Uncatalyzed Meerwein-Ponndorf-Oppenauer-Verley Reduction of Aldehydes and Ketones under Supercritical Conditions

Article abstract of DOI:10.1021/jo035251f

When a solution of a carbonyl compound in alcohol (primary or secondary) is heated to ca. 300 °C, a disproportionation reaction, in which a carbonyl compound is reduced to the corresponding alcohol and the alcohol is oxidized to the corresponding ketone, takes place. This uncatalyzed variation of the Meerwein-Ponndorf-Oppenauer-Verley reaction gives, in certain cases, e.g., reduction of acetophenone or benzaldehyde by i-PrOH, almost quantitative yields. Yields are higher with secondary alcohols such as i-PrOH than with a primary alcohol such as EtOH. The reactions were also performed in a flow system by passing at a slow rate the same solutions through a glass or a metal coil heated to elevated temperatures. Ab initio calculations performed at the B3LYP/6-31G* level show that thermodynamically i-PrOH is a more potent reducing agent than EtOH by ca. 4 kcal/mol. The computations also show that in cases of aromatic carbonyl compounds, part of the deriving force is obtained from the entropy change of the reaction. The major contributor to the high yield, however, is the excess alcohol used, which shifts the equilibrium to the right. Calculated entropy of activation as well as isotopic H/D labeling suggest a cyclic transition state.

Full text of DOI:10.1021/jo035251f

Un ca ta lyzed Meer w ein -P on n d or f-Op p en a u er -Ver ley Red u ction
of Ald eh yd es a n d Keton es u n d er Su p er cr itica l Con d ition s
Lena Sominsky, Esther Rozental, Hugo Gottlieb, Aharon Gedanken,* and Shmaryahu Hoz*
Department of Chemistry, Bar-Ilan University, Ramat-Gan, Israel 52900
shoz@mail.biu.ac.il
Received August 27, 2003
When a solution of a carbonyl compound in alcohol (primary or secondary) is heated to ca. 300 °C,
a disproportionation reaction, in which a carbonyl compound is reduced to the corresponding alcohol
and the alcohol is oxidized to the corresponding ketone, takes place. This uncatalyzed variation of
the Meerwein-Ponndorf-Oppenauer-Verley reaction gives, in certain cases, e.g., reduction of
acetophenone or benzaldehyde by i-PrOH, almost quantitative yields. Yields are higher with
secondary alcohols such as i-PrOH than with a primary alcohol such as EtOH. The reactions were
also performed in a flow system by passing at a slow rate the same solutions through a glass or a
metal coil heated to elevated temperatures. Ab initio calculations performed at the B3LYP/6-31G*
level show that thermodynamically i-PrOH is a more potent reducing agent than EtOH by ca. 4
kcal/mol. The computations also show that in cases of aromatic carbonyl compounds, part of the
deriving force is obtained from the entropy change of the reaction. The major contributor to the
high yield, however, is the excess alcohol used, which shifts the equilibrium to the right. Calculated
entropy of activation as well as isotopic H/D labeling suggest a cyclic transition state.
In tr od u ction
Despite its seniority in chemistry, a literature survey
of recent years shows that this reaction does not cease
to attract the attention of the chemical community. New
uses and improvements are continually suggested,5-8 as
well as new catalysts such as borinic acid derivatives,9
magnesium phosphates,10 Ir, Rh complexes,11,12 and
chromium derivatives.13
The disproportionation reaction between carbonyls and
alcohols involving hydrogen transfer between the two (eq
1) was discovered about 70 years ago by Meerwein,
Ponndorf, and Verley in the context of the reduction of
the carbonyl compound (aldehyde or ketone) to the
corresponding alcohol.1-3 About 10 years later the revers-
In the course of our studies of reactions under super-
critical conditions we have discovered that mixing a
carbonyl compound (aldehyde or ketone) with alcohol at
elevated temperatures induces the disproportionation
typical of the Meerwein-Ponndorf-Oppenauer-Verley
reaction. We report here the results of a study regarding
the scope and limitation of this variation of the H transfer
reaction with a brief examination of the mechanistic
options. In light of the world’s need for industrial green
chemistry, we have also explored the feasibility of a
continuous flow method. However, no attempt was made
to optimize the method with respect to time, temperature,
ible nature of the reaction was realized by Oppenauer,
who illuminated the oxidation capability of the reaction.4
This reaction is usually catalyzed by trivalent aluminum
in an assumed cyclic transition state (eq 2).
