Thermal hydrogenolysis of nitrogen-containing (hetero)arenes at elevated temperature and pressure

Article abstract of DOI:10.1002/ejoc.200400077

The behaviour of the aromatic nitrogen derivatives pyridine P and quinoline Q - and for comparison benzonitrile and azobenzene - was studied under elevated pressure (20 to 30 bar) in a flow reactor with an atmosphere of hydrogen at temperatures ranging from 773 to 1173 K. Compounds were introduced into the reactor admixed with benzene. Efficient removal of the heteroelement was observed, accompanied by deep gasification of the carbon skeletons, at higher temperatures. P reacts much faster than benzene, while Q is a little less reactive than P. The main products were methane, hydrogen cyanide and ammonia. Mechanisms are discussed based on induced decomposition by hydrogen atoms. The facile reaction of P is explained by hydrogen abstraction to give the 2-pyridyl radical followed by ring opening, a reaction with much better thermochemistry than that for benzene and the phenyl radical. Reaction via (partially) hydrogenated intermediates may also play a part. Exocyclic nitrogen - such as in benzonitrile - is readily displaced by hydrogen, while for azobenzene spontaneous fission of the rather weak C-N bond is the major rate- (and product-) determining step. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2004.

Full text of DOI:10.1002/ejoc.200400077

FULL PAPER
Thermal Hydrogenolysis of Nitrogen-Containing (Hetero)arenes at Elevated
Temperature and Pressure
Mariusz K. Cieplik[a] and Robert Louw*[b]
Keywords: Hydrogenolysis / Nitrogen heterocycles / NH3 / HCN / Methanisation
The behaviour of the aromatic nitrogen derivatives pyridine
based on induced decomposition by hydrogen atoms. The fa-
P and quinoline Q − and for comparison benzonitrile and azo- cile reaction of P is explained by hydrogen abstraction to give
benzene − was studied under elevated pressure (20 to 30 bar)
in a flow reactor with an atmosphere of hydrogen at temper-
atures ranging from 773 to 1173 K. Compounds were intro-
duced into the reactor admixed with benzene. Efficient re-
moval of the heteroelement was observed, accompanied by
deep gasification of the carbon skeletons, at higher temper-
atures. P reacts much faster than benzene, while Q is a little
the 2-pyridyl radical followed by ring opening, a reaction
with much better thermochemistry than that for benzene and
the phenyl radical. Reaction via (partially) hydrogenated in-
termediates may also play a part. ‘‘Exocyclic’’ nitrogen − such
as in benzonitrile − is readily displaced by hydrogen, while
for azobenzene spontaneous fission of the rather weak C−N
bond is the major rate- (and product-) determining step.
less reactive than P. The main products were methane, hy- ( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
drogen cyanide and ammonia. Mechanisms are discussed
Germany, 2004)
Introduction
graded to methane, and therefore this ‘‘dry’’ gasification
has potential applications in waste management technology.
Halogens can be tolerated even at high concentrations, as
these Ϫ and other Ϫ substituents are displaced by hydrogen
at much higher rates than arene rings are degraded.[1] Un-
der reductive conditions, organic chlorine derivatives ex-
clusively give HCl, unlike combustion, which also produces
free chlorine. Furthermore, ‘‘dioxins’’ (polychlorinated di-
benzo-para-dioxins and -furans, PCDD/Fs) cannot form,
and if present, will be detoxified, at least by hydro-dechlori-
nation to their parent compounds, dibenzo-p-dioxin (DD)
and dibenzofuran (DF).
The reactivity of DD and DF has also been dealt with
recently.[3] DD is relatively easily converted in a hydrogen
atmosphere at around 1000 K; this reaction is about a fac-
tor of 10 faster than that of chlorobenzene, which gives ben-
zene and HCl. Interestingly, in addition to CO, DF is an
important product. Both processes are sensitive to the con-
centration of hydrogen atoms, which appear to react by ad-
dition to (ϪCl or ϪO) substituted carbon atoms as a first
step.
