þ*V W

(2)

The I2 numbers were determined according to the ASTM norm D4607-94 [35]. 2.4.3. Infrared spectroscopy and scanning electron microscopy The surface functionality groups were analyzed by Fourier Transformed Infrared Spectroscopy (FTIR) using the KBr technique. The experiments were conducted in a Varian 660-IR spectrometer in the wave number range of 4000e500 cm1. The porosity development was also studied with a field emission scanning electron microscope (FE-SEM) type DSM 982 Gemini (Carl Zeiss Ltd., Oberkochen, Germany) equipped with an annular high brightness in-lens-SE detector for high resolution and true surface imaging. A laterally mounted SE detector (EverhartThornley-type) provides topographical contrast (sensitive for SE þ BSE). The beam accelerating voltage was 10 kV. 2.4.4. Surface area determination with N2 and CO2 adsorption isotherms The specific surface areas were calculated applying the BET model to the N2 and CO2 adsorption isotherms in a non-classical relative pressure range (0.001e0.01 p/p0) using the Rouquerol plot [36], whereas the micropore areas were calculated in the range of 0.2e0.5 p/p0 with a t-plot. Both results showed a correlation coefficient r2 > 0.99. In the case of CO2-isotherms the p0 of CO2 was set at 26,142 torr. The N2 isotherms were measured in a Quantachrome NOVA 1200 at 196 C and for CO2 a Quantachrome NOVA 2000e at 0 C was employed. Prior to the analysis, the activated carbon samples were degassed (Master prep MP-2-220, Quantachrome) at 250 C under high vacuum (1 mPa) using a controlled temperature program over a period of 20 h following the program presented in Table 1. 3. Results and discussion 3.1. Pyrolysis The decomposition of the biomass components under pyrolytic conditions has been widely studied. Most publications show that

Table 1 Controlled temperature program under high vacuum (1 mPa) for sample conditioning previous to the isotherm measurement. Temperature ( C)

Duration (min)

Rate (K min1)

35 50 80 100 130 200 250

30 30 36 45 120 240 600

1 1 1 1 1 2 2

56


cellulose has the highest decomposition rate and a low carbon production in contrast to lignin, which decomposes over a wide temperature range but with low rates and high yields. Xylan decomposes slower than cellulose but faster than lignin, and it yields a relatively low amount of carbon [33,37]. The parameters employed in this work led to a carbon production in mass fraction of 16.6%, 24.8% and 47.0% for cellulose, xylan, and lignin, respectively (Fig. 1). The carbon yields of the mixtures showed that there was no interaction between the components during the pyrolysis process (Fig. 2), as the yields of the mixtures equal the sum of yields of the single components. The “calculated yield” given as mass fraction (Equation (3)) was estimated by multiplying the carbon yield of the original substance (yi) by the proportion in which this substance is present in the mixture (xi):

ycalc ¼ xC yC þ xx yx þ xL yL

(3)

The lack of interaction can also be observed in the decomposition rate profiles of the mixtures, where the dm/dT curves of the pure substances are superimposed (Fig. 3). The decomposition rates were slightly slower than the rates of the pure substances due to limited heat transfer phenomena caused by the early char formation of xylan and/or cellulose. This is also found in the studies by Raveendran et al. [38], Stefanidis et al. [39] and Yang et al. [40]. The thermogravimetric analysis (TGA) of the chars (Fig. S1 e supplementary material) show how xylan and lignin continue to decompose at temperatures higher than 600 C due to decomposition of remaining organics or gasification of the chars, according to the Boudouard reaction C þ CO2 / CO. The decomposition of cellulose ceased at temperatures lower than 450 C, meaning that an almost complete decomposition of cellulose was achieved at the carbonization temperature used in this work. This was observed not only from the mass loss but also from the decomposition rate, which reached a maximum of 0.02% C1 at approximately 750 C. On the other hand, the decomposition rates of xylan and lignin were 0.03% C1 and 0.08% C1 at 655 C and 740 C, respectively. As expected, there was a strong reduction of the oxygen and hydrogen content during the carbonization stage due to decomposition reactions that generate H2O, CO2, CO, CH4, H2 and shortchain hydrocarbons (volatiles) as well as condensable hydrocarbons (Table 2) [41]. The ash content increased due to its relatively inert character under pyrolytic conditions. Fig. 4 shows SEM images of the char samples. Seen in these images, cellulose particles are smooth and show no evidence of

Fig. 1. Pyrolysis char mass fraction yield; the error bars represent the standard deviation from at least three repetitions of the pyrolysis process (C ¼ cellulose; X ¼ xylan; L ¼ kraft lignin).

Fig. 2. The correlation between calculated vs. measured yields of the pyrolysis chars (PC) shows no interaction between the components during the carbonization process.

porosity. In contrast, xylan and lignin show a textured surface with a developed broad macroporosity and “bubbles” which were formed during pyrolysis.

