Hydrodehalogenation of halogenated aryl ketones under multiphase conditions. 6. pH effect on the chemoselectivity and preliminary mechanistic investigation

Article abstract of DOI:10.1021/jo982337d

A model halogenated aryl ketone (4-chloropropiophenone 1) was subjected to mild catalytic hydrodehalogenation (p[H₂] = 1 arm, T = 50°C) in a multiphase system consisting of isooctane and an alkaline aqueous phase, in the presence of onium salts as phase-transfer agents. The chemoselectivity of the reaction was sharply influenced by the KOH concentration of the aqueous phase: thus, the formation of 1-phenyl-l-propanol 4 was favored in a highly alkaline pH range (>13.5 by 50% to 5% KOH(aq)). Instead, up to 95% selectivity toward phenyl ethyl ketone 2 was observed if the pH was controlled (8 < pH < 13) by adjusting the KOH(aq) concentration in the range between 5 and 1%. Further investigation of the pH effect indicated that an acidic aqueous phase directed the reaction chemoselectivity toward the formation of products arising from aromatic ring reduction (i.e., cyclohexyl ethyl ketone 7 and 1-cyclohexyl-1-propanol 8) rather than from carbonyl hydrogenation or hydrogenolysis. Also aliphatic amines, used in place of the PT onium salt, afforded good selectivities toward 2, although not as high as the ones obtained with Aliquat 336.

Full text of DOI:10.1021/jo982337d

3
934
J . Org. Chem. 1999, 64, 3934-3939
Hyd r od eh a logen a tion of Ha logen a ted Ar yl Keton es u n d er
Mu ltip h a se Con d ition s. 6. p H Effect on th e Ch em oselectivity a n d
P r elim in a r y Mech a n istic In vestiga tion
Alvise Perosa, Maurizio Selva, and Pietro Tundo*,†
Dipartimento di Scienze Ambientali, Universit a` Ca’ Foscari, Dorsoduro 2137, 30123 Venezia, Italy
Received November 30, 1998
A model halogenated aryl ketone (4-chloropropiophenone 1) was subjected to mild catalytic
hydrodehalogenation (p[H ] ) 1 atm, T ) 50 °C) in a multiphase system consisting of isooctane
2
and an alkaline aqueous phase, in the presence of onium salts as phase-transfer agents. The
chemoselectivity of the reaction was sharply influenced by the KOH concentration of the aqueous
phase: thus, the formation of 1-phenyl-1-propanol 4 was favored in a highly alkaline pH range
(>13.5 by 50% to 5% KOHaq). Instead, up to 95% selectivity toward phenyl ethyl ketone 2 was
observed if the pH was controlled (8 < pH < 13) by adjusting the KOHaq concentration in the
range between 5 and 1%. Further investigation of the pH effect indicated that an acidic aqueous
phase directed the reaction chemoselectivity toward the formation of products arising from aromatic
ring reduction (i.e., cyclohexyl ethyl ketone 7 and 1-cyclohexyl-1-propanol 8) rather than from
carbonyl hydrogenation or hydrogenolysis. Also aliphatic amines, used in place of the PT onium
salt, afforded good selectivities toward 2, although not as high as the ones obtained with Aliquat
3
36.
In tr od u ction
We previously reported that by operating in a multi-
phase system composed of an alkaline aqueous phase and
a hydrocarbon solvent such as isooctane, and in the
presence of onium salts and of a phase-transfer (PT)
agent, halogenated aromatics undergo facile and selective
Pd-catalyzed hydrodehalogenation to afford the corre-
and a wide range of reduction products was observed in
the absence of the onium salt. Moreover, we also showed
that an accurate control of the ratio PT agent/platinum
was necessary in order to achieve the optimal compro-
mise between selectivity and reaction time. A similar
behavior, although with slower rates, was observed also
for other hydrophilic PT agents such as triethylbenzy-
lammonium chloride (TEBA) and a polyethyleneglycol
sponding dehalogenated aromatics, under mild conditions
3 2 2 2
such as CH (CH )11O(CH CH O)23H (Brij 35). Accord-
-3
(
p[H
2
] ) 1 atm, T ) 50 °C).1 Even sterically hindered
ingly, both the chemical structure and the properties of
the PT agent were suggested to play a role in determining
the outcome of the hydrodehalogenation reaction.