Recently, the hydrogenolysis of benzene (B) and its hom-
ologues has been discussed.[1,2] Benzene (B) reacts readily
at around 1200 K in a quartz tubular flow reactor with
20Ϫ30 bar of hydrogen to give high yields of methane on a
second timescale. This process appears to proceed via (re-
versible) hydrogen atom addition and isomerisation, that is,
hydrogenation to the C6H8 species and subsequent splitting
into ·CH3 and ·C5H5 (e.g. cyclopentadienyl) radicals.[1]
Naphthalene (N) reacts about 10 times faster; this is under-
stood Ϫ also on a thermokinetic basis Ϫ overall to be an
easier ring-opening hydrogenolysis reaction.[2] Anthracene
(A) reacts only marginally faster than N; this is explained
by the extra energy barrier required to split the central ring,
leaving degradation of a side ring as the only effective chan-
nel for the reaction.[2]
Activated carbon (AC), ‘‘the ultimate PAH’’ is consider-
ably more resistant; its lifetime in its gasification to methane
is over an hour at 1250 K (970 °C).[2] This highly condensed
structure is less reactive with H·/H2, and effective break-
down of the aromatic rings is slow. Nevertheless, in the vap-
our phase even quite complex PAHs can be ultimately de-
DF is much more resistant to degradation. Moreover,
rates do not seem to depend on the hydrogen (atom) con-
centration, as hydrogenolysis with 1Ϫ35 bar of H2 and a
recent report[4] on the thermolysis in argon show. Naphtha-
lene is an important product throughout, and a thermolysis
mechanism for DF involving CϪO fission and isomeris-
ation has been advanced.
[a]
Energy Research Centre of the Netherlands,
P. O. BOX 1, 1755 ZG Petten, The Netherlands
Fax: (internat.) ϩ 31-224568487
E-mail: cieplik@ecn.nl
[b]
Leiden Institute of Chemistry, Gorlaeus Laboratories, Leiden
University,
P. O. Box 9502, 2300 RA Leiden, The Netherlands
Fax: (internat.) ϩ 31-71-5274451
E-mail: R.Louw@chem.leidenuniv.nl
Since nitrogen is a fairly common ingredient in organic
materials, especially those related to biomass or crude oil,
Eur. J. Org. Chem. 2004, 3011Ϫ3016
DOI: 10.1002/ejoc.200400077
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3011

M. K. Cieplik, R. Louw
FULL PAPER
it is worth investigating the fate of N-containing benzene
derivatives under the same conditions. In addition to het-
eroaromatic nitrogen compounds Ϫ pyridine (P) and
quinoline (Q) Ϫ our investigation includes benzonitrile
(BN) and azobenzene (AB) as examples of ‘‘exocyclic’’ N
derivatives.
Figure 2. Observed by-products in the hydrogenolysis of quinoline
Results and Discussion
Pyridine
The yields were low, less at higher degrees of conversion,
and never exceeded 1Ϫ2 mol% of the quinoline input. In-
dole (IN) and benzonitrile (BN) were relatively prominent.
In one case (at the lowest temperature) traces of carbazole
(C), 1,2,3,4- (1-THQ) and 5,6,7,8-tetrahydroquinoline (5-
THQ) and o-methylbenzonitrile (MB) were also found.
To compare the relative stability of N and Q several runs
were performed using a 1:1 mixture, again diluted in ben-
zene (ca. 5 mol% each), with a H2 pressure of 30 bar and a
residence time of about 6 s. The recoveries of both sub-
strates are depicted in Figure 3, which shows that Q appears
to react several times faster than N.
Five runs with pyridine were performed employing a mo-
lar inflow ratio of P/benzene/hydrogen of 0.05:1:12 at 20
bar with a residence time of about 4.5 s. The recoveries of
P are depicted in Figure 1. The main products were meth-
ane and ethane in a ratio of around 2. The formation and
survival of ethane can be ascribed to the relatively mild con-
ditions under which the conversion of P occurred: benzene
is essentially inert at T Ͻ 1100 K.[1] Higher molecular-
weight products included traces of (in decreasing order)
pyrrole, methylpyridine and quinoline (Q), but only in the
experiment performed at the lowest temperature. The nitro-
gen atom of pyridine ended up almost exclusively in the
form of HCN, reaching ‘‘synthetic’’ yields; at 1013 K, with
60% conversion of P, 55% of the pyridine input was found
as HCN.
Quinoline: HCN and NH3 Formation
HCN and NH3 were identified in the outflowing gases of
the reaction of Q. Figure 4 shows that the HCN yield re-
aches about 70 mol% of the inflow of Q under a pressure
of 20 bar and at 1173 K. Parallel to this, a 10Ϫ15% yield
of ammonia was observed. Under a pressure of 30 bar
yields of NH3 increased to about 35 mol% (dotted line in
Figure 4).
