Investigation of the mechanism of the intramolecular Scholl reaction of contiguous phenylbenzenes

Article abstract of DOI:10.1021/jo0526744

Two mechanisms of the Scholl reaction were investigated in the series 1, 2, ..., n-oligophenylbenzenes (n = 2, 3, 4, 6) at the B3LYP/6-31G(d) level of theory. A mechanism involving generation of a radical cation followed by C-C bond formation and dehydrogenation is unlikely on the basis of unfavorable energies of activation. A mechanism involving generation of an arenium cation followed by C-C bond formation and dehydrogenation is energetically feasible. An explanation for the facile polycondensation of hexaphenylbenzene to hexa-peri-hexabenzocoronene, where six new aryl-aryl bonds are formed, is provided. Kinetic simulations based on the calculated activation energies of the arenium cation mechanism predict that intermediates will not accumulate; this is supported by mass balance experiments. Reaction optimization studies suggest that PhI(O₂CCF₃)₂/BF₃OEt₂ or MoCl₅ are superior to FeCl₃ or AlCl ₃/CuCl₂.

Full text of DOI:10.1021/jo0526744

Investigation of the Mechanism of the Intramolecular Scholl
Reaction of Contiguous Phenylbenzenes
Pawel Rempala, Ji ˇr ´ı Kroul ´ı k, and Benjamin T. King*
UniVersity of NeVada, Reno, Department of Chemistry, Reno, NeVada 89557
king@chem.unr.edu
ReceiVed December 30, 2005
Two mechanisms of the Scholl reaction were investigated in the series 1, 2, ..., n-oligophenylbenzenes
(n ) 2, 3, 4, 6) at the B3LYP/6-31G(d) level of theory. A mechanism involving generation of a radical
cation followed by C-C bond formation and dehydrogenation is unlikely on the basis of unfavorable
energies of activation. A mechanism involving generation of an arenium cation followed by C-C bond
formation and dehydrogenation is energetically feasible. An explanation for the facile polycondensation
of hexaphenylbenzene to hexa-peri-hexabenzocoronene, where six new aryl-aryl bonds are formed, is
provided. Kinetic simulations based on the calculated activation energies of the arenium cation mechanism
predict that intermediates will not accumulate; this is supported by mass balance experiments. Reaction
optimization studies suggest that PhI(O
This is a full account of our work reported partially as a communication previously (Rempala, P.; Kroul ´ı k,
2
CCF
3
)
2
/BF
3
2 5 3 3 2
‚OEt or MoCl are superior to FeCl or AlCl /CuCl .
J.; King, B. T. J. Am. Chem. Soc. 2004, 126, 15002-15003).
Introduction
works well for dendritic oligo- and polyphenylene compounds.
M u¨ llen and co-workers have used this reaction to convert
dendritic oligophenylenes to the corresponding symmetric
hydrocarbons forming as many as 126 new C-C bonds in a
Dehydrogenative aryl-aryl bond formation can be ac-
complished in many ways. Treatment with a Lewis acid (1 f
2
,3
4
5
2
), base (3 f 4a + 4b), or oxidant (5 f 6), irradiation in
1
2-15
6
single step.
In contrast to such reactions, the simple
the presence of oxidant (7 f 8), melting under vacuum (9 f
7
cyclization of o-terphenyl, which requires forming just one C-C
1
1
0), and heterogeneous catalysis at high temperature (11 f
8
-10
bond, was not successful in our hands using a variety of Scholl
2)
were used successfully (Scheme 1). These reactions are
1
6
conditions.
attractive because functionality is usually not required.
The Scholl reaction, understood as the condensation of
aromatics under the influence of a Lewis acid, often AlCl3,
Hexa-peri-hexabenzocoronene (HBC, 10) is synthesized in
excellent yields from hexaphenylbenzene (C6Ph6, 13, Scheme
11
1
7,18
2
a).
M u¨ llen and co-workers have shown that this reaction
(
1) Rempala, P.; Kroul ´ı k, J.; King, B. T. J. Am. Chem. Soc. 2004, 126,
5002-15003.
2) Scholl, R.; Mansfeld, J. Ber. Dtsch. Chem. Ges. 1910, 43, 1734-
746.
3) Balaban, A. T.; Nenitzescu, C. D. Dehydrogenation Condensation
1
1
(11) Li, J. J. Name Reactions: A Collection of Detailed Reaction
(
Mechanisms, 2nd ed.; Springer-Verlag: New York, 2003.
(12) Simpson, C. D.; Brand, J. D.; Berresheim, A. J.; Przybilla, L.; R a¨ der,
H. J.; M u¨ llen, K. Chem. Eur. J. 2002, 8, 1424-1429.
(13) Iyer, V. S.; Wehmeier, M.; Brand, J. D.; Keegstra, M. A.; M u¨ llen,
K. Angew. Chem., Int. Ed. Engl. 1997, 36, 1604-1607.
(14) Watson, M. D.; Fechtenk o¨ tter, A.; M u¨ llen, K. Chem. ReV. 2001,
101, 1267-1300.
(
of Aromatics (Scholl and Related Reactions). In Friedel-Crafts and Related
Reactions; Olah, G., Ed.; Wiley: New York, 1964; Vol. 2, pp 979-1047.
(
(
4) Bradley, W.; Sutcliffe, F. K. J. Chem. Soc. 1952, 1247-1251.
5) Pummerer, R.; Prell, E.; Rieche, A. Ber. Dtsch. Chem. Ges. 1926,
5
9, 2159-2161.
(15) Simpson, C. D.; Mattersteig, G.; Martin, K.; Gherghel, L.; Bauer,
R. E.; R a¨ der, H. J.; M u¨ llen, K. J. Am. Chem. Soc. 2004, 126, 3139-3147.
(16) Kroul ´ı k, J.; Robertson, C. R.; Hilton, C. L.; Korinek, J. D.; Gortari,
L. M.; Rempala, P.; King, B. T. Manuscript in preparation.
(17) Ito, S.; Wehmeier, M.; Brand, J. D.; K u¨ bel, C.; Epsch, R.; Rabe, J.
P.; M u¨ llen, K. Chem. Eur. J. 2000, 6, 4327-4342.
(
6) Mallory, F. B.; Wood, C. S.; Gordon, J. T.; Lindquist, L. C.; Savitz,
M. L. J. Am. Chem. Soc. 1962, 84, 4361-4362.
(
(
(
7) Clar, E.; Ironside, C. T.; Zander, M. J. Chem. Soc. 1959, 142-148.
8) Scola, D. A. J. Chem. Eng. Data 1964, 9, 405.
9) Copeland, P. G.; Dean, R. E.; McNeil, D. J. Chem. Soc. 1960, 1687-
1
689.
(18) K u¨ bel, C.; Eckhardt, K.; Enkelmann, V.; Wegner, G.; M u¨ llen, K.
J. Mater. Chem. 2000, 10, 879-886.
(10) Hansch, C.; Geiger, C. F. J. Org. Chem. 1958, 23, 477-478.
1
0.1021/jo0526744 CCC: $33.50 © 2006 American Chemical Society
Published on Web 06/16/2006
J. Org. Chem. 2006, 71, 5067-5081
5067