Herein we describe further results in the study of the
hydrodehalogenation reaction: particularly, concerning
the effect on the chemoselectivity observed by varying
the pH of the aqueous phase. Attempts to rationalize our
observations by proposing a possible mode of action of
the PT agent will also be discussed.
aryl halides could be rapidly reduced to the corresponding
2
aromatic hydrocarbons. When polyfunctionalized sub-
strates such as halogenated aryl ketones were subjected
to those conditions, we observed that the presence of the
PT agent induced very high chemoselectivity: hydro-
dehalogenation took place without any further reduction
4
of the carbonyl or of the aromatic moiety. For example,
the reaction of p-chloropropiophenone carried out over
Pd/C in the presence of Aliquat 336 (tricaprylmethyl-
ammonium chloride, A336) afforded selectively pro-
piophenone when the onium salt was used.
Resu lts a n d Discu ssion
More recently we reported that also by using Pt/C as
a catalyst, the chemoselectivity of the multiphase hydro-
dehalogenation of haloaromatics was sharply modified
by the presence of a PT agent. Thus, while using A336,
the hydrodehalogenation of haloaromatic ketones could
yield selectively the corresponding benzylic alcohols
1
. Mu ltip h a se System . To study the effect of the
KOHaq concentration on the multiphase hydrodehalogen-
ation reaction, a series of reactions was conducted by
choosing 4-chloropropiophenone 1 as a model substrate.
Experiments were run at 50 °C: a mixture of 7 mL of an
isooctane solution of the substrate (0.07 M), a KOHaq
solution (4 mL) of varying concentrations (from 5 to 0.5%
in weight), the 5% Pt/C catalyst (Cl/Pt ratio of 32.6), and
the PT agent (see Tables 1 and 2 for the amounts) was
5
(without reduction of the hydroxyl or aromatic moiety),
†
Phone: +39 041 257 8642. Fax: +39 041 257 8620. e-mail:
tundop@unive.it.
magnetically stirred while H
2
was bubbled at atmo-
(1) Marques, C. A.; Selva, M.; Tundo, P. J . Org. Chem. 1993, 58,
5
3
1
2
256.
spheric pressure (approximately 1 mL/min) into the
reaction vessel. The choice of the reaction conditions and
of the PT agents was dictated by the previous results that
gave the best selectivity toward phenyl propanol 4.
The results are reported in Table 1, where the reaction
(2) Marques, C. A.; Selva, M.; Tundo, P. J . Org. Chem. 1994, 59,
430.
(3) Marques, C. A.; Selva, M.; Tundo, P. J . Mol. Catal. A: Chemical
5
995, 96, 301.
(
4) Marques, C. A.; Selva, M.; Tundo, P. J . Org. Chem. 1995, 60,
430.
5) Selva, M.; Tundo, P.; Perosa, A. J . Org. Chem. 1998, 63, 3266.
(
times refer to when maximum selectivity toward 4 and
1
0.1021/jo982337d CCC: $18.00 © 1999 American Chemical Society
Published on Web 05/07/1999

Hydrodehalogenation of Halogenated Aryl Ketones
J . Org. Chem., Vol. 64, No. 11, 1999 3935
Ta ble 1. Hyd r od eh a logen a tion of
4
-Ch lor op r op iop h en on e 1 u n d er Mu ltip h a se Con d ition s
a
(
Isoocta n e/KOHa q), Usin g 5% P t/C
products (% by GC)
time
run
aq phase
PT agent
(min)
2
3
4
5
6
7
8
5
1
2
KOHaq 50% A336 53 mg
KOHaq 50% TEBA 18 mg 720
60
6
5
94
95
85
80
3
5
3
4
5
6
7
KOHaq 5%
KOHaq 5%
KOHaq 1%
KOHaq 1%
KOHaq 1%
A336 72 mg
TEBA 18 mg 360 15
A336 72 mg
TEBA 18 mg
none
360 12
1
2
2
1
1
2
3
17 95
20 89
17 51
10
31
6
11
a
Conditions: T ) 50 °C; Cl/Pt ) 32.6; Pt/C (5%) 0.0586 g; 7
mL of 0.07 M 1 in isooctane; 4 mL of alkaline aqueous phase. Data
refer to complete substrate conversion.
