Green Acetylation of Solketal and Glycerol Formal by Heterogeneous Acid Catalysts to Form a Biodiesel Fuel Additive

Article abstract of DOI:10.1002/cssc.201402070

A glut of glycerol has formed from the increased production of biodiesel, with the potential to integrate the supply chain by using glycerol additives to improve biodiesel properties. Acetylated acetals show interesting cold flow and viscosity effects. Herein, a solventless heterogeneously catalyzed process for the acetylation of both solketal and glycerol formal to new products is demonstrated. The process is optimized by studying the effect of acetylating reagent (acetic acid and acetic anhydride), reagent molar ratios, and a variety of commercial solid acid catalysts (Amberlyst-15, zeolite Beta, K-10 Montmorillonite, and niobium phosphate) on the conversion and selectivities. High conversions (72–95 %) and selectivities (86–99 %) to the desired products results from using acetic anhydride as the acetylation reagent and a 1:1 molar ratio with all catalysts. Overall, there is a complex interplay between the solid catalyst, reagent ratio, and acetylating agent on the conversion, selectivities, and byproducts formed. The variations are discussed and explained in terms of reactivity, thermodynamics, and reaction mechanisms. An alternative and efficient approach to the formation of 100 % triacetin involves the ring-opening, acid-catalyzed acetylation from solketal or glycerol formal with excesses of acetic anhydride.

Full text of DOI:10.1002/cssc.201402070

CHEMSUSCHEM
FULL PAPERS
DOI: 10.1002/cssc.201402070
Green Acetylation of Solketal and Glycerol Formal by
Heterogeneous Acid Catalysts to Form a Biodiesel Fuel
Additive
Jennifer R. Dodson, Thays d C. M. Leite, Nathꢀlia S. Pontes, Bianca Peres Pinto, and
Claudio J. A. Mota*[a]
A glut of glycerol has formed from the increased production of
biodiesel, with the potential to integrate the supply chain by
using glycerol additives to improve biodiesel properties. Acety-
lated acetals show interesting cold flow and viscosity effects.
Herein, a solventless heterogeneously catalyzed process for the
acetylation of both solketal and glycerol formal to new prod-
ucts is demonstrated. The process is optimized by studying
the effect of acetylating reagent (acetic acid and acetic anhy-
dride), reagent molar ratios, and a variety of commercial solid
acid catalysts (Amberlyst-15, zeolite Beta, K-10 Montmorillonite,
and niobium phosphate) on the conversion and selectivities.
High conversions (72–95%) and selectivities (86–99%) to the
desired products results from using acetic anhydride as the
acetylation reagent and a 1:1 molar ratio with all catalysts.
Overall, there is a complex interplay between the solid catalyst,
reagent ratio, and acetylating agent on the conversion, selec-
tivities, and byproducts formed. The variations are discussed
and explained in terms of reactivity, thermodynamics, and re-
action mechanisms. An alternative and efficient approach to
the formation of 100% triacetin involves the ring-opening,
acid-catalyzed acetylation from solketal or glycerol formal with
excesses of acetic anhydride.
