Inhibition of the Cobalt Acetate/Bromide-Catalyzed Hydrogen Peroxide Oxidation of 4-tert-Butyltoluene

Article abstract of DOI:10.1021/op0340614

The hydrogen peroxide oxidation of 4-tert-butyltoluene to 4-tert-butylbenzaldhyde, an important fragrance intermediate, catalyzed by cobalt acetate and bromide in acetic acid was investigated again. The initial stages of the reaction appear to be rapid and quantitative, but after approximately 25-30% conversion, the process ceases. Overoxidation to 4-tert-butylbenzoic acid does not occur. It appears that both water and the aldehyde product itself inhibit further oxidation. An engineering solution to remove the product continuously seems required for process optimization.

Full text of DOI:10.1021/op0340614

Organic Process Research & Development 2003, 7, 879−882
Inhibition of the Cobalt Acetate/Bromide-Catalyzed Hydrogen Peroxide
Oxidation of 4-tert-Butyltoluene
Ahmed A. Amin and James K. Beattie*
School of Chemistry, UniVersity of Sydney, Sydney NSW 2006, Australia
Abstract:
to benzaldehydes catalyzed by cobalt acetate and bromide
ion. In their hands, sodium perborate monohydrate was more
7
The hydrogen peroxide oxidation of 4-tert-butyltoluene to 4-tert-
butylbenzaldhyde, an important fragrance intermediate, cata-
lyzed by cobalt acetate and bromide in acetic acid was
investigated again. The initial stages of the reaction appear to
be rapid and quantitative, but after approximately 25-30%
conversion, the process ceases. Overoxidation to 4-tert-butyl-
benzoic acid does not occur. It appears that both water and
the aldehyde product itself inhibit further oxidation. An
engineering solution to remove the product continuously seems
required for process optimization.
effective than aqueous hydrogen peroxide, and benzyl
bromide was suggested as an important intermediate. There
was little overoxidation to benzoic acids under the mild
conditions used at 45 °C.
8
Further to our interest in cobalt acetate complexes, we
have investigated the hydrogen peroxide oxidation of 4-tert-
butyltoluene with the ultimate goal of identifying the active
catalytic species. Here, we report observations on the
stoichiometry and rate of the reaction.
Experimental Section
Introduction
Materials. Chemicals were used as received from the
following suppliers: 4-tert-butyltoluene, 4-tert-butylbenzal-
dehyde, 4-tert-butylbenzoic acid, 4-tert-butylbenzyl bromide,
and manganese(II) acetate tetrahydrate (Fluka); 4-tert-
butylbenzyl alcohol, 4-tert-butylphenol, benzophenone, co-
balt(II) acetate, cobalt(III) acetate tetrahydrate, and sodium
perborate monohydrate (Aldrich); hydrogen peroxide (29-
The partial oxidation of 4-tert-butyltoluene to 4-tert-
butylbenzaldehyde is of interest for the use of the aldehyde
as an intermediate in the production of 4-tert-butyl-R-
methyldihydrocinnamaldehyde, trade names Lilestral (BBA)
or Lilial (Givaudan), an important fragrance compound used
in large quantities in soap and cosmetic perfumes.
1
Various methods have been investigated. Gerber et al.
3
2% w/w), glacial acetic acid, and acetic anhydride (AJAX
applied the cobalt(II) acetate-bromide-catalysed air oxida-
tion system that has been intensively examined and applied
for the autoxidation of alkyl aromatics to aromatic carboxylic
Chemicals); and sodium bromide (May and Baker).
Hydrogen peroxide was standardized prior to use with a
standard potassium permanganate solution supplied by
Merck.
2
acids, especially p-xylene to terephthalic acid. By using
lower temperatures than those applied for the oxidation to
carboxylic acids, they were able to achieve a selectivity to
the benzaldehyde of 70% with conversions of 50% of the
substrate. Naturally enough, however, there was overoxida-
tion to the benzoic acid which increased as the aldehyde
accumulated.
