Structures of reactive nitrenium ions: Time-resolved infrared laser flash photolysis and computational studies of substituted N-methyl-N-arylnitrenium ions

Article abstract of DOI:10.1021/ja001184k

A series of para-substituted N-methyl-N-phenylnitrenium ions (N-(4-biphenylyl)-N-methylnitrenium ion, N-(4-chlorophenyl)-N-methylnitrenium ion, N-(4-methoxyphenyl)-N-methylnitrenium ion, and N-(4-methylphenyl)-N-methylnitrenium ion) were generated through photolysis of the appropriately substituted 1-aminopyridinium salt. Laser flash photolysis using UV - vis detection as well as photoproduct analysis verified that the expected nitrenium ions were formed cleanly and rapidly following photolysis. Laser flash photolysis with time-resolved infrared detection allowed for structural characterization of the nitrenium ions through observation of a symmetrical aromatic C=C stretch in the region 1580-1628 cm^-1. The specific frequencies reflect the degree of quinoidal character present in each phenylnitrenium ion (i.e., the degree to which the nitrenium ion resembles a 4-iminocyclohexa-2,5-dienyl cation). The 4-methoxy derivative shows the highest frequency C=C stretch, indicating that this strongly π-electron-donating substituent imparts more quinoidal character, and the 4-chloro derivative shows the lowest frequency C=C stretch, suggesting that it possesses the least quinoidal character. Quantum calculations using density functional theory (BPW91/cc-pVDZ) were carried out on the same nitrenium ions. The theoretically derived IR frequencies showed excellent quantitative agreement with the experiment. The computed structures show significant bond length alternation in the phenyl rings, shortened C-N bond lengths, and substantial positive charge delocalization into the phenyl rings. All of these effects are more pronounced with increasing π-donating character of the ring substituent. Arylnitrenium ions are well described as 4-iminocyclohexa-2,5-dienyl cations.

Full text of DOI:10.1021/ja001184k

J. Am. Chem. Soc. 2000, 122, 8271-8278
8271
Structures of Reactive Nitrenium Ions: Time-Resolved Infrared Laser
Flash Photolysis and Computational Studies of Substituted
N-Methyl-N-arylnitrenium Ions
Sanjay Srivastava,† Patrick H. Ruane,† John P. Toscano,*,† Michael B. Sullivan,‡
Christopher J. Cramer,*,‡ Dominic Chiapperino,§ Elizabeth C. Reed,§ and
Daniel E. Falvey*,§
Contribution from the Department of Chemistry, Johns Hopkins UniVersity, 3400 North Charles Street,
Baltimore, Maryland 21218-2685, Department of Chemistry and Supercomputer Institute, UniVersity of
Minnesota, 207 Pleasant Street SE, Minneapolis, Minnesota 55455-0431, and Department of Chemistry
and Biochemistry, UniVersity of Maryland, College Park, Maryland 20742-2021
ReceiVed April 5, 2000. ReVised Manuscript ReceiVed June 12, 2000
Abstract: A series of para-substituted N-methyl-N-phenylnitrenium ions (N-(4-biphenylyl)-N-methylnitrenium
ion, N-(4-chlorophenyl)-N-methylnitrenium ion, N-(4-methoxyphenyl)-N-methylnitrenium ion, and N-(4-
methylphenyl)-N-methylnitrenium ion) were generated through photolysis of the appropriately substituted
1-aminopyridinium salt. Laser flash photolysis using UV-vis detection as well as photoproduct analysis verified
that the expected nitrenium ions were formed cleanly and rapidly following photolysis. Laser flash photolysis
with time-resolved infrared detection allowed for structural characterization of the nitrenium ions through
observation of a symmetrical aromatic CdC stretch in the region 1580-1628 cm-1. The specific frequencies
reflect the degree of quinoidal character present in each phenylnitrenium ion (i.e., the degree to which the
nitrenium ion resembles a 4-iminocyclohexa-2,5-dienyl cation). The 4-methoxy derivative shows the highest
frequency CdC stretch, indicating that this strongly π-electron-donating substituent imparts more quinoidal
character, and the 4-chloro derivative shows the lowest frequency CdC stretch, suggesting that it possesses
the least quinoidal character. Quantum calculations using density functional theory (BPW91/cc-pVDZ) were
carried out on the same nitrenium ions. The theoretically derived IR frequencies showed excellent quantitative
agreement with the experiment. The computed structures show significant bond length alternation in the phenyl
rings, shortened C-N bond lengths, and substantial positive charge delocalization into the phenyl rings. All
of these effects are more pronounced with increasing π-donating character of the ring substituent. Arylnitrenium
ions are well described as 4-iminocyclohexa-2,5-dienyl cations.