(5) Fujita, M.; Takarada, Y.; Sugimura, T.; Tai, A. Chem. Commun.
1997, 17, 1631-1632.
(6) Akamanchi, K. G.; Chaudhari, B. A. Tetrahedron Lett. 1977, 38,
6925-6928.
(7) Campbell, E. J .; Zhou, H.; Nguyen,S. T. Org. Lett. 2001, 3, 2391-
2393.
(8) Ooi, T.; Otsuka, H.; Miura, T.; Ichikawa, H.; Maruoka, K. Org.
Lett. 2002, 4, 2669-2672.
(9) Ishihara, K.; Kurihara, H.; Yamamoto, H. J . Org. Chem. 1997,
62, 5664-5665.
(10) Aramendia, M. A.; Borau, V.; J imenez, C.; Marinas, J . M.;
Romero, F. J . Catal. Lett. 1999, 58, 53-58.
(11) Ajjou, A. N. Tetrahedron Lett. 2001, 42, 13-15.
(12) Suzuki, T.; Morita, K.; Tsuchida, M.; Hiroi, K. J . Org. Chem.
2003, 68, 1601-1602.
(1) Meerwein, H.; Schmidt, R. J ustus Liebigs Ann. Chem. 1925, 39,
221.
(2) Pondorf, W. Angew. Chem. 1926, 39, 138.
(3) Verley, M. Bull. Soc. Chim. Fr. 1925, 37, 871.
(4) Oppenauer, R. V. Recl. Trav. Chim. 1937, 56, 137.
(13) Schrekker, H, S.; de Bolster, M. W. G.; Orru, R. V. A.;
Wessjohann, L. A. J . Org. Chem. 2002, 67, 1975.
10.1021/jo035251f CCC: $27.50 © 2004 American Chemical Society
Published on Web 02/06/2004
1492
J . Org. Chem. 2004, 69, 1492-1496

Uncatalyzed Meerwein-Ponndorf-Oppenauer-Verley Reduction
TABLE 1. Yield s of Cor r esp on d in g Alcoh ols in th e
Red u ction of Va r iou s Ca r bon yl Com p ou n d s by i-P r OH
a n d EtOH u n d er Sta n d a r d Con d ition s
TABLE 3. Yield s in Red u ction of m -Meth oxy
Ben za ld eh yd e by i-P r OH a s a F u n ction of Rea ction
Tem p er a tu r e (Sta n d a r d Con d ition s Excep t for
Tem p er a tu r e)
yield (%)
T (°C)
substrate
PhCOMe
m-MeOC6H4CHO
m-PhOC6H4CHO
EtOH
i-PrOH
240
32
260
61
280
89
300
99
14
52
67
93
85
>98
yield (%)
the reducing agent. On the other hand, whereas cyclo-
hexanone benefited only marginally from the addition of
the base in EtOH, the yield in i-PrOH jumped from 50%
to 100% as a result of added base.
TABLE 2. Effect of Ad d ed Ba se on th e Red u ction Yield
u n d er th e Sta n d a r d Con d ition s
ROH
NaOH
yield (%)
PhCOMe
Because of the ambiguity of the results it was sus-
pected that catalytic sites on the metallic surface of the
reaction container enhance the reaction. We have there-
fore conducted a series of experiments using three
different cells. Two of them were cells fabricated from
Swagelock parts made of either stainless steel or copper
(see Experimental Section). In addition the reaction was
also carried out in a glass ampule. We have found no
differential effect of the surface material on the reaction
yield.
