Abstract
This work focused on the solution-sensitivity and the comprehensive mechanism of lignin breakdown during the phosphoric acid-acetone pretreatment process using density functional theory calculations. The structures and properties of a-O-4 lignin, β-5-3 lignin, and β-βlignin were detected, which showed that the bond length follows the order: a-O-4 bond < β-5-3 bond < β-βbond, but a-O-4 lignin is more sensitive to solvent molecule thanβ–β lignin and β-5-3 lignin. The decomposition mechanism of a-O-4 lignin, β-5-3 lignin, and β–β lignin in different solutions showed that a-O-4 lignin decomposes much more easily than β-5-3 lignin and β–β lignin, acting as the most accessible and susceptible point of lignin. Further, the selectivity of decomposition of lignin depends markedly on the synergy of solution and position. The physical origin of the structure-selectivity of lignin in different solution can be rationalized in terms of both thermodynamics and kinetics. The reactions investigated in this work constitute a large database for understanding the chemistry of a-O-4 lignin, β-5-3 lignin, and β-βlignin, and their decomposition in different solutions.
Download PDF
Full Article
Solution-Sensitivity and Comprehensive Mechanism of Lignin Breakdown during the Phosphoric Acid-Acetone Pretreatment Process
Wu Qin,* Zong-Ming Zheng,* Peng Kang, Changqing Dong, and Yongping Yang
This work focused on the solution-sensitivity and the comprehensive mechanism of lignin breakdown during the phosphoric acid-acetone pretreatment process using density functional theory calculations. The structures and properties of -O-4 lignin, β-5-3 lignin, and β-β lignin were detected, which showed that the bond length follows the order: -O-4 bond < β-5-3 bond < β-β bond, but -O-4 lignin is more sensitive to solvent molecule thanβ–β lignin and β-5-3 lignin. The decomposition mechanism of -O-4 lignin, β-5-3 lignin, and β–β lignin in different solutions showed that -O-4 lignin decomposes much more easily than β-5-3 lignin and β–β lignin, acting as the most accessible and susceptible point of lignin. Further, the selectivity of decomposition of lignin depends markedly on the synergy of solution and position. The physical origin of the structure-selectivity of lignin in different solution can be rationalized in terms of both thermodynamics and kinetics. The reactions investigated in this work constitute a large database for understanding the chemistry of -O-4 lignin, β-5-3 lignin, and β-β lignin, and their decomposition in different solutions.
Keywords: Lignin; Lignocellulose; Biomass; Pretreatment; DFT
Contact information: National Engineering Laboratory for Biomass Power Generation Equipment, School of Renewable Energy Engineering, North China Electric Power University, Beijing, 102206, PR China;
* Corresponding author: qinwugx@126.com; zhengzm@ncepu.edu.cn
INTRODUCTION
Biofuel has received increasing attention due to the instability of petroleum prices and the development of government policies in many countries. To enhance biofuel production from biomass, two biological approaches to lignin biosynthetic pathway modification are employed, i.e. metabolic perturbation and the fundamental redesign of lignin (Guerriero et al. 2013). In addition, the pretreatments are used to reduce cell wall recalcitrance and promote sugar release during enzymatic hydrolysis. The extent and severity of biomass recalcitrance have been mainly attributed to lignin composition and amount. In particular, the S/G ratio affects cell wall deconstruction during pretreatment (Koo et al. 2012). Increasing the S/G ratio from 0.56 to 1.06 after phosphoric acid fractionation promotes the enzymatic hydrolysis efficiency (Kang et al. 2011). Varanasi concluded that with higher pretreatment times using 1-ethyl-3-methylimidazolium acetate ([C2mim] [OAc]). (6–12 h), it is easier to remove G-lignin at 120 oC and to remove S-lignin at 160 oC (Varanasi et al. 2012). The formation of new carbon–carbon bonds during acid-pretreatment leads to the disappearance of the G6 correlation and the obvious shift of G2 correlation due to the steric hindrance of the methoxyl group in the S units (Wang et al. 2012). Based on the solubility of cellulose, hemicellulose, and lignin in solvent-phosphoric acid and organic solvent-acetone, concentrated phosphoric acid fractionation is an effective pretreating method leading to the enhancement of enzymatic digestibility (Zhang et al. 2006; Zhang et al. 2007).
The relative-abundance ranges of the common linkages in monolignol are β–O–4 (45 to 50%), 5–5 (18 to 25%), β–5 (9 to 12%), α-linkages (6 to 8%), and β–β (3%) (Watts et al. 2011). Molecular simulation can provide atomistic detail and accurate energies of reactions and processes involving lignin, providing fundamental insight in biomass conversion. (Sangha et al. 2012). The density functional theory (DFT) was utilized to evaluate a potential reaction mechanism leading to the formation of coniferyl alcohol α-linkages in lignin (Watts et al. 2011).
