SYNTHESIS OF EPICHLOROHYDRIN

FROM DICHLOROPROPANOLS
Kinetic Aspects of the Process

L. Ma, J. W. Zhu
, X. Q. Yuan and Q. Yue

Institution of Chemical Engineering, East China University of Science and Technology, Shanghai,
200237, China.

Abstract: The reactions involved in the industrial production of epichlorohydrin from dichloro-
propanols (DCP, including 1,3-dichloro-2-propanol and 2-3-dichloro-1-propanlol) have been kine-
tically investigated. The kinetics of dehydrochlorination of DCP has been studied in the presence
of caustic soda, by potentiometric techniques. The kinetic parameters of the reactions have been
determined. The epichlorohydrin hydrolysis has been studied by measuring the decrease of the
reagents during the time using titration. The whole study indicate that both dehydrochlorination
and hydrolysis reaction can be considered as second order kinetic reaction. It is also indicated
that the kinetic constant for dehydrochlorination reaction is far bigger than that for hydrolysis
reaction in the same conditions.
Keywords: dichloropropanol; epichlorohydrin; kinetics; dehydrochlorination; hydrolysis;
activation energy.

INTRODUCTION equations (1) and (2) (Braun, 1932, 1943;

Originally, epichlorohydrin(EPCH; IUPAC
name:1-chloro-2,3-epoypropane) was formed
by Berthelot in 1854 (Berthelot and Luca,
1856), and by Clarke and Hartman (1941),
using caustic soda for hydrolysis with the


Correspondence to: product of the reaction happened between
Professor J.W. Zhu, hydrogen chloride and crude glycerol.
Institution of Chemical
Engineering, East China
catalyst
University of Science and C3 H8 O3 þ HCl ! C3 H6 Cl2 O
Technology, Shanghai,
200237, China. þ H2 O
E-mail: jwzhu@ecust.edu.cn
DOI: 10.1205/cherd07061
C3 H6 Cl2 O þ NaOH
0263–8762/07/
$30.00 þ 0.00 ! C3 H5 ClO þ NaCl þ H2 O
Chemical Engineering
At present, energy source shortages caused
Research and Design
obtaining chemical materials from biomass,
Trans IChemE, and it was successful. Because of the high
Part A, December 2007 price of the petroleum, the glycerol process
of producing EPCH has lower cost than the
# 2007 Institution
synthesis from petroleum products. And
of Chemical Engineers
now, the process of producing EPCH from
glycerol was considered by many big chemi-
cal groups, such as SOLVAY, DOW,
DUPONT, SHANG HAI HUAYI GROUP, and
so on. Then, EPCH will be mainly prepared
in industry from the DCP, followed by
dehydrochlorination of DCP, according to
Fairbourne et al., 1930; Horsley, 1965):
The basic medium necessary for the
second reaction can be obtained by using a
caustic soda solution, afforded by Huayi
Group, Shanghai. Together with the dehydro-
chlorination reaction, hydrolysis always
occurs, according to the scheme:
EPCH þ NaOH þ H2 O
! C3 H8 O3 þ NaCl (3)
(1)
But the hydrolysis reaction lowers the yields;
EPCH must be removed as quickly as poss-
ible from the reaction environment. The dehy-
drochlorination reaction is performed in a
(2) distillation column, and EPCH is flashed out
with stream to shorten the contact time to
prevent hydrolysis.
Designing suitable reactors for dehydro-
chlorination reaction and the distillation
column and choosing the optimal operating
conditions require the determination of
kinetic parameters for the reactions involved
in the process and the evaluation of
vapour–liquid equilibrium data. But both the
kinetics of dehydrochlorination and hydrolysis
in this process are also poorly described in
literatures.
In the present paper, the kinetics of dehy-
drochlorination of DCP and the hydrolysis of

