Dehydration of benzyl alcohols in phosphonium ionic liquids: Synthesis of ethers and alkenes

Article abstract of DOI:10.1002/adsc.201100445

The dehydration of benzylic alcohols has been studied in several phosphonium ionic liquids in the absence of any metal catalysts. Benzyl ethers and alkenes were obtained from primary and secondary benzylic alcohols in good to excellent yields for these reactions. Commercially available hydrophobic phosphonium ionic liquids containing the trihexyl(tetradecyl)phosphonium cation paired with six different anions were used for the reactions under microwave irradiation. The interaction of the substrate with the ionic liquid was investigated using different NMR techniques, such as NOESY NMR. The effects of cation and anions on the behaviour of these ionic liquids in the reactions were studied in order to understand the mechanism. A catalytic cycle is proposed involving activation of the benzyl alcohol by the phosphonium cation. Copyright

Full text of DOI:10.1002/adsc.201100445

FULL PAPERS
DOI: 10.1002/adsc.201100445
Dehydration of Benzyl Alcohols in Phosphonium Ionic Liquids:
Synthesis of Ethers and Alkenes
Hassan A. Kalviria and Francesca M. Kertona,*
a
Centre for Green Chemistry and Catalysis, Department of Chemistry, Memorial University of Newfoundland, St. Johnꢀs,
Newfoundland, Canada A1B 3X7
Fax : (+1)-709-864-3702; phone: (+1)-709-864-8089; e-mail: fkerton@mun.ca
Received: May 30, 2011; Revised: August 6, 2011; Published online: November 16, 2011
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adcs.201100445.
Abstract: The dehydration of benzylic alcohols has ionic liquid was investigated using different NMR
been studied in several phosphonium ionic liquids in techniques, such as NOESY NMR. The effects of
the absence of any metal catalysts. Benzyl ethers and cation and anions on the behaviour of these ionic liq-
alkenes were obtained from primary and secondary uids in the reactions were studied in order to under-
benzylic alcohols in good to excellent yields for these stand the mechanism. A catalytic cycle is proposed
reactions. Commercially available hydrophobic phos- involving activation of the benzyl alcohol by the
phonium ionic liquids containing the trihexyl(tetra- phosphonium cation.
decyl)phosphonium cation paired with six different
anions were used for the reactions under microwave Keywords: alkenes; benzyl ethers; catalysis; dehy-
irradiation. The interaction of the substrate with the dration; NOESY NMR; phosphonium ionic liquids
Introduction
phosphonium ionic liquids can participate in reac-
tions.[13–15] We have seen this in our own research
Imidazolium ionic liquids have been widely studied where Pd(II) salts can be reduced to yield Pd(0)
and used in many types of reactions.[1] In some cases, nanocrystals in phosphonium ionic liquids without
imidazolium ionic liquids have acted as both catalyst adding any reducing agents.[13] In this case, the phos-
and solvent.[2–4] Previously, we reported dehydrative phonium cation acts as the reducing agent.
etherifications of benzylic alcohols in imidazolium
Many phosphonium ionic liquids are commercially
ionic liquids, specifically, [BMIm]PF6 (1-butyl-3-meth- available and have been successfully used as reaction
ylimidazolium hexafluorophosphate) in the presence media[8,10,16] and catalysts[17–19] in a variety of organic
of Pd catalysts.[5] We chose to study such reactions in syntheses. They have also found industrial uses.[20,21]
ionic liquids because these solvents display an excep- To the best of our knowledge, reports concerning the
tional ability to stabilize ionic intermediates. Phos- use of phosphonium ionic liquids as catalysts contain
phonium ionic liquids have been studied to a lesser limited details about the mechanism of catalysis. For
extent although they possess some advantages over the oxychlorination reaction,[18] Perosa et al. proposed
imidazolium ionic liquids. Phosphonium ionic liquids that the nitrate anion is acting as the catalyst.
can be prepared without halide contamination, which
Dehydration reactions are becoming increasingly
is an important factor for the halogen-sensitive re- important in order to reduce the oxygen content of
agents.[6,7] Furthermore, phosphonium ionic liquids biomass feedstocks and convert them into more val-
are more thermally stable than imidazolium and qua- uable materials for chemical and allied industries.