3.2. Activation The cellulose and lignin charcoals had a similar resilience against KOH in contrast to xylan (Fig. 5). This explains the high activated carbon yields obtained from these substances, when the yield is calculated based on the initial amount of charcoal. This is also valid for the mixture of lignin and cellulose. This behavior was also observed by Guo et al. [23] independently of the activation temperature after activating xylan, cellulose, and kraft lignin with H3PO4. In contrast, when the yield is calculated based on the initial amount of raw sample, lignin definitely produces more activated carbon (Fig. 5). This is consistent with the theory [42] and with the TGA analysis results, which showed high thermal stability of lignin (Fig. 3) and lignin char (Fig. S1). A lack of interaction between mixture components was also identified after comparing the yields resulting from the activation after obtaining a linear correlation between calculated (Equation (3)) and measured yields (Fig. 6). This was noticed especially when the yield is calculated based on the initial biomass. Cagnon et al. [2] arrived to a similar conclusion after studying the contributions of the biomass main components on the properties of steam activated carbons. Regarding the elemental composition of the activated carbons, there was not only a decrease of the carbon content (gasification reactions), but also an increase in oxygen due to oxidation reactions (Table 2). Khezami et al. [32] observed the presence of acidic groups after performing a temperature programmed desorption on chemically activated carbons from biomass model substances. All their samples evidenced pronounced CO and CO2 peaks corresponding to oxygen-containing surface functional groups such as carboxylic acids, lactones, or phenols. The slight reduction of nitrogen content present after the activation could be explained due to the formation of potassium cyanide during the alkaline activanchez et al. [43,44] explain tion as reported by Ref. [43]. Robau-Sa the cyanide formation as a result of using nitrogen as a reaction medium, however the participation of the gaseous nitrogen needs the presence of a catalyst like iron to promote the formation of cyanide. Otherwise, the nitrogen in KCN originates from the char.


57

Fig. 3. Decomposition rate profiles (dm/dT) of the substrates and mixtures; a) cellulose þ xylan mixture with the curves of pure compounds, b) cellulose þ lignin mixture with the curves of pure compounds, c) xylan þ lignin mixture with the curves of pure compounds and d) cellulose þ xylan þ lignin mixture with the curves of the pure compounds. Table 2 Element mass fractions (%) of the parent materials, pyrolysis chars and activated carbons. Oxygen was calculated by difference.

Cellulose (C) Untreated Pyrochar Activated Carbon Xylan (X) Untreated Pyrochar Activated Carbon Lignin (L) Untreated Pyrochar Activated Carbon CþX Pyrochar Activated Carbon LþX Pyrochar Activated Carbon LþC Pyrochar Activated Carbon CþXþL Pyrochar Activated Carbon

C

H

N

S

ODIF

Ash

44.2 92.6 85.6

6.8 1.7 0.3

0.1 0.2 0.5

0.1 0.2 0.2

48.7 4.8 13.2

0.08 0.54 0.2

43.8 74.4 84.9

6.8 1.4 0.7

0.1 0.4 0.6

0.1 0.1 0.3

42.3 5.1 12.6

6.95 18.6 0.92

64.4 86.4 86.0

5.8 1.7 0.9

0.3 0.8 0.6

1.4 1.1 0.2

25.5 4.7 11.7

2.62 5.32 0.6

80.2 87.0

1.5 0.6

0.3 0.6

0.1 0.1

8.7 11.5

9.22 0.17

82.8 76.9

1.6 0.7

0.6 0.5

0.7 0.1

5.8 21.4

8.47 0.41

87.6 87.1

1.8 0.7

0.7 0.5

0.7 0.1

5.3 11.5

3.88 0.07

85.2 87.8

1.6 0.7

0.6 0.6

0.6 0.1

5.2 10.7

6.84 0.13

The drastic decrease of the ash content can be explained by the formation of potassium silicates or aluminum silicates [45].

3.3. Surface area and adsorption properties The use of N2 to determine surface areas is a common technique found in literature, however it is not the ideal adsorptive to measure the microporous surface area. According to Garrido et al. [46], at 196 C it is not energetically favorable for N2 to enter micropores (pore diameter < 2 nm) and instead of measuring the surface area covered by the monolayer, it gives information on the energetic interactions between the adsorbate and the active sites [36]. Oxygen groups present on the carbon surface interact strongly with the N2 molecules due to its intrinsic quadrupole moment [47]. Additionally, the strong adsorption potentials in confined micropores results in a high uptake of the N2 adsorptive, which corresponds to the primary filling of microporous at very low relative pressure [36]. In contrast to N2, an analysis with CO2 addresses both of these issues by reaching equilibrium faster (analysis taking place at higher temperatures) and having a gas uptake at much higher pressure, allowing for a rapid characterization of microporous carbon-like materials [48]. Fig. 7 shows the N2 and CO2 isotherms of the activated carbon from the single substances. The N2 isotherms can be classified as type I isotherms and displayed an almost vertical step from the origin that gradually sloped up to a plateau until 0.5 p/p0. This is consistent with the high micropore volumes presented by all samples (Table 3). Additionally, cellulose was observed to have adsorbed the highest amount of N2, followed by lignin and xylan. This agreed with the respective surface areas of the activated carbons (Fig. 8). The isotherms of the mixtures (Fig. S2 - supplementary material) were in between the isotherm of cellulose and lignin, indicating that mixing other compounds with cellulose had