Quinoline
Analogously, runs were performed with about 4.6 mol%
Q in benzene under a pressure of 20 bar. The recoveries of
Q have been plotted in Figure 1; it appears that Q reacts
only marginally slower than P. Several nitrogen-containing
compounds were identified in the liquid product mixtures
(see Figure 2).
Benzonitrile and Azobenzene
As mentioned in the previous section, BN is a by-product
of the reaction of Q. Four experiments were performed to
Figure 1. Recoveries of pyridine and quinoline
3012
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3011Ϫ3016

Thermal Hydrogenolysis of Nitrogen-Containing (Hetero)arenes
FULL PAPER
Figure 3. Naphthalene and quinoline recoveries at 30 bar H2
Figure 4. HCN and NH3 formation from the reaction of quinoline
see if Ϫ and how rapid Ϫ the cyano group could be re- The calculated (extrapolated) first-order rate constants for
moved under the thermal hydrogenolysis (TH) conditions P at 1250 K have a range of over 2 orders of magnitude.
used. The results show that this very strong ArϪCN bond, However, when leaving out data extrapolated from meas-
BDE 131 kcal·molϪ1 [5]
can be hydrogenolysed with some urements taken at much higher temperatures (Entries
,
70% efficiency under relatively mild conditions (1123 K, 20 2,6Ϫ8) the k values are around 1Ϫ2 sϪ1. The value (also
bar H2). The only gaseous N-containing product was found extrapolated) determined in this work (Entry 9) is decidedly
to be HCN.
larger; depending on its temperature dependence, k ഠ 5 sϪ1
Azobenzene, when co-fed into the reactor with BN, was (for an experimental Ea of 40 kcal·molϪ1 [4]
to even 200
)
completely converted at 973 K. The azo-function is very (assuming Ea ϭ 75 kcal·molϪ1).[6] If we accept the shock-
likely converted to N2, but this has not been checked. Ne- tube data of Entries 6 and 7 as representative for unimolec-
ither NH3 nor N-containing organics could be detected, ex- ular homolysis of the (weakest) CϪH bond in P, the (ex-
cept for traces of aniline at lower temperatures.
trapolated) rate constant is some 20 times less than those
for thermolysis, and is essentially negligible compared with
our value for hydrogenolysis.[7Ϫ13]
Rates
Table 1 summarises the available literature data on the
The limited data on Q (Entries 10Ϫ12) show similar k
thermolytic reactions of P (Entries 1Ϫ8) and Q (10Ϫ12). values for shock-tube and regular thermolysis (despite a
Eur. J. Org. Chem. 2004, 3011Ϫ3016
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3013

M. K. Cieplik, R. Louw
FULL PAPER
Table 1. Summary of kinetic data for the reactivity of P and Q
Entry/[Ref.]/Technique
Reaction
A [sϪ1
]
Ea [kcal·molϪ1] k (at 1250 K) [sϪ1] T range [K]
1/[4]/static vs. time in Ar (0.012 bar)
pyridine Ǟ products
7.3 ϫ 1012
1.6 ϫ 1016
6.3 ϫ 106
3.8 ϫ 1012
73.8
0.88
1135Ϫ1200
2/[7]/shock tube in Ar (0.2Ϫ4 bar)
pyridine Ǟ products
100.3
39.5
0.044
1700Ϫ2000
3/[6]/stirred flow reactor, bath: pyridine
pyridine Ǟ products
0.77
1173Ϫ1273
4/[8]/static vs. time in He
70.0
2.12
1223Ϫ1323
pyridine Ǟ products
5/[9]/static vs. time, bath: pyridine (0.08Ϫ0.13 bar) 1.26 ϫ 1015
pyridine Ǟ 2-pyridyl ϩ H·
85.1
1.60
1100Ϫ1220
6/[10]/shock tube in Ar
7.94 ϫ 1015
98.0
0.055
1300Ϫ1800
pyridine Ǟ 2-pyridyl ϩ H·
7/[11]/shock tube in Ne (0.1Ϫ0.5 bar)
pyridine Ǟ 2-pyridyl ϩ H·
8/[12]/shock tube in He (2.2Ϫ3.4 bar)
pyridine Ǟ products
7.69 ϫ 1014
95.0
0.018
1711Ϫ2415
6.54 ϫ 109/8.1 ϫ 1010
52.5/63.5
5.50/0.620
Ͼ5 (Ϫ200)
1615Ϫ2032/1725Ϫ2335
1073Ϫ1103
9/this work: pyridine (see text)
rate constant kP for hydro- Ϫ
genolysis (20 bar)
10/[13]/shock tube in Ar (0.93 bar)
quinoline Ǟ products
1.00 ϫ 1013
1.59 ϫ 1013
2.1 ϫ 108
75.5
0.61
1275Ϫ1700
1135Ϫ1200
1130Ϫ1273
1073Ϫ1103
11/[4]/static vs. time in Ar (0.012 bar)