Rempala et al.
SCHEME 1. Examples of Dehydrogenation Coupling
Reactions of Aromatic Compoundsa
could be obtained along with HBC (Scheme 3b). While HBC
was unequivocally characterized, the partially cyclized inter-
mediates were not. These putative intermediates can be attributed
to low solubility of the reactants and intermediates in the polar
AlCl3/NaCl melt.
We aim to explain the ease of polycondensation, elusiveness
of intermediates, and lack of generality of the Scholl reaction
by examining its mechanism using computational and experi-
1
mental chemistry. Aside from our own communication, there
is apparently only one computational study of the Scholl
2
0
reaction.
3
In the Scholl reaction, oxidative aryl-aryl coupling reactions
are commonly effected by a transition metal halide, which can
act both as a Lewis acid and an oxidant, as developed by
2
1-23
Kovacic and co-workers.
arenium cation mechanism was proposed as early as 1935 and
A mechanism similar to the
24
2
5
3
refined by Nenitzescu and Balaban. In their review, the
arenium cation and radical cation
posed (Scheme 4).
3
,26
mechanisms were juxta-
In the arenium cation mechanism, an aromatic compound (18)
is protonated, and the resulting electrophilic σ complex (19)
reacts with another aromatic core to generate a new C-C bond.
Deprotonation gives intermediate 20, and dehydrogenation
restores aromaticity to afford the condensed product, 21. The
isolation of a dihydro analogue of 20 has been reported.27
Brønsted acids are implicated as catalysts in the Scholl
reaction. Traces of HCl or water accelerate the reaction,3
presumably by reaction with AlCl3 to generate H[AlCl3OH] and
HCl. Indeed, the reaction proceeds in anhydrous HF,2 which
is not an oxidizing acid. Of course, any acid, Lewis or Brønsted,
could form the key arenium intermediate. Our calculations
focused on Brønsted acids because they are likely to be present
in the reaction mixture and because they are computationally
tractable.
,24
8,3
In the second proposed mechanism, a radical cation, 22, is
first generated by one-electron oxidation of the starting material.
Its subsequent electrophilic C-C bond-forming reaction with
another aromatic group, deprotonation, and formal loss of H
atom from intermediate 23, follow.
a
Key: (a) ref 2, (b) ref 4, (c) ref 5, (d) ref 6, (e) ref 7, (f) refs 8-10.
is intramolecular. In one case, an intermediate lacking two C-C
bonds (14) was obtained in low yield when insufficient oxidant
was used (Scheme 2a).18 Intermediates with two and four new
CC bonds were also reported as present in the reaction mixture
based on FD-MS but were not isolated. An early report on
the preparation of HBC from C6Ph6 under Scholl conditions
The generation of aryl radical cations under oxidizing
conditions is precedented. When treated with Lewis acids or
with concentrated sulfuric acid, many PAHs give paramagnetic
1
9
3
solutions, and therefore, radical cations may be present in the
(
NaCl/AlCl3 melt, air) appeared in the patent literature and
Scholl reaction. This, however, does not necessarily mean they
are productive reactive intermediates. Phenols do probably
claimed that three incompletely cyclized structures (15, 16, 17)
SCHEME 2. Two18,19 HBC Syntheses
5068 J. Org. Chem., Vol. 71, No. 14, 2006

Intramolecular Scholl Reaction of Contiguous Phenylbenzenes
SCHEME 3. Proposed Mechanisms of the Scholl Reaction:
acid, Lewis or Brønsted, is required and the reaction is formally
an oxidation.
(a) Arenium Cation Mechanism; (b) Radical Cation
Mechanism
Many of their results agree well with our work on radical
cation mechanism. Specifically, the C-C bond formation of
radical cations is endothermic by ∼22 kcal/mol, its reverse
reaction is characterized by a very low energy barrier (2 kcal/
mol), the gas-phase formation of radical cations using CuCl2 is
strongly endergonic, and bonds are contiguously formed.20
Cases involving heterocycles, like the condensation of 1,2-
dipyrimidyl-3,4,5,6-tetra(4-tert-butylphenyl)benzene (24),30 re-
quire special consideration (Scheme 4). We propose that an
arenium mechanism also applies. A protonated or doubly
protonated pyrimidine ring could act as an electrophile, explain-
ing bond formation to the pyrimidine rings. The N-protonated
PAH core might be insufficiently electrophilic to attack the
adjacent phenyl ring. Protonation of the phenyl rings will be
difficult since the basic nitrogen sites will be protonated and
bear a positive charge. Even if protonation of the phenyl
substitutents was possible, the N-protonated PAH core would
be a poor nucleophile, inhibiting electrophilic attack. The
chelating substrate complicates the mechanistic analysis. Dif-
ferent reactants, FeCl3 or AlCl3/CuCl2, give different product
distributions, presumably because the metals coordinate differ-
ently. It is notable that 25 is not converted to 26 by treatment
with AlCl3/CuCl2, indicating that 25 is not an intermediate in
the AlCl3/CuCl2 induced conversion of 24 to 26.
SCHEME 4. Scholl Condensation on Heterocycles
Computational Details
Density functional calculations were performed using Gaussian
33 and Gaussian 98 programs. The B3LYP functional
together with standard 6-31G(d) basis set (Cartesian d functions)
was used. DFT calculations on radical ions are problematic;
1
32
33-35
0
36
3
7
(
30) Gregg, D. J.; Bothe, E.; H o¨ fer, P.; Passaniti, P.; Draper, S. M. Inorg.
Chem. 2005, 44, 5654-5660.
31) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
undergo coupling via a radial mechanism. The mechanism pro-
posed by Pummerer, dimerization of phenoxy radical, followed
by tautomerization,29 is widely accepted. This mechanism is,
however, unlikely in substrates that do not form long-lived
radicals. For both mechanisms (Scheme 3a,b), generalization
to intramolecular case, where R is linker, is easily envisioned.
Di Stefano et al.20 recently investigated the intramolecular
condensation of o-terphenyl or C6Ph6, coupled with reduction
of CuCl2, at the B3LYP/3-21G level of theory. The investigation,
which focused on a radical cation mechanism, was based on
two postulates: oxidants are required but Brønsted acids are
not required. We prefer a less restrictive set of postulates: an
(
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K.
N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.;
Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.;
Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels,
A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz,
P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson,
B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03;
Revision B.02. Gaussian, Inc.: Pittsburgh, PA, 2003.
(19) Halleux, A. L. US 3,000,984, 1961.
(
20) Di Stefano, M.; Negri, F.; Carbone, P.; M u¨ llen, K. Chem. Phys.
(32) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.
D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi,
M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; S. Clifford, S.;
Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Rega,
N.; Salvador, P.; Dannenberg, J. J.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Baboul, A.
B.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.;
Chen, W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-Gordon, M.;
Replogle, E. S.; Pople, J. A. Gaussian 98; Revision A.11.4. Gaussian, Inc.:
Pittsburgh, PA, 2002.
2
005, 314, 85-99.
(21) Kovacic, P.; Koch, F. W. J. Org. Chem. 1965, 30, 3176-3181.
(22) Kovacic, P.; Jones, M. B. Chem. ReV. 1987, 87, 357-379.
(23) Kovacic, P. Reactions of Aromatics with Lewis Acid Metal Halides.
In Friedel-Crafts and Related Reactions; Olah, G., Ed.; Wiley: New York,
965; Vol. 4, pp 111-126.
24) (a) Baddeley, G. J. Chem. Soc. 1950, 994-997. (b) Baddeley, G.;
Kenner, J. J. Chem. Soc. 1935, 303-309.
25) Balaban, A. T.; Nenitzescu, C. D. Acad. Repub. Pop. Rom., Fil.
1
(
(
Cluj. Stud. Cercet. Chim. 1959, 7, 521-529. Nenitzescu, C. D.; Balaban,
A. T. Chem. Ber. 1958, 91, 2109-2116.
(
26) Rooney, J. J.; Pink, R. C. Trans. Faraday Soc. 1962, 58, 1632-
1
641.
(
27) Clar, E. Ber. Dtsch. Chem. Ges. 1930, 63, 112-120.
(33) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(
67.
28) Simons, J. H.; McArthur, R. E. J. Ind. Eng. Chem. 1947, 39, 364-
(34) Becke, A. D. Phys. ReV. A 1988, 38, 3098-3100.
3
1
(35) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785-789.
(36) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28, 213-
222.
(
29) Pummerer, R.; Luther, F. Ber. Dtsch. Chem. Ges. 1928, 61, 1102-
107.
J. Org. Chem, Vol. 71, No. 14, 2006 5069