58


Fig. 4. SEM images in two different magnifications (1000 and 10,000 for a) and b) and 3000 and 10,000 for c) of the pyrolysis carbons from a) cellulose. b) xylan. and c) kraft lignin.

a negative impact on its adsorption potential. Xylan showed the lowest adsorption. The CO2 isotherms from the activated carbons followed a similar trend to the N2 isotherms: cellulose and lignin showed the highest adsorption and xylan the lowest (Fig. 7). Nevertheless, the isotherms of the single substances and those of the mixtures (Fig. S3 - supplementary material) were much closer together, indicating that the micropore size distribution was similar in all cases. The surface areas calculated from the N2 ðSN2 Þ and CO2 ðSCO2 Þ isotherms are shown in Fig. 8. As expected, the values for SCO2 were lower than SN2 by a factor of 0.7e0.8. The highest surface areas were found for cellulose (SN2 : 2221 m2 g1; SCO2 : 1778 m2 g1) and lignin (SN2 : 2093 m2 g1; SCO2 : 1485 m2 g1). In contrast, xylan showed the lowest surface area (SN2 : 1949 m2 g1; SCO2 : 1360 m2 g1) even when compared to the mixtures. Shown in Table 3, the total surface area as well as the micropore area, which comprised up to 91e99% of the total area, were hindered under the presence of more than one component. This could be explained by the different thermal stabilities from the carbons as shown by the

TGA (Fig. S1- supplementary material). Chars remain stable at temperatures lower than 600 C, but temperatures higher than this will incur a slow decomposition, which could result in a pore enlargement or blockage due to re-condensation processes. This decomposition can be accelerated or can happen at lower temperatures in the presence of KOH due to the catalytic character of potassium [49,50]. The I2 number and MB adsorption results (Fig. 9) show a similar trend to the specific surface areas, indicating the applicability of these analytical methods to gain information on surface area and microporosity. Nonetheless, the correlation between these measurements is relatively low (Fig. 10) and it can be inaccurate to derive the exact surface area from either MB or I2 number. Further investigation regarding adsorption thermodynamics and kinetics is required to understand the lack of correlation between these measurements. Regarding the interaction among biomass components, it was observed that activated carbons from lignin and xylan alone showed the lowest adsorption potentials, yet the adsorption


59

Fig. 5. Activated carbon mass fraction yield based on the substrate (light grey) and on the pyrolysis char (dark grey) (C ¼ cellulose; X ¼ xylan; L ¼ kraft lignin).

Fig. 7. N2 (top) and CO2 (bottom) isotherms of the activated carbons from the single substances.

where the measured and calculated values were very similar and in accordance with the model. Fig. 6. Calculated vs. measured mass fraction yields of the activated carbon.

3.4. SEM image analysis potential and surface areas increase when mixed together. This may also occur under the addition of cellulose to the mixture. Similarly, the presence of xylan had a negative effect on surface area development and adsorption properties of cellulose. This leads to the conclusion that biomass components have an influence on each other. To support the previous statement, the adsorption results and specific surface areas were calculated following Equation (3) where, in this case, y represents either MB, I2 number, SN2 or SCO2 (Fig. 11). In the case of MB and SN2 , the measured and calculated values showed a considerably low correlation, which therefore supports the idea of a possible interaction between components. On the other hand, both I2 number and SCO2 showed a relatively high correlation between the measured and calculated values, however not as high as those presented by the yields (Fig. 6). To validate whether the activated carbon properties depend on the mixture composition, Equation (3) was also applied to calculate properties from a beech wood activated carbon produced under equal conditions. The organic constituents of wood were calculated according to the van Soest method [51] and are presented in Table 4. In the case of SN2 and MB, the data for wood are very different to the correlation curve. The contrary occurs in the case of the I2 number,