quinoline Ǟ products
77.9
44.9
0.37
12/[8]/static vs. time in He
quinoline Ǟ products
13/this work: quinoline
(see text)
2.9
rate constant kQ for hydro- Ϫ
genolysis (30 bar)
ca. 0.6 kP
large variation in rate parameters), but not unlike those for
In Scheme 1 the (reversible) hydrogen abstraction from
the thermolysis of P. Whatever their significance, our ob- P/Q/B by hydrogen atoms to give the radicals 2-P·, 2-Q·
served rate (constant) for hydrogenolysis (compare Fig- and phenyl (Ph·) is considered. This ‘‘assisted pyrolysis’’
ure 1) is about a factor of 1.5Ϫ2 less than that for P, al- with fragmentation of Ph· is important in the high tempera-
though the hydrogen pressure used was higher (30 rather ture pyrolysis of benzene,[14] but insignificant under our
than 20 bar). Note that Bruinsma et al.[4] have reported a conditions of hydrogenolysis. For P (and Q) the situation is
comparable relative rate for static pyrolysis (Entries 1 and different due to different thermochemistries. The heats of
11).
formation of Ph· and 2-P· (78 and 86.5 kcal·molϪ1, respec-
Whereas Q reacts a bit more slowly than P and faster tively) differ by 8.5 kcal·molϪ1. For the ring-opened species
than N, by far the largest rate difference is that between P 5 and 1 (with estimated heats of formation of 142 and 113
and benzene B. At 20 bar H2, 50% benzene is converted at kcal·molϪ1, respectively),[5] only 26 kcal·molϪ1 is required
a temperature of about 1250 K,[1] while P reaches the same for the isomerisation of 2-P· compared with 64 kcal·molϪ1
degree of conversion at about 1000 K (Figure 1). In free for Ph·. This allows a facile reaction of P, via 2-P·, 1 and 2
energy terms, this suggests that the hydrogenolysis of P is Ϫ and/or their closed-shell hydrogenation products Ϫ to,
about 13 kcal·molϪ1 less ‘‘costly’’ than that of B. Since the for example, acetylene and HCN. To put this on a more
CϪH bond neighbouring the nitrogen of P is (only) about quantitative kinetic basis, consider equilibrium (1):
5 kcal·molϪ1 weaker than a benzene CϪH bond,[14] the
major contribution to the difference in rates must be from
P ϩ H· ǟǞ 2-P· ϩ H2
(1)
rate determining steps other than CϪH bond scission Ϫ a
conclusion consonant with the observed overall rates (rate
constants) given in Table 1.
2-P· Ǟ products
(2)
The value of K1 can be estimated as follows. The reaction,
based on a CϪH bond strength[14] of about 106 kcal·molϪ1
Mechanisms
,
The proposed reaction rationales are presented in is endothermal by about 3 kcal. On the other hand, the
Schemes 1 and 2. Also aimed at understanding the forma- entropy term (at 1000 K) is about 8 cal·molϪ1·KϪ1 posi-
tion of the detected products these schemes illustrate a tive,[15] leading to log K1 ഠ 1. The logarithm of the equilib-
(major part of) the TH pathways of P and Q.
rium ratio [H·]/[H2] in 20 bar hydrogen is Ϫ9.3 at 1000 K
3014
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3011Ϫ3016

Thermal Hydrogenolysis of Nitrogen-Containing (Hetero)arenes
FULL PAPER
Scheme 1. TH of P, Q and B Ϫ comparison of the hydrogen abstraction pathways
(about Ϫ8.3 at 1100 K, and about Ϫ7.3 at 1200 K).[16]
In Scheme 2 another possibility is presented based on the
Consequently, the (log) ratio of [2-P·]/[P] is Ϫ 8.3. For log addition of hydrogen atoms, leading to hydrogenation of
k2, reasonable parameters [15] are log A2 ϭ 14 and E2 ϭ 30, the aromatic rings; the resulting intermediates are easily de-
hence log k2 ϭ 7.4 at 1000 K. This leads to an overall first- graded further, either by spontaneous (CϪC, CϪN) bond
order rate constant of about 0.1 sϪ1 for the conversion of homolysis or via radical species derived therefrom, as illus-
P at 1000 K, which is very close to the observed rate (Fig- trated next.