Introduction
strated that appropriately substituted arylnitrenium ions can be
selectively trapped by guanine bases in DNA.2,6,8,11-13 This
trapping event is thought to lead to carcinogenic mutations.14
Arylnitrenium ions are short-lived, electrophilic species,1-4
which have received considerable attention in recent years due
to their suspected role in chemical carcinogenesis.5-8 It is known
that certain aromatic amines, such as 2-acetylaminofluorene,
are enzymatically converted into sulfate esters of the corre-
sponding N-hydroxylamines. It is now clear that in aqueous
media such esters spontaneously eliminate a sulfate anion,
producing an arylnitrenium ion.9,10 It has further been demon-
Despite this biochemical interest there is very little detailed
experimental information on arylnitrenium ion structures. The
inherent instability of these species usually renders them
unsuitable for X-ray crystallography or high-resolution NMR.
(However, certain highly stabilized nitrenium ions have been
examined with these techniques.15) Most of the knowledge
regarding reactive arylnitrenium ion structures comes from quan-
tum chemical calculations.16-23 Phenylnitrenium ion (PhNH+)
is the simplest example of the arylnitrenium ions and has thus
† Johns Hopkins University.
‡ University of Minnesota.
§ University of Maryland.
(1) Simonova, T. P.; Nefedov, V. D.; Toropova, M. A.; Kirillov, N. F.
Russ. Chem. ReV. 1992, 61, 584-599.
(9) Novak, M.; Kahley, M. J.; Lin, J.; Kennedy, S. A.; James, T. G. J.
Org. Chem. 1995, 60, 8294-8304.
(2) McClelland, R. A.; Gadosy, T. A.; Ren, D. Can. J. Chem. 1998, 76,
1327-1337.
(10) Novak, M.; VandeWater, A. J.; Brown, A. J.; Sanzebacher, S. A.;
Hunt, L. A.; Kolb, B. A.; Brooks, M. E. J. Org. Chem. 1999, 64, 6023-
6031.
(3) McClelland, R. A. Tetrahedron 1996, 52, 6823-6858.
(4) Abramovitch, R. A.; Jeyaraman, R. In Nitrenium Ions; Scriven, E.
F. V., Ed.; Academic: Orlando, FL, 1984; pp 297-357.
(5) Kadlubar, F. F. In DNA Adducts of Carcinogenic Amines; Hemminki,
K., Dipple, A., Shuker, D. E. G., Kadlubar, F. F., Segerba¨ck, D. and Bartsch,
H., Ed.; Oxford University Press: Oxford, UK, 1994; pp 199-216.
(6) Humphreys, W. G.; Kadlubar, F. F.; Guengerich, F. P. Proc. Natl.
Acad. Sci. U.S.A. 1992, 89, 8278-8282.
(11) Novak, M.; Kennedy, S. A. J. Am. Chem. Soc. 1995, 117, 574-
575.
(12) Meier, C.; Boche, G. Tetrahedron Lett. 1990, 31, 1693-1696.
(13) Famulok, M.; Boche, G. (University of Marburg, Germany). Angew.
