Investigation of the kinetics and mechanism of the glycerol chlorination reaction using gas chromatography-mass spectrometry

Article abstract of DOI:10.2298/JSC1001101L

As a primary by-product in biodiesel production, glycerol can be used to prepare an important fine chemical, epichlorohydrin, by the glycerol chlorination reaction. Although this process has been applied in industrial production, unfortunately, less attention has been paid to the analysis and separation of the compounds in the glycerol chlorination products. In this study, a convenient and accurate method to determine the products in glycerol chlorination reaction was established and based on the results the kinetic mechanism of the reaction was investigated. The structure of main products, including 1,3-dichloropropan-2-ol, 2,3-dichloropropan-1-ol, 3-chloro-1,2-propanediol, 2-chloro-1,3-propanediol and glycerol was ascertained by gas chromatography-mass spectrometry and the isomers of the products were distinguished. Apidic acid was considered as the best catalyst because of its excellent catalytic effect and high boiling point. The mechanism of the glycerol chlorination reaction was proposed and a new kinetic model was developed. Kinetic equations of the process in the experimental range were obtained by data fitting and the activation energies of each tandem reaction were 30.7, 41.8, 29.4 and 49.5 kJ mol^-1, respectively. This study revealed the process and mechanism of the kinetics and provides the theoretical basis for engineering problems. 2009 copyright (CC) SCS.

Full text of DOI:10.2298/JSC1001101L

J. Serb. Chem. Soc. 75 (1) 101–112 (2010)
UDC 547.426.1:543.544.3:543.51:531.3+541.124
JSCS–3945
Original scientific paper
Investigation of the kinetics and mechanism of the
glycerol chlorination reaction using gas
chromatography–mass spectrometry
1
1
XIUQUAN LING1, DINGQIANG LU1,2 , JUN WANG , MINGXIN LIANG ,
*
SHUMIN ZHANG1, WEI REN1, JIANHUI CHEN1 and PINGKAI OUYANG1
1State Key Laboratory of Materials-Oriented Chemical Engineering, College of Life Science
and Pharmaceutical Engineering, Nanjing University of Technology, Nanjing 210009 and
2Jiangsu Provincial Institute of Materia Medica, Nanjing 210009, China
(Received 21 November 2008, revised 11 June 2009)
Abstract: As a primary by-product in biodiesel production, glycerol can be
used to prepare an important fine chemical, epichlorohydrin, by the glycerol
chlorination reaction. Although this process has been applied in industrial pro-
duction, unfortunately, less attention has been paid to the analysis and separa-
tion of the compounds in the glycerol chlorination products. In this study, a
convenient and accurate method to determine the products in glycerol chlori-
nation reaction was established and based on the results the kinetic mechanism
of the reaction was investigated. The structure of main products, including 1,3-
-dichloropropan-2-ol, 2,3-dichloropropan-1-ol, 3-chloro-1,2-propanediol, 2-chlo-
ro-1,3-propanediol and glycerol was ascertained by gas chromatography–mass
spectrometry and the isomers of the products were distinguished. Apidic acid
was considered as the best catalyst because of its excellent catalytic effect and
high boiling point. The mechanism of the glycerol chlorination reaction was
proposed and a new kinetic model was developed. Kinetic equations of the
process in the experimental range were obtained by data fitting and the activa-
tion energies of each tandem reaction were 30.7, 41.8, 29.4 and 49.5 kJ mol-1,
respectively. This study revealed the process and mechanism of the kinetics
and provides the theoretical basis for engineering problems.
Keywords: glycerol; monochloropropanediol; dichloropropanol; chlorination
reaction; gas chromatography–mass spectrometry; kinetic model.
INTRODUCTION
Epichlorohydrin is an important fine chemical, which is widely used to pre-
1,2
pare organic chemical raw materials such as epoxy resin. In China, the pro-
duction capacity of epichlorohydrin had reached 497 thousand tons until 2008
*Corresponding author. E-mail: ludingqiang@njut.edu.cn
doi: 10.2298/JSC1001101L
101
Available online at www.shd.org.rs/JSCS/
___________________________________________________________________________________________________________________________
2009 Copyright (CC) SCS