Reaction Procedures. In a typical experiment, a 100 mL
round-bottom flask was equipped with a water cooling
condenser, water bath, thermometer, nitrogen inlet and outlet,
overhead stirrer, glass stirrer rod, and poly(tetrafluoroethyl-
ene) paddle. It was charged with glacial acetic acid (50 g),
sodium bromide (4.8 mmol), and cobalt(II) acetate tetrahy-
drate (2.0 mmol). The system was then purged with nitrogen
at ambient temperature for 15 min. Substrate (32.9 mmol)
was added to the mixture. The dark blue solution was
warmed to 45 °C, and the peroxygen source was then added
over 0, 15, 30, 60, or 120 min. Aqueous hydrogen peroxide
Electrochemical methods have been employed. Bosma et
3
al. used direct oxidation on graphite electrodes in methanol
to obtain high yields of the dimethyl acetal, which could be
hydrolyzed to the aldehyde product. Bejan and co-workers
have described electrochemical assisted air oxidation which
4,5
increases the rate of the autoxidation reaction. The end
product of the oxidation is still the benzoic acid, but the
aldehyde concentration can be maximized.6
(50, 75, 100, or 200 mmol) was added by a peristaltic pump,
while sodium perborate monohydrate was added through the
nitrogen inlet in four aliquots. In some experiments, the
reaction was left to proceed at 45 °C for a further 2 h after
the final addition of peroxide. The solutions were subse-
quently cooled to ambient temperature and analyzed by
HPLC.
In the direct antecedent to the present work, Jones et al.
described the use of peroxide oxidants for selective oxidation
(
(
(
1) Gerber, T. I. A.; Wiechers, A.; Young, R. S.; Zeelie, B. S. Afr. J. Chem.
1
997, 50, 82.
2) Partenheimer, W. In Catalysis of Organic Reactions; Scaros, M. G., Prunier,
M. L., Eds.; Marcel Dekker: New York, 1995; Vol. 62, p 307.
3) Bosma, C.; Gouws, S.; Loyson, P.; Zeelie, B. S. Afr. J. Chem. 1999, 52,
In various experiments, the temperature, time of reaction,
cobalt(II) acetate tetrahydrate and sodium bromide catalyst
1
33.
(
4) Bejan, D.; Lozar, J.; Falgayrac, G.; Savall, A. Catal. Today 1999, 48, 363.
(7) Jones, C. W.; Hackett, A.; Pattinson, I.; Johnstone, A.; Wilson, S. L. J.
Chem. Res. (M) 1996, 2501.
(5) Trevin, S.; Savall, A.; Bejan, D. PCT Intl. Appl. France, WO 02/33151,
2
002.
(8) Beattie, J. K.; Hambley, T. W.; Klepetko, J. A.; Masters, A. F.; Turner, P.
Polyhedron 1998, 17, 1343.
(
6) Lozar, J.; Bejan, D.; Savall, A. J. Appl. Electrochem. 2002, 32, 839.
1
0.1021/op0340614 CCC: $25.00 © 2003 American Chemical Society
Vol. 7, No. 6, 2003 / Organic Process Research & Development
•
879
Published on Web 10/15/2003

significance that the unaccounted material disappears when
the reaction becomes inhibited, but no unidentified peaks in
the HPLC traces were observed at this point.
The minor oxidation products 4-tert-butylbenzyl bromide,
4
-tert-butylbenzyl alcohol and 4-tert-butylphenol were all
observed at less than 1 mmol each. These products never
appeared in more than these small quantities.
The yield of the over-oxidation product 4-tert-butylben-
zoic acid also remained small at less than 2 mmol under
these and most of the other conditions investigated. It only
2 2
increased to about 5 mmol when a large excess of H O was
added slowly over an hour. It also increased to 4 mmol when
the temperature was raised from 45 to 60 °C, without any
increase in the yield of aldehyde, so 45 °C was used as the
temperature for the remaining experiments.
The most striking features of these results is the rapid
but incomplete formation of the 4-tert-butylbenzaldehyde
product. The stoichiometry of the reaction requires 2 mol of
hydrogen peroxide to produce 1 mol of aldehyde from the
toluene substrate. Under the conditions of this experiment,
the 50 mmol of peroxide should produce 25 mmol of
aldehyde, leaving 8 mmol of unoxidized toluene. But only
Figure 1. Mass-time data for the oxidation of 33 mmol of
-tert-butyltolune by the addition of 50 mmol of H over 60
min: (×) mass balance, (9) 4-tert-butyltoluene, (b)4-tert-
butylbenzaldehyde, (2) 4-tert-butylbenzoic acid, (1)4-tert-bu-
tylphenol, ([)4-tert-butylbenzyl bromide, and (+)4-tert-butyl-
benzyl alcohol.
