The synthesis of butene glycol ethers with aluminium triflate

Article abstract of DOI:10.1139/V06-086

The use of aluminium triflate as a ring-opening catalyst for butene oxide (BuO) was evaluated in the presence of different alcohols such as methanol, ethanol, n-propanol, n-butanol, 2-propanol, 2-methyl-1-propanol, and 2-methyl-2-propanol. The reaction with methanol was studied kinetically by varying the temperature, catalyst concentration, and methanol - butene oxide molar ratio. These reactions yielded two major products (2-methoxy-1-butanol and 1-methoxy-2-butanol) in a approximate ratio of 1:1. It was noted that at low catalyst concentrations (<5 ppm), low temperatures (<90 °C), and a MeOH-BuO molar ratio of 8:1, the selectivity of the reaction could be kinetically manipulated to shift the product ratio towards 1-methoxy-2-butanol, the α-alkoxyalcohol. This result was confirmed by an experimental design program. Statistical calculations using the data from the experimental design identified a feasible region in which reactions with methanol could be carried out, which would lead to slightly higher selectivities to 1-methoxy-2-butanol. This region shows that the methanol - butene oxide ratio should be 8:1, the temperature between 80 and 85 °C, and the catalyst concentration between 3.9 and 5 ppm. These reaction conditions were used to carry out a test reaction with methanol and an extended series of alcohols. All the alcohols, except for 2-methyl-2-propanol, reacted with butene oxide under these conditions, with the selectivity to the α-alkoxyalcohol higher than to the β- alkoxyalcohol. To obtain a ring-opening reaction with 2-methyl-2-propanol, it was found that a higher catalyst concentration (approximately 10 ppm) and a lower alcohol - butene oxide ratio (6:1) at a temperature of 80 °C were necessary. This reaction led to a mixture of 1-tert-butoxy-2-butanol and 2-tert-butoxy-1-butanol with the selectivity to the α-alkoxyalcohol being somewhat higher because of the steric influence of the bulky tert-butoxy group.

Full text of DOI:10.1139/V06-086

8
59
The synthesis of butene glycol ethers with
aluminium triflate
Yvette M. Terblans, Michelle Huyser, Desmond A. Young, and Michael J. Green
Abstract: The use of aluminium triflate as a ring-opening catalyst for butene oxide (BuO) was evaluated in the pres-
ence of different alcohols such as methanol, ethanol, n-propanol, n-butanol, 2-propanol, 2-methyl-1-propanol, and
2
-methyl-2-propanol. The reaction with methanol was studied kinetically by varying the temperature, catalyst concen-
tration, and methanol – butene oxide molar ratio. These reactions yielded two major products (2-methoxy-1-butanol
and 1-methoxy-2-butanol) in a approximate ratio of 1:1. It was noted that at low catalyst concentrations (<5 ppm), low
temperatures (<90 °C), and a MeOH–BuO molar ratio of 8:1, the selectivity of the reaction could be kinetically manip-
ulated to shift the product ratio towards 1-methoxy-2-butanol, the α-alkoxyalcohol. This result was confirmed by an
experimental design program. Statistical calculations using the data from the experimental design identified a feasible
region in which reactions with methanol could be carried out, which would lead to slightly higher selectivities to
1
8
-methoxy-2-butanol. This region shows that the methanol – butene oxide ratio should be 8:1, the temperature between
0 and 85 °C, and the catalyst concentration between 3.9 and 5 ppm. These reaction conditions were used to carry out
a test reaction with methanol and an extended series of alcohols. All the alcohols, except for 2-methyl-2-propanol,
reacted with butene oxide under these conditions, with the selectivity to the α-alkoxyalcohol higher than to the β-
alkoxyalcohol. To obtain a ring-opening reaction with 2-methyl-2-propanol, it was found that a higher catalyst concen-
tration (approximately 10 ppm) and a lower alcohol – butene oxide ratio (6:1) at a temperature of 80 °C were neces-
sary. This reaction led to a mixture of 1-tert-butoxy-2-butanol and 2-tert-butoxy-1-butanol with the selectivity to the
α-alkoxyalcohol being somewhat higher because of the steric influence of the bulky tert-butoxy group.
Key words: aluminium triflate, alcohols, butene oxide, ring opening, α-alkoxyalcohol, β-alkoxyalcohol.