102
LING et al.
and it is estimated that demand for it will have reached 700 thousand tons in
2012. Currently, the methods to produce epichlorohydrin include the high-tempe-
3,4
rature chlorination of propylene route and the allyl acetate route, which depend
on the petroleum industry. Then, with the petroleum energy crisis in recent years
and the soar of oil prices, the price of epichlorohydrin on the international market
rose. Due to the increase in the production of the biodiesel industry in recent
years and the consequential increase in its by-product glycerol (one tenth of the
yield of biodiesel), there is an oversupply of glycerol and a slump in its price on
the market. The efficient and reasonable utilization of glycerol has become a bot-
tleneck problem for the healthy development of the biodiesel industry chain. Di-
chloropropanol (DCP), as the raw material of epichlorohydrin production, can be
produced by the reaction between glycerol and hydrogen chloride. Thus, the pro-
cess of preparing epichlorohydrin from glycerol would allow mankind to be less
dependent on petroleum and promote the development of the biomass energy in-
dustry, which has great economic and social values.
The most important step in the synthesis of DCP from glycerol is the chlori-
5,6
nation reaction glycerol. This process has been introduced on the industrial
7–9
scale
but, unfortunately, less attention has been paid to the analysis and sepa-
ration of the compounds produced in the reaction. Therefore, there is an urgent
need for the development of an accurate method to determine the compositions in
this complex reaction system. To date, the primary method for analyzing mono-
10
chloropropanediol (MCP) and DCP is gas chromatography. Schuhmacher et al.
and Crews et al. determined 1,3-dichloropropan-2-ol (1,3-DCP) by gas chroma-
11,12
tography–mass spectrometry (GC–MS).
Furthermore, Boden et al. and Chung
et al. both reported methods for the simultaneous analysis of 3-chloro-1,2-propa-
13,14
nediol (3-MCP) and 1,3-DCP by GC–MS.
However, the above-mentioned
methods not only required the samples to be derivatized before analysis, but also
they cannot recognize the isomers present in the reaction system. In addition, the
formation of the intermediate MCP complicates the reaction system and causes
great difficulties for further research on the dynamics of the reaction. In this stu-
dy, a GC–MS method that can simultaneously ascertain the composition, inclu-
ding glycerol, 3-MCP, 2-chloro-1,3-propanediol (2-MCP), 1,3-DCP and 2,3-di-
chloropropan-2-ol (2,3-DCP), in the chlorination reaction system was developed
without the necessity of derivatization, which effectively simplified the analysis
process and enabled the isomers in products to be distinguished, thus providing a
fast and convenient method for further study of the dynamics of the reaction.
Hitherto, there have been only a few reports concerning the kinetics of the
15
glycerol chlorination reaction. Siano et al. found that propionic acid was the
best catalyst for this reaction, although its boiling point is only slightly higher
than that of acetic acid, which is used in the traditional process. They considered
that an oxonium group was formed during the glycerol chlorination process and
Available online at www.shd.org.rs/JSCS/
___________________________________________________________________________________________________________________________
2009 Copyright (CC) SCS