ternary ammonium salts.[8,9] Also, tetraalkylphospho- Low selectivity and use of harsh reaction conditions
nium salts are much less acidic than imidazolium are the main problems associated with dehydration
ionic liquids.[10,11] Additionally, imidazolium ionic liq- reactions of alcohols. For example, sulfuric acid and
uids are not completely inert in some reactions (re- p-toluenesulfonic acid are among the strong acids that
agents can potentially interact with the p-system and/ are used for these reactions,[22,23] which need a neu-
or the acidic hydrogens on the ring periphery).[12] tralization process at the end of the pipeline. Hetero-
However, it has also recently been reported that geneous acid catalysts have also been used with some
3178
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 3178 – 3186

Dehydration of Benzyl Alcohols in Phosphonium Ionic Liquids: Synthesis of Ethers and Alkenes
success.[24,25] Furthermore, using alcohols as feedstocks the ionic liquid and our initial studies showed greater
for the preparation of ethers is desirable, as waste stability for phosphonium ionic liquids. As a starting
production (mainly alkali metal salts) and use of toxic point, 4-methoxybenzyl alcohol was used as the sub-
alkylating agents are two hazards encountered in clas- strate. Several phosphonium ionic liquids containing
sical ether syntheses.[22]
[P666,14]+, trihexyl(tetradecyl)phosphonium cation, and
Through our previous studies,[5] we became interest- six different counter ions were used as the reaction
ed in using phosphonium ionic liquids as reaction media (for their complete names and more informa-
media in the dehydration of benzylic alcohols to yield tion see Table 1). Among these phosphonium ionic
benzyl ethers and alkenes. Unlike our previous stud- liquids, [P666,14]DBS was revealed as having the great-
ies using imidazolium ionic liquids, we have since est compatibility with the reactions studied in terms
found that there is no need for Pd catalysts in phos- of higher yields and ease of product isolation.
phonium ionic liquids for these dehydration reactions
and the phosphonium cation can catalyze these reac- cation of 4-methoxybenzyl alcohol in [P666,14]DBS in
tions. the presence and absence of Pd(CH3CN)2Cl2 (5 mol%
Figure 1 shows the progress of dehydrative etherifi-
AHCTUNGTRENNUNG
Herein, dehydration reactions of benzylic alcohols with respect to 4-methoxybenzyl alcohol). For this ex-
are reported that can be performed within a short re- periment, separate reaction mixtures were prepared
action time via microwave irradiation. The reaction with an alcohol to ionic liquid mole ratio of 0.65 (Æ
conditions are very simple and highly atom economic, 0.05). The reaction mixtures were heated under mi-
since there is no need for metal catalysts and alkylat- crowave conditions at 1208C for different reaction
ing agents and the ionic liquids are recyclable. Also a times as indicated in Figure 1. Higher yields were ob-
catalytic cycle for the role of phosphonium ionic liq- served at shorter reaction times in the presence of
uids in these dehydration reactions is presented.
palladium, however in the absence of palladium
higher yields could be reached at longer reaction
times. In the presence of palladium at prolonged reac-
tion times product decomposition occurs[5] and benzyl
chloride is formed presumably from the chloride pres-
Results and Discussion
In our previous studies, very low conversions were ob- ent in the palladium precursor. Without palladium,
served in the absence of palladium for dehydrative the reaction progress can be seen to plateau. We attri-
etherification reactions in imidazolium ionic liquids,[5] bute this to the increasing water content of the reac-
however, we have since discovered that the reaction tion mixture. Under the reaction conditions studied
can be performed without any metal catalyst in phos- no considerable improvement in yields was observed
phonium ionic liquids (Scheme 1). Furthermore, in the presence of water scavengers, such as molecular
[BMIm]PF6 was not completely stable under the reac- sieves, and anhydrous sodium sulfate. Kappe et al.
tion conditions, as HF was released via hydrolysis of have previously noted that molecular sieves are not
Scheme 1. Dehydrative etherification reaction of 4-methoxybenzyl alcohol in [P666,14]DBS and [BMIm]PF6.
Table 1. Names and some general physicochemical properties of the trihexyl(tetradecyl)phosphonium ionic liquids that were
used in this work. Data were collected from the cited references.
Ionic liquid
Anion name
Physical state at room
temperature
Approximate thermal
stability [8C]
A
chloride
bromide
liquid, Tg =À568C[11]
335[27]
320[16c]
380[27]
350[28]
360[27]
340[16c]
N
U
ACHTUNGTRENNUNG
G
liquid[16c]
]
G
bis(trifluoromethanesufonyl)amide
dodecylbenzenesulfonate
dicyanamide
liquid, Tg =À768C[11]
liquid[16c]
liquid, Tg =À678C[11]
E
]A
bis(2,4,4-trimethylphenyl)phosphinate
liquid[16c]
Adv. Synth. Catal. 2011, 353, 3178 – 3186
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3179

Hassan A. Kalviri and Francesca M. Kerton
FULL PAPERS
(Supporting Information) shows that higher yields
were obtained at higher alcohol concentrations.
Several substrates were screened for this reaction
and the results are presented in Table 2. Table 2
shows that higher yields were obtained for electron-
rich substrates (entries 1–5). No conversion was ob-
served for control reactions, neat alcohol and no ionic
liquid, under these conditions. Since the reactions
were performed under batch conditions in closed ves-
sels, the released water accumulated in the system
and based on the Le Chꢂtelierꢀs principle the reaction
was prohibited from achieving completion. No etheri-
fication was observed for aliphatic alcohols or phe-
nols. However, asymmetric etherification was possible
between phenols and benzyl alcohols but with low
Figure 1. Reaction progress for dehydrative etherification of
4-methoxybenzyl alcohol at 1208C in [P666,14]DBS in the
presence of 5 mol% PdACTHNUTGRNE(UNG CH3CN)2Cl2 and with no palladium. yields and poor selectivities. For example, a 16%
yield was obtained for phenyl benzyl ether from the
condensation of 4-methoxybenzyl alcohol and 4-me-
thoxyphenol at 1508C (for GC-MS analysis see Fig-
ure S1 in the Supporting Information).