Résumé : On a évalué l’utilité du triflate d’aluminium comme catalyseur pour effectuer l’ouverture du cycle de
l’oxyde de butène (BuO) en présence de divers alcools, dont le méthanol, l’éthanol, le n-propanol, le n-butanol, le
2
-propanol, le 2-methyl-1-propanol et le 2-methyl-2-propanol. On a étudié la cinétique de la réaction avec le méthanol
en faisant varier la température, la concentration de catalyseur et le rapport molaire méthanol – oxyde de butène. Ces
réactions conduisent à la formation de deux produits principaux (le 2-méthoxybutan-1-ol et le 1-méthoxy-2-butanol)
dans un rapport approximatif de 1 : 1. On a noté que, à de faibles concentrations de catalyseurs (<5 ppm), de faibles
températures (<90 °C) et un rapport molaire MeOH–BuO de 8 : 1, la sélectivité de la réaction peut être manipulée
d’une façon cinétique pour déplacer le rapport des produits en faveur du le 1-méthoxy-2-butanol, le α-alkoxyalcool. Ce
résultat a été confirmé par un programme de modélisation expérimental. Des calculs statistiques à l’aide de données de
modélisations expérimentales ont permis d’identifier une région appropriée dans laquelle les réactions avec le méthanol
pourraient se produire pour conduire à des sélectivités légèrement supérieures en le 1-méthoxy-2-butanol. Ces calculs
montrent que le rapport méthanol – oxyde de butène devrait être de 8 : 1, la température entre 80 et 85 °C et la
concentration en catalyseur entre 3,9 et 5 ppm. On a utilisé ces conditions pour évaluer la réaction avec le méthanol et
une série importante d’alcools. Dans ces conditions, tous les alcools réagissent avec l’oxyde de butène, à l’exception de
l’alcool tert-butylique, et dans tous les cas la sélectivité favorise le α-alkoxyalcool par rapport au β-alkoxyalcool. Pour
obtenir une réaction d’ouverture de cycle avec le 2-methyl-2-propanol, il est nécessaire d’utiliser une concentration en
catalyseur plus élevée (environ 10 ppm) et un rapport alcool – oxyde de butène plus faible (6 : 1), à une température
de 80 °C. Cette réaction conduit à un mélange de 1-tert-butoxy-2-butanol et de 2-tert-butoxy-1-butanol dans lequel la
sélectivité en faveur du α-alkoxyalcool est légèrement supérieure en raison de l’influence stérique du groupe tert-bu-
toxy.
Mots clés : triflate d’aluminium, oxyde de butène, ouverture de cycle, α-alkoxyalcool, β-alkoxyalcool.
[
Traduit par la Rédaction] Terblans et al. 866
Received 14 February 2006. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 9 June 2006.
1
Y.M. Terblans, M. Huyser, D.A. Young, and M.J. Green. Homogeneous Catalysis and Organic Synthesis, SASOL Technology
(
Pty) Ltd., Research and Development Division, SASOL, 1 Klasie Havenga Road, Sasolburg, 1947, South Africa.
1Corresponding author (e-mail: yvette.terblans@sasol.com).
Can. J. Chem. 84: 859–866 (2006)
doi:10.1139/V06-086
© 2006 NRC Canada

8
60
Can. J. Chem. Vol. 84, 2006
Scheme 1. Ring opening of butene oxide with an alcohol.
O
Al(CF SO )
3 3
β
α
β
OH
3
α
HO
β
RO
+
+
R
α
OH
OR
Butene oxide
Alcohol
α-Alkoxyalcohol
β-Alkoxyalcohol
Introduction
α-alkoxyalcohol, whereas the opposite was observed for
higher temperatures and increased catalyst concentrations.
These results were verified by statistical calculations on data
obtained from an experimental design program.
Ring-opening reactions of epoxides with alcohols, which
have been successfully carried out for many decades, can be
either base- or acid-catalyzed (1). In recent years the use of
triflate salts (also known as trifluoromethane sulfonate salts)
has become increasingly popular as catalysts for these reac-
tions. The catalytic acitivity of rare earth triflates (e.g.,
Eu(CF SO ) , La(CF SO ) , Lu(CF SO ) , Nd(CF SO ) ,
Experimental
Unless otherwise stated, all chemicals used in this study
were purchased from Sigma-Aldrich and used without fur-
ther purification. Aluminium triflate was prepared as de-
3
3 3
3
3 3
3
3 3
3
3 3
Sm(CF SO ) , Sc(CF SO ) , Y(CF SO ) , and Ce(CF SO ) )
3
3 3
3
3 3
3
3 3
3
3 4
scribed in the literature (12). Stock solutions of Al(CF SO )
has been investigated for the ring opening of epoxides by al-
cohols (2). Scandium triflate was found to be an exception-
ally active catalyst for the ring opening of epoxides.
Yb(CF SO ) has also been reported as a catalyst for inter-
3
3 3
(
500 ppm) were prepared in each alcohol and the desired
quantity thereof was weighed for every reaction. Kinetic re-
actions were carried out in a Parr autoclave. Gas chromatog-
raphy was carried out using a Shimadzu GC-17A equipped
with a flame ionization detector and a PONA capillary col-
umn (50 m × 0.21 mm × 0.5 µm). Each sample was weighed
into a GC vial and n-heptane was used as an internal stan-
dard.
3
3 3
molecular ring-opening reactions of epoxides with alcohols
to give β-alkoxyalcohols in good to high yields with high
regio- and stereo-selectivity (3).