Chem., Int. Ed. Engl. 1989, 28, 468-469.
(14) Hoffman, G. R.; Fuchs, R. P. P. Chem. Res. Toxicol. 1997, 10, 347-
359.
(7) Novak, M.; Kennedy, S. A. J. Phys. Org. Chem. 1998, 11, 71-76.
(8) McClelland, R. A.; Ahmad, A.; Dicks, A. P.; Licence, V. J. Am.
Chem. Soc. 1999, 121, 3303-3310.
(15) Boche, G.; Andrews, P.; Harms, K.; Marsch, M.; Rangappa, K. S.;
Schimeczek, M.; Willecke, C. J. Am. Chem. Soc. 1996, 118, 4925-4930.
10.1021/ja001184k CCC: $19.00 © 2000 American Chemical Society
Published on Web 08/11/2000

8272 J. Am. Chem. Soc., Vol. 122, No. 34, 2000
Scheme 1. Nitrenium Ion Structure and Behavior
SriVastaVa et al.
Chart 1. Nitrenium Ions
that can be conveniently calculated from quantum mechanics
(singlet-triplet state energy gaps, equilibrium geometries, etc.)
are extremely difficult to determine experimentally, especially
for transient molecules in solution.
been the focus of several theoretical investigations. A recent
computational study of phenylnitrenium ion using density
functional theory (DFT) predicts this species to be a planar
singlet in its ground state with the lowest energy triplet state
21.2 kcal/mol higher in energy.16 An analysis of the computed
geometry reveals the singlet to possess a shortened C-N bond
and significant bond alternation in the phenyl ring. This result,
together with the predicted charge distribution, suggests that
the structure of PhNH+ resembles the 4-iminocyclohexa-2,5-
dienyl structure 1b more closely than structure 1a, where the
positive charge is localized on N (Scheme 1). The latter is
consistent with the observation that aromatic nitrenium ions react
with nucleophiles (e.g., N3-, Cl-, H2O) on the aromatic ring
rather than on nitrogen24-26 (although for reasons that remain
obscure, guanine, glutathione, and some aromatic amines are
observed to react at nitrogen).
Experimentally, it is possible to directly detect reactive
arylnitrenium ions using laser flash photolysis (LFP) and related
photochemical methods. In its traditional implementation, the
LFP experiment provides a time-resolved UV-visible absorp-
tion spectrum of the transient species of interest (LFP-TRUV).
From this it is possible to obtain accurate lifetimes for the
intermediate and rate constants for its reactions with added traps.
A number of substituted arylnitrenium ions have been character-
ized in this manner, although the parent system, PhNH+ remains
elusive. Generally speaking, arylnitrenium ions are observed
to react with anionic nucleophiles (N3-, Cl-, etc.) at or near
the diffusion limit. Neutral nucleophiles such as alcohols and
water react several orders of magnitude more slowly, with rate
constants that are sensitive to substitution on the ring of the
arylnitrenium ion.3,26-29
Recently, several of us have communicated results from time-
resolved infrared laser flash photolysis (LFP-TRIR) studies on
diphenylnitrenium ion.30 The LFP-TRIR experiment is concep-
tually similar to LFP-TRUV in that the transient intermediates
are generated using a short pulse of laser light. The TRIR
experiment, however, uses IR absorption rather than UV-vis
to detect the transient species. Thus, it is possible to detect key
IR bands for the transient species and compare them to spectra
derived from quantum calculations. Unlike the situation with
UV-vis spectra, it is a relatively simple matter to compute IR
spectra for ground-state molecules. In the case of Ph2N+ the
TRIR experiment revealed several bands. The band at 1392
cm-1 is assigned to a composite C-NdC stretch on the basis
of isotope substitution experiments. Detected in the same
experiment was a band at 1568 cm-1 due to an aromatic ring
stretching.