A formal kinetic model treating the depolymerization and hydrodeoxygenation of wheat straw lignin in ethanol with formic acid as hydrogen source was developed and has been validated in a continuous stirred tank reactor (CSTR) at varying temperatures between 633 K and 673 K (Forchheim et al. 2012; Gasson et al. 2012). The dispersion-corrected DFT was used to model the interactions of imidazolium chloride ionic liquid anions and cations with cellulose and the lignin polyphenol models, respectively. The results suggested that extended π-systems of benzimidazolium ionic liquids lead to stronger interactions between lignin and the anions and cations of the ionic liquid (Janesko 2011). The Cα–O (Cα–Cβ) bond dissociation enthalpies of 40 to 44 (57 to 62) kcal/mol were obtained with DFT calculations (Younker et al. 2012). A diverse set of lignin linkages encompassing 65 lignin model compounds were obtained using the DFT approach (Parthasarathi et al. 2011). The valence band analysis of cellulose and lignin by DFT provides a powerful basis for a detailed interpretation of spectroscopic data (Haensel et al. 2012). However, the detailed physiochemical properties of lignin in phosphoric acid, acetone and water are still unknown. The lignin solubility properties in these different solutions have not been systematically explained yet. These unsolved problems hindered the development and optimization of lignocellulose pretreatment, preventing the selective and efficient conversion of lignin for advantage applications.
Herein, -O-4 lignin, β-5-3 lignin, and β–β Lignin were chosen as research models to do DFT calculations to investigate the properties of lignin in different solutions (water solution, acetone solution, and 85% w/w phosphoric acid solution) in relation to the phosphoric acid-acetone pretreatment process. In additions, we also detected the decomposition mechanism of these lignins, and the position effect and the different linkage effect on the decomposition of these lignins, in different solutions. This work aimed to reveal a fundamental database for understanding the chemistry of different lignin and the decomposition of such lignins in different solutions. Results will benefit the optimization of lignocellulose phosphoric acid-acetone pretreatment process for suitable applications.
COMPUTATIONAL METHODS
Lignin is formed through nonselective condensation of three phenolic alcohol units (monolignols) (Dorrestijn et al. 2000): p-coumaryl alcohol, coniferyl alcohol, and sinapyl alcohol. The most dominant linkage in lignin is arylglycerol-β-aryl ether (β-O-4), which is by far the most common linkage and much research has been dedicated to it. Our preliminary work was set to detect the property and breakdown of β-O-4 in different solutions. We herein focus on the other three important linkages in softwood lignin (Karhunen et al. 1995): -O-4 lignin, β-5-3 lignin, and β–β lignin. The optimized structure of -O-4 lignin, β-5-3 lignin, and β–β lignin contains two alcohols-coniferyl monomers (Haensel et al. 2012), in which the two alcohols (coniferyl monomers) are bonded to each other with the related -O-4 linkage, β-5-3 linkage, and β–β linkage, respectively. Although the stoichiometry of native lignin is not well reflected, we have chosen this constitution in order to keep the system sufficiently small for computational reasons and to have the typical linkage included as that of the simplest model compound for β-O-4 linkage for theoretical investigations (Gilbert and Gajewski 1982; Klein and Virk 1983). Then a large molecular structure model containing six alcohols-coniferyl monomers with the five related linkages was built to detect the position effect on the decomposition of -O-4 lignin, β-5-3 lignin, and β–β lignin. Moreover, two optimized structures with two different linkages were built to investigate linkage effects on the decomposition of lignin in different solutions. One contains -O-4 linkage and β-5-3 linkage next to each other among three alcohols-coniferyl monomers, and the other contains β-5-3 linkage and β–β linkage next to each other among three alcohols-coniferyl monomers. The lignin models were optimized in different solutions by using a solvation model with the appropriate dielectric constant, ɛ. Specific interactions between lignin and the solvent molecules were not considered in this study, since the solvent is represented by a polarizable continuum with a particular dielectric constant. The absolute errors in bond lengths with respect to the equilibrium geometrical para-meters were less than 0.1 Å. Values of the solvent dielectric constants ranging from ɛ = 1 (gas phase) to ɛ = 80.1 (water) were used, including ɛ = 20 (acetone) and ɛ = 12.7 (85% w/w phosphoric acid solution), in reference to the work of Wang and Anderko (2001) and Munson (1964). Then the molar volume for each lignin was computed in solution by making the solvent reaction field self-consistent with the solute electrostatic potential. During geometric optimization, all atoms in these models were allowed to relax freely.