The porosity development is outlined in the SEM images (Fig. 12). In the case of cellulose, the surface of the pyrolysis char was relatively smooth compared to that of the xylan and lignin particles. This enabled a better contact with KOH, facilitating the formation of micropores. As previously mentioned, lignin develops “bubbles” on its surface during the pyrolysis, which could explain ~ ero its low micropore areas. Using x-ray diffraction, Raymundo-Pin et al. [52] discovered that K2CO3 is formed from KOH at around 400 C and the whole KOH is consumed at about 600 C. In the presence of carbon, it is possible that K2CO3 had decompose into metallic potassium at temperatures lower than its decomposition point (891 C) [28,53]. Metallic (elementary) potassium (Equation (1)) is extremely unstable and can react explosively with different agents, expanding the carbon lattices and creating pores. These are solid/solid and liquid/solid reactions therefore they are dominated by diffusion rates, which take place faster on thin layers (bubbles) rather than in the bulk body of the carbon particle, and therefore expanding the pore size (Fig. 13). Consequently, the micropore formation can occur more efficiently. Compared to lignin and cellulose, the adsorption of xylan char was poor and its surface area was considerably lower. Fig. 12b shows an ampler porosity distribution as well as a broader pore size

60

C. Rodriguez Correa et al. / Biomass and Bioenergy 97 (2017) 53e64 Table 3 Micropore volume, area and ratio between the micropore and total area calculated from the N2 isotherm. Sample

Micropore Volume (cm3 g1)

Micropore Area (m2 g1)

Ratio of Micropore Area/Total Area

Cellulose (C) Xylan (X) Lignin (L) C+X X+L L+C C+X+L

0.86 0.74 0.77 0.85 0.85 0.86 0.83

2116 1826 1912 2114 2107 2136 2076

0.95 0.94 0.91 0.96 0.94 0.94 0.99

Fig. 8. Specific surface areas calculated from N2 isotherm at 77 K ðSN2 Þ and from CO2 isotherm ðSCO2 Þ at 273.15 K (C ¼ cellulose; X ¼ xylan; L ¼ kraft lignin).

Fig. 9. I2 number and maximum methylene blue (MB) adsorption (C ¼ cellulose; X ¼ xylan; L ¼ kraft lignin). The error bars represent the standard deviation of at least three measurements.

diameter. Additionally, the carbon content of the chars obtained from cellulose, lignin and the mixtures containing them were considerably higher than that of xylan chars. This means that there were more sites available for the reaction between KOH and C. 3.5. FTIR and group functionalities on the surface Adsorption processes can be divided into physisorption and chemisorption. FTIR was used to study the functional groups on the surface, especially those containing oxygen, in order to determine which type of adsorption was predominant. Figs. 14 and 15 show a

Fig. 10. Correlation of MB (squares) and I2 number (circles) with SCO2 (black) and SN2 (white).

comparison between the substrate and the activated carbons. In almost all spectra, the bands between 3200 and 3600 cm1 were present. These are typically attributed to the stretching of eOH groups from adsorbed water. Cellulose and hemicellulose (Fig. 14) are polymers composed mainly by linked hexoses. Glucose is the main monomer of cellulose and hemicellulose is a mixture of glucose, xylose, mannose, galactose, arabinose, fructose, glucuronic acid, and galacturonic acid in different proportions depending on the source [54]. Therefore, it was possible to observe a strong stretching of C]O bonds, the stretching and deformation of CeOeC bonds, as well as of CeOe(H) at 1636, 1246, and 1033 cm1 for cellulose. For xylan, the corresponding bands were 1616, 1252 and 1044 cm1. Fig. 14 also shows the characteristic bands of lignin located at 1426, 1463, 1513, and 1595 cm1, which indicated the existence of aromatic rings and CeH bonds. The appearance of bands at 1269 and 1368 cm1 indicated that guaiacyl and syringyl groups were present. According to the findings of Liu et al. [55], this particular lignin could have been extracted from some type of hardwood. Compared to xylan and cellulose, the stretching of the “fingerprint” bands of lignin (500e1595 cm1) may have been due to compounds containing methoxyleOeCH3, CeOeC stretching, and C]C stretching of aromatic ring as reported by Yang et al. [33]. The FTIR spectra of the activated carbons show notably less bands, which indicates that during the carbonization and activation, most of the surface groups had decomposed. This implies that during an adsorption process, the adsorbate was reversibly attached to the surface by van der Waals forces (physisorption) instead of chemically bound (chemisorption). The activated carbons from cellulose and lignin also showed the band corresponding to the stretching of hydroxyl groups at around 3350 cm1. This band was absent in the activated carbon made from xylan. Since this band is commonly attributed to adsorbed water, it can be said the activated carbons from xylan have a higher hydrophobic character


61

Fig. 11. Calculated vs. measured properties by considering the property of the mixture as the sum of the properties of the single components; a) I2 number, b) MB adsorption, c) SCO2 and d) SN2 . The black circles correspond to the results of the model substances and the empty squares to the results of beech wood.

