Fully recyclable Br?nsted acid catalyst systems

Article abstract of DOI:10.1039/d0gc04199h

Homogeneous and heterogeneous sulfonic acid catalysts are some of the most common catalysts used in organic chemistry. This work explores an alternative scheme using a fully recyclable polymeric solvent (a poly-α-olefin (PAO)) and soluble PAO-anchored polyisobutylene (PIB)-bound sulfonic acid catalysts. This PAO solvent is nonvolatile and helps to exclude water by its nonpolar nature which in turn drives reactions without the need for distillation of water, avoiding the need for excess reagents. This highly nonpolar solvent system uses polyisobutylene (PIB) bound sulfonic acid catalysts that are phase-anchored in solvents like PAO. The effectivenes of these catalysts was demonstrated by their use in esterifications, acetalizations, and multicomponent condensations. These catalysts and the PAO solvent phase show excellent recyclability in schemes where products are efficiently separated. For example, this non-volatile polymeric solvent and the PIB-bound catalyst can be recycled quantitatively when volatile products are separated and purified by distillation. In other cases, product purification can be effected by product self-separation or by extraction.

Full text of DOI:10.1039/d0gc04199h

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
Fully recyclable Brønsted acid catalyst systems†
Cite this: Green Chem., 2021, 23,
266
Christopher B. Watson, Adrianna Kuechle and David E. Bergbreiter
*
1
Homogeneous and heterogeneous sulfonic acid catalysts are some of the most common catalysts used
in organic chemistry. This work explores an alternative scheme using a fully recyclable polymeric solvent
(a poly-α-olefin (PAO)) and soluble PAO-anchored polyisobutylene (PIB)-bound sulfonic acid catalysts.
This PAO solvent is nonvolatile and helps to exclude water by its nonpolar nature which in turn drives
reactions without the need for distillation of water, avoiding the need for excess reagents. This highly
nonpolar solvent system uses polyisobutylene (PIB) bound sulfonic acid catalysts that are phase-anchored
in solvents like PAO. The effectivenes of these catalysts was demonstrated by their use in esterifications,
acetalizations, and multicomponent condensations. These catalysts and the PAO solvent phase show
excellent recyclability in schemes where products are efficiently separated. For example, this non-volatile
polymeric solvent and the PIB-bound catalyst can be recycled quantitatively when volatile products are
separated and purified by distillation. In other cases, product purification can be effected by product self-
separation or by extraction.
Received 11th December 2020,
Accepted 20th January 2021
DOI: 10.1039/d0gc04199h
rsc.li/greenchem
and reuse can be accomplished. In addition, a homogeneous
Brønsted acid catalyst requires a solvent. Since the solvent is
Introduction
Brønsted acid catalysts have been part of the repertoire of che- typically the majority of the mass of most reactions, it of par-
mists for over 200 years. Sulfuric acid is a common example of ticular concern in terms of sustainability and green
7
such Brønsted catalysts. It is used in large scale processes like chemistry.
the isomerization of cyclohexanone oxime to caprolactam and
in petroleum refining. Sulfuric acid is also used in organic problems in green and sustainable chemistry.
The issues associated with solvent use are well-recognized
8
–12
There have
chemistry as a catalyst for esterification, acetalization, ether been many attempts to alleviate this problem. For example,
formation, and dehydration reactions. In many cases, an in- there are now a range of more benign solvents which are less
1
3
expensive catalyst like sulfuric acid is not recycled. harmful as generated wastes. Since these solvents are bioder-
Nonetheless, though this acid catalyst is inexpensive, issues of ived, they produce no net greenhouse gas (GHG) in contrast to
recyclability, sustainability, and safety have led to the wide- petroleum derived solvents if combustion is used for solvent
spread use of solid inorganic or organic alternatives like poly- disposal. However, regardless of the source for the carbons in
1
,2
styrene-bound sulfonic acids and Nafion as catalysts. These the solvent, the best atom economic scenario for a solvent
insoluble analogs of sulfuric acid allow for products to be sep- used in any reaction including a homogeneous Brønsted acid
arated from the acid catalysts by filtration and make catalyst catalyzed process would be to use a solvent that is readily
reuse feasible. However, these heterogeneous catalysts can recyclable. Recycling not only minimizes GHG, it also reduces
have drawbacks when compared to their homogeneous the need to remake more chemicals – another objective for
1
4
analogs. For instance, the activity of heterogeneous Brønsted green chemistry. Unfortunately, even when recycling is prac-
acid catalysts can be limited by diffusion and recycling is not ticed, it most often involves an extra energy intensive distilla-
always as feasible as might be expected. While these catalysts tion step to the overall process.
3
–6
are continuing to be improved,
alternative systems remain
Our prior work has shown that poly(α-olefin)s (PAOs) have
1
5–17
of interest. One alternative to these recyclable heterogeneous the potential to be highly recyclable polymeric solvents.
acid catalysts would be a fully recyclable homogeneous cata- These materials are commercially available on a large scale
lyst. The problem in this case is in part how catalyst separation and are relatively inexpensive. PAOs have other advantages too.
They have low toxicity, are nonvolatile, are reportedly bio-
degradable in some cases, and have higher flash points than
conventional organic solvents. PAOs also have disadvantages
Department of Chemistry, Texas A&M University, College Station, Texas 77842-3012,
USA. E-mail: bergbreiter@tamu.edu
18
in that they are simply alkanes and as alkanes they are poor
solvents for polar catalysts, reactants, or substrates.
