Boyer's Reaction and Transetherification: Mechanism and New Perspectives

Article abstract of DOI:10.1002/chir.22581

The progress and stereochemistry of Boyer's reaction were analyzed using several simple, chiral, alcoholic substrates, a variable amount of BiBr₃ and different solvents. Basic solvents inhibit the reaction, while cyclohexane works very well; thus, it was our choice for the present study. In contrast to previous works, BiBr₃ behaves as a true catalyst, being not consumed during the reaction. Although poisoning of the catalyst occurs to some extent, it does not prejudice the reaction yields (>90%). Gas chromatography/mass spectrometry (GC-MS) monitoring of the reaction revealed that, for example, in the presence of alcohol rac-1, isomeric ethers 4 transetherificate to 3. We propose a unifying mechanistic model for both Boyer's and transetherification reactions, in which the electronic properties of n-adducts intermediates, formed by combination of bismuth(III) of BiBr₃ and oxygen atoms of alcohols and ethers, play the key role for both the reactivity and the stereochemical outcome of the reaction.

Full text of DOI:10.1002/chir.22581

CHIRALITY 28:269–275 (2016)
Special Issue Article
Boyer’s Reaction and Transetherification: Mechanism and New
Perspectives
1
CATERINA FRASCHETTI,1 MAURIZIO SPERANZA,1 LAURA GUARCINI,1 GRAZIELLA ROSELLI,2 AND ANTONELLO FILIPPI
*
1Dipartimento di Chimica e Tecnologie del Farmaco – Sapienza, University of Rome, Rome, Italy
2School of Science & Technology (Chemistry Division), University of Camerino, Camerino, Italy
ABSTRACT
The progress and stereochemistry of Boyer’s reaction were analyzed using sev-
eral simple, chiral, alcoholic substrates, a variable amount of BiBr3 and different solvents. Basic
solvents inhibit the reaction, while cyclohexane works very well; thus, it was our choice for the
present study. In contrast to previous works, BiBr3 behaves as a true catalyst, being not
consumed during the reaction. Although poisoning of the catalyst occurs to some extent, it does
not prejudice the reaction yields (>90%). Gas chromatography/mass spectrometry (GC–MS)
monitoring of the reaction revealed that, for example, in the presence of alcohol rac-1, isomeric
ethers 4 transetherificate to 3. We propose a unifying mechanistic model for both Boyer’s and
transetherification reactions, in which the electronic properties of n-adducts intermediates,
formed by combination of bismuth(III) of BiBr3 and oxygen atoms of alcohols and ethers, play
the key role for both the reactivity and the stereochemical outcome of the reaction. Chirality
28:269–275, 2016. © 2016 Wiley Periodicals, Inc.
KEY WORDS: ethers; stereoselectivity; bismuth bromide; benzylation; Lewis acid catalysis
INTRODUCTION
reaction and to analyze its main potentialities. Moreover, it is
not clear if a stoichiometric amount of BiBr3 is really needed,
which is another point we want to investigate. Finally, CCl4
certainly is not the more desirable solvent, especially on
large-scale applications; therefore, first of all, we looked for
an alternative, more ecologically friendly solvent.
After our previous work on the stereochemistry of Boyer’s
reaction (Eq. 1) published some years ago,1 we now return
to the same topic, because in our opinion the potentialities
of the reaction has not been completely explored as a conse-
quence of the erroneous attribution of its stereochemical out-
come in the seminal work,2 as well as in other works and
reviews published in subsequent years.3–5
MATERIALS AND METHODS
ROH þ R’OH ꢀB→HiBrO3 ROR’
(1)
All the alcoholic substrates, as well as solvents, catalyst BiBr3, and
internal standard were purchased from commercial sources and used
without any further purification. The reaction mixture (each compo-
nent 0.625 M in 4 ml closed vial) was continuously stirred and sys-
tematically sampled by taking a small amount of solution, and then
analyzed by a gas chromatography/mass spectrometry (GC–MS) sys-
tem equipped with a chiral column (CP-Chirals-Dex CB, L 25 m, ID
0.25 mm, df 0.25 μm) in the 40–170°C temperature range. The syn-
thesis of an isomeric mixture of ethers 3 and 4, not commercially
available, was carried out by means of the same above simple proce-
dure in 20-ml vials. The raw mixture of ethereal diastereoisomers was
filtered, dried on anhydrous Na2SO4, and purified by preparative GC
(provided with an OV-17 column on Chromosorb WAW 80-100 mesh)
at 170°C. Control tests were performed by means of a linear ion trap
mass spectrometer coupled with an electrospray source (ESI-MS-IT -
Applied Biosystems, Foster City, CA, Mod. 3200Q LC/MS/MS).