Introduction
The transesterification of fats and oils to biodiesel produces
approximately 10 wt% of glycerol as a byproduct.[1] With an in-
creasing demand for biodiesel, especially in Europe and the
Americas, there is the potential for glycerol to become an im-
portant feedstock for the chemical industry if sustainability
concerns of biodiesel production are addressed. In Brazil, since
2010 it has been mandatory for diesel to be blended with 5%
biodiesel. The use of crude glycerin from biodiesel production
for traditional pharmaceutical and cosmetic uses is limited as
a result of its low purity; approximately 20% is a mixture of
water, unrecovered methanol, and dissolved salts from neutral-
ization of the basic catalyst. The chemical transformation of
glycerol into value-added products has therefore attracted
a great deal of interest, with a variety of routes available.[1,2]
One opportunity is the production of biodiesel fuel addi-
tives, thereby enabling the whole supply chain to be integrat-
ed. Owing to solubility problems, polymerization, and its de-
composition at high temperatures, glycerol cannot be added
directly to the fuel. However, several studies have shown that
glycerol-derived fuel additives can enhance the biodiesel fuel
specifications, including the viscosity, cetane number, cloud
point, oxidation stability, and emissions.[3,4] Glycerol acetals and
ketals are of particular interest, with the former produced by
the acid-catalyzed reaction of glycerol with aldehydes, and the
latter by reaction with ketones. Upon reaction with ketone, the
main ketal formed is a five-membered ring, whereas a mixture
of five- and six-membered cyclic acetals are produced from re-
actions with aldehydes. These products may also have further
applications as surfactants, flavorings, solvents, and plasticizers
among others.[5,6]
We previously demonstrated that solketal (1), which is the
product of the reaction of acetone with glycerol, improved the
octane number and reduced gum formation upon addition to
gasoline,[7] whereas glycerol acetals improved the cold flow
properties,[3] although there were problems of solubility of
glycerol formal in biodiesel. Recent work has demonstrated
the synthesis of acetylated solketal as a potential additive for
biodiesel, with improvement in the viscosity and cloud point
of the biodiesel beyond that of the non-acetylated version.[8]
However, optimization of the synthetic process was not per-
formed: large amounts of solvent were used and, without the
use of a heterogeneous catalyst, the workup procedures were
intensive. Herein, we study the effect of various parameters, in-
cluding acetylating reagent, reagent ratios, and a variety of
solid acid catalysts, on the formation of acetylated solketal
with this extended to glycerol formal; this demonstrates the
potential for a greener, solvent-free, heterogeneous catalyzed
process (Scheme 1). A complex relationship between the vari-
ous factors is demonstrated.
Results
[a] Dr. J. R. Dodson, T. d. C. M. Leite, N. S. Pontes, Dr. B. Peres Pinto,
Prof. C. J. A. Mota
The effect of three variables on the acetylation of 1 and glycer-
ol formal (3) were studied: the acetylating reagent, the molar
ratio of ketal or acetal to the acetylating agent, and the effect
Institute of Chemistry, Federal University of Rio de Janeiro
Cidade Universitꢀria—Rio de Janeiro, 21941-909 (Brazil)
E-mail: cmota@iq.ufrj.br
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2728 – 2734 2728

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
lyst-15 achieved the highest con-
versions for the acetylation of
both 1 and 3 with acetic acid,
reaching 56% with a 1:1 molar
ratio and 86% with a 1:3 molar
ratio for 1. K-10 montmorillonite,
zeolite Beta, and niobium phos-
phate were all poor catalysts for
the reaction with conversions of
only 17–47% after 2 h. However,
only traces of 8 were formed, in
contrast to the reactions with
acetic anhydride.
The only striking difference
between the use of 3 and 1 was
the selectivity to the desired
products when acetic acid was
Scheme 1. a) Synthesis of solketal (1) from glycerol and acetone, and subsequent acetylation to form ketal 2.
b) Synthesis of acetals 3a and 3b from glycerol and formaldehyde (9), and subsequent acetylation to form acetals
4a and 4b.
of using different solid acid catalysts. For the formation of 2,
the catalysts tested were niobium phosphate, Amberlyst-15, K-
10 Montmorillonite, and zeolite Beta, whereas, for 4, niobium
phosphate and Amberlyst-15 were tested because these cata-
lysts showed the largest variations in conversions and selectivi-
ties.
The results show that high conversions and good selectivi-
ties to the acetylated products were achieved under solvent-
less conditions and by using solid acid catalysts (Figure 1).
However, variations were observed between different catalysts,
reagents, and reagent ratios. Blank reactions were also studied
in the absence of the solid acid catalysts for comparison pur-
poses. These showed lower and slower rates of conversions
(Figures 1 and 2).
Due to the nature of the ketal and acetal reagents, in the
case of the acetylation of 1, only the five-membered ring prod-
uct was observed, whereas for 3 a mixture of five- and six-
membered ring products were formed (Scheme 1). Almost
identical trends for the effect of reagent, reagent molar ratios,
and catalyst were observed for both the ketal and acetal.