1580 Vol 85 (A12) 1580– 1585

 SYNTHESIS OF EPICHLOROHYDRIN FROM DICHLOROPROPANOLS
EPCH have been studied, both in the presence of caustic
soda.
EXPERIMENTAL
Equipment, Reagents and Operating Conditions
Both the kinetics of dehydrolorination and hydrolysis reac-
tions was studied in presence of caustic soda. Dehydrochlor-
ination of DCP in aqueous basic solution is a fast reaction
occurring with the elimination of hydrogen chloride. In a
dilute caustic soda solution, the reaction rate has been deter-
mined by using a glass complex electrode to measure the
decrease of conductivity of the solution with time according
to the method described by Jin et al. (2005). The conductivity
varies with concentration of ions in the dilute electrolyte sol-
ution, and was recorded by the conductivity recorder with
the measure accuracy of 0.01 mS cm21 and the response
time of 0.01 s, which make sure that the conductivity
measurement system has high control accuracy. The temp-
erature control device shown in Figure 1 makes sure of the
temperature-control with the measure accuracy of 0.01 K.
The dehydrochlorination reaction took place in a thermo
stated stirred batch reactor. Fifty millilitres of basic solution
containing about 0.01 mol L21 of NaOH was put into the reac-
tor. When the reaction temperature was reached, 0.25 ml of
the organic reagent was injected into the solution using a
glass syringe, and the potentiometric response was
recorded. Runs made with the two pure dichloropropanols
isomers (1,3-DCP and 2,3-DCP) have shown identical reac-
tivity. The kinetic data collected in this paper refer to an indus-
trial mixture of the two isomers mentioned. The main
component of these mixtures is 1,3-DCP, with little 2,3-DCP
only (Krafft et al., 2005).
Hydrolysis reaction was studied in a different procedure.
The reactor for hydrolysis reaction study is almost the
same as the reactor shown in Figure 1 but with different
dimensions: diameter of 100 mm, height of 100 mm, and
the diameter of the anchor-like agitator is 45 mm. The
Figure 1. Conductivity measurement device: (1) conductivity recor-
der; (2) temperature recorder; (3) thermocouple in the reactor near
the wall; (4) glass complex electrode; (5) thermocouple for the tin
bath; (6) glass syringe; (7) oil; (8) anchor-like agitator with the diam-
eter of 20 mm; (9) agitator to churn up the oil; (10) cylinder reactor
with the diameter of 50 mm, the height of 50 mm and four baffles to
strengthen the mixing effect; (11) temperature control device.
Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A12): 1580–1585
1581
temperature control device is also the same as the one
used in dehydrochlorination reaction study. Since this is a
very slow reaction, the reaction rate was determined by
using titration to measure the decrease of the hydroxyl ion
concentration as a consequence of the reaction, and the
reaction took place in a thermo stated stirred batch reactor.
Three hundred millilitres of basic solution containing about
0.2 mol L21 of NaOH was put into the reactor. When the reac-
tion temperature was reached, one equivalent of EPCH per
equivalent of base was used. Runs were made in the temp-
erature range of 322.85 K–344.25 K and in the rotation rate
range of 270 –410 r min21.
RESULTS
Dehydrochlorination Reaction
In the dehydrochlorination reaction, the hydroxyl ion con-
centration decreases rapidly during the reaction, and
caused the conductivity decrease rapidly, which was
recorded by the conductivity recorder. The experimental
results, in fact, can be well represented by assuming a
second order kinetic model, of the type:
r1 ¼ k1 ½OH  ½DCP (4)
The whole data processing was taken according to the
method mentioned by Li (2004).
Let
1
M¼
(b  a)

h 
1 ai
 ln  (k0  k/ )  (k0  kt )  (5)
kt  k/ b
The experimental data were obtained for runs made in
different operating conditions and the kinetic constants
have been calculated on the basis of the second order kinetic
model.
Figure 2 shows the values of M versus time for the dehy-
drochlorination reactions made at different temperatures.
The linearity of the curves justifies the assumption of the
second order for dehydrochlorination reaction, and we deter-
mined the constant k by the slope of the straight line.
Figure 3 shows the exponential-like variation of the kinetic
constant with the temperature, compared with Figures 7–12
in literature (Hu, 1999), we can conclude that the dehydro-
chlorination reaction obeys the general reaction kinetics.
In Figure 4, an Arrhenius plot for this reaction is also
reported. From Figure 4, one can compare the kinetic con-
stants of the reaction in different temperature range:
293 K  313 K and 313 K  333 K. According to the Arrhe-
nius equation (Laidler, 1987a; Kooij, 1893)
k ¼ AeEa=RT (6)
One can calculate the values of Ea and A for each tempera-
ture range, respectively. We can then conclude that the dehy-
drochlorination reaction rate in the presence of caustic soda
in different temperature ranges can be expressed as:
293 K  313 K:
r1 ¼ 1:77  107 e41,000=RT ½DCP½OH   (7)
1582 MA et al.

