The effect of pressure on hydrogen transfer reactions with quinones

Article abstract of DOI:10.1002/chem.200305686

The effect of pressure on the oxidation of hydroarenes 3-9 with 2,3-dichloro-5,6-dicyano-1,4-quinone (DDQ; 1a) or o-chloranil (10), leading to the corresponding arenes, has been investigated. The activation volumes were determined from the pressure dependence of the rate constants of these reactions monitored by on-line UV/Vis spectroscopic measurements in an optical high-pressure cell (up to 3500 bar). The finding that they are highly negative and only moderately dependent on the solvent polarity (ΔV ^?=-13 to -25 in MTBE and -15 to -29 cm³mol ^-1 in MeCN/AcOEt, 1:1) rules out the formation of ionic species in the rate-determining step and is good evidence for a hydrogen atom transfer mechanism leading to a pair of radicals in the rate-determining step, as was also suggested by kinetic measurements, studies of kinetic isotope effects, and spin-trapping experiments. The strong pressure dependence of the kinetic deuterium isotope effect for the reaction of 9,10-dihydroanthracene 5/5-9,9,10,10-D4 with DDQ (1a) can be attributed to a tunneling component in the hydrogen transfer. In the case of formal 1,3-dienes and enes possessing two vicinal C-H bonds, which have to be cleaved during the dehydrogenation, a pericyclic hydrogen transfer has to considered as one mechanistic alternative. The comparison of the kinetic deuterium isotope effects determined for the oxidation of tetralin 9/9-1,1,4,4-D₄/9-2,2,3,3-D₄/9-D ₁₂ either with DDQ (1a) or with thymoquinone 1c indicates that the reaction with DDQ (1a) proceeds in a stepwise manner through hydrogen atom transfer, analogously to the oxidations of 1,4-dihydroarenes, whereas the reaction with thymoquinone 1c is concerted, following the course of a pericyclic hydrogen transfer. The difference in the mechanistic courses of these two reactions may be explained by the effect of the CN and Cl substituents in 1a, which stabilize a radical intermediate better than the alkyl groups in 1c. The mechanistic conclusions are substantiated by DFT calculations.

Full text of DOI:10.1002/chem.200305686

FULL PAPER
The Effect of Pressure on Hydrogen Transfer Reactions with Quinones
Frank Wurche,[a] Wilhelm Sicking,[a] Reiner Sustmann,[a] Frank-Gerrit Kl‰rner,*[a] and
Christoph R¸chardt[b]
Abstract: The effect of pressure on the
oxidation of hydroarenes 3 9 with 2,3-
dichloro-5,6-dicyano-1,4-quinone
radicals in the rate-determining step, as
was also suggested by kinetic measure-
ments, studies of kinetic isotope effects,
and spin-trapping experiments. The
strong pressure dependence of the ki-
netic deuterium isotope effect for the
reaction of 9,10-dihydroanthracene 5/5-
9,9,10,10-D4 with DDQ (1a) can be at-
tributed to a tunneling component in
the hydrogen transfer. In the case of
formal 1,3-dienes and enes possessing
tive. The comparison of the kinetic
deuterium isotope effects determined
for the oxidation of tetralin 9/9-1,1,4,4-
D4/9-2,2,3,3-D4/9-D12 either with DDQ
(1a) or with thymoquinone 1c indi-
cates that the reaction with DDQ (1a)
proceeds in a stepwise manner through
hydrogen atom transfer, analogously to
the oxidations of 1,4-dihydroarenes,
whereas the reaction with thymoqui-
none 1c is concerted, following the
course of a pericyclic hydrogen trans-
fer. The difference in the mechanistic
courses of these two reactions may be
explained by the effect of the CN and
Cl substituents in 1a, which stabilize a
radical intermediate better than the
alkyl groups in 1c. The mechanistic
conclusions are substantiated by DFT
calculations.
(DDQ; 1a) or o-chloranil (10), leading
to the corresponding arenes, has been
investigated. The activation volumes
were determined from the pressure de-
pendence of the rate constants of these
reactions monitored by on-line UV/Vis
spectroscopic measurements in an opti-
cal high-pressure cell (up to 3500 bar).
The finding that they are highly nega-
tive and only moderately dependent on
the solvent polarity (DV° = ꢀ13 to
ꢀ25 in MTBE and ꢀ15 to
ꢀ29 cm3 molꢀ1 in MeCN/AcOEt, 1:1)
rules out the formation of ionic species
in the rate-determining step and is
good evidence for a hydrogen atom
transfer mechanism leading to a pair of
ꢀ
two vicinal C H bonds, which have to
be cleaved during the dehydrogenation,
a pericyclic hydrogen transfer has to
considered as one mechanistic alterna-
Keywords: DFT calculations ¥ iso-
tope effects ¥ kinetics ¥ oxidation ¥
quinones ¥ radicals
Introduction
1) Direct hydride transfer, leading to a pair of ions from
which the observed products can be formed by proton
transfer.[1,2]
2) Hydrogen atom transfer, leading to a pair of free radi-
cals in the rate-determining step, followed by fast reac-
tions such as disproportionation of the free radicals or
single-electron transfer (SET) and subsequent proton
transfer, producing the observed arene and hydroqui-
none.[4,5]
Quinones are important oxidizing reagents in synthetic[1,2]
and biological processes.[3] For example, the oxidation of hy-
droarenes by quinones such as 2,3-dichloro-5,6-dicyano-1,4-
quinone (DDQ, 1a) or p-chloranil (1b) is frequently used
for the synthesis of aromatic compounds.[1,2] Five mecha-
nisms for the quinone oxidation of hydroarenes have been
proposed (Scheme 1):
3) SET from the hydroarene to the quinone, followed by a
proton transfer, producing the same pair of radicals as
the hydrogen atom transfer, which reacts further as de-
scribed under (2).[6]
[a] Dr. F. Wurche, Dipl.-Ing. W. Sicking, Prof. Dr. R. Sustmann,
Prof. Dr. F.-G. Kl‰rner
4) A pericyclic ene reaction, leading to a covalently bound
adduct, which is subsequently cleaved by a retro-ene re-
action to yield the observed products after keto enol
tautomerization.[7]
Institut f¸r Organische Chemie der Universit‰t Duisburg-Essen
Universit‰tsstrasse 5, 45117 Essen (Germany)
Fax : (+49)201-183-4252
E-mail: frank.klaerner@uni-essen.de
5) Finally, a pericyclic dyotropic hydrogen transfer.[8] This is
limited, however, to hydroarenes possessing two vicinal
[b] Prof. Dr. C. R¸chardt
Institut f¸r Organische Chemie und Biochemie
der Universit‰t Freiburg
Albertstrasse 21, 79104 Freiburg i. Br. (Germany)
ꢀ
C H bonds that can both be cleaved in the product-de-
termining step.[9]
2707
Chem. Eur. J. 2004, 10, 2707 2721
DOI: 10.1002/chem.200305686
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

FULL PAPER
bilities of these cations[13]
a
huge difference in the reactivity
of 5 and 6 would be expected in
the case of the direct hydride
transfer, contrary to the experi-
mental observations.[4] The
large primary kinetic isotope ef-
fects (kH/kD = 8.3 and 7.85)
measured for the reactions of
5/5-9,9,10,10-D4 and 6/6-9,9-D2,
respectively, at 508C[4] indicate
that the hydrogen is transferred
in the rate-determining transi-
tion state and rule out a single-
electron transfer in the rate-de-
termining step preceding the
proton transfer in a fast subse-
quent process. A distinction be-
tween a hydrogen atom transfer
and a proton-coupled electron
transfer, as discussed for the
benzyl/toluene, methoxy/metha-
nol, and phenoxy/phenol self-
exchange reactions,[14] is not ex-
perimentally feasible, so we do
not consider this possibility
below. Further evidence for the
hydrogen atom transfer is pro-
vided by spin-trapping experi-
ments. In the reactions of 5 and
6 with DDQ (1a), intermediary
radicals could be trapped by ni-
trosobenzene, producing the
more stable radicals 12 and 13
detected by ESR spectrosco-
py.[4]
Scheme 1. Mechanistic alternatives for quinone dehydrogenation of hydroaromatic compounds: 1) direct hy-
dride transfer, 2) hydrogen atom transfer, 3) single-electron transfer, 4) ene and retro-ene reaction, and 5) peri-
cyclic dyotropic hydrogen transfer.
From the observations that the oxidation of hydroarenes
with quinones follows second-order kinetics, that the rate of
reaction is enhanced by increasing solvent polarity but only
moderately, and that products of carbenium ion rearrange-
ments are occasionally observed, Braude, Linstead, et al.[1]
have proposed the direct hydride transfer (1) as the most
probable mechanism. Further support for this proposal,
widely accepted even today,[10] is provided by the lack of
rate effects when free radical initiators or inhibitors were
added. This, indeed, excludes free radical chain reactions
quite safely, but certainly not a stoichiometric non-chain ho-
molytic process as discussed for the hydrogen atom transfer
(mechanism (2)).[4] The findings that the effect of solvent
polarity on the rate of reaction is small[11] and comparable
to the effects observed for free radical reactions[12] and that
xanthene 6 reacts more rapidly than 5 with DDQ (1a) only
by a factor of 16 at 258C (DDG° = 2 kcalmolꢀ1)[4] is good
evidence in favor of the hydrogen atom transfer and against
the direct hydride transfer, in which the carbenium ions de-
rived from 5 and 6 would have to be generated in the rate-
determining step. Because of the large difference in the sta-
To gain further insight into the mechanism of the quinone
oxidation we investigated the effects of pressure on the re-
actions of the hydroarenes 3 9 (Scheme 2) with DDQ (1a)
and o-chloranil (10) and on the primary kinetic isotope
effect kH/kD in the reaction of 5/5-9,9,10,10-D4 with 1a.
Compounds 1a and 10 are reduced to 2a and 11 in the reac-
tion. Comparison between the kinetic isotope effects kH/kD
in the reactions of the tetralin derivatives 9/9-1,1,4,4-D4/9-
D12 either with 1a, previously unknown and first reported
here, or with thymoquinone 1c, which have been studied by
Brower et al.,[9] allows a distinction to be made between the
potential mechanisms (2) and (5).
The effect of pressure on quinone dehydrogenations: deter-
mination of volumes of activation and reaction: Pressure in
2708
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2707 2721