Under almost all reaction conditions, triacetin (8) and small
quantities of mono- (5) and diacetin were formed; this was
also observed by Garcꢂa et al. in the acetylation of 1 in triethyl-
amine.[8]
Effect of the acetylating reagent
Across all catalysts, acetic anhydride displayed a more rapid re-
action and higher conversion than acetic acid as an acetylating
reagent, without the need for an excess amount (Figure 2).
The reactions with acetic anhydride were complete within
2 h, and generally reached maximum conversion (90–100%)
within 5 min (Figure 2a). Only niobium phosphate displayed
slower reaction rates, and a lower conversion with a 1:1 molar
ratio, although the selectivity to the desired product was gen-
erally higher than that of the other catalysts.
In contrast, the use of acetic acid resulted in slower reac-
Figure 1. a) Conversion and b) selectivity to the acetylated products of ketal
1 or acetal 3 by using different catalysts and molar ratios by using either
acetic acid after 2 h or acetic anhydride after 1 h of reaction.
tions and lower conversions with much greater variation be-
tween the catalysts (Figure 2b). Of the four catalysts, Amber-
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2728 – 2734 2729

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
Figure 3. Conversion and product distribution from the reaction of 3 and
acetic acid at a molar ratio of 1:3 with Amberlyst-15. The selectivities of the
5, 6, and glycerol are qualitative.
Effect of the acetal/acetylating reagent molar ratio
In general, the use of a higher molar ratio of acetylating re-
agent resulted in a higher total conversion (Figure 1). In all of
the reactions with acetic anhydride, the conversion reached
96–100% after 2 h of reaction with a 1:3 molar ratio of acetal/
acetylating reagent. The improvement was particularly noticea-
ble for niobium phosphate, although the rate of conversion
was still slower than that for the other catalysts. Generally,
small increases in conversion were also observed for a higher
molar ratio with acetic acid; Amberlyst-15 showed a much
larger increase in conversion with a higher molar ratio.
Figure 2. Kinetics of the acetylation of 1 at 608C by using four different
solid acid catalysts, or none, and different acetylating agents at a molar ratio
of 1:3: a) acetic anhydride and b) acetic acid.
Particularly striking is the impact of the molar ratio on the
increased selectivity to 8 when using acetic anhydride, espe-
cially with Amberlyst-15 as a catalyst (Figure 4). With no excess
of anhydride, the selectivity to the desired products was high:
83% for the ketal (2) and 92% for the acetals (4a, 4b) com-
bined. However, when an excess of acetic anhydride was pres-
ent, following an initial rapid conversion to the desired prod-
ucts, almost complete conversion to 8 occurred with a concur-
rent reduction in the concentration of the product (Figure 4b
and c). This was especially rapid and complete in the case of
the reaction of 3 with acetic anhydride in the presence of Am-
berlyst-15, for which the selectivity from acetal to 8 occurred
over 5 min. All of the other catalysts showed an increase in the
selectivity to 8 with excess acetic anhydride. However, the
extent of the conversion of acetal or ketal product to 8 was
not as rapid or as extensive with K10 Montmorollinite, zeolite
Beta, or niobium phosphate (Figure 4a). In the case of niobium
phosphate for the acetylation of the acetal (3), no byproducts
were formed at a molar ratio of either 1:1 or 1:3. Very little 5
or diacetin was observed for any of the samples.
used as the acetylating agent. In general, across the reactions
high selectivities to the desired products 2, 4a, and 4b were
observed. However, when acetic acid was used as the acetylat-
ing agent for 3, a range of byproducts were observed. These
included glycerol, 5, and four peaks with the most intense sig-
nals at m/z 71, 87, 117, and 147. These were identified as glyc-
erol ethers (6), and were identical to those observed during
the hydrolysis of 3.[9] However, in contrast to the reactions with
acetic anhydride, 8 was not identified. The selectivity of these
byproducts combined ranged from 50 to 80% during the reac-
tion, based on the quantity of product formed. Observations
indicate that the concentrations of glycerol and 6 increase in
selectivity in parallel at the beginning of the reaction followed
by a shift in the selectivity towards the desired acetylated
products and 5 as the reaction proceeded (Figure 3). Both Am-
berlyst-15 and niobium phosphate show the same trends in
product distribution, but with a slower rate for the lower
molar ratio reactions and when using niobium phosphate. In
contrast, the reaction of 1 with acetic acid formed the product Discussion
(2) in high selectivity with minor, but increasing, levels of 5
Effect of the acetylating reagent on conversion and selectiv-
ity
and glycerol and traces of 8 with longer reaction times. The 6
were not observed.