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
d0gc04199h
1266 | Green Chem., 2021, 23, 1266–1273
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
1
9
Polymeric solvents like PAOs are not new. However, their involved synthesis of a PAO-soluble sulfonic acid from PIB that
recycling most often requires substantial additional solvent is commercially available as material derived from 100%
use or further purification. For example, the most common renewable feedstock. These initial efforts to synthesize a
polymeric solvent, poly(ethylene glycol) (PEG), is commonly polyisobutylene (PIB) bound arylsulfonic acid from alkene-
recycled by precipitation with excess ether. An alternative functionalized PIB via electrophilic aromatic substitution and
scheme would be to use distillation to separate products from then sulfonation failed due to depolymerization of the PIB in
2
4
a polymeric solvent and then to simply reuse the solvent in arene solvents. Thus, we turned our attention to the syn-
another reaction. However, while there is one report where pro- thesis of a PIB-bound alkylsulfonic acid. This methanesulfonic
ducts were separated from a polymeric propylene glycol acid analog was synthesized in two steps via a thio-ene reac-
2
0
solvent by distillation, the solvent still had to be filtered tion followed by an oxidation. That process afforded the PIB-
before reuse and there was some evidence for solvent bound sulfonic acid 2 in 47% yield (Scheme 1). The loading of
degradation.
The work here expands on the potential of soluble poly- analyzed by titration and was shown to be 0.74 mmol of –SO
meric catalysts and polymeric solvents and addresses the per g of PIB, consistent with an M of the PIB-CH SO H 2 of
the sulfonic acid on this analog of methanesulfonic acid was
3
H
n
2
3
green chemistry issues above in several ways. First, we show 1400 Da.
that a polyisobutylene (PIB)-bound alkyl- and arylsulfonic acid We then turned to reexamine the problems of synthesis and
catalysts’ high solubility in a PAO phase keeps them separate sulfonation of an aryl functionalized PIB. While we were
from a polar phase and makes them recyclable. This phase unable to develop a synthesis of a PIB analog of toluene sulfo-
n
selective solubility requires a suitable M for the PIB phase nic acid from alkene-terminated PIB, we were successful in sul-
anchor. Thus while a smaller PIB450-bound sulfonic acid fonating the more readily accessible anisole-terminated PIB
might leach to a small extent into a polar extraction solvent that could be prepared by reaction of an alkene-terminated
3
like CH CN, an otherwise similar PIB1000-bound sulfonic acid fully renewable PIB with anisole in the presence of catalytic
has less than 0.2% leaching. Using a higher molecular weight sulfuric acid. Sulfonation of this PIB-bound anisole group with
PIB starting material requires using a more viscous PIB species oleum afforded a 62% yield of the PIB-bound acid 3
in the synthesis of the catalyst (PIB1000 is highly viscous with a (Scheme 1). The loading of the arylsulfonic acid was analyzed
2
1,22
viscosity of ca. 190 cSt at 100 °C)
species is irrelevant in the reactions since it is present at rela- PIB, consistent with an M
tively low concentration. As a result, the viscosity of the reac- The polymer products from both these syntheses have M
tion solution is essentially the same as the much lower vis- values that are higher than that expected from the approxi-
the viscosity of the PIB by titration and was shown to be 0.62 mmol of –SO H per g of
3
n
3
of the PIB-ArSO H 3 of 1600 Da.
n
2
3
cosity of the PAO432 (4.09 cSt at 100 °C). PAO432 was used mately 1000 Da starting material. We’ve noted this effect
because it has good balance of very low (<0.1 wt%) leaching before in other chemistry modifying PIB groups to make cata-
2
5
and modest viscosity vis-à-vis PAOs of lower or higher mole- lysts. We ascribe the increase in M of both of these products
n
1
6
n
cular weight. Second, while added solvents can be used to to fractionation of the PIB where in leaching of lower M poly-
extract products from a PAO phase, we show that the low vola- mers into extraction solvents occurs during the separations
tility of a PAO solvent and the PIB-bound catalysts facilitates during this synthesis. The product sulfonic acids 2 and 3 are
product isolation by distillation. Volatile products are typically firmly phase anchored, are non-volatile, and do not leach to a
purified by distillation anyway. Thus, separation by distillation measureable extent in other polar solvents. This phase selec-
of a volatile product from a nonvolatile polymeric solvent tive solubility was noted too in prior work where these PIB-
under reduced pressure is a simple but effective way to reuse a bound sulfonic acids were used to solubilize silica nano-
2
6
polymeric catalyst and polymeric solvent. Third, we show that particles in alkane solvents. While they can be purified and
these PAO solutions of recyclable Brønsted acid catalysts have isolated from low molecular weight solvents like heptane or
general applicability in that they catalyze reactions such as THF during their synthesis, we have not attempted to and
acetal formation, a sulfonic acid catalyzed multicomponent believe it would be difficult to separate them from PAO
condensation to form a 3,4-dihydropyrimidin-2-one, and ester- because of their phase selective solubility in PAO versus other
ifications. Finally, in the case of esterifications, we show that solvents.
the self-removal of water from the PAO phase modestly shifts
the predicted equilibrium to favor the ester product.
To further test the separability of the PIB-bound sulfonic
acids 2 and 3 from a polar phase we carried out experiments
Results and discussion
Conventional strong organic acids such as p-toluenesulfonic
acid and methanesulfonic acid have little to no solubility in
PAOs. Thus, to explore PAOs as recyclable media for Brønsted
acid catalysis, we had to first synthesize sulfonic acids that
could be anchored and recycled in PAOs. Our initial attempts Scheme 1 Synthesis of PIB-bound alkyl- and arylsulfonic acids 2 and 3.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 1266–1273 | 1267

View Article Online
Paper
Green Chemistry
using biphasic solutions of methanol and PAO432 that con-
tained either a methanol-anchored azo dye or a PAO432
-
anchored azo dye. In the first case, we use methanol contain-
ing the azo dye, p-methyl red 4a, and PAO432. In the second
case, we used a biphasic solution of methanol and PAO432 con-
taining a PIB-bound p-methyl red dye 4b. The low molecular
weight azo dye 4a either in the neutral or protonated form has
no solubility in the 30-carbon PAO432 solvent. Likewise, the
Scheme 2 Acetalization promoted by PIB-bound Brønsted acid cata-
lysts PIB-SO
3
H (2 or 3) in reactions that were recycled 10 times with
average isolated yields of 91% and 90% respectively.