The local-density approximation (LDA) is more accurate than the generalized gradient approximation (GGA) in describing a system with π–π interactions and other van der Waals systems (Vosko et al.1980; Janotti et al. 2001; Qin et al. 2010). Thus, all calculations were performed using DFT calculations with LDA for a previously generated exchange-correlation functional and a double-numeric quality basis set with polarization functions (DNP). Brillouin zone integration was performed at the gamma point. Calculations used an energy convergence tolerance at 1×10-6 Ha and gradient convergence at 1×10-6 Ha Å−1. A new formulation for the linear (LST) and quadratic synchronous transit (QST) methods was used to search the transition states and to investigate lignin fragment decomposition (Govind et al. 2003).
RESULTS AND DISCUSSION
Properties of -O-4 Lignin, β-5-3 Lignin and β–β Lignin in Different Solutions
Firstly, the geometric and electronic structures of -O-4 lignin, β-5-3 lignin, and β–β lignin in different solutions are illustrated in Fig. 1. From Fig. 1a, it can be observed that the molar volume for -O-4 lignin in gas phase, aqueous solution, acetone solution, and phosphoric acid solution is 258.482 cm3/mol, 277.395 cm3/mol, 274.527 cm3/mol, and 229.023 cm3/mol, respectively. The length of the -O-4 bond in gas phase, aqueous solution, acetone solution, and phosphoric acid solution is 1.469 Å, 1.474 Å, 1.473 Å, and 1.477 Å, respectively, while the bond population is 0.475, 0.459, 0.460, and 0.452, respectively. The bond length has an inverse relationship with the bond population. In comparison with -O-4 linkage in gas phase, it seems that the interaction between solvent molecule and lignin lengthens the -O-4 bond length and decreases the bond population on -O-4. Figure 1a depicts the highest occupied molecular orbit (HOMO) and the lowest unoccupied molecular orbit (LUMO) of -O-4 lignin in different solutions. As can be observed from Fig. 1a, LUMO and HOMO of -O-4 lignin in gas phase differ greatly from those in solutions. In gas phase, both HOMO and LUMO are mainly attributable to the left part of -O-4 lignin. However, in the solutions, the left part of -O-4 lignin contributes to the LUMO and the right part contributes to the HOMO. The energy level of HOMO and LUMO follows the order: gas phase > aqueous solution < acetone solution < phosphoric acid solution. The energy gap (Egap) between HOMO and LUMO for lignin in gas phase, aqueous solution, acetone solution, and phosphoric acid solution is 3.387 eV, 3.262 eV, 3.261 eV, and 3.260 eV, respectively. Results show that these solvent molecules can obviously modify the electronic property of -O-4 lignin, for which solvent molecules decrease the HOMO, the LUMO, and the Egap, while compared with those of -O-4 lignin in gas phase. And the interaction between phosphoric acid solution and -O-4 lignin can greatly change the properties of -O-4 lignin, especially the molar volume of -O-4 lignin.
For β-5-3 lignin shown in Fig. 1b, the molar volume in gas phase, aqueous solution, acetone solution, and phosphoric acid solution is 253.074, 265.350, 304.081, and 284.239 cm3/mol, respectively. The length of β-5-3 bond is 1.565 Å in gas phase compared to 1.546 Å in aqueous solution, acetone solution, and phosphoric acid solution. These results show that the interaction between lignin and solvent molecule lengthens the β-5-3 bond. Also, Fig. 1b illustrates that the bond population on β-5-3 in gas phase is 0.724, whereas the bond population on β-5-3 aqueous solution, acetone solution, and phosphoric acid solution is 0.722. In Fig. 1b, the left part of β-5-3 lignin contributes to the LUMO while the right part contributes to the HOMO. The energy level of HOMO and LUMO follows the order: gas phase < aqueous solution < acetone solution < phosphoric acid solution. The Egap for lignin in gas phase, aqueous solution, acetone solution, and phosphoric acid solution is 3.294 eV, 3.301 eV, 3.300 eV, and 3.300 eV, respectively. Compared with the effect on the structural and electronic properties of -O-4 lignin by these solvent molecules, the effect on the electronic properties of β-5-3 lignin is smaller, which implies the β-5-3 lignin is more stable in these solutions.