The much more effective acetylation of the acetal and ketal
with acetic anhydride versus acetic acid was not unexpected. It
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2728 – 2734 2730

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
mation of 8 is endothermic in the case of acetic acid, whereas
it is exothermic when acetic anhydride is used; this accounts
for the lower conversions and requirement for greater excesses
of reagent for acetic acid than that with acetic anhydride.[10]
Furthermore, in acetylation reactions for glycerol, a similar
trend for the solid acid catalysts was observed as those for the
ketal and acetal in this study; niobium phosphate was much
less active at 608C than Amberlyst-15, zeolite Beta, or K-10
Montmorillonite.[10] In addition, Amberlyst-15 showed a higher
selectivity to 8 than the other catalysts with acetic acid, that is,
greater acetylation correlating with this study.[10]
The formation of glycerol, 5, and 6 during the reactions of 3
and acetic acid can be explained by the formation of water
from the reaction of acetic acid and the acetals, which are not
removed from the reaction medium under the conditions used
(Scheme 2). Water present can result in acid-catalyzed hydroly-
sis of both the initial acetals and the acetylated products; this
accounts for the formation of glycerol and 5, respectively. By-
product 6 will result from the reaction of free glycerol with the
initial acetal; these compounds have also been observed
during the hydrolysis of 3, but not 1, as described previously.[9]
It appears that 6 formed favorably during the initial period of
the reaction, but was gradually converted into the acetylated
product; this indicates that the latter is thermodynamically fa-
vored, especially with excess acetic acid. With Amberlyst-15,
a higher product/ether ratio is observed after 2 h than that
with niobium phosphate; this is presumably due to the gener-
al lower reactivity of the latter catalyst. Monoacetin (5) could
form from either the reaction of acetic acid and glycerol or the
ring-opening hydrolysis of the acetylated acetal; this accounts
for the increasing selectivity to 5 with reaction time as greater
quantities of products 4a and 4b are formed (Scheme 2). As
mentioned previously, the lack of formation of 8 under these
reaction conditions can be attributed to the low reactivity of
acetic acid for the acetylation of glycerol.[10] Interestingly, by-
product 6 was not observed during the acetylation of 1 with
acetic acid. We attribute this to the faster reaction of 1, leading
to rapid formation of the more stable final acetylated product.
Indeed, during the hydrolysis of 1, no 6 or other byproducts
were observed, in contrast to that of 3.[9]
Formation of triacetin (8)
The conversion of 2, 4a, and 4b quantitatively into 8 with Am-
berlyst-15 and, to a lesser degree, with zeolite Beta and K-10
Montmorillonite with an excess of acetic anhydride was unex-
pected. Ring-opening hydrolysis could occur if some of the
acetic acid formed from the initial reaction of acetic anhydride
also reacted to form water. However, the high reactivity of
acetic anhydride; the lack of 5 or diacetin observed in the reac-
tion mixtures, particularly at a 1:1 molar ratio of reagents; and
the variation between the different catalysts suggests that an
alternative mechanism may be in operation. Ethers are well
known as protecting groups, and therefore, cleavage reactions
are extremely important; cyclic ethers such as tetrahydrofuran
(THF) have been widely studied. These are known to be
cleaved by mixtures of acetic anhydride with concentrated sul-
Figure 4. Increase in selectivity to 8 with higher molar ratios of acetic anhy-
dride: a) Selectivity to 8 at different reagent molar ratios for all catalysts
after 120 min of reaction. Variation in the selectivity to 8 and acetylated
products for b) 1 and c) 3 at a reagent molar ratio of 1:1 (unfilled markers)
or 1:3 (filled markers) with Amberlyst-15 and acetic anhydride.
is well known that acetic anhydride is a better acetylating
agent, as observed for the acetylation of glycerol among other
alcohols.[4,10] This has been suggested to be because the for-
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2728 – 2734 2731

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
acetone from 1). The slower con-
version to 8 for the ketal system
may be due to steric hindrance
for this final step. If the molar
ratio is only 1:1, little diacetin is
formed; this indicates that acety-
lation of the alcohol is more fa-
vorable than that of ring-open-
ing acetylation. The presence of
a strong Brønsted acid is neces-
sary for the formation of the acy-
lium ion, and this could explain
the variation in the trend for dif-
ferent solid acid catalysts used in
this study.