PIB-bound azo dye 4b has no solubility in methanol. To test protection of benzyl alcohol with 3,4-dihydro-2H-pyran (DHP)
the phase selective solubility of the PIB-bound sulfonic acids 2 (Scheme 2). Similar reactions have been carried out before
and 3, we physically mixed a solution of 4a in methanol with a with a heterogeneous Brønsted acid catalyst (Amberlyst 15)
PAO432 solution of 2 or 3. The resulting biphasic mixtures that using the same 1 : 1.2 ratio of reactants, and 1 h reaction time
formed showed that the dye in the methanol phase was not at room temperature. Similar 93–98% yields were obtained
protonated by the PIB-bound acids (Fig. 1) based on the using 10 mol% of a catalyst. Catalyst reuse was not described
2
7
absence of any of the red protonated dye. We then added a in this paper. Using 2 or 3, we were able to show that this
similar PIB-bound azo dye (4b) that was soluble in heptane to reaction could be carried out at ambient temperature using
PAO432 containing 2 or 3. In this case, the dye readily proto- 1.2 mmol of DHP and 1.0 mmol of benzyl alcohol with
nated. Then we added 4a to this second biphasic mixture. No 5 mol% of 2 or 3. After 1 h of stirring, the 10 mL of the PAO432
protonation of 4a was seen under conditions where 4b was reaction mixture was shaken with 3 mL of acetonitrile and the
fully protonated by 2 or 3. The visual absence of leaching of resulting biphasic mixture was transferred to a centrifuge tube.
the acids 2 and 3 into the methanol phase and the absence of After centrifugation, the bottom PAO432 layer was removed
protonation of 4a was confirmed by UV-visible analysis of the with a pipette, charged with fresh substrates, and reused. This
methanol phase which showed no detectable protonated 4a. PAO phase containing 2 or 3 was also easily recycled. Ten
Other experiments not shown in Fig. 1 showed that CH
3
SO
3
H
cycles were carried out with no observable decrease in the
did protonate 4a but that CH SO
3
3
H did not protonate 4b dis- volume of the PAO phase. The acetonitrile phases containing
solved in PAO432. This shows that the PIB-bound sulfonic acids the product were combined and washed twice with 5 mL ali-
have the preferential solubility in PAO432 necessary if they are quots of heptane and the acetonitrile was then removed at
to be used as recyclable Brønsted acid catalysts.
reduced pressure to afford the product with 91% and 90%
Next, the use of PAO as a solvent with both PIB-bound sulfo- average isolated yield respectively for 2 and 3 over 10 cycles.
1
13
nic acid catalysts 2 or 3 was explored using as an example the The product THP ether was characterized by H and C NMR
spectroscopy and did not contain any detectable PAO432, 2, or
3
. This showed that the acid catalysts and PAO432 solvent
system was recyclable when products were separated using a
liquid/liquid extraction.
The recyclability of PIB-bound acids 2 and 3 was further
tested by comparing the PIB-bound sulfonic acids 2 and 3 to a
commercially available hydrocarbon soluble sulfonic acid,
p-dodecylbenzenesulfonic acid 5 (Fig. 2). p-Dodecylbenzene-
sulfonic acid, like 2 and 3, was soluble in heptane, however 5
had limited solubility in PAO432 presumably because it is not
hydrophobic enough for this PAO solvent. Thus, our initial test
to compare the recyclability of 2 and 3 versus 5 in the DHP pro-
tection of benzyl alcohol (Scheme 2) was studied in heptane to
insure that we could use the catalysts 2, 3, and 5 at the same
5
mol% catalyst loading. The results in Fig. 2 show that over
Fig. 1 From left to right; (1) biphasic mixture of p-methyl red 4a dis- the course of five cycles, the conversion of benzyl alcohol to
solved in methanol (top) with a PAO432 phase (bottom), (2) biphasic
mixture of methanol (top) with the PIB-bound p-methyl red 4b dis-
solved in a PAO432 phase (bottom), (3) a biphasic mixture of p-methyl
red 4a dissolved in methanol (top) and p-methyl red 4b dissolved in a
the THP ether was in the range of 90–100% for the PIB-bound
acids. In contrast, the conversion of benzyl alcohol to THP
ether decreased to ca. 10% after five cycles using 5. Similar
results were seen using PAO432 though those studies required
PAO432 phase (bottom) with PIB-bound acid 2 added, and (4) a biphasic
mixture of p-methyl red 4a dissolved in methanol (top) and p-methyl longer reaction times due to the diminished solubility of 5 in
red 4b dissolved in a PAO432 phase (bottom) with PIB-bound acid 3
added. The color and lack of absorption at 514 nm in the methanol
phase in cases (3) and (4) show no leaching of PIB-bound sulfonic acids
PAO432 versus heptane. These studies using PAO432 versus
heptane showed catalysts 2 and 3 could be reused while the
catalyst 5 leached into the acetonitrile solvent used to isolate
2
or 3 into the methanol phase occurs. The color and absorption at 514
the product. These studies also showed that some heptane was
lost in each cycle while the PAO432 was fully recoverable and
in the PAO432 phase in cases (3) and (4) shows the protonation of the
PIB-bound dye by the PIB-bound sulfonic acids readily occurred.
1268 | Green Chem., 2021, 23, 1266–1273
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
excess urea. The recrystallized product was then washed with
two 5 mL portions of heptane to give a final product in 82%
and 81% average isolated yield for reactions using 2 or 3,
respectively. The final product was free of excess urea, PAO432
1
and the PIB sulfonic acids based on H NMR spectroscopy.
The use of a variety of heterogeneous Brønsted acids was
explored in this case and Amberlyst 15 DRY, a heterogeneous
2
9
acid Brønsted acid catalyst, was the most effective catalyst. In
these studies, this heterogeneous catalyst afforded products in
a yield comparable to what we observed with 2 or 3 in PAO.