Rempala et al.
SCHEME 5. Arenium Cation Pathways for the
Condensation of o-Terphenyl with Ortho or Para
Protonations and Cis or Trans C-C Bond Formation
therefore, the BHandHLYP functional was also tested in the case
of o-terphenyl, as it is capable of locating transition states on radical
ion potential surfaces that are difficult to find with the B3LYP
functional.38 Standard cutoff values were used in geometry opti-
mizations. Stationary points were subjected to frequency calcula-
tions and assigned as minima (all real frequencies) or transition
states (exactly one imaginary frequency); this also provided G298
.
2
For doublet species (radical ions), the expectation values of S were
within 0.03 of the exact value, 0.75. Use of molecular symmetry
was deliberately avoided in radical cations. Wave function stability
calculations were performed for o-terphenyl doublet species. Single-
point solvation calculations (evaluating ∆Gsolv, with SCFVAC
option) were performed with polarizable continuum model,3 as
implemented in Gaussian 03, with a dielectric constant of 8.93
9,40
2 2
(CH Cl ). Atomic radii optimized for the HF level of theory were
applied (radii ) UAHF). No scaling factor was used for frequencies
in the evaluations of thermodynamic functions, and no attempt was
made to treat any low-frequency vibrations as internal rotations.
Results and Discussion
o-Terphenyl Coupling: Arenium Cation Mechanism. In
the arenium cation mechanism for the intramolecular C-C
bond-forming reaction of o-terphenyl, protonation of a peripheral
phenyl ring at sites leading to favorable charge density ortho
(position 3) and para (position 5) with respect to the new C-C
bond was considered. Although these are not the most basic
site in o-terphenyl, the differences in calculated proton affinities
between various sites are small. Since these protonated o-
terphenyl tautomers, which are also rotamers, are in rapid
equilibrium, the Curtin-Hammet principle applies:41 the reaction
pathway is defined by lowest absolute energy transition states,
not the relative population of tautomers or by their relative
activation energies. Both trans and cis geometries of the
electrophilic intermediates and transition states were considered.
The full mechanism of triphenylene formation via the ortho
protonation, trans pathway is shown in Scheme 5.
To identify the rate-limiting step, proton transfer and dehy-
drogenation transition states must also be considered. Because
the exact nature of these elementary reactions is unclear,
calculations of these transition states were not attempted. Rather,
activation parameters for related processes were taken from the
literature. Proton transfer in a model hydronium ion-benzene
system was calculated recently42 to have a barrier in vacuo of
4
kcal/mol; proper treatment of the problem required sophisti-
cated calculations, which are currently not feasible for larger
systems. Instead, this value was taken as the energy of proton-
q
q
transfer transition states (∆G and ∆H ) relative to higher energy
of the two species that this transition state connects. The
experimental value for the rate constant of the pseudo-first-
order loss of deuterium from C6H5D in 83 wt % H2SO4 is 3.45
-
4
-1
43
×
10
s
at 25 °C, which corresponds to a half-life of 33
In our mechanism, the presence of superacid is postulated, and
we adopt therefore the value from ab initio calculations.
The strength of the catalytic acid is an important thermody-
namic parameter. The relative energies of the starting material
min, indicating process somewhat less feasible than that
characterized by 4 kcal/mol activation barrier. However, the
measured rate was strongly dependent on the acidity function.
(o-terphenyl) and the final product (triphenylene + oxidation
(
37) Saettel, N. J.; Oxgaard, J.; Wiest, O. Eur. J. Org. Chem. 2001,
1
429-1439.
product of hydrogen, or H2 itself) do not depend on acid
strength, since the acid is a catalyst. The same holds for all
neutral intermediates considered (see below). The effect of acid
strength on the reaction coordinate diagram is straightforward.
The use of weaker acids results in more endergonic initial
protonation; the use of stronger acids results in more exergonic
initial protonation. We chose an acid that gives a thermoneutral
initial protonation: protonated o-terphenyl.
(
38) Oxgaard, J.; Wiest, O. J. Phys. Chem. A 2001, 105, 8236-8240.
39) Miertu sˇ , S.; Scrocco, E.; Tomasi, J. Chem. Phys. 1981, 55, 117-
(
1
29.
(40) Cossi, M.; Scalmani, G.; Rega, N.; Barone, V. J. Chem. Phys. 2002,
1
17, 43-54.
(
41) Curtin, D. Y. Rec. Chem. Prog. 1954, 15, 111-128.
(42) Kryachko, E. S.; Nguyen, M. T. J. Phys. Chem. A 2001, 105, 153-
1
55.
(43) Gold, V.; Satchell, D. P. N. J. Chem. Soc. 1955, 3619-3622.
5070 J. Org. Chem., Vol. 71, No. 14, 2006

