284
as explained previously [62]. Further AIMD simulations showed that the
oxocarbenium ions were found to be stable for 2.5 ps in the reaction environment
containing water [62] (Fig. 6(ii)). However, AIMD simulations are computationally
very expensive. Another alternative approach to this is Car-Parrinello AIMD
(CPMD) which combines quantum mechanics and classical MD. However, CPMD
calculation takes a longer time. Metadynamics [81] is used as a method to accelerate
the dynamics of CPMD calculations. To understand the solvent dynamics effects on
the rDA reaction of partially saturated 2-pyrones containing electron-withdrawing
and electron-donating substituents, CPMD simulation in conjunction with
metadynamics was used. In the presence of solvent phase, the RO and decarboxylation
barrier were lowered, and the effect was more pronounced for the 2-pyones
containing electron-donating substituent [69]. The decrease in the barrier was
attributed to the differential stabilization of the transition and reactant state due to
the structuring and dynamics of solvent molecules [69].
4 Pyrolysis Chemistry
High degree of polymerization in cellulose and hemicellulose renders use of DFT
simulations in studying the pyrolysis reaction and mechanism. Moreover, pyrolysis
reactions are carried out in condensed phase, whereas most of the DFT simulations
are done in gas phase. Recent studies of pyrolysis chemistry were performed by
AIMD simulations in condensed environment and at finite temperature [82]. A
simulation cell of periodically repeated two units of cellulose Iβ was used to mimic
the solid cellulose by Agarwal et al. [82] Using CPMD metadynamics, they found
that at 327 °C, ring contraction of glucopyranose to glucofuranose led to
depolymerization with a lowest free energy barrier of 20 kJ/mol (Fig. 7(i)) and at
600 °C precursor to levoglucosan was formed with a free energy barrier of 36 kcal/
mol [82] (Fig. 7(ii)). On the other hand, Mettler et al. used α-cyclodextrin as a
model compound for cellulose [83] and suggested direct formation of furanic
compounds from cellulose instead of cellulose decomposition to glucose [83]. They
observed that the cleavage of glycosidic linkage between two glucose units leads to
furan ring formation followed by formation of carbonyl group at C1 position
(Fig. 7(iii)). This is followed by RO of the pyran ring. In order to satisfy the valency
of pyran O, it undergoes ring closure again but at C2 which is followed by
dehydration to form the furan ring [83]. Therefore, for small-scale cellulose
pyrolysis system having ~100 atoms, CPMD-metadynamics appears to be a useful
approach. This limits detailed description of pyrolysis of cellulose in complex
environment.
The reactive force field (ReaxFF) [84, 85] is another promising approach that can
model complex reactive system. Zheng et al. constructed a large cellulose model
having 7572 atoms containing six long chains and each containing 60 glycopyranoses
for conducting ReaxFF simulation [86]. The simulations were performed at
500–1400 K for 250 ps. The ReaxFF simulation results showed that the reaction
S. Gupta
as explained previously [62]. Further AIMD simulations showed that the
oxocarbenium ions were found to be stable for 2.5 ps in the reaction environment
containing water [62] (Fig. 6(ii)). However, AIMD simulations are computationally
very expensive. Another alternative approach to this is Car-Parrinello AIMD
(CPMD) which combines quantum mechanics and classical MD. However, CPMD
calculation takes a longer time. Metadynamics [81] is used as a method to accelerate
the dynamics of CPMD calculations. To understand the solvent dynamics effects on
the rDA reaction of partially saturated 2-pyrones containing electron-withdrawing
and electron-donating substituents, CPMD simulation in conjunction with
metadynamics was used. In the presence of solvent phase, the RO and decarboxylation
barrier were lowered, and the effect was more pronounced for the 2-pyones
containing electron-donating substituent [69]. The decrease in the barrier was
attributed to the differential stabilization of the transition and reactant state due to
the structuring and dynamics of solvent molecules [69].
4 Pyrolysis Chemistry
High degree of polymerization in cellulose and hemicellulose renders use of DFT
simulations in studying the pyrolysis reaction and mechanism. Moreover, pyrolysis
reactions are carried out in condensed phase, whereas most of the DFT simulations
are done in gas phase. Recent studies of pyrolysis chemistry were performed by
AIMD simulations in condensed environment and at finite temperature [82]. A
simulation cell of periodically repeated two units of cellulose Iβ was used to mimic
the solid cellulose by Agarwal et al. [82] Using CPMD metadynamics, they found
that at 327 °C, ring contraction of glucopyranose to glucofuranose led to
depolymerization with a lowest free energy barrier of 20 kJ/mol (Fig. 7(i)) and at
600 °C precursor to levoglucosan was formed with a free energy barrier of 36 kcal/
mol [82] (Fig. 7(ii)). On the other hand, Mettler et al. used α-cyclodextrin as a
model compound for cellulose [83] and suggested direct formation of furanic
compounds from cellulose instead of cellulose decomposition to glucose [83]. They
observed that the cleavage of glycosidic linkage between two glucose units leads to
furan ring formation followed by formation of carbonyl group at C1 position
(Fig. 7(iii)). This is followed by RO of the pyran ring. In order to satisfy the valency
of pyran O, it undergoes ring closure again but at C2 which is followed by
dehydration to form the furan ring [83]. Therefore, for small-scale cellulose
pyrolysis system having ~100 atoms, CPMD-metadynamics appears to be a useful
approach. This limits detailed description of pyrolysis of cellulose in complex
environment.
The reactive force field (ReaxFF) [84, 85] is another promising approach that can
model complex reactive system. Zheng et al. constructed a large cellulose model
having 7572 atoms containing six long chains and each containing 60 glycopyranoses
for conducting ReaxFF simulation [86]. The simulations were performed at
500–1400 K for 250 ps. The ReaxFF simulation results showed that the reaction
S. Gupta