EtOH
EtOH
i-PrOH
i-PrOH
-
+
-
+
14
42
93
78
Cyclohexanone
EtOH
EtOH
i-PrOH
i-PrOH
-
+
-
+
36
41
50
100
Another important factor in this reaction is the tem-
perature. Table 3 displays the effect of the reaction
temperature on the yield of the alcohol in the reduction
of m-methoxy benzaldehyde in the range 240-300 °C
under the standard reaction conditions.
or yield. We were rather interested in exploring the
boundaries and possibilities provided by this new ap-
proach.
Resu lts a n d Discu ssion
Clearly, the reaction is enhanced by the temperature.
Assuming that the reverse reaction, namely, oxidation
of the benzyl alcohol by the acetone, produced in the first
step, is negligible and that the reaction is pseudo-first-
order in the substrate, we can calculate the reaction rate
constants at each temperature and derive from these the
activation energy. Rate constants of 0.0007, 0.0017, and
0.0041 s-1 were obtained for the temperatures 240, 260,
and 280 °C (the value at 300 °C was too close to
equilibrium). These values lead to an activation energy
of 24.5 kcal/mol. These values should be considered as
rough estimates since the reaction time of 1.5 h is
measured from the moment the reaction vessel is placed
in the oven and does not reflect the actual time the
reaction mixtures were incubated at the said tempera-
tures.
Not unexpectedly, the alcohol giving the best results
among the simple aliphatic alcohols was i-PrOH. Given
in Table 1 are the results of a comparative study of the
reduction of three carbonyl substrates by i-PrOH and
EtOH. All of the reactions were performed under the
same conditions, namely, T ) 300 °C, t ) 1.5 h, with
substrate concentration ca. 10 wt % in the appropriate
alcohol (standard conditions).
As can be seen from the results depicted in Table 1,
i-PrOH is by far superior to EtOH. It should be empha-
sized that the high yields obtained with i-PrOH imply
that after 1.5 h, the reaction has essentially reached
complete, or nearly complete, equilibration. The gap
between the effectiveness of the two alcohols is therefore
much larger than the simple yields ratio and would have
been better manifested at shorter reaction times. Of
course, to obtain quantitative comparison, initial rates
or initial yields have to be determined. However, as
stated above, this was not the objective of this scope
study.
Another conclusion that can be derived from the data
in Table 1 is that although in our standard conditions
we have used a reaction time of 1.5 h, for practical
purposes, reaction times should be tailored to the specific
reactant. Thus, for example, acetophenone, which yielded
93% after 1.5 h, yielded 85% of the corresponding alcohol
already after only 0.5 h.
It is common, well-established knowledge that the
Meerwein-Ponndorf-Oppenauer-Verley reaction is cata-
lyzed by base. We have examined the sensitivity of our
reaction to base by adding a small amount of NaOH
(0.025 M in the i-PrOH solution) to the reaction mixture.
Table 2 shows that although the addition of base im-
proved the results in the case of EtOH reducing ac-
etophenone, it slightly lowered the yield with i-PrOH as
Table 4 lists the results for a variety of additional
carbonyl compounds.
All of the above reactions were done in batch. This is
a major drawback of the method for industrial uses.
Hence, we have explored the applicability of this reaction
to a continuous flow system. The system consists of a
syringe pump that pushes the reaction mixture into a
spiral column placed in an oven and then through an
orifice to a receiver where the mixture is collected and
cooled. The results are given in Table 5. The temperature
recorded was measured at the center of the oven, and
the flow rate of the liquid reaction mixture at the exit of
the pump was set to 2 mL/h.
Obviously, the yield in this case depends on the flow
rate. This was demonstrated in the following experi-
ments. 2-Butanone was dissolved in i-PrOH to a concen-
tration of 15 wt %. The temperature at the center of the
oven was maintained at 350 °C. In the first experiment
the flow rate was 2 mL/h, yielding 81% of the starting
material and 19% of 2-BuOH. When the flow rate was
J . Org. Chem, Vol. 69, No. 5, 2004 1493

Sominsky et al.