Hydrogen Transfer Reactions
2707 2721
larger than the expansion of
volume resulting from the dis-
sociation. Thus, the overall
effect, called electrostriction,
leads to negative volumes of ac-
tivation and reaction.[17] In asso-
ciative formation of ions–in
the alkylation of pyridine deriv-
atives (Menshutkin reaction)[18]
or in the polar [2+2] cycloaddi-
tion of a vinyl ether to tetracya-
noethylene (TCNE),[19] for ex-
ample–both
effects
are
volume-contracting and sum to
give highly negative volumes of
activation and reaction, which
are strongly dependent on the
solvent polarity. In nonpolar
solvents the volume contraction
caused by the effect of electro-
striction is larger than in polar
solvents. For example, the acti-
vation volume of the [2+2] cy-
cloaddition of n-butyl vinyl
ether to TCNE involving
a
zwitterionic intermediate was
determined to be less negative
in
acetonitrile
(DV°
=
ꢀ23 cm3 molꢀ1) than in CCl4
(DV° = ꢀ44 cm3 molꢀ1).[19] Sim-
ilar results were obtained for
the alkylation of pyridine with
Scheme 2. The hydroaromatic compounds used as hydrogen or deuterium donors: a) formal 1,4-dienes,
b) formal 1,3-dienes, c) formal enes, and d) the quinones used as oxidizing reagents and the resulting hydroqui-
nones.
benzyl bromide (DV°
=
ꢀ29.2 cm3 molꢀ1 in acetonitrile
the range of 1 20 kbar influences the rate and equilibrium
position of many chemical reactions.[15] Processes accompa-
nied by a decrease in volume–such as association, ioniza-
tion, and cyclization–are accelerated by pressure (volume
of activation DV° <0) and the equilibria are shifted toward
the side of the products (volume of reaction: DV<0). The
reverse reactions–such as homolytic dissociation, neutrali-
zation of charges, and cycloreversion–lead to an increase in
volume. In these reactions, pressure induces a deceleration
and a shift of equilibrium toward the side of the reactants
(DV°, DV>0). The overall change in the molar volume
during a reaction, as determined experimentally from the
volumes of activation and of reaction, consists of the
changes in the intrinsic volumes of the reacting molecules
DVW (VW = van der Waals volume) and of the changes in
the empty space between the molecules (void and expansion
volume),[16] which is required for thermally induced motions
and collisions in the liquid state. How important it is to con-
sider the whole ensemble of molecules and not single mole-
cules for the explanation of the pressure effect can be dem-
onstrated in the effect of electrostriction,[17] which we also
need for the explanation of the pressure effects reported
here. In a heterolytic bond dissociation, the attractive inter-
action between the newly generated ions and the solvent
molecules leads to a contraction of volume generally much
and DV° = ꢀ39.1 cm3 molꢀ1 in toluene).[18] To distinguish
between the mechanistic alternatives of the quinone dehy-
drogenation–the direct hydride transfer leading to a pair of
ions (comparable to the intermediary zwitterion in the polar
[2+2] cycloaddition) and the hydrogen atom transfer lead-
ing to a pair of non-charged radicals (Scheme 1: mecha-
nisms (1) and (2), respectively)–we determined the activa-
tion volumes of the reactions of DDQ (1a) with the hydro-
arenes 3 9 and of o-chloranil (10) with 5 and 6 in MTBE
(tert-butyl methyl ether: ET = 34.7, e = 4.5 D) and in the
more polar mixture (1:1) of acetonitrile (ET = 45.6, e =
35.5 D) and ethyl acetate (ET = 38.1, e = 6.0 D) as sol-
vents. The low solubilities of some of the hydroarenes did
not allow us to use pure acetonitrile as solvent. The rate
constants k1 of the reactions of the quinones 1a and 10 with
the hydroarenes 3 9 over the 1 3500 bar pressure range
were determined under pseudo-first-order kinetics condi-
tions with an excess of the hydroarene by on-line UV/Vis
monitoring of the decrease in the quinone concentration in
an optical 4 kbar high-pressure cell. For the example of the
reaction of o-chloranil (10) with 6, the order of the reaction
was determined by measurement of the dependence of the
pseudo-first-order rate constant kobs on the hydroarene con-
centration of 6 to be, indeed, second order. Details of the
determination of pressure-dependent rate constants k1 are
2709
Chem. Eur. J. 2004, 10, 2707 2721
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

F.-G. Kl‰rner et al.
FULL PAPER
described in the Experimental Section. Since the slope of
lnk1 against pressure (p) is not linear in the range of 1 to
3500 bar investigated here, as illustrated for the pressure de-
pendence of the dehydrogenation of 1,4-cyclohexadiene (3)
with DDQ (1a) in Figure 1, the activation volumes were de-
termined from the slopes of these curves at p = 0 by use of
the quadratic Equation (3) given in the Experimental Sec-
tion.
partial molar volume of the transition state (V° = (V(1a)+
V(3)+DV°) can be calculated from these partial molar vol-
umes and the activation volume. The van der Waals volumes
(VW[22]) of the starting and transition structures involved in
this reaction (defined by the space occupied by the van der
Waals spheres of the atoms) were calculated from the Carte-
sian coordinates of the molecular structures predicted by
quantum mechanical calculations (vide infra). The ratio
VW/V or VW°/V° is defined as the packing coefficient of a
starting or transition structure(h and h°) and provides infor-
mation about the packing of the molecular structures in the
liquid state, and hence of how much empty space is between
the molecules.[16]
DFT calculations on the mechanism of hydrogen transfer:
Density functional calculations (B3LYP) were performed on
the mechanism of the hydrogen transfer reactions from 1,4-
cyclohexadiene (3) to DDQ (1a) and o-chloranil (10). Gaus-
sian 98[23] was used with two basis sets (6 31G* and
6 31G**), and restricted or unrestricted calculations were
carried out where appropriate. Ground states and transition
structures (TSs) were characterized by vibrational analysis
and in each case correspond to minima or maxima (one
imaginary frequency) on the potential energy surface.
Table 2 displays the enthalpies (298.15 K), dipole moments,
and DH values relative to starting materials. Two transition
structures were considered in each case (Figure 3 and
Figure 4), one showing an extended (TS1) and the other one
a compact structure (TS2). For each reaction, TS2 is lower in
energy: by 1.6 kcalmolꢀ1 (1a and 6 31G**) and
0.8 kcalmolꢀ1 (10 and 6 31G**), respectively. Similar differ-
ences, but slightly higher in absolute values, are obtained
with the 6 31G* basis set. The preference for the compact
structures might reflect favorable p p interactions of the p-
electrons of the electron-deficient quinone with those of 1,4-
cyclohexadiene. Alternatively, it might be argued that the
structure of TS2 enables stabilizing CH p interactions of
one of the hydrogen atoms of the CH2 group of cyclohexa-
diene with the p-electrons of the quinone. The closest
Figure 1. Pressure dependence of the second-order rate constants k1 of
the dehydrogenation of 3 with 1a in tert-butyl methyl ether (MTBE) and
in acetonitrile/ethyl acetate (MeCN/AcOEt) at 25.18C.
The activation volumes of the dehydrogenation of the hy-
droarenes 3 9 with DDQ (1a) or o-chloranil (10) are listed
in Table 1, together with the other activation parameters en-
thalpy (DH°), entropy (DS°), and Gibbs enthalpy (DG°) of
activation, which were taken from the literature.[20] The un-
known parameters for the reaction of 5 and 6 with 10 were
obtained from the temperature dependence of the rate con-
stants at atmospheric pressure measured in this work.
The complete volume profile for the reaction between
DDQ (1a) and 1,4-cyclohexadiene (3) was determined as a
representative example (Figure 2). The partial molar vol-
umes (V) of the reactants (1a, 3) and products (2a, benzene
(15)) were obtained from density measurements in MTBE
and in an MeCN/AcOEt mixture (1:1), the solvents also
used for determination of the activation volumes. The reac-
tion volume (DV = (V(2a)+V(15))ꢀ(V(1a)+V(3)) and
(within the scope of Eyring transition state theory) also the
ꢀ
CH¥¥¥Csp2 distance is 3.08 pm, and the angle C H¥¥¥Csp2 is
1498. This arrangement is typical for CH p interactions.[24]
Another interesting feature concerns the increases in the
dipole moments of the transition structures in relation to
those of the starting materials. While 3 has no dipole
moment, the calculated m of 1a (DDQ) is 4.0 D and of 10
Table 1. Volumes of activation (DV°) and activation parameters of the dehydrogenation of hydroarenes to the corresponding arenes by DDQ (1a) and
by o-chloranil (10) (MTBE = tert-butyl methyl ether, MeCN = acetonitrile, AcOEt = ethyl acetate).
DV°MTBE
DV°MTBE=AcOEt
[cm3 molꢀ1
DH°[a]
DS°[a]
[E.U.]
DG°25 o
C
[a]
Quinone
H-donor
T
[8C]
[cm3 molꢀ1
]
]
[kcalmolꢀ1
]
[kcalmolꢀ1
]
DDQ (1a)
3
4
5
7
8
9
25.1
25.0
25.1
39.9
39.9
64.9
39.9
25.1
25.1
ꢀ24.4ꢁ0.2
ꢀ24.8ꢁ0.2
ꢀ25.9ꢁ0.3
ꢀ25.3ꢁ0.4
ꢀ25.1ꢁ0.5
ꢀ29.5ꢁ0.5
ꢀ28.9ꢁ0.4
ꢀ26.9ꢁ0.8
ꢀ28.4ꢁ0.5
ꢀ29.4ꢁ0.5
ꢀ26.7ꢁ0.4[c]
ꢀ22.9ꢁ0.2
ꢀ18.1ꢁ 0.5[c]
ꢀ15.5ꢁ0.3[c]
11.7ꢁ0.1
10.8ꢁ0.1
11.7ꢁ0.1
ꢀ28.2ꢁ0.4
ꢀ28.7ꢁ0.7
ꢀ28.1ꢁ0.3
ꢀ29.9ꢁ3.6[b]
ꢀ25.6ꢁ2.4
20.1ꢁ0.1
19.4ꢁ0.2
20.1ꢁ0.1
12.9ꢁ1.2[b]
16.2ꢁ0.8
22.6 ꢁ1.2[b]
23.8ꢁ0.9
5 9,9-Me2
ꢀ20.0ꢁ0.3
ꢀ16.4ꢁ0.3
ꢀ13.0ꢁ0.2
10
5
6
13.3ꢁ0.3[d]
12.1ꢁ0.3[d]
ꢀ24.0ꢁ1.0[d]
ꢀ25.6ꢁ1.0[d]
20.4ꢁ0.3[d]
19.7ꢁ0.3[d]
[a] Ref.[20], solvent dioxane. [b] Ref.[21]. [c] In acetonitrile. [d] This work.
2710
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2707 2721