Figure 2. A plot of M versus time for the dehydrochlorination reaction
performed in the temperature range of 293 K –333 K: reaction made
in caustic soda solutions with [OH-] ¼ 0.01 mol L21, [DCP] ¼ 0.05
mol L21; the kinetic constants have been calculated assuming a
second order kinetic model.
Figure 4. The Arrhenius plot for the kinetic constants of the dehydro-
chlorination reaction: reaction made in caustic soda solutions with
[OH-] ¼ 0.01 mol L21, [DCP] ¼ 0.05 mol L21.

neglected, and the reactor could be considered as a well

313 K  333 K:
r1 ¼ 8:97  1020 e123,200=RT ½DCP½OH  
Hydrolysis Reaction
Hydrolysis of EPCH lowers the yields of the process. It is
important to know the kinetic parameters of this reaction.
At first, the influence of rotation rate on hydrolysis reaction
in the reactor was studied, and the kinetic data of the hydroly-
sis reaction obtained at different rotation rate at 333.55 K
were shown in Table 1. The rotation rate was measured by
the digital infrared tachometer.
In Table 1, the value of the kinetic constant did not vary with
the value of the rotation rate. This indicates that the influence
of rotation rate on hydrolysis reaction in the reactor could be
stirred batch reactor.
Runs were then made in the temperature range of
(8) 322.85 K  345.25 K at the rotation rate of 271 r min21. The
kinetic constant for the hydrolysis reaction can be calculated
by the determination of the decrease of the hydroxyl ion con-
centration. Both of the initial concentration of NaOH and
EPCH are 0.2314 mol L21.
The kinetic constants have been calculated on the basis of
the second order kinetic model. The experimental results can
be well represented by assuming a second order kinetic
model, of the type:
r2 ¼ k2 ½OH  ½EPCH (9)
Figure 5 shows the values of 1/CNaOH versus time for runs
made at different temperatures. The linearity of the curves
justifies the assumption of the second order for the hydrolysis

Table 1. Kinetic data of the hydrolysis reaction obtained at different

rotation rate at 333.55 K: the kinetic constants have been calculated
assuming a second order kinetic model; both of the initial concen-
tration of NaOH and EPCH are 0.2314 mol L21.

N(r min21) Time (min) [OH-] (mol L21) k (L mol21min21)

170 0.00 0.2314 0.4944

2.00 0.1871
6.00 0.1393
10.00 0.1089
14.00 0.09114
18.00 0.07565
271 0.00 0.2314 0.5054
2.750 0.1747
7.30 0.1261
11.00 0.1026
15.00 0.08350
20.60 0.06539
408 0.00 0.2314 0.4929
3.50 0.1630
Figure 3. A plot of k versus time for the dehydrochlorination reaction 8.30 0.1172
performed in the temperature range of 293 K –333 K: reaction 15.00 0.08450
made in caustic soda solutions with [OH-] ¼ 0.01 mol L21, 22.00 0.06589
[DCP] ¼ 0.05 mol L21.

Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A12): 1580–1585
 SYNTHESIS OF EPICHLOROHYDRIN FROM DICHLOROPROPANOLS
Figure 5. A plot of 1/CNaOH versus time for the EPCH hydrolysis
reaction with the rotation rate at 271 r min21: the kinetic constants
have been calculated assuming a second order kinetic model; both
of the initial concentration of NaOH and EPCH are 0.2314 mol L21.
reaction. In Figure 6, an Arrhenius plot for the kinetic constant
of hydrolysis reaction is also drawn. The kinetic parameters,
including Ea and A, for hydrolysis reaction are also calculated
from the data shown in Figure 6. Then, the kinetic expression
of EPCH hydrolysis reaction can be simplified as follows:
r2 ¼ 5:66  1010 e70 790=RT ½EPCH½OH   (10)
To prove that equation (10) could be well applied for the
description of hydrolysis reaction, another run was also con-
ducted at different conditions (both for temperature and the
initial concentration of NaOH and EPCH) from the runs
shown in Figure 5. The reaction results show in Table 2.
The hydrolysis reaction kinetic constant at 318.3 K was cal-
culated according to equation (10). And the calculated value
of k is 0.1354, which is almost the same as the experimental
value in Table 2, with the relative error of 1.81%, which is less
than 5%. This means equation (10) could be well applied for
the description of hydrolysis reaction.
Figure 6. An Arrhenius type plot for the kinetic constant k2.
Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A12): 1580–1585
1583
Table 2. Kinetic data obtained for hydrolysis reaction at 318.3 K: with
the rotation rate at 271 r min21.
Time [OH-] [EPCH] k (L mol21
T (K) (min) (mol L21) (mol L21) min21)
318.3 0 0.1097 0.1316 0.1379
1.00 0.1088 0.13078
6.00 0.09929 0.12122
11.33 0.09074 0.11267
17.30 0.08400 0.10593
22.00 0.07797 0.09990
DISCUSSION
The experiments described in this paper, allowed the deter-
mination of the kinetic parameters of the reactions involved in
the EPCH production process. The rates of reactions
involved in the progress are identical, as has been seen
experimentally.
The Boltzmann factor Ea/RT is considered as the statisti-
cal distribution of energies among the reacting molecules.
And the pre-exponential factor A represents the rate of
activated molecules converting into the product molecules
(Laidler, 1987b). Both the activation energy and pre-
exponential factor are temperature dependent. In fact, the
activation could vary with the temperature, but could be con-
sidered as changeless at a certain temperature range, and it
was also shown in Figure 4. Apparently, the rate of activated
molecules converting into product molecules at temperature
range 313 K  333 K is big than that at the temperature
range 293 K  313 K. But in the hydrolysis reaction,
the Arrhenius plot is linear; the activation energy was a
constant at the experimental temperature range.
For the EPCH formation, the ring closure reaction, occurs
according to an internal nucleophilic substitution (SN2) mech-
anism preceded by a base catalyzed dissociation equilibrium.
It is belong to the Williamson reaction (Williamson, 1852;
Ullmann and Sponagel, 1905; Dermer, 1934; Allen and
Gates, 1945; Baker and Martin, 1960; Fieser and Fieser,
1961), and is similar to the reaction happens between propy-
lene chlorohydrins and caustic soda, as described by Patai
(1967), Weissberger (1964) and Elderfield et al. (1950).
The reaction mechanism can be represented as follows:
Protonation of the hydroxyl makes a better C –O –H–O –C
group. This step is very slow and reversible. This is the rate
determining step. The lone pairs on the oxygen make it a
Lewis base. Simultaneous formation of C– O–C bond and
cleavage of the C–Cl bond allows the loss of the good leav-
ing group, a neutral water molecule.
The reaction rate, in this case is
rdehy ¼ k  Ke  ½OH  ½R (11)
where k is the kinetic constant and Ke is the equilibrium
constant for the formation of the intermediate ion. [R] is the
reagent concentration. The reaction strongly depends on
this equilibrium constant and therefore depends also on the
enthalpy change for the intermediate ion formation. The two
1584