2
In fact, Boyer’s reaction is relevant from a different point of
view. First, the number of synthetic routes to ethers is rather
low, if compared to other common functional groups;
therefore, development of further synthetic approaches is
highly desirable. Second, the classical methods, first of all
Williamson’s reaction, often need extreme environmental con-
ditions (i.e., alkaline media, which exclude base-sensitive com-
pounds as substrates), while Boyer’s method needs only
mildly acidic conditions and room temperature. Third, Boyer’s
reaction makes direct use of alcoholic substrates, a class of
substances relatively easy to synthesize or available from com-
mercial sources. Fourth, in Boyer’s reaction several derivatives
of benzylic alcohol have been used,2,4 but if other alcoholic
substrates are also effective, it is still an open question. Fifth,
even limiting the reaction to benzylic substrates, these are
widely employed as protecting groups of alcoholic func-
tions,6–9 therefore the availability of “soft” methods such as
Boyer’s one, where benzylic alcohols are used as reactant, ap-
pears very desirable. The erroneous conclusion of Boyer et al.4
about the loss of configuration of chiral alcoholic substrates
probably limited the application and the success of the method.
The aim of this work was to investigate the mechanism of the
© 2016 Wiley Periodicals, Inc.
[This article is part of the Thematic Issue: “Chirality in Separation Science and
Molecular Recognition Honoring Prof. F. Gasparrini.” See the first articles for
this special issue previously published in Volumes 27:9, 27:10, 27:11, 27:12,
28:1, 28:2 and 28:3. More special articles will be found in this issue as well as
in those to come.]
*Correspondence to: Antonello Filippi, Dipartimento di Chimica e Tecnologie
del Farmaco - Sapienza - University of Rome, P. le Aldo Moro,5–00185, Rome,
Italy. antonello.filippi@uniroma1.it
Received for publication 5 August 2015; Accepted 29 December 2015
DOI: 10.1002/chir.22581
Published online 28 February 2016 in Wiley Online Library
(wileyonlinelibrary.com).

270
FRASCHETTI ET AL.
Owing to the heterogeneous nature of the catalytic process and poison-
pentyl ether among the reaction products, pointing out the in-
ability of the aliphatic secondary alcohols rac-2P to self-
condense to ethers, different from the “activated,” benzylic alco-
hol rac-1, under the same conditions. However, after a few
hours the reaction in CH2Cl2 stops, probably because of cata-
lyst poisoning, as pointed out by the persistence of a substantial
amount of reagents after a couple of hours, when the composi-
tion of the reaction mixture does not change any more (Fig. 1),
despite an equimolar amount of BiBr3 was employed.
ing effects (see below), the reproducibility of reaction rates of replicated
experiments is rather low (relative error about 10%). Therefore, reaction
kinetics cannot be evaluated and the experimental data are discussed
on the grounds of reaction progress, i.e., relative increasing/decreasing
trends of reagents and products along reaction time.
RESULTS AND DISCUSSION
Choice of an Alternative Reaction Solvent
All the experiments specifically addressed to the selection of
alternative reaction solvents were performed under the same
concentration (0.625 M) and temperature (25°C) conditions al-
ready employed in previous works, and using racemic mixture
of reactants. In acetonitrile and in ethyl acetate, Boyer’s reac-
tion does not take place. Indeed, equimolar mixtures of the ra-
cemate of 2-pentanol (rac-2P) and of the racemate of 1-
phenylethanol (rac-1)6 in acetonitrile were completely de-
graded to several subproducts when BiBr3 was added in an
equimolar amount, while they were completely inert in ethyl ac-
etate under the same conditions. Probably the basic nature of
these solvents inhibits the acidity of bismuth in BiBr3 towards
alcoholic substrates. Accordingly, in CH2Cl2, a less basic sol-
vent, 1-phenylethyl 2-pentyl ethers (3) and di-1-phenylethyl
ethers (4), are promptly formed from the combination of rac-
2P and rac-1 and by reaction of rac-1 with itself. Figure 1
shows the composition of the reaction mixture at the corre-
sponding reaction time. In this figure, as well as in the other
ones reported in this article and in the supporting information
section, the experimental points represent mean values ob-
tained from at least three replicated experiments. Because of
the intrinsic poor reproducibility, about 10% relative error
should be taken into account. In order to specify the stereo-
chemistry of the ethereal products, a two-letter subscript (XY)
was associated with the corresponding representative number,
where X indicates the R/S configuration of the benzylic center,
while Y states the stereochemistry of the other chiral carbon
atom of the ether. Note that production of ethers 3 is not
stereoselective (3RR:3SS:3RS:3SR = 1:1:1:1). Ethers 4 are also
initially formed without any significant stereoselectivity, but
they seem to undergo an epimerization process
We found that in acetonitrile, ethyl acetate, CH2Cl2, as well
as in CCl4,1 the salt BiBr3 disperses rather than dissolves, sug-
gesting the occurrence of a heterogeneous catalytic process.