Aluminium triflate is a Lewis acid that has been used as a
catalyst in organic synthesis, for reactions including Friedel–
Crafts acylation and alkylation (4, 5), the epoxidation of
olefins with iodosylbenzene (6), the preparation of esters
and lactones (7), as initiators for cationic polymerization (8),
and as co-catalysts in the polymerization of olefins (9). Alu-
minium triflate has also been used as a catalyst for the ring-
opening reactions of epoxides in the presence of alcohols
and amines (10–12). Some work has also been carried out in
which different epoxides and alcohols have been employed
Typical experimental procedure for the ring opening of
butene oxide
Methanol (192.25 g, 6 mol) and butene oxide (72.11 g,
mol) were combined in a 300 mL Parr reactor. On closing
1
the reactor, it was flushed with nitrogen and evacuated to re-
move all air present. This flushing procedure was repeated
three times.
A catalyst bomb was placed under vacuum prior to the
run. The desired amount of the stock solution (2.644 g,
(
13). In a recent study, we described the reaction of butene
oxide with methanol, catalyzed by aluminium triflate (14).
We now report on a kinetic study in which the effect of reac-
tion variables (i.e., catalyst concentration, temperature, and
reagent molar ratio) on the rate of conversion of butene
oxide and selectivity to the major products (the α- and β-
alkoxyalcohols) was investigated. The study was further
expanded to include reactions with ethanol, n-propanol, n-
butanol, 2-propanol, 2-methyl-1-propanol, and 2-methyl-2-
propanol, with the aim of comparing the activity of
Al(CF SO ) as a ring-opening catalyst of butene oxide in
5
ppm) was weighed and transferred via a syringe to the
bomb, which was then connected to the inlet of the auto-
clave. Next, a high-pressure nitrogen line was connected to
the opposite side of the catalyst bomb after which stirring in
the reactor commenced. The reactor was subsequently
heated to 100 °C after which the catalyst solution was dis-
charged into the reactor and a sample taken immediately
(
0 min) and weighed. The sample was weighed to determine
the mass removed from the reactor. Further sampling from
the reactor was carried out at the following intervals: 5, 10,
3
3 3
the presence of different alcohols (Scheme 1).
Due to the fact that Al(CF SO ) is a Lewis acid, its use
1
5, 20, 30, 60, and 120 min and each sample was weighed.
3
3 3
Saturated NaOH–water solution (0.5 mL) was added to all
samples to quench the reaction.
as a ring-opening catalyst for butene oxide leads to the for-
mation of both isomers of the corresponding butene glycol
ether, i.e., α- and β-alkoxyalcohols. Previous studies have
shown that the two isomers form in a molar ratio of close to
one with the β-alkoxyalcohol in a slight excess compared
with the α-alkoxyalcohol (10–12, 14). In a recent communi-
cation, three examples were provided in which the selectiv-
ity to the α-alkoxyalcohol is higher when using butene oxide
and several others in the case of styrene oxide (13). In this
study we show that it is possible to slightly increase the se-
lectivity to the α-alkoxyalcohol with a simultaneous de-
crease in selectivity to the β-alkoxyalcohol by choosing the
correct reaction conditions. Lower temperatures and catalyst
concentrations were found to increase the selectivity to the
Results and discussion
Butoxylation of methanol
Catalyst concentration
In the first part of this study, the influence of catalyst con-
centration on the conversion of butene oxide and the selec-
tivity to the major products (the α- and β-alkoxyalcohols)
was studied over a range of catalyst concentrations. Included
in this was the uncatalyzed reaction. The conversion of bu-
tene oxide is shown in Table 1. The data showed that the
rate of conversion of the uncatalyzed reaction (0 ppm) was
©
2006 NRC Canada

Terblans et al.
861
Table 1. Conversion of butene oxide at different Al(CF SO ) catalyst concentrations.
3
3 3
Conversion (mol%) of butene oxide
Time
Al(CF SO )
Al(CF SO )
Al(CF SO )
Al(CF SO )
3
3 3
3
3 3
3
3 3
3
3 3
(
min)
(0 ppm)
(1 ppm)
(5 ppm)
(10 ppm)
5
3.00
6.00
11.00
16.00
21.00
29.00
41.00
16.00
36.00
46.00
50.00
61.00
77.00
88.00
74.00
97.00
100.00
100.00
100.00
100.00
100.00
85.00
100.00
100.00
100.00
100.00
100.00
100.00
1
1
2
3
6
0
5
0
0
0
1
20
Note: T = 100 °C; MeOH–BuO = 6:1.
Table 2. Selectivity to products at different Al(CF SO ) catalyst concentrations.
3
3 3
1
-Methoxy-2-butanol
2-Methoxy-1-butanol
Catalyst concentration (ppm)
Selectivity (%)
Yield (%)
Selectivity (%)
Yield (%)
0
1
5
49.60
49.75
43.39
42.98
10.42
30.35
43.39
42.98
42.94
42.25
48.61
49.02
9.02
25.77
48.61
49.02
1
0
Note: T = 100 °C; MeOH–BuO = 6:1.