Intramolecular Scholl Reaction of Contiguous Phenylbenzenes
TABLE 1. Activation Parameters and Energies (kcal/mol) for C-C Bond-Forming Transition States of the Arenium Cation Mechanism in the
o-Terphenyl Condensation
q
q
∆
G C-C
∆G relative acidities of
corresponding protonation sites
G relative to the
TS description
bond-forming step
most stable tautomer
2
2
3
3
9t: trans,ortho protonation
9c: cis,ortho protonation
6t: trans,para protonation
6c: cis,para protonation
14.2
19.1
11.8
17.9
0.2
0.2
0
14.4
19.3
11.8
17.9
0
The formation of triphenylene requires loss of H2 after C-C
bond formation and deprotonation. As in the case of proton
transfer discussed above, the exact nature of this step is unclear,
and therefore, calculations were not attempted. Rather, a
reasonable value for the activation barrier was assigned based
on prior experimental and computational work. The experimental
gas-phase activation barrier of the uncatalyzed dehydrogenation
of 1,4-cyclohexadiene is 43.8 kcal/mol.44 The measured con-
densed phase activation barrier for the oxidative dehydrogena-
tion of 1,4-dihydronaphthalene by 1,4-benzoquinone in ethox-
ybenzene is 18.9 kcal/mol.45 Oxidative dehydrogenation by
quinones is faster when the product is aromatic.46 The enthalpy
of activation for dehydrogenation of 9,10-dihydroanthracene by
and 17.9 kcal/mol (para protonation), and higher endergonici-
ties: +9.9 kcal/mol (ortho) and +13.4 kcal/mol.
4
1
Again, the Curtin-Hammett principle appliessneither
simple comparison of the basicities of different protonation sites
in tautomers 28 (ortho) or 35 (para), nor comparison the
activation barriers for C-C bond formation from 28 and 35, is
appropriate. Rather, the absolute values of the corresponding
C-C bond forming transition state energies define the preferred
reaction pathway. Table 1 summarizes these parameters for the
four C-C bond-forming TS shown in Scheme 5. The transition
state resulting from para protonation with a trans geometry, 36t,
is preferred, followed closely by 29t.
o-Terphenyl Coupling: Radical Cation Mechanism. Cy-
clization of the o-terphenyl radical cation was considered next.
Its formation would require use of a strong oxidantsthe
o-terphenyl ionization potential from photoelectron spectroscopy
47
q
q
acidified nitrobenzene is 6.8 kcal/mol sboth ∆G and ∆H of
oxidation were set to this value. Oxidation of 1,2-dihydroaro-
matics should be easier than of 1,4-dihydroaromatics, if
concerted. We have conservatively assumed the barrier to
dehydrogenation to be 6.8 kcal/mol.
4
8
x
+
is 7.99 eV and the E°
1
/2
is +1.30 V vs Ag|Ag in
4
9
acetonitrile. As was the case for the protonation mechanism,
two geometries (trans and cis) were considered. Cyclization of
the o-terphenyl radical cation 42 to give the trans isomer 44t
was calculated to be endergonic by 23.5 kcal/mol and cyclization
to give the cis isomer 44c was calculated to be endergonic by
28.5 kcal/mol.
The dehydrogenation of 2,3-dihydrotriphenylene with loss of
H2 is calculated to be exergonic by -27.4 kcal/mol. Dehydro-
genation will be more exergonic when coupled with oxidation
process, e.g., p-quinone + H2 f hydroquinone reaction, which
is calculated to be exergonic by -31 kcal/mol at the same level
of theory. Since the oxidant and its product of reaction with
hydrogen appear in the net equation of the Scholl reaction, the
oxidant strength influences the net energetics of the reaction.
The oxidant strength will affect the relative energies of the
dihydroaromatic and fully benzenoid intermediates or products.
In the case of multistep reactions this becomes more interesting,
as the oxidant strength may shift the position of the rate-
determining transition state toward the final product (see
C6Ph6 section).
The thermochemistry of the generation of the o-terphenyl
radical cation from o-terphenyl via electron transfer to a one-
electron oxidant can be controlled by choice of the oxidant and
reactions conditions. Oxidants strong enough to warrant exer-
4
+
gonicity of this process are available, e.g., Ce . However, the
free energy barriers for both the trans and cis C-C bond-
forming steps of the radical cation were substantial: 27.3 and
30.0 kcal/mol, respectively. The structures of these transition
states strongly resemble those of the cyclized radical cation
products, to which they are also close in energy (very small
barriers for reverse reactions), providing an illustration of the
Hammond postulate. Calculations at the BHandHLYP/6-31G-
(d) level provide similar results, with slightly lower free energy
expenditure (21.6 kcal/mol) and barrier (25.3 kcal/mol) to C-C
bond formation predicted (trans geometry).
The fair comparison of the arenium cation and the radical
cation mechanism for the conversion of o-terphenyl to triph-
enylene requires judicious choice of both the acid and the
oxidant. To accomplish this, we chose reagents in such way
that the initial step for both reactions, protonation or one-electron
+
oxidation, is thermoneutral. Thus, BH ) 28 and O ) oxidant,
with the strength such that ∆G(O + H2 f OH2) ) ∆G(2B +
Identifying the nature of the Scholl reactionsarenium cation
or radical cation s is an objective of this work. As mentioned
before, these fundamentally different mechanisms can be directly
compared if the acid and oxidant are chosen to provide the
identical net energetic outcome for both reactions. The use of
•
+
+
2
11 + H2 f 2BH + 211). The complete thermochemical
argument is included in the Supporting Information.
In the case of the trans transition state geometry and ortho
protonation, the C-C bond-forming activation barrier was
calculated to be 14.2 kcal/mol, and the step was endergonic,
+
protonated o-terphenyl as an acid (BH ) and the radical cation
+
4.7 kcal/mol. This compares to a 11.8 kcal/mol barrier and a
.1 kcal/mol endergonicity for trans geometry and para proto-
of o-terphenyl as oxidant (O) satisfy this criterion. Side-by-
side reaction coordinate diagrams with these thermodynamic
assumptions are presented in Figure 1.
8
nation (Scheme 5). Pathways leading to cis products result in
higher activation energies: 19.1 kcal/mol (ortho protonation)
(
47) Cristiano, M. L. S.; Gago, D. J. P.; Gonsalves, A. M., d’A. R.;
(
44) (a) Rico, R. J.; Page, M.; Doubleday, C., Jr. J. Am. Chem. Soc.
Johnstone, R. A. W.; McCarron, M.; Varej a˜ o, J. M. T. B. Org. Biomol.
Chem. 2003, 1, 565-574.
1
1
992, 114, 1131-1136. (b) Benson, S. W.; Shaw, R. Trans. Faraday Soc.
967, 63, 985-992.
(48) Dewar, M. J. S.; Goodman, D. W. J. Chem. Soc., Faraday Trans.
2 1972, 68, 1784-1788.
(45) Braude, E. A.; Jackman, L. M.; Linstead, R. P. J. Chem. Soc. 1954,
3
548-3563.
(49) Mizuno, K.; Ichinose, N.; Tamai, T.; Otsuji, Y. Tetrahedron Lett.
1985, 26, 5823-5826.
(46) Stoos, F.; Ro cˇ ek, J. J. Am. Chem. Soc. 1972, 94, 2719-2723.
J. Org. Chem, Vol. 71, No. 14, 2006 5071