GLYCEROL CHLORINATION KINETICS BY GC–MS
103
16
deduced a dynamic model of the chlorination tandem reactions. However, there
were some defects in their hypothesis and the model cannot completely accura-
tely describe the glycerol chlorination process. In this study, the chlorination re-
action model has been improved, which reflected the mechanism of the reaction
and offered a theoretical basis for the industrial production of epichlorohydrin.
EXPERIMENTAL
Materials and instruments
All chemicals used in the present work, viz., glycerol, acetic acid, propanoic acid, malo-
nic acid, succinic acid and apidic acid (all purchased from Ludu, China) were of analytical
reagent (A.R.) grade.
Gas chromatographic and mass spectrometric analysis
The GC–MS analyses were performed using a CP 3800-Saturn 2200 gas chromatograph–
–mass spectrometry instrument (Varian, Middelburg, The Netherlands). The GC analyses were
performed using an SP-6890 gas chromatography instrument (Lunan Ruihong, Shandong,
China), equipped with a KR-9 capillary column (30 m×0.32 mm×1 μm). The injector and fla-
me ionization detector (FID) temperature were 200 and 280 °C, respectively. The oven tempe-
rature was held at 190 °C; N2 was the carrier gas (1.1 mL min-1). The injected volume was
0.60 μL with the split ratio set at 60:1.
Experimental apparatus
Glycerol chlorination reaction experiments were realized in a self-designed glass appara-
tus shown in Fig. 1. Glycerol and catalyst were fed into the glass-jacketed reactor, and the
external circulation oil bath controller maintained the reaction mixture at the predetermined
temperature. Then hydrogen chloride gas, previously dried using a gas dryer, was introduced
into the reactor. A porous glass fritter and strong mechanical stirring assured that the gas–li-
quid interface contacted well. The excess hydrogen chloride gas was fed through a protection
bottle and absorbed by alkali liquor in the exit gas absorption bottle. Reaction mixture
samples were withdrawn through a valve at the bottom of the reactor for gas chromatographic
analysis.
Fig. 1. Schematic diagram of glyce-
rol chlorination reaction apparatus.
1. hydrogen chloride cylinder; 2. gas
dryer; 3. reactor; 4. porous fritter; 5.
mechanical stirrer; 6. circulation oil
bath controller; 7. condenser; 8. pro-
tection bottle; 9. exit gas absorber;
10. sampling valve.
Available online at www.shd.org.rs/JSCS/
___________________________________________________________________________________________________________________________
2009 Copyright (CC) SCS

104
LING et al.
RESULTS AND DISCUSSION
The total ion chromatogram of a sample obtained under the optimal chroma-
tographic conditions is shown in Fig. 2. It depicts that there were five compo-
nents in the glycerol chlorination products and all components in the sample were
well separated during 6 min without additional derivatization. These five compo-
nents were detected with the mass spectrum detector and the mass spectrograms
are shown in Fig. 3.
Fig. 2. Total ion chromatogram of the products of glycerol chlorination.
1,3-DCP and 2,3-dichloropropan-1-ol (2,3-DCP) are isomers and, having the
same molecular weight, great difficulties are encountered in distinguishing them.
Generally, because molecules of alcohols are often fragmented completely under
electron impact (EI), the molecular ion peaks of 1,3-DCP and 2,3-DCP can hard-
ly appear in their mass spectrogram. According to the laws of alcohol β-cracking
and halide substituent rupture, the molecule of 1,3-DCP may produce a fragment
ion with m/z 79 under EI due to the hydroxyl group linking with the β-carbon,
while that of 2,3-DCP may produce an m/z 62 fragment peak due to the hydroxyl
group linking with the α-carbon. On the other hand, m/z 81 and 64 fragment ion
peaks, with one-third of the relative abundance of the m/z 79 and 62 peaks, could
appear next to these two peaks, respectively, because of the existence of the iso-
37
tope Cl in the sample. The possible fragmentation pathways of 1,3-DCP and
Available online at www.shd.org.rs/JSCS/
___________________________________________________________________________________________________________________________
2009 Copyright (CC) SCS

GLYCEROL CHLORINATION KINETICS BY GC–MS
105
2,3-DCP are shown in Fig. 4. The fragment ion peaks m/z 81 and 79 are both pre-
sent in the mass spectrogram of component a while fragment ion peaks m/z 64
and 62 are present in the spectrogram of component b, Fig. 3. Hence, it can be
inferred that component a in the total ion chromatogram is 1,3-dichloropropan-2-ol
and component b is 2,3-dichloropropan-1-ol.
Fig. 3. Mass spectrograms of components a and b.
Available online at www.shd.org.rs/JSCS/
___________________________________________________________________________________________________________________________
2009 Copyright (CC) SCS

106
LING et al.
Fig. 3 continued. Mass spectrograms of components c, d and e.
Available online at www.shd.org.rs/JSCS/
___________________________________________________________________________________________________________________________
2009 Copyright (CC) SCS