For β–β lignin shown in Fig. 1c, the molar volume in gas phase, aqueous solution, acetone solution, and phosphoric acid solution is 277.341 cm3/mol, 282.345 cm3/mol, 258.456 cm3/mol, and 258.508 cm3/mol, respectively. The length of β–β bond in aqueous solution, acetone solution, and phosphoric acid solution is the same, namely 1.563 Å, while the bond population in all cases is 0.720. The right part of the β–β lignin fragment contributes to the LUMO while the left part contributes to the HOMO. The HOMO energy of β–β lignin in different solutions follows the order: gas phase > aqueous solution > acetone solution > phosphoric acid solution, while the LUMO energy follows the order: gas phase < aqueous solution < acetone solution < phosphoric acid solution. The solution effect changes the HOMO and LUMO of β–β lignin in a perverse way. The Egap of β–β lignin in gas phase, aqueous solution, acetone solution, and phosphoric acid solution is 3.955 eV, 3.933 eV, 3.901 eV, and 3.877 eV, respectively. While comparing the solution effect on the property of -O-4 lignin, β-5-3 lignin, and β–β lignin with each other, it was found that -O-4 lignin is more sensitive to solvent molecule than β–βlignin and β–β lignin is more sensitive to solvent molecule than β-5-3 lignin.
Fig. 1. Stable configurations and the related HOMO and LUMO for -o-4 lignin, β-5-3 lignin, and β-βlignin in different solutions
Decomposition of -O-4 Lignin, β-5-3 Lignin, and β-β Lignin in Different Solutions
Following the property analysis of -O-4 lignin, β-5-3 lignin, and β-β lignin in different solutions, the decomposition of these lignin in different solutions was investigated. Figure 2a shows the calculated potential energy profiles for the decomposition of -O-4 bonds in aqueous solution, acetone solution, and phosphoric acid solution. The reaction initiates from the optimized geometries of -O-4 lignin. The barrier (Ea) for the -O-4 lignin decomposition in aqueous solution, acetone solution, and phosphoric acid solution is 1.481 eV, 1.333 eV, and 1.279 eV, respectively, and the related reaction energy (Er) in those solutions is 1.096 eV, 1.151 eV, and 1.194 eV, respectively. The -O-4 lignin decomposition in aqueous solution is the most energetically close, while the decomposition in phosphoric acid solution is the easiest, which is inconsistent with the electronic structure (bond length and bond population) of lignin in different solutions. A qualitatively similar one-step reaction mechanism was obtained for -O-4 decomposing in different solutions. Supposing the reaction rate (r) for lignin decomposition is r = k [Clignin]n (k is the reaction rate constant, [Clignin] is the concentration of lignin), and k = Aexp(-Ea/RT), r for the decomposition of -O-4 lignin in phosphoric acid solution is about 7.893 times as large as that in aqueous solution and 1.992 times as large as that in acetone solution at room temperature (298 K).
Figure 2b shows the same energy profiles as Fig. 2a but for the decomposition of the β-5-3 bond. The one-step reactions initiates from the optimized β-5-3 lignin in aqueous solution, acetone solution, and phosphoric acid solution, respectively. Both the Ea and the Er for β-5-3 decomposition in aqueous solution, acetone solution, and phosphoric acid solution are 4.999 eV and 4.856 eV, 5.006 eV and 4.874 eV, and 5.014 eV and 4.882 eV, respectively. The Ea follows the increasing order: aqueous solution < acetone solution < phosphoric acid solution, while the Egap follows a decreasing order, which is opposite to the phenomenon found during -O-4 lignin decomposition process. The value of r for the decomposition of β-5-3 lignin in aqueous solution is about 0.270 times larger than that in acetone solution and 0.585 times larger than that in phosphoric acid solution under room temperature.
Figure 2c shows the same energy profiles as Fig. 2a and b for the decomposition of β–β lignin. The Eaand the Er for β–β lignin decomposition in these three solutions are 4.277 eV and 3.768 eV, 4.257 eV and 3.770 eV, and 4.195 eV and 3.804 eV, respectively. While the Ea decreases in these solutions, the Er increases. The value of r for the decomposition of β–β lignin in phosphoric acid solution is about 3.194 times as large as that in aqueous solution and 2.384 times as large as that in acetone solution under room temperature.
Electronic analysis shows that during the decomposition reactions (in Fig. 2a, b, and c), one alcohols-coniferyl monomer in the model -O-4 lignin, β-5-3 lignin, and β–β lignin contributes to hole transfer at the valence band (related to HOMO), and the other alcohols-coniferyl monomer contributes to the electron transfer at the conduction band (related to LUMO) at initial state and final state. However, both electron and hole transfer occur at the atoms related to the lengthen -O-4 linkage, β-5-3 linkage, and β–β linkage at the transition state. The active electronic state favors the breakdown of these linkages.