4
2 2
O
concentrations, and oxidant type and concentration were
changed or one of the reagents was omitted. In other
experiments, 4-tert-butyltoluene was replaced by other
substrates or was mixed with expected products. In a few
experiments, the solvent was replaced or mixed with acetic
anhydride or water.
Analytical Methods. The HPLC instrument comprised
the Waters 600 controller, Rheodyne 7725 rotary injection
valve, and Waters 486 tunable absorbance detector operating
at 210 nm. The data were collected through a Waters
Millennium HPLC. Serva Octadecyl Si ) 100 or Altech
Altima C18-086 columns were used. The flow rate was 1.0
mL/min. The mobile phase A was water with 0.05% (v/v; 4
mM) trifluoroacetic acid; the mobile phase B was acetonitrile
with 0.05% (v/v; 4 mM) trifluoroacetic acid. The solvent
program with the Octadecyl Si ) 100 column was started at
A:B (50:50) for 8 min and then stepped to A:B (40:60) held
for 12 min and then to A:B (50:50) for another 10 min. The
solvent program with the Altech Altima C18-086 column
was started at A:B (50:50) for 10 min, then stepped to A:B
9
mmol of aldehyde are formed, and this almost all appears
in the first 30 min when only 25 mmol of peroxide had been
added. Indeed, in the first 15 min, the yield of aldehyde was
virtually quantitative, with 12 mmol of peroxide producing
almost 6 mmol of aldehyde.
The effect of the rate of addition of hydrogen peroxide
on the yield of aldehyde was examined. If the 50 mmol of
peroxide were added all at once at the start of the reaction,
the yield was only about 3 mmol. But if they were added
gradually over 15, 30, 60, or 120 min, just 7-9 mmol of
aldehyde were produced, most in the early stages as described
above. It appears that, after the initial almost quantitative
reaction, the process is inhibited and the remaining peroxide
is ineffective for the oxidation of the toluene. Neither,
however, does it oxidize the aldehyde further to the benzoic
acid.
(
20:80), held for 20 min, and then stepped to A:B (50:50)
Increasing the amount of hydrogen peroxide to 75, 100,
or 200 mmol did not improve the yield of aldehyde, and at
the higher concentrations there was considerable loss of
material to unidentified products, with poor mass balances.
Catalysts. Both bromide ion and a source of cobalt are
required for the catalytic oxidation. In the absence of
bromide, no products were observed. With 1 mmol sodium
bromide, about 5 mmol of aldehyde were produced, while,
with 2-6 mmol of bromide, the maximum yield of 8-9
mmol was found, decreasing to 7 mmol of aldehyde with 8
mmol of added bromide. Hence, the turnover number of the
bromide as cocatalyst is no more than 5.
for another 15 min. Benzophenone was used as the internal
standard.
Product identification was by the use of authentic
compounds and retention times. Mixtures of all the expected
products and the substrate in different concentrations were
injected to the HPLC system, and standard curves were
obtained. The unconverted substrate and the yield of all
products were measured by comparing the peak area of each
compound with the peak area of the standards.
Results
General. Figure 1 illustrates many of the principal
features of the results. It shows the mass balance, remaining
substrate, and products observed when 50 mmol of hydrogen
peroxide were added gradually over 60 min to oxidize 32.9
mmol of 4-tert-butyltoluene and the reaction was then
allowed to continue for another 75 min.
The benzylic bromide compound 4-tert-butylbenzylbro-
mide, however, is not an intermediate in the reaction. When
16.5 mmol were used under the same conditions, in place
of the toluene substrate, less than 1 mmol each of the
aldehyde, acid, and phenol products were found, with 12.5
mmol of the bromide recovered unchanged.
The mass balance is reasonably good, with over 90% of
the starting material accounted for. There may be some
In the absence of cobalt(II) acetate, less than 1 mmol of
aldehyde was produced, together with 2 mmol of 4-tert-
880
•
Vol. 7, No. 6, 2003 / Organic Process Research & Development

Figure 2. Effect of the presence of 75 mmol of acetic anhydride
on the conditions shown in Figure 1.
