Intramolecular aromatic 1,5-hydrogen transfer in free radical reactions III. Reactivity of diaryl ketones, ethers, thioethers, sulfoxydes, and sulfones. An experimental and theoretical study

Article abstract of DOI:10.1021/ol027301t

(Matrix presented) Experiments show that free radical hydrogen shift is significant in the Pschorr cyclization of diphenyl ethers (X = O) and thioethers (X = S) and does not take place with sufoxides (X = SO) and sulfones (X = SO₂). DFT calculations of the product ratios, activation energies, rate constants for H-transfers, and ring-closings at the UB3PW91/6-31G(d,p) level are in excellent agreement with the experimental results reported here and elsewhere in the literature.

Full text of DOI:10.1021/ol027301t

ORGANIC
LETTERS
2003
Vol. 5, No. 8
1175-1178
Intramolecular Aromatic 1,5-Hydrogen
Transfer in Free Radical Reactions III.
Reactivity of Diaryl Ketones, Ethers,
Thioethers, Sulfoxydes, and Sulfones.
An Experimental and Theoretical Study
,†
,‡
Sandor Karady,* Jordan M. Cummins,† J. J. Dannenberg,* Emma del Rio,§
Peter G. Dormer,† Benjamin F. Marcune,† Robert A. Reamer,† and
,§
Tomas L. Sordo*
Process Research Department, Merck Research Laboratories, Merck and Company,
P.O. Box 2000, Rahway, New Jersey 07065, Department of Chemistry,
City UniVersity of New York-Hunter College and Graduate School,
695 Park AVenue, New York, New York 10021, and UniVersidad de OViedo,
Departamento de Quimica Fisica y Analitica, 33006 OViedo,
Principado de Asturias, Spain
sandor_karady@merck.com
Received November 18, 2002
ABSTRACT
Experiments show that free radical hydrogen shift is significant in the Pschorr cyclization of diphenyl ethers (X ) O) and thioethers (X ) S)
and does not take place with sufoxides (X ) SO) and sulfones (X ) SO ). DFT calculations of the product ratios, activation energies, rate
2
constants for H-transfers, and ring-closings at the UB3PW91/6-31G(d,p) level are in excellent agreement with the experimental results reported
here and elsewhere in the literature.
After a century of Pschorr chemistry, the free radical aromatic
hydrogen shift was first clearly recognized in our attempt to
use the Pschorr cyclization for the preparation of a substituted
fluorenone.1 It was unexpected to find that the chemistry of
the benzophenone radical system is dominated by intramo-
lecular hydrogen transfer as demonstrated by the dual
products formed in Pschorr cyclization, the Sandmeyer
reaction, and hydro-, hydroxy-, and iododediazoniation
reactions.2 The estimated rate of the hydrogen shift was 106-
107 s-1. Subsequently, we have shown that the rearrangement
can be initiated by photolysis or by tin reduction of an
iodobenzophenone and it can be terminated by reaction with
aromatic solvents.3 Molecular orbital calculations established
the activation free energy for hydrogen transfer at 8.5 kcal
and the estimated rate of the hydrogen shift was in good
agreement with the experiments.4
Recently, several groups reported on the observation of
analogous hydrogen migrations.5 Notable among these is the
elegant and detailed study of Hanson and his group,6
confirming the hydrogen translocation rate at 1.2 × 106 s-1.
† Merck Research Laboratories.
‡ Hunter College.
Most of these studies were concentrated on the substituted
benzophenone system. One could speculate that the conjuga-
tion and electronic effect of the carbonyl is essential to this
rearrangement and therefore the all-aromatic hydrogen shift
is limited to benzophenones.
§ Universidad Oviedo.
(1) Yasuda, N.; Huffman, M. A.; Ho, G.-J.; Xavier, L. C.; Yang, C.;
Emerson, K. M.; Tsay, F.-R.; Li, Y.; Kress, M. H.; Rieger, D. L.; Karady,
S.; Sohar, P.; Abramson, N. L.; DeCamp, A. E.; Mathre, D. J.; Douglas,
A. W.; Dolling, U.-H.; Grabowski, E. J. J.; Reider, P. J. J. Org. Chem.
1998, 63, 5438-5446.
10.1021/ol027301t CCC: $25.00 © 2003 American Chemical Society
Published on Web 03/27/2003