Table 4 Characterization of beech wood composition (mass fraction %) obtained with the van Soest method (ADF, ADL, NDF) [47] as well as adsorption and textural properties of the activated carbon produced chemically with KOH from beech wood. Cellulose (ADF - ADL) Hemicellulose (NDF - ADF) Lignin (ADL) Methylene blue adsorption (mg g1) Iodine number (mg g1) SBET;N2 ðm2 g1 Þ

57.8 19.4 14.7 747 1666 2483

in contrast to those from cellulose and lignin. Especially in the spectra of the activated carbons obtained from lignin and cellulose, few weak bands were present between 3400 and 1500 cm1, which can be explained by a small concentration of residual oxygencontaining groups (ketones, aldehydes, lactones or carboxyl groups). The presence of a small concentration or even the complete lack of surface groups may be credited to the high activation temperature that was used. Similar observations were presented by Guo et al. [23], with results from the Boehm titration of activated carbons produced at 500 C. Moreover, all activated carbons show bands around 1600 cm1 and 1080 cm1, indicating the stretching of C]C bonds ascribed to aromatic carbon groups. Fig. 15 presents a comparison of the FTIR spectra corresponding to the activated carbons of the different mixtures. It is noteworthy to mention that the activated carbons spectra from mixtures that contain xylan show similar peaks to either lignin or cellulose with the exception of the mixture cellulose and xylan. From the

spectrum of this sample, no functional groups could be detected due to the absence of strong transmission bands. 4. Conclusions During the pyrolysis step, the components had decomposed independently without detectable mutual interaction. However, thermally stable behavior of the components was only observed at temperatures lower than 600 C (reaction temperature). Regarding the activation step of the model compounds, the highest surface area and adsorption potential was found for cellulose, followed closely by lignin. Xylan showed the lowest surface areas, microporosity, and adsorption potential. It was observed that when cellulose was mixed with other components, its textural and adsorption properties had decreased considerably, especially when mixed with xylan. This can be explained by the strong thermal instability shown by xylan pyrolysis char, which could lead to pore expansion or blockage. CO2 isotherms indicated that the ultra microporosity might not be as influenced by the mixture composition. Nonetheless, calculations applying density functional theory are necessary for a better assessment. Additionally, no correlation was found when comparing measured and calculated activated carbon properties (e.g. surface areas or methylene blue adsorption potential). This indicates that the presence of more than one component had altered the properties of the final product. This, in turn, could mean that the activated carbon properties depend strongly on the composition of the initial feedstock. Finally, the FTIR diagrams showed little to no functional groups on the surfaces of the activated carbon, meaning

62


Fig. 12. SEM images at two different magnifications (30,000 and 75,000) of activated carbon from a) cellulose, b) xylan, c) alkali kraft lignin, and d) the mixture of alkali kraft lignin and cellulose.

KOH

KOH

Lignin “bubble” Fig. 13. Scheme describing the porosity formation of lignin.

Pore


63

Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.biombioe.2016.12.017. References

Fig. 14. FTIR spectra of the parent materials.

Fig. 15. FTIR spectra of the activated carbons from the single substances and their mixtures (C ¼ cellulose; X ¼ xylan; L ¼ kraft lignin).

that the iodine and methylene blue adsorption occur mainly due to physisorption phenomena.

Acknowledgements The author would like to thank Ms. Doreen Neumann-Walter and Ms. Yvonne Gil-Pascual for their help with the BET measurements, as well as to Mr. Wilhelm Habicht for the SEM images. Also, especial thanks to Mr. Jonas Sage for the production of the activated carbons and for performing the adsorption tests and to Mr. Daniel Smart and Ms. Hannah Oliphant for proof reading this work.