Fig. 1. Ring-opened intermediate.
Al(CF SO )
time and are tabulated in Table 2. Upon considering the in-
fluence of the catalyst concentration on the selectivity to the
products, it is observed that the selectivity to 1-methoxy-2-
butanol was slightly higher than to 2-methoxy-1-butanol.
However, as the catalyst concentration was increased above
H
3
3 3
O
O
2
.5 ppm, the selectivity to the 2-methoxy-1-butanol was
R
slightly higher than to 1-methoxy-2-butanol. This observa-
tion shows that as the catalyst concentration is increased, the
formation of the sterically more hindered isomer becomes
more facile.
significantly slower than those of the catalyzed reactions.
The increase in the rates from 0 to 1 ppm and to 5 ppm can
clearly be seen; thus, as the catalyst concentration was in-
creased, the rate of conversion also increased. However,
hardly any difference was observed in the rate of conversion
when the catalyst concentration was increased from 5 to
Temperature
The second part of the study with methanol focused on
the influence of reaction temperature on the rates of conver-
sion, as well as the selectivity to the products. The conver-
sion data are shown in Table 3.
1
0 ppm. At these two catalyst concentrations, the rate of the
reaction was very fast and the reactions were completed af-
ter approximately 10 min. The low catalyst concentration
needed for a fast reaction is due to the fact that the Lewis
acid (Al(OCF SO ) ) coordinates to the epoxide, which leads
to further activation of the already active electrophile. It was
also further demonstrated that the transfer of the Al(OCF SO )
The temperature data display a similar trend as the cata-
lyst concentration data, i.e., the rate of conversion, increased
as the reaction temperature was increased. The data also
showed that the increase in the rate of conversion was not as
significant when the temperature was increased from 100 to
120 °C, as was the case when it was increased from 60 to
80 °C and to 100 °C. This is due to the fact that the rates be-
come very fast at the higher temperatures. It was also seen
that the higher the temperature, the higher the concentration
of by-products, i.e., dimers and trimers (total yield approxi-
mately 10%, selectivity to dimers approximately 7% and to
trimers approximately 3%).
The selectivities to the alkoxyalcohols and the yields of
these products were calculated after 30 min of reaction time.
The data are shown in Table 4. It was observed that, al-
though the ratio of the major products is approximately 1, at
temperatures below 80 °C, the selectivity to 1-methoxy-2-
butanol is slightly higher than to 2-methoxy-1-butanol,
whereas the opposite is observed at temperatures above
80 °C. At 90 °C, the selectivity to 1-methoxy-2-butanol is
3
3 3
3
3 3
from the ring-opened intermediate (Fig. 1) to the next butene
oxide molecule is energetically favourable (14).
In these reactions two major products (the α- and β-
alkoxyalcohols, 1-methoxy-2-butanol and 2-methoxy-1-
butanol) were formed. By-products such as dimers, diols,
and in some instances, trimers, were also formed during
these reactions with a total yield of dimers, diols, and tri-
mers of approximately 8%, selectivity to dimers of approxi-
mately 4%, selectivity to diols of approximately 1%, and
selectivity to the trimers of approximately 3%. All products
were identified using GC–MS. It was noticed that the higher
the catalyst concentration, the higher the concentration of
by-products.
The selectivities to the α- and β-alkoxyalcohols and yields
of these products were calculated after 30 min of reaction
©
2006 NRC Canada

8
62
Can. J. Chem. Vol. 84, 2006
Table 3. Conversion of butene oxide at different temperatures.
Conversion (mol%) of butene oxide
Time (min)
5
60 °C
80 °C
90 °C
100 °C
120 °C
16.05
36.43
53.47
69.64
74.33
88.50
95.86
16.68
54.31
79.95
81.95
89.17
93.32
97.15
31.67
73.62
97.22
100.00
100.00
100.00
100.00
59.18
94.39
99.55
100.00
100.00
100.00
100.00
88.80
100.00
100.00
100.00
100.00
100.00
100.00
1
1
2
3
6
0
5
0
0
0
1
20
Note: Al(CF SO ) catalyst concentration = 5 ppm; MeOH–BuO = 6:1.
3
3 3
Table 4. Selectivity to products at different temperatures.
-Methoxy-2-butanol
1
2-Methoxy-1-butanol
Selectivity (%)
Temperature (°C)
Selectivity (%)
Yield (%)
Yield (%)
6
8
9
00
20
0
0
0
56.01
55.17
46.00
43.39
43.33
41.63
49.20
46.00
43.39
43.33
43.99
44.83
45.94
48.61
50.67
32.70
39.97
45.94
48.61
50.67
1
1
Note: Al(CF SO ) catalyst concentration = 5 ppm; MeOH–BuO = 6:1.
3
3 3
Table 5. Conversion of butene oxide with Al(CF SO ) at different methanol – butene oxide ratios.