Ta ble 2. Hyd r od eh a logen a tion of
4
-Ch lor op r op iop h en on e 1 u n d er Mu ltip h a se Con d ition s
in th e P r esen ce of a Bu ffer ed Aqu eou s P h a sea
F igu r e 1. Multiphase hydrodechlorination of 4-chloropropio-
phenone in the presence of 5% KOH (entry 3, Table 1).
products (% by GC)
run
buffer pH
time (h)
2
4
5
7
8
7 + 8
Sch em e 1
1
2
3
4
7.00
5.00
2.00
1.00
7
7
6
6
9
14
0
46
35
24
8
9
17
18
24
33
26
27
42
49
9
18
16
0
a
Conditions: T ) 50 °C; Cl/Pt ) 32.6; Pt/C (5%) 0.0586 g; 7
mL of 0.07 M 1 in isooctane; aqueous phase buffered at pH 7.00,
5
.00, 2.00, 1.00. Data refer to complete substrate conversion.
2
is reached, at complete substrate conversion. We report
only the data referring to the amounts of PT agent found
to afford the highest selectivity, which were determined
experimentally by running a series of reactions. All data
were verified by conducting each reaction at least three
negative effects of the HCl formed during the hydro-
5
times. As was previously reported, by using 50% KOHaq
,
6-8
dehalogenation reaction
but also allowed for control
the selectivity toward 4 was high (94% for A336, entry
; and 95% for TEBA, entry 2) and p-chlorophenylpro-
of the reaction selectivity simply by accelerating the rate
of path “a” (hydrodehalogenation of the aryl ring) with
respect to “b” (carbonyl hydrogenation), and not by
modifying the reaction pathway.
A further and more convincing evidence for this
behavior was gathered by conducting the hydrodehalo-
genation of 1 with 1% KOHaq. Under such conditions, the
reaction could be selectively and easily directed toward
the dehalogenated aryl ketone 2. At complete substrate
conversion (after 17 min) we observed 95% and 89% of
1
panol 3 was the byproduct (6% for A336 and 5% for
TEBA). By lowering the KOHaq concentration from 50%
to 5%, the hydrodehalogenation reaction of 1 still afforded
the dehalogenated aryl alcohol, although with lower
selectivity compared to the preceding case: after 6 h,
using A336 and TEBA, the yield of 4 was 85% and 80%,
respectively (entries 3 and 4). Interestingly, the dehalo-
genated aryl ketone 2 was now the main other product
(12% and 15% with A336 and TEBA, respectively), and
2
, using A336 (entry 5) and TEBA (entry 6), respectively,
no p-chlorophenylpropanol 3 was observed.
the remaining product being 4 (3% and 10%, respec-
tively). In this case, by merely diminishing the concen-
tration of the base, chlorine removal took place without
any further reduction process. This fact indicated that
the rate of conversion of 1 into 2 was fast relative to the
successive carbonyl hydrogenation to afford 4. Figure 2
shows the progress of the reaction with 1% KOHaq
described in entry 5 of Table 1. By carefully monitoring
the reaction, the formation of 4 could be prevented, and
with 1% KOHaq the unwanted byproducts were kept
below 3%.
Figure 1 shows the progress of the hydrodehalogena-
tion reaction of 1 in the presence of 5% KOHaq, under
the conditions described in entry 3 of Table 1.
The fact that ketone 2, and not p-chlorophenylpropanol
3
, was observed upon lowering the KOHaq concentration
from 50% to 5% confirmed our previously reported
5
hypothesis of two distinct pathways for the reaction of
1
to 4 (Scheme 1).
In the case of KOHaq 50% the rates of the two parallel
reactions (path “a” and path “b”) were competitive, and
the key to reduce unwanted byproducts was to favor path
As we previously noted,5 it is a synergistic effect
between the presence of the base and of the onium salt
which accounts for the chemoselective outcome of the
multiphase hydrodechlorination. In fact, when the multi-
“a” by adjusting the platinum/PT agent ratio and the
reaction time. Instead, in the case of 5% KOHaq, path “a”
was faster than path “b” and the reaction afforded 2
without concomitant formation of 3; the ketone 2 then
further reacted to give 4 selectively. These results
indicated that the selectivity toward dehalogenation
could be controlled by a combined effect of the Pt/PT
agent ratio and of the concentration of KOH. In other
words, the presence of the base not only minimized the
(6) Hoke, J . B.; Gramiccioni, G. A.; Balko, E. N. Appl. Catal. B: Env.
992, 1, 285.
1
(
(
7) Pasett, P.; Gentile, F. J . Org. Chem. 1995, 60, 1351.