[1] P.J.M. Suhas, M.M.L. Carrott, Ribeiro Carrott, Lignin - from natural adsorbent to activated carbon: a review, Bioresour. Technol. 98 (2007) 2301e2312, http:// dx.doi.org/10.1016/j.biortech.2006.08.008. [2] B. Cagnon, X. Py, A. Guillot, F. Stoeckli, G. Chambat, Contributions of hemicellulose, cellulose and lignin to the mass and the porous properties of chars and steam activated carbons from various lignocellulosic precursors, Bioresour. Technol. 100 (2009) 292e298, http://dx.doi.org/10.1016/ j.biortech.2008.06.009. [3] J. Guo, A. Chong Lua, Characterization of chars pyrolyzed from oil palm stones for the preparation of activated carbons, J. Anal. Appl. Pyrol. 46 (1998) 113e125, http://dx.doi.org/10.1016/S0165-2370(98)00074-6. [4] J. Guo, A.C. Lua, Effect of heating temperature on the properties of chars and activated carbons prepared from oil palm stones, J. Therm. Anal. Calorim. 60 (2000) 417e425, http://dx.doi.org/10.1023/A:1010137308378. [5] J. Guo, A.C. Lua, Textural and chemical characterisations of activated carbon prepared from oil-palm stone with H2SO4 and KOH impregnation, Microporous Mesoporous Mater. 32 (1999) 111e117, http://dx.doi.org/10.1016/ S1387-1811(99)00096-7. [6] W.M.A.W. Daud, W.S.W. Ali, Comparison on pore development of activated carbon produced from palm shell and coconut shell, Bioresour. Technol. 93 (2004) 63e69, http://dx.doi.org/10.1016/j.biortech.2003.09.015. [7] X. Li, C. Han, X. Chen, C. Shi, Preparation and performance of straw based activated carbon for supercapacitor in non-aqueous electrolytes, Microporous Mesoporous Mater. 131 (2010) 303e309, http://dx.doi.org/10.1016/ j.micromeso.2010.01.007. [8] T. Robinson, B. Chandran, P. Nigam, Effect of pretreatments of three waste residues, wheat straw, corncobs and barley husks on dye adsorption, Bioresour. Technol. 85 (2002) 119e124, http://dx.doi.org/10.1016/S09608524(02)00099-8. [9] D. Kalderis, S. Bethanis, P. Paraskeva, E. Diamadopoulos, Production of activated carbon from bagasse and rice husk by a single-stage chemical activation method at low retention times, Bioresour. Technol. 99 (2008) 6809e6816, http://dx.doi.org/10.1016/j.biortech.2008.01.041. [10] A.H. Basta, V. Fierro, H. Saied, A. Celzard, Effect of deashing rice straws on their derived activated carbons produced by phosphoric acid activation, Biomass Bioenergy 35 (2011) 1954e1959, http://dx.doi.org/10.1016/ j.biombioe.2011.01.043. [11] Y. Chen, Y. Zhu, Z. Wang, Y. Li, L. Wang, L. Ding, et al., Application studies of activated carbon derived from rice husks produced by chemical-thermal process - a review, Adv. Colloid Interface Sci. 163 (2011) 39e52, http:// dx.doi.org/10.1016/j.cis.2011.01.006. [12] H. Demiral, G. Gündüzoǧlu, Removal of nitrate from aqueous solutions by activated carbon prepared from sugar beet bagasse, Bioresour. Technol. 101 (2010) 1675e1680, http://dx.doi.org/10.1016/j.biortech.2009.09.087. [13] H.L. Mudoga, H. Yucel, N.S. Kincal, Decolorization of sugar syrups using commercial and sugar beet pulp based activated carbons, Bioresour. Technol. 99 (2008) 3528e3533, http://dx.doi.org/10.1016/j.biortech.2007.07.058. [14] K.J. Cronje, K. Chetty, M. Carsky, J.N. Sahu, B.C. Meikap, Optimization of chromium(VI) sorption potential using developed activated carbon from sugarcane bagasse with chemical activation by zinc chloride, Desalination 275 (2011) 276e284, http://dx.doi.org/10.1016/j.desal.2011.03.019. [15] B.S. Girgis, L.B. Khalil, T.A.M. Tawfik, Activated carbon from sugar cane bagasse by carbonization in the presence of inorganic acids, J. Chem. Technol. Biotechnol. 61 (1994) 87e92, http://dx.doi.org/10.1002/jctb.280610113. [16] S. Nanda, R. Azargohar, J.A. Kozinski, A.K. Dalai, Characteristic studies on the pyrolysis products from hydrolyzed Canadian lignocellulosic feedstocks, BioEnergy Res. 7 (2013) 174e191, http://dx.doi.org/10.1007/s12155-013-9359-7. [17] K. Hammes, R.J. Smernik, J.O. Skjemstad, A. Herzog, U.F. Vogt, M.W.I. Schmidt, Synthesis and characterisation of laboratory-charred grass straw (Oryza sativa) and chestnut wood (Castanea sativa) as reference materials for black carbon quantification, Org. Geochem. 37 (2006) 1629e1633, http://dx.doi.org/ 10.1016/j.orggeochem.2006.07.003. [18] L.C. Bornemann, R.S. Kookana, G. Welp, Differential sorption behaviour of aromatic hydrocarbons on charcoals prepared at different temperatures from grass and wood, Chemosphere 67 (2007) 1033e1042, http://dx.doi.org/ 10.1016/j.chemosphere.2006.10.052. [19] A.H. Basta, V. Fierro, H. El-Saied, A. Celzard, 2-Steps KOH activation of rice straw: an efficient method for preparing high-performance activated carbons, Bioresour. Technol. 100 (2009) 3941e3947, http://dx.doi.org/10.1016/ j.biortech.2009.02.028. [20] J. Alvarez, G. Lopez, M. Amutio, J. Bilbao, M. Olazar, Physical activation of rice husk pyrolysis char for the production of high surface area activated carbons, Ind. Eng. Chem. Res. 54 (2015), http://dx.doi.org/10.1021/acs.iecr.5b01589, 150702091000005. [21] D. An, Y. Guo, B. Zou, Y. Zhu, Z. Wang, A study on the consecutive preparation of silica powders and active carbon from rice husk ash, Biomass Bioenergy 35