ure 1): ca. 50% reaction in 4.5 s implies a unimolecular rate
constant of about 0.15 sϪ1. By applying this approach to hold for P, which reacted only a bit more quickly than Q.
the reaction of P at 1100 K using the same parameters, a An analysis of available thermochemical data does not
Inasmuch as this scenario is relevant, the same should
rate constant of about 5 sϪ1 is obtained (compare Table 1). lead to clear-cut ‘‘advantageous’’ pathways for the hydro-
In the case of quinoline the reaction via the ring-opened genolysis of these or analogous (partially hydrogenated) N
isomer 3 is not as straightforward as it is for pyridine; the derivatives. Whereas the full hydrogenation of P to piperi-
splitting off of benzyne (4), requires much more energy dine is exothermal by 45 kcal·molϪ1 Ϫ about 4.5 kcal less
(some 30 kcal·molϪ1) than the loss of acetylene in the case than that of B to cyclohexane Ϫ NϪH or NϪC bonds, on
of pyridine. A possible mechanism for this process is the the other hand, are about 6Ϫ8 kcal·molϪ1 weaker than the
hydrogenation of 3 to the substituted benzene, which is then corresponding CϪC and CϪH bonds.[5] The same can be
subject to simple and rapid ‘‘desubstitution’’.
said for dihydroarene-based pathways as discussed for N.[2]
Scheme 2. TH of Q by hydrogenation
Eur. J. Org. Chem. 2004, 3011Ϫ3016
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3015

M. K. Cieplik, R. Louw
FULL PAPER
In the competition reaction involving Q and N(aphtha-
lene), the latter appeared to react only several times more
slowly, and compared with rates observed in our earlier
study, at least an order of magnitude more rapidly, by a
comparable hydrogenation/ring opening/fragmentation
scenario.[2] The reason for this difference is as yet unknown.
For similar reactivities, the hydrogenolysis would be ex-
pected to occur on both rings of Q. However, no aliphatic
pyridine derivatives, except the earlier mentioned 5-THQ,
were detected in the liquids produced in the experiments
with quinoline, whereas products arising from hydro-
genolysis of the ‘‘pyridine’’ ring were clearly present. One
explanation could be that the aliphatic pyridine derivatives
are even more vulnerable to TH, and get readily converted
even at mild temperatures.
In the proposed rationale (Scheme 2) both HCN and am-
monia are direct products from the hydrogenolysis of Q.
The question of whether HCN can be hydrogenolysed to
form ammonia (observed at high degrees of conversion of
Q), which seems very unlikely on thermodynamic grounds,
cannot yet be answered on the basis of the present results.
Azobenzene reacted completely at the lowest temperature
of this study (973 K), with no intermediate products
formed. One reason for this is the character of the azo
bond. Reaction by addition of hydrogen to the nitrogen
double bond appears to be energetically unfavourable. Hy-
drogenolysis by ipso-addition of a hydrogen atom may well
occur, but spontaneous homolysis of the rather weak CϪN
bond is the most likely first step.
Experimental Section
Setup: The high-pressure set-up as well as the experimental pro-
cedures have been described elsewhere.[1]
Sampling and Analyses of Inorganic Products: Inorganic products
were trapped in an ice-cooled, magnetically-stirred water trap.
Upon detection, by treating the solution obtained with a standard
solution of FeII/FeIII (0.1 mol each as chlorides), hydrogen cyanide
was quantified by potentiometric titration (0.1 mol AgNO3, Titri-
sol standard solution) on an automatic Mettler 25 titrator.
Addition of HCl to the aqueous solution obtained followed by
evaporation in a rotavap gave a white, subliming crystalline prod-
uct. The substance was analysed by IR spectroscopy and the spec-
trum matched that of analytically pure ammonium chloride. Am-
monia was quantified by pH-metric titration with a 0.01  HCl
solution (Titrisol).