In terms of the catalytic activi-
ty, the trends for the catalysts in
this study mirror those observed
for the acetylation of glycerol
with acetic acid and acetic anhy-
dride; niobium phosphate dis-
plays a lower catalytic activity,
whereas Amberlyst-15 is distin-
Scheme 2. Reaction scheme showing possible routes for the formation of monoacetin (5), glycerol, and glycerol
ether 6 byproducts observed during the reaction of 3 with acetic acid.
furic acid or with Lewis acids such as ZnCl2, FeCl3, and MgBr2
resulting in ring opening and the formation of 1,4-diacetoxy-
butane.[11–13] Recently solid heteropolyacids, such as 12-tung-
stophosphoric acid and also sulfamic acid (NH2SO3H), have
shown to be effective for this reaction.[14,15] In these studies,
acetic anhydride was shown to be vital for ring cleavage to
occur; a small concentration of acetic acid was necessary to
prevent nucleophilic attack of THF at the acetylated THF oxo-
cations. Little work has been done on dioxanes or dioxolanes;
however, solid heteropolyacids catalysts were tested on the
cleavage of 1,4-dioxane, with a conversion of 2% for the heter-
opolyacid and 86% for sulfamic acid to 1,2-diacetoxy-
ethane.[14,15] This suggests that a similar mechanism to ether
cleavage could explain the formation of 8 through nucleophilic
attack of 2, 4a, or 4b on the acylium cation (7) formed from
acetic anhydride (Scheme 3). Typically, the product of ether
ring-cleavage reactions results in acyl group addition at the
oxygen atom and acetate addition at the a carbon; however,
further reaction could occur to form 8 and formaldehyde (9; or
guished by its significantly increased acetylating activity com-
pared with those of K-10 Montmorillonite and zeolite Beta
with acetic acid.[10] The acid-catalyzed reaction of glycerol with
acetone to form 1 also shows a catalytic activity trend in the
order of Amberlyst-15>zeolite Beta>K-10 Montmorillonite.[5]
There are huge challenges in the measurement of the acid
strength of solid acids, with little agreement on a preferred
method and variations in the acid strengths and its relation-
ship with catalytic activity observed.[16–18] The presence of
metals, cations, and confinement effects can all make Brønsted
acid quantification difficult. A recent study on the free energy
relationship with acidic strength of solid acid catalysts showed
that the acid strength varied in the order Amberlyst-15>zeo-
lite Beta>niobic acid>K-10 Montmorillonite, whereas n-butyl-
amine thermodesorption showed strong/weak acidity ratios in
the order of zeolite Beta>K-10 Montmorillonite>niobic
acid.[18,19] Amberlyst-15 could not be measured by this method
due to its low thermal stability. Niobium phosphate is expect-
ed to have a higher acid strength than that of niobic acid from
NH3 adsorption studies;[20] how-
ever, this technique measures
both Lewis and Brønsted acid
sites, which may not both be
catalytically active.[16] Overall,
these acid strength trends agree
with the catalytic activities ob-
served in this study and the
other studies highlighted above.
However, they do not account
for the much lower activity of
niobium phosphate in general or
the much higher activity of Am-
berlyst-15 for the conversion of
Scheme 3. Proposed mechanism for the formation of 8 during the reaction of 3 and acetic anhydride.
the acetylated product to 8. This
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2728 – 2734 2732

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
hyde (5.4 mol) and Amberlyst-15 (50 g) at 708C. Before use, the re-
agents were purified by vacuum distillation.
may be due to the effect of the reaction solution on the cata-
lytic activity. For instance, the activity of niobium phosphate
reduces dramatically in polar solvents,[20] whereas anion stabi-
lizing solvents, such as CF3COOH, have been shown to dramat-
ically increase the acidic strength of Nafion, which is a sulfonat-
ed solid acid, and its ability to protonate hydrocarbons by the
formation of strong hydrogen bonds by CF3COOH with the
anion of the acidic site.[17] It can be postulated that the remark-
able conversion of the acetylated products 2, and 4 to 8 in the
presence of Amberlyst-15, and the much higher conversion in
the presence of acetic acid with the same catalyst, could be
due to a similar anion-stabilizing effect upon interaction with
CH3COOH formed in situ from acetylation by acetic anhydride.