Alkenes were produced as the major products for
dehydration reactions of secondary benzylic alcohols
which have b-hydrogens (Table 3). Different catalytic
systems have been previously reported for this trans-
formation. For example, recently Re2O7 or iodine
under solvent-free conditions were reported as effi-
cient catalysts for the dehydration of benzylic alco-
hols.[29,30] Also, aromatic alkenes can have biological
properties. For example, some of the products in this
paper (methoxyarylalkenes) are building blocks of
asarones, which have bactericide and pesticide activi-
ties and can be used to treat diphtheria and ty-
phoid.[31] Since the alkene bond in indene is more
stable than that in 4-methoxystyrene and styrene, (in-
ternal double bonds which are more substituted are
more stable than terminal double bonds),[22] higher
yields were observed for the dehydration of 1-indanol
than 1-(4-methoxyphenyl)ethanol and 1-phenyletha-
Figure 2. Plot of ln[4-methoxybenzyl alcohol] vs. time for
dehydrative etherification in [P666,14]DBS at 1208C.
Table 2. Etherification reaction of different benzylic alcohol
substrates in [P666,14]DBS under microwave irradiation.
efficient water scavengers under microwave-assisted
reaction conditions.[26]
In the aforementioned experiments, the concentra-
tion of 4-methoxybenzyl alcohol was determined
1
based on H NMR spectra of the reaction mixtures.
Entry
X
Yield[a] [%]
Time [min]/Temp. [8C]
For these experiments, a plot of ln[4-methoxybenzyl
alcohol] vs. time gives a straight line (Figure 2).
Therefore, the ionic liquid-catalyzed dehydrative
etherification reaction under these conditions is first
order with respect to benzyl alcohol.
The etherification reaction was also performed on
reaction mixtures with different concentrations of 4-
methoxybenzyl alcohol. These results were in agree-
ment with the first order nature of the reaction
(Table S1 in the Supporting Information). Table S1
1
2
3
4
5
6
7
4-MeO
3-MeO
4-EtO
4-Me
2-Me
H
72
53
50
61
63
34
26
35/120
25/200
20/140
20/180
20/180
60/180
45/170
4-Cl
[a]
1
Yields were calculated using H NMR spectroscopy with
acetophenone or mesitylene as standards.
3180
asc.wiley-vch.de
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 3178 – 3186

Dehydration of Benzyl Alcohols in Phosphonium Ionic Liquids: Synthesis of Ethers and Alkenes
Table 3. Dehydration reactions of secondary benzylic alcohols that have b-hydrogens.
Entry
1
Substrate
Alkene
Ionic liquid
Yield[a] [%]
96
Time [min]/Temp. [8C]
[P666,14]DBS
30/120
2
3
4
(1)
(1)
(1)
(1’)
(1’)
(1’)
N
99
30/120
30/120
30/120
99[b]
30
5
[P666,14]DBS
67
45/160
6
7
(2)
(2)
(2’)
(2’)
N
85
16
45/160
45/160
8
[P666,14]DBS
70
20/120
9
10
(3)
(3)
(3’)
(3’)
N
63
58
20/120
20/120
[a]
1
Yields were calculated using H NMR spectroscopy with acetophenone or mesitylene as standards.
[b]
Recycled [P666,14]Br was used. The reaction mixture from entry 2 was vacuum distilled and the IL reloaded with (1) and
reused.
nol. The lower yield in entry 5 compared to entry 6 is, ionic liquids containing the same cation as
possibly, because of polymerization of styrene in [P666,14]DBS but differing anions. The pH of these
[P666,14]DBS. After reaction, most mixtures were trans- ionic liquids was measured to probe the effect of acid-
parent dark yellow-orange solutions while for entry 5 ity on the reaction mechanism. The results have been
the reaction mixture was opaque. No attempt was summarized in Figure 3. Figure 3 shows that all these
made to fully characterize the insoluble by-product. phosphonium ionic liquids have acidic properties and
The product from the reaction mixture in Table 3, cover roughly three different acidic pH ranges. Al-
entry 2 was vacuum distilled (3 h, 70 mtorr and though the acidic properties of these ionic liquids can
1108C) and the IL was reloaded with (1) and a 99% come from possible impurities,[21] it has been shown
yield was obtained (Table 3, entry 3). For a picture of that these ionic liquids have an inherent acidic
the vacuum distillation apparatus see Figure S2 in the nature.[15,32] 1H and 31P NMR spectra and elemental
Supporting Information. This recycled ionic liquid analysis did not indicate the presence of any impuri-
was also used for dehydrative etherification of 4-me- ties in these ionic liquids. Surprisingly, high and low
thoxybenzyl alcohol and a 71% yield was obtained, yields were observed for different ionic liquids across
which is comparable with the reaction in fresh the pH range 1.5–4.5. These results highlight the point
[P666,14]DBS (Table 2, entry 1). It is interesting that the that there is no correlation between the pH of the
ionic liquid can be re-used in a different dehydration ionic liquids and the reaction yield. Therefore, the re-
to that which for it was originally used in without any actions are not acid catalyzed. The method that was
problems.
used for the pH measurements did not directly mea-
sure the acidity of the ionic liquids (see the Experi-
mental Section for more details).[32] However, it pro-
vides a reasonable estimate for the relative acidities
of the ionic liquids. Also, no correlation between the
Effect of pH
Dehydrative etherification of 4-methoxybenzyl alco- pH of the ionic liquids and yield was seen in the de-
hol was also performed in five other phosphonium hydration of secondary benzylic alcohols (Table 3).