However, unlike our system where the yields were comparable
in cycles 1 and 7, yields in a reaction with a recycled hetereoge-
neous catalysts decreased steadily through five recycling steps.
Esterifications were the third class of Brønsted acid cata-
lyzed reactions studied. Esterifications were of particular inter-
est to us because we hypothesized that they could highlight
Fig. 2 Five cycles of a DHP protection of benzyl alcohol using three additional advantages of a PAO432 solvent. Specifically, esters
different acids; (from left to right) p-dodecylbenzenesulfonic acid 5, the
PIB-bound alkylsulfonic acid 2, and the PIB-bound arylsulfonic acid 3.
are often volatile compounds. Thus, we thought that we could
combine the separation of the ester products from a PAO432
solution of 2 or 3 with a purification of these products taking
advantage of the nonvolatility of the PAO solvent and catalyst.
reusable over these 10 cycles. These results suggest that the
This was accomplished removing the products from the reac-
PIB phase-anchor is required if these soluble sulfonic acid cat-
tion solution at reduced pressure after the equilibrium was
alysts are to be effectively recyclable in an alkane solvent, and
that PAO432 is a better choice than heptane as a recyclable
established.
An additional feature of esterifications is that the reactant
solvent in these liquid/liquid separations.
acid and alcohol are in equilibrium with the product ester and
A second example using 2 or 3 in PAO432 involved a multi-
water. While the product ester and water are somewhat favored
component reaction that forms a 3,4-dihydropyrimidin-2-one
2
8
thermodynamically versus the reactant acid and alcohol, the
equilibrium constants are not much different from one. Thus
these reactions are typically driven to completion by using
excess alcohol or by removing the water product as it forms
derivative (Scheme 3). This reaction used 3 mmol of urea,
2
1
mmol of benzaldehyde, and 2 mmol of ethyl acetoacetate in
0 mL of PAO432 and was carried out at 90 °C. In this case, we
were able to effect a reaction even though one component was
a highly polar substrate. The solubility of urea in PAO432 was a
concern since separate experiments showed that urea was
essentially insoluble in PAO432 even at 90 °C. However, in the
actual reaction, benzaldehyde and ethyl acetoacetate are
present at concentrations of 0.2 M. In the event, we assume
these compounds polarize the PAO phase enough to effect
some urea solubilization in the PAO432 solvent system allowing
the reaction to occur. Regardless of the explanation, the result
was that reaction proceeded to completion over 12 h. In this
case, the product was insoluble in the PAO432 solvent and pre-
cipitated as it formed. Thus, the product could be isolated by
centrifugation. The remaining PAO432 solution containing the
catalyst 2 or 3 was decanted and reused in six more cycles with
fresh substrates. The solids from the seven cycles were com-
bined and purified by recrystallization from ethanol to remove
using distillation and a Dean–Stark trap. In the case of PAO432
,
we hypothesized that the immiscibility of water in PAO432 and
the nonpolar nature of the PAO432 solvent might make these
reactions more facile by self-separation of the water from the
PAO432 phase. This could allow us to carry out esterifications
without excess reagents.
Our initial studies compared the reactivity of the sulfonic
acids 2 and 3 in PAO432 to a common sulfonic acid catalyst
p-toluene-sulfonic (p-TsOH). Since p-TsOH does not dissolve
in PAO432, this comparison involved the reaction of propionic
acid with benzyl alcohol in PAO432 with 2 or 3 or in toluene
with p-TsOH. The results of this kinetic comparison are shown
in Fig. 3. These data show that the acids 2 and 3 in PAO432 are
kinetically comparable to a conventional p-TsOH catalyst in
toluene in esterification chemistry suggesting that PIB-bound
sulfonic acids have acidity like their low molecular weight
counterparts, a result consistent with the results in Fig. 2
above. We also briefly explored the potential of quantifying the
acidity of these PIB-bound sulfonic acids using Hammett indi-
3
0
31
cator dyes or P NMR spectroscopy probes of Brønsted
3
1
acidity. Unfortunately, the Hammett indicator dyes are in-
soluble in PAO and even a relatively hydrophobic trialkyl-phos-
phine oxide probe of Brønsted acidity – trioctylphosphine
oxide – did not dissolve in PAO. Thus, we can only qualitatively
compare 2 and 3 with other sulfonic acids (cf. Fig. 2 and 3).
Scheme 3 Multicomponent condensation reaction that was recycled
seven times with an average isolated yield of 82% and 81% per cycle
using 2 and 3, respectively.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 1266–1273 | 1269

View Article Online
Paper
Green Chemistry
Scheme 4 Esterification of hexanoic acid using methanol. Recycled 10
times with 2 and 3 to give 80% and 81% averaged isolated yield
respectively.
10 mL of PAO432 at 70 °C for 5 h. The results showed that the
conversion of alcohol to ester under these conditions as
1
measured by H NMR is better than the calculated values in all
cases. While the differences are modest for ethyl and methyl
acetate, the observed conversions are much better than the
predicted conversions for methyl and ethyl octanoate. The
result of higher conversion in these PAO solvent systems is
Fig. 3 Comparison of esterification rates for reactions using PIB-bound
acids 2 or 3 in PAO432 with a esterification using p-toluenesulfonic acid consistent with our hypothesis that the exclusion of water can
catalyst in toluene.
help drive these reactions by Le Chatelier’s Principle. The
results also show that even relatively polar materials such as
methanol and acetic acid can be used in a PAO solvent system
The conversion of acids to esters in a solution of PAO432 to give good conversions to ester products without the use of
with catalysts 2 and 3 was then tested and compared with the excess reagents or a Dean–Stark apparatus.