In passing to cyclohexane as a solvent, the general reactivity
pattern observed in CH2Cl2 qualitatively holds (cf. Figs. 1 vs.
2), but in this apolar, hydrophobic solvent the reaction seems
much more lively. In fact, reaction yields increase, as indi-
cated either by the lower residual amount of substrate rac-
2P at the end of the reaction, and by fast decreasing of the
benzylic reagent (rac-1) up to its complete disappearance.
At the beginning of the reaction, in cyclohexane at 25°C, be-
sides the formation of “symmetric” ethers 4 from rac-1,
“asymmetric” ethers 3 mainly arise from the condensation of
rac-1 with rac-2P (Scheme 1). However, for a reaction time
longer than 2 hours (i.e., in the absence of rac-1): 1) the rela-
tive amount of ethers 3 slowly increases at the expense of both
rac-2P and the “symmetric” ethers 4, suggesting the occur-
rence of the transetherification reaction 4 → 3 (Scheme 1);
2) diastereomeric distribution of ethers 4 changes favoring
←
←
(4RR→4RS→4SS) favoring the enantiomeric homochiral forms
(4RR e 4SS) over the meso heterochiral one 4RS (inset in
Fig. 1). It is also worth noting the absolute absence of di-2-
Scheme 1. Competitive etherification and transetherification reactions un-
der Boyer’s conditions.
Fig. 1. Reaction in CH2Cl2 between the racemate of 2-pentanol (rac-2P)
and the racemate of 1-phenylethanol (rac-1) in the presence of an equimolar
Fig. 2. Reaction in cyclohexane between the racemate of 2-pentanol (rac-2P)
and the racemate of 1-phenylethanol (rac-1) in the presence of an equimolar
amount of BiBr3. Full circle: rac-2P; empty circle: rac-1; full triangle: 3 (3RR
:
amount of BiBr3. Full circle: rac-2P; empty circle: rac-1; full triangle: 3 (3RR:
3
SS: 3RS: 3SR = 1: 1: 1: 1); empty triangle: 4. Inset: (4RR + 4SS)/4RS
.
3SS: 3RS: 3SR = 1: 1: 1: 1); empty triangle: 4. Inset: (4RR + 4SS)/4RS.
Chirality DOI 10.1002/chir

NEWS FROM BOYER’S REACTION
271
the homochiral forms (4RR e 4SS) over the meso heterochiral
one (4RS), as already observed in CH2Cl2, although to a
higher extent (cf. insets in Figs. 1, 2), confirming the occur-
←
←
rence of 4RR→4RS→4SS epimerization.
The residual amount of rac-2P and 4 after 24 h (Fig. 2) sug-
gests that poisoning of the catalyst cannot be excluded also in
cyclohexane. As in CH2Cl2, also in cyclohexane di-2-pentyl
ethers are not formed at all, and BiBr3 is suspended rather than
dissolved in the reaction medium. Therefore, cyclohexane
seems a good alternative to CCl4, and thus we decided to carry
on our work making use of cyclohexane as a solvent.
Role of BiBr3
Before proceeding, control experiments were performed in
order to ascertain the catalytic nature of the process in cyclo-
hexane and the susceptibility of the alcoholic reagents to
BiBr3. First of all, a suspension of BiBr3 in cyclohexane was
stirred at room temperature for 2 h. By means of a centrifuge,
the solid was separated and the liquid phase was then diluted
in methanol and analyzed by ESI-MS-IT, revealing the com-
plete lack of any signal attributable to bismuth-containing
compounds. Even more, a mixture of rac-1 and rac-2P
dissolved in such a liquid phase does not react after more
than 24 h. Similarly, the stirred 1:1 solution of rac-1 and
rac-2P in cyclohexane proved unreactive (even warming
the solution up to 65°C), until BiBr3 was added. Therefore,
BiBr3 is a true catalyst for the reaction of Figure 2, and in
our opinion it acts in the solid state (heterogeneous cataly-
sis), in contrast to previous conclusions.4,5 Furthermore, a cy-
clohexane solution of a mixture of simple primary and
secondary alcohols (1-pentanol, 1-heptanol, 2-butanol, 2-
pentanol, 2-heptanol, and 2-decanol) proved completely inert,
even in the presence of BiBr3, pointing out that “activated” al-
cohols are needed for the reaction. The progressive increase
of the (4RR + 4SS)/4RS ratio (inset in Fig. 2) suggests that a
similar reactivity check for 1-phenylethanol (1) deserves
more attention, as shown in the next section.