Adv. Synth. Catal. 2011, 353, 3178 – 3186
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3181

Hassan A. Kalviri and Francesca M. Kerton
FULL PAPERS
[P666,14]Cl, while for [P666,14]Br, 4.5 wt% was repor-
ted.[16c,34] However, we did not measure the water con-
tent of the ionic liquids in our study. This factor could
have a dramatic impact on the efficiency of the ionic
liquids in dehydration reactions. In these reactions
water is produced and the more hydrophobic the
ionic liquid is, the higher should be the yield obtained.
In order to confirm the negative effects of water, the
dehydrative etherification of 4-methoxybenzyl alcohol
in [P666,14]DBS was performed in the presence of in-
tentionally added 20 wt% water. The reaction was
carried out under microwave conditions (15 min at
1208C) with an alcohol to IL mole ratio of 0.7. A
24% yield was obtained for this reaction while, under
the same reaction conditions, a 64% yield was
achieved for the reaction without added water. There-
fore, the use of hydrophobic ionic liquids is a driving
force for enhancing the progress of the reactions to-
wards the products.
Figure 3. Percent yield for dehydrative etherification reac-
tion of 4-methoxybenzyl alcohol in six different phosphoni-
um ionic liquids. Horizontal axis shows the measured pH for
the corresponding neat ionic liquids. [P666,14]: trihexyl(tetra-
decyl)phosphonium; DBS: dodecylbenzene sulfonate; Br:
bromide; [Tf2N]: bis(trifluoromethanesulfonyl)amide; Cl:
chloride; [(i-C8)2PO2]: bis(2,4,4-trimethylpentyl)phosphinate;
[N(CN)2]: dicyanamide
Effect of Ionicity
Another possibility is that the ionic properties of the
ionic liquids could be a driving force for these reac-
tions via their ability to stabilize ionic intermediates.
Some of the ionic liquids, for example [P666,14]Cl, have
more molecular-related properties and can have a
more closely packed network extended structure with
the cations and anions interacting closely. On the
Significantly lower yields were obtained in [P666,14]Cl other hand, some other anions, like [NTf2]À are less
compared with the other two ionic liquids studied.
coordinating and can lead to more ionic properties
within the ionic liquids, that is, the cation and anion
are less closely associated with one another within the
Chloride Ion Levels
liquid phase.[35] Therefore, [P666,14
]ACHTUNGRTENNUGN
Seddon et al. have reported the importance of chlo- than the chloride ionic liquid, as it will be better able
ride impurities on the physical properties of ionic liq- to stabilize the ionic intermediates. The less ionic
uids.[33] We wondered if the low yield obtained using nature of the chloride ionic liquid was confirmed
[P666,14]Cl was due to chloride ions and if the other when a reaction mixture of [P666,14]Cl was vacuum dis-
two ionic liquids ([P666,14
]
[(i-C8)2PO2] and [P666,14
]
tilled at 80 mtorr (3 h at 808C). The presence of this
[N(CN)2]) which gave low yields were contaminated ionic liquid in the gas phase demonstrates that
with chloride ions. Therefore, chloride measurements [P666,14]Cl can behave in a molecular fashion, where
were performed on the six phosphonium ionic liquids there is significant bonding between the chloride and
1
studied. All contained less than 0.3 wt% chloride, phosphonium ions. H and 31P NMR spectra (Support-
except [P666,14]Cl which contained 8.02 wt% chloride ing Information, Figure S3) of the distilled material
in agreement with its formulation.
showed the presence of [P666,14]Cl in addition to the
desired product. However, under the same vacuum
distillation conditions no peaks related to the
ionic liquid were observed for samples prepared in
Effect of Water
AHCTNURTGE[GNNUN P666,14]AHCTUNGTRENN[GNU NTf2] or [P666,14]DBS. Thus, these two ionic liq-
All six phosphonium ionic liquids studied are hydro- uids are less volatile and more ionic in nature com-
phobic, however their water capacities are not the pared with [P666,14]Cl. Therefore, they can provide
same. Water solubility in some of the ionic liquids is more suitable environments for ionic reaction inter-
increased due to the formation of reverse micelles. mediates and yields are thereby enhanced.