GLYCEROL CHLORINATION KINETICS BY GC–MS
107
Fig. 4. Fragmentation pathways of 1,3-dichloropropan-2-ol and
2,3-dichloropropan-1-ol under electron impact.
3-MCP and 2-chloro-1,3-propanediol (2-MCP) are also isomers and neither
of their molecular ion peaks can appear in the EI mass spectrogram. The mole-
cule of 2-MCP may produce a fragment ion with m/z 62 under EI due to hydroxyl
group and chlorine atom linking with their α-carbon and β-carbon respectively,
while that of 3-MCP may produce a m/z 79 fragment ion due to hydroxyl group
and chlorine atom linking with their α-carbon and γ-carbon, respectively, accord-
ing to the laws of alcohol β-cracking and halide substituent rupture. In addition,
m/z 81 and 64 fragment ion peaks, with one-third of the relative abundance of the
m/z 79 and 62 peaks, may appear next to these two peaks, respectively, also be-
37
cause of the existence of the Cl isotope in the sample. The possible fragmen-
tation pathways of 3-MCP and 2-MCP are shown in Fig. 5. The fragment ion
peaks m/z 79 and 81 are both present in the mass spectrogram of component c
while the fragment ion peaks m/z 62 and 64 are both present in the mass spectro-
gram of component d. Hence it can be inferred that component c in the total ion
chromatogram is 3-chloro-1,2-propanediol and component d is 2-chloro-1,3-pro-
panediol.
There are three hydroxyl groups in the molecule of glycerol meaning that the
molecular ion peak must be absent in its EI mass spectrogram. The molecule of
glycerol may produce a fragment ion with m/z 61 under EI according to the law
of alcohol β-cracking. The possible fragmentation pathway of glycerol is shown
in Fig. 6. These fragment ion peaks are all present in the mass spectrograms of
component e. Hence, it can be inferred that component e in the total ion chroma-
togram is glycerol.
Tesser et al. indicated that there was no relationship between the acidity
15
strength of a catalysts and its catalytic activity, while Phillipe et al. proposed
17
that a variety of carboxylic acid could catalyze the chlorination reaction. In the
Available online at www.shd.org.rs/JSCS/
___________________________________________________________________________________________________________________________
2009 Copyright (CC) SCS

108
LING et al.
present study, different kinds of lower carboxylic acids, such as acetic acid,
propanoic acid, malonic acid, succinic acid and adipic acid, were utilized as ca-
talysts in the experiments. A comparison of the catalytic effect of a variety of ca-
talysts is depicted in Fig. 7, from which it can be seen that acetic acid, propanoic
acid and adipic acid displayed better catalytic effects that the other investigated
catalysts. However, the low boiling point of acetic acid and propanoic acid caused
severe volatilization loss of these acids, which lowered the rate of the reaction.
To overcome these shortcomings, apidic acid was selected as the chlorination catalyst.
Fig. 5. Fragmentation pathways of 3-chloro-1,2-propanediol and
2-chloro-1,3-propanediol under electron impact.
Fig. 6. Fragmentation pathways of glycerol under electron impact.
According to the reaction products glycerol chlorination determined by GC–
MS, the net tandem reaction can be schematized in the following manner:
(1)
Available online at www.shd.org.rs/JSCS/
___________________________________________________________________________________________________________________________
2009 Copyright (CC) SCS

GLYCEROL CHLORINATION KINETICS BY GC–MS
109
Fig. 7. Comparison of a variety of catalysts on the conversion of dichloropropanol.
The evolution of each component in the reaction product under the optimum
condition is depicted in Fig. 8. It can be observed that the concentration of 2-MCP
hardly increased when glycerol was still present in the reaction mixture. After a
sufficiently long time, the amount of glycerol decreased to a low constant value.
Tesser et al. considered that conversion of 2-MCP to 2,3-DCP could be neglect-
ted, namely reaction path 5 (Eq. (1)) does not occur. Reaction paths 2 and 4 (Eq.
(1)) can be considered as being irreversible, due to the low accumulation of 2-
15
MCP and 2,3-DCP throughout the whole reaction process. Moreover, reaction
1 can also be considered as being irreversible because rate of reaction 1 from
glycerol to 3-MCP is very high and glycerol can finally be completely converted.
According to these hypotheses, the glycerol chlorination reaction model can be
modified to the following:
(2)
Thus, the kinetic model can be reduced to the following differential equa-
tions:
Available online at www.shd.org.rs/JSCS/
___________________________________________________________________________________________________________________________
2009 Copyright (CC) SCS