Fig. 3. Reaction of (ꢀ)-(S)-1-phenylethanol (1S) in the presence of an
equimolar amount of BiBr3 dispersed in cyclohexane. Timescale starting
from the introduction of BiBr3. Full circle: 1S; empty circle: 1R; full triangle:
4SS; empty triangle: 4RR; square: 4RS. Inset: circle 4SS/4RR; square
(4RR + 4SS)/4RS
.
Reaction Between (ꢀ)-(R)-2-Pentanol and the Racemate of
1-Phenylethanol in Cyclohexane
Having in mind the results of Figure 2, in the next step of
our work we focused on the behavior of the other alcoholic re-
actant rac-2P. However, taking into account the fast racemi-
zation of 1S (Fig. 3), we decided to make direct use of the
racemate rac-1, rather than its pure enantiomer 1S, in all
the experiments below. Moreover, being the reaction of
rac-1 with itself, to yield ethers 4, faster than its combination
with rac-2P, to yield ethers 3 (cf. 4 vs. 3 yields within the
first hour in Fig. 2), with the aim to maintain the concentra-
tion of rac-1 as low as possible, we also decided to modify
the original Boyer’s procedure as follows: rac-1 was dis-
solved in cyclohexane and added dropwise to a cyclohexane
solution containing the chiral alcohol (ꢀ)-(R)-2-pentanol
(2PR; Aldrich, Milwaukee, WI, R/S > 98/2) in the presence
of an equimolar amount of solid BiBr3. As a matter of fact,
such a procedure lead to a lower production of “symmetric”
ethers 4 in favor of the desired “asymmetric” 3 ones
(cf. Figs. 2 vs. 4). It is worth noting that: 1) ethers 3
are formed as 3RR e 3SR (in 1:1 ratio), whose 2-pentyl
Reaction of (ꢀ)-(S)-1-Phenylethanol in Cyclohexane
In order to deeply investigate the reactivity of 1 and the
intrinsic behavior of “symmetric” ethers 4 (without the
interference of other reactants), the chiral alcohol
(ꢀ)-(S)-1-phenylethanol (1S; Fluka, Buchs, Switzerland, R/
S < 1/99), instead of its racemate, was dissolved in cyclo-
hexane. At room temperature, the inertness of this stirred
solution was ascertained for 2 h by GC–MS monitoring.
Then dispersing an equimolar amount of BiBr3 in the same
solution promptly opens a very lively reactivity scenario
moiety maintains its original chirality of 2PR; 2) 3RS
,
3
SS, as well as di-2-pentyl ethers, were not found among
the reaction products; 3) GC–MS analysis reveal that, un-
like phenylethanol (Fig. 3), the chiral aliphatic substrate
2PR does not undergoes any racemization process.
←
(Fig. 3). Indeed, the results point out the fast 1S→1R race-
BiBr3 as a “True” Catalyst
mization of the starting substrate and the substantial lack
of diastereoselectivity in the formation of ethers 4 (open
squares in the inset of Fig. 3 not far from unit within the first
One more question to be addressed before a mechanistic
hypothesis can be formulated is whether the catalytic process
really needs an equimolar amount of BiBr3, as proposed by
Boyer et al.2 To this end, several experiments were per-
formed in cyclohexane adding rac-1 dropwise to solutions
of 2PR containing 10%, 20%, or 50% of BiBr3 with respect to
the molar concentration of the alcoholic reactants (Figs. 1-SM,
2-SM and 3-SM, respectively; Supporting Information). The re-
sults of these experiments can be summarized as follows.
Within 24 h: 1) in all cases the total molar amount of ethers pro-
duced (3 + 4) is higher than that of the catalyst employed, 2)
most of rac-1 is consumed, but some 2PR remain unreacted;
3) the higher the amount of BiBr3 used, the higher the reaction
Chirality DOI 10.1002/chir
←
←
1.5 h), whose subsequent 4RR→4RS→4SS epimerization, after
disappearance of either the starting alcohol 1S and the formed en-
antiomer 1R, goes on to make (4RR + 4SS)/4RS = 4.2. In the
meantime, the initial enantiomeric ratio 4SS/4RR = 1.4 de-
creases to 1.0 (full circles in the inset of Fig. 3).
It should also be noted that once the starting alcohol 1S and
its enantiomer 1R disappear and without any other alcoholic
reagent in solution, the total amount of ethers 4 do not
change any more, confirming the occurrence of the above-
mentioned 4 → 3 transetherification when rac-2P is avail-
able (Fig. 2 and Scheme 1).