2
Figure 3. Effect of the presence of 100 mmol of H O on the
conditions shown in Figure 1.
butylbenzyl bromide and a trace of the benzyl alcohol. With
1
mmol of cobalt(II) acetate, the yield of the aldehyde was
about 4 mmol, increasing to 10 mmol with 2 mmol cobalt-
II) and decreasing to 8 mmol with 3 mmol of cobalt. Here
inhibit the reaction completely. This was despite the observa-
tion that the addition of this water converted the blue cobalt-
(
(II) species to a pink one. The rate profile was altered, with
again, the turnover number with respect to cobalt is no greater
than 5.
formation of the aldehyde product again occurring more
slowly over the full 60 min of peroxide addition (Figure 3).
Hence, it appears that the oxidation is not completely
inhibited by water but that it converts the cobalt to a less
active form.
Jones et al. made an indirect test of this hypothesis by
comparing the reactivity of sodium perborate monohydrate
and sodium perborate tetrahydrate on the conversion and
aldehyde yield of this oxidation. The monohydrate gave
higher conversion (38.6 to 26.5%) and higher selectivity to
the aldehyde (75.5 to 66.7%), consistent with its lower water
content. They also reported that the perborates gave consid-
erably higher selectivity to the aldehyde than they found with
The cobalt can be supplied as cobalt(III). Crude “cobalt-
(III) acetate” prepared by ozone oxidation was equally
effective as the equivalent amount of cobalt(II).
Manganese(II) acetate is not a cocatalyst, unlike its effect
in cobalt acetate/bromide autoxidation reactions conducted
at higher temperatures. Addition of 2 mmol of manganese-
(
(
II) acetate completely inhibited the reaction, and manganese-
II) acetate in place of cobalt(II) also produced no products.
Search for Inhibitors. Water appears to be an inhibitor
of the oxidation reaction. A solution of cobalt(II) acetate in
acetic acid is pink, indicative of octahedral cobalt(II).
Addition of bromide ion causes the solution to turn blue,
indicative of tetrahedral cobalt(II). As the reaction approaches
the inhibition point, it turns pink again.
This is not due to a loss of bromide ion, for the effect is
independent of the bromide concentration over the range 2-6
mM.
3
5% hydrogen peroxide (75.5 and 66.7 vs 40.5%). We did
not observe this. Under similar conditions the sodium
perborate monohydrate yielded only 6.4 mmol of aldehyde
with a selectivity of 54% compared with hydrogen peroxide
oxidation which yielded 9.4 mmol of aldehyde with a
selectivity of 80%.
Water is a product of the reaction, according to the
stoichiometry
The above observations indicate that the presence of water
is not sufficient to inhibit the reaction completely, so the
search was continued for other inhibitors. 4-tert-Butylphenol
is observed as a minor product and could act as a free radical
trap inhibitor of the oxidation process. But, addition of 5
mmol of the phenol did not completely inhibit the reaction,
which still produced 2 mmol of the benzaldehyde product,
while adding 1 mmol of phenol gradually with the hydrogen
peroxide over 60 min yielded 7.5 mmol of aldehyde. Hence,
the trace amounts of phenol produced in the reaction are
insufficient to cause the inhibition. (The reaction was
completely inhibited, however, in the presence of 1.3 mmol
of TEMPO, a more efficient free radical trap.)
The presence of 10 mmol of the other minor products
4-tert-butylbromide or 4-tert-butylbenzoic acid had no effect
on the formation of the benzaldehyde product. When 10
mmol of the product 4-tert-butylbenzaldhyde were treated
over 120 min in the absence of the toluene, 9.4 mmol were
recovered, confirming the evidence shown in Figure 1 that
the aldehyde is not oxidized further under these conditions.
ArCH + 2H O f ArCHO + 3H O
3
2
2
2
Hence, the formation of the maximum amount of ap-
proximately 10 mmol of aldehyde produces 30 mmol of
water. But in addition as 20 mmol of H
2 2
O are being added,
another 88 mmol of water are introduced with the 30% H
O
2 2
reagent. (The 2 mmol of cobalt(II) acetate also contribute 8
mmol of water as the tetrahydrate.) Thus, approximately 100
mmol of water appear to be sufficient to convert an active
tetrahedral cobalt(II) catalyst into an inactive octahedral form.
Attempts to sequester the water with acetic anhydride
were unsuccessful, as the reaction between water and acetic
anhydride is much slower than the oxidation reaction. The
presence of acetic anhydride, however, did affect the reaction
profile, slowing the initial rate of benzaldehyde product
formation but not affecting the final yield of ∼10 mmol
(Figure 2).