8) Armandia, M. A.; Borau, V.; J imenez, C.; Marinas, J . M.;
Sempere, M. E.; Urbano, P. Appl. Catal. 1988, 43, 47.

3
936 J . Org. Chem., Vol. 64, No. 11, 1999
Perosa et al.
The final pH was chosen as the reference parameter since
it provides a measure for the effect on selectivity of the
KOHaq concentration. All data refer to complete substrate
conversion. To allow a faster interpretation of the results,
the different base concentrations are shown in paren-
theses. The observed trend reaches a maximum yield of
2
(95%) in correspondence with a KOHaq concentration
of 1%. As can be seen, if the KOHaq concentration was
2
1
% or higher, the final pH of the reaction remained >
3.5, implying that the HCl developed by the reaction
was neutralized by the base, which remained in excess.
By using 1% KOHaq the final pH of the aqueous phase
was still alkaline; however during the reaction it de-
creased from 13.2 to 9.4 pH units, at complete conversion
after 17 min, and the selectivity toward 2 reached 95%.
From this value onward of KOHaq concentration, a
further decrease of the base amount caused the final pH
of the aqueous phase to turn acidic. Thus, using KOHaq
F igu r e 2. Multiphase hydrodechlorination of 4-chloropropio-
phenone in the presence of 1% KOH (entry 5, Table 1).
0
.75%, 0.50%, and 0.25%, a pH of 2.26, 1.44, and 1.17,
respectively, was measured after complete dechlorina-
tion, and the selectivity toward 2 dropped to 60%, 38%,
and 25%, respectively.
Hence, the rate of formation of the products was
strictly dependent on the alkalinity of the aqueous
phase: this allowed us to tune the chemoselectivity of
the multiphase catalytic hydrodehalogenation reaction.
At low base concentrations (1% e KOHaq e 2%), the rate
of formation of the dehalogenated ketone 2 was faster
than the successive carbonyl reduction, while higher
concentrations of KOH promoted the rapid formation of
the dehalogenated alcohol 4. If one refers to path “a” of
Scheme 1, when KOHaq was used in the range of 1 to
2%, the first step 1 f 2 was faster than the second one
F igu r e 3. Multiphase hydrodechlorination of 4-chloropropio-
phenone: selectivity toward 2 at different KOH concentrations
and at complete substrate conversion.
(2 f 4); while when using a higher KOHaq concentration
50% to 5%) the second step 2 f 4 competed with the
(
first step 1 f 2. Therefore, an increase of the KOHaq
concentration caused a major enhancement of the rate
of carbonyl hydrogenation over the rate of aryl dehalo-
genation.
phase hydrodehalogenation of 4-chloropropiophenone was
carried out in the presence of 1% KOH, but without the
PT agent, the chemoselectivity of the reaction was
sharply lowered: after 17 min conversion was complete,
but we observed only 51% of 2, along with 31% of 4, 11%
of propylbenzene 6, and 7% of aliphatic products (entry
To obtain a good control of chemoselectivity, the
reaction environment had to remain alkaline until after
complete substrate conversion. This fact indicated that
the lower limit of the amount of KOHaq necessary to
achieve good selectivities was approximately the mini-
mum concentration required to ensure the complete
neutralization of the HCl formed in the hydrodehalogen-
ation reaction, in our case 1%. It was also in agreement
with the observation, reported by others,8 that the
presence of a base inhibits carbonyl hydrogenolysis of
acetophenone, in favor of its reduction to the correspond-
ing alcohol.
7
, Table 1). On the other hand, we already reported that
in the absence of the base, only a range of aliphatic
products was observed, both with and without a PT
agent.5
2
. p H Effect. These facts allowed us to conclude that
the decrease of the KOHaq concentration from 50% to 1%
had a dramatic effect on the reaction selectivity: in
particular, at KOHaq ) 1% and by tuning the amount of
the PT agent, the halo ketone 1 could be hydro-
dechlorinated with up to 95% selectivity without concur-
rent reduction of the carbonyl function.
To investigate more indepth the effect of the base
concentration on selectivity, a series of reactions was then
conducted by further progressively lowering the KOHaq
concentration from 1% to 0.25% and by measuring the
pH of the aqueous phase once the substrate was com-
pletely converted. Then the hydrodehalogenation of 1 was
also conducted in the absence of KOH in the aqueous
phase, and an acidic final pH of 1.0 was measured,
consistent with HCl produced by the reaction being
quantitatively absorbed in the aqueous phase. The
results are shown in Figure 3, where the maximum
selectivity toward 2 (% yield by GC, after the indicated
time) is plotted against the final pH of the aqueous phase.