272
yield and the extent of both 4RR→4RS→
FRASCHETTI ET AL.
←
←
←
←
4
SS epimerization and
racemization, ii) 4RR→4RS→4SS epimerization, iii) 4 → 3
transetherification, as well as the lower absolute yields of
products 3 and 4), according to the H2O inhibition
hypothesis.
4 → 3 transetherification; 4) adding further BiBr3 (30% stoi-
chiometric) after 24 or 26 h (see footnotes in Figs. 1-SM and
2-SM, respectively), when the starting mixture practically is
no more reactive, the reaction promptly restarts; 5) on the
contrary, addition of further rac-1 after 24 h does not mod-
ify the reactivity of the system (see footnote in Fig. 2-SM).
This evidence point out that BiBr3 behaves as a true cata-
lyst towards alcoholic and ethereal reagents, but the incom-
pleteness of the reaction, even when equimolar BiBr3 was
employed (Figs. 2, 4), suggests progressive poisoning of
the catalyst, probably due to the formation of H2O as a coproduct
of the reaction.
To test this hypothesis, ancillary experiments were per-
formed adding an equimolar amount of H2O to the starting
mixture of Figure 2 after 10 min (Fig. 4-MS) or at the begin-
ning (Fig. 5-MS) of the reaction. Although replicated experi-
ments provided rather scattered results (relative error about
20%), comparison of Figure 2 and Figures 4-SM/5-SM un-
equivocally reveals the progressive decreasing of the general
Other Observations
According to what Boyer et al. found in CCl4,2 benzyl alco-
hol proved less reactive than phenylethanol (1) in promoting
the benzylation of alcohols, being about one order of magni-
tude slower, even increasing the temperature up to 65°C.
Instead, warming the reaction mixtures containing rac-1 lead
to its substantial dehydration, yielding styrene.
In cyclohexane, all the isomeric 1,2-, 1,3- and 1,4-
metylcyclohexanols maintain their configuration at the
alcoholic carbon atom when reacted with rac-1 in the pres-
ence of BiBr3 at room temperature, and the reaction is
about as fast as that of simple secondary aliphatic alcohols
(2-butanol, 2-pentanol, 2-eptanol, and 2-decanol), with reac-
tion yields >90%.
Boyer’s Reaction Mechanism
←
reactivity pattern (namely, the lower extent of: i) 1S→1R
As a consequence of the wrong interpretation of the ste-
reochemical outcome of the reaction, Boyer et al2 pro-
posed an erroneous model for the benzylation of
alcohols, based on the loss of the stereochemical integrity
of the substrate and the consumption of the catalyst,4
which is in contrast to our findings. As the simplest con-
ceivable mechanistic interpretation of our experiments,
←
we propose that fast 1S→1R racemization (Fig. 3) is a con-
sequence of the reversible coordination of the acidic bis-
muth atom of BiBr3 to the oxygen atom of 1R and 1S
yielding the corresponding n-adducts IR and IS (left side,
Scheme 2). In fact, it is well know that bismuth(III) is a
soft Lewis acid10–13 able to form coordination complexes
with several class of basic compounds, including oxygen-
ated substances.14–17 Within IR/IS adducts the C-O bond
←
is weakened enough to induce the extended 1S→1R race-
mization of the alcohol (Scheme 2) owing to “benzylic” sta-
bilization of the incipient carbocation. With aliphatic or
alicyclic alcohols, similar IS and IR intermediates (R = alkyl,
cycloakyl) can also be formed, but the lack of a similar sta-
bilizing effect makes racemization not energetically feasi-
ble under the same environmental conditions.
Fig. 4. Reaction between the racemate of 1-phenylethanol (rac-1) in cyclo-
hexane added dropwise to a solution of (ꢀ)-(R)-2-pentanol (2PR) in cyclohex-
ane containing equimolar BiBr3. Full circle: 2PR; empty circle: rac-1
(dropwise added in about 15 minutes); full triangle: 3 (3RR:3SR ~ 1:1); empty
triangle: 4. Inset: (4RR + 4SS)/4RS
.
At room temperature, development of the positive
charge at the “alcoholic” carbon atom of IR/IS adducts
(Scheme 2, R = phenyl) is not so extended to promote de-
hydration to the corresponding alkene (e.g., to obtain sty-
rene from 1-phenylethanol, warming the solution up to
refluxing is needed), but enough to allow the nucleophilic
attack by any alcoholic substrate leading to the corre-
sponding ether (right side, Scheme 2; e.g., 3 from alcohol
2, and 4 from alcohol 1) with retention of the configura-
tion of the nucleophilic moiety and elimination of a mole-
cule of water containing the oxygen atom of IR/IS.
