Aerobic alcohol oxidations mediated by nitric acid

Article abstract of DOI:10.1002/anie.201105620

A touch of acid: Catalytic amounts of HNO₃ can trigger the aerobic oxidation of alcohols in the presence of the solid acid amberlyst-15. The desired oxidation cycle, mediated by (H)NO_x species, is in kinetic competition with the detr

Full text of DOI:10.1002/anie.201105620

DOI: 10.1002/anie.201105620
Shuttling Oxygen
Aerobic Alcohol Oxidations Mediated by Nitric Acid**
Christof Aellig, Christophe Girard, and Ive Hermans*
The selective oxidation of organic substrates plays a pivotal
The final goal would be to develop a process which uses nitric
[
1,2]
role in the chemical value-chain.
Despite the industrial
acid as an oxygen carrier in which O is the true terminal
2
relevance, the scientific understanding of oxyfunctionaliza-
tion technologies lags behind the current state-of-the-art.
Radical-chain autoxidations are a classical example of large-
oxidant. However, the proposed systems suffer from various
shortcomings, such as limited turnover. In addition, HCl leads
to corrosive reaction conditions and poses recycling issues;
highly dispersed carbon-based materials do not seem very
appealing as oxidation catalysts from a stability or even safety
point of view.
[3]
scale technology which is still not completely understood.
Nitric acid based oxidations are another example of poorly
understood processes of tremendous industrial importan-
[
2a,4]
ce.
The production of adipic acid (3 Mt/year)—a building
Herein, the aerobic oxidation of benzyl alcohol
(PhCH OH) to benzaldehyde (PhCHO), by HNO , is inves-
block for nylon-6,6—from a mixture of cyclohexanol/cyclo-
hexanone, the synthesis of glyoxylic acid (80 kt/year)—used
in the synthesis of vanillin and Amoxicillin—as well as the
oxidation of dimethyldisulfide to methanesulfonic acid (30 kt/
year)—used in electroplating and detergent formulations—
are just a few examples from bulk and commodity chemis-
2
3
tigated. This partial oxidation is not only a model reaction of
[
12]
interest to (fine-)chemical synthesis, the selectivity of this
reaction is also very sensitive to radical side-reactions which
easily cause over-oxidation of PhCHO to benzoic acid
[
13]
(PhCOOH).
Selective aerobic alcohol oxidations with
[
1,2a,5]
try.
However, oxidations with nitric acid are also
catalytic amounts of HNO could offer an attractive metal-
3
[
6]
extensively used for the synthesis of fine-chemicals. The
free alternative to precious-metal-catalyzed systems.
To investigate the HNO -based PhCH OH oxidation and
reason why HNO -based oxidations are so valuable is the fact
3
3
2
that nitric acid is an inexpensive oxidant that can achieve
remarkable selectivities. One disadvantage of this technology
is the strongly acidic nature of nitric acid, especially when
applied in high concentration. Another disadvantage is the
stoichiometric reduction of the HNO to NO and N O. In a
to decouple the dual role of HNO —being both an oxidant
3
and a strong acid—sub-stoichiometric amounts of HNO were
3
used in the presence of a solid acid as (co-)catalyst. Screening
of various solid acids, including most conventional zeolites
(e.g., MFI, BEA, FAU, and MOR topologies), revealed the
ion-exchange resin amberlyst-15 in protonic form
3
x
2
[
7]
large-scale process (such as adipic acid synthesis), NO and
x
ꢀ
1
N O are separated, and the NO is recycled into nitric acid.
(4.7 meqg ) as a promising candidate. Actually, zeolites
gave disappointing results, probably due to strong sorption of
HNO and NO species, inhibiting the reaction. Our standard
2
x
This ex situ NO recycling however results in large recycle
x
streams. Moreover, on a smaller scale—such as typically
encountered in fine-chemical industry—it usually does not
3
x
experiments are performed in a bubble column reactor using
pay to recycle the NO . Nitrous oxide, being a severe
25 mL 1,4-dioxane as solvent, [PhCH OH] = 500 mm,
x
2
0
greenhouse gas, cannot be recycled and should be (catalyti-
[HNO ] = 125 mm, 150 mg of amberlyst-15 (corresponding
3
0
[1]
+
cally) decomposed prior to emission. More recently, N O
to [H ]
= 28 mm), and 1 atm of O (bubbling through at
2
amberlyst 2
ꢀ
1
arising from adipic acid production has also been used to
100 mLmin ). An induction period was observed which
could be eliminated by the addition of small amounts of
[
1,8]
oxidize olefins to valuable ketone products.