Further evidence for the alkalinity-dependent chemo-
selectivity trend in the hydrodehalogenation of 1 derived
from our previous observation that, when using a neutral
aqueous phase in the presence of A336 (which caused
the final pH of the reaction mixture to become acidic: pH
) 1.0), the selectivity toward 2 and 4 decreased, interest-
ingly in favor of products arising from the reduction of
the aromatic ring: thus, cyclohexyl ethyl ketone 7 (21%)
and cyclohexylpropanol 8 (44%) were obtained, which did
5
not react further. This preferential reduction of the aryl
ring seemed interesting also from a synthetic standpoint.
It was only observed when A336 was used in the absence
of KOHaq; instead, mainly propylcyclohexane 5 was
obtained by removing the PT agent. It should be noted
that products 7 and 8 appeared rapidly during the

Hydrodehalogenation of Halogenated Aryl Ketones
J . Org. Chem., Vol. 64, No. 11, 1999 3937
Ta ble 3. Hyd r od eh a logen a tion of
4
-Ch lor op r op iop h en on e 1 u n d er Mu ltip h a se Con d ition s
a
in th e P r esen ce of Differ en t Am in es
products (% by GC)
run
amine
Et2NH
Et3N
n-Bu3N
(PhCH2)3N
n-C8H17NH2
time (min)
2
4
others
1
2
3
4
5
55
45
105
60
83
83
82
83
82
15
15
16
16
15
2
2
2
1
3
55
a
Conditions: T ) 50 °C; Pt/C 5% 0.010 g; amine used in 0.2 M
ration with respect to 1; 0.07 M solution of 1 in isooctane (7 mL);
aqueous phase KOH 1% (4 mL). Data refer to complete substrate
conversion.
F igu r e 4. Multiphase hydrodechlorination of 4-chloropropio-
phenone: selectivity toward 2 at different KOH concentrations
using triethylamine as modifier of the catalyst surface.
reaction, indicating that the observed chemoselectivity
depended specifically on the alkalinity of the aqueous
phase and was therefore not determined simply by
different reaction rates.
This fact, and the effect of the pH on selectivity
discussed earlier, prompted us to investigate the influ-
ence of acidic pH’s on the outcome of the hydrodehalo-
genation reaction of 1 in the presence of A336. Under
phase-transfer catalysis conditions, the catalytic reduc-
tion of aromatic ketones to yield alicyclic ketones and
lower than usual) was used. In this way, the reaction was
slowed and complete consumption of the substrate re-
quired 45-105 min (entries 1-5); in the presence of 1%
KOHaq, the use of different amines allowed the dehalo-
genation to occur with selectivities of 80-83% toward the
ketone 3. Although the reaction was not as chemoselec-
tive as in the case of the onium salts, it followed a trend
similar to the one observed using A336. This behavior
was further confirmed by running the hydrodechlorina-
tion reaction in the presence of triethylamine, with
varying KOHaq concentrations. The results are plotted
in Figure 4. Similar to what was observed previously
alcohols was claimed in the presence of [1,5-HDRhCl]
2
as a homogeneous catalyst and of a buffered aqueous
9
solution. Therefore, a series of reactions was conducted,
(compare Figure 3), the reaction outcome was markedly
with an aqueous phase buffered at pH ) 1.00, 2.00, 5.00,
and 7.00; the experiments were run under the conditions
similar to those used for the reaction performed with pure
dependent upon the pH of the aqueous phase: in par-
ticular, the highest selectivity toward the formation of 3
was obtained with 1% KOHaq. Instead, when the concen-
tration was either reduced or increased the selectivity
dropped, and ketone hydrogenation took over. While, the
lipophilicity of the amine appeared to have no appreciable
effect on directing the chemoselectivity of the reaction:
both hydrophilic diethylamine and hydrophobic triben-
zylamine afforded almost identical yields of 3 and 4 after
water (unbuffered solution) where a combined yield of
5
6
5% was reached for 7 + 8. The results are collected in
Table 2.