Several authors stated that the coordination number of
Bi(III) complexes range from three to nine,18–20 allowing
ligand-to-ligand intermolecular interactions around the co-
ordination sphere of Bi(III). In these optics, following the
hypothesis of IR/IS intermediates, preferential formation
of ethers 4 over 3 ones at the beginning of the reaction
(Fig. 2) could be the consequence of the greater local
concentration of the aromatic alcohol 1 around adducts
IR/IS due to the capability of bismuth to form π-complexes
Fig. 5. BiBr3 induced transetherification of diasteromeric di-1-phenylethyl
ethers in cyclohexane with an equimolar mixture of 1-pentanol (1P) and 2-
pentanol (rac-2P). Full circle: 4; empty circle: rac-2P; full triangle: 1P;
empty triangle: 3; square: 5.
Chirality DOI 10.1002/chir

NEWS FROM BOYER’S REACTION
273
Scheme 2. Mechanism and stereochemistry of the etherification of alcohols under Boyer’s conditions.
with the aromatic ring of 1-phenylethanol.4 Of course, sim-
ilar π-interactions are precluded in aliphatic and alicyclic
alcohols.
The stereochemistry of the ethereal products does not
allow a more intimate description of the nucleophilic step
of the reaction (frontside vs. backside nucleophilic attack
to IR/IS). Anyway, once formed, ethers 4 undergo a par-
←
←
tial epimerization (4RR→4RS→4SS). The simplest mecha-
nistic interpretation of this process (left side, Scheme 3)
is that the acidic bismuth atom of BiBr3 coordinates the
oxygen atom of 4 forming n-adduct intermediates (IIXY
(X,Y = R and/or S)). The n-interaction weakens the C-O-
Scheme 3. Mechanism and stereochemistry of the transetherification reaction under Boyer’s conditions.
Chirality DOI 10.1002/chir

274
FRASCHETTI ET AL.
C bonds of IIXY allowing inversion of configuration of both
the X and Y benzylic moieties. Note that also “asymmetric”
basic solvent, ranks in the middle. In all these solvents BiBr3
is not soluble, and all our experimental findings converge to-
ward BiBr3 acting as a true heterogeneous catalyst, which is
not consumed during the reaction, but undergoes a poisoning
process to a little extent (reaction yield >90%), probably in-
duced by coproduced water, owing to the formation of the
n-complex H2O—BiBr3. Accordingly, a general lower reactiv-
ity is observed when equimolar H2O is added to the reaction
mixture. Similar Lewis acid–base n-interactions could explain
the lack of reactivity in basic media. We propose that, in cyclo-
hexane, n-adducts (IR/IS) between the bismuth atom of BiBr3
and the oxygen atom of alcoholic substrates play a key role in
both the catalytic efficiency and the stereochemical outcome
of the reaction. We guess that, within IR/IS adducts, weaken-
ing of the C-O bond of “activated” alcoholic substrates (able to
stabilize cationic α-carbon atom; e.g., 1) allows both racemi-
zation of the alcoholic moiety in chiral IR/IS and nucleophilic
attack by another alcoholic substrate to form the correspond-
ing ether. Self-reaction of alcohol 1 leads to isomeric ethers
ethers 3 epimerize, inverting exclusively the configuration
←
at its benzylic moiety (e.g., via IRS→ SS
I
in Scheme 3), but
←
the process (e.g., 3RR→
3
SR) is only barely detectable by
GC–MS analysis, because epimerization does not change
very much the 3RR:3SR = 1:1 distribution. In this view, in-
termediates IIXY play a key role in the mechanism of the
transetherification reaction (right side, Scheme 3; next
section), which is very similar to that shown in Scheme
2, with the only difference being that the leaving group
is a molecule of alcohol instead of H2O.
Transetherification Reaction
With the aim to investigate the transetherification reaction
to a deeper extent, a mixture of ethers 4 (specifically synthe-
sized, purified, and dissolved in cyclohexane) was added
dropwise to a solution containing BiBr3 and an equimolar
mixture of 1-pentanol (1P) and the racemate of 2-pentanol
(rac-2P). The results (Fig. 5) confirm the view, pointed out
before, about the reaction of ethers 4 slower than that of
the alcoholic substrate 1 towards the same rac-2P (cf. Figs. 5
vs 4). Moreover, the primary alcohol 1P is about four times
more reactive than the secondary analog rac-2P for the
transetherification reaction under the same conditions
(cf. product 3 vs. 5 in Fig. 5). Both of these results re-
flect the steric demand of intermediates I and II (Schemes 2
and 3) when undergoing nucleophilic attack by alcoholic
reactants.