Scheme 1
In this paper we investigate the generality of this phe-
nomenon, examining the role of the bridge between the two
interacting aromatic systems. We report here the results of
our experimental and theoretical study on the chemistry of
diaryl ethers, -thioethers, -sulfoxides, and -sulfones, in
comparison with benzophenones.
a third of the product was generated by rearrangement (3a),
while with the thioethers only 10% rearrangement (3b) was
observed.
Pschorr cyclization with the corresponding sulfoxide 1c10
generated a mixture of sulfide (2b), sulfoxide (2c), and
sulfone (2d), all derived from the direct ring closure product
2c by a disproportionation process. Only direct ring closure
to 2d was observed in the reaction of diazo sulfone 1d.11
There was no indication of hydrogen shift in the last two
systems. In all these reactions, about 25% reduction product
4a, 4b, 4c, or 4d accompanied the ring closure.
The diphenylmethane system (1f) was also tested but in
this case no ring closure was observed and the dominant
product was 4f, probably the result of a favorable intramo-
lecular hydrogen shift from the dibenzylic methylene group.
The mechanism of the hydrogen transfer and the various
rate constants considered for the theoretical calculations are
outlined in Scheme 2.
The copper-catalyzed Pschorr cylicization was used to
evaluate the extent of free radical hydrogen migration
operating in a system.7 Since the reactions were carried out
in water, the coating of the catalyst by the precipitating
product prevented the completion of the reaction. When
isopropyl alcohol or tert-butyl alcohol was used as a
cosolvent, hydrogen transfer from the solvent became the
dominant pathway, generating 4 as the major product.
Addition of Celite to the solution of 18 to physically adsorb
the precipitating product resulted in complete reaction in a
few hours.9 As is summarized in Table 1, in the ether series
It is clear from Table 1 that the experimentally observed
ratios of products 2a, 3a and 2b, 3b are in good agreement
with the calculated ratios. Furthermore, in the case of
sulfoxide 1c and sulfone 1d, where hydrogen transfer was
not observed, calculations clearly favor rapid ring closure.
The DFT calculations used the B3PW91 hybrid functional.
The details of the calculations are the same as described in
our previous paper4 except that the Gaussian 98 program
was used. This method combines Becke’s 3-parameter
functional12 with the nonlocal correlation provided by the
Perdew-Wang expression.13 According to recent reports,14
only hybrid functionals can provide an accurate description
for the systems with hydrogen bonds. Since the H‚‚‚C
interaction that occurs along the reaction path might bear
some resemblance to an H-bonding interaction, we deemed
(2) Karady, S.; Abramson, N. L.; Dolling, U.; Douglas, A. W.;
McManemin, G. J.; Marcune, B. J. Am. Chem. Soc. 1995, 117, 5425-
5426.
(3) Cummins, J. M.; Dolling, U.; Douglas, A. W.; Karady, S.; Leonard,
W. R.; Marcune, B. Tetrahedron Lett. 1999, 34, 6153-6156.
(4) Sordo, T. L.; Dannenberg, J. J. J. Org. Chem. 1999, 64, 1922-
1924.
(5) (a) Qian, X.; Mao, P.; Yao, W.; Guo, X. Tetrahedron Lett. 2002,
43, 2995-2998. (b) Qian, X.; Cui, J.; Zhang, R. Chem. Commun. 2001,
24, 2656-2657. (c) Sengupta, S.; Bhattacharyya, S. Tetradedron Lett. 2001,
42, 2035-2037. (d) Cioslowski, J.; Piskorz, P.; Mocrieff, D. J. Org. Chem.
1998, 63, 4052-4054.
(6) Chandler, S. A.; Hanson, P.; Taylor, A. B.; Walton, P. H.; Timms,
A. W. J. Chem. Soc., Perkins Trans. 2 2001, 214-228.
(7) The reactions were monitored and the product ratios were determined
by liquid chromatography. The isolated products were identified by H1 and
13C NMR and mass spectometry.
(8) Standard synthetic methods were used to prepare the starting
diazonium fluoroborates utilizing the following scheme:
was stirred for several hours. After methylene chloride was added, the
mixture was filtered, the layers were separated, the organic layer was dried,
concentrated, and analyzed by liquid chromatography, and NMR Pure
substrates were isolated by preparative TLC, utilizing silica gel plates.
(10) Sulfoxyde 1c was prepared by the scheme shown in ref 8 except
the nitrosufide 5 (X ) S) was oxidixed to the corresponding sufoxyde 5
(X ) SO) with Mg perphthalate in CH2Cl2, MeOH mixture.
(11) The sulfone 1d was also prepared as shown in ref 8, but the
nitrosufide intermediate 5 (X ) S) was oxidized to the corresponding sufone
5 (X ) SO2) with m-chloro-perbenzoic acid in CH2Cl2.
(12) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(9) Typical experimental procedure for Pschorr cyclization is as
follows: To a solution of 310 mg of diazonium fluoborate 1a in 100 mL
of 1 N sulfuric acid was added 1 g of Celite and the mixture was purged
with nitrogen. Powdered cuprous oxide was then added and the mixture
(13) Perdew, J. P.; Wang, Y. Phys. ReV. B 1992, 45, 13244.
(14) (a) Mamann, D. R. Phys. ReV. B 1997, R10157. (b) Maerker, C.;
Schleyer, P. v. R.; Liedl, K. R.; Ma, T.-K.; Quack, M.; Suhn, M. A. J.
Comput. Chem. 1997, 18, 1695.
1176
Org. Lett., Vol. 5, No. 8, 2003