TABLE 4. Yield s of Cor r esp on d in g Alcoh ols in th e
Red u ction of Va r iou s Ca r bon yl Com p ou n d s by i-P r OH a t
Sta n d a r d Con cen tr a tion a n d Tem p er a tu r e
A clear outcome of the computational data is that
i-PrOH is a better reducing agent than EtOH, probably
stemming from the additional stabilizing Me-CdO
interaction present in acetone in comparison with ac-
etaldehyde. This is reflected in the exothermicity of the
reaction of acetaldehyde with i-PrOH (4.1 kcal/mol) and
is manifested, although to a lower extent, in the reaction
of PhCHO compared to PhCOMe. These two reactions
are endothermic probably because the conjugative inter-
action between the carbonyl group and the ring is lost in
the product. However, the reaction of PhCOMe is more
endothermic (by ca. 2 kcal/mol) as a result of the
additional Me group. The endothermicity of the reaction
suggests that the reaction energy will tilt the equilibrium
constant (8.8 at 573 K) in favor of the reactants. This,
however, can be balanced by a change in entropy that
will render the reaction favoring the products in terms
of free energy. Using the ab initio calculated thermody-
namic data we indeed obtain that at 300 °C the overall
∆G ) -0.6 kcal/mol. This value translates at T ) 573 K
to K ) 1.7. Thus, the entropy factor shifts the equilibrium
to the right and hence there is a small preference in
terms of free energy to the products. The major “push”
in the product direction comes from the mass effect.
Having a large excess of the reducing alcohol over the
substrate is the main cause for having such a high yield
in the reaction.
In the classical mechanism of the Meerwein-Ponndorf-
Oppenauer-Verley reaction, the transition state encom-
passes aluminum alkoxide in which a hydride is trans-
ferred to the carbonyl carbon in a cyclic array. Clearly,
this is only a descriptive presentation of the mechanism
and the degree of ionicity at the transition state is rather
small. The major ionic component is introduced in the
ground state in the Al-O bond. However, the gas phase
tolerates much less ionicity than the condensed phase,
and therefore we should expect either a radical mecha-
nism or a neutral cyclic transition state. The latter could
be envisioned as the traditional transition state in which
an H atom assumes the position of the Al atom.
increased to 14 mL/h, the yield dropped drastically to only
2% of the desired alcohol and 98% of the starting
material.
Mech a n istic Asp ects. Because one of the major
factors determining the yield of these reactions is the
thermodynamic deriving force, we have performed an ab
initio calculation at the B3LYP/ 6-31G* level14 for the
isodesmic reaction15 shown for the case of benzaldehyde
and EtOH in eq 3.
A neutral noncyclic transition state will most probably
involve two consecutive H radical transfer steps. The first
H transfer will be from the central C-H unit of i-PrOH
to the oxygen of the carbonyl group and the second from
the OH of i-PrOH to the carbonyl carbon (eq 4).
PhCHO + EtOH h PhCH2OH + CH3CHO (3)
The results are given in Table 6.
(14) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J . R.; Zakrzewski, V. G.; Montgomery, J . A.,
J r.; Stratmann, R. E.; Burant, J . C.; Dapprich, S.; Millam, J . M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J .;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,
C.; Clifford, S.; Ochterski, J .; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J . B.; Cioslowski, J .; Ortiz, J . V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J .; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; J ohnson, B. G.; Chen,
W.; Wong, M. W.; Andres, J . L.; Head-Gordon, M.; Replogle, E. S.;
Pople, J . A. Gaussian 98, Revision A.4; Gaussian, Inc.: Pittsburgh,
PA, 1998.
This differs from the traditional cyclic transition state
in which the hydrogen is exchanged between two carbons.
A convenient way to resolve the two options is to use
i-PrOD instead of i-PrOH. Since in the cyclic transition
state hydrogen is transferred between two carbon atoms,
such a transition state will produce no deuterium labeling
at the benzylic position of a substrate such as benzalde-
hyde. On the other hand, as was pointed out above, a
sequence of H transfers will produce PhCHDOH from the
same substrate.
(15) An isodesmic equation is one in which there are equal numbers
of each type of bond on each side. Clark, T. A Handbook of Computa-
tional Chemistry; Wiley: New York, 1985; p 97.
1494 J . Org. Chem., Vol. 69, No. 5, 2004

Uncatalyzed Meerwein-Ponndorf-Oppenauer-Verley Reduction
TABLE 5. Resu lts of Red u ction Exp er im en ts Usin g i-P r OH in th e F low System a
a
Yields are given as percents of the starting material.