For example, a high water capacity, higher than
14 wt%, was reported for [P666,14]AHCUNTGREN[NUG (i-C8)2PO2] and
3182
asc.wiley-vch.de
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 3178 – 3186

Dehydration of Benzyl Alcohols in Phosphonium Ionic Liquids: Synthesis of Ethers and Alkenes
Interactions of Substrate with Phosphonium Cation
d, b and c and fewer interactions with the methylenes
adjacent to the phosphorus a (Figure 4). Regarding
NMR studies were performed on samples of the reac- the choice of ionic liquid used in this study, of the
tion mixtures and neat ionic liquids in order to collect ionic liquids used by us and that afforded good yields
more data to use in proposing a reaction mechanism. (Table 2 and Table 3), [P666,14]AHCNTUGRENUN[G NTf2] has the lowest
1H NMR spectra of the phosphonium ionic liquids viscosity. Therefore, better molecular tumbling in this
were obtained with higher resolution when using deu- ionic liquid provides lower signal broadening and
terated acetone rather than CDCl3 as the solvent higher resolution NMR spectra, especially for the
(Supporting Information, Figure S4 and Figure S5). neat sample used in the NOESY NMR experiment.
This demonstrates the proton exchange ability of all Furthermore, [P666,14]AHCTUNGERN[NUG NTf2] does not have alkyl chains
the methylene groups on the alkyl chains within the within its anion, which simplifies its NMR spectra
ionic liquids. The acidic nature of the protons on the compared with those of [P666,14]DBS.
methylene groups adjacent to the phosphorus atom
1H NMR spectra of neat [P666,14
]ACHTUNGRTEN[NGNU NTf2] with differ-
has been reported previously,[15] but significant levels ent concentrations of 4-methoxybenzyl alcohol were
of exchange are also observed for protons farther in agreement with the results from NOESY NMR
from the phosphorus atom. In order to investigate the (Supporting Information, Figure S6). The resolution
1
interactions of alcohols with phosphonium ionic liq- of the H NMR spectra decreased at higher alcohol
uids, NOESY (nuclear Overhauser effect spectrosco- concentration for the methylene groups farther from
py) NMR was used (Figure 4). The integrations in the the phosphorus (Supporting Information, Figure S6,
NOESY spectrum showed the greatest interactions of spectrum 3), while for the methylene groups adjacent
the CH2 benzylic protons with the methylene groups to the phosphorus, better resolution was obtained at
Figure 4. NOESY spectrum (selected region) of 4-methoxybenzyl alcohol in [P666,14]ACTHGUNETRN[NUG NTf2]. D2O was used as NMR lock sol-
vent in a coaxial NMR tube. Numbers on the contours in the Figure indicate integrals for the cross-peak resonances.
Adv. Synth. Catal. 2011, 353, 3178 – 3186
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3183

Hassan A. Kalviri and Francesca M. Kerton
FULL PAPERS
1
higher alcohol concentrations. This means that, in CD3OD and D2O (1:1 volume ratio) and a H NMR
these systems, there are more interactions and ex- spectrum obtained. A solution of NaOD (0.4M,
change between the alcohol and the methylenes la- 50 mL) was added to the abovementioned solution
belled b, c and d (Figure 4 and Supporting Informa- and after 40 min a second spectrum obtained. The in-
tion, Figure S4) than methylenes labelled a.
tegral ratios for the different methylene groups were
All the aforementioned discussion has been sum- normalized relative to the terminal methyl hydrogens.
marized in a catalytic cycle that is proposed in The results showed that protons of the three methyl-
Figure 5. Further support for the presence of ionic in- ene groups adjacent to the phosphorus, Ha, Hb and Hc
termediates can be obtained by considering the reso- in Figure 4, in the ionic liquid had undergone deuteri-
lution and chemical shifts of the protons in the um exchange (their integrals decreased over time).
1H NMR spectra of the ionic liquids in the presence 10% of “H” had undergone exchange for each of
and absence of 4-methoxybenzyl alcohol (Supporting these but was negligible for all other H environments.
Information,
Figure S6).
A
significant
shift
The stability of phosphonium cations in phosphoni-
(0.14 ppm) to lower frequency was observed in the um ionic liquids toward bases strongly depends on the
presence of alcohol for the protons adjacent to the size and strength of the base.[36,37] In the case of larger
phosphorus (protons labelled a) than the other pro- bases or reagents, the inner protons of the cation, Ha,
tons (0.08 ppm for protons labelled b and c). This are less accessible and the more accessible protons,
demonstrates an increase in the partial electron densi- Hc, etc., have significant interactions with these re-
ty next to the above-mentioned protons with increas- agents. Therefore, in the case of benzylic alcohols and
ing alcohol concentration. This would be expected for the mechanism proposed here it is plausible in that,
ionic intermediates such as those proposed in because of steric hindrance, the inner protons, Ha,
Figure 5.
have lower interactions with the benzylic protons. The
A similar mechanism could be followed by secon- proposed mechanism matches with the NMR evi-
dary alcohols with b-hydrogens. However, as intramo- dence, but more experimental and possibly theoretical
lecular elimination is kinetically favoured for these evidence is needed to exclude the possibility of an
molecules over intermolecular substitution, these re- ylide-catalyzed reaction, where Ha is abstracted by
actions yield alkenes rather than ethers.
the alcohol. Ylides have been shown to act as reactive
A deuterium isotope exchange experiment was also intermediates in a variety of organic reactions and
performed on a sample of [P666,14]ACHTNUGRTENNUNG
[NTf2], the same they can also activate alcohols.[38] At this time we
ionic liquid that was used for the NOESY NMR ex- have obtained no evidence (e.g., 31P NMR data) to
periment. The procedure was modified from the support ylide formation, but it cannot be excluded. It
method that was used by Chu et al. for [P666,14]Cl.[15] is likely that any reactive intermediates, for example,
The ionic liquid, 16.6 mg, was dissolved in 600 mL an ylide, would be present on a very short time scale
before reacting and therefore would not be observa-
ble by NMR spectroscopy.