110
LING et al.
dc
dc
dc
3
1
2
= −k c − k c ;
= k c − k c + k c − k c ;
= k c
2 1
1 1
2 1
1 1
3 2
5 4
4 2
dt
dc
dt
dc
dt
(3)
4
5
= k c − k c ;
= k c
3 2
5 4
4 2
dt
dt
Fig. 8. Evolution of the composition in the glycerol chlorination process
under optimum conditions.
The kinetic constants at various temperatures were calculated by non-linear
regression in Matlab, based on the kinetic model and the data of the time evo-
lution of the composition, reported in Table I.
TABLE I. Rate constants of the positive reactions at different temperatures
t / °C
k1×102 / min-1
k2×104 / min-1
k3×103 / min-1
k4×105 / min-1
90
1.23
1.35
2.01
2.56
3.16
4.59
6.59
9.07
2.42
4.18
5.39
5.03
3.12
8.70
11.10
11.37
100
110
120
Then, according to the Arrhenius equation:
E
a
(4)
ln k = −
+ b
RT
the slopes, namely the activation energy of the tandem reactions, were deter-
mined by plotting –ln k as the ordinate against 1/RT as the abscissa. The obtained
Available online at www.shd.org.rs/JSCS/
___________________________________________________________________________________________________________________________
2009 Copyright (CC) SCS

GLYCEROL CHLORINATION KINETICS BY GC–MS
111
–1
–1
–1
values were: E (1) = 30.7 kJ mol , E (2) = 41.8 kJ mol , E (3) = 29.4 kJ mol
a
a
a
–1
and E (4) = 49.5 kJ mol . The kinetic constants were then introduced into kine-
a
tic equations and the results were in good agreement with experimental data, as
shown in Fig. 9, which demonstrates that this kinetic model can predict the chlo-
rination process behavior very well.
Fig. 9. Comparison between the experimental data and the behavior
predicted by the proposed model.
CONCLUSIONS
A convenient and accurate gas chromatography–mass spectrometry method
that can simultaneously determine the composition of glycerol chlorination pro-
ducts was first established. The possibility of distinguishing the isomers of mono-
chloropropanediol and dichloropropanol in the reaction products from the mass
spectra of the individual products provided the basis for further study of the reac-
tion kinetics and industrial production.
The dynamic behavior of the glycerol chlorination reaction was investigated
and a new dynamic model was proposed. According to regression fitting of the
experimental data, kinetic equations were obtained and the activation energy of
–1
each positive tandem reaction was calculated as follows: E (1) = 30.7 kJ mol ,
a
–1
–1
–1
E (2) = 41.8 kJ mol , E (3) = 29.4 kJ mol and E (4) = 49.5 kJ mol . The fit-
a
a
a
ting curves were in good agreement with the experimental data.
Acknowledgements. This work was supported by the National Natural Science Foun-
dation of China (20676060) and National Basic Research Program of China (2009CB724700).
Available online at www.shd.org.rs/JSCS/
___________________________________________________________________________________________________________________________
2009 Copyright (CC) SCS