The decomposition of -O-4 linkage is more accessible than that of β-5-3 linkage and β–β linkage, corresponding to the previous report that the decomposition of -O-4 linkage needs far lower barrier energy than the decomposition of C–C linkages (β-5-3 linkage and β–β linkage) (Cho and Parthasarathi 2010; Kim et al. 2011). Therefore, it is suggested that -O-4 linkage can act as the vulnerable point of lignin. The r value for -O-4 lignin decomposition in aqueous solution, acetone solution, and phosphoric acid solution is about 137, 143, and 145 times as large as those for β-5-3 lignin decomposition, and is about 108, 114, and 113 times as large as those for β–β lignin decomposition under room temperature, respectively. The -O-4 bond in lignin breaks more easily than the glucosidic bond in cellulose, while the β-5-3 bond and β–β bond are more stable than the glucosidic bond in aqueous solution, acetone solution, and phosphoric acid solution (Kang et al. 2012). According to the value of r, solutions have greater effect on -O-4 and β–β decompositions than on β-5-3 decomposition. These results of DFT calculations on lignin model will be accurate enough to predict the bond dissociation energies and reactivity trend, as mentioned previously (Cho 2010).
Fig. 2. Calculated potential energy profiles given in eV for the decompositions of (a) -O-4 linkage, (b) β-5-3 linkage, and (c) β–β linkage in water, acetone, and phosphoric acid solution. The symbols Ea and Er denote the barrier energy and reaction energy, respectively.
Decomposition of -O-4 Lignin, β-5-3 Lignin and β-β Lignin at Different Position in Different Solutions
Two lignin fragments with five -O-4 linkages and five β-5-3 linkages were modeled respectively. The break of -O-4(1 or 1́), -O-4(2 or 2́), and -O-4(3), as well as the break of β-5-3(1 or 1́), β-5-3(2 or 2́), and β-5-3(3) were considered. Figure 3 presents the calculated potential energy profiles given in eV for the decomposition of -O-4 (1 or 1́) (the head -O-4 linkage), -O-4(2 or 2́) and -O-4(3) (the center -O-4linkage), and the decomposition of β-5-3(1 or 1́) (the head β-5-3 linkage), β-5-3(2 or 2́), and β-5-3(3) (the center β-5-3 linkage) in different solutions. According to Ea and Er for the decomposition of -O-4(1 or 1́), -O-4(2 or 2́), and -O-4(3), it can be observed that the decomposition of the center -O-4linkage (-O-4(3)) is more accessible than the decomposition of -O-4(1 or 1́) and -O-4(2 or 2́) in the solutions and especially in the acetone solution. This result implies that -O-4 lignin usually breaks from the center -O-4 linkage. The result will favor the dissolution of lignin into shorter fragments during phosphoric acid treatment process, and the easier decomposition of -O-4 lignin in acetone solution ensures that the temperature is low enough to avoid further breakdown of lignin fragments during acetone treatment.
Fig. 3. Calculated potential energy profiles given in eV for (a) the decomposition of -O-4 lignin (including -O-4(1), -O-4(2), and -O-4(3)), and (b) the decomposition of β-5-3 (including β-5-3(1), β-5-3(2), and β-5-3(3)) in water, acetone, and phosphoric acid solution. The IS, TS, and FS denote the initial state, transition state, and final state, respectively.
Figure 3b shows that all Ea values for the decomposition of β-5-3(1 or 1́), β-5-3(2 or 2́), and β-5-3(3) in acetone solution and phosphoric acid solution are much larger than those for the decomposition of -O-4(1 or 1́),-O-4(2 or 2́), and -O-4(3), suggesting that β-5-3 lignin is more stable than -O-4 lignin. However, the position effect together with certain solution effects greatly threatens the stability of β-5-3 lignin, as shown by the fact that the Ea for the decomposition of β-5-3(2) in water is only 2.860 eV, far lower than those for the other cases. On the whole, the independent effect of position on the decomposition of β-5-3 lignin and -O-4 lignin is relatively small in the same solution system, and the independent effect of solution for the same β-5-3 linkage and -O-4 linkage decomposition is also small. As far as pretreatment of lignin in solution is concerned, the combination of position effect and solution effect would be an important factor for the dissolution and sedimentation processes.
Different Linkage Effect on the Decomposition of Lignin in Different Solutions
Lignin has been described as a random, three-dimensional network polymer comprised of variously linked phenylpropane units. Different linkages at realistic conditions may be another important factor affecting the pretreatment of lignin. To discuss different linkage effect on the decomposition of lignin in different solutions, we randomly bonded -O-4 and β-5-3 in one lignin system and bonded β-5-3 and β–β in another lignin system, which is illustrated in Fig. 4.