Conclusions
A simple and atom-economic method for the conver-
sion of benzylic alcohols to their corresponding ethers
and alkenes has been developed. The method is com-
patible with the principles of green chemistry, since
alkylating agents and metal catalysts are avoided.
Phosphonium ionic liquids, which are more thermally
stable and occasionally cheaper than imidazolium
ionic liquids have been used as both the reaction
media and catalysts. The ionic liquids are recyclable
through vacuum distillation of the reaction mixtures.
Both solvent screening and NMR studies support the
formation of an ionic intermediate. The protons in b,
g and d positions of the alkyl chains in the ionic liq-
uids are seen via NOESY NMR to undergo ready ex-
Figure 5. Proposed catalytic cycle for dehydrative etherifica-
change with the acidic protons of the alcohol. In the
tion of benzylic alcohols in phosphonium ionic liquids. The
presence of NaOD, CD3OD (a smaller alcohol) and
D2O, a, b and g protons undergo deuterium exchange.
Theses results show that phosphonium ionic liquids
counterion of the ionic liquid has been omitted for clarity.
H+ abstraction may also be occurring at the C a to the P
atom.
3184
asc.wiley-vch.de
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 3178 – 3186

Dehydration of Benzyl Alcohols in Phosphonium Ionic Liquids: Synthesis of Ethers and Alkenes
to identify the products formed (e.g., Figure S7 to Fig-
can act as the catalyst and reaction media in dehydra-
tion reactions. Further studies are needed in order to
fully understand the mechanism and in order to scale
up the reactions and increase substrate scope. One
possible route for increasing yields and improving
scale-up potential would be to use a supported ver-
sion of these ionic liquids and perform dehydration
reactions in a continuous-flow manner.
ure S12). Product separation can be performed either by sol-
vent extraction or distillation. Solvent extraction is possible
for [P666,14]Cl with a 3:1 volume ratio of heptane:water. No
degradation was observed for the ionic liquids during the re-
1
actions studied. H NMR spectra of the ionic liquids, before
and after the reactions, did not show any changes in the rel-
ative intensities and chemical shifts of the resonances assign-
able to the ionic liquids. Also high resolution mass spec-
trometry on [P666,14]DBS did not show any differences be-
tween the mass spectra of the ionic liquid before and after
the reaction. The same procedure can be used for the dehy-
dration reactions of alcohols 1–3 (Table 3) to yield alkenes.
Yields were calculated relative to the alcohols using
1H NMR spectroscopy with acetophenone (for [P666,14]DBS)
or mesitylene (for [P666,14]Br and [P666,14]Cl) as the internal
standards.
Experimental Section
General Information and Instrumentation
The ionic liquids [P666,14]DBS (Cyphos phosphonium IL 202)
and [P666,14]Cl (Cyphos phosphonium IL 101) were gift sam-
ples from Cytec Industries and the other ionic liquids were
purchased from Sigma–Aldrich. Ionic liquids were purified
following the procedure of Walsby, Clyburne and co-work-
ers.[14] Alcohols were purchased from Alfa Aesar and used
as received. A Biotage Initiator 2.5 microwave reactor was
employed for the reactions. 0.5–2 mL microwave vials with
PTFE seals were used for all the experiments. A Bruker
AVANCE 500 MHz and a Bruker AVANCE III 300 MHz
spectrometer equipped with XBB probe were used for the
General Procedure for pH Measurements of the Ionic
Liquids
A procedure similar to that reported by Deng et al. was
used to measure the pH of the ionic liquids.[32] In each case,
5.00 mL deionized water were added to 0.5 mmol ionic
liquid and the mixture was vigorously stirred for 2 min. The
mixtures were passed through silica plugs and the aqueous
phases were titrated with sodium hydroxide titrant. Each
sample was titrated three times.
1
NMR experiments. H NMR spectra were referenced to tet-
ramethylsilane as an internal standard and 31P NMR spec-
trum was referenced to external 81% H3PO4. GC-mass spec-
tra were recorded on an Agilent 7890 A GC system coupled
with an Agilent 5975C MS detector that was equipped with
a capillary column DB-5 (column length: 30.0 m and column
diameter: 0.25 mm). Elemental analysis and chloride ion
measurements of the ionic liquids were performed by Cana-
dian Microanalytical Service Ltd. (Delta, BC, Canada).
Supporting Information
1
Additional H and 31P NMR spectra, Table S1, a photo of
the vacuum distillation apparatus and some examples of
GC-mass spectra are available in the Supporting Informa-
tion.