theoretical values that could be calculated from NIST data
The use of this system was then tested for its recyclability
Tables 1 and 2). These calculations used the ΔHformation for using the esterification of hexanoic acid with methanol
3
2
(
these materials as an approximation of the ΔG of the reaction. with 2 and 3 (Scheme 4). The reaction was carried out in
In these calculations we assume the entropy term should be PAO432 using the same 1 : 1 mol ratio of the carboxylic acid :
small for a reaction of two substrate molecules forming two alcohol and 1 mol% of the PIB-bound catalyst in 10 mL of
molecules of product. Thus, we assume that the ΔH value is PAO432
roughly equal to the ΔG of the reaction. Using these assump-
The ester product was separated from the PAO432 solvent,
.
tions, we calculated a theoretical equilibrium constant and the catalyst, and any unreacted acid by a distillation of the
theoretical conversion for reactions forming methyl and ethyl product at reduced pressure. Each additional cycle required
esters at 70 °C.
only addition of fresh alcohol and carboxylic acid substrate.
These reactions used 10 mmol of the carboxylic acid and This reaction was carried out five times without any decrease
0 mmol of the alcohol with 1 mol% of catalysts 2 or 3 in in product yield for both acids 2 and 3. The average isolated
yield of methyl hexanoate was 80% and 81% for catalysts 2 and
1
3
respectively.
Esterifications have been catalyzed using the heterogeneous
using catalyst 2 and a 1 : 1 molar ration of acid : alcohol as measured by Brønsted acid catalyst Amberlyst 15 before. However these
H NMR spectroscopy. Calculated conversions are shown in parentheses reactions typically used excess alcohol (a 6 : 1 mole ratio of the
Table 1 Conversions of carboxylic acids to esters with various alcohols
3
3
1
for the methyl and ethyl esters
alcohol to the carboxylic acid) to achieve yields like those in
Table 1. The Amberlyst 15 catalyzed reactions also employed a
slightly higher 80 °C reaction temperature and the catalyst was
less recyclable.
Catalyst 3
Acetic acid
Hexanoic acid
Octanoic acid
Methanol
Ethanol
Propanol
94% (84%)
89% (77%)
86%
91% (73%)
95% (78%)
93%
91% (55%)
84% (69%)
80%
Benzyl alcohol
80%
87%
73%
Conclusions
Table 2 Conversions of carboxylic acids to esters with various alcohols Two PIB-bound sulfonic acids that were highly phase-selec-
using catalyst 3 and a 1 : 1 molar ration of acid : alcohol as measured by tively soluble in alkane solvents were synthesized. Both PIB-
1
H NMR spectroscopy. Calculated conversions are shown in parentheses
bound alkyl- and arylsulfonic acids showed good catalytic
activity in known sulfonic acid catalyzed reactions. The work
for the methyl and ethyl esters
presented here shows that a PAO432 solvent system with these
Catalyst 2
Acetic acid
Hexanoic acid
Octanoic acid
acids is effective in a variety of different reactions with an
assortment of reagents and that these soluble acid catalysts
and PAO432 can be repeatedly recycled. This is an improve-
ment on classical solvents such as heptane, because the PAOs
Methanol
Ethanol
Propanol
94 (84%)
83% (77%)
87%
88% (73%)
97% (78%)
80%
89% (55%)
95% (69%)
81%
Benzyl alcohol
77%
81%
65%
1270 | Green Chem., 2021, 23, 1266–1273
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
do not suffer from leaching or evaporative losses. This PAO solution was then washed 3 times with 50 mL portions of
solvent system modestly facilitates the esterification reactions acetonitrile and concentrated to give 5.1 g of a viscous brown
1
by expelling water due to its hydrophobicity. Finally, we oil; H NMR (400 MHz, CDCl
3
) δ 5.67 (br, chemical shift is con-
show that polymeric solvents like these PAO solvent systems centration dependent due to trace water), 3.23 (dd, J = 3.7,
can not only be used in schemes where filtration and extrac- 14.2 Hz, 1H), 3.00 (dd, J = 8.9, 14.2 Hz, 1H), 2.18 (br, 1H),
tion are used to isolate products, but that they are also 0.84–1.6 (m).
useful when volatile products are isolated and purified by
distillation.
Synthesis of PIB-bound arylsulfonic acid 3
Anisole-terminated PIB (10 g, 10 mmol) was prepared using a
3
5
literature procedure. The product of this synthesis was dis-
solved in dichloromethane (100 mL) in a 250 mL round-bot-
tomed flask. The reaction flask was then charged with 20%
fuming sulfuric acid (10 mL). This mixture was stirred over-
night and then concentrated under reduced pressure. The
viscous oily residue was dissolved in heptane (100 mL) and
washed with 3 times with 50 mL portions of acetonitrile to
remove residual sulfuric acid. The heptane phase was then
Experimental
General procedures
n w
Alkene terminated PIB 1 (M = 1048 Da, M = 1886 Da, PDI =
1
.8, 81% alpha vinylidene content, viscosity at 100 °C = 196
cSt) was obtained from the TPC group and used without purifi-
2
1
cation. The PAO432 used was obtained from Exxon-Mobil (vis-
2
3
cosity at 100 °C = 4.09 cSt) and used without purification.
Acetonitrile (>99.9%), heptane (99%), azobisisobutyronitrile
98%), 3,4-dyhydro-2H-pyran (99%), toluene (99.8%), aceto-
concentrated under reduced pressure to give 7.4 g of a viscous
1
purple oil. H NMR (400 MHz, CDCl
3
) δ 7.91 (d, J = 2.2 Hz,
(
1
H), 7.59, 7.02 (dd, J = 8.8, 2.2 Hz, 2H), 3.82 (s, 3H), 3.76 (br,
chemical shift is concentration dependent due to trace water),
.80 (s, 2H), 0.84–1.6 (m).
nitrile (99.5%) and ethyl acetoacetate (99%) were purchaced
from Sigma-Aldrich. Tetrahydrofuran, methanol (99.8%), and
urea were purchased from Fisher Chemical. Octanoic acid
1
Recycling of PAO/acid in the dyhydropyran protection of
benzyl alcohol
(
98%), anisole (99%) were purchased from Tokyo Chemical
Industry. 1-Propanol (99%), hydrogen peroxide (35 wt%) were
purchased from Acros organics. Benzyl alcohol (98%), hexa-
noic acid (98%) were purchased from Merck. Dodecylbenzene
sulfonic acid (97%) was purchased from spectrum chemical.