In the presence of β-5-3, the decomposition of -O-4 in water, acetone, and phosphoric acid solution requires the overcoming of an barrier energy of 1.743 eV, 1.818 eV, and 1.792 eV, respectively, and these values are higher than those for a single -O-4 decomposition investigated in Fig. 2a. Conversely, in the presence of -O-4, the decomposition of β-5-3 in water, acetone, and phosphoric acid solution needs the barrier energies of 4.921 eV, 4.935 eV, and 4.946 eV, respectively, which are lower than those for a single β-5-3 decomposition investigated in Fig. 2b. The β-5-3 linkage favors the stability of -O-4, since when -O-4 and β-5-3 are combined next to each other in one lignin system, the -O-4 bond-length decreases from 1.474 Å to 1.462 Å in water solution, from 1.473 Å to 1.163 Å in acetone solution, and from 1.477 Å to 1.463 Å in phosphoric acid solution, respectively. However, -O-4 depresses the stability of β-5-3, though β-5-3 bond-length decreases from 1.546 Å to 1.519 Å in the three solution. Differing from the change of Ea in a perverse way for the β-5-3 & -O-4 lignin system, both β–β and β-5-3 decrease the Ea for each other’s decomposition in the β-5-3 & β–β lignin system. β-5-3 in the β-5-3 & β–β system lengthens from 1.590 Å to 1.563 Å, β-5-3 in the β-5-3 & β–β system lengthens from 1.546 Å to 1.555 Å, in water, acetone, and phosphoric acid solution, respectively.
Such a position effect on linkage breakdown can also be analyzed in the previous discussion about bond strength for the diverse set of lignin linkages (Cho 2010) and bond dissociation enthalpies for lignin (Younker et al 2011; Younker et al 2012). Taking the β-5-3 & -O-4 lignin system and the β-5-3 & β–β system into consideration, β-5-3 and β–β are also far more stable than -O-4.
Fig. 4. Calculated potential energy profiles given in eV for the decomposition of lignin linkage under the effect of different linkage in water, acetone, and phosphoric acid solution: (a) decomposition of -O-4 and β-5-3; (b) decomposition of β–β and β-5-3. The IS, TS, and FS denote the initial state, transition state, and final state, respectively.
Thermodynamical Factors in Lignin Decomposition
The reactions investigated in this work constitute a large database for under-standing the chemistry of different types of lignin and their decomposition in different solutions. For example, the effect of thermodynamics on reaction barriers is able to be examined. The reaction barriers against the reaction energies of the -O-4, β-5-3, and β–β breakage reactions in different solutions are plotted in Fig. 5. Here the reaction energy is defined as the energy difference between the final state and the initial state, which is close to the reaction enthalpy. The barrier vs. enthalpy relationship is known traditionally as the Brönsted-Evans-Polanyi (BEP) correlation (Liu and Hu. 2001; Michaelides et al. 2003; Bligaard et al. 2004). However, for the cleavage of different linkage types of lignin in different solutions, whether the BEP relation holds or not is still unclear.
Fig. 5. Plot of Ea against Er for (a) a different lignin bond (-O-4, β-5-3, and β–β) breaking in the same solution and (b) the same lignin bond breaking in different solutions
As shown in Fig. 5a, a good linear relationship was identified for different linkage (-O-4, β-5-3, and β–β) breakage reactions in the same solution. This striking feature simply reflects the fact that a single BEP relation can be established for these different lignin decomposition paths in the same solution. However, three distinct regions are identified for the same linkage breaking in different solutions, as shown in Fig. 5b, each corresponding to a class of bond-breaking reactions. BEP relation can also be established for -O-4, β-5-3, and β–β bond breakage. The upper-left region is dominated by the β-5-3 bond breakage, the bottom-right region is mainly about the -O-4, and the region in-between is for the β–β bond breakage. Inside the upper-left region and the bottom-right region, a positive linear BEP relationship holds for most of the reactions, while the region in-between a negative BEP linear relationship. There is also a minor distinction between reactions in different solutions. As the -O-4, β-5-3, and β–β bonds are known to possess quite different bond polarities, the figure shows that Ea not only depends on thermodynamics, as represented by the Er, but also is determined by the intrinsic bond polarity. Under similar thermodynamic conditions, the polar bond -O-4 breaks first. Except for this, the thermodynamics is often the most important factor that determining the barrier height. The β-5-3 bond breakage is usually a more highly endothermic reaction than the β–β bond breakage. The thermodynamic factor determines that the β–β bond breakage is usually preferred over the β-5-3 bond, although the β-5-3 bond is more polar than the β–β bond. This is what the authors found in lignin decomposition.