Further work is needed to understand the detailed mecha-
nistic process and the specific influence of the solution on the
catalytic activities of the solid acids. Nevertheless, alongside
the formation of the acetylated acetals, the observations in
this study could open up an alternative route to the selective
formation of 8 from glycerol, which has a range of applications
from cosmetics to fuel additives,[4] through the initial formation
of an acetal followed by ring opening with acetic anhydride in
the presence of Amberlyst-15. In the case of 3, 100% selectivi-
ty to 8 was observed; the reaction was more rapid and in-
volved a lower ratio of acetic anhydride than that for the
equivalent conversion and selectivity directly from glycerol.[10]
Byproduct 9 formed should be recyclable. It would be interest-
ing to test whether the same product could be formed with
a lower excess of acetic anhydride.
To study the acetylation of the ketal and acetal, typically the dis-
tilled reagent (75 mmol) was added to a round-bottomed flask
along with either acetic anhydride or acetic acid (1:3 or 1:1 molar
ratio) and the catalyst (1.5 mmol acid sites). The mixture was
heated in an oil bath at 608C for 2 h. Four different catalysts were
used in this study with varying acidic properties (Table 1). Prior to
each reaction, an activation procedure was used to remove water
and impurities (Table 1).
Table 1. Characterization data and pretreatment temperature of the cata-
lysts.
Catalyst
Pretreatment temperature
[8C][a]
Acidity
[mmolgÀ1
]
[b]
Amberlyst-15
K-10 Montmorillon- 150
ite
120
4.7[c]
0.6
zeolite Beta
niobium phosphate
300
90
1.4
0.3
[a] Rate=10 8CminÀ1. Time in activation temperature=1 h. [b] Measured
by n-butylamine adsorption at 1508C.[21] [c] Informed by the producer
(Rohm and Haas).
Samples (50 mL) were removed at intervals over 2 h to analyze the
products formed. The final product generally contained a mixture
of 8 and the acetylated acetal or ketal. As observed by Garcꢂa
et al., it was not possible to separate 8 by vacuum distillation or
solvents, including water, ethanol, methanol, acetone, dichlorome-
thane, ethyl acetate, chloroform, heptane, toluene, and hexane.[8]
Conclusions
Characterization of samples: The conversion and selectivity of the
reactions were followed by GC by using an Agilent Model 6850 in-
strument. Prior to injection, aliquots (50 mL) of the reaction solution
were added to dichloromethane (1 mL) and immediately filtered.
An Agilent HP-5 capillary column (30 mꢃ0.25 mm) with a 0.25 mm
methyl silicone stationary phase was used with helium as a carrier
gas. Aliquots (0.2 mL) were injected and a heating program was
used to heat from 70 to 2008C at 108CminÀ1. The separated com-
pounds were identified by mass spectrometry on a quadrupole
mass spectrometer, Model 5973 Network Agilent, with impact ioni-
zation of electrons at 70 eV. The reagents, acetylated products, and
8 were quantified. Monacetin (5), glycerol, and 6 were identified
by GC-MS.
This work demonstrated the effective acetylation of glycerol
acetal and ketal under a solventless, heterogeneously catalyzed
system. Higher selectivities and conversions to the desired
products were obtained by using acetic anhydride as the ace-
tylation reagent and a 1:1 molar ratio of the reagents with Am-
berlyst-15, zeolite Beta, and K-10 Montmorillonite; these were
all effective catalysts. There is a complex interplay between the
solid catalyst, reagent ratio, and acetylating agent on the con-
version, selectivities, and byproducts formed. An alternative
approach to the highly selective formation of 8 was found
through the ring-opening, acid-catalyzed acetylation from 1 or
3 with higher molar ratios of acetic anhydride in the presence
of Amberlyst-15. Further tests on the use of the acetylated
products as additives to biodiesel will be performed.