Hydrogen Transfer Reactions
2707 2721
ꢀ0.6 electrons and deficiencies
of corresponding magnitude of
electrons in 3. If, however, the
charge of +0.4 on the migrat-
ing hydrogen atom is attributed
to DDQ and taken away from
3, then the overall process does
not show hydride transfer char-
acter, but, in contrast, hydrogen
atom transfer character. An im-
portant property of the transi-
tion structures is the position of
the transferred hydrogen atom.
According to the DFT calcula-
ꢀ
tions the C H bond length of
the CH2 hydrogen atoms in 1,4-
cyclohexadiene is 1.10 ä, while
this bond is stretched to 1.33 ä
for (3+10)TS1 and 1.37 ä for
(1a+3)TS1. The elongation of
ꢀ
the C H bond length is more
pronounced in TS2, being
1.41 ä both in (3+10)TS2 and
in (1a+3)TS2. The OH bonds
undergoing
formation
are
1.27 ä in (3+10)TS1 and
1.22 ä in (1a+3)TS1, being
shorter and identical (1.19 ä)
in the compact TS2s. The 6
31G** basis set provides com-
parable, sometimes slightly
smaller distances (see Figure 3
and Figure 4). The conclusion is
Figure 2. Volume profile[a] of the reaction between 1,4-cyclohexadiene (3) and DDQ (1a). The van der Waals
volumes VW were calculated from the Cartesian coordinates of the ground and transition structures resulting
from quantum mechanical calculations by use of the MOLVOL computer program.[22]
3.4 D. Whereas the transition structures for hydrogen trans-
fer to o-chloranil (10) show increases in m of about 1 D
((3+10)TS1 and (3+10)TS2), these increases are more than
4 D ((1a+3)TS2) and more than 6 D ((1a+3)TS1) in the
case of DDQ (1a). This indicates a net electron flow be-
tween the reactants, in particular in the case of the more
electron-deficient DDQ. Natural population analysis (NPA)
of the charge distributions for (1a+3)TS1 and (1a+3)TS2
indicates surpluses of negative charge on 1a of ꢀ0.5 to
that the hydrogen atom is midway between starting materi-
als and products. In all the transition structures, the negative
vibration corresponds to the transfer of the hydrogen atom
from 3 to 1a or 10. Although CAS(2,2) is approximate and
crude,[25] we determined the biradical characters (BRCs) of
the transition structures (see Table 2). Depending on the
basis set, the values range from 16 to 34%. As observed in
other examples,[26] such low BRCs also yield closed-shell
wave functions (S2 = 0) when determined by unrestricted
Table 2. B3LYP DFT calculations on the mechanism of hydrogen transfer; basis sets 6 31G* (A) and 6 31G** (B).
H (298.15 K) DH°[a]
[a.u.]
m
%BRC
3Eꢀ1E
H (298.15 K) DH°[a]
[a.u.]
m
%BRC
3Eꢀ1E
[kcalmolꢀ1
]
[D] CAS(2,2)/6 31G* [kcalmolꢀ1
A
]
[kcalmolꢀ1
]
[D] CAS(2,2)/6 31G* [kcalmolꢀ1
B
]
Structure
1a
3
ꢀ1485.0185[b]
ꢀ233.2897[b]
4.0
0.0
ꢀ1485.0185[b]
4.0
0.0
ꢀ233.3022[b]
(1a+3)TS1 ꢀ1718.2868[c] 13.4
10.6
8.5
3.4
4.7
4.9
0.4
7.3
5.0
0.8
3.3
34
23
8.2
14.6
ꢀ1718.3024[c] 11.5
10.4
8.3
3.4
4.7
4.9
0.4
7.3
5.0
0.8
3.3
30
21
8.9
15.3
(1a+3)TS2 ꢀ1718.2893[c] 11.9
ꢀ1718.3050[c]
ꢀ2219.7395[b]
9.9
10
ꢀ2219.7395[b]
(3+10)TS1 ꢀ2453.0069[c] 14.0
30
18
7.3
19.0
ꢀ2453.0223[c] 12.2
ꢀ2453.0236[b] 11.4
ꢀ232.6894[c]
25
16
8.3
19.7
(3+10)TS2 ꢀ2453.0077[b] 13.5
16
ꢀ232.6783[c]
ꢀ1485.6153[c]
ꢀ1485.6151[c]
ꢀ2220.3480[c]
ꢀ2220.3378[c]
17a
17b
18a
18b
ꢀ1485.6210[c]
ꢀ1485.6206[c]
ꢀ2220.3535[c]
ꢀ2220.3434[c]
[a] Energy relative to reactants. [b] RB3LYP. [c] UB3LYP.
2711
Chem. Eur. J. 2004, 10, 2707 2721
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

F.-G. Kl‰rner et al.
FULL PAPER
Figure 3. Calculated structures of the hydrogen transfer reaction from
1,4-cyclohexadiene (3) to DDQ (1a): [1] (U)B3LYP/6 31G*,
[2] (U)B3LYP/6 31G**.
Figure 4. Calculated structures of the hydrogen transfer reaction from
1,4-cyclohexadiene (3) to o-chloranil (10): [1] (U)B3LYP/6 31G*.
[2] (U)B3LYP/6 31G**.
DFT calculations. Further support for the closed-shell char-
acter of the transition structures is provided by the differen-
ces between the singlet and triplet biradical energies. These
range from 7 to about 20 kcalmolꢀ1 (Table 2), depending on
basis set and the quinone under consideration. Such big dif-
ferences are a measure of the closed-shell character of the
systems.[26] Products of the hydrogen transfer are radicals
(16, 17 (Figure 3), and 18 (Figure 4)). Radicals 17 and 18
may exist in two conformations with respect to the orienta-
tion of the OH bond (a and b in the formulas). While the
enthalpy difference between 17a and 17b is marginally in
favor of 17a (0.1 kcalmolꢀ1 (6 31G*) and 0.3 kcalmolꢀ1 (6
31G**)), the difference between 18a and 18b amounts to
6.4 kcalmolꢀ1 (6 31G*) or 6.3 kcalmolꢀ1 (6 31G**) in favor
of 18a. This is due to a strong intramolecular OH¥¥¥O hydro-
gen bond. If the enthalpy of reaction is calculated for 1+
3!16+17, values of 9.2 or 9.3 kcalmolꢀ1 (6 31G*) or 6.5 or
6.8 kcalmolꢀ1 (6 31G**) are obtained. The same calculation
on the basis of the more stable structure 18a for the reac-
tion 1+10!16+18a gives an enthalpy of reaction of
1.8 kcalmolꢀ1 (6 31G*) or ꢀ0.8 kcalmolꢀ1 (6 31G**). On
the basis of the less stable structure, these values have to be
corrected by about +6 kcalmolꢀ1. The reactions are there-
fore slightly endothermic or thermoneutral, with activation
enthalpies of slightly above 10 kcalmolꢀ1. Final product for-
mation, consisting of the disproportionation of the radicals
to benzene and the corresponding hydroquinones, is a
strongly exothermic process. A 6 31G* calculation, per-
formed for the disproportionation of 16 and 17, provides
ꢀ45.6 kcalmolꢀ1. No TS could be located for this process.
Discussion
The oxidation of the hydroarenes 3 9 with DDQ (1a) or o-
chloranil (10) is accelerated by pressure, indicating negative
activation volumes (Table 1). Similar results have been ob-
tained by Brower et al. for the dehydrogenation of 1,4-cyclo-
hexadiene (3) and tetralin (9) by thymoquinone (1c) (DV°
= ꢀ33 (758C)[9] and ꢀ28 cm3 molꢀ1 (1758C)[27], respective-
ly). The negative volumes of activation indicate an associa-
tion in the rate-determining steps of these oxidations. The
values of DV° determined in this work, however, are only
moderately dependent on the solvent polarity and even
slightly more negative in the more polar medium (MeCN/
AcOEt) than in MTBE, contrary to expectations for the for-
mation of charged species. These observations are good evi-
dence against the mechanisms of direct hydride transfer and
single-electron transfer in the rate-determining step
(Scheme 1 mechanisms (1) and (3)) and are further strong
2712
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2707 2721