CONCLUSION
- Water, acetone, and phosphoric acid solutions can greatly modify the electronic properties of -O-4 lignin, for which solvent molecules decrease the HOMO, LUMO, and Egap, compared to those of -O-4 lignin in the gas phase. The interaction between phosphoric acid solution and -O-4 lignin can greatly change the properties of -O-4 lignin, especially the molar volume of -O-4 lignin. The solution effect changes the HOMO and LUMO of β–β lignin in a perverse way, leading to the decrease of Egap of β–β lignin. However, solvent molecules affect the electronic properties of β-5-3 lignin slightly. The -O-4 lignin is more sensitive to solvent molecule than β–β lignin while the β–β lignin is more sensitive to solvent molecule than β-5-3 lignin.
- -O-4 lignin decomposes much more easily than β-5-3 lignin and β–β lignin. The r value for -O-4 lignin decomposition in aqueous solution, acetone solution, and phosphoric acid solution is about 137, 143, and 145 times as large as those for β-5-3 lignin, and about 108, 114, and 113 times as large as those for β–β lignin decomposition under room temperature, respectively. Solutions have greater effect on -O-4 and β–β decomposition than on β-5-3 decomposition.
- The independent effect of position on the decomposition of β-5-3 lignin and -O-4 lignin is relatively small in the same solution system, and the independent effect of solution for the same β-5-3 linkage and -O-4 linkage decomposition is also small. But the position effect together with certain solution effects greatly threatens the stability of β-5-3 lignin.
- β-5-3 can enhance the stability of -O-4, while -O-4 depresses the stability of β-5-3 at the same time. However, β–β and β-5-3 decrease the Ea for each other’s decomposition in the β-5-3 and β–β lignin system.
- A linear BEP relation, a thermodynamic rule, holds clearly for different lignin decomposition reactions in different solutions, which may be used to qualitatively predict the general pattern in lignin dissociation reactions, such as decomposition sites and the reaction precursor.
ACKNOWLEDGEMENTS
The authors thank the financial support for this work provided by National Natural Science Foundation of China (51106051) and National Key Technology R&D Program (2012BAA09B01).
REFERENCES CITED
Bligaard, T., Norskov, J. K., Dahl, S., Matthiesen, J., Christensen, C. H., and Sehested, J. (2004). “The Brønsted–Evans–Polanyi relation and the volcano curve in heterogeneous catalysis,” Journal of Catalysis 224(1), 206-217.
Cho, D. W., and Parthasarathi, R. (2010). “Gaining insight to more efficiently convert plant biomass into biofuels,” The Journal of Organic Chemistry 75, 6549-6562.
Forchheim, D., Gasson, J. R., Hornung, U., Kruse, A., and Barth, T. (2012). “Modeling the lignin degradation kinetics in an ethanol/formic acid solvolysis approach. Part 1. Kinetic model development,” Industrial & Engineering Chemistry Research 51(32), 10595-10606.
Forchheim, D., Gasson, J. R., Hornung, U., Kruse, A., and Barth, T. (2012). “Modeling the lignin degradation kinetics in an ethanol/formic acid solvolysis approach. Part 2. Validation and transfer to variable conditions,” Industrial & Engineering Chemistry Research 51(46), 15053-15063.
Gilbert, K. E., and Gajewski, J. J. (1982). “Coal liquefaction model studies: Free radical chain decomposition of diphenylpropane, dibenzyl ether, and phenethyl phenyl ether via .beta.-scission reactions,” The Journal of Organic Chemistry 47, 4899-4902.
Govind, N., Petersen, M., Fitzgerald, G., King-Smith, D., and Andzelm, J. (2003). “A generalized synchronous transit method for transition state location,” Computational Materials Science 28(2), 250-258.
Guerriero, G., Sergeant, K., and Hausman, J. F. (2013). “Integrated -Omics: A powerful approach to understanding the heterogeneous lignification of fibre crops,” International Journal of Molecular Sciences 14(6), 10958-10978.
Haensel, T., Reinmöller, M., Lorenz, P., Beenken, W. J. D., Krischok, S., and Ahmed, S. I. U. (2012). “Valence band structure of cellulose and lignin studied by XPS and DFT,” Cellulose 19(3), 1005-1011.
Janesko, B. G. (2011). “Modeling interactions between lignocellulose and ionic liquids using DFT-D,” Physical Chemistry Chemical Physics 13(23), 11393-11401.
Janotti, A., Wei, S. H., and Singh, D. J. (2001). “First-principles study of the stability of BN and C,” Physical Review B 64(17), 4107-4111.
Kang, P., Zheng, Z. M., Qin, W., Dong, C. Q., and Yang, Y. P. (2011). “Efficient fractionation of Chinese white poplar biomass with enhanced enzymatic digestibility and modified acetone- soluble lignin,” BioResources 6(4), 4705-4720.
Karhunen, P., Rummakko, P., Sipilä, J., Brunow, G., and Kilpeläinen, I. (1995). “Dibenzodioxocins; A novel type of linkage in softwood lignins,” Tetrahedron Letters 36(1), 167-170.