To date the
formation mechanism of N O in nitric acid based oxidations
NaNO (Figure 1, curves a–c). This observation—known from
2
2
has not been elucidated, making it a difficult to decrease its
formation.
Recently, aerobic oxidation reactions which use sub-
[
9–11]
stoichiometric amounts of HNO have been proposed.
In
3
those systems, strong acids (such as hydrochloric acid), or
carbon-based materials (such as active carbon), are added to
achieve a catalytic turnover in HNO . Although the reaction
3
mechanism is currently unknown, the idea is very attractive.
[
*] C. Aellig, C. Girard, Prof. Dr. I. Hermans
ETH Zurich, Institute for Chemical and Bioengineering
Wolfgang-Pauli-Strasse 10, CH-8093 Zurich (Switzerland)
E-mail: hermans@chem.ethz.ch
Homepage: http://www.hermans.ethz.ch
[**] We acknowledge the financial support from ETH Zurich (grant ETH-
18 09-2).
Figure 1. Effect of the NaNO concentration on the PhCH OH con-
2
2
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201105620.
version at 808C: a–c) at 1 atm O (0, 5, and 15 mm NaNO ,
2
2
respectively); d,e) at 20 atm O (0 and 5 mm NaNO , respectively).
2
2
Angew. Chem. Int. Ed. 2011, 50, 12355 –12360
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12355

Communications
[
12]
the literature —points towards the crucial role of HNO ,
2
ꢀ
which is formed upon protonation of NO2 . Inhibition of the
reaction in the presence of urea—a known HNO trap —is
[
14]
2
in line with this hypothesis. NO and NO can alternatively be
2
used to shorten the induction period (see Supporting Infor-
mation), pointing towards an efficient interconversion of the
[
15]
nitrogen species.
Interesting to note is that under N2
bubbling (1 atm) at 808C, the conversion levels off after 6 h
to 50%, leading to a D[PhCH OH]/[HNO ] ꢁ 2. Under the
2
3 0
same conditions, but in presence of 1 atm O , a conversion of
2
8
0% can be achieved in 6 h (see curve b in Figure 1).
Nevertheless, the initial reaction rate is approximately 40%
lower in the presence of O , compared to N (see Supporting
2
2
Information). These observations indicate that oxygen plays
an ambiguous role in this system. On the one hand O is able
2
to trigger a recycling of the active species and thereby
increase the final conversion, on the other hand it reduces the
Figure 2. PhCH OH conversion versus time for various amounts of
2
reaction rate. At elevated O pressures (e.g., 20 atm in a high-
amberlyst: a–d): 50, 75, 100, and 150 mg (5 mm NaNO ). The insert
shows the dependency of the initial rate on the amberlyst amount.
2
2
pressure autoclave) the induction period is significantly
longer than at atmospheric pressure, even in presence of
initiator (Figure 1, curves e and d). This elongated induction
period is probably caused by a shift of the equilibrated
decomposition of HNO to NO towards the left-hand-side at
[17]
of 1,4-dioxane ), and benzoic acid (Figure 3). Interestingly,
benzyl nitrite (PhCH ONO) can be identified as a reaction
3
2
2
higher O pressure [Reaction (1)]. Indeed, NO is able to
intermediate: its concentration initially increases, then
decreases. The benzyl nitrite is formed in the equilibrated
esterification reaction PhCH OH + HNO QPhCH ONO +
2
2
[
4,16]
form HNO by Reaction (2),
the key oxidizing species in
2
the system, and a shift of the equilibrium in Reaction (1) will
hence reduce the HNO2 concentration. Note that the
2
2
2
H O. Traces of benzyl nitrate, benzyl benzoate, dibenzylether,
2
amberlyst catalyst accelerates the HNO decomposition to
and benzyl formate (formed out of PhCH OH and formic
3
2
[
18]
NO (see Supporting Information), thereby speeding up the
acid, a decomposition product of 1,4-dioxane ) are detected
by GC-MS. The most important by-products arise from the
use of dioxane as solvent and can be avoided when working in
2
formation of active species.