Another transetherification experiment was performed by
dropwise addition of a mixture of ethers 3 (specifically syn-
thesized, purified, and dissolved in cyclohexane) to a suspen-
sion of BiBr3 in cyclohexane containing an equimolar mixture
of 1-pentanol (1P) and the racemate of 2-hexanol. After more
than 24 h, only a minor amount of 1-phenylethyl 1-pentyl
ethers were formed, according to: 1) “benzylic” moiety of
ethers 3 susceptible to nucleophilic reagents, in contrast to
the “alkylic” 2-pentyl one; 2) primary 1-pentanol more effective
than secondary 2-hexanol in promoting the transetherification
reaction.
←
←
4, which partially epimerize (4RR→4RS→4SS) while undergo-
ing a parallel transetherification process (e.g., 4 → 3 in the
presence of 2-pentanol). Thus, it is reasonable that the “acti-
vated” benzylic carbon atoms of alcohol 1 and its symmetric
ether 4 behave mechanistically similarly, the former being
more reactive than the latter owing to steric factors within
the corresponding n-complexes I and II, respectively.
To date, there are only a few examples of transetherification21–24
reactions in the literature, despite that development of new strat-
egies for the synthesis of ethers under mild conditions is highly
desirable. However, in the last decade interest for this reaction
has grown, mainly for its potential industrial applications.25–27
To this end, work is in progress to investigate the efficiency of
other bismuth(III)-based compounds in the transetherification
reaction. Finally, the reactivity of tert-butyl methyl ether in the
transetherification reaction also promises that other tertiary
alcoholic and/or ethereal substrates (not only benzylic ones)
can react under Boyer’s conditions, which could be very useful
in the production of high-value sugars from poor materials such
as natural polysaccharides.
Further similar transetherification experiments using tert-
butyl methyl ether gave more than an appreciable amount
of tert-butyl 2-pentyl ether (about 10%, from rac-2P), and
tert-butyl 1-pentyl ether (about 30%, from 1P), showing the
aptitude, under Boyer’s conditions, at breaking the C-O bond
of ethers bearing tertiary α-carbon atoms. In fact, under
the same conditions, di-phenyl ether, allyl phenyl ether,
and di-octyl ether proved completely unreactive.
ACKNOWLEDGMENTS
This work was supported by the Ministero dell’Istruzione
dell’Università e della Ricerca (MIUR-PRIN 2010-2011: CUP
B81J1200283001, Prot. 2010ERFKXL_006. Annito Di Marzio
is gratefully acknowledged for technical assistance.
Transetherification reaction, besides its growing interest
for synthetic purposes,21–24 deserves a deeper future investi-
gation due its potential application in the production of added
value chemicals from cheap, abundant, natural polysaccha-
rides, e.g., cellulose,25 lignin,25–27 especially if mediated by a
“green” metal like bismuth as a catalyst.
SUPPORTING INFORMATION
Additional supporting information may be found in the
online version of this article at the publisher’s web-site.
LITERATURE CITED
1. Fraschetti C, Speranza M, Filippi A. Does a chiral alcohol really racemize
when its OH group is protected with Boyer’s reaction? Chirality
2010;22:88–91.
2. Keramane E, Boyer B, Roque J, Pavia A. BiBr3 an efficient catalyst for the
benzylation of alcohols: 2-phenyl-2-propyl, a new benzyl-type protecting
group. Tetrahedron Lett 2000;41:2891–2894.
CONCLUSION
For Boyer’s reaction, a useful process yielding ethers di-
rectly from alcohols under mild conditions, cyclohexane
proved a solvent as good as CCl4, in terms of reactivity and
product yields, but much more convenient for both laboratory
and industrial applications. Basic solvents, such as acetoni-
trile and ethyl acetate do not work at all, while CH2Cl2, a less
Chirality DOI 10.1002/chir
3. Keramane E, Boyer B, Roque J. Reactivity of bismuth(III) halides towards
alcohols.
A
tentative to mechanistic investigation. Tetrahedron
2001;57:1909–1916.

NEWS FROM BOYER’S REACTION
275
4. Keramane E, Boyer B, Roque J. BiBr3-catalyzed benzylation of alcohols.
17. Steward CA, Calabrese JC, Arduengo AJ. Synthesis and structure of the
first 20-Bi-9 system: a discrete nine-coordinate 20-electron bismuth. J
Am Chem Soc 1985;107:3397–3398.
Stereochemistry
and
mechanistic
investigations.
Tetrahedron
2001;57:1917–1921.