Rempala et al.
SCHEME 6. Radical Cation Mechanism for o-Terphenyl
FIGURE 1. Comparison of reaction coordinates.
The key difference between the arenium cation mechanism
and the radical cation mechanism is that the cyclized Wheland
intermediate 37t (and its analogues) is much lower in energy
(
∼15 kcal/mol) than the cyclized radical cation 44t (and its ana-
logues) (Scheme 6). This difference is also manifest in the cor-
responding deprotonated forms 39/32 and 46. Since these inter-
mediates define the rate-determining proton transfer and oxida-
tion barriers, the arenium cation mechanism appears to be
favored.
1,2,3-Triphenylbenzene: Arenium Cation Mechanism. In
1,2,3-triphenylbenzene, 48, two aryl-aryl bonds must be formed
to produce dibenzo[fg,op]naphthacene 71 (Scheme 7). We
expect that 1-phenyltriphenylene 56 is the fully benzenoid
intermediate in this process. Since the dihydro intermediate is
much higher in energy relative to the fully benzenoid product
+
H2, deprotonation and oxidation to restore aromaticity will
precede formation of the second C-C bond. If this were not
the case, the simultaneous loss of aromatic stabilization in
multiple benzene rings would be energetically unreasonable.
Since in larger systems the consideration of all transition-
state possibilities becomes prohibitively time-consuming, only
trans transition-state geometries, which were the most favorable
for o-terphenyl, were calculated. Different sites of initial
protonation were considered to ensure that Curtin-Hammett
scenarios were properly modeled.
We now adopt the thermodynamic standard of loss of H2;
e.g., no oxidant is used. This is the equivalent of reporting
energies with respect to a standard hydrogen electrode. We
continue to use reference acid that provides a thermoneutral
first proton-transfer step. For para protonation, formation of the
first aryl-aryl bond requires overcoming an 11.2 kcal/mol free
energy barrier and is endergonic by 5.9 kcal/mol. For ortho
protonation, formation of the first aryl-aryl bond requires
overcoming a 13.4 kcal/mol energy barrier and is endergonic
by 5.8 kcal/mol. These calculated free energy barriers are about
mol (ortho). In contrast, when the triphenylene fragment is
protonated, activation barriers rise to 16.0 and 20.5 kcal/mol
(para and ortho, respectively) and reaction is endergonic: 9.8
(para) and 4.6 kcal/mol (ortho) (Figure 2).
However, the triphenylene moiety is much more basic than
the phenyl group, presumably because the resulting charge is
more delocalized. Ortho protonation of the triphenylene moiety
to give cation 80 results in a structure 14.7 kcal/mol more stable
(∆G) than ortho protonation at the phenyl group to give cation
65. Once again, the Curtin-Hammet principle applies: para
protonation of the triphenylene moiety gives the lowest energy
C-C bond-forming TS (Table 2).
The results for 1,2,3-triphenylbenzene indicate that introduc-
tion of the second ary-aryl bond by the arenium cation mech-
anism will be easier than formation of the first. Since the path-
ways intersect at the fully benzenoid intermediate 56, the actual
pathway is probably a combination of the pathways shown. The
two most favorable pathways are shown in Figure 2.
0
.7 kcal/mol lower than the barriers for analogous o-terphenyl
geometries, indicating that the additional phenyl group stabilizes
the transition states relative to the protonated hydrocarbon.
The formation of the second C-C bond, the conversion of
5
6 to 71, involves more possible protonation sites. To examine
influence of the choice of the protonation sites, two series were
considered: phenyl group protonation ortho and para relative
to the nascent bond (i.e., 64 and 72) and triphenylene core
protonation ortho and para relative to the nascent bond (i.e., 79
and 86). Protonation of the phenyl group results in remarkably
low free energy activation barriers to CC bond formation, 6.3
and 8.4 kcal/mol for para and ortho protonation, respectively,
and the reaction becomes exergonic: -1.6 (para) and -4.5 kcal/
5072 J. Org. Chem., Vol. 71, No. 14, 2006

Intramolecular Scholl Reaction of Contiguous Phenylbenzenes
SCHEME 7. Arenium Cation Mechanisms for Condensation of 1,2,3-Triphenylbenzene
TABLE 2. Activation Parameters for the Arenium Cation Mechanism in 1-Phenyltriphenylene (kcal/mol) (Scheme 7)
∆
G, acidities
of protonation sites,
q
q
∆
G , C-C bond-forming
G , relative to the
TS
step, relative to 65, 73, 80, 87
relative to 87
most stable tautomer
o-phenyl protonation, 66
p-phenyl protonation, 74
o-triphenylene protonation, 81
p-triphenylene protonation, 88
8.4
6.3
20.5
16.0
14.8
14.1
0.2
23.3
20.4
20.7
16.0
0.0
1
,2,3-Triphenylbenzene: Radical Cation Mechanism. The
mol endergonicity (full convergence of geometry of 96t could
not be achieved), the latter yielding 29.3 kcal/mol barrier and
28.1 kcal/mol endergonicity. These numbers indicate that the
radical cation mechanism in the C-C bond formation step
requires overcoming barriers larger than in the worst considered
case of the arenium cation mechanism. The formation of the
second C-C bond, 56 f 71 via 102, was also investigated.
The activation barrier and endergonicity remained high, at 24.2
and 20.3 kcal/mol, respectively. It is unlikely that radical cation
mechanism is feasible for either o-terphenyl or 1,2,3-triphenyl-
benzene. Our thermochemistry data for 94, 95t and 96t are in
good agreement with the B3LYP/3-21G calculations by Di
Stefano et al. published recently on o-terphenyl radical cation.20
radical cation mechanism was disfavored with respect to the
arenium mechanism in the condensation of o-terphenyl. To
determine if the radical pathway is increasingly favored in larger
systems, this mechanism was examined in condensation of 1,2,3-
triphenylbenzene (Scheme 8).
Both trans and cis geometries were considered for the
formation of the first C-C bond (95t and 95c), the former
resulting in 30.0 kcal/mol free energy barrier and 27.4 kcal/
1,2,3,4-Tetraphenylbenzene: Arenium Cation Mechanism.
The computational results for o-terphenyl and 1,2,3-triphenyl-
benzene suggest the following two mechanistic pathways merit
consideration. In the first pathway, the phenyl group adjacent
to the developing PAH system is protonated ortho to the new
bond and C-C bond formation occurs via a trans geometry.
Deprotonation and dehydrogenation follow. Next, the phenyl
group adjacent to the growing PAH core is protonated and the
cycle continues. Ortho protonation was chosen, as it consistently
results in largest exergonicity (lowest endergonicity) of the CC
bond-forming reaction, although activation barriers are lower
for para protonation. The second pathway differs from the first
in that the growing PAH system is protonated. Both pathways
are illustrated in Scheme 9.
The first C-C bond-forming step in the condensation of
1
,2,3,4-tetraphenylbenzene (107) in the ortho protonation path-
FIGURE 2. Reaction profiles for the most favorable arenium cation
mechanisms described in Scheme 7.
q
way requires 13.4 kcal/mol ∆G (110) and is endergonic by
J. Org. Chem, Vol. 71, No. 14, 2006 5073

Rempala et al.
SCHEME 8. Radical Mechanism for the Condensation of 1,2,3-Triphenylbenzene
SCHEME 9. Stationary Points on the o-Phenyl Protonation (Blue) and p-PAH Protonation (Red) Pathways for Arenium
Cation Condensation of 1,2,3,4-Tetraphenylbenzene
5074 J. Org. Chem., Vol. 71, No. 14, 2006

Intramolecular Scholl Reaction of Contiguous Phenylbenzenes
5
.8 kcal/mol. This is consistent with results obtained for
o-terphenyl and 1,2,3-triphenylbenzene.
Figure 3 compares reaction profiles for the o-phenyl and
p-PAH protonation mechanisms with the thermodynamic as-
sumptions that H2 is evolved and protonated 1,2,3,4-tetraphe-
nylbenzene, 109 (or acid of identical strength), serves as an acid.
This figure indicates that either the dehydrogenation TSs (114,
1
38) or the deprotonation TSs (112, 136) immediately after the
formation of the first C-C bond is rate determining. Because
common intermediates are involved, the actual optimal pathway
could be a combination of the o-phenyl protonation and the
p-PAH protonation pathways.
It is noteworthy that the first complete Scholl step (107 to
1
1
15) is thermoneutral, while subsequent condensations (115 to
23 and 123 to 131) are exergonic. A similar pattern is observed
for C6Ph6 and we believe that this phenomenon has major
consequences for the course of this type of reactions.
Hexaphenylbenzene: Arenium Cation Mechanism. The
ultimate goal of this study is consideration of the condensation
of C6Ph6 to HBC. The same mechanistic schemes described for
1,2,3,4-tetraphenylbenzene were applied to this reaction: o-
phenyl protonation with trans C-C bond formation and para
protonation of the PAH core (initially a phenyl group) with trans
FIGURE 3. Reaction profiles for mechanisms described in Scheme
+
1
1. BH ) 109, O ) none, free H
2
evolution.
6 6
SCHEME 10. Stationary Points on the o-Phenyl (Blue) Pathway for the Arenium Cation Condensation of C Ph to HBC
J. Org. Chem, Vol. 71, No. 14, 2006 5075

