Kinetic modelling of methyl formate hydrolysis in the presence of formic

acid as a homogenous catalyst

1

O. Jogunola, 1T. Salmi, 1,2J.-P. Mikkola

bo Akademi University, Dep. Chemical Engineering, Biskopsgatan 8, 20500, bo/Turku
Finalnd; Fax. +35822154479, E-mail: jolatund@abo.fi; 2Ume University, Dep. Chemistry,
90187, Ume, Sweden.
1

Methyl formate (MeFo) hydrolysis, a prominent industrial process for producing formic
acid (FA), a versatile, environmentally benign chemical, accounts for more than 70% of the
world production capacity. Nowadays, new uses are being sought for the acid by researchers
and the chemical industry due to its relatively strong acidity, reducing power, and ecofriendliness. The methyl formate hydrolysis process is a slow, endothermic and equilibriumlimited reaction, which produces methanol (MeOH) and FA in almost equimolar amount.
However, FA produced in the reaction catalysed that same reaction. HCl and H 2SO4 have
been reported to speed up the reaction, but these acids have to be separated from the reaction
mixture and this adds to the cost of energy. The endeavour is to use formic acid as an initial
charge to speed up its own reaction.
The hydrolysis experiments (with or with FA catalyst) were done in a stirred laboratoryscale, almost isothermal batch reactor at 80 110OC and 20 bar nitrogen pressure using a
constant initial water-to ester molar ratio (H2O/MeFo = 1.8). The hydrolysis of methyl
formate can be simply represented as:
HCOOCH3 (A) + H2O (B) HCOOH (C) + CH3OH (D)
In the absence of the catalyst, the reaction is autocatalysed and the reaction rate can be
expressed as:
1
r (k k C C )(C A C B
CC C D )
KC
Using FA catalyst in the form of an initial charge, the rate of the reaction becomes:
C C
r k (C A C B C D )(K d C C ) 0.5
KC
where Kd is the dissociation constant of formic acid. The model equations were solved by
difference method. The software, ModEst solved the model equations and minimized the
objective function (SSQ = (ymodel-yexp)2 by adjusting the three parameters, k 0 , K 0C , and E a
with the Levenberg-Marquardt-simplex method, while H ro was fixed at +5.44 kJ/mol. The
kinetic and equilibrium parameters included in the rate equations were estimated from the
experimental data by non-linear regression analysis. The predictive power R2 (coefficient of
determining between predicted and experimental values) is given as: R 2 1

(c
(c

i ,exp, k

ci , k ) 2

i ,exp, k

c ,i , k ) 2

where ci ,k is the model prediction concentration, ci ,k is the observation mean concentration.

Results
In the absence of the catalyst, the reaction was slow as indicated by the induction period.
Furthermore, autocatalysis (S-shaped) was pronounced as shown in Fig. 1. Upon addition of
the catalyst, the induction period ceased to exist and the reaction rate was sped up. However,

the product yield was suppressed. It can be deduced from Fig. 1 that the optimal initial acidto-ester molar ratio, which will increase the reaction rate without having a drastic effect on the
yield is 0.1. The model prediction of the reaction rate of both the autocatalysed process and
the FA-catalysed system (FA/MeFo = 0.1) is depicted in Fig. 2.
0.02

ri (mol/g min)

Conversion (mol-%)

30

20
FA/Me Fo
FA/Me Fo
FA/Me Fo
FA/Me Fo
FA/Me Fo

10

=
=
=
=
=

0.0
0.1
0.2
0.05
0.15

0.016
0.012
0.008
0.004
0
0

0
0

50

100

150

200

50

100

150

Time (min)

250

Autocatalysis

Time (min)

Formic acid catalysed

Figure 1. The effect of the acid catalyst on

Figure 2. Comparison of the reaction rate of the

the reaction rate and equilibrium conversion

FA-catalysed system with that of the autocatalysed

process at 90oC; H2O/MeFo = 1.8

at 80 C; H2O/MeFo = 1.8

The results of the kinetic modelling for the two systems is depicted in Table 1.
Table 1. Comparison of the parameters between autocatalyzed and FA-catalyzed reactions
k
k
k
Reaction types
KC
Ea
E a
E a'
(kg/mol min)

Autocatalysed
FA-catalyzed

0.03

(kg2/mol2 min)

(kJ/mol)

0.39

88.2

0.12

(kJ/mol)

67.8

(kJ/mol)

66.4

0.2
0.2

FA-catalysed reaction is more than three times faster than the autocatalysed reaction and
their activation energies are in the correct range. The equilibrium constant remains the same.
Some of the fits of the two models to the experimental results are depicted in Fig. 3.
20

Ci (mol/kg)

Ci (mol/kg)

20

15
10
5
0
0

50

100

150

15
10
5
0
0

Time (min)

100

200

Time (min)

A ( expt )

B ( expt )

C ( expt )

A ( mo del)

B ( mo del)

C ( mo del)

Figure 3a. The model fit to experimental data at 90oC;

FA/MeFo = 0.1 and H2O/MeFo = 1.8

A (m ode l)
D (m ode l)
C (e xpt)

B (m ode l)
A (e xpt)
D (e xpt)

C (m ode l)
B (e xpt)

Figure 3. The model fit to experimental data at 90oC;

FA/MeFo = 0.0 and H2O/MeFo = 1.8

Conclusion
The hydrolysis of methyl formate with or without the formic acid catalyst was accompanied
in a batch reactor. The models developed for the two processes were able to predict the
experimental results successfully. This approach can also be used for other alkyl formates.
Reference
[1] O. Jogunola, T. Salmi, J. Wrn, J.-P. Mikkola, Kinetic studies of alkyl formate
hydrolysis using formic acid as a catalyst, J. Chem. Technol. Biotechnol. (2011) DOI
10.1002/jctb.2714
[2] O. Jogunola, T. Salmi, K. Ernen, J. Wrn, M. Kangas, J.-P. Mikkola, Reversible
autocatalytic hydrolysis of alkyl formate: kinetic and reactor modelling, Ind. Eng. Chem.
Res. 49 (2010) 4099-4106.