5. Leonard N, Wieland C, Mohan S. Applications of bismuth(III) compounds
in organic synthesis. Tetrahedron 2002;58:8373–8397.
6. David S, Hanessian S. Regioselective manipulation of hydroxyl groups via
18. Briand GG, Burford N. Bismuth compounds and preparations with biolog-
ical or medicinal relevance. Chem Rev 1999;99:2601–2658 (and references
cited therein).
organotin derivatives. Tetrahedron 1985;41:643–663.
7. Wessel HP, Iversen T, Bundle DR. Acid-catalysed benzylation and
19. Lange KCH, Klapotke TM. In: Patai S, editor. The chemistry of organic ar-
senic, antimony and bismuth compounds. Wiley, Academic Press: New
York; 1994. p315–366.
allylation by alkyl trichloroacetimidates.
J Chem Soc Perkin Trans
1985;1:2247–2250.
20. Genge ARJ, Levason W, Reid G. Bismuth(III) thioether chemistry: the
synthesis and structure of [Bi4Cl12(MeSCH2CH2CH2SMe)4]n·nH2O, a
highly unusual network involving Bi4Cl4 rings and bridging dithioether
ligands. Chem Commun 1998;2159–2160.
21. Salehi P, Irandoost M, Seddighi B, Behbahani FK. Tahmasebi DP.
Transetherification of allylic and benzylic ethers in the presence of
ferric ion Synth Commun 2000;30:1743–1747.
22. Cheun Y, Koag MC, Kou Y, Warnken Z, Lee S. Transetherification-medi-
ated E-ring opening and stereoselective “Red-Ox” modification of
furostan. Steroids 2012;77:276–281.
23. Li X, Liu B, Xu X, Chmielewski PJ. DDQ-Supported Alkoxylation of 2-Aza-
21-carbaporphyrin and Noncatalyzed Transetherification of Its 3,21-
Dialkoxy Derivatives. J Org Chem 2012;77:8206–8219.
24. Yang L, Seshan K, Li Y. Transetherification of guaiacol to O-
ethoxyphenol with gamma Al2O3 as a catalyst in supercritical ethanol.
Catal Commun 2013;30:36–39.
25. Karagoz S, Bhaskar T, Muto A, Sakata Y. Comparative studies of oil
compositions produced from sawdust, rice husk, lignin and cellulose
by hydrothermal treatment. Fuel 2005;84:875–884.
26. Yuan Z, Cheng S, Leitch M, Xu C. Hydrolytic degradation of alkaline
lignin in hot-compressed water and ethanol. Biores Technol
2010;1019308–9313.
27. Gosselink, RJA. Lignin as a renewable aromatic resource for the chemical
industry. PhD Thesis, Wageningen University, Wageningen, NL (2011)
ISBN_: 978–94-6173-100-5.
8. Liotta LJ, Ganem B. Selective benzylation of alcohols and amines under
mild conditions. Tetrahedron Lett 1989;30:4759–4762.
9. Poon KWC, Dudley GB. Mix-and-heat benzylation of alcohols using a
bench-stable pyridinium salt. J Org Chem 2006;71:3923–3927.
10. Pearson RG. Hard and soft acids and bases.
J Am Chem Soc
1963;85:3533–3539.
11. Pearson RG. Recent advances in the concept of hard and soft acids and
bases. J Chem Educ 1987;64:561–567.
12. Repichet S, Leroux C, Hernandez P, Dubac J, Desmurs Jr. Bismuth(III)
trifluoromethanesulfonate: an efficient catalyst for the sulfonylation of
arenes. J Org Chem 1999;64:6479–6482 (and references cited therein).
13. Sanderson J, Bayse CA. The Lewis acidity of bismuth(III) halides: a DFT
analysis. Tetrahedron 2008;64:7685–7689.
14. Ogawa T, Yoshikawa A, Wada H, Ogawa C, Ono N, Suzuki H.
Tris(2-methoxyphenyl)bismuthane as dehydrating agent with high tem-
plate ability: an efficient single step synthesis of macrocycle diesters from
diacid anhydrides and glycols.
1995;1407–1408: (and references cited therein).
J
Chem Soc Chem Commun
15. Kallal K, Coin C, Dunach E, Postel M. Bismuth(III) carboxylates as cata-
lysts in the oxidative cleavage of epoxides and α-hydroxyketones. A tenta-
tive route to α, ω-dicarboxylic acids. New J Chem 1997;21:495–499.
16. Habata Y, Fujishiro F, Akabori S. Syntheses of crown ether-esters with
(benzyloxy)methyl side arms using SbPh3 and BiPh3 as templates. J
Chem Soc Perkin Trans 1996;1:953–957.
Chirality DOI 10.1002/chir