Herein we describe experimental TRIR and TRUV studies
on a series of substituted methylarylnitrenium ions (2-5, Chart
1). These measurements are compared with spectra for the
singlet states calculated using state-of-the-art DFT methods.
Excellent quantitative agreement between theoretically calcu-
lated and experimentally measured IR spectra is obtained. This
study yields three major conclusions regarding the nitrenium
ions. First, the observation of the characteristic IR bands in the
TRIR, coupled with the isolation of nitrenium ion products,
removes any doubt that arylnitrenium ions are the direct products
from photolysis of the 1-aminopyridinium ions. Second, the
calculated IR spectra for the singlet state nitrenium ions show
a much better match with the experimental IR spectra than the
calculated spectra for the corresponding triplet-state nitrenium
ions. This confirms the long-held belief that simple aromatic
nitrenium ions 2-5 are ground-state singlets, rather than triplets.
Finally, analysis of the stretching frequencies and the theoreti-
cally derived bond lengths shows that the arylnitrenium ions
all possess considerable quinoidal character (i.e., their structures
are closer to 1b than 1a). This tendency is enhanced with
π-donating substitutents.
One of the problems that has plagued studies of nitrenium
ions, and indeed all short-lived intermediates, is finding
meaningful ways of comparing theoretical calculations with
experiments. The sorts of data readily available from experi-
ments (UV-vis absorption spectra, reaction rate constants, stable
product distributions) are notoriously difficult and/or expensive
to compute to the requisite accuracy. Likewise, the parameters
(16) Cramer, C. J.; Dulles, F. J.; Falvey, D. E. J. Am. Chem. Soc. 1994,
116, 9787-9788.
(17) Novak, M.; Lin, J. J. Org. Chem. 1999, 64, 6032-6040.
(18) Cramer, C. J.; Falvey, D. E. Tetrahedron Lett. 1997, 38, 1515-
1518.
Results
(19) Falvey, D. E.; Cramer, C. J. Tetrahedron Lett. 1992, 33, 1705-
1. Synthesis of Photochemical Precursors to the Nitrenium
Ions. The N-methyl-N-arylnitrenium ions (2-5) are generated
through photolysis of appropriate 1-aminopyridinium salt
derivatives.31 The latter can be prepared using the synthetic route
in Scheme 2. The appropriate N-methylaniline derivative
(7-10) is first nitrosated with HCl/NaNO2 to give N-nitrosoa-
niline derivatives (11-14) and then reduced with Zn, providing
the 1,1-disubstituted hydrazine derivative (e.g., 19). This was
then condensed with 2,4,6-trimethylpyrylium tetrafluoroborate
1708.
(20) Ford, G. P.; Herman, P. S. J. Am. Chem. Soc. 1989, 111, 3987-
3996.
(21) Ford, G. P.; Herman, P. S.; Thompson, J. W. J. Comput. Chem.
1999, 20, 231-243.
(22) Li, Y.; Abramovitch, R. A.; Houk, K. N. J. Org. Chem. 1989, 54,
2911-2914.
(23) Sullivan, M. B.; Brown, K.; Cramer, C. J.; Truhlar, D. G. J. Am.
Chem. Soc. 1998, 120, 11778-11783.
(24) Gassman, P. G.; Campbell, G. A. J. Am. Chem. Soc. 1971, 93,
2567-2569.
-
(25) Gassman, P. G.; Campbell, G. A.; Frederick, R. C. J. Am. Chem.
Soc. 1972, 94, 3884-3891.
(20), providing the BF4 salts of 15-18. The latter can be
isolated as crystalline solids and were characterized by the
customary spectroscopic methods. In the case of 15 the
hydrazine derivative 19 was isolated and characterized. For the
(26) Ren, D.; McClelland, R. A. Can. J. Chem. 1998, 76, 78-84.
(27) Robbins, R. J.; Yang, L. L.-N.; Anderson, G. B.; Falvey, D. E. J.
Am. Chem. Soc. 1995, 117, 6544-6552.