Benzaldehyde (99%) and thioacetic acid (97%) were purchased
from Alfa Aesar. Acetic acid (97%) was purchased from
Millipore Sigma. Fuming sulfuric acid (20%) was purchased
from Beantown Chemical. Propionic acid (99.9%) was pur-
chased from Mallinckrodt. Formica acid (88%) was purchased
from Aqua Solutions. Ethanol (200 proof) was purchased from
Koptec. All reagents and solvents were used without further
purification.
Benzyl alcohol (108 mg, 1.0 mmol) and 3,4-dyhydro-2H-pyran
(
DHP) (101 mg, 1.2 mmol) were dissolved in PAO432 (10 mL)
containing 0.05 mmol of 2 or 3. The reaction mixture was
stirred at room temperature for 1 h. The reaction mixture was
then placed into a test tube, and acetonitrile (3 mL) was
added. The resultant biphasic mixture was then thoroughly
mixed and then separated by centrifugation for 5 min at 1000
rpm. The acetonitrile and PAO432 layers were then separated
using a pipette. The PAO432 phase containing the PIB-bound
acid 2 or 3 was recycled by addition with fresh alcohol and
DHP. This process was carried out 10 times with no measur-
able loss of the PAO solvent. The acetonitrile phases from
these 10 cycles were then combined, washed twice with 5 mL
Synthesis of PIB-bound alkylsulfonic acid 2
Alkene terminated polyisobutylene 1 (10 g, 10 mmol) (M
w
=
portions of heptane, and then concentrated to give an average
1
1
000) was dissolved in a 50 : 50 mixture of heptane and isolated yield of 91% and 89% for 2 and 3 respectively.
H
ethanol (200 mL) in a 250 mL round-bottomed flask. To this NMR (400 MHz, CDCl ) δ 7.41–7.29 (m, 5H), 4.82 (d, J = 12 Hz,
3
was added azobisisobutyronitrile (0.165 g, 1 mmol) and thioa- 1H), 4.72 (t, J = 3.4, 1H), 4.53 (d, J = 12 Hz, 1H), 4.00–3.92 (m,
cetic acid (2.15 mL, 30 mmol). The reaction mixture was then 1H), 3.63–3.55 (m, 1H), 1.94–1.54 (m, 6H).
irradiated overnight with a 254 nm UV lamp. After stirring
Comparison of acid recyclability in the 3,4-dyhydro-2H-pyran
protection of benzyl alcohol
overnight, water (50 mL) was added to induce biphasic separ-
ation of this latent biphasic solvent mixture. The aqueous
3
4
phase was separated, and the organic phase was washed 3 Benzyl alcohol (108 mg, 1.0 mmol) and 3,4-dyhydro-2H-pyran
times with 50 mL portions of 90% aqueous ethanol. The (DHP) (101 mg, 1.2 mmol) were dissolved in 10 mL of heptane
organic phase was then concentrated under reduced pressure containing 0.05 mmol of 2, 3, or 5. This solution was stirred at
and the resulting viscous residue was dissolved in 100 mL of room temperature for 1 h. After 1 h, a small aliquot of the
THF in a 250 mL round-bottomed flask. This thioester was heptane phase was dissolved in CDCl
3
and the conversion was
then hydrolyzed and oxidized using 35% hydrogen peroxide determined by H NMR spectroscopy. The remaining heptane
3.7 mL) and 88% formic acid (23 mL). This reaction mixture was extracted with 5 mL of acetonitrile to separate the product,
1
(
was stirred overnight at 60 °C. After stirring overnight, the reac- from heptane. The heptane phase containing the acid catalysts
tion mixture was concentrated under reduced pressure and the 2, 3, or 5 was charged with fresh alcohol and DHP and the
residue was dissolved in 100 mL of heptane. This heptane process was repeated for 5 cycles.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 1266–1273 | 1271

View Article Online
Paper
Green Chemistry
Recycling of PAO/acid in a multicomponent synthesis of 3,4-
dihydropyrimidin-2-one
catalyst was charged with fresh hexanoic acid and methanol
and the procedure was repeated. This process was repeated 5
times, affording methy hexanoate with an average isolated
Benzaldehyde (212 mg, 2 mmol), ethyl acetoacetate (260 mg,
1
yield of 80% and 81% for 2 and 3, respectively. H NMR
2
mmol), and urea (180 mg, 3 mmol) were dissolved in 10 mL
(
400 MHz, CDCl ) δ 3.60 (s, 3H), 3.23 (t, J = 7.6 Hz, 2H),
3
of PAO432 in a capped vial. Then 0.5 mmol of either 2 or 3 was
added. The resulting mixture was then stirred at 90 °C for
1.59–1.52 (m, 2H), 1.3–1.19 (m, 4H), 0.83 (t, 3H).
1
2 h. Then the mixture was transferred to a centrifuge tube
Calculated conversion in esterifications using a 1 : 1 ratio of
carboxylic acid/alcohol
and centrifuged for 5 min at 1000 rpm. The PAO432 and acid
were then decanted and reused in a subsequent reaction cycle
with fresh reagents. The solid product and excess urea were
from all 7 cycles was combined and recrystallized from ethanol
to remove excess urea. The recrystallized 3,4-dihydropyrimidin-
The change in the Gibbs free energy was approximated as the
heat of formation of the products minus the reactants with the
assumption that the change in entropy should be minimal.