Experimental Procedure for Dehydrative
Etherification Reaction
In a typical reaction, [P666,14]DBS (0.5070 g) was degassed
under vacuum. 4-Methoxybenzyl alcohol (0.0979 g) was
added to the ionic liquid in 0.5–2.0 mL Biotage microwave
vials and capped under a nitrogen flow. The solution was
stirred for a few minutes before putting the vials in the mi-
crowave reactor. The reaction mixture was heated for
35 min at 1208C. The absorptivity on the MW was set at
“high” for all the samples and the microwave power reached
43–45 W (in this case) during the reaction. No colour
change was observed in the reaction mixture upon heating,
however droplets of water could be observed near the top
of the microwave vial after reactions. Yields were calculated
Acknowledgements
We acknowledge support from NSERC and Canada Founda-
tion for Innovation in the form of a Discovery Grant, a Re-
search Tools and Instrument Grant and a Leaders Opportu-
nity Fund Award (F. M. K.). We also thank Memorial Uni-
versity for funding and Cytec Industries Inc. for samples of
[P666,14]DBS and [P666,14]Cl. Dr. Cꢀline Schneider (C-CART)
is thanked for assistance with some NMR experiments.
based on 1H NMR spectra using mesitylene or acetophe-
1
none as standards. H NMR chemical shifts related to CH2 References
benzylic (in benzyl ethers) and alkene bonds (in alkenes)
have been presented in the Supporting Information,
Table S2 and Table S3 and the data are comparable with re-
ported values.[5,39] In the case of styrene and 4-methoxystyr-
ene, the products were compared with commercially avail-
able standards. The progress of the reactions was monitored
using thin layer chromatography (TLC). GC-MS analysis
could be performed on the TLC spots extracted with diethyl
ether or direct extraction of the reaction mixture, if possible,
[1] J. P. Hallett, T. Welton, Chem. Rev. 2011, 111, 3508–
3576.
[2] T. Welton, Coord. Chem. Rev. 2004, 248, 2459–2477.
[3] S. Chowdhury, R. S. Mohan, J. L. Scott, Tetrahedron
2007, 63, 2363–2389.
[4] R. Sheldon, Chem. Commun. 2001, 2399–2407.
[5] H. A. Kalviri, C. F. Petten, F. M. Kerton, Chem.
Commun. 2009, 5171–5173.
Adv. Synth. Catal. 2011, 353, 3178 – 3186
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3185

Hassan A. Kalviri and Francesca M. Kerton
FULL PAPERS
[6] L. G. Bonnet, B. M. Kariuki, Eur. J. Inorg. Chem. 2006,
437–446.
[7] B. Kirchner, (Ed.), Ionic Liquids, in: Top. Curr. Chem.
[20] N. V. Plechkova, K. R. Seddon, Chem. Soc. Rev. 2008,
37, 123–150.
[21] C. J. Bradaric, A. Downard, C. Kennedy, A. J. Robert-
son, Y. Zhou, Green Chem. 2003, 5, 143–152.
[22] M. B. Smith, J. March, (Ed.), Marchꢁs Advanced Organ-
ic Chemistry, John Wiley & Sons, Hoboken, New
Jersey, 2007, pp 1513–1516, 529–532 and 1497–1501.
[23] E. C. Wijnbelt, E. H. Hekelaar, Patent WO 2010110841,
2010.
2009, 290, 345.
[8] E. Janus, W. Stefaniak, Catal. Lett. 2008, 124, 105–110.
[9] K. Tsunashima, M. Sugiya, Electrochem. Commun.
2007, 9, 2353–2358.
[10] B. D. Powell, G. L. Powell, P. C. Reeves, Lett. Org.
Chem. 2005, 2, 550–553.
[24] L. R. Pizzio, P. G. Vazquez, C. V. Caceres, M. N.
Blanco, E. N. Alesso, M. I. Erlich, R. Torviso, L. Fin-
kielsztein, B. Lantano, G. Y. Moltrasio, J. M. Aguirre,
Catal. Lett. 2004, 93, 67–73.
[25] A. Tarlani, A. Riahi, M. Abedini, M. M. Amini, J.
Muzart, J. Mol. Catal. A: Chem. 2006, 260, 187–189.
[26] M. Baghbanzadeh, C. O. Kappe, Aust. J. Chem. 2009,
62, 244–249.
[27] Z. S. Breitbach, D. W. Armstrong, Anal. Bioanal.
Chem. 2008, 390, 1605–1617.
[28] L. Yu, D. Garcia, R. Ren, X. Zeng, Chem. Commun.
2005, 2277–2279.
[29] T. J. Korstanje, J. T. B. H. Jastrzebski, R. J. M. Klein
Gebbink, ChemSusChem 2010, 3, 695–697.
[11] R. E. Del Sesto, C. Corley, A. Robertson, J. S. Wilkes,
J. Organomet. Chem. 2005, 690, 2536–2542.
[12] A. Sharma, R. Kumar, N. Sharma, V. Kumar, A. K.
Sinha, Adv. Synth. Catal. 2008, 350, 2910–2920.