4
2
HNO Ð 4 NO þ 2 H O þ O
ð1Þ
ð2Þ
3
2
2
2
pure PhCH OH. Note that the maximum benzyl nitrite
2
concentration is significantly higher at lower reaction temper-
atures (e.g. 34 mm at 608C versus 21 mm at 808C); when the
reaction is performed with less amberlyst (e.g. 50 mg instead
of 150 mg), the benzyl nitrite concentration remains below
the GC-MS detection limit.
NO þ H O Ð N O þ H O ! HNO þ HNO
2
2
2
4
2
3
2
Gas–liquid mass transfer is found to play an important
role for this reaction: when the reaction is performed under a
static O pressure (in a round-bottom flask with an oxygen
2
balloon, conventional lab practice), the end conversion is
2
0% lower than in a bubble column at the same pressure but
with enhanced gas–liquid mass-transfer.
The reaction is quasi first-order in HNO , although the
3
initial rate is clearly leveling off at increasing [HNO ] (see
3
0
Supporting Information). This observation points towards an
initiating role of nitric acid, rather than a pure stoichiometric
reaction. The reaction rate is proportional to the amberlyst
ꢀ
1
loading, up to 4 mgmL after which it levels off (Figure 2).
Although this could in principle point towards gas–liquid
mass-transfer limiting the reaction, an alternative mechanistic
explanation is proposed below. Fluid–solid mass-transfer
effects can be neglected as 1 mm amberlyst beads have the
same activity as the powdered catalyst (thus, the activity is
independent of the fluid–solid interfacial area). The solid
catalyst can be recycled several times without loss in activity.
Overall, the selectivity towards benzaldehyde (PhCHO) is
very high, ranging from 95% at 808C to 98.5% at 608C. The
most important by-products are benzyl nitrite, 2-benzyl-1,3-
dioxolane (from acetalyzation of benzaldehyde with 1,2-
dihydroxyethane, formed in the acid catalyzed decomposition
Figure 3. Time evolution of the most important by-products at 808C
under standard conditions ([PhCH OH] =500 mm, [NaNO ] =5 mm,
[HNO ] =125 mm, 150 mg of amberlyst and 1 atm O ).
3 0 2
2
0
2 0
Doubling the initial alcohol concentration barely affects
the rate at low amberlyst loadings, whereas first-order
behavior is observed for higher catalyst loadings. Addition
1
2356 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 12355 –12360

of conventional radical trappers, such as 2,4,6-tri-tert-butyl-
phenol, did not significantly affect the reaction rate or
selectivity, excluding the involvement of peroxyl or alkoxyl
radicals as important reaction intermediates.
corresponding Arrhenius activation energy of (12 ꢂ 1) kcal
ꢀ1
mol excludes mass-transfer limitations as an explanation for
the leveling off in reaction rate at higher amberlyst loadings
(see Figure 2). Using only 2 mg amberlyst per mL, the
ꢀ1
Mass spectrometry was used to monitor the evolution of
NO (m/z 30) and N O (m/z 44) in the gas phase. N O was also
activation energy is significantly higher, (33 ꢂ 1) kcalmol .
This observation, together with a change in reaction order in
alcohol (see above), points towards a shift in rate-determin-
ing-step when going from a low-to-high amberlyst loading.
Interestingly, a higher end-conversion can be achieved at
lower temperature.
2
2
monitored with transmission IR spectroscopy of the gas phase
under static O pressure of 1 atm (Figure 4; n vibrational
2
3
To understand the precise role of the benzyl nitrite ester
(
PhCH ONO)—experimentally identified as a reaction inter-
2
mediate (see above)—we synthesized it according to a
[
19]
literature procedure and studied its acid-catalyzed decom-
position (see Supporting Information). A stoichiometric
amount of benzaldehyde is formed, confirming that the
nitrite ester is a pivotal intermediate. The reaction is first-
order in both nitrite and amberlyst. In the temperature range
4
0–708C, the rate constant is well described by the Arrhenius
6
ꢀ1
ꢀ1 ꢀ1
expression (3 ꢂ 2) ꢀ 10 exp(ꢀ(13ꢂ1) kcalmol /RT) m s .