Rempala et al.
Ph to
SCHEME 11. Stationary Points on the p-PAH Protonation Pathway (Red) for the Arenium Cation Condensation of C
6
6
HBC
C-C bond formation. The results are summarized in Schemes
triphenylene, dibenzonaphthacene, tribenzoperylene, diben-
5
0
1
0 and 11 and Figure 4.
zophenanthropentaphene (cf. benzene, 3.34 kcal/mol)). We
1
An optimal combination of two pathways, as common
propose, as before, that this “slippery slope” phenomenon is
intermediates are present, can be chosen. Table 3 shows, for
both pathways, the energies of highest transition states, relative
to the proceeding fully benzenoid compound, for each of the
six bonds formed. These data indicate that the formation of the
first and last three bonds formation will follow the o-phenyl
protonation pathway, while the second and third aryl-aryl bond
formation will follow the p-PAH pathway. The oxidation step
responsible for the success of the Scholl condensation of
dendritic oligophenylenes.
The stereochemistry of the fully benzenoid intermediates (e.g.,
15, 167, etc.) merits discussion. While the o-phenyl protonation
2
pathway leads to C -symmetrical benzenoid intermediates, the
p-PAH protonation pathway leads to C symmetrical benzenoid
s
intermediates. The fully benzenoid intermediates 14 and 190
in the formation of the first C-C bond will be rate determining
are exceptionssthe C structure is preferred in both cases.
2
q
(
∆G rel, 21.4 kcal/mol); the subsequent cyclizations will be faster
Nonetheless, energy differences between C and C structures
2
s
q
(
-
∆G rel < 17 kcal/mol). The overall reaction is exergonic by
are low, and they are most likely in rapid equilibrium. To keep
93.8 kcal/mol, which is -15.6 kcal/mol/C-C bond, assuming
the two pathways readily comparable, the C geometries were
2
hydrogen evolution (no oxidant).
used in both schemes.
The dehydrogenation steps become more exergonic with the
progress of the reactionsthis can be attributed to the increasing
electron delocalization in the intermediates, all of which are
fully benzenoid. This increasing exergonicity can be related to
the increasing resonance energy per π electron (REPE) of the
growing PAH core (3.43, 3.53, 3.55, 3.50, 3.63 kcal/mol for
Other pathways may existsthe formation of C-C bonds is
not necessarily contiguous. To examine this possibility, the strain
energy of all fully benzenoid minima on the potential energy
surface for the conversion of C Ph to HBC were calculated
(Scheme 12). Strain was defined by a homodesmotic reaction
with triphenylene and o-terphenyl as unstrained references.
Indeed, the least strained intermediates lie on the contiguous
pathway, supporting the contiguous formation of C-C bonds.
6
6
5
0
(50) Randi c´ , M.; Guo, X. New J. Chem. 1999, 23, 251-260.
5076 J. Org. Chem., Vol. 71, No. 14, 2006

Intramolecular Scholl Reaction of Contiguous Phenylbenzenes
6 6
FIGURE 4. Energetics comparison of the o-phenyl protonation (blue) and p-PAH protonation (red) pathways of the condensation of C Ph to
HBC. Acid ) 154, H evolution.
2
TABLE 3. Maximum Free Energy TSs between Fully Benzenoid
Compoundsa
The rate-determining TS is the oxidation (159, 202, Schemes
0 and 11, Figure 4), or possibly the proton transfer, following
1
ortho pathway
p-PAH pathway
the first C-C bond formation. Regardless of the exact nature
of this TS, it is clear that the rate-determining TS51 occurs in
the course of the formation of the first C-C bond. It follows
that the intermediates should not accumulate during the course
of the reaction: after the formation of the first C-C bond, all
intermediates will be consumed faster than they are formed.
Moreover, each step is increasingly exergonic. This observation,
together with the Hammond postulate, implies that each
cyclization will be faster than the preceding one.
q
q
bond
formation
G rel,
∆G rel,
kcal/mol
kcal/mol
structure
structure
1
2
3
4
5
6
21.4
15.7
15.3
13.2
12.9
9.2
159
166
170
178
189
193
22.9
12.2
12.2
16.7
14.2
9.4
202
209
216
223
230
233
a
The lowest energy TS for each step is underlined.
52
Simulation of a simplified kinetic scheme, which involves
only the fully benzenoid hydrocarbons and the highest energy
transition states of the lowest energy pathway (Table 3), supports
Of course, TS barriers determine the favored reaction pathway,
but these are likely to be related by the relative energies of the
fully benzenoid intermediates.
It is interesting to notice that the strain in the proposed scheme
generally decreases with the number of C-C bonds. The strain
actually becomes negative for 190 and 10. This indicates that
electronic stabilization due to increased delocalization is not
taken into account in our definition of strain.
11
-1 -1
this analysis. A preexponential factor of 10
M
s
was used.
The simulation, based on 1 mM C6Ph6, 6 mM oxidant, and
excess acid, suggests that no intermediates will accumulate in
-
8
quantities exceeding 5×10 M (Figure 7).
Analytical treatment of a simple irreversible model, A1 f
A2 f A3, with rate constants k1 and k2, and [A1]t)0 ) [A1]0,
[A2]t)0 ) [A3]t)0 ) 0, offers some insight to why intermediates
do not accumulate. This scheme would describe the condensa-
tion of 1,2,3-triphenyl benzene in the presence of an excess of
acid and oxidant. The analytical solution for the concentration
The influence of the catalytic acid is easily understood (Figure
5). The relative energies of the neutral species are not affected
by choice of the acid. A weak acid will render the key protonated
intermediates inaccessible. A very strong acid would cause the
protonated intermediates to become energetic sinks, which
would slow the formation of the dihydro precursors.
-
k1t
-k2t
of intermediate A2 is [A2] ) [A1]0k1(e
- e )/(k2 - k1).
Keeping k1 constant and varying k2 from ∞ to 0, [A2]max, the
maximum concentration of the intermediate A2, spans a range
from 0 to [A1]0, as expected. For the case k2 ) k1, [A2]max equals
The more complex role of the oxidant is revealed by Figure
-
1
-1
[
A1]0/e at tmax ) k1 ) k2 . As the first step becomes rate
6. A weak oxidant not only lowers overall exergonicity, but
limiting (k2 > k1), [A2]max not only decreases, but the peak
concentration occurs earlier. If k2 and k1 are governed by
Arrhenius behavior and their preexponential factors are equal,
can shift the position of the rate determining transition state.
As a consequence, a weak oxidant would be beneficial if
partially condensed intermediates are desired.
Synthetic Study and Kinetic Modeling. The exceedingly
low solubility of HBC hinders the experimental investigation
of its formation from C6Ph6. However, analysis of the end
reaction mixture provided some mechanistic insight.
(
51) Espenson, J. H. Chemical Kinetics and Reaction Mechanisms, 2nd
ed.; McGraw-Hill: New York, 1995.
52) Chemical Kinetics Simulator; Chemical Kinetics Simulator v. 1.01,
IBM Corp., 1996.
(
J. Org. Chem, Vol. 71, No. 14, 2006 5077