Acknowledgements
We would like to thank Capes, FAPERJ, CNPq and PRH-ANP for
their financial support.
Experimental Section
Amberlyst-15 was purchased from Room and Hass, K-10 Montmor-
illonite was from Fluka, niobium phosphate was supplied by
CBMM (Brazil), and zeolite Beta was obtained from Zeolyst. The
other reagents used were analytical grade and purchased from
Sigma–Aldrich.
Keywords: acetylation · glycerol · heterogeneous catalysis ·
reaction mechanisms · waste prevention
[1] A. Behr, J. Eilting, K. Irawadi, J. Leschinski, F. Lindner, Green Chem. 2008,
Ketal 1 and acetals 3a and 3b were synthesized by using a modi-
fied procedure reported by da Silva et al.[5] The reactions were per-
formed under batch conditions. Typically, glycerol (500 g, 5.4 mol)
was stirred with acetone or a 37% aqueous solution of formalde-
10, 13.
[2] F. Jꢄrꢅme, J. Barrault, Eur. J. Lipid Sci. Technol. 2011, 113, 118.
[3] P. H. R. Silva, V. L. C. GonÅalves, C. J. A. Mota, Bioresour. Technol. 2010,
101, 6225.
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2728 – 2734 2733

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
[4] X. Liao, Y. Zhu, S.-G. Wang, Y. Li, Fuel Process. Technol. 2009, 90, 988.
[5] C. X. A. da Silva, V. L. C. Goncalves, C. J. A. Mota, Green Chem. 2009, 11,
38.
[14] Y. Izumi, K. Iida, K. Usami, T. Nagata, Appl. Catal. A 2003, 256, 199.
[15] B. Wang, Y. Gu, W. Gong, Y. Kang, L. Yang, J. Suo, Tetrahedron Lett. 2004,
45, 6599.
[6] P. Ferreira, I. M. Fonseca, A. M. Ramos, J. Vital, J. E. Castanheiro, Appl.
Catal. B 2010, 98, 94.
[16] A. Corma, Chem. Rev. 1995, 95, 559.
ˇ
[17] D. Farcas¸iu, A. Ghenciu, G. Marino, K. D. Rose, J. Am. Chem. Soc. 1997,
[7] C. J. A. Mota, C. X. A. da Silva, N. Rosenbach, Jr., J. Costa, F. da Silva,
Energy Fuels 2010, 24, 2733.
[8] E. Garcꢂa, M. Laca, E. Pꢄrez, A. Garrido, J. Peinado, Energy Fuels 2008, 22,
4274.
[9] L. P. Ozorio, R. Pianzolli, M. B. S. Mota, C. J. A. Mota, J. Braz. Chem. Soc.
2012, 23, 931.
[10] L. N. Silva, V. L. C. GonÅalves, C. J. A. Mota, Catal. Commun. 2010, 11,
1036.
119, 11826.
[18] V. L. C. GonÅalves, R. C. Rodrigues, R. LorenÅato, C. J. A. Mota, J. Catal.
2007, 248, 158.
[19] R. Tourinho, I. Oliveira, P. Arroyo, C. A. Mota, Catal. Lett. 2011, 141, 1472.
[20] P. Carniti, A. Gervasini, S. Biella, A. Auroux, Chem. Mater. 2005, 17, 6128.
[21] V. L. C. GonÅalves, B. P. Pinto, J. C. da Silva, C. J. A. Mota, Catal. Today
2008, 133–135, 673.
[11] R. L. Burwell, Chem. Rev. 1954, 54, 615.
[12] D. J. Goldsmith, E. Kennedy, R. G. Campbell, J. Org. Chem. 1975, 40,
3571.
[13] Y. Tanoue, M. Hamada, N. Kai, T. Nagai, K. Sakata, M. Hashimoto, S.-I.
Morishita, J. Heterocycl. Chem. 2000, 37, 1351.
Received: February 17, 2014
Revised: May 4, 2014
Published online on July 8, 2014
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2728 – 2734 2734