During the nitrite decomposition—performed under N
2
atmosphere—N O formation is detected by IR spectroscopy,
2
identifying this step as the nitrous oxide source. Although the
initial nitrite decomposition rate is not affected by the
Figure 4. Gas-phase transmission IR spectra from a reaction per-
formed at 808C under static O pressure of 1 atm (spectra shown after
2
0
.5, 1, 2, 3, and 5 h; see Supporting Information): n vibrational mode
presence of NO , at higher reaction times, NO2 slightly
2
3
of N O.
enhances the decomposition rate, due to the formation of
HNO and HNO in Reaction (2) which increases the proton
2
3
2
(i.e. the catalyst) concentration. More important is that
mode), unambiguously assigning the m/z signal of 44 to
approximately 20% less N O was found in the gas phase
2
nitrous oxide. No signal for nitric oxide was observed at
when the nitrite decomposition was studied in a dioxane
ꢀ
1
1
876 cm , probably due to the small dipole-moment change
solution saturated with 0.5% NO in an He gas flow, prior to
2
of the NO-stretch. Despite the yellow color of the liquid and
the m/z signal of 46 determined in the liquid phase, no NO2
could be detected in the gas phase, confirming that it is
strongly absorbed in the dioxane solution. The addition of
water to the reaction mixture (in addition to the water already
the addition of amberlyst. Benzyl nitrate (PhCH ONO ),
2
2
although being detected with GC-MS as a trace product, was
found to be unreactive at 808C for several hours and should
hence be considered as a spectator in equilibrium with
PhCH OH plus HNO , rather than a true reaction inter-
2
3
added in the 65% HNO solution) prevents the formation of
mediate.
3
ꢀ1
the yellow color and inhibits the reaction. Presumably, water
traps NO in Reaction (2). This observation suggests NO as
Note that at high amberlyst loadings (i.e., 6 mgmL ), the
ꢀ1
observed activation energy (i.e. (12 ꢂ 1) kcalmol ) corre-
2
2
an important reaction intermediate.
Figure 5 shows the temperature dependence of the overall
oxidation reaction catalyzed by 6 mg amberlyst per mL. The
sponds with the activation energy of the nitrite decomposition
ꢀ1
reaction (i.e. (13 ꢂ 1) kcalmol , see above). This agreement,
together with the observed first-order in alcohol, is an
important indication that this acid-catalyzed decomposition
of
the
nitrite
ester
(formed
by
PhCH OH +
2
HNO QPhCH ONO + H O) is the rate-determining-step
2
2
2
(RDS) in the overall mechanism, at least at high catalyst
loadings. This hypothesis explains why the reaction rate levels
off after the nitrite intermediate reaches its maximum
concentration (i.e. after 1 h, see curve b in Figure 1 and
Figure 2). Under the assumption that the nitrite decomposi-
tion is rate-determining, the rate constant of this decompo-
sition can be estimated from the overall reaction rate (e.g.
ꢀ1
0
.26m h after 0.5 h, see curve b in Figure 1), the correspond-
ing nitrite and proton concentration stemming from amber-
+
lyst (i.e. [nitrite] = 20 mm, see Figure 2, and [H ]
=
amberlyst
ꢀ
1
ꢀ1
2
8 mm) to be (0.13 ꢂ 0.03)m s . This value is in excellent
agreement with the value obtained from our nitrite decom-
Figure 5. Temperature dependency of the PhCH OH conversion in the
2
position study (see Supporting Information), (0.11 ꢂ
presence of 5mm NaNO under standard conditions (a–d: 60, 70, 80,
2
ꢀ
1
ꢀ1
and 908C, respectively).
0.02)m s .
Angew. Chem. Int. Ed. 2011, 50, 12355 –12360
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12357

Communications
ꢀ
1
At low amberlyst loadings (i.e., 2 mgmL ), HNO
2 NO þ O
2
! 2 NO
2
ð5Þ
ð6Þ
3
decomposition (according to the stoichiometry of Reac-
tion (1)) is likely to be the RDS, explaining the zero order
in alcohol. The experimental activation energy in this range
NO þ HNO ! NO þ HNO
3
2
2
ꢀ
1
(
i.e. 33 ꢂ 1 kcalmol ) is significantly lower than the reported
Note that, although the temperature dependence of HNO
consumption by Reactions (3) and (4) are similar, HNO
formation (by nitrite decomposition) is more accelerated at
higher temperature than NO formation by Reaction (5)
2
which actually becomes slower at higher T. This hypothesis
ꢀ
1 [20]
HNO self-decomposition barrier of 38.7 kcalmol , point-
3
ing towards an amberlyst-catalyzed decomposition (see
Supporting Information). This shift in RDS also explains
why the nitrite ester could not be detected when the reaction
is performed with a low amberlyst loading (see above) as its
decomposition is faster than its formation.