Hydrogen Transfer Reactions
2707 2721
support for the hydrogen atom
transfer (mechanism (2)). The
complete volume profile shown
for the reaction of 1a with 3 in
Figure 2 indicates that the partial
molar volume of the transition
state (calculated from the partial
molar volumes of the reactants
and the activation volume within
the scope of the Eyring transition
state theory) is the same within
1 cm3 molꢀ1 in both solvents, sug-
gesting that there is no significant
effect of electrostriction in the
rate-determining transition state.
Experimentally
determined
values of activation enthalpies
for hydrogen transfer are report-
ed in Table 1 and may be com-
pared with the DFT results
shown in Table 2. The DH° value
Scheme 3. a) Formation of the electron transfer complex of N-phenyldihydronicotinamide (19) with p-chlora-
nil (1b), and b) the products of the hydrogen transfer reaction.
of 11.7 kcalmolꢀ1 for the reaction of DDQ (1a) with 1,4-cy-
clohexadiene (3) is almost identical to the 6 31G*-calculat-
ed value for the compact (1a+3)TS2; the 6 31G** result
for the same structure is 9.9 kcalmolꢀ1. A compact structure
is also suggested by the measurement of the highly negative
activation volumes. The calculated activation enthalpy via
the stretched transition structure is higher and deviates from
the experimentally ascertained value more strongly. There
are no experimental data for the reaction of 3 with o-chlora-
nil (10). However, results for the reaction of 9,10-dihydroan-
thracene (5) are presented in Table 1. Compound 5 shows
the same activation energy as 3 in the reaction with DDQ.
Thus, it may be assumed that the dehydrogenation of 3 by
10 should provide an identical value to the reaction of 5
with 10. The conclusion is that 10 is less reactive than DDQ.
Indeed, the DFT calculations yielded a higher DH° for the
hydrogen transfer to o-chloranil, comparable to the experi-
mentally established difference in activation energy for the
reaction of 5. The DFT results are compatible with the acti-
vation volume measurements (i.e., with a hydrogen atom
transfer mechanism). There is a polar component in the
mechanism, for which the increase in dipole moment in the
TS of the reaction of DDQ with 3 is evidence. As this is
much less pronounced for the reaction with o-chloranil, it
may be concluded that the actual mechanism might also be
dependent on the specific system. Therefore, systems in
which a mechanism other than hydrogen atom transfer may
be followed might exist. In the quinone oxidation of N-
alkyl- or N-aryldihydronicotinamides with p-chloranil (1b),
for example (Scheme 3), an intermediate–most likely the
complex resulting from an electron transfer–could be de-
tected in the UV/Vis spectrum of the mixture of N-phenyl-
dihydronicotinamide (19) with p-chloranil (1b) in acetoni-
trile (lmax = 448 nm).[28] At 378C this complex disappears
within 60 min and 20 and 21 are formed as final products, as
described in the Experimental Section.
with 1,4-cyclohexadiene (3) and that the activation parame-
ters of both reactions at 1 bar are the same rule out the ene-
retro-ene mechanism (Scheme 1, mechanism 4)). In the hy-
pothetical ene reaction of 1a with 5, the aromaticity of one
of the benzene rings in 5 has to be given up. Therefore, this
reaction should be retarded and its activation barrier should
be substantially higher than that of the corresponding reac-
tion of 1a with 3; comparable, for example, to the electrocy-
clic ring-closures of (E,Z,E)-2,4,6-octatriene[29] and the cor-
responding 4,5-benzo-anellated derivative,[30] which differ in
the temperature of reaction by 1008C and in the enthalpy of
activation by DDH° = 3.7 kcalmolꢀ1.
The large primary kinetic isotope effect observed for
DDQ oxidation of the hydroarenes 5 and 6 suggested an in-
vestigation of its pressure dependence. For the reaction of
9,10-dihydroanthracene 5/5-9,9,10,10-D4 with DDQ (1a) at
258C and 1 bar in MTBE the kinetic isotope effect was de-
termined to be kH/kD = 10.8, decreasing to kH/kD = 5.0 at
3000 bar and 258C (Figure 5). Accordingly, the volume of
Figure 5. Effect of pressure on the kinetic isotope effect (kH/kD) of hydro-
gen transfer reactions: a) reaction of leuco crystal violet with p-chloranil
in acetonitrile at 26.58C,[31] b) reaction of DDQ (1a) with 5 in MTBE at
258C, c) proton transfer from benzoic acid to diphenyldiazomethane at
26.58C.[31]
The findings that the reaction of DDQ (1a) with 9,10-di-
hydroanthracene (5) is not significantly slower than that
2713
Chem. Eur. J. 2004, 10, 2707 2721
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

F.-G. Kl‰rner et al.
FULL PAPER
activation of the oxidation of 5-9,9,10,10-D4 with 1a was cal-
culated from the pressure dependence of the rate constants
(Table 3) to be DV° = ꢀ38.6ꢁ0.5 cm3 molꢀ1, substantially
that also in these reactions no ionic intermediates are
formed in the rate-determining step and that they proceed
through hydrogen atom transfer, comparably to the DDQ
oxidations. In the oxidation of
5-9,9-Me2 and 6 with quinones
1a or 10, the aromatization of
the intermediary radicals 22
and 23 by H atom transfer is
blocked by the methyl groups
or the ring oxygen atom
(Scheme 4). The finding of 9,9-
dimethylanthrone 26 and xan-
thone 27 as final products in
the presence of traces of water
suggests that the cations 24
and 25 are involved as inter-
mediates in these reactions. In
the reactions of 5-9,9-Me2 with
1a and of 6 with 10, spin-trap-
ping experiments with nitroso-
Table 3. Pressure-dependent second-order rate constants k2 (in 10ꢀ2 mꢀ1 sꢀ1) for the reaction of 1a with various
hydrogen donors in tert-butyl methyl ether.[a]
P
3
5
5 9,9,10,10-D4
5 9,9-Me2
4
7
8
[kbar]
k2
k2
k2
k2
k2
k2
k2
1
500
3.28
5.37
8.2
3.81
6.73
10.7
15.5
23.0
31.4
43.1
58
0.35
0.80
1.52
2.53
4.05
6.06
8.63
0.561
0.76
1.10
1.44
1.88
2.29
2.63
15.0
0.253
0.0435
25.2
37.5
58.2
82.6
113
144
195
0.417
0.627
0.988
1.51
2.20
3.15
0.0735
0.112
0.157
0.224
0.329
0.425
1000
1500
2000
2500
3000
3500
T[8C]
12.1
18.8
27.5
37.7
49.6
25.1
25.1
25.0
39.9
25.0
39.9
39.9
[a] Initial concentrations: [1a] = 1.15 1.17î10ꢀ3 m; [3] = 1.87î10ꢀ2 m; [5] = 3.03î10ꢀ2 m; [5 9,9,10,10-D4] =
3.04î10ꢀ2 m; [5 9,9-Me2] = 1.01î10ꢀ1 m; [4] = 1.45î10ꢀ2 m; [7] = 1.04î10ꢀ1 m; [8] = 2.04î10ꢀ1 m.
more negative than that of undeuterated
5
(DV°
=
benzene produced the long-living nitroxide radicals 13 and
14, which could be characterized by ESR (Experimental
Section). These results are further evidence that hydrogen
atom transfer also occurs in the rate-determining steps in
these reactions, producing the radicals 22 and 23, which can
ꢀ25.9 cm3 molꢀ1). Similar results were obtained by Isaacs
et al.[31] for the oxidation of leuco crystal violet with p-chlor-
anil (1b), in which the kinetic isotope effect decreases from
kH/kD = 11 at 1 bar to nearly 8 at 2000 bar (DV° = ꢀ25
(H) and ꢀ35 cm3 molꢀ1 (D)).
Contrary to these findings, the
kinetic isotope effect on the re-
action of benzoic acid and di-
phenyldiazomethane (proceed-
ing by
a
rate-determining
proton transfer) is almost pres-
sure-independent (kH/kD = 4.5
(1 bar) and 4.7 (1090 bar) and
the activation volumes for the
two isotopic species are essen-
tially
identical
(DV°
=
ꢀ13 cm3 molꢀ1).[31] This finding
indicates that the ordinary ki-
netic isotope effect is not signif-
icantly pressure-dependent. The
strong pressure dependence of
the kinetic isotope effect ob-
served for the quinone oxida-
tions of 9,10-dihydroanthracene
5/5-9,9,10,10-D4 and leuco crys-
tal violet can be attributed to a
tunneling component in the hy-
Scheme 4. The oxidation of 5-9,9-Me2 with 1a and 6 with 10, in which the aromatization by H transfer in the
secondary reaction step is blocked, with spin trapping of the intermediary radicals 15 and 16 by nitrosoben-
zene.
drogen transfer and can be used as a criterion for tunneling,
in addition to the known kinetic indicators such as large ki-
netic isotope effects[32] and nonlinear Arrhenius plots[33]
(large differences in the preexponential factors and activa-
tion energies for H vs. D transfer).
either be trapped by nitrosobenzene or, in the absence of a
trapping reagent, can react further with excesses of the qui-
nones to produce the cations 24 or 25, which are converted
into the final products 26 or 27 by water addition and qui-
none oxidation. Accordingly, oxidations of both hydroarenes
either with DDQ (1a) or o-chloranil (10) follow the same
mechanistic course. The less negative activation volumes of
the o-chloranil oxidations may be explained by quantum
mechanical calculations performed for the reactions of 1,4-
cyclohexadiene (3) with either DDQ (1a) of o-chloranil
(10). Accordingly, the dipole moment of the transition state
The activation volumes (DV°) of the reactions of 9,10-di-
hydroanthracene (5) and xanthene (6) with o-chloranil (10)
are substantially less negative than those of the correspond-
ing reactions with DDQ (1a) (Table 1). The tendency of the
solvent dependence of DV° of the reactions of 10, however,
is similar to that observed for the reactions of 1a, indicating
2714
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2707 2721