Using the heats of formation from NIST (see ESI†) the equili-
brium constant for the reaction was calculated and this equili-
brium used to calculate the expected conversion at 70 °C.
2
-one was then washed twice with 5 mL portions of heptane to
remove any residual. The product was isolated in average iso-
lated yield of 82% and 81% for 2 and 3 respectively. H NMR
1
(
400 MHz, CDCl ) δ 7.51 (br, chemical shift is concentration
3
dependent, 1H), 7.35–7.29 (m, 5H), 5.53 (br, chemical shift is
concentration dependent, 1H), 5.43 (dd, J = 2.5 Hz, 1H), 4.1
Author contributions
(
dq, J = 2.5, 7.3 Hz, 2H), 2.38 (s, 3H), 1.19 (t, J = 7.3 Hz, 3H),
Christopher B. Watson contributed to this work by conceptual-
ization, formal analysis, investigation, methodology, visualiza-
tion, and writing. Adrianna Kuechle contributed to this work
by investigation. David E. Bergbreiter contributed to this work
by con-ceptualization, funding acquisition, project adminis-
tration, resources, supervision, and writing.
2
8
mp = 201–203 °C.
Rate comparison for esterification of propionic acid with
benzyl alcohol using 2, 3, and p-toluenesulfonic acid
Propionic acid (740 mg, 10 mmol) and 10 mmol of benzyl
alcohol (1.08 g, 10 mmol) were dissolved in 10 mL of PAO432
(using 0.1 mmol of 2 or 3 as a catalyst) or in toluene (using
0
.1 mmol of p-TsOH as a catalyst). The reaction was then
Conflicts of interest
stirred at 70 °C. Aliquots of the reaction were removed period-
ically and analyzed by H NMR spectroscopy. The conversion There are no conflicts to declare.
percentage was determined from the integration ratio of the
1
alcohol protons at 4.6 δ and the ester protons at 5.0 δ.
Acknowledgements
General esterification procedure
Support of this research by the Robert A. Welch Foundation
A carboxylic acid (10 mmol) and an alcohol (10 mmol) were
added to a 20 mL vial and dissolved in PAO432 (10 mL) contain-
ing 0.1 mmol of either 2 or 3. The vial was then capped, and
the reaction mixture was heated to 70 °C and stirred for 5 h.
(Grant A-0639). The donation of poly(α-olefin)s by Exxon
(Dr Wenning Han) and of polyisobutylene by the TPC Group
(Dr Michael Nutt) are gratefully acknowledged.
3
After 5 h, an aliquot was taken, and dissolved in CDCl . The
conversion of starting material to product was then deter-
mined by comparison of the integral of the alcohol for singlets
at ca. 4.6 δ or 3.4 δ (benzyl alcohol and methanol, respectively),
triplet at ca. 3.6 δ (propanol) and quartet at ca. 3.7 (ethanol)
and the ester proton singlet at ca. 5.0 δ and 3.6 δ (benzyl esters
and methyl esters, respectively), triplet at ca. 4.1 δ (propyl
Notes and references
1
F. Liu, K. Huang, A. Zheng, F.-S. Xiao and S. Dai,
Hydrophobic solid acids and their catalytic applications in
green and sustainable chemistry, ACS Catal., 2018, 8, 372–391.
J. H. Clark, Solid acids for green chemistry, Acc. Chem. Res.,
2
3
1
esters) and quartet at ca. 4.1 δ (ethyl esters) using H NMR
spectroscopy.
2002, 35(9), 791–797.
M. L. Granados, A. C. Alba-Rubio, I. Sadaba, R. Mariscal,
I. Mateos-Apariciob and A. Heras, Poly(styrenesulphonic)
acid: an active and reusable acid catalyst soluble in polar
solvents, Green Chem., 2011, 13, 3203–3212.
Recycling of PAO/acid in the esterification of hexanoic acid
with methanol
Hexanoic acid (1.16 g, 10 mmol) and methanol (320 mg,
4 W. Song, Y. Zhang, A. Varyambath, J. S. Kim and I. Kim,
Sulfonic acid modified hollow polymer nanospheres with
tunable wall-thickness for improving biodiesel synthesis
efficiency, Green Chem., 2020, 22, 3572–3583.
5 X. Zhang, Y. Zhao, S. Xu, Y. Yang, J. Liu, Y. Wei and
Q. Yang, Polystyrene sulphonic acid resins with enhanced
1
0 mmol) were dissolved in PAO432 (10 mL) in a 25 mL round-
bottomed flask. Then 0.1 mmol of either 2 or 3 was added.
The reaction was carried out at 70 °C for 5 h. After 5 h, the
product was isolated and purified by vacuum distillation at
reduced pressure. The remaining solution of PAO432 and the
1272 | Green Chem., 2021, 23, 1266–1273
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
acid strength via macromolecular self-assembly within con- 21 Polyisobutylene Overview, https://www.tpcgrp.com/pro-
fined nanospace, Nat. Commun., 2014, 5, 3170. ducts/polyisobutylene-pib, (Accessed January 13, 2021).
H. Tang, N. Li, F. Chen, G. Li, A. Wang, Y. Cong, X. Wang 22 Polyisobutene/Polyisobutylene (PIB), https://www.basf.com/
6
and T. Zhang, Highly efficient synthesis of 5-hydroxy-
methylfurfural with carbohydrates over renewable cyclopen-
tanone-based acid resin, Green Chem., 2017, 19, 1855–1860.
D. Prat, J. Hayler and A. Wells, A survey of solvent selection
guides, Green Chem., 2014, 16, 4546–4551.
J. H. Clark and S. J. Tavener, Alternative solvents: shades of
green, Org. Process Res. Dev., 2007, 11, 149–155.
P. Pollet, E. A. Davey, E. E. Urena-Benavides, C. A. Eckert 24 C. B. Watson, D. Tan and D. E. Bergbreiter, Enthalpy-driven
global/en/products/segments/industrial_solutions/perform-
ance_chemicals/business/fuel-and-lubricant-solutions/poly-
isobutenes–pib-/highly-reactive-pib.html, (Accessed January
13, 2021).