All these experimental observations point towards an
acid-catalyzed decomposition of the nitrite ester, involving
the protonation of the N-atom (Scheme 1), leading to the
[22]
is confirmed by the lower nitrite ester concentration at higher
temperatures. Because of this situation, the [HNO]/[NO
concentration ratio will increase at increasing temperature,
favoring N O formation (termination) as seen by IR spec-
troscopy. As an example, at 908C, 40% more N O is found
]
2
2
2
than at 808C, both at 60% alcohol conversion. This explains
why a higher end conversion can be achieved at a lower T (see
Figure 5).
Scheme 2 summarizes the mechanistic insights. It shows
the NO and HNO cycling and emphasizes that O is the
x
x
2
terminal oxidant; HNO only triggers an aerobic oxidation
3
cycle.
Scheme 1. Acid-catalyzed decomposition of PhCH ONO to PhCHO
2
and HNO; PhCH ONO seems unreactive under similar conditions.
2
2
formation of HNO. Because the nitrate ester cannot be
protonated it is unreactive under similar reaction conditions.
HNO is a reactive substance that dimerizes very rapidly to
hyponitrous acid (HON=NOH) which decomposes to water
[
21]
and N O [Reaction (3)].
The overall rate constant for
2
Reaction (3)
can
be
described
by
k3(T) = 6.9 ꢀ
5
ꢀ1
ꢀ1 ꢀ1 [22]
1
0 exp(ꢀ2.8 kcalmol /RT) m s .
In the presence of
NO , HNO can be converted into NO (detected by mass
2
spectrometry) by Reaction (4), regenerating HNO2.
Scheme 2. HNO is initiating a NO /HNO mediated oxidation cycle.
HNO þ HNO ! HON¼NOH ! N O þ H O
ð3Þ
ð4Þ
3
x
x
2
2
HNO þ NO ! NO þ HNO2
2
As reported above, water has a detrimental effect on the
Reaction (4) is driven by the weak HꢀNO bond of only
reaction as it traps the NO in Reaction (2) and hence favors
2
ꢀ
1
8
ꢀ1
4
9 kcalmol , leading to a k (T) = 6 ꢀ 10 exp(ꢀ2 kcalmol /
the formation of N O. We suspect this effect limits the TON in
4
2
ꢀ1
ꢀ1 [23]
RT)m s .
Kinetic competition between Reactions (3)
(H)NO under conditions when the nitrite decomposition is
x
and (4) explains why less N O is formed during nitrite
rate-determining (i.e., high amberlyst loadings and hence
2
decomposition in the presence of NO (see above). If water
high reaction rates). Karl–Fischer titration of H O in the
2
2
is added to the reaction mixture, NO is rapidly trapped by
system shows a gradual increase of the water content from
2
Reaction (2), preventing it from reacting with HNO and the
reaction is quenched (see above).
approximately 0.5 w% just after addition of the HNO to
3
1.8 w% after 6 h (corresponding to 80% conversion).
This observation prompted us to remove water in situ.
Addition of molecular sieves to the reaction resulted in
inhibition, probably due to strong adsorption of (H)NOx,
present in the liquid phase. Therefore a gas recirculation
reactor was constructed (Figure 6) in which the gas phase is
constantly pumped over a fix-bed with molecular sieve 4 ꢁ to
remove the water via the gas phase.
NO can be re-oxidized by O to NO2 according to
2
[
22]
Reaction (5). In the absence of O , NO can probably also
2
[
24]
react with HNO3,
regenerating HNO2 and NO2 [Reac-
tion (6)], explaining why even under an inert N atmosphere a
2
D[PhCH OH]/[HNO ] ꢁ 2 can be achieved. Note that rapid
2
3 0
re-oxidation of NO to NO is crucial to prevent HNO from
2
dimerizing and forming N O. This explains why a higher end-
2
conversion can be reached at higher O pressure, given a
certain temperature (see Figure 1).