(28) Sukhai, P.; McClelland, R. A. J. Chem. Soc., Perkin Trans. 2 1996,
1529-1530.
(30) Srivastava, S.; Toscano, J. P.; Moran, R. J.; Falvey, D. E. J. Am.
Chem. Soc. 1997, 119, 11552-11553.
(31) Balaban, A. T. Tetrahedron 1968, 24, 5059-5065.
(29) Robbins, R. J.; Laman, D. M.; Falvey, D. E. J. Am. Chem. Soc.
1996, 118, 8127-8135.

Structures of ReactiVe Nitrenium Ions
Scheme 2. Precursor Synthesis
J. Am. Chem. Soc., Vol. 122, No. 34, 2000 8273
Scheme 3. Photolysis of Precursors and Stable Products
Figure 1. Transient UV-vis absorption spectra generated from LFP
(308 nm, 10 ns 20 mJ/pulse) of 11 in CH3CN taken 2, 5, 10, 20, and
40 µs after the laser pulse.
Scheme 4. Mechanism of Chloride Addition
other members of this series, it proved more efficient to trap
the corresponding hydrazine derivatives in situ by adding an
appropriate amount of the pyrylium salt to the reaction mixture.
2. Stable Photoproducts. The photolysis of 1-aminopyri-
dinium salts has been shown to produce nitrenium ions through
heterolytic scission of the N-N bond. This has previously been
demonstrated for some 1-(N,N-diarylamino)pyridinium ions,32
the 1-aminopyridinum ions,33,34 and several other related
systems.35-40 The precursors used in this study show photo-
chemical behavior consistent with this generalization. Each
decomposes rapidly when irradiated with UV light to produce
2,4,6-trimethylpyridine 21 (Scheme 3). This product was
observed by both GC/MS and 1H NMR of the photolysis
mixtures. Formation of this product verifies that the anticipated
N-N bond scission occurs.
Scheme 5. Products from the Chloro/MeOH
In addition to 21, products characteristic of nitrenium ions
were also detected by GC/MS. Specifically, photolysis of 15-
18 in the presence of ca. 5-20 mM chloride ion gives chloride
adducts 22-25. (Scheme 3) These result from addition of Cl-
at the ring position ortho to the nitrenium ion center. Nucleo-
philes such as chloride are also known to add para to the
nitrenium ion center.25,41 However, when there is a substituent
in this position, we assume that the initial para adduct (29) is
formed reversibly, readily eliminating the labile chloride ion.
The initial ortho adduct 30, however, can readily tautomerize
to give the stable aniline derivative 22-25. This mechanism is
illustrated in Scheme 4. The lability of the chloride in structure
29 is further established by photolysis of the 4-chloro derivative
Scheme 6. Chloro/Methoxy Mechanism
16 in MeOH (Scheme 5). In this case a mixture of N-methyl-
4-chloro-2-methoxyaniline 31 and its isomer, N-methyl-4-
methoxy-2-chloroaniline 24, is obtained. The latter is the result
of initial attack of MeOH on the para position, followed by
elimination of Cl- and subsequent re-addition of the latter at
the ortho position, as illustrated in Scheme 6.
(32) Moran, R. J.; Falvey, D. E. J. Am. Chem. Soc. 1996, 118, 8965-
8966.
(33) Srivastava, S.; Kercher, M.; Falvey, D. E. J. Org. Chem. 1999, 64,
5853-5857.
3. LFP-TRUV. Pulsed laser (355 nm, 4 mJ, 10 ns) photolysis
of the 4-phenyl precursor 15 in CH3CN gives a short-lived
visible absorption band with λmax ) 460 nm, as shown in Figure
1. The species responsible for this band has a lifetime of 24 (
4 µs in dry CH3CN.42 It is assigned to N-methyl-N-(4-biphenylyl)-
nitrenium ion 2 on the basis of the following observations. (1)
Addition of the nucleophiles Cl- and MeOH to the LFP samples
(34) Takeuchi, H.; Higuchi, D.; Adachi, T. J. Chem. Soc., Perkin Trans.