7
8
9
23 Low Viscosity Polyalphaolefins, https://www.exxonmobil-
chemical.com/en/products/synthetic-base-stocks/low-vis-
cosity-polyalphaolefins, (Accessed January 13, 2021).
and C. L. Liotta, Solvents for sustainable chemical pro-
cesses, Green Chem., 2014, 16, 1034–1055.
polyisobutylene depolymerization, Macromolecules, 2019,
52, 3042–3048.
1
0 D. J. C. Constable, C. Jimenez-Gonzalez and 25 C. Chao and D. E. Bergbreiter, Highly organic phase
R. K. Henderson, Perspective on solvent use in the pharma-
ceutical industry, Org. Process Res. Dev., 2007, 11, 133–137.
1 F. Pena-Pereira, A. Kloskowskim and J. Namiesnik,
soluble polyisobutylene-bound cobalt phthalocyanines as
recyclable catalysts for nitroarene reduction, Catal.
Commun., 2016, 77, 89–93.
1
Perspectives on the replacement of harmful organic sol- 26 N. Rosenfeld, C. Chao, C. B. Watson, M. P. Kumar,
vents in analytical methodologies: a framework toward the
implementation of a generation of eco-friendly alternatives,
Green Chem., 2015, 17, 3687–3705.
S. T. Madrahimov and D. E. Bergbreiter, Solubilization of
silica nanoparticles in alkanes and poly(α-olefin)s by func-
tionalized polyisobutylene oligomers, J. Polym. Sci., 2020,
58, 1144–1152.
1
2 F. P. Byrne, S. Jin, G. Paggiola, T. H. M. Petchey, J. H. Clark,
T. J. Farmer, A. J. Hunt, C. R. McElroy and J. Sherwood, 27 A. Bongini, G. Cardillo, M. Orena and S. Sandria, A simple
Tools and techniques for solvent selection: green solvent
and practical method for tetrahydropyranylation of alco-
selection guides, Sustainable Chem. Processes, 2016, 4, 7.
hols and phenols, Synthesis, 1979, 618–620.
1
1
1
3 C. J. Clarke, W. Tu, O. Levers, A. Bröhl and J. P. Hallett, 28 T. Jin, H. Wang, C. Xing, X. Li and T. Li, An efficient one–
Green and sustainable solvents in chemical processes,
Chem. Rev., 2018, 118, 747–800.
4 M. Poliakoff, P. Licence and M. W. George, A new approach
pot synthesis of 3,4-dihydropyrimidin-2-ones catalyzed by
methanesulfonic acid, Synth. Commun., 2004, 34, 3009–
3016.
to sustainability: a Moore’s law for chemistry, Angew. 29 S. R. Jetti, D. Verma and S. Jain, An efficient one-pot green
Chem., Int. Ed., 2018, 57, 12590–12591.
protocol for the synthesis of 5-unsubstituted 3,4-dihydro-
pyrimidin-2(1H)-ones using recyclable Amberlyst 15 DRY as
a heterogeneous catalyst via three-component Biginelli-Like
reaction, Int. Scholarly Res. Not., 2012, 2012, 480989.
30 L. P. Hammet and M. A. Paul, A series of simple basic indi-
cators, III, the zero point of the acidity function scale,
J. Am. Chem. Soc., 1932, 56, 827–829.
5 S. Thavornpradit, J. M. Killough and D. E. Bergbreiter,
Minimizing solvent waste in catalytic reactions in highly
recyclable hydrocarbon solvents, Org. Biomol. Chem., 2020,
1
8, 4248–4256.
1
6 M. L. Harrell, T. Malinski, C. Torres-López, K. Gonzalez,
J. Suriboot and D. E. Bergbreiter, Alternatives for conven-
3
1
tional alkane solvents, J. Am. Chem. Soc., 2016, 138, 14650– 31 A. Zheng, S. B. Liu and F. Deng, P NMR chemical shifts
1
4657.
of phosphorus probes as reliable and practical acidity
scales for solid and liquid catalysts, Chem. Rev., 2017, 117,
12475–12531.
1
7 T. J. Malinski, H. S. Bazzi and D. E. Bergbreiter,
Sustainable hydrocarbon oligomer solvent systems for
sequestration of trace organics from water, ChemSusChem, 32 NIST Chemistry Webbook, https://webbook.nist.gov/,
019, 12, 416–419. (Accessed January 13, 2021).
8 L. R. Rudnick, Synthetics, Mineral Oils, and Bio-based 33 J. Y. Park, D. K. Kim and J. S. Lee, Esterification of free fatty
2
1
1
2
Lubricants: Chemistry and Technology, CRC Press, Boca
acids using water-tolerable Amberlyst as a heterogeneous
Rotan, 1st edn, 2005.
catalyst, Bioresour. Technol., 2010, 101, S62–S65.
9 D. J. Heldebrant, H. N. Witt, S. M. Walsh, T. Ellis, 34 D. E. Bergbreiter and S. D. Sung, Liquid/liquid biphasic
J. Rauscher and P. G. Jessop, Liquid polymers as solvents
for catalytic reductions, Green Chem., 2006, 8, 807–815.
recovery/reuse of soluble polymer– supported catalysts,
Adv. Synth. Catal., 2006, 348, 1352–1366.
0 B. A. Roberts, G. W. V. Cave, C. L. Raston and J. L. Scott, 35 J. Li, S. Sung, J. Tian and D. E. Bergbreiter, Polyisobutylene
Solvent-free synthesis of calix[4]resorcinarenes, Green
Chem., 2001, 3, 280–284.
supports—a non-polar hydrocarbon analog of PEG sup-
ports, Tetrahedron, 2005, 61, 12081–12092.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 1266–1273 | 1273