Using this set-up it is possible to significantly speed up the
reaction, while also increasing the end conversion to 100%
2
(Figure 7). Karl–Fischer titration confirms that this gas
1
2358 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 12355 –12360

both steps are acid catalyzed. These mechanistic insights will
be used as a starting point to investigate other HNO -based
3
oxidation reactions, such as adipic acid synthesis. The present
study shows that the reaction of the alcohol substrate with
HNO leads to the reactive (H)NO species. Most likely this is
2
x
the reason why cyclohexanol is needed as a (co-)substrate in
the adipic acid synthesis as pure cyclohexanone is rather inert
[
7]
under the same conditions.
Experimental Section
Standard experiments were carried out in a bubble column reactor
using 1,4-dioxane (25 mL; Fluka, p.a. ꢃ 99.5%) as
a solvent,
PhCH OH (500 mm; Acros, 99%), HNO (125 mm; Merck, p.a.
2
3
6
5%), NaNO2 (5 mm; Sigma–Aldrich, ꢃ 99%), and amberlyst-15
(150 mg; Sigma–Aldrich). High-pressure reactions were performed in
Figure 6. Gas recirculation reactor: a) heated tank-reactor, b) O reser-
2
a 100 mL stainless steel Parr reactor equipped with a glass insert and a
polyether ether ketone (PEEK) propeller. Products were quantified
using GC (HP-FFAP column, Flame-Ionization-Detector), or—in the
case of the thermally unstable nitrite—with GC-MS with cool-on-
column injection (HP-5 column). More information on the exper-
imental procedures can be found in the ESI.
voir connected to reactor by a check-valve, c) jacketed packed column
with 4 ꢀ molecular sieves (kept at reaction temperature to avoid
solvent condensation), d) condenser, e) condensate recycling, f) dried-
gas recycling, g) membrane pump, and h) gas disperser.
Received: August 8, 2011
Published online: October 21, 2011
Keywords: alcohols · nitric acid · oxidation · reaction kinetics ·
.
selectivity
[
[
1] F. Cavani, J. H. Teles, ChemSusChem 2009, 2, 508 – 534.
2] G. Franz, R. Sheldon, Oxidation, Ullmannꢀs Encycl. Ind. Chem.,
Wiley-VCH, Weinheim, 2000.
[
3] See, for example, a) I. Hermans, T. L. Nguyen, P. A. Jacobs, J.
Peeters, ChemPhysChem 2005, 6, 637 – 645; b) I. Hermans, P. A.
Jacobs, J. Peeters, Chem. Eur. J. 2006, 12, 4229 – 4240; c) U.
Neuenschwander, F. Guignard, I. Hermans, ChemSusChem
Figure 7. PhCH OH conversion in the presence of 5mm NaNO at
2
2
2
010, 3, 75 – 84.
8
08C versus time for a reaction performed in a) a bubble column
[
4] M. Thiemann, E. Scheibler, K. W. Wiegand, Nitric Acid, Nitrous
Acid, and Nitrogen Oxides, Ullmannꢀs Encycl. Ind. Chem.,
Wiley-VCH, Weinheim, 2000.
reactor and in the gas recirculation reactor with drying of the gas
ꢀ1
phase (gas recirculation rate of b) 75 and c) 150 mLmin ).
[
[
[
[
5] C. Seidel, BASF expands methanesulfonic acid plant in Ludwig-
shafen, Press Information, May 11th 2010.
6] See, for example, H. Osato, M. Kabaki, S. Shimizu, Org. Process
Res. Dev. 2011, 15, 581 – 584, and references therein.
7] M. T. Musser, Adipic Acid, Ullmannꢀs Encycl. Ind. Chem.,
Wiley-VCH, Weinheim, 2000.
8] a) I. Hermans, K. Janssen, B. Moens, A. Philippaerts, B.
Van Berlo, J. Peeters, P. A. Jacobs, B. F. Sels, Adv. Syn. Catal.
2007, 349, 1604 – 1608, and references therein; b) K. Banert, O.
Plefka, Angew. Chem. 2011, 123, 6295 – 6298; Angew. Chem. Int.