64


(2011) 1227e1234, http://dx.doi.org/10.1016/j.biombioe.2010.12.014. [22] M. Jagtoyen, F. Derbyshire, Activated carbons from yellow poplar and white oak by H3PO4 activation, Carbon N. Y. 36 (1998) 1085e1097, http://dx.doi.org/ 10.1016/S0008-6223(98)00082-7. [23] Y. Guo, D.A. Rockstraw, Physical and chemical properties of carbons synthesized from xylan, cellulose, and kraft lignin by H3PO4 activation, Carbon N. Y. 44 (2006) 1464e1475, http://dx.doi.org/10.1016/j.carbon.2005.12.002. [24] H.P. Boehm, E. Diehl, W. Heck, R. Sappok, Surface oxides of carbon, Angew. Chem. Int. Ed. Engl. 3 (1964) 669e677, http://dx.doi.org/10.1002/ anie.196406691. [25] H.P. Boehm, Some aspects of the surface chemistry of carbon blacks and other carbons, Carbon N. Y. 32 (1994) 759e769, http://dx.doi.org/10.1016/00086223(94)90031-0. [26] J. Romanos, M. Beckner, T. Rash, L. Firlej, B. Kuchta, P. Yu, et al., Nanospace engineering of KOH activated carbon, Nanotechnology 23 (2011) 15401, http://dx.doi.org/10.1088/0957-4484/23/1/015401. denas, J. Juan-Juan, D. Cazorla-Amoro s, A. Linares-Solano, About [27] M.A. Lillo-Ro reactions occurring during chemical activation with hydroxides, Carbon N. Y. 42 (2004) 1365e1369, http://dx.doi.org/10.1016/j.carbon.2004.01.008. denas, D. Cazorla-Amoro s, A. Linares-Solano, Understanding [28] M. Lillo-Ro chemical reactions between carbons and NaOH and KOH: an insight into the chemical activation mechanism, Carbon N. Y. 41 (2003) 267e275, http:// dx.doi.org/10.1016/S0008-6223(02)00279-8. , J.M. Calo, D. Cazorla-Amoro s, A. Linares-Solano, Carbon [29] D. Lozano-Castello activation with KOH as explored by temperature programmed techniques, and the effects of hydrogen, Carbon N. Y. 45 (2007) 2529e2536, http:// dx.doi.org/10.1016/j.carbon.2007.08.021. [30] J. Wang, S. Kaskel, KOH activation of carbon-based materials for energy storage, J. Mater. Chem. 22 (2012) 23710, http://dx.doi.org/10.1039/ c2jm34066f. [31] H. Marsh, F. Rodríguez-Reinoso, Activated Carbon, Elsevier, 2006. http://dx. doi.org/10.1016/B978-008044463-5/50020-0. [32] L. Khezami, A. Chetouani, B. Taouk, R. Capart, Production and characterisation of activated carbon from wood components in powder: cellulose, lignin, xylan, Powder Technol. 157 (2005) 48e56, http://dx.doi.org/10.1016/ j.powtec.2005.05.009. [33] H. Yang, R. Yan, H. Chen, D.H. Lee, C. Zheng, Characteristics of hemicellulose, cellulose and lignin pyrolysis, Fuel 86 (2007) 1781e1788. [34] F. Ahmad, W.M.A.W. Daud, M.A. Ahmad, R. Radzi, The effects of acid leaching on porosity and surface functional groups of cocoa (Theobroma cacao)-shell based activated carbon, Chem. Eng. Res. Des. 91 (2013) 1028e1038, http:// dx.doi.org/10.1016/j.cherd.2013.01.003. [35] American Society for Testing and Materials, D4607 94 standard test method for determination of iodine number of activated carbon, ASTM Int. 94 (2011) 1e5, http://dx.doi.org/10.1520/D4607-94R11.2. [36] J. Rouquerol, P. Llewellyn, F. Rouquerol, Is the bet equation applicable to microporous adsorbents? Stud. Surf. Sci. Catal. (2007) 49e56, http:// dx.doi.org/10.1016/S0167-2991(07)80008-5. [37] L. Burhenne, J. Messmer, T. Aicher, M.-P. Laborie, The effect of the biomass components lignin, cellulose and hemicellulose on TGA and fixed bed pyrolysis, J. Anal. Appl. Pyrol. 101 (2013) 177e184, http://dx.doi.org/10.1016/ j.jaap.2013.01.012. [38] K. Raveendran, Pyrolysis characteristics of biomass and biomass components, Fuel 75 (1996) 987e998, http://dx.doi.org/10.1016/0016-2361(96)00030-0. [39] S.D. Stefanidis, K.G. Kalogiannis, E.F. Iliopoulou, C.M. Michailof, P.A. Pilavachi, A.A. Lappas, A study of lignocellulosic biomass pyrolysis via the pyrolysis of cellulose, hemicellulose and lignin, J. Anal. Appl. Pyrol. 105 (2014) 143e150, http://dx.doi.org/10.1016/j.jaap.2013.10.013.