Rempala et al.
SCHEME 12. Possible Pathways Connecting Benzenoid
Intermediates in Condensation of C Ph
6
6
FIGURE 5. Influence of the acid strength (defined as ∆
r
G (BH+(sol)
+
+
T(sol) f B(sol) +TH (sol)), T ) terphenyl).
r
FIGURE 6. Influence of the oxidant strength (defined as ∆ G (O(sol)
+
H2(sol) f OH2(sol))).
a 2 kcal/mol difference in activation energy at 298 K, i.e., E2q
q
-
E1 ) -2 kcal/mol would make k2 faster than k1 by a factor
0. This would result in the maximum concentration of A2 being
3
only 3% of [A1]0.
Another interesting system is based on the sequence of
second-order irreversible processes: a + o f b + r, b + o f
c + r, c + o f d + r, ..., f + o f g + r, where o ) oxidant,
r ) reduced form, [a]t)0 ) 1, and k1 ) k2 ) ... ) k6. Both
numerical solutions of the differential equations and stochastic
simulation of this reaction system demonstrate that the initial
concentration of the oxidant [o]t)0 determines the distribution
of the intermediates and the final product at t f ∞ (Figure 8).
Interestingly, the maximum concentration of the intermediate
“
e” is achieved when [o]t)0 ) 4.0, which is the stoichiometric
5078 J. Org. Chem., Vol. 71, No. 14, 2006

Intramolecular Scholl Reaction of Contiguous Phenylbenzenes
TABLE 4. Oxidation of C6Ph6 to HBC
mole ratio
oxidant/HBC
recovered
C6Ph6 (%)
trial
yield HBC (%)
MoCl5 in CH2Cl2
1
2
3
4
0.97
3.0
6.0
6.0
41.5
68.8
99.3
94.2
54.8
28.6
2.3
12
PhI(O2CCF3)2/BF3‚Et2O in CH2Cl2
5
6
7
8
1.0
3.0
5.9
9.0
5.3
34.7
81.1
77.8
86.9
71.0
23.6
8.0
CuCl2 /AlCl3 in CS2
9
0
1
1.0
6.0
6.9
18.8
60.3
91.9
76.2
33.1
1
1
12
6.0
18
Cu(OTf)2/AlCl3 in CS2
0.0
1
1
2
90.3
30.0
FeCl3 in CH2Cl2
117.7
3a
FIGURE 7. Concentrations of important species from kinetic simula-
tions.
a
Intractable mixture.
FIGURE 9. Mass balance summary (Table 4).
5
-8), CuCl2 with AlCl3 in CS2 at rt18 (entries 9-11), Cu(OTf)2
17,61
with AlCl3 in CS2 at rt (entry 12), and FeCl3 in CH2Cl2 at rt
(entry 14). The variable stoichiometry oxidations are sum-
marized in Table 4. In all cases except entry 13, mass balances
were nearly quantitative (Figure 9).
FIGURE 8. Ultimate (t f ∞) concentrations as a function of initial
concentration assuming a sequence of irreversible steps without a rate-
limiting step.
MoCl5/CH2Cl2 is a homogeneous medium for the Scholl
5
7,62
reaction
and converts C6Ph6 to HBC in good yield. A
disadvantage of this reagent is that chlorinated impurities are
formed, as indicated by MALDI-TOF MS and combustion
analyses of the crude product.
ratio for formation of “e” (this also applies to other intermedi-
ates). This intermediate “e” corresponds to compound 14
reported by M u¨ llen and co-workers.18
(
53) Kumar, S.; Manickam, M. Chem. Commun. 1997, 1615-1616.
These mathematical models and simulations show that small
reductions in the activation energy of the consecutive steps limit
accumulation of intermediates. The progressive lessening of
activation barrierssthe slippery slope effectshas two implica-
tions: (1) many consecutive reactions occur readily and (2)
intermediates will not accumulate. Consecutive intermediates
show decreasing maximum concentrations (Figure 7), as
expected for a scheme in which the first step is rate determining.
To test these predictions, a series of experiments were
performed in which the quantity of oxidant was varied and
reaction mixtures were carefully analyzed. Several oxidizing
(54) Waldvogel, S. R.; Wartini, A. R.; Rasmussen, P. H.; Rebek, J., Jr.
Tetrahedron Lett. 1999, 40, 3515-3518.
(55) Waldvogel, S. R.; Fr o¨ hlich, R.; Schalley, C. A. Angew. Chem., Int.
Ed. Engl. 2000, 39, 2472-2475.
(
56) Waldvogel, S. R. Synlett 2002, 622-624.
(57) Waldvogel, S. R.; Aits, E.; Holst, C.; Fr o¨ hlich, R. Chem. Commun.
2002, 1278-1279.
58) Kramer, B.; Averhoff, A.; Waldvogel, S. R. Angew. Chem., Int.
Ed. 2002, 41, 2981-2982.
59) Kramer, B.; Fr o¨ hlich, R.; Bergander, K.; Waldvogel, S. R. Synthesis
(
(
2003, 91-96.
(60) Takada, T.; Arisawa, M.; Gyoten, M.; Hamada, R.; Tohma, H.; Kita,
Y. J. Org. Chem. 1998, 63, 7698-7706.
(61) Fechtenk o¨ tter, A.; Tchebotareva, N.; Watson, M.; M u¨ llen, K.
5
3-59
systems were investigated: MoCl5 in CH2Cl2 at rt
1
(entries
Tetrahedron 2001, 57, 3769-3783.
(62) Kovacic, P.; Lange, R. M. J. Org. Chem. 1963, 28, 968-972.
6
0
-4), PhI(O2CCF3)2/BF3‚Et2O in CH2Cl2 at -40 °C (entries
J. Org. Chem, Vol. 71, No. 14, 2006 5079