Table 3. The ratio between the dehydrochloriantion reaction rates and the hydrolysis reaction rates.
Temperature/K 293.15 313.15
Ratio 63.55 29.14
DCP isomers which can give EPCH, have the different reac-
tivity. Apparently, in the reaction mentioned in progress, the
reactivity of 1,3-DCP is much higher than the reactivity of
2,3-DCP. This is due to the inductive effects (Clarki et al.,
1989) and space effect. In the structure figure, both halogen-
oalkyls in 1,3-DCP increased the chlorine mobility or the
negative charge on the oxygen, and the hydroxyl could
attack 1-C, and 3-C, but only one halogenoalkyl in 2,3-DCP
could increase the negative charge on the oxygen, and the
hydroxyl could only attack 2-C. At the same time, it is hard
for the hydroxyl to attack 2-C in 2,3-dicloropropanol due to
the space effect (Salaun, 2000), which is even not exist in
1,3-DCP.
So the probability of intermediate ion formation is evidently
not the same. And the activation energy value for the dehy-
drochlorination reaction in the presence of caustic soda is
equal to the true activation energy (E) plus the enthalpy
change (DHR ) in intermediate ion formation, that is:
Eapp ffi E þ DHR (12)
The hydrolysis or ring opening reaction is a base catalysed
reaction occurring according to the scheme suggested by
Patai (1967) and Sakakibara et al. (1992).
For this mechanism, kinetic law of the following type is
justified:
rhydr ¼ k 0  Ke0  ½OH  ½R (13)
where k0 is the kinetic constant, K e0 is the equilibrium
constant, and [R] is the organic reagent concentration.
Finally, according to the kinetic model obtained, that is,
equations (8) and (10), assuming that the initial concentration
of DCP and EPCH were the same, the ratio between
the dehydrochloriantion reaction rates and the hydrolysis
reaction rates were calculated, as shown in Table 3.
In Table 3, apparently, it indicates that enhancing tempera-
ture is good for improving dehydrochlorination reaction rate
when the reaction temperature is above 313.15 K, but it will
also enhance the hydrolysis reaction rate, therefore a suit-
able reaction temperature should be considered in the indus-
trial production. But for a low contact time, the problem could
be solved, because for a low contact time, the probability of
hydrolysis reaction was greatly reduced, which means that
the yields of EPCH could hardly be affected by the hydrolysis
reaction (Carra et al., 1979).
Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A12): 1580–1585
MA et al.
333.15 353.15 363.15
96.13 280.65 458.80
Although decreasing the temperature is also good for
improving dehydrochlorination reaction rate when the reac-
tion temperature is below 313.15 K, the absolute reaction
rate is low; and such kind of low temperature is not practical
in industrial producing.
CONCLUSION
(1) Both of the dehydrocholorination reaction and hydrolysis
reaction were proved to be the second order kinetic
reaction, and could be well described by the second
kinetic model, which could be applied in reactor design
and reaction simulation, obtained in this paper.
(2) In the high temperature range (313.15 K –363.15 K),
enhancing temperature is good for dehydrocholorination
reaction, but it will also enhance the hydrolysis reaction
rate.
(3) If we consider the high temperature of 363.15 K, which is
the reaction temperature in industrial process, and for low
contact time, yields of EPCH could hardly be affected by
the hydrolysis reaction. Therefore, low contact time is one
of the important factors during designing the distillation
column and choosing the optimal operating conditions
for the process of synthesis of EPCH from DCP.
NOMENCLATURE
a the initial concentration of caustic soda in
dehydrochlorination reaction, mol L21
b the initial concentration of DCP in dehydrochlorination
reaction, mol L21
A preexponential factors of dehydrochlorination reaction, 1
CNaOH concentration of caustic soda, mol L21
E true activation energy for the EPCH synthesis, J mol21
Ea apparent activation energy, J mol21
k1 kinetic constant of dehydrochlorination reaction,
L mol21s21
k2 kinetic constant of hydrolysis reaction, L mol21min21
Ke equilibrium constant for the formation of the
intermediate ion during dehydrocholorination reaction, 1
Ke0 equilibrium constant for the formation of the
intermediate ion during hydrolysis reaction, 1
mm millimeter, 1
[OH-] hydroxyl ion concentration, mol L21
[DCP] dichloropropanols concentration, mol L21
[EPCH] epichlorohydrin concertration, mol L21
r1 reaction rate of dehydrochlorination reaction,
mol L21s21
r2 reaction rate of hydrolysis reaction, mol L21min21
rdehy reaction rate of dehydrochlorination reaction,
mol L21s21
rhydr reaction rate of hydrolysis reaction, mol L21s21
R general constant of gases, J mol21K21
t time, min or s
T absolute temperature, K
Greek symbols
kt the conductivity of solution in dehydrochlorination
reaction at any time, mS cm21
k0 the initial conductivity of solution in dehydrochlorination
reaction, mS cm21
 SYNTHESIS OF EPICHLOROHYDRIN FROM DICHLOROPROPANOLS
k/ the infinite conductivity of solution in
dehydrochlorination reaction, mS cm21
DHR enthalpy change in intermediate ion formation during
the synthesis reaction, KJ mol21
REFERENCES
Allen, C.F.H. and Gates, J.W., 1945, a-Butoxynitrobenzene, Organic
Syntheses, 25: 9.
Baker, R.H. and Martin, W.B., 1960, Side reaction in the Williamson
synthesis. II, J Org Chem, 25: 1496.
Berthelot and Luca, 1856, Ann Chem, 48(3): 306, 311.
Braun, G. 1932, Preparation of epichloro- and epibromohydrins, J Am
Chem Soc, S4: 1248–1250.
Braun, G., 1943, Glycerol a,g-Dibromohydrin, Organic Syntheses,
Coll., 2: 308.
Carra, S., Santacesaria, E., Morbidelli, M. and Cavalli, L., 1979,
Synthesis of propylene oxide from propylene chlorohydins—I,
Chem Eng Sci, 34: 1123–1132.
Clark, S.A.C., Bawagan, A.O. and Brion, C.E., 1989, The valence
orbital momentum distributions and binding energy spectra of
dimethyl ether by electron momentum spectroscopy: an investi-
gation of the methyl inductive effect[J]. Chemical Physics, 137:
407.
Clarke, H.T. and Hartman, W.W., 1941, Org Syn Coll, 1: 233.
Claus, 1877, Chem Ber, 10: 557.
Dermer, O.C., 1934, Metallic salts of alcohols and alcohol analogs,
Chem Rev, 14: 409.
Elderfield, R.C. et al., 1950, Synthesis of branched-chain d-
isopropylaminoamyl and 4-isopropylaminobutyl ethers and of
derived bromides, J Amer Chem Soc, 72: 1334.
Fairbourne, Gibson and Stephens, 1930, Chemistry and Industry, 49:
1021.
Fieser, L.F. and Fieser, M., 1961, Advanced Organic Chemistry, 113,
306 (Reinhold, New York, USA).
Horsley, H.L., 1965, Encyclopedia of Chemical Technology, in
Othmer, K. (ed.), 595 (Wiley, New York, USA).
Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A12): 1580–1585
1585
Hu, Y., 1999, Physical Chemistry, 3rd edition, 312 (Higher Education
Press, Shanghai, China).
Jin, L., Wu, S. and Chen, D., 2005, Physical chemistry experi-
ments[M]. Shanghai, East China University of Science and
Technology Press, 1: 55– 58.
Kooij, D.M., 1893, Z Phys Chem, 12: 155.
Krafft, P., Gilbeau, P., Gossenlin, B. and Claessens, S., 2005,
Process for producing dichloropropanlol from glycerol-the
glycerol coming eventually from the conversion of animal fats
in the manufacture of biodiesel[P], SOLVAY, WO/2005/054167,
7– 10.
Laidler, K.J., 1987a, Chemical Kinetics, 3rd edition, 42 (Harper &
Row, New York, USA).
Laidler, K.J., 1987b, Chemical Kinetics, 3rd edition, 48 (Harper &
Row, New York, USA).
Li, S., 2004, Measurement of reaction rate constant in ethyl acetate
saponification, Journal of Luoyang Teachers College, 2: 54– 55.
Patai, S., 1967, The Chemistry of Functional Groups, 43–49 (Inter-
science Publishers, Wiley, New York, USA).
Sakakibara, M., Katsumata, K., Watanabe, Y., Toru, T. and Ueno, Y.,
1992, Synthesis, 4: 377.
Salaun, J., 2000, Small Ring Compounds in Organic SynthesisV1
(Springer-Verlag, Berlin, Germany).
Ullmann, F. and Sponagel, P., 1905, Chem Ber, 38: 2211.
Weissberger, A., 1964, The Chemistry of Heterocyclic Compounds,
180, 210 (Interscience, New York, USA).
Williamson, A.W., 1852, J Chem Soc, 4: 229.
ACKNOWLEDGEMENT
The authors wish to acknowledge the financial support by Shanghai
Huayi (Group) Company.
The manuscript was received 20 April 2007 and accepted for
publication after revision 16 August 2007.