[13] H. A. Kalviri, F. M. Kerton, Green Chem. 2011, 13,
681–686.
[14] T. Ramnial, S. A. Taylor, M. L. Bender, B. Gorodetsky,
P. T. K. Lee, D. A. Dickie, B. M. McCollum, C. C. Pye,
C. J. Walsby, J. A. C. Clyburne, J. Org. Chem. 2008, 73,
801–812.
[15] M. Tseng, H. Kan, Y. Chu, Tetrahedron Lett. 2007, 48,
9085–9089.
[16] a) T. Itoh, K. Kude, S. Hayase, M. Kawatsura, Tetrahe-
dron Lett. 2007, 48, 7774–7777; b) V. I. Parvulescu, C.
Hardacre, Chem. Rev. 2007, 107, 2615–2665; c) K. J.
Fraser, D. R. MacFarlane, Aust. J. Chem. 2009, 62, 309–
321; d) L. Ford, F. Atefi, R. D. Singer, P. J. Scammells,
Eur. J. Org. Chem. 2011, 942–950; e) J. McNulty, J. J.
Nair, A. Robertson, Org. Lett. 2007, 9, 4575–4578; f) J.
McNulty, S. Cheekoor, T. P. Bender, J. A. Coggan, Eur.
J. Org. Chem. 2007, 1423–1428; g) J. McNulty, J. J. Nair,
S. Cheekoori, V. Larichev, A. Capretta, A. J. Robert-
son, Chem. Eur. J. 2006, 12, 9314–9322; h) J. McNulty,
S. Cheekoori, J. J. Nair, V. Larichev, A. Capretta, A. J.
Robertson, Tetrahedron Lett. 2005, 46, 3641–3644;
i) D. A. Gerritsma, A. Robertson, J. McNulty, A. Cap-
retta, Tetrahedron Lett. 2004, 45, 7629–7631; j) H. Cao,
H. Alper, Org. Lett. 2010, 12, 4126–4129; k) Q. Yang,
H. Alper, J. Org. Chem. 2010, 75, 948–950; l) H. Cao,
L. McNamee, H. Alper, Org. Lett. 2008, 10, 5281–5284;
m) Q. Yang, A. Robertson, H. Alper, Org. Lett. 2008,
10, 5079–5082; n) H. Cao, L. McNamee, H. Alper, J.
Org. Chem. 2008, 73, 3530–3534.
[30] G. Stavber, M. Zupan, S. Stavber, Tetrahedron Lett.
2006, 47, 8463–8466.
[31] R. Kumar, A. Sharma, N. Sharma, V. Kumar, A. K.
Sinha, Eur. J. Org. Chem. 2008, 5577–5582.
[32] X. Cui, S. Zhang, F. Shi, Q. Zhang, X. Ma, L. Lu, Y.
Deng, ChemSusChem 2010, 3, 1043–1047.
[33] K. R. Seddon, A. Stark, M. Torres, Pure Appl. Chem.
2000, 72, 2275–2287.
[34] J. Martak, S. Schlosser, Sep. Purif. Technol. 2007, 57,
483–494.
[35] a) M. J. Earle, J. M. S. S. Esperanca, M. A. Gilea, J. N.
Canongia Lopes, L. P. N. Rebelo, J. W. Magee, K. R.
Seddon, J. A. Widegren, Nature 2006, 439, 831–834;
b) K. Ueno, H. Tokuda, M. Watanabe, Phys. Chem.
Chem. Phys. 2010, 12, 1649–1658; c) V. Ermolaev, V.
Miluykov, I. Rizvanov, D. Krivolapov, E. Zvereva, S.
Katsyuba, O. Sinyashin, R. Schmutzler, Dalton Trans.
2010, 39, 5564–5571.
[36] T. Raminal, D. D. Ino, J. A. C. Clyburne, Chem.
Commun. 2005, 325–327.
[37] S. Sowmiah, V. Srinivasadesikan, M. C. Tseng, Y. H.
Chu, Molecules 2009, 14, 3780–3813.
[38] For example, see: L. Kꢃrti, B. Czakꢄ, in: Strategic Ap-
plications of Named Reactions in Organic Synthesis
Background and Detailed Mechanisms, Elsevier Aca-
demic Press, California, 2005, pp 16, 27, 104.
[39] J. S. Yadav, B. V. Subba Reddy, K. Shiva Shankar, T.
Swamy, Tetrahedron Lett. 2009, 51, 46–48.
[17] S. R. Sarda, W. N. Jadhav, S. U. Tekale, G. V. Jadhav,
B. R. Patil, G. S. Suryawanshi, R. P. Pawar, Lett. Org.
Chem. 2009, 6, 481–484.
[18] M. Noe, A. Perosa, M. Selva, L. Zambelli, Green
Chem. 2010, 12, 1654–1660.
[19] S. R. Sarda, W. N Jadhav, A. S. Shete, K. B. Dhopte,
S. M. Sadawarte, P. J. Gadge, R. P. Pawar, Synth.
Commun. 2010, 40, 2178–2184.
3186
asc.wiley-vch.de
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 3178 – 3186