Scheme 2
it appropriate to use this type of functional.15 B3PW91
eliminates the large spin contamination generally encountered
with unrestricted Hartree-Fock (UHF) wave functions.15
Stable structures were fully optimized and transition states
located at the UB3PW91/6-31G(p,d) level with use of the
GAUSSIAN 98 series of programs.16 Standard integration
grids were employed. We calculated the vibrational frequen-
cies of all stationary points to characterize them and obtained
the zero-point vibrational energies (ZPVE) and ∆H, ∆S, and
∆G values. These calculations used the harmonic oscillator,
rigid rotor, and ideal gas approximations at room temperature
(298.15 K) and 1 atm of pressure.
for ring closure as C-C bond formation can be either above
or below the ring at each of the two ortho positions. We
multiplied the rate constants calculated as described above
by the appropriate statistical factor (2 for H-transfer and 4
for ring closure) to obtain the values given in Table 1.
The product ratios are calculated by using the following
assumptions (see Scheme 2 for definitions of the rate
constants):
(1) The meta-substituent does not affect the ∆Gq values
for H-transfer or ring closure. Thus, all H-transfers have the
same kh. Due to statistics this is kh/2 for each of the two
ortho H’s.
(2) The unrearranged and rearranged radicals ring-close
exclusively to product 2 or 3, respectively, with identical
rate constants of kc.
All calculations were performed on the parent system,
which does not have methyl or chloro substituents.
The calculation of the rate constants and kinetic analyses
depends on several assumptions. Specific rate constants, k,
are calculated from the ∆Gq values, using the expression
(kBT/h)exp(-∆Gq/RT), which comes from absolute rate
theory. In this expression, kB is the Boltzmann constant, and
h is Planck’s constant. At 298 K (kBT/h) ) 6.21 × 1012.
Since the TS values for H-transfer are planar (except for the
sulfone, see below), there are two equivalent TS’s for
H-transfer, corresponding to transfer of an H-atom from
either of the ortho positions of the ring. Since the TS for
ring closure is not planar, there are four equivalent TS values
(3) The rearranged radical ring-closes with identical rate
constants of kc/2 to each of the products 2 and 3.
(4) Once the ring is closed, it does not reopen. Thus the
product ratio (2/3) is determined upon ring closure. There is
no experimental evidence for this, but ring opening should
be endoergonic.
(5) H-abstraction is not competitive with the H-transfer
and ring closure. In the cases where this last assumption does
not hold (see the experimental results), the abstraction is
assumed to occur with equal facility from the unrearranged
and rearranged radicals.
Table 1. Experimental and Calculated Product Ratios (2/3) and Calculated Activation Parameters (kcal/mol) and Rate Constants (s-1
)
at 298 K
ratio 2:3
H-transfer
ring closure
reacting system
ether (1a )
thioether (1b)
sulfoxide (1c)
sulfone (1d )
exptl
66:33
calcd
∆Hq
∆Gq
kh
∆Hq
∆Gq
kc
66:34
95:5
>99:<1
>99:<1
55:45b
67:33
6.7
6.2
11.8
13.3
6.7
8.8
8.7
13.4
14.9
8.5
4.8 × 106
5.0 × 106
1.9 × 103
1.7 × 102
7.9 × 106
2.0 × 105
8.0
5.7
7.9
8.5
8.4
8.6
10.1
8.3
9.9
10.7
10.2
11.5
1.1 × 106
2.0 × 107
1.5 × 106
3.8 × 106
8.4 × 105
1.0 × 105
90:10
no rearr
no rearr
55:45a
ketone (1e)
diphenylmethane (1f)
no rearrc
8.7
10.7
a See ref 2. b See ref 4. c No ring closure was observed.
Org. Lett., Vol. 5, No. 8, 2003
1177