Deliberate addition of 100 mmol of water to the reaction
system reduced the yield of aldehyde to 5 mmol but did not
Vol. 7, No. 6, 2003 / Organic Process Research & Development
•
881

But, when 5 mmol of the aldehyde product were added
to the toluene oxidation system, only 7 mmol of toluene
oxidation occurred; when 10 mmol of aldehyde were added,
only 4 mmol of toluene oxidation were observed, and when
Second, we do not find that aqueous hydrogen peroxide
is inferior to sodium perborate as oxidant. We have no
explanation for this difference between our results and the
earlier report. We do concur with their observations that the
system loses its potency as the color changes from blue to
pink and that further addition of cobalt and bromide is
ineffectual.
20 mmol of aldehyde were added, the oxidation of toluene
to aldehyde ceased, although there was considerable unac-
counted loss of the toluene substrate.
These observations suggest that the 4-tert-butylbenzal-
dehyde product is also an inhibitor of the oxidation of the
toluene. They are consistent with additional observations of
the oxidation of 4-tert-butylalcohol. When 20 mmol of the
alcohol were oxidized, 14 mmol of aldehyde were produced,
only 1 mmol of unreacted alcohol was recovered and 5 mmol
of material were lost. When 10 mmol of the alcohol were
added to the toluene oxidation, only 14 mmol of aldehyde
were found, again suggesting that the rapid formation of the
aldehyde from the alcohol then inhibits further oxidation of
the toluene.
The presence of an inhibitor in addition to water was
tested in the following way. A reaction was run under the
usual conditions. After 1 h, approximately 9 mmol of
aldehyde had been formed and the reaction solution had
turned pink. At this stage, additional aliquots of 2 mmol of
cobalt(II) acetate tetrahydrate and 4.8 mmol of sodium
bromide were added, causing the solution to turn blue again.
Further addition of 50 mmol of hydrogen peroxide over the
next hour, however, yielded no additional aldehyde product.
Whatever the identity of the inhibitors, they appear to act
by the decomposition of the hydrogen peroxide. No excess
peroxide was detected in the reaction solutions when tested
by flow injection analysis for the oxidation of iodide by
peroxide.
1
In a closely related work, Gerber et al. studied the
dioxygen oxidation of tert-butyltoluene catalyzed by the
similar cobalt(II) acetate-bromide system in acetic acid.
There are a number of important differences between these
two systems, however. The dioxygen oxidation requires
higher temperatures, with their experiments conducted
between 55 and 105 °C, with an optimum temperature of
about 90 °C. Second, they observe further oxidation to the
benzoic acid, a common feature of such autoxidation
reactions. Third, the reaction is enhanced, not inhibited, by
the presence of manganese acetate. Finally, they claim to
observe Co(III) in the catalytic system; there is no evidence
of the characteristic green cobalt(III) color in the present
work. They do observe, however, that addition of water,
which diminishes the concentration of tetrahedral cobalt(II),
diminishes its catalytic activity. Hence, there are significant
differences between the hydrogen peroxide and the dioxygen
oxidation mechanisms, although they may share some
common features.
We conclude the following: (a) The oxidation proceeds
by a free radical mechanism, completely inhibited by
TEMPO and by manganese(III) acetate. (b) Over-oxidation
to benzoic acid is minimal. (c) Both cobalt(II) (or cobalt-
(III)) and bromide are required, but p-tert-butylbenzyl
bromide is not an intermediate. (d) The rate of the oxidation
is slowed but not quenched by the accumulation of excess
water. (e) The oxidation is inhibited by accumulation of the
p-tert-butylbenzaldehyde product. (f) Increased yields might
be obtained if the benzaldehyde product was removed as it
were being produced and if less water was added with the
peroxide oxidant.
Discussion
7
Jones et al. appear to be the only other workers to have
examined this system in detail. Our results differ from theirs
in two important respects.
First, we do not find that 4-tert-butylbromide is an
important intermediate. Their assignment of this role to the
bromide is speculative and not based on any substantial
evidence. There is an essential role for the bromide ion: the
catalysis does not occur without it. It does not involve the
Received for review May 26, 2003.
OP0340614
4-tert-butylbromide, however, for adding that to the system
does not enhance product formation.
882
•
Vol. 7, No. 6, 2003 / Organic Process Research & Development