Kim, S., Chmely, S. C., Nimlos, M. R., Bomble, Y. J., Foust, T. D., Paton, R. S., and Beckham, G. T. (2011). “Computational study of bond dissociation enthalpies for a large range of native and modified lignins,” Journal of Physical Chemistry Letters 2, 2846-2852.
Klein, M. T., Virk, P. S. (1983). “Model pathways in lignin thermolysis 1. Phenethyl-phenylether,”Industrial & Engineering Chemistry Fundamentals 22, 35-45.
Koo, B. W., Min, B. C., Gwak, K. S., Lee, S. M., Choi, J. W., Yeo, H., and Choi, I. G. (2012). “Structural changes in lignin during organosolv pretreatment of Liriodendron tulipifera and the effect on enzymatic hydrolysis,” Biomass and Bioenergy 42, 24-32.
Liu, Z. P., and Hu, P. (2001). “General trends in CO dissociation on transition metal surfaces,” Journal of Chemical Physics 114(19), 8244(1-4).
Michaelides, A., Liu, Z. P., Zhang, C. J., Alavi, A., King, D. A., and Hu, P. (2003). “General linear relationships between activation energies and enthalpy changes for reactions at surfaces,” Journal of the American Chemistry Society 125(13), 3704-3705.
Munson, R. A. (1964). “Dielectric constant of phosphoric acid,” Journal of Chemical Physics 40, 2044-2046.
Parthasarathi, R., Romero, R. A., Redondo, A., and Gnanakaran, S. (2011). “Theoretical study of the remarkably diverse linkages in lignin,” The Journal of Physical Chemistry Letters 2(20), 2660-2666.
Qin, W., Li, X., Bian, W. W., Fan, X. J., and Qi, J. Y. (2010). “Density functional theory calculations and molecular dynamics simulations of the adsorption of biomolecules on graphene surfaces,” Biomaterials 31(5), 1007-1016.
Sangha, A. K., Petridis, L., Smith, J. C., Ziebell, A., and Parks, J. M. (2012). “Molecular simulation as a tool for studying lignin,” Environmental Progress & Sustainable Energy 31(1), 47-54.
Varanasi, P., Singh, P., Arora, R., Adams, P. D., Auer, M., Simmons, B. A., and Singh S. (2012). “Understanding changes in lignin of Panicum virgatum and Eucalyptus globulus as a function of ionic liquid pretreatment,” Bioresource Technology 126, 156-161.
Vosko, S. J., Wilk, L., and Nusair, M. (1980). “Accurate spin-dependent electron liquid correlation energies for local spin density calculations: A critical analysis,” Canadian Journal of Physics 58(8), 1200-1211.
Wang, P., and Anderko, A. (2001). “Computation of dielectric constants of solvent mixtures and electrolyte solutions,” Fluid Phase Equilibria 186, 103-122.
Wang, K., Yang, H., Yao, X., Xu, F., and Sun, R. C. (2012). “Structural transformation of hemicelluloses and lignin from triploid poplar during acid-pretreatment based biorefinery process,” Bioresource Technology 116, 99-106.
Watts, H. D., Mohamed, M. N., and Kubicki, J. D. (2011). “Evaluation of potential reaction mechanisms leading to the formation of coniferyl alcohol α-linkages in lignin: A density functional theory study,” Physical Chemistry Chemical Physics 13(47), 20974-20985.
Younker, J. M., Beste A., and Buchanan, A. C. (2011). “Computational study of bond dissociation enthalpies for substituted β-O-4 lignin model compounds,” Chemphyschem 12, 3556-3565.
Younker, J. M., Beste, A., and Buchanan III, A. C. (2012). “Computational study of bond dissociation enthalpies for lignin model compounds: β-5 arylcoumaran,” Chemical Physics Letters 545, 100-106.
Zhang, Y. H. P., Cui, J. B., Lynd, L. R., and Kuang, L. R. (2006). “A transition from cellulose swelling to cellulose dissolution by o-phosphoric acid: Evidence from enzymatic hydrolysis and supramolecular structure,” Biomacromolecules 7(2), 644-648.
Zhang, Y. H. P, Ding, S. Y., Mielenz, J. R, Cui, J, Elander, R. T., Laser, M., Himmel, M. E., McMillan, J. D., Lynd, L. R. (2007). “Fractionating recalcitrant lignocellulose at modest reaction conditions,” Biotechnology and Bioengineering 97, 214-223.
Article submitted: October 4, 2013; Peer review completed: November 21, 2013; Revised version received and accepted: November 27, 2013; Published: December 4, 2013.