By using these assumptions, the product fraction of 2
becomes (kc + kh/2)/(kc + kh), where the numerator represents
the sum of the rate constants for formation of 2 from
unrearranged and rearranged radicals, while the denominator
represents the sum of all reactions emanating from the
original radical. This expression becomes (1 + kh/2kc)/(1 +
kh/kc) upon division of numerator and denominator by kc.
The calculated product ratios are collected with the rate
constants in Table 1.
angles in the ground states of these species are more acute
than in the others, requiring a greater distortion to achieve a
planar transition state for H-transfer. In fact, the TS for
H-transfer in the sulfone is not planar. (2) The C-H‚‚‚O
interactions which favor the transition state of the benzophe-
none series preferentially stabilize the ground states of the
sulfone and sulfoxide. The C-H‚‚‚O distances calculated
for these species are consistent with other calculated17 and
observed18 C-H‚‚‚O interactions.
The calculated product ratios are in remarkably good
agreement with the experimental observations. The calculated
value for kc for the benzophenone radical (8.4 × 105 L/s ×
mol) is in excellent agreement with our previous estimate
and with the more precise value (8.0 ( 0.9 × 105 L/s ×
mol) reported by Hanson.6 However, the corresponding
values for kh (7.9 × 106 × and 1.2 ( 0.2 × 106 L/s × mol)
differ by a somewhat larger amount. Use of the Hanson’s
rate constants would lead to a predicted product ratio (2e/
3e) of 70:30 rather than 55:45.
The three compounds for which H-transfer is predicted
and observed all have similar activation energies for H-
transfer (as does diphenylmethane for which there are no
experimental data). On the other hand, the sulfone and
sulfoxide both have significantly higher activation energies
for this process. The higher barriers to H-transfer are likely
due to a combination of two related factors: (1) the C-S-C
The extent of H transfer depends on the ratio of the rates
for H-transfer and ring closure. Since the rates for ring
closure are much less sensitive to the structural variations
studied here than those for H-transfer, this ratio is more
sensitive to the H-transfer rate. H-transfer must occur through
a planar (or nearly so) transition state. Those factors that
stabilize such a structure such as C-H‚‚‚O interactions and
relatively unstrained C-X-C angles in the planar structure
will favor H-transfer.
The conclusion of this paper is that free radical hydrogen
shift takes place readily with aromatic ketones, ethers, and
sufides and is expected to proceed with similar ease in other
aromatic systems where the coplanarity required for the
transition is not restrained. Therefore, we can expect that
hydrogen transfer will participate in the free radical chemistry
of a variety of aromatic systems.
Acknowledgment. This work was supported (in part) by
grants from PSC-CUNY (Hunter) and DGICYT (SAF2001-
3596) (Oviedo).
(15) Cioslowski, J.; Liu, G.; Moncrieff, D. J. Org. Chem. 1996, 61, 4111.
Cioslowski, J.; Liu, G.; Matinov, M.; Piskorz, P.; Moncrieff, D. J. Am.
Chem. Soc. 1996, 118, 5261.
(16) 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.; Clifford, S.;
Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma; K.;
Salvador, P.; Dannenberg, J. J.; Malick, D. K.; Rabuck, A. D.; Raghavachari,
K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Baboul, A. G.; 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; Gaussian, Inc.: Pittsburgh, PA, 2001.
Supporting Information Available: Spectral data and
structure correlation for relevant compounds. This material
is available free of charge via the Internet at http://pubs.acs.org.
OL027301T
(17) (a) Turi, L.; Dannenberg, J. J. J. Phys. Chem. 1993, 97, 7903. (b)
Turi, L.; Dannenberg, J. J. J. Phys. Chem. 1993, 97, 12197. (c) Turi, L.;
Dannenberg, J. J. J. Chem. Mater. 1994, 6, 1313.
(18) (a) Peterson, K. I.; Klemperer, W. J. J. Chem. Phys. 1984, 81, 3482.
(b) Desiraju, G. R. Acc. Chem. Res. 1991, 24, 290.
1178
Org. Lett., Vol. 5, No. 8, 2003