As expected, at the pH’s tested, the ketone reduction
became competitive with aromatic ring reduction. Even
though the trend confirmed that lower final pH’s (pH <
1
) favored ring reduction products (compare entries 1-4
5
5 min of reaction (82-83% and 15%: Table 3, entries 1
of Table 2), a selectivity as high as that with pure water
65% combined), toward cyclohexyl ketone 7 and cyclo-
and 5, respectively). Therefore, it seems that the concen-
tration and the interaction of the PT agent with the
supported metal catalyst, rather than its partitioning (or
lipophilicity) between the two phases, are the key factors
which determined the observed reaction pathway.
(
hexyl alcohol 8, was never reached. In no case was 5
observed. It was therefore concluded that, when using
pure water, the gradual decrease of pH during the
progress of the reaction promoted reduction of the aryl
moiety, while protecting the carbonyl function. Therefore,
in this case also, the effect of pH on chemoselectivity
appears to be important.
4
. Mech a n ism of Action of th e P T Agen t. As was
previously mentioned, before addition of the PT agent to
the aqueous-organic system, the carbon-supported metal
catalyst was uniformly distributed as a suspension in
both phases. Instead, once the PT agent was added, the
Pt (or Pd)/C catalyst appeared to become coated by it,
and to settle at the phase boundary at rest; upon stirring
it was completely transferred into the organic phase.
The interactions between the carboxyl and carbonyl
3
. Effect of th e P T Agen t. While onium salts, even
of aliphatic amines, are preferentially soluble in the
aqueous phase, a favorable partitioning of aliphatic
amines into the organic phase should, on the other hand,
be expected under multiphase conditions. Therefore, if
an amine was used, in place of an onium salt, a modifica-
tion of the catalyst surface present in the organic phase
should be expected as well, and a consequent effect on
the chemoselectivity of the reaction should be observable.
To this aim, amines with different solubility properties
in the aqueous and organic phases were chosen: diethyl-,
triethyl-, tributyl-, tribenzyl-, and octylamine were tested
in a 0.2 M ratio with respect to 4-chloropropiophenone
10
groups present on the active carbon surface and the PT
agent (onium salt) formed a lipophilic film on the Pt/C
catalyst, thereby favoring its partitioning into the organic
phase where the reaction takes place, and probably
generating a suitable microenvironment for the reaction.
Under basic conditions, the first step in Scheme 2 takes
place in the aqueous phase and gives the formation of
carboxylate anions on the surface of the carbon support
1
. Table 3 reports the results. Since under the usual
+
of the Pt catalyst. In the second step, exchange of K with
conditions conversion of 1 was complete after 17 min, in
order to follow with higher accuracy any variations of the
reaction outcome, a lower mol % of Pt catalyst (5 times
+
the cation (Q ) of the onium salt (or with the ammonium
+
nitrogen, HNR
3
, in the case of the amine) makes the
catalyst surface lipophilic. The supported catalyst is then
(9) J anuszklewicz, K. R.; Alper, H. Organometallics 1983, 2, 1055.
(10) Swiatkowskj, A.; Wieczorek, M. Chem. Anal. 1984, 25, 565.

3
938 J . Org. Chem., Vol. 64, No. 11, 1999
Perosa et al.
F igu r e 5. Pictorial representation of the hypothetical arrangement of substrate 1 on the catalyst surface in the absence of (a)
+
and in the presence of (b) the onium salt Q .
Sch em e 2
“
extracted”11 into the organic phase, thereby favoring
the aqueous phase, provided that a suitable onium salt
or aliphatic amine was present. An acidic aqueous phase
directed chemoselectivity toward the formation of prod-
ucts arising from aromatic ring reduction 7 and 8 rather
than ketone hydrogenolysis, but a buffered acidic phase
did not improve the previously reported yields.
-
+
formation of the -COO Q species (Scheme 2). The same
shift does not occur under neutral or acidic conditions.
+
The large lipophilic onium cation Q modifies the
catalyst because, by completely coating the carbon sur-
face, it finds itself near the active Pt sites.
This change might directly affect the selectivity of the
hydrodehalogenation reaction of the substrate in two
ways: (a) by an interaction of Q with the incoming
In addition to preventing catalyst poisoning by trans-
ferring the HCl to the alkaline aqueous phase where it
is neutralized, the PT agent appeared to coat the particles
of supported metal catalyst, thereby transferring it to the
organic phase where the reaction took place. The polar
PT agent film formed on the Pt/C catalyst, thus providing
a reaction microenvironment for the substrate, with
functional group-specific catalytic sites.