Hydrogen Transfer Reactions
2707 2721
of the reaction of 3 with 10 is calculated (Table 2)–with m
= 4.7 D and 4.9 D, respectively–to be substantially smaller
than that of the reaction of 3 with 1a (m = 10.6 D and
8.5 D, respectively). Thus, the contraction of volume caused
by the effects of electrostriction, and hence the activation
volume, would be expected to be smaller in the reaction of
3 with 10.
results already rule out the pericyclic hydrogen transfer as
the only mechanism. Competition between a stereospecific
pericyclic reaction and a non-stereoselective stepwise reac-
tion is one mechanistic possibility. However, the observed
stereoselectivity could be also explained if the subsequent
hydrogen transfer were to proceed rapidly in the pair of rad-
icals (generated in the first step) in relation to its dissocia-
tion from the solvent cage. Also, the kinetic isotope effects
observed for the oxidation of the deuterated 1,2-dihydro-
naphthalene derivatives with 1a and 10 (Scheme 5) do not
allow clear-cut conclusions concerning the mechanism of
these reactions. The finding that the reactions of both deriv-
atives 7-1,1-D2 and 7-2,2-D2 show primary isotope effects
and that the effects observed for the reactions of 7-1,1,2,2-
D4 are larger than those found with 7-1,1-D2 or 7-2,2-D2 is
Finally, in the case of the oxidation of the formal dienes 7,
8 and ene 9, a pericyclic hydrogen transfer has to be consid-
ered as a potential mechanism in addition to the possibility
of the hydrogen atom transfer, as already discussed in the
Introduction (Scheme 1, mechanisms (2) and (4)). Accord-
ing to the findings obtained for pericyclic and stepwise cy-
cloadditions (the latter involving diradical intermediates),
the activation volume of the pericyclic hydrogen transfer
would be expected to be more negative than that of a step-
wise hydrogen atom transfer. The activation volumes of the
DDQ oxidations of the formal 1,4-dienes 3 5 and the
formal 1,3-dienes 7, 8 and ene 9 do not show any significant
difference (Table 1). This finding suggests the conclusion
that all oxidations studied here follow the same mechanistic
course. Independent support for this conclusion comes from
an investigation of the kinetic isotope effect observed in the
DDQ oxidation of the tetralin derivatives 9/9-1,1,4,4-D4 and
9-D12. Heesing et al.[34] have carefully studied the stereo-
chemical course and the kinetic isotope effects for the oxi-
dation of 1,2-dihydronaphthalene (7) and 1,3-cyclohexa-
diene (28) with the deuterated derivatives shown in
Scheme 5 and various quinones such as 1a and 10. The oxi-
dation of cis- and trans-7-1,2-D2 and ꢀ28-5,6-D2 proceeds
with 90% and 98% cis stereoselectivity, respectively. These
ꢀ
in accord with a pericyclic reaction in which both C H
bonds are cleaved in the rate-determining step. The struc-
tures of two potential radical intermediates 29 and 30
(which can be obtained from 7 through H atom transfer) are
both resonance-stabilized, however, so the H atom transfer
from 7 may be not selective, yielding both radicals. This
would also explain why the kinetic isotope effects observed
for the oxidation of 7-1,1-D2 and 7-2,2-D2 with DDQ (1a)
and o-chloranil (10) are quite different from each other.
Evidently, the less reactive 10 is more selective, leading to
the radical 30 preferentially, whereas the more reactive 1a is
less selective, leading to the two radicals 29 and 30 in nearly
equal amounts (Scheme 6). The experimental data thus do
not allow an unambiguous conclusion as to whether the qui-
none oxidation of 1,2-dihydronaphthalene proceeds concert-
edly by a pericyclic hydrogen transfer or in stepwise manner
with a hydrogen atom transfer in the first step .
Scheme 6. Potential radical intermediates involved in the quinone oxida-
tion of 1,2-dihydronaphthalene 7 and tetralin 9.
In the case of the oxidation of tetralin 9 with thymoqui-
none 1c on one hand and with DDQ (1a) on the other
(leading to 1,2-dihydronaphthalene 7 as the primary isolable
product), a distinction between the two mechanistic alterna-
tives on the basis of the kinetic deuterium isotope effects
was feasible.[35] The hydrogen atom transfer would be ex-
pected here to lead exclusively to the benzyl resonance-sta-
1
ꢀ
bilized radical 31 through C H bond cleavage and not to
Scheme 5. Specifically deuterated 1,2-dihydronaphthalene, 1,3-cyclohexa-
diene, and tetralin derivatives used for investigation of the stereochemi-
cal course and the kinetic isotope effect of the quinone oxidation of
formal 1,3-dienes (n. d. = not determined).
2
ꢀ
the ordinary secondary alkyl radical 32 through C H bond
cleavage, whereas in the pericyclic hydrogen transfer both
adjacent C H bonds have to be cleaved. Brower et al.
[9]
ꢀ
2715
Chem. Eur. J. 2004, 10, 2707 2721
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

F.-G. Kl‰rner et al.
FULL PAPER
found primary deuterium isotope effects of kH/kD = 3.0 and
2.3 for the reaction of thymoquinone 1c with the two deu-
terated tetralin derivatives 9-1,1,4,4-D4 and 9-2,2,3,3-D4, re-
spectively, and a larger value of kH/kD = 5.7 for the reaction
of perdeuterated tetralin 9-D12, as would be expected for
diradical intermediate, so that in the case of the chloroprene
cyclodimerization the stepwise process is competitive with
the concerted pathways. The different reactivity and course
of reaction observed for DDQ (1a) and thymoquinone 1c
can be explained in terms of a substituent effect, analogous-
ly to the cyclodimerizations of 1,3-butadiene and chloro-
prene.
ꢀ
the cleavage of both C H bonds in the rate-determining
step. In contrast with these results, we found the deuterium
isotope effect of the reactions of DDQ (1a) with 9-1,1,4,4-
D4 and 9-D12 to be nearly identical, indicating that only the
1
ꢀ
C
H bond is broken in the rate-determining step in this re-
Conclusion
action. This surprising finding suggests that the reaction of
tetralin 9 with the less reactive thymoquinone 1c (which re-
quires a temperature of reaction 1008C higher than that
with DDQ (1a)) proceeds via a pericyclic hydrogen transfer,
whereas stepwise hydrogen atom transfer occurs in the reac-
tion with DDQ (1a) (Scheme 7). The different behavior of
The highly negative activation volumes and their moderate
dependence on solvent polarity observed for oxidation of
hydroarenes with DDQ (1a) and o-chloranil (10) are good
evidence for a hydrogen atom transfer mechanism, as also
suggested from kinetic measurements, studies of kinetic iso-
tope effects, and spin-trapping
experiments.[4,20] The strong
pressure dependence of the ki-
netic deuterium isotope effect
found for the reaction of 9,10-
dihydroanthracene 5/5-1,1,4,4-
D4 with DDQ (1a) can be used
as a criterion for tunneling. Fi-
nally, comparison of the kinetic
deuterium isotope effects deter-
mined for the oxidation of tet-
ralin 9/9-1,1,4,4-D4/9-2,2,3,3-D4/
9-D12 either with DDQ (1a) or
with thymoquinone 1c[9] indi-
cates that the reaction with
DDQ (1a) proceeds in stepwise
fashion through hydrogen atom
transfer in the rate-determining
first step, whereas the reaction
with thymoquinone 1c is con-
certed, following the course of
a pericyclic dyotropic hydrogen
transfer (Scheme 7). The differ-
ence in the mechanistic courses
of these two reactions may be
Scheme 7. Potential mechanisms of the dehydrogenation of tetralin 9 with a) thymoquinone 1c and b) DDQ
(1a), respectively, suggested from kinetic isotope effects.
explained by the effect of the
CN and Cl substituents in 1a,
which stabilize a radical inter-
thymoquinone 1c and DDQ (1a) may be explained by the
different stabilities of the radicals derived from 1c and 1a
after the hypothetical hydrogen atom transfer from 9 to 1c
or 1a. The radical derived from 1a should experience addi-
tional resonance stabilization by the CN and Cl,[36] substitu-
ents absent from the radical derived from 1c. This also ex-
plains, inter alia, the high reactivity of DDQ (1a) in relation
to all other quinones. A similar substituent effect is ob-
served for the cyclodimerizations of 1,3-butadiene and chlo-
roprene.[37] The [4 +2] cyclodimerization of 1,3-butadiene
proceeds at 1208C predominantly in a concerted fashion,
whereas in the cyclodimerization of chloroprene–occurring
readily at room temperature–a stepwise [4+2] cycloaddi-
tion involving a diradical intermediate competes with con-
certed pathways. Evidently, the Cl substituent stabilizes the
mediate better than the alkyl groups in 1c. DFT calculations
on the hydrogen transfer from 3 to DDQ (1a) and o-chlor-
anil (10) are in agreement with these conclusions.
Experimental Section
1
IR: Bio-Rad FTS 135. UV: J&M Tidas FG/Cosytec/RS 422. H NMR, 13C
NMR: Bruker DRX 500, Varian XL 200; the undeuterated proportion of
the solvent was used as an internal standard. MS: Fisons Instruments
VG ProSpec 3000. High-pressure equipment: up to 14 kbar: A. Hofer,
M¸lheim a. d. Ruhr; A. W. Birks, Department of Mechanical and Indus-
trial Engineering at the Queen×s University of Belfast; up to 4 kbar: Die-
ckers, Willich; Sitec round optical cell with sapphire windows, Z¸rich.
GC: Hewlett Packard 5890 Series II equipped with DB1 (25 m) and
OV 17 (30 m) capillary columns (d. 0.32 mm; film 0.25 mm, 2 mLminꢀ1
2716
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2707 2721