Rempala et al.
a
Little is known about the stoichiometry or the mechanism of
the reaction of MoCl5 with aromatics. It is not fully clear
whether the reaction is a two-electron (Mo + 2e f Mo ;
6 6
SCHEME 13. Products from Oxidation of C Ph
5+
3+
5
+
3+
+
5+
Mo + 2[H] f Mo + 2H ) or one-electron process (Mo
4
+
5+
4+
+
+
e f Mo ; Mo + [H] f Mo + H ). Preparation of
MoCl4 by reaction of MoCl5 with benzene, where chlorobenzene
6
3
is generated, was reported. We have isolated MoCl3, which
was identified by X-ray crystallography, from the reaction
mixture after quenching with methanol. However, the presence
of this compound might arise from the reduction of MoCl5 or
MoCl4 with methanol.
Cyclodehydrogenation by PhI(O2CCF3)2/BF3‚Et2O is usually
a
Conditions: (a) MoCl5/CH2Cl2, rt, 24 h; (b) CuCl2/AlCl3/CS2, rt, 24
6
0
performed at low temperatures (-40 °C) for 1.5 h. In our
experiments, we extended reaction time to 3.5 h and the
temperature was allowed to rise slowly to -10 °C. Good yields
of 10 were obtained under these conditions.
h; (c) FeCl3/CH2Cl2, rt, 24 h; (d) PhI(COCF3)2/BF3‚Et2O, -40 °C, 3 h.
showing significant level of chlorination (C, 90.56; H, 4.38;
Cl, 1.29).
Oxidations using Cu(OTf)2, CuCl2, and FeCl3 were performed
for comparison. Reactions with CuCl2 for 12 h afforded lower
conversions, probably due to its low solubility in CS2. Cu(OTf)2
was ineffective, giving no reaction after 12 h, also probably
due to its low solubility. The experimental details of the
oxidation of the parent C6Ph6 with FeCl3 have not previously
been reported to our best knowledge,14 but this reagent has been
used for substituted hexaphenylbenzenes. Because this oxidant
For all reaction conditions, intermolecular condensation
products could be detected. For example, a peak at m/z ) 1038,
which corresponds to with condensed product 241 formed from
two hexabenzocoronene units (Scheme 13), was observed. This
dimer was observed in all samples, including those obtained
using substoichiometric oxidant.
Concluding Remarks
1
7,61
is stronger and more soluble, the usual reaction time
is
limited to 1 h. Analysis of the sample taken from the reaction
mixture after 1 h confirmed the presence of the final product,
HBC. Extending the reaction time to 12 h produced intractable
material, which could not be characterized, even by MALDI-
or LD-TOF mass spectroscopy. The mass balance of the FeCl3
experiment exceeded 100%, suggesting that the product was
contaminated by insoluble iron oxides.
Mechanistically relevant information was derived from gravi-
metric analyses of oxidations with substoichiometric oxidant.
If partially oxidized intermediates (e.g., 15, 167) accumulate
during the course of reaction, they will be present when a
substoichiometric amount of oxidant is consumed.
For this reason, experiments with variable amounts of
oxidants were performed. These experiments require quantitative
mass balance data. Insoluble HBCwas isolated by filtration and
exhaustive washing; purity was checked by MALDI-TOF MS
and combustion analysis. Unreacted starting material C6Ph6 was
isolated and its purity verified by HPLC with tandem UV-vis
photodiode array (PDA) and evaporative light scattering (ELS)
detectors and MALDI-TOF MS. Based on reported solubility
of partially oxidized 14, we anticipated that other partially
oxidized intermediates would be soluble. In all cases, the yield
of HBC and recovery of C6Ph6 accounted for all mass, within
experimental uncertainty. In the case of oxidation by substo-
ichiometric CuCl2 (entries 9 and 10), we observed formation
of a yet-unidentified soluble compound (∼3%, ELS) the UV
spectrum of which may correspond to a true intermediate. No
other intermediates could be detected, and are therefore only
present in trace (<1%) amounts.
The radical cation and arenium cation mechanisms of the
Scholl reaction were investigated computationally. A comparison
of these mechanisms, based on the thermodynamic assumption
that the net energetic outcome is identical, reveals that the
arenium cation mechanism involves lower energy transition
states than the radical cation mechanism. In larger systems, the
radical cation mechanism is more strongly disfavored. These
findings are consistent with the literature reports of the Scholl
reaction proceeding in acidic media that do not promote radical
cation formation (anhydrous HF). We believe that the Scholl
reaction of unsubstituted oligophenylenes proceeds by an
arenium cation mechanism, not a radical cation mechanism.
The thermodynamic calculations and kinetic analysis of the
conversion of C6Ph6 to HBC suggest that intermediates will not
accumulate in the course of the reaction. This contrasts with
the results obtained by Halleux and by M u¨ llen and co-workers.
Halleux, however, did not report details of separation and
characterization. M u¨ llen and co-workers characterized the
isolated intermediatesthere is no doubt to its formation. This
illustrates a shortcoming of our calculationsswe are not
modeling the actual system. The oxidant used, CuCl2, is
insoluble in carbon disulfide and the oxidation occurs on the
CuCl2 surface. The various intermediates certainly interact
differently with the surface. In homogeneous reactions, we are
unable to observe the accumulation of intermediates. We observe
the formation of an HBC dimer or dimers, 241, under all reaction
conditions.
2
8
Several oxidation methods were tested and compared, which
provided the opportunity for reaction optimization. The homo-
geneous oxidizing systems, MoCl5 in CH2Cl2 and PhI(O2CCF3)2/
BF3‚Et2O in CH2Cl2, were shown to be effective for the Scholl
cyclization of C6Ph6 to HBC.
Analysis of the insoluble crude HBC products from reactions
using MoCl5 by MALDI-TOF MS led to another interesting
finding. Besides of major peak of HBC at m/z ) 522, minor
components corresponding to chlorinated derivatives of 10 (m/z
)
556, 590, 624) could be detected. This finding was cor-
roborated by combustion analysis of the crude reaction product,
Experimental Section
General Methods. Reactants and reagents were used as received.
CH
and CS
2
Cl
2
was distilled from CaH
2
and stored over molecular sieves,
. Good ventilation, protective
(63) Larson, M. L.; Moore, F. W. Inorg. Chem. 1964, 3, 285-286.
2
was stored over CaCl
2
5080 J. Org. Chem., Vol. 71, No. 14, 2006

Intramolecular Scholl Reaction of Contiguous Phenylbenzenes
equipment, and absence of sources of ignition are vital for the safe
mmol) in CH
2
Cl
2
(5 mL) was added dropwise an equimolar CH
2
-
handling of CS
2
due to its toxicity and low flash point.
Cl solution of PhI(O
4). The reaction mixture was stirred for 3.5 h while the temperature
was allowed to rise to -10 °C. The reaction was quenched by
MeOH and an aqueous solution of Na
2
2
CCF and BF ‚Et O at -40 °C (see Table
3
)
2
3
2
Mass spectra were recorded on a MALDI-TOF MS instrument
in reflector mode using an R-cyano-4-hydroxycinnamic acid matrix.
HPLC analyses were done using a photodiode array detector and
evaporative light scattering detector, employing a reversed-phase
2 3
SO . The resulting mixture
was worked up as described for MoCl
5
oxidations.
C
18 column and a mobile-phase methanol-water (92:8).
General Procedure: Oxidation by MoCl , Cu(II) Salts, FeCl
To a solution of C Ph (0.100 g, 0.187 mmol) in 10 mL of solvent
Acknowledgment. This work was supported by an award
from the Research Corporation, the University of Nevada, the
National Science Foundation (CHE-0449740), and the Office
of Naval Research. Acknowledgment is made to the Donors of
the American Chemical Society Petroleum Research Fund for
partial support of this research.
5
3
.
6
6
was added an appropriate amount of oxidant (see Table 4), and the
reaction mixture was stirred under nitrogen at rt for 12 h. Methanol
(
10 mL) was added and stirring continued for 1 h. The mixture
was diluted with CH Cl and methanol, and the insoluble HBC was
collected by fitration. The solids were exhaustively washed with
MeOH, CH Cl , MeOH, H O, MeOH, CH Cl , THF, benzene, and
CH Cl . The filtrate was diluted with H O and extracted with CH
Cl . The composition of the resulting CH Cl -soluble residue was
analyzed by HPLC.
General Procedure: Oxidation by PhI(O
To a solution of 1,2,3,4,5,6-hexaphenylbenzene (0.100 g, 0.187
2
2
Supporting Information Available: Optimized structures (Car-
tesian coordinates), energies, and thermodynamic arguments for the
comparison of the arenium cation and radical cation mechanisms.
This material is available free of charge via the Internet at
http://pubs.acs.org.
2
2
2
2
2
2
2
2
2
-
2
2
2
2
CCF
)
3 2
/BF
3
2
‚Et O.
JO0526744
J. Org. Chem, Vol. 71, No. 14, 2006 5081