112
LING et al.
И З В О Д
ИСПИТИВАЊЕ КИНЕТИКЕ И МЕХАНИЗМА РЕАКЦИЈЕ ХЛОРОВАЊА ГЛИЦЕРОЛА
ГАСНО–МАСЕНОМ СПЕКТРОМЕТРИЈОМ
XIUQUAN LING1, DINGQIANG LU1,2, JUN WANG1, MINGXIN LIANG1, SHUMIN ZHANG1,
JIANHUI CHEN1 и PINGKAI OUYANG1
1State Key Laboratory of Materials-Oriented Chemical Engineering, College of Life Science and Pharma-
ceutical Engineering, Nanjing University of Technology, Nanjing 210009 и 2Jiangsu Provincial
Institute of Materia Medica, Nanjing 210009, China
Као примарни споредни производ у производњи биодизела, глицерол се може употре-
бити за синтезу важне супстанце епихлорхидрина, који настаје као интермедијерни производ
у процесу хлоровања глицерола. Мада се тај процес користи у индустријским условима,
мала пажња је била поклоњена анализи и сепарацији хлорираних производа реакције. У
овом раду је испитана кинетика и механизам реакције и описан одговарајући и прецизан
метод одређивања производа реакције. Применом гасно–масене спектрометрије утврђено је
да производ реакције садржи 1,3-дихлоро 2-пропанол, 2,3-дихлоро 1-пропанол, 3-хлоро-1,2-
-пропандиол, 2-хлоро-1,3-пропандиол и глицерол. Због високе тачке кључања и добрих ката-
литичких особина адипинска киселина се показала као најбољи катализатор у овој реакцији.
Предпостављен је механизам и развијен је нови кинетички модел ове реакције. Фитовањем
експерименталних података добијене су следеће вредности за енергије активације поједи-
начних ступњева: 30,7, 41,8, 29,4 и 49,5 kJ mol-1. Овај рад, разјашњавајући кинетику и меха-
низам процеса, поставља теоријску основу за инжењеризацију овог процеса.
(Примљено 21. новембра 2008, ревидирано 11. јуна 2009)
REFERENCES
1. S. H. Lee, D. R. Park, H. Kim, J. Lee, J. C. Jung, S. Y. Woo, W. S. Song, M. S. Kwon, I.
K. Song, Catal. Commun. 9 (2008) 1920
2. G. Lewandowski, M. Bartkowiak, E. Milchert, Oxid. Commun. 31 (2008) 108
3. B. M. Bell, J. R. Briggs, R. M. Campbell, S. M. Chambers, P. D. Gaarenstroom, J. G.
Hippler, B. D. Hook, K. Kearns, J. M. Kenney, W. J. Kruper, D. J. Schreck, C. N. The-
riault, C. P. Wolfe, Clean-Soil Air Water 36 (2008) 657
4. S. H. Lee, D. R. Park, H. Kim, J. Lee, J. C. Jung, S. Park, K. M. Cho, I. K. Song, React.
Kinet. Catal. Lett. 94 (2008) 71
5. E. C. Britton, R. L. Heindel, US 2,144,612 (1939)
6. E. C. Britton, H. R. Slagh, US 2,198,600 (1940)
7. D. J. Schreck, W. J. Kruper Jr., R. D. Varjian, M. E. Jones, R. M. Campbell, K. Kearns,
B. D. Hook, J. R. Briggs, J. G. Hippler, WO 2006,020,234 (2006)
8. P. Kubicek, P. Sladek, I. Buricova, WO 2005,021,476 (2005)
9. P. Krafft, C. Franck, I. De Andolenko, R. Veyrac, WO 2007,054,505 (2007)
10. J. Gaca, G. Wejnerowska, Anal. Chim. Acta 540 (2005) 55
11. R. Schuhmacher, J. Nurmi-Legat, A. Oberhauser, M. Kainz, R. Krska, Anal. Bioanal.
Chem. 382 (2005) 366
12. C. Crews, G. Le Brun, P. A. Brereton, Food Addit. Contam. 19 (2002) 343
13. L. Boden, M. Lundgren, K. E. Stensio, M. Gorzynski, J. Chromatogr. A 788 (1997) 195
14. W. C. Chung, K. Y. Hui, S. C. Cheng, J. Chromatogr. A 952 (2002) 185
15. R. Tesser, E. Santacesaria, M. Di Serio, G. Di Nuzzi, V. Fiandra, Ind. Eng. Chem. Res. 46
(2007) 6456
16. D. Siano, E. Santacesaria, V. Fiandra, R. Tesser, G. Di Nuzzi, M. Di Serio, WO
2006,111,810 (2006)
17. K. Phillipe, G. Patrick, G. Benoit, C. Sara, WO 2005,054,167 (2005).
Available online at www.shd.org.rs/JSCS/
___________________________________________________________________________________________________________________________
2009 Copyright (CC) SCS

Copyright of Journal of the Serbian Chemical Society is the property of National Library of Serbia and its
content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for individual use.