2 1991, 1525-1529.
(35) Takeuchi, H.; Koyama, K. J. Chem. Soc., Perkin Trans. 1 1988,
2277-2281.
(36) Takeuchi, H.; Hayakawa, S.; Tanahashi, T.; Kobayashi, A.; Adachi,
T.; Higuchi, D. J. Chem. Soc., Perkin Trans. 2 1991, 847-855.
(37) Abramovitch, R. A.; Evertz, K.; Huttner, G.; Gibson, H. H.; Weems,
H. G. J. Chem. Soc., Chem. Commun. 1988, 325-327.
(38) Abramovitch, R. A.; Shi, Q. Heterocycles 1994, 37, 1463-1466.
(39) Takeuchi, H.; Watanabe, K. J. Phys. Org. Chem. 1998, 11, 478-
484.
(42) Most of the N-methyl-N-arylnitrenium ions generated in this way
decay following first-order kinetics. However, in the absence of trapping
agents, the product mixtures are complicated, and the lifetimes probably
reflect more than one decay pathway, including hydride shift from the methyl
group to the nitrenium center. Chiapperino, D.; Falvey, D. E. J. Phys. Org.
Chem. 1997, 10, 917-924.
(40) Takeuchi, H.; Hayakawa, S.; Murai, H. J. Chem. Soc., Chem.
Commun. 1988, 1287-1289.
(41) Gassman, P. G.; Campbell, G. A. J. Am. Chem. Soc. 1972, 94,
3891-3896.

8274 J. Am. Chem. Soc., Vol. 122, No. 34, 2000
SriVastaVa et al.
Table 1. Observed Rates of Decay for 4-Substituted
N-Methylphenylnitrenium Ions 2-5 in Acetonitrile (TRUV) and in
Acetonitrile-d3 (TRIR)
kobs (s-1
)
by TRUV
by TRIR
2
3
4
5
4.16 × 104
1.05 × 106
1 × 104 a
1.21 × 105
1.25 × 106
5.98 × 104
4.92 × 105
1.35 × 105
a Decay does not follow first-order kinetics; the half-life is
>100 µs.
Table 2. Electronic State and Zero-Point Vibrational Energies and
Singlet-Triplet Splittings for Para-Substituted
Methylphenylnitrenium Ionsa
Figure 2. TRIR difference spectra observed over the first 1.0 µs
following laser photolysis (266 nm, 10 ns, 0.4 mJ) of pyridinium salts
15-18 in argon-saturated acetonitrile-d3.
para
electronic
state
ZPVE
(kcal/mol)
S-T
substituent
E
splittingb
Me
1A′
3A
-365.295 75
-365.266 85
-785.596 96
-785.568 72
-557.041 24
-557.010 20
-440.521 65
-440.485 88
0.154 75
0.152 93
0.119 34
0.117 52
0.207 87
0.205 86
0.160 61
0.158 74
-17.0
-17.7
-18.2
-21.3
diminishes its lifetime. A pseudo-first-order kinetic analysis
provides rate constants of 8 × 109 and 3.7 × 105 M-1 s-1 for
Cl- and MeOH, respectively. Diffusion-limited reaction with
halides and slower reactions with alcohols are characteristic of
arylnitrenium ions.27,29,43 (2) The absorption band is of a similar
shape and position to those reported by McClelland et al. for
structurally similar 4-biphenylylnitrenium ions.44-46 (3) Life-
times measured in O2- and N2-purged solution are indistinguish-
able. Arylnitrenium ions react slowly, if at all, with O2.