TABLE 6. Ab In itio (B3LYP /6-31G*) Com p u ted En er gies
for Mod el Com p ou n d s Used in th e Isod esm ic Rea ction
a n d Th eir Rea ction En er gies (∆E) w ith i-P r OH a n d EtOH
Su m m a r y a n d Con clu sion s
We have shown that at elevated temperatures the
normally catalyzed Meerwein-Ponndorf-Oppenauer-
Verley reaction takes place in a noncatalytic fashion. It
may be driven nearly to completion, even in cases of
endothermic reactions by using a large excess of the
alcohol. The reaction is usually clean and does not yield
wastes to be desposed of; the product carbonyl compound
derived from the sacrificial alcohol can be easily sepa-
rated and used in a different reaction. On the basis of
isotopic labeling results we conclude that the transition
state of the rate-determining step is cyclic.
∆E (kcal/mol)
compound
i-PrOH
acetone
EtOH
MeCHO
PhCH2OH
PhCHO
cyclohexanol
cyclohexanone
PhCH(OH)CH3
PhCOCH3
HF (au)
i-PrOH
EtOH
-194.3533105
-193.1556937
-155.0342889
-153.8301224
-346.7705987
-345.5734556
-311.0902461
-309.8912915
-386.089635
-384.8959876
-4.1
0.3
0.0
4.4
3.3
6.6
-0.8
2.5
Heating benzaldehyde with i-PrOD for 1 h under
standard conditions gave 87% of benzyl alcohol where no
D atoms were incorporated into the benzylic position.
This implies that in the reduction the carbonyl carbon
obtains the hydrogen from the central carbon of i-PrOH
as in the classical mechanism and therefore strongly
suggests that the reaction takes place in a single step
via a cyclic transition state rather than through a two
consecutive H transfer steps. The large negative entropy
of activation calculated for the reaction points also to a
highly ordered transition state.
The mechanism of the base-catalyzed reaction has not
been explored and may either take place via a single-
step cyclic transition state taking place at the surface of
the reaction vessel or involve either a hydride transfer
from the alkoxide to the carbonyl carbon (eq 5) or H
radical transfer to the carbonyl oxygen (eq 6). The latter
produces a radical anion of acetone, which may be
involved in a multitude of other elementary processes.
The role of the metal catalyst is, most probably, to
bridge the two reacting parties in the conformation,
which will enable the transfer of hydride. The increased
number of collisions at high temperature and the higher
energy content of the molecules compensates for the
absence of the bridging metal in our case. The alcoholic
proton most likely functions as the metallic catalyst by
forming a hydrogen bond between the alcohol and the
carbonyl function. An indication of such a bonding can
be found in Table 4 where electron-withdrawing substit-
uents on the aromatic nucleus drastically reduce the
yield.
Exp er im en ta l Section
The static reaction was carried out in a 2-mL closed cell,
which was assembled from stainless steel Swagelok parts. A
3/8 in. union part was capped from one side by a standard
plug. The reaction mixture was introduced into the union and
plugged from the other side. The cell was filled with 0.8-0.9
mL of liquid. At the end of the reaction the cell was cooled in
the air. All manipulations with the reactants or products in
the deuterated 2-propanol experiments were conducted inside
a glovebox under nitrogen.
In the flow experiments we have used either copper,
stainless steel, or glass tubings. The stretched length of the
J . Org. Chem, Vol. 69, No. 5, 2004 1495

Sominsky et al.
metallic coils was about 2 m, whereas the glass coil was about
150 cm. The tubings were coiled and placed in an oven. The
diameter of the tubings was 1/4 in.
Su p p or tin g In for m a tion Ava ila ble: Gaussian archive
files for structures appearing in Table 6. This material is
available free of charge vi the Internet at http://pubs.acs.org.
Reaction mixtures were analyzed using a 300-MHz NMR
spectrometer. The accuracy of the integration is better than
5%.
J O035251F
1496 J . Org. Chem., Vol. 69, No. 5, 2004