Kinetic Modeling of Hardwood Prehydrolysis

Water prehydrolysis is effected by the acetic acid released from labile acetyl groups

present in the hemicelluloses. The extent of xylan hydrolysis is dependent upon

the hardwood source, temperature, time, acidity and liquor-to-solid ratio. Control

of the performance of the prehydrolysis kraft process necessitates an understanding

of the kinetics of wood fractionation during the course of prehydrolysis. The aimis to

establish a relationship between the intensity of prehydrolysis and the purification

and delignification efficiency and selectivity under given subsequent pulping conditions.

Assuming that the xylan is completely accessible to water such that no diffusional

restrictions affects the rate, and the local rate of xylan dissolution is proportional

to the xylan concentration, the kinetics should follow first order in xylan. It

has been shown that xylan contains fractions of different reactivity as there is a

distinct slowing of the rate at about 70% conversion [29,34–36]. Transport limitations,

different portions of the xylan with different reactivities and changing interfacial

areas have been considered as possible explanations of the observed rate behavior

[36,37]. Assuming xylan to be composed of two fractions, each of which

reacts according to a homogeneous first-order kinetics law, the rate of xylan hydrolysis

can be expressed as Eq. (142), according to the proposal of Veeraraghavan

[38]and Conner [37]. The model is based on the presence of two types of xylan

that hydrolyze via parallel first-order reactions:

_

dX

dt _ zX _ kf _ X _ Xf _1 _ zX __ ks _ X _ Xs _142_

in which X is the fraction of original xylan remaining in the solid residue, Xf and

Xs are the fractions of the fast- and slowly-reacting xylan remaining in the residue,

zX is the fraction of the readily reacting xylan exemplified by Eq. (143):

zX _

_ X _ Xs _

X _

Xf

X _143_

and kf,X and ks,X are the rate constants.

A nonlinear regression (least squares) analysis is applied to determine the best

values for the coefficients in the integrated form of Eq. (142), to calculate the dissolution

of the wood components for each set of reaction conditions:

X _ zX _ Exp _ kf _ X _ t _ __1 _ zX __ Exp _ ks _ X _ t _ _ _144_

The same algorithm can be used by modeling the (total) weight loss kinetics of

the wood residue as the removal of xylan and its degradation products (xylo-oligomers,

xylo-monomers and furfural) constitutes the bulk of the weight loss:

W _ zW _ Exp __ kf _ W _ t __1 _ zW __ Exp _ ks _ W _ t _ _ _145_

in which W is the mass fraction of the remaining solid wood, Wf and Ws are the

fractions of the fast- and slowly-reacting wood components remaining in the residue,

zW is the fraction of the readily reacting wood components expressed by

Eq. (146):

zW _

_ W _ Ws _

W _

Wf

W _146_

and kf,W and ks,W are the rate constants.

The kinetics of weight loss and xylan hydrolysis is demonstrated for the water

prehydrolysis of beech wood [39]. Beech wood was ground to finely divided pieces

of average particle size 4 mm to minimize diffusion-controlled mass transfer. The

composition of the beech wood is summarized in Tab. 4.42.

Tab. 4.42 Composition of beech wood (Fagus sylvatica) [39].

Anhydride Percentage [o.d. wood]

Glucose 42.6

Xylose 19.5

Arabinose 0.7

Galactose 0.8

Mannose 1.1

Rhamnan 0.5

Acetyl 4.5

Uronic 3.1

Klason lignin 21.0

Acid soluble lignin 3.5

Extractivesa) 1.8

Ash 0.4

Sum 99.5

a) Ethanol and DCM extractives.

330 4 Chemical Pulping Processes

Prehydrolysis was performed at a liquor-to-solid ratio of 10:1 using a series of

three 450-mL Parr autoclaves at three different temperature levels, 140 °C, 155 °C,

and 170 °C. Heating-up time was kept below 20 min and corrected for isothermal

conditions using the following expression [Eq. (147)]:

tTo _

tT 0

tTt

Exp

EA

R _

TTt _

_ _ T 0___ dt _147_

in which Tt is the temperature during heating-up, T0 the target temperature, tT0

the reaction time at target temperature, and EA the activation energy. The use of

corrected prehydrolysis time requires an iterative procedure. First, a value is

assumed, and the isothermal reaction time is calculated for each of the prehydrolysis

experiments. A revised activation energy may then be calculated from an

Arrhenius plot. The iterative procedure is repeated until the change in the activation

energy becomes insignificant.

The results of weight-loss kinetics are shown in Fig. 4.96.

0 500 1000 1500 2000

140.C 155.C 170.C

Wood Yield, %

Time at temperature, min

Fig. 4.96 Mass of wood residue versus reaction time (corrected

for isothermal conditions) for water prehydrolysis of

beech wood. Liquor-to-solid ratio 10:1 (according to [39]).

The kinetic parameters for the fast weight-loss reaction from water prehydrolysis

of beech and Eucalyptus saligna are compared in Tab. 4.43.

4.2 Kraft Pulping Processes 331

332 4 Chemical Pulping Processes

Tab. 4.43 Kinetic parameters for the weight-loss of beech wood

and eucalypt according to Eq. (145).

Wood species I:s ratio Temperature

°C

(1 – zW) kf,W

[min–1]

Ks,W

[min–1]

Ef,W

[kJ mol–1]

Дата добавления: 2015-10-21