The other derivatives 3-5 were also studied by TRUV. In
these cases, lower wavelength excitation (266 or 308 nm) was
used to generate the nitrenium ion. The 4-chloro derivative (3
generated from 16) absorbs at 340 nm, and its lifetime is 950
ns in CH3CN. This species reacts with Cl- at the diffusion limit
(2.8 × 1010 M-1 s-1) and more slowly with MeOH (6.8 × 107
M-1 s-1). The 4-methoxy derivative (4 generated from 17)
absorbs at 320 nm. This spectrum shows good agreement with
those reported by McClelland for similar 4-alkoxynitrenium
ions.28,47 In the absence of added traps, nitrenium ion 4 lives
for over 100 µs and decays in a kinetically complex manner.
When the nucleophilic trap Cl- is added, the decay is observed
to conform to first-order kinetics and a rate constant of 1.7 ×
1010 M-1 s-1 is measured. The 4-methyl derivative (5 from 18)
shows an absorption maximum at 330 nm and a lifetime of 7.4
µs in CH3CN. It, too is observed to react with Cl- near the
diffusion limit (2.3 × 1010 M-1 s-1) and more slowly with
MeOH (6.0 × 107 M-1 s-1).
Cl
1A′
3A′′
1A
Ph
3A
MeO
1A′
3A
a BPW91/cc-pVDZ level; electronic energies and ZPVE in hartrees,
S-T splittings (including ZPVE) in kcal/mol. b A negative splitting
indicates the singlet state is below the triplet.
consistent with those experimentally determined from TRUV-
LFP studies (Table 1). The observed rate constants for nitrenium
ion decay were found to increase significantly upon continuous
irradiation of the solution, presumably due to trapping of the
nitrenium ion by stable products photogenerated during the
experiment. One possibility for this trapping agent is 2,4,6-
trimethylpyridine, which is expected to trap the nitrenium ions
at or near the diffusion limit. Thus, values given in Table 1
represent upper limits for the observed decay rates of 2-5 in
argon-saturated acetonitrile-d3 solutions. The lifetimes follow
the expected trend for a series of 4-substituted cations, with
the 4-methoxy derivative 4 being the most stable (τ ) 16.7 µs)
and the 4-chloro-derivative 3 having the shortest lifetime (τ )
0.80 µs). (3) The decay rate constants are unaffected by the
presence of O2. (4) The signal for biphenylylnitrenium ion 2 is
quenched by nucleophiles such as chloride ion. A pseudo-first-
order kinetic analysis using the 1584 cm-1 IR band provides a
value of 1.4 × 1010 M-1 s-1 for kCl-, in good agreement with
the value obtained from TRUV analysis.
4. LFP-TRIR. Figure 2 shows the TRIR difference spectra
observed from 1700 to 1500 cm-1 over the first 1.0 µs following
laser photolysis of the 1-aminopyridinium salts 15-18 in Ar-
saturated CD3CN. The negative bands observed near 1640 and
1520 cm-1 in Figure 2 are due to the depletion of the pyridinium
precursors. The positive bands at 1628-1584 cm-1 are assigned
to the substituted singlet N-methyl-N-phenylnitrenium ions 2-5
on the basis of the following. (1) The experimentally observed
frequencies are in excellent agreement with the aromatic CdC
symmetric stretching frequencies calculated using density
functional theory (see below). (2) The observed rate of decay
for each of these signals indicates that the lifetimes of 2-5 are
5. DFT Calculations. The absolute energies, zero-point
energies, and singlet-triplet splittings for the equilibrium
geometries of the four para-substituted N-methyl-N-phenylni-
trenium ions (2-5) are provided in Table 2. Computed harmonic
IR frequencies and intensities for two different normal modes
in both spin states of the four nitrenium ions are provided in
Table 3; the modes are best characterized as the C-N stretching
mode and the symmetric combination of the two aromatic
(H)C-C(H) stretches (animations of both modes for 2-5 are
provided as Supporting Information). The latter mode is
computed to be quite intense and corresponds to the mode
(43) Davidse, P. A.; Kahley, M. J.; McClelland, R. A.; Novak, M. J.