Ed. 2011, 50, 6171 – 6174.
recirculation reactor limits the water content to 0.5 w%. It is
important to emphasize that the observed effect is not due to
enhanced gas–liquid mass-transfer as increasing the O2
superficial flow rate in a bubble column reactor, or operating
the gas recirculation reactor without molecular sieve, did not
lead to an improvement.
From this study, it can be concluded that HNO is able to
3
start off a chain oxidation of benzyl alcohol, mediated by
[
9] A. Levina, S. Trusov, J. Mol. Catal. 1994, 88, L121 – L123.
(
H)NO species; O is however the terminal oxidant. HNO
x 2 2
[
10] a) Y. Kuang, N. M. Islam, Y. Nabae, T. Hayakawa, M. Kakimoto,
reacts with the alcohol substrate to form a nitrite ester which
decomposes, acid catalyzed, to the benzaldehyde product and
HNO. The reaction of HNO with NO , regenerates HNO and
Angew. Chem. 2010, 122, 446 – 450; Angew. Chem. Int. Ed. 2010,
49, 436 – 440; b) Y. Kuang, H. Rokubuichi, Y. Nabae, T. Hay-
2
2
akawa, M. Kakimoto, Adv. Synth. Catal. 2010, 352, 2635 – 2642.
forms NO which is oxidized to NO (propagation reactions).
[11] a) F. Bleger, O. Simon, A. Schouteeten, WO 2009/092734, 2009;
2
b) F. Bleger, O. Simon, A. Schouteeten, US 2011/0012056A1,
HNO can dimerize and yield N O in a termination reaction. A
2
2
011.
12] a) Y. Ogata, Y. Sawaki, F. Matsunaga, H. Tezuka, Tetrahedron
966, 22, 2655 – 2664; b) D. S. Ross, C.-L. Gu, G. P. Hum, R.
significant increase in chain length, and hence an improved
yield, can be achieved at lower temperature, or upon drying of
the gas phase. This chain reaction is catalyzed by the solid acid
amberlyst-15. Depending on the acid catalyst loading, either
[
1
Malhotra, Int. J. Chem. Kinet. 1986, 18, 1277 – 1288; c) S. R.
Joshi, K. L. Kataria, S. B. Sawant, J. B. Joshi, Ind. Eng. Chem.
Res. 2005, 44, 325 – 333.
the HNO or the nitrite decomposition is rate determining;
3
Angew. Chem. Int. Ed. 2011, 50, 12355 –12360
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12359

Communications
[
[
[
[
[
[
13] I. Hermans, J. Peeters, L. Vereecken, P. Jacobs, ChemPhysChem
[19] M. P. Doyle, J. W. Terpstra, R. A. Pickering, D. M. LePoire,
J. Org. Chem. 1983, 48, 3379 – 3382.
[20] H. F. Cordes, N. R. Fetter, J. A. Happe, J. Am. Chem. Soc. 1958,
2007, 8, 2678 – 2688.
14] Y. Wada, K. Morimoto, T. Goibuchi, H. Tomiyasu, Radiochim.
Acta 1995, 68, 233 – 243.
15] A. I. Kazakov, Yu. I. Rubtsov, L. P. Andrienko, G. B. Manelis,
Russ. Chem. Bull. 1987, 36, 1999 – 2002.
80, 4802 – 4808.
[21] V. Shafirovich, V. S. Lymar, Proc. Natl. Acad. Sci. USA 2002, 99,
7340 – 7345.
16] Y. Kameoka, R. L. Pigford, Ind. Eng. Chem. Fundam. 1977, 16,
[22] NIST Chemical Kinetics Database, http://kinetics.nist.gov/
163 – 169.
kinetics/.
17] K. P. C. Vollhardt, N. E. Shore, Organic Chemistry Structure and
Function, 4th ed., Freeman, New York, 2002.
18] M. A. Beckett, I. Hua, Environ. Sci. Technol. Environ. Sci.
Technol. 2000, 34, 3944 – 3953.
[23] W. Tsang, J. T. Herron, J. Phys. Chem. Ref. Data 1991, 20, 609 –
663.
[24] R. Svensson, E. Ljungstrçm, Int. J. Chem. Kinet. 1988, 20, 857 –
866.
1
2360 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 12355 –12360