+
substrate, i.e. by modifying its “solvation shell” near the
catalyst and favoring approach and reduction of the
haloaromatic portion of the substrate (Figure 5a,b); (b)
by an effect on the coordination sphere of Pt due to the
+
proximity of Q which changes its catalytic activity.
Therefore, the modified catalyst surface may interact
with the substrate in a way that provides preferential
reduction of one functional group respect to another (e.g.
hydrodehalogenation vs CO reduction).
The multiphase system therefore offers a wide versa-
tility since it allows to tune selectivity by changing simple
parameters such as KOHaq concentration, PT agent, and
reaction conditions. It thus offers the opportunity for new
synthetic applications, thanks also to the facile separa-
tion of the products and to the small amounts of reagents
required.
In this regard, it should be pointed out that the
selectivity effect, which is obtainable by adding the PT
agent, is very large relative to the amount which is used,
making its efficiency very high. By forcing the reaction
to take place in this microenvironment, made up by the
supported metal catalyst and the PT agent, one is
effectively steering it precisely to the reactive site.
Further investigations are underway to take advantage
of these features for synthetic applications, e.g. to employ
the hydrodehalogenation of haloaromatic ketones for the
Con clu sion s
Under the here reported multiphase conditions, the
outcome of the Pt-catalyzed hydrodehalogenation of
4
1
-chloropropiophenone 1 could be directed to afford either
-phenyl-1-propanol 4 or phenyl ethyl ketone 2 selec-
tively, simply by appropriately tuning the alkalinity of
(11) In analogy to the proposed mechanism for extraction of car-
boxylic acids into an organic phase by a PT catalyst in the presence of
-
OH . Cf. Dehmlow, E. V.; Dehmlow, S. S. Phase Transfer Catalysis.
In Monographs in Modern Chemistry, 2nd rev. ed.; Ebel, H. F., Ed.;
Weinheim; Deerfield Beach, FL; Basel: Verlag Chemie, 1983; pp 32-
4
1 and refs 34 and 35 therein.

Hydrodehalogenation of Halogenated Aryl Ketones
J . Org. Chem., Vol. 64, No. 11, 1999 3939
formation of products arising from aromatic ring reduc-
tion (compare 7 and 8) of substituted haloaromatics.
mmol), 4.0 mL of aqueous solution (alkaline by KOH or at
controlled pH by using a buffer solution12), 0.0586 g of 5% Pt/C
(0.015 mmol of Pt), and 0.0725 g of PT agent (0.20 mmol).
Hydrogen was bubbled at atmospheric pressure through the
Exp er im en ta l Section
organic phase at a rate of about 1 mL/min. The reaction flask
was thermostated at 50 ( 0.1 °C, and the mixture was
magnetically stirred at 1000 rpm with a bar 2.5 cm in length
and 0.6 cm in diameter.
The reaction course was followed by gas chromatography.
Conversions are referenced to the internal standard (n-decane).
All reagents and solvents were ACS grade and were used
without further purification. The 5% Pt/C was from Fluka, Art.
no. 80982; its surface area was 700-800 m /g. GC analyses
were performed on a Varian GC 3400 using a fused silica
capillary column (30 m × 0.25 mm) with SPB-5 as the liquid
phase (film thickness 0.25 mm). GC/MS analyses were per-
formed on a HP 5971 mass detector coupled to a HP 5890 gas
chromatograph fitted with a 30 m × 0.25 DB5 capillary
column.
2
Ack n ow led gm en t. This work was supported by
INCA (Interuniversity Consortium Chemistry for the
Environment) and MURST (Ministero dell’Universita`
e Ricerca Scientifica e Tecnologica), Fondo 40%.
The structure of all compounds obtained (2-8) was con-
firmed by mass spectroscopy.
Gen er a l P r oced u r e for th e Hyd r od eh a logen a tion Re-
a ction s. All reactions were carried out in a 25 mL three-
necked flask equipped with a condenser and a system for the
bubbling of hydrogen.
If not otherwise indicated, they were carried out on 7.0 mL
of an isooctane solution of the 4-chloropropiophenone 1 (0.49
J O982337D
(
12) Buffer solutions were prepared as follows: pH ) 1.00: 0.13 M
HCl/0.050 M KCl; pH ) 2.00: 0.03 M citric acid/0.0082 M HCl/0.061
M NaCl; pH ) 5.00: 0.096 M citric acid/0.20 M NaOH.; pH ) 7.00:
2 4
0.029 M NaOH/0.050 M KH PO .