Am. Chem. Soc. 1994, 116, 4513-4514.
observed experimentally in the region just above 1600 cm-1
.
(44) McClelland, R. A.; Kahley, M. J.; Davidse, P. A. J. Phys. Org.
Chem. 1996, 9, 355-360.
The C-N stretch is computed to be much less intense and is
not observed in the experimental spectrasit is discussed below,
however, because it provides a different measure of the degree
of quinoidal character in the arylnitrenium ion. Yet another
measure of quinoidal character is provided by the degree of
bond length alternation about the aromatic ring.5 Heavy atom
(45) McClelland, R. A.; Kahley, M. J.; Davidse, P. A.; Hadzialic, G. J.
Am. Chem. Soc. 1996, 118, 4794-4803.
(46) McClelland, R. A.; Davidse, P. A.; Hadzialic, G. J. Am. Chem. Soc.
1995, 117, 4173-4174.
(47) Ramlall, P.; McClelland, R. A. J. Chem. Soc., Perkin Trans. 2 1999,
225-232.

Structures of ReactiVe Nitrenium Ions
J. Am. Chem. Soc., Vol. 122, No. 34, 2000 8275
Table 3. Selected Harmonic Vibrational Frequencies (cm-1) and
Intensities (km/mol) for Para-Substituted Methylphenylnitrenium
Ionsa
al.43,47 for similar arylnitrenium ions which lack only the
N-methyl substituent.
Any remaining doubt that the predicted nitrenium ions are
generated is removed by the observation of the characeristic
IR bands when photolysis of the same precursors is carried out
(Table 1). In the LFP-TRIR experiments, the observed lifetimes
are consistently shorter than those measured by LFP-TRUV.
This quantitative disagreement is the result of the requirement
that the TRIR experiment be conducted at concentrations 20
times higher than those used in the TRUV experiments. We
have observed that the lifetime of the nitrenium ion diminishes
as the initial concentration of its photoprecursor is increased.
In any case, the same general trend is observed with both data
sets, where the decay rate constants follow the order 3 > 5 >
2 > 4.
frequency (intensity)
para
substituent
electronic
state
C-N stretch
sym arom C-C stretch
Me
1A′
1531 (29)
1465 (2)
1524 (24)
1465 (0.4)
1535 (18)
1619 (287) [1616]b
3A
1592 (136)
Cl
1A′
3A′′
1A
1604 (269) [1604]b
1575 (96)
Ph
1619 (183) [1612]b
1601 (480) [1584]b,c
1593 (10)
3A
1446 (76)
1596 (305) c
MeO
1A′
1550 (9)
1460 (7.7)
1629 (241) [1628]b
1607 (90)
3A
a BPW91/cc-pVDZ level. b Experimental value. c Normal mode for
the substituent ring.
Arylnitrenium ions have been predicted to possess significant
iminocyclohexadienyl cation-like character (Scheme 1).30 The
most obvious manifestation of this comes from the computed
Cipso-N bonds lengths (r8 in Table 4). For the singlets r8 )
1.320 ( 0.003 Å, placing it much closer to typical values for
CdN (1.28 Å) than for C-N (1.40 Å).51 The data from the
4-substituted phenylnitrenium ions 2-5 allow an examination
of electronic factors that influence the structure and reactivity
of phenylnitrenium ions. IR spectra (Figure 2) clearly show a
shift in the aromatic CdC stretch that is sensitive to substitution
in the 4-position. Introduction of electron-donating groups
(MeO, Me, Ph) at the 4-position enhances delocalization of the
positive charge from the N atom into the ring. This results in
the nitrenium ions having more 4-iminocyclohexa-2,5-dienyl
cation-like character with a subsequent upward shift in fre-
quency. 4-Chloro-substituted nitrenium ion 3 has an aromatic
CdC stretch centered at 1604 cm-1. The presence of