Chemicals: The following chemicals were utilised (purity given in
%): benzene (Merck p.a., 99%, distilled), monobromobenzene
(Baker, Ͼ 99.5%), naphthalene (Janssen Chimica, Ͼ 99%), benzo-
nitrile (Merck, Ͼ 99%), azobenzene (Ͼ 99%, recrystallised), n-pen-
tane (Baker Analysed, Ͼ 99%), quinoline (Ͼ 99%, distilled in
vacuo), pyridine (Merck,
Ͼ 99%), hydrogen (Air Products,
99.995%), methane (Air Products, 99.995%), nitrogen (Air
Products 99.995%).
[1]
W. B. van Scheppingen, M. K. Cieplik, R. Louw, Eur. J. Org.
Chem. 2001, 11, 2101Ϫ2106.
M. K. Cieplik, R. Louw, W. B. van Scheppingen, Eur. J. Org.
[2]
Chem. 2001, 23, 4517Ϫ4524.
Also, benzonitrile reacted for a great part under relatively
mild conditions. Similar to the results of earlier stud-
ies,[17,18] the facile removal of the ϪCN functionality under
hydrogenolysis conditions, despite the very strong
areneϪCN bond, can be understood in terms of the (revers-
ible) addition of a hydrogen atom to the carbon of the cy-
ano function, followed by fragmentation to HCN and the
phenyl radical.
[3]
M. K. Cieplik, O. J. Epema, R. Louw, Eur. J. Org. Chem. 2002,
16, 2792Ϫ2799.
O. S. L. Bruinsma, P. J. J. Tromp, H. J. J. de Sauvage Nolting,
[4]
J. A. Moulijn, Fuel 1988, 67, 334Ϫ340.
[5]
NIST thermochemical database at http://webbook.NIST.gov.
[6]
T. J. Houser, M. Hull, R. M. Alway, T. Biftu, Int. J. Chem.
Kinet. 1980, 12, 569Ϫ574.
H. I. Leidreiter, H. Gg. Wagner, Z. Phys. Chem. (Munich)
[7]
1987, 153, 99Ϫ103.
[8]
A. E. Axworthy, V. H. Dayan, G. B. Martin, Fuel 1978, 57,
29Ϫ35.
[9]
E. Ikeda, J. C. Mackie, J. Anal. Appl. Pyrolysis 1995, 34,
47Ϫ63.
J. C. Mackie, M. B. Colket III, P. F. Nelson, J. Phys. Chem.
Conclusions
[10]
High-pressure thermal hydrogenolysis has proven to be
an efficient way to convert (hetero)aromatics. Basic product
patterns are common for all compounds studied and can be
briefly described as ‘‘mineralization’’ and, at sufficiently
high temperatures, ‘‘gasification’’. The nitrogen atoms of
pyridine P and quinoline Q are converted into HCN and
NH3 in reactions which are (much) faster than the degra-
dations of benzene or naphthalene. Decomposition is in-
duced by hydrogen atoms. Thermochemical considerations
show that reversible formation of 2-pyridyl radicals fol-
lowed by ring opening to a cyano-pentadienyl radical Ϫ 1
in Scheme 1 Ϫ is much easier than the analogous process
with benzene. Alternatively, (partial) hydrogenation of P
and Q may occur, also followed by ring opening and
further fragmentation.
1994, 94, 4099Ϫ4102.
[11]
R. D. Kern, J. N. Yong, J. H. Kiefer, J. N. Shah, Proc. Int.
Symp. Shock Tubes Waves 1987, 16, 437Ϫ440.
[12]
H. U. R. Memon, K. D. Bartle, Int. J. Energy Res. 2000, 24,
1141Ϫ1159.
[13]
A. Laskin, A. J. Lifshitz, J. Phys. Chem. A 1998, 102, 928Ϫ946.
C. Barckholtz, T. A. Barckholtz, C. M. Hadad, J. Am. Chem.
Soc. 1999, 121, 491Ϫ500.
S. W. Benson, Thermochemical Kinetics, Wiley, New York,
[14]
[15]
1968.
[16]
D. R. Stull, H. Prophet (Project Directors), JANAF Thermo-
chemical Tables, 2nd ed., National Bureau of Standards,
Washington, DC, USA, 1971, NSRDS-NBS 37.
K. van Nierop, R. Louw, Recl. Trav. Chim. Pays-Bas 1977,
96, 176Ϫ178.
K. van Nierop, F. J. Kuyers, W. F. M. Vonk, R. Louw, J. Chem.
Soc., Perkin Trans. 2 1977, 1062Ϫ1066.
[17]
[18]
Received February 5, 2004
3016
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3011Ϫ3016