3
3 3
Conversion (mol%) of butene oxide
MeOH–BuO MeOH–BuO MeOH–BuO
(4:1) (6:1) (8:1)
76.69 59.18 57.44
Time
MeOH–BuO MeOH–BuO
Me–BuO
(14:1)
(
min)
(10:1)
(12:1)
5
55.53
96.37
56.75
96.20
57.03
92.55
100.00
1
1
0
5
100.00
100.00
94.39
100.00
96.57
100.00
100.00
100.00
Note: T = 100 °C; catalyst concentration = 5 ppm.
virtually equal to the selectivity to 2-methoxy-1-butanol.
These observations correlate well with those already ob-
served for the catalyst concentration. It was observed that as
the temperature is increased, the formation of the sterically
more hindered isomer becomes more facile; the same trend
is observed for the catalyst concentration.
trimers (total yield approximately 10%, selectivity to dimers
approximately 7% and to trimers approximately 3%).
The selectivities to the alkoxy alcohols and the yields of
these products were calculated after 30 min and the data are
shown in Table 6.
The data show that at lower MeOH–BuO ratios, the selec-
tivity to 2-methoxy-1-butanol is slightly higher, however,
this slight excess decreases and the selectivity to 1-methoxy-
Reactant molar ratio
2
-butanol starts to increase. At a 14:1 ratio of MeOH–BuO,
The last part of the study with methanol focused on the
influence of the molar ratio of the reagents on the rates of
conversion, as well as the selectivity to the alkoxyalcohol
products. The conversion data are shown in Table 5. In this
study the ratio of MeOH–BuO was varied from 4:1 to 14:1.
the selectivity to the two products is virtually the same. This
is due to the fact that at lower MeOH–BuO ratios, the for-
mation of the sterically hindered product 2-methoxy-1-
butanol is more facile than at higher MeOH–BuO ratios.
It was observed that as the ratio of MeOH–BuO is in-
creased, the rate of the reaction decreases, since the transfer
of the Al(CF SO ) from the ring-opened intermediate
Experimental design
All the results obtained from this kinetic study led us to
formulate an experimental design program in which to opti-
mize conditions that would lead to shifting the selectivity to
the formation of more 1-methoxy-2-butanol and less 2-
methoxy-1-butanol. All work done in the experimental de-
sign program can be found in the Supplementary material2
to the publication. After all the data had been accumulated
3
3 3
(
Fig. 1) to butene oxide depends on the concentration of bu-
tene oxide in the reaction mixture. Since the increase in the
MeOH–BuO ratio will lead to a lowering in butene oxide
concentration, a drop in rate was observed as expected. It
was observed that the lower the MeOH–BuO ratio, the
higher the concentration of by-products such as dimers and
2
Supplementary data for this article are available on the journal Web site (http://canjchem.nrc.ca) or may be purchased from the Depository
of Unpublished Data, Document Delivery, CISTI, National Research Council Canada, Ottawa, ON K1A 0R6, Canada. DUD 5044. For more
information on obtaining material refer to http://cisti-icist.nrc-cnrc.gc.ca/irm/unpub_e.shtml.
©
2006 NRC Canada

Terblans et al.
863
Table 6. Selectivity to products at different methanol – butene oxide ratios.
1
-Methoxy-2-butanol
2-Methoxy-1-butanol
Selectivity (%)
MeOH–BuO molar ratio
Selectivity (%)
Yield (%)
Yield (%)
4
6
8
0:1
2:1
4:1
:1
:1
:1
41.48
42.54
43.10
44.12
43.83
44.93
41.48
42.54
43.1
44.12
43.83
44.93
48.52
47.46
46.90
45.88
46.17
45.07
48.52
47.46
46.9
45.88
46.17
45.07
1
1
1
Note: T = 100 °C; catalyst concentration = 5 ppm.
Table 7. Conversion of butene oxide with Al(CF SO ) in the presence of different alcohols.
3
3 3
Conversion (mol%) of butene oxide
Time (min)
5
MeOH
EtOH
n-PrOH
n-BuOH
i-PrOH
i-BuOH
t-BuOH
31.73
47.09
59.00
69.85
80.13
88.50
20.94
37.5
46.27
53.69
64.77
73.74
17.00
28.03
38.03
43.81
49.53
60.46
13.87
21.36
29.47
33.15
39.04
47.12
11.63
17.36
21.10
25.89
28.45
34.50
5.63
8.71
9.57
10.58
3.24
15.91
0.80
0.12
0.03
1.13
0.07
0.46
1
1
2
3
6
0
5
0
0
0
and the relevant calculations had been carried out to deter-
mine the selectivities to 1-methoxy-2-butanol and 2-
methoxy-1-butanol, the data were statistically evaluated by
fitting response surface models. A response surface model
describes the response as a function of the influential vari-
ables and yields a regression model. The significance of the
regression model is evaluated through an analysis of vari-
ance table. This was done for each time interval. As a simi-
lar result was obtained for each time interval, only the
results obtained for the 5 min interval are discussed here; the
rest of the statistical calculations can be found in the Supple-
Fig. 2. Overlay contour plot presenting a feasible area.