[40] H. Yang, R. Yan, H. Chen, C. Zheng, D.H. Lee, D.T. Liang, In-depth investigation of biomass pyrolysis based on three major components: hemicellulose, cellulose and lignin, Energy Fuels 20 (2006) 388e393, http://dx.doi.org/10.1021/ ef0580117. [41] C. Di Blasi, G. Signorelli, C. Di Russo, G. Rea, Product distribution from pyrolysis of wood and agricultural residues, Ind. Eng. Chem. Res. 38 (1999) 2216e2224, http://dx.doi.org/10.1021/ie980711u. [42] T. Faravelli, A. Frassoldati, G. Migliavacca, E. Ranzi, Detailed kinetic modeling of the thermal degradation of lignins, Biomass Bioenergy 34 (2010) 290e301, http://dx.doi.org/10.1016/j.biombioe.2009.10.018. n, M.A. Lillo-Ro denas, M.A. Montes-Mor [43] E. Fuente, R.R. Gil, R.P. Giro an, M.J. Martin, et al., Evidence for the presence of cyanide during carbon activation by KOH, Carbon N. Y. 48 (2010) 1032e1037, http://dx.doi.org/10.1016/ j.carbon.2009.11.022. nchez, A. Aguilar-Elgue zabal, J. Aguilar-Pliego, Chemical activation [44] A. Robau-Sa of Quercus agrifolia char using KOH: evidence of cyanide presence, Microporous Mesoporous Mater. 85 (2005) 331e339, http://dx.doi.org/10.1016/ j.micromeso.2005.07.003. [45] D.W. Mckee, C.L. Spiro, K.E.J.P.G. Lamby, Catalytic effects of alkali metal salts in the gasification of coal char, Symp. Coal Gasif. (1982) 74e86. http://www.osti. gov/scitech/biblio/5568073 (Accessed 5 May 2016). [46] J. Garrido, A. Linares-Solano, J.M. Martin-Martinez, M. Molina-Sabio, F. Rodriguez-Reinoso, R. Torregrosa, Use of N2 vs. CO2 in the characterization of activated carbons, Langmuir 3 (1987) 76e81, http://dx.doi.org/10.1021/ la00073a013. [47] J. Pikunic, P. Llewellyn, R. Pellenq, K.E. Gubbins, Argon and nitrogen adsorption in disordered nanoporous carbons: simulation and experiment, Langmuir 21 (2005) 4431e4440, http://dx.doi.org/10.1021/la047165w. [48] H. Marsh, W.F.K. Wynne-Jones, The surface properties of carbon-I the effect of activated diffusion in the determination of surface area, Carbon N. Y. 1 (1964) 269e279, http://dx.doi.org/10.1016/0008-6223(64)90281-7. [49] N. Mims, , Exxon Research and Engineering Co., C.A. Linden, J.K. Pabst, Alkalicatalyzed carbon gasification. I. Nature of the catalytic sites, Am. Chem. Soc. Div. Fuel Chem. Prepr. (United States) 25 (3) (1980). http://www.osti.gov/ scitech/biblio/6735971 (Accessed 23 February, 2016). [50] C.A. Mims, J.K. Pabst, Alkali catalyzed carbon gasification - 2. Kinetics and mechanism, ACS Div. Fuel Chem. Prepr. 25 (1980) 263e268. http://www.osti. gov/scitech/biblio/7021434 (Accessed 23 February, 2016). [51] P.J. Van Soest, J.B. Robertson, B.A. Lewis, Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition, J. Dairy Sci. 74 (1991) 3583e3597, http://dx.doi.org/10.3168/jds.S00220302(91)78551-2. ~ ero, P. Azais, T. Cacciaguerra, D. Cazorla-Amoro s, A. Linares[52] E. Raymundo-Pin guin, KOH and NaOH activation mechanisms of multiwalled Solano, F. Be carbon nanotubes with different structural organisation, Carbon N. Y. 43 (2005) 786e795, http://dx.doi.org/10.1016/j.carbon.2004.11.005. [53] F. Kapteijn, G. Abbel, J.A. Moulijn, CO2 gasification of carbon catalysed by alkali metals. Reactivity and mechanism, Fuel 63 (1984) 1036e1042, http:// dx.doi.org/10.1016/0016-2361(84)90184-4. , T. Heinze, Xylan and xylan derivatives - biopolymers with [54] A. Ebringerova valuable properties, 1. Naturally occurring xylans structures, isolation procedures and properties, Macromol. Rapid Commun. 21 (2000) 542e556, http://dx.doi.org/10.1002/1521-3927(20000601)21:93.0.CO;2-7. [55] Q. Liu, S. Wang, Y. Zheng, Z. Luo, K. Cen, Mechanism study of wood lignin pyrolysis by using TGeFTIR analysis, J. Anal. Appl. Pyrol. 82 (2008) 170e177, http://dx.doi.org/10.1016/j.jaap.2008.03.007.