Hydrogen Transfer Reactions
2707 2721
He flow FID); preparative GC: Varian Aerograph P 90 with WLD
(Varian), column: 1.7 m, i. d. 1 cm, packed with OV17 on Volaspher A2,
carrier gas: He (90 mLminꢀ1), 2008C.
in 6 was not feasible because of hydrogen/deuterium exchange occurring
during measurement at 20 eV.
b) 6 9,9-D2: m.p. 1018C, yield: 73%. The degree of deuteration in 6-9,9-
D2 was determined in the same way as described for 5-9,9,10,10-D4: d =
4.05 (s, H2 9, I = 0.0336), 7.02 7.07 (m, H-1, 3, 6, 8, I = 1.0000), 7.17
7.23 ppm (m, H-2, 4, 5, 7, I = 0.9826). From these intensities the content
of deuterium at C-9 was calculated to be 96.6% (estimated error 2%).
Chemicals: 9,10-Dihydroanthracene (1a), xanthene (6), DDQ, and o-
chloranil (10) were obtained from Aldrich and purified by repeated crys-
tallization from acetone/water 1:1 for 1a, ethanol for 6, and glacial acetic
acid for DDQ and 10, respectively, and additional sublimation under re-
duced pressure. Samples used for kinetic measurements had melting
points in agreement with published values. 1,4-Cyclohexadiene (3), 1,2-di-
hydronaphthalene (7), 9,10-dihydrophenanthrene (8), and tetralin (9) (all
from Aldrich) were purified by distillation under reduced pressure and
stored under Argon. No significant impurities were detected by GC.
1,2,3,4-Tetrahydronaphthalene-1,1,4,4-[D4] (9 1,1,4,4-D4): This compound
was prepared by repetitive treatment of 9 with NaH in [D6]DMSO by
the method described by Bergman et al.[41] The product was purified by
distillation under reduced pressure and preparative GC (tR = 10 min).
The degree of deuteration in 9-1,1,4,4-D4 was determined in the same
way as described for 5-9,9,10,10-D4: d = 1.52 (s, H-2, 3, I = 1.9861), 2.56
(s, H-1, 4, I = 0.0670), 6.94 6.96 (m, H-5, 8, I = 1.000), 7.01 7.03 ppm
(m, H-6, 7, I = 1.0067). From these intensities the content of deuterium
at C-1, 4 was calculated to be 96.7% (estimated error 2%).
Synthesis
1,4-Dihydronaphthalene (4): Naphthalene was reduced with sodium in
ethanol by the procedure described by Pajak and Brower.[9] The crude
product (14.0 g), consisting of a mixture of naphthalene, 1,4-dihydronaph-
thalene (4), and 1,2-dihydronaphthalene (7) (12.5%:82.3%:4.8% GC),
was dissolved in diethyl ether (130 mL) and added dropwise to a solution
of mercury(ii) acetate (34.2 g) and 25 drops of glacial acetic acid in water
(130 mL). After the mixture had been stirred for 24 h, the precipitated
addition product was filtered off. The residue was washed with water
(400 mL) and ether (250 mL). A stirred suspension of the product in pe-
troleum ether (200 mL) was heated under reflux for four hours. After fil-
tration the product was dried under reduced pressure to give 25.5 g. To
recover 4, hydrochloric acid (30%, 350 mL) was added to the stirred mix-
ture of the addition product (25.5 g) in dichloromethane (400 mL), and
the mixture was stirred at room temperature for two hours. The precipi-
tated colorless solid was removed by filtration. The aqueous phase was
extracted with three portions (150 mL) of diethyl ether. The combined
organic layers were washed with saturated aqueous NaHCO3 (200 mL)
and water (200 mL) and dried (MgSO4). After evaporation of the sol-
vents in vacuo, the yellow solid was purified by column chromatography
(150 g silica gel, dichloromethane) and distillation in vacuo (25 mbar) to
give pure 4 (5.1 g, m.p. 268C, ref.[38] m.p. 26 278C). GC analysis: 98.5%
1,2,3,4-Tetrahydronaphthalene-1,1,2,2,3,3,4,4,5,6,7,8-[D12] (9-D12): This
compound was prepared according to the synthesis of 1,2,3,4-tetrahydro-
naphthalene from naphthalene as described by Blum et al.[42] A mixture
of RhCl3¥3H2O (0.1 g), Aliquat-336 (methyltridecylammonium chloride,
0.22 g), D2O (10 mL), and 1,2-dichloroethane (5 mL) was vigorously
shaken in a Paar hydrogenation apparatus at 308C under a deuterium
pressure of 1 bar for 60 min. After addition of a solution of [D8]naphtha-
lene (Aldrich, 5.0 g, 98% D) in 1,2-dichloroethane (10 mL), the mixture
was shaken for an additional 72 h under a deuterium pressure of 1 bar at
308C. The reaction mixture was extracted twice with 20 mL portions of
1,2-dichloroethane, and the combined organic layers were washed twice
with 50 mL portions of water and dried over MgSO4. After distillation
under reduced pressure, the product was further purified by preparative
GC (tR = 10 min). The degree of deuteration in 9-D12 was determined
from the intensities (i) of the 1H NMR signals of the deuterated 9-D12
(0.327m in [D6]benzene) relative to the intensities of the signals of the in-
ternal standard dioxane (0.520m). d = 1.49 (s, H-2, 3, I = 0.3270), 2.53
(s, H-1, 4, I = 0.3183), 3.35 (dioxane, I = 8.000), 6.95 (m, H-6, 7, I =
0.0511), 7.03 (m, H-6, 7, I = 0.1006). From these intensities the content
of deuterium at C-1, 4 was calculated to be 87.9%, at C-2, 3 87.6%, C-5,
8 95.8%, and at C-6, 7 91.7% (estimated error 2%).
1
of 4 and 1.5% of naphthalene. H NMR (200 MHz; CDCl3): d = 3.38 (s,
4H; H-1, H-4); 5.90 (s, 2H; H-2, H-3); 7.05 7.25 ppm (m, 4H; H-Ar).
9,9-Dimethylanthrone (26): This compound was obtained by methylation
of anthrone with methyl iodide as described by Herberg.[39] The crude
product was purified by column chromatography (silica gel, cyclohexane/
dichloromethane 2:1, yield: 12%). m.p. = 1018C; 1H NMR (200 MHz;
Kinetic measurements: The rate constants of the oxidation of the hydro-
aromatic compounds 3 9 with DDQ (1a) and o-chloranil (10) were de-
termined by UV/Vis spectroscopy under pseudo-first-order conditions
with at least 16-fold molar excess of the hydroaromatic compound. The
conversion of 1a to DDQ-H2 (2a) during the reaction with the hydroaro-
matic compounds could be monitored by the decreases in the UV/Vis ab-
sorptions of 1a and 10, respectively, at wavelengths of l = 400 nm and l
= 450 nm, at which 2a or 1,2,3,4-tetrachloro-5,6-dihydroxybenzene (11)
do not absorb light, as illustrated for 1a and 2a in Figure 6.
3
4
ꢀ
CDCl3): d = 1.74 (s, 6H; CH3), 7.42 (td, J = 7.5 Hz, J = 1.3 Hz, 2H;
H-3,6), 7.64 (td, 3J = 7.5 Hz, 4J = 1.3 Hz, 2H; H-2,7), 7.69 (dd, 3J =
7.7 Hz, J = 1.4 Hz, 2H; H-1,8), 8.36 ppm (dd, J = 7.9 Hz, J = 1.5 Hz,
2H; H-4,5); IR (KBr): n˜ = 1656 (s) (C=O), 1597 cmꢀ1 (s) (ArC=C).
4
3
4
9,9-Dimethyl-9,10-dihydroanthracene (7): Sublimed AlCl3 (8.2 g,
61.3 mmol) was dissolved at 08C in diethyl ether (20 mL). After the addi-
tion of LiAlH4 (1.2 g, 31.1 mmol), a solution of 26 (3.9 g, 17.5 mmol) in
diethyl ether (30 mL) was added, and the reaction mixture was heated
under reflux for 20 minutes. After cooling to room temperature, the
green colored reaction mixture was quenched with water. The water
layer was extracted three times with diethyl ether, the combined organic
layers were dried over MgSO4, and the solvent was removed in vacuo.
The crude product was purified by column chromatography (silica gel,
CCl4) and recrystallized twice from ethanol at 58C to give 7 (1.0 g, 27%).
1
ꢀ
m.p. 498C; H NMR (500 MHz; CDCl3): d = 1.64 (s, 6H; CH3), 4.11 (s,
4H; H-10), 7.21 7.56 ppm (m, 8H; H Ar); 13C NMR (126 MHz; CDCl3):
ꢀ
ꢀ
d = 28.8 ( CH3), 35.8 (C-10), 39.2 (C-9), 124.1 (C-1), 125.8, 126.4 (C-
2,3), 127.8 (C-4), 135.6 (C-4a), 144.9 ppm (C-8a).
9,10-Dihydroanthracene-9,9,10,10-[D4] (5-9,9,10,10-D4) and xanthene-9,9-
[D2] (6-9,9-D2): These were prepared by catalytic hydrogen/deuterium
exchange in [D1]ethanol/sodium ethoxide as described by Gerst and R¸-
chardt.[40]
Figure 6. UV/Vis spectra of DDQ (1a) and DDQ-H2 (2a) (each c = 6î
10ꢀ4 m) in tert-butyl methyl ether.
a) 5 9,9,10,10-D4: M.p. 1068C, yield: 97%. The degree of deuteration in
5 9,9,10,10-D4 was determined from the intensities (I) of the 1H NMR
signals of the deuterated 6 corrected by means of the intensities meas-
ured for the corresponding signals of the undeuterated 5 at d = 3.94
3.96 (m, H2 9, 10, I = 0.0359), 7.20 7.22 (m, H-1, 4, 5, 8, I = 1.0000),
7.29 7.32 ppm (m, H-2, 3, 6, 7, I = 0.9881). From these intensities the
content of deuterium at C-9, 10 was calculated to be 96.4% (estimated
error 2%). Exact mass spectroscopic analysis of the content of deuterium
Equation (1) was used to calculate pseudo-first-order rate constants kobs
and the bimolecular rate constant k2, from kobs and the initial concentra-
tion of the hydroarene, which can be assumed to be constant due to the
large excess of the hydroarene.
ln ðA=A0Þ ¼ kobst with kobs ¼ k2½H-donorŠ
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ð1Þ
2717
Chem. Eur. J. 2004, 10, 2707 2721
www.chemeurj.org