DESIGN-EXPERT Plot
5.00
4.25
3.50
Prim Isom/BuO: 0.350
Overlay Plot
Design Points
X = A: Temp
Y = C: Cat conc
Sec Isom/BuO: 0.490
BuO reacted:0.180
Actual Factor
Feasible area
2
mentary material to the publication.
The results obtained for the 5 min interval were evaluated
to determine the simultaneous effect of the variables on the
responses. An overlay plot (a three-point design plot) of the
predicted trends from the models in Fig. 2 was used to deter-
mine this effect.
2.75
3
2.00
6
0.00 67.50 75.00 82.50 90.00
A: Temp
Figure 2 depicts the feasible region (white area demar-
cated by three blue lines) for temperature and catalyst con-
centration where the 1-methoxy-2-butanol – butene oxide
ratio is greater than 0.45, the 2-methoxy-1-butanol – butene
oxide ratio is smaller than 0.35, and the mass of butene ox-
ide reacted is greater than 0.18 mol simultaneously. This fea-
sible region is shown for a methanol – butene oxide ratio at
its centre value (8:1). It is clear that the criteria are satisfied
simultaneously for temperatures between 80 and 85 °C and
catalyst concentrations between 3.9 and 5 ppm.
The statistical interpretation of the data indicates that for
slightly higher selectivities to 1-methoxy-2-butanol and
lower selectivities to 2-methoxy-1-butanol, lower tempera-
tures and catalyst concentrations should be used. This would
represent the feasible region as indicated by the overlay plot
tions led to the expansion of the study to include other alco-
hols. The conditions identified for methanol were used to
carry out reactions with ethanol, n-propanol, 2-propanol, n-
butanol, 2-methyl-1-propanol, and 2-methyl-2-propanol, to
test the activity of Al(OCF SO ) as an epoxide ring-
opening catalyst in the presence of these alcohols. A reac-
tion was also carried out with methanol at these reaction
conditions, to verify that these were indeed suitable for
slightly higher selectivities to 1-methoxy-2-butanol. The re-
action conditions chosen were T = 80 °C, catalyst concentra-
tion = 4 ppm, and ROH–BuO = 8:1.
The conversion data obtained for the different alcohols are
shown in Table 7.
The results show that an increase in the carbon chain
length of the alcohol has an influence on the rate of the
reaction: the longer the carbon chain, the lower the reaction
rate. This was evident when the alcohol was changed from
3
3 3
(
T = 80–85 °C, catalyst concentration = 3.9–5 ppm, and
MeOH–BuO ratio = 8:1).
Butoxylations of various alcohols
The results obtained from the methanol butoxylation reac-
©
2006 NRC Canada

8
64
Can. J. Chem. Vol. 84, 2006
Table 8. Selectivity to major products with different alcohols.
α-Alkoxyalcohol
β-Alkoxyalcohol
Alcohol
Selectivity (%)
Yield (%)
Selectivity (%)
Yield (%)
Sel α : Sel β
MeOH
EtOH
PrOH
n-BuOH
i-PrOH
i-BuOH
t-BuOH
51.17
59.90
62.27
66.21
59.37
39.00
0
41.002 52
38.797 23
30.842 33
25.848 38
16.890 77
1.263 6
40.83
32.10
26.73
24.79
16.63
0
32.72
20.79
13.24
9.68
4.73
0
1.3:1
1.9:1
2:1
2.7:1
3.5:1
—
0
0
0
—
C → C → C → C n-alcohols. When the sterically more
Table 9. Conversion of BuO with t-BuOH at different tempera-
tures.
1
2
3
4
hindered alcohols (2-propanol, 2-methyl-1-propanol, and 2-
methyl-2-propanol) were used, the rate decreased even fur-
ther, up to the point where negligible reaction was observed
when 2-methyl-2-propanol was used.
Conversion (mol%) of butene oxide
Time (min)
5
80 °C
100 °C
120 °C
The selectivities to the alkoxyalcohols were calculated af-
ter 30 min of reaction time and are shown in Table 8. The
yields of these products were also calculated at this time, as
well as the ratio of the selectivity of the α-alkoxyalcohol to
the the selectivity of the β-alkoxyalcohol (ratio Sel α : Sel β);
these values are also shown in Table 8.
The nature of the alcohol plays a role on the product se-
lectivity, clearly demonstrationg a greater selectivity for the
α-alkoxyalcohol with increasing chain length and (or) steric
bulk of the alcohol. No products were formed when using
tert-butanol.
At these conditions, irrespective of the alcohol used, the
reactions favoured the formation of the α-alkoxyalcohol
above that of the β-alkoxyalcohol, in that more of the former
and less of the latter were formed. The longer the carbon
chain of the alcohol, the slower the rate of formation of the
β-alkoxyalcohol. Results showed that even with methanol as
the alcohol, 2-methoxy-1-butanol was only detected after
9.04
13.60
17.06
22.14
30.75
44.33
65.31
15.10
28.05
39.62
45.17
54.43
71.72
85.48
24.62
45.32
64.60
78.29
90.07
98.64
99.45
1
1
2
3
6
1
0
5
0
0
0
20
Note: Al(CF SO ) catalyst concentration = 10 ppm; t-BuOH–BuO = 6:1.
3
3 3
action attempted at T = 80 °C, t-BuOH–BO = 8:1, and a cat-
alyst concentration of 4 ppm showed no activity. Although
the possibility of impurities in the feedstock acting as cata-
lyst poisons cannot be ruled out at this stage, it has previ-
ously been indicated that higher catalyst concentrations are
required with longer chain and bulkier alcohols (13).
The selectivities to the major products, as well as their
yields, were calculated after 30 min of reaction time. These
results are shown in Table 10.
5
min. The ratio of 1-methoxy-2-butanol (the α-alkoxy-
alcohol) to 2-methoxy-1-butanol (the β-alkoxyalcohol) was,
on average, 1.3. With an increase of one carbon atom in the
chain length of the alcohol to ethanol, 2-ethoxy-1-butanol,
the β-alkoxyalcohol was only detected after 10 min. The ra-
tio of the two monomeric products (1-ethoxy-2-butanol to 2-
ethoxy-1-butanol; α-alkoxyalcohol to β-alkoxyalcohol) was,
on average, 1.9.
As the temperature was increased, the selectivity to the
major products decreased. The ratio of the α-alkoxyalcohol
to the β-alkoxyalcohol was approximately 2:1 at all three
temperatures. This is expected as the sterically hindered al-
cohol (t-BuOH) would prefer to attach to the primary carbon
(α-C) of butene oxide, rather than to the secondary carbon
(β-C). This ratio is lower than expected (n-BuOH, 8:1) and
can be rationalized as follows. The formed β-alkoxyalcohol
is a primary alcohol and would react faster with the next
molecule of BuO to produce dimers. Experimentally this is
verified by large concentrations of the by-products such as
dimers and trimers that are formed (yields between 15% and
20%). It can thus be concluded that an increase in the tem-
perature leads to an increase in the conversion of the β-
alkoxyalcohol to dimers and trimers.
As reactions with t-BuOH are generally known to be diffi-
cult because of the steric hinderance of the tert-butyl group,
and as reaction conditions for the ring opening of butene ox-
ide with t-BuOH were identified, it was decided to employ
these reaction conditions in reactions with i-PrOH and i-
BuOH, which have previously shown low activity. The influ-
ence of the different branched alcohols at the same set of re-
action conditions on the rate of conversion of the
butoxylation reaction is shown in Table 11.
It should be noted that the reaction conditions used here
were optimized for methanol as the alcohol and are not nec-
essarily the optimum conditions when using the other alco-
hols. It was therefore decided to further investigate the reaction
conditions for reaction with 2-methyl-2-propanol to deter-
mine whether any reaction will take place when
Al(CF SO ) is used as the catalyst. If any reaction was ob-
3
3 3
served, these conditions would also be used to carry out re-
actions with 2-propanol and 2-methyl-1-propanol.
The reaction conditions chosen were T = 80, 100, and
20 °C, catalyst concentration = 10 ppm, and t-BuOH–
1
BuO = 6:1. The butene oxide conversion data are shown in
Table 9. As expected, it was observed that as the temperature
was increased, the reaction rate increased. Interestingly, at
the higher catalyst concentration of 10 ppm and a lower
t-BuOH–BO ratio of 6:1, a reaction between t-BuOH and
BO was observed at 80 °C. Previously, the butoxylation re-
©
2006 NRC Canada

Terblans et al.
865
Table 10. Selectivities to products at different temperatures with t-BuOH.
-tert-Butoxy-2-butanol 2-tert-Butoxy-1-butanol
1
Temperature (°C)
Selectivity (%)
Yield (%)
Selectivity (%)
Yield (%)
8
1
1
0
00
20
62.20
59.50
51.00
19.13
32.39
45.94
26.80
27.50
29.90
8.24
14.97
26.93
Note: Al(CF SO ) catalyst concentration = 10 ppm; t-BuOH–BuO = 6:1.
3
3 3
Table 11. Conversion of BuO with different branched alcohols.
Conversion (mol%) of butene oxide
Time (min)
5
t-BuOH
i-BuOH
i-PrOH
15.10
28.05
39.62
45.17
54.43
71.72
85.48
33.89
53.06
70.50
77.91
86.42
95.84
99.09
35.67
49.96
71.80
82.47
88.02
96.53
99.65
1
1
2
3
6
0
5
0
0
0
1
20
Table 12. Selectivity to products with different branched alcohols.
α-Alkoxyalcohol
β-Alkoxyalcohol
Alcohol
Selectivity (%)
Yield (%)
Selectivity (%)
Yield (%)
t-BuOH
i-BuOH
i-PrOH
45.10
52.1
49.10
24.56
45.026
43.22
5.90
33.90
34.30
14.10
29.30
30.19
With t-BuOH, which is sterically more hindered than
i-BuOH and i-PrOH, the rate of the butoxylation reaction
with t-BuOH was slower. It was also observed that the rate
with i-BuOH (a primary alcohol with branching) was similar
to that with i-PrOH (a shorter chain secondary alcohol).
The selectivities to the major product and their yields
were calculated at 30 min. These results are shown in Ta-
ble 12.
For each of the three alcohols, the selectivity to the α-
alkoxyalcohol (secondary alcohol) is higher than to the β-
alkoxyalcohol (primary alcohol). This was also previously
observed when the chain length of the alcohol was in-
creased. The selectivity with t-BuOH was lower than for the
other two alcohols, which correlates with the slower reaction
rate, as the sterically more hindered alcohol will react
slower. The selectivity for i-BuOH is higher than for
i-PrOH, as the former alcohol is a primary alcohol and the
latter a secondary alcohol. In all of these reactions, by-
products were formed that are dimers and trimers.
(<80 °C) are required to achieve this ring-opening reaction. For
branched alcohols such as 2-propanol, 2-methyl-1-propanol,
and 2-methyl-2-propanol, either slightly higher catalyst con-
centrations (approximately 10 ppm) or temperatures (ap-
proximately 120 °C) are required because of increased steric
bulk and (or) hinderance.
It was also observed that the regioselectivity to the two
major products (α- and β-alkoxyalcohols) can be kinetically
manipulated to slightly favour either the one or the other,
by changing the reaction conditions. For the reaction with
methanol, it was observed that by choosing the correct
reaction conditions, i.e., low catalyst concentrations and
temperatures, the selectivity is slightly shifted to the α-
alkoxyalcohol and vice versa. By using these conditions for
the other alcohols, it was observed that the regioselectivity is
shifted towards the α-alkoxyalcohol, except for 2-methyl-2-
propanol where no reactivity was observed. However, by
choosing the correct set of reaction conditions for 2-methyl-
2-propanol to obtain activity (catalyst concentration of ap-
proximately 10 ppm and an alcohol–BuO molar ratio of 6),
it was observed that the α-alkoxyalcohol is the preferred
product. The same result is obtained when using 2-propanol
and 2-methyl-1-propanol at the reaction conditions em-
ployed for 2-methyl-2-propanol.
Conclusion
The ring-opening reaction of butene oxide with different
alcohols (methanol, ethanol, propanol, n-butanol, 2-
propanol, 2-methyl-1-propanol, and 2-methyl-2-propanol)
can easily be achieved in the presence of low concentrations
of Al(CF SO ) as the catalyst. For short chain primary alco-
Acknowledgements
3
3 3
hols, e.g., methanol, ethanol, propanol, and n-butanol, very
low catalyst concentrations (<5 ppm) and temperatures
The authors would like to thank SASOL Technology (Pty)
Ltd. for the opportunity to publish this work. We would also
©
2006 NRC Canada

8
66
Can. J. Chem. Vol. 84, 2006
like to acknowledge our colleague, Mr. R. Coetzer, for per-
forming all the statistical calculations in the experimental
design program.
7. Y. Ishii and T. Nakano. EP 1078922 A1. Daiceil Chemical
Company Industries Ltd. 2001.
8
. O. Farooq and W. Minn. US Patent 5 084 586. Minnesota
Mining and Manufacturing Company. 1992.
9
. S.J. Mclain. WO 98/03559. Du Pont. 1997.
References
1
0. M.J. Green. US Patent 4 605 769. BP Chemicals Limited.
1986.
11. M.J. Green. US Patent 4 762 952. BP Chemicals Limited.
1987.
12. D. Falgoux, D. Simoulin, and M. Pascal-Mousselard. US Pat-
ent 4 543 430. BP Chemicals Ltd. 1983.
13. D.B.G. Williams and M. Lawton. Org. Biomol. Chem. 3, 3269
(2005).
1
2
3
4
. T.W. Graham Solomons. Organic chemistry. 5th ed. John
Wiley and Sons Inc., New York. 1992. p. 778.
. A. Berkessel, E. Ashkenazi, and M.R.M. Andreae. Appl.
Catal. A, 254, 27 (2003).
. S. Kobayashi, M. Sugiura, H. Kitagawa, and W.W.-L. Lam.
Chem. Rev. 102, 2227 (2002).
. G.A. Olah, O. Farroq, S.M.F. Farina, and J.A. Olah. J. Am.
Chem. Soc. 110, 2560 (1988).
14. Y.M. Terblans, J.J. Huyser, M. Huyser, M.J. Green, D.A.
Young, and M.S. Sibiya. Can. J. Chem. 83, 854 (2005).
5
6
. S. Kobayashi and I. Komoto. Tetrahedron, 56, 6463 (2000).
. Y. Yang, Y.F. Diederich, and J.S. Valentine. J. Am. Chem. Soc.
1
12, 7826 (1990).
©
2006 NRC Canada