F.-G. Kl‰rner et al.
FULL PAPER
In the reaction of 1,4-cyclohexadiene (3) with DDQ (1a) the decrease in
the absorption A at three different wavelengths (400, 415, and 430 nm)
with time was used to calculate the pseudo-first-order rate constants, by
use of Equation (1), as kobs
=
(1.548ꢁ0.005) sꢀ1 at 400 nm, (1.540ꢁ
0.004) sꢀ1 at 415 nm, and (1.544ꢁ0.004) sꢀ1 at 430 nm ([3]
= 1.87î
10ꢀ2 m, [1a] = 1.16î10ꢀ3 m), which agree within limits of experimental
error. All other rate constants of the reactions of 3 9 with 1a were calcu-
lated only from the decrease in the absorption at 430 nm and those with
10 at 437 nm.
The reaction order was determined by measurement of kobs at various H-
donor concentrations. From a plot of log kobs versus log [H-donor] ac-
cording to Equation (2), a slope of one indicates that the reaction order
is one in terms of the H-donor following second order kinetics, as shown
for the reaction of xanthene (6) with 10 as example (Figure 7).
log kobs ¼ log k2 þ xlog ½H-donorŠ
ð2Þ
Figure 8. Schematic view of the high-pressure optical cell with pressure
vessel: 1) high-pressure optical cell, 2) volume 2.2 mL, optical path
14 mm, 3) lid, 4) thermocouple, 5) sapphire windows, optical width
=
6 mm, 6) heating system filled with water, 7) quartz lenses, 8) optical
fiber link to the UV/Vis spectrometer, 9) heating insulation, 10) enclosed
volume, 11) antiextrusion ring (Cu/Be alloy), 12) O-ring, 13) nut,
14) PTFE seal, 15) coupling for the PTFE tube, 16) PTFE tube with reac-
tion mixture, 17) hydraulic liquid: isooctane/decalin (1:1), 18) pressure
vessel (pmax = 7 kbar, volume = 25 mL), 19) heating system, filled with
silicon oil, 20) thermocouple, and 21) connection to the motor-driven
press, which keeps the pressure constant (ꢁ1 bar) during the experiment.
Figure 7. Plot of log kobs versus log [6] for the reaction of xanthene (6)
with o-chloranil (10). The slope was determined to be x = 1.03ꢁ0.03.
The pressure dependence of the rate constants was measured in a high-
pressure optical cell (pmax = 4 kbar, SITEC) which was connected to a
4 kbar vessel, heated by an external oil bath thermostated to ꢁ0.28C,
and pressurized by a motor-driven press (Dieckers). To follow the time
dependence of the reactions of the quinones 1a and 10 with one of the
hydrocarbons 3 9 and of 1b with N-phenyldihydronicotinamide (19) at
elevated pressures and temperatures, pre-thermostated solutions of the
quinone and the hydrocarbon were mixed and immediately transferred
into the high-pressure optical cell, which was connected to a diode-array
UV/Vis spectrometer by a fiber optic link. With this equipment, the sche-
matic cross section of which is depicted in Figure 8, the decrease in the
quinone concentration could be monitored on-line continuously, as illus-
trated in Figure 9 for the reaction of 3 with 1a in tert-butyl methyl ether
at 25.18C and 500 bar.
For the reaction of 1b with 19 at 37.08C, the time-dependent decrease in
the absorption of the electron transfer complex at 448 nm is depicted in
Figure 10. The first spectrum was obtained one minute after mixing of
the reaction solutions of 1b (c = 1.05î10ꢀ3 m) and 19 (c = 2.21î10ꢀ2 m)
in acetonitrile in the high-pressure optical cell. From the time depen-
dence of the decrease in the UV/Vis absorption at 378C and 1 bar, the
pseudo-first-order rate constant was ascertained to be kobs = (3.9ꢁ0.2)î
10ꢀ2 sꢀ1. Because of the high sensitivity of the reaction to the presence of
small amounts of oxygen in the high-pressure optical cell it was not possi-
ble also to determine the pressure dependence of the pseudo-first-order
rate constants of the oxidation reaction with the required accuracy to
obtain the volume of activation of the reaction.
Figure 9. Array of 290 UV/Vis spectra of the reaction of DDQ (1a) (c =
1.15î10ꢀ3 m) with 1,4-cyclohexadiene (3; c = 1.89î10ꢀ2 m) in tert-butyl
methyl ether at 25.18C and 500 bar.
Since the slope of the plot of ln k2 against pressure p is not linear in the
1 3500 bar range investigated here (Figure 11), the volumes of activation
(DV°) were determined from the pressure dependence of the rate con-
stants k2 at a constant temperature T by use of the quadratic Equa-
tion (3).
For kinetic measurements at atmospheric pressure,
a thermostated
ln kp ¼ a þ bp þ cp2 and DV° ¼ DV°0 ¼ ꢀbRT
ð3Þ
tandem quartz cell, which allowed thermostating of the separate solutions
of the reactants, was used. The tandem quartz cell was connected to a
diode-array UV/Vis spectrometer by a fiber optic link. The temperature
of the reaction solution, both in the high-pressure cell and in the tandem
quartz cell, was checked by a calibrated resistance thermometer PT-100.
Results are listed in Table 3, Table 4, Table 5, and Table 6
ESR spin-trapping: For the spin-trapping experiments, solutions of H-
donor, quinone, and nitrosobenzene in benzene were prepared (2.0î
2718
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2707 2721

Hydrogen Transfer Reactions
2707 2721
Table 5. Pressure-dependent second-order rate constants k2 (in
10ꢀ2 Mꢀ1 sꢀ1) for the reactions of 10 with 5 and 6 in tert-butyl methyl
ether (MTBE) and acetonitrile (MeCN).[a]
p
5MTBE
5MeCN
6MTBE
6MeCN
[kbar]
k2
k2
k2
k2
1
500
0.67
0.98
1.26
1.69
2.09
2.58
0.40
0.52
0.76
1.01
1.39
1.80
2.60
2.36
3.17
3.94
4.86
6.32
7.37
9.40
10.9
25.2
1.87
2.47
3.53
4.31
5.95
7.42
9.84
11.8
25.2
1000
1500
2000
2500
3000
3500
T [8C]
3.22
25.1
25.1
[a] Initial concentrations: [10] = 1.05î10ꢀ3 m; [5] = 9.73î10ꢀ2 m; [6] =
2.21î10ꢀ2 m.
Figure 10. Array of UV/Vis spectra of the reaction of N-phenyldihydroni-
cotinamide (19; c = 2.21î10ꢀ2 m) with p-chloranil (1b; c = 1.05î10ꢀ3 m)
in acetonitrile at 37.08C.
at room temperature immediately
after mixing. Without the addition of
the quinones, but under otherwise
identical conditions, no ESR spectrum
was observed. For results see Table 7
and Figure 12.
Molar volumes: For each substance,
the densities of solutions of seven dif-
ferent concentrations were deter-
mined. Measurements were performed
in
a temperature range from 20 to
508C in steps of 58C. For each temper-
ature, the value of fV (apparent molar
volume of a substance at a given con-
centrations) was calculated by using
Equation (4), with c [molLꢀ1] = con-
centration of the solution, M [gmolꢀ1
]
]
= molar mass of the solute, d [gcmꢀ3
=
density of the solution, and d0
[gcmꢀ3] = density of the pure solvent.
M
d0
1000 dꢀd0
ð4Þ
ꢀ ¼
ꢀ
c
d0
The partial molar volume (V) is calcu-
lated by linear extrapolation of fV to a
concentration c = 0 [Eq. (5)].
Figure 11. Pressure dependence of the rate constants k2 of the oxidation of the hydroarenes 3 9 with 1a in
tert-butyl methyl ether (left) and in MeCN/AcOEt (right).
V ¼ lim ꢀv
ð5Þ
c!0
10ꢀ1 m 5 9,9-Me2, 7.8î10ꢀ2 m 1a, and 7.8î10ꢀ2 m nitrosobenzene for trap-
ping of the 9,9-dimethyl-9-hydroanthranyl radical (22), and 1.9î10ꢀ1 m 6,
7.8î10ꢀ1 m 10, and 7.8î10ꢀ1 m nitrosobenzene for trapping of the xanthyl
radical (23)). Equal aliquots were mixed and degassed by use of several
freeze-pump-thaw cycles. The spectra of the spin adducts were recorded
The van der Waals volumes VW of 1a and 3 were calculated by use of the
MOLVOL[22] computer program with the van der Waals radii of hydrogen
(1.17 ä), carbon (1.8 ä), nitrogen (1.8 ä), oxygen (1.53 ä), and chlorine
(1.8 ä).
The packing coefficients of the start-
Table 4. Pressure-dependent second-order rate constants k2 (in 10ꢀ2 mꢀ1 sꢀ1) for the reaction of 1a with various
ing and the transition structures, h and
h°, were calculated by using Equa-
hydrogen donors in MeCN/AcOEt.[a]
tions (6) and (7).
P
3
5
5 9,9,Me2
10
7
8
9[b]
[kbar]
k2
k2
k2
k2
k2
k2
k2
h ¼ Vw=V
h° ¼ V°W
ð6Þ
ð7Þ
1
500
3.76
2.52
4.86
7.21
0.293
0.428
0.61
0.80
1.00
1.12
1.21
6.25
0.428
0.053
0.101
6.6
12.4
18.6
25.7
36.4
52.6
68.6
25.1
10.7
19.4
27.3
43.0
55.9
80.9
0.732
1.19
1.55
2.16
2.84
3.26
0.096
0.164
0.239
0.364
0.529
0.773
0.169
0.264
0.371
0.533
0.711
1.05
1000
1500
2000
2500
3000
3500
T [8C]
11.6
16.2
25.9
36.2
1.32
64.9
Acknowledgement
25.1
39.9
25.0
39.9
39.9
[a] Initial concentrations: [1a] = 1.15 1.17î10ꢀ3 m; [3] = 1.87î10ꢀ2 m; [5] = 3.03î10ꢀ2 m; [5 9,9-Me2] =
This work was supported by the Deut-
sche Forschungsgemeinschaft and the
Fonds der Chemischen Industrie.
1.01î10ꢀ1 m; [4] = 1.43î10ꢀ2 m; [7] = 1.04î10ꢀ1 m; [8] = 2.04î10ꢀ1 m; [9] = 4.59î10ꢀ1 m. [b] In pure acetoni-
trile.
2719
Chem. Eur. J. 2004, 10, 2707 2721
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

F.-G. Kl‰rner et al.
FULL PAPER
Table 6. Temperature-dependent second-order rate constants k2 (in
10ꢀ2 mꢀ1 sꢀ1) for the reactions of 10 with 5 and 6 in tert-butyl methyl
ether.[a]
Rogers, J. Am. Chem. Soc. 1991, 113, 7761 7762; W. Grimme, K.
Pohl, J. Wortmann, D. Frowein, Liebigs Ann. Chem. 1996, 1905
1916.
[9] J. Pajak, K. R. Brower, J. Org. Chem. 1985, 50, 2210 2216.
[10] J. March, M. B. Smith, March×s Advanced Organic Chemistry,
5th ed., Wiley, New York, 2001, Chapter 19.
T [8C]
5 k2
6 k2
1.44
2.38
4.44
7.32
20.2
25.9
35.7
43.4
49.4
0.427
0.731
1.44
2.45
4.06
[11] The rate of the dehydrogenation of 9,10-dihydroanthracene 5 with
DDQ (1a) is enhanced by the change of the solvent from dioxane
(ET
=
36.0 kcalmolꢀ1
,
e
=
2.9 D) to acetonitrile (ET
=
45.6 kcalmolꢀ1, e = 3.9 D) only by a factor of 3.6 (Ref. [4]).
10.7
[a] Initial concentrations: [10] = 1.05î10ꢀ3 m; [5] = 9.72î10ꢀ2 m; and
[6] = 2.33î10ꢀ2 m.
[12] K. U. Ingold, Pure Appl. Chem. 1997, 69, 241 244; C. Reichardt,
Solvents and Solvent Effects in Organic Chemistry, Wiley-VCH,
Weinheim, 2003, chapter 5.3.4.
[13] H. Mayr, M. Patz, Angew. Chem. 1994, 106, 990 1010; Angew.
Chem. Int. Ed. Engl. 1994, 33, 934 958.
[14] J. M. Mayer, D. A. Hrovat, J. L. Thomas, W. T. Borden, J. Am.
Chem. Soc. 2002, 124, 11142 11147.
Table 7. Comparison of the coupling constants a of the spin adducts 12,
13, and 14.
Coupling
14 [G]
12 [G][43]
13 [G][43]
[15] Monographs: R. van Eldik, F.-G. Kl‰rner, High Pressure Chemistry,
Wiley-VCH, Weinheim, 2002; W. B. Holzapfel, N. S. Isaacs, High-
Pressure Techniques in Chemistry and Physics, Oxford University
Press, New York, 1997; R. van Eldik, C. D. Hubbard, Chemistry
under Extreme and Non-Classic Conditions, Wiley, New York, 1997;
K. Matsumoto, R. M. Acheson, Organic Synthesis at High Pressure,
Wiley, New York, 1991; W. J. le Noble, Organic High Pressure
Chemistry, Elsevier, Amsterdam, 1988; N. S. Isaacs, Liquid Phase
High Pressure Chemistry, Wiley, Chichester, 1981; Reviews: F.
Wurche, F.-G. Kl‰rner in High Pressure Chemistry (Eds.: R. van El-
dik, F.-G. Kl‰rner), Wiley-VCH, Weinheim, 2002, 41 96; F.-G.
Kl‰rner, F. Wurche, J. Prakt. Chem. 2000, 342, 609 636; A. Drljaca,
C. D. Hubbard, R. van Eldik, T. Asano, M. V. Basilevsky, W. J.
le Noble, Chem. Rev. 1998, 98, 2167 2289; K. Matsumoto, M.
Kaneko, H. Katsura, N. Nayashi, T. Uchida, R. M. Acheson, Hetero-
cycles 1998, 47, 1135 1177; F.-G. Kl‰rner, Chem. Unserer Zeit 1989,
23, 53 63; R. van Eldik, T. Asano, W. J. le Noble, Chem. Rev. 1989,
89, 549 688; W. J. le Noble, Chem. Unserer Zeit 1983, 17, 152 162;
T. Asano, W. J. le Noble, Chem. Rev. 1978, 78, 407 489.
aN
ao-H
am-H
ap-H
ab-H
10.45
2.61
0.90
2.61
0.90
10.92
2.70
0.98
2.70
0.98
2.056
10.91
2.66
0.95
2.66
0.95
g factor
2.0056
2.0057
[16] T. Asano, W. J. le Noble, Rev. Phys. Chem. Jpn. 1973, 43, 82 91.
[17] W. J. le Noble, H. Kelm, Angew. Chem. 1980, 102, 841 856; Angew.
Chem. Int. Ed. Engl. 1980, 19, 841.
[18] Y. Kondo, M. Onishi, N. Tokura, Bull. Chem. Soc. Jpn. 1972, 45,
3579; Y. Kondo, M. Shinzawa, N. Tokura, Bull. Chem. Soc. Jpn.
1977, 50, 713.
[19] J. Jouanne, H. Kelm, R. Huisgen, J. Am. Chem. Soc. 1979, 101, 151.
[20] C. Hˆfler, Ph.D. Dissertation, Universit‰t Freiburg, 1997.
[21] R. Paukstat, M. Brock, A. Heesing, Chem. Ber. 1985, 118, 2579
2593.
Figure 12. ESR spectrum of the spin adduct of the 9,9-dimethyl-9-hydro-
anthranyl radical to nitrosobenzene. Sweep-width 80 Gauss; microwave
frequency 9.417 GHz; centerfield 3355 Gauss; modulation 0.08 Gauss.
[22] The Van der Waals volumes VW can be calculated by the computer
program MOLVOL: U. Artschwager-Perl, Ph.D. Dissertation, Ruhr-
Universit‰t Bochum, 1989. This program uses the Cartesian coordi-
nates of a molecular structure resulting from a force-field or quan-
tum mechanical calculation and can be obtained on request.
[23] Gaussian 98 (Revision A.7), M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski,
J. A. Montgomery Jr., R. E. Stratmann, J. C. Burant, S. Dapprich,
J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J.
Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli,
C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q.
Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. Cioslowski, J. V. Ortiz, B. B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin,
D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
C. Gonzalez, M. Challacombe, P. M. W. Gill, B. G. Johnson, W.
Chen, M. W. Wong, J. L. Andres, M. Head-Gordon, E. S. Replogle,
J. A. Pople, Gaussian, Inc., Pittsburgh, PA, 1998.
[1] E. A. Braude, R. P. Linstead, J. Chem. Soc. 1954, 3, 3544 3547;
E. A. Braude, L. M. Jackman, R. P. Linstead, J. Chem. Soc. 1954, 3,
3548 3574; J. R. Barnard, L. M. Jackman, J. Chem. Soc. 1960,
3110 3115; E. A. Braude, L. M. Jackman, R. P. Linstead, J. S. Shan-
non, J. Chem. Soc. 1960, 3116 3122; E. A. Braude, L. M. Jackman,
R. P. Linstead, G. Lowe, J. Chem. Soc. 1960, 3123 3138; A. M.
Creighton, L. M. Jackman, J. Chem. Soc. 1960, 3138 3147; L. M.
Jackman, Adv. Org. Chem. 1960, 329 366.
[2] H.-D. Becker, A. B. Turner, in The Chemistry of the Quinoid Com-
pounds, Vol. II (Eds.: S. Patai, Z. Rappoport), Wiley, New York,
1988, 1351 1384.
[3] L. Stryer, Biochemistry, W. H. Freeman, New York, 2002; D. Voet,
J. D. Voet, Biochemistry, Wiley, New York, 1995.
[4] C. Hˆfler, C. R¸chardt, Liebigs. Ann. Recl. 1996, 183 188.
[5] C. R¸chardt, M. Gerst, J. Ebenhoch, Angew. Chem. 1997, 109,
1474 1498; Angew. Chem. Int. Ed. Engl. 1997, 36, 1407 1430.
[6] Z. M. Hashish, I. M. Hoodless, Can. J. Chem. 1976, 54, 2261 2265.
[7] B. M. Jacobson, J. Am. Chem. Soc. 1980, 102, 886 887.
[8] L. A. Paquette, M. A. Kesselmayer, R. D. Rogers, J. Am. Chem.
Soc. 1990, 112, 284 291; L. A. Paquette, G. A. O×Doherty, R. D.
[24] M. Nishio, The CH/p Interaction, 1st ed., Wiley-VCH, New York,
1998.
[25] F. Jensen, Introduction to ComputationalChemistry , Wiley, New
York, 1999.
[26] R. Sustmann, W. Sicking, R. Huisgen, J. Am. Chem. Soc. 2003, 125,
14425 14434.
[27] K. R. Brower, J. Org. Chem. 1982, 47, 1889 1893.
[28] A. Sewing, Diploma Thesis, Universit‰t Freiburg, 1997.
2720
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2707 2721

Hydrogen Transfer Reactions
2707 2721
[29] E. N. Marvell, G. Caple, B. Schatz, W. Pippin, Tetrahedron 1973, 29,
3781.
[30] H. Heimgartner, H.-J. Hansen, H. Schmidt, Helv. Chim. Acta 1972,
55, 1385.
the size of the activation volume may also depend on the quinone
structure, as found for DDQ (1a) and o-chloranil (10).
[36] J. J. Brocks, H.-D. Beckhaus, A. L. J. Beckwith, C. R¸chardt, J. Org.
Chem. 1999, 64, 1935 1943.
[31] N. S. Isaacs, K. Javaid, E. Rannala, J. Chem. Soc. Perkin Trans. 2
1978, 709 711.
[37] F.-G. Kl‰rner, B. Krawczyk, V. Ruster, U. K. Deiters, J. Am. Chem.
Soc. 1994, 116, 7646 7657.
[32] L. Melander, W. H. Saunders, Reaction Rates of Isotopic Molecules,
Wiley, New York 1980.
[33] G. Brunton, D. Griller, L. R. C. Barclay, K. U. Ingold, J. Am. Chem.
Soc. 1976, 98, 6803 6811.
[38] R. Huisgen, G. Seidl, Chem. Ber. 1963, 96, 2730 2739.
[39] C. Herberg, H. D. Beckhaus, C. R¸chardt, Chem. Ber. 1994, 127,
2065 2072.
[40] M. Gerst, C. R¸chardt, Tetrahedron Lett. 1993, 34, 7733 7736.
[41] P. B. Comita, M. R. Berman, C. B. Moore, R. G. Bergman, J. Phys.
Chem. 1981, 85, 3266 3276.
[42] J. Blum, I. Amer, A. Zoran, Y. Sasson, Tetrahedron Lett. 1983, 24,
4139 4142.
[43] S. Terabe, R. Konaka, J. Chem. Soc. Perkin Trans. 2 1972, 2163
2172.
[34] M. Brock, H. Hintze, A. Heesing, Chem. Ber. 1986, 119, 3727 3736.
[35] The activation volumes of both reactions are highly negative (DV°
= ꢀ28 (reaction of 9 with 1c at 1758C) and ꢀ26.7 cm3 molꢀ1 (reac-
tion of 9 with 1a at 64.98C in MeCN/AcOEt), but the comparison
of these data does not provide further insight into the mechanism.
Since the two values were determined at quite different tempera-
tures, the temperature dependence of the activation volume cannot
be neglected here (See: B. S. El’yanov, E. M. Gonikberg, J. Chem.
Soc. Faraday Trans. 1 1979, 75, 172 191; B. S. El’Yanov, E. M. Vasyl-
vitskaya, Rev. Phys. Chem. Jpn. 1980, 50, 169 184). Furthermore,
Received: November 5, 2003
Published online: April 14, 2004
2721
Chem. Eur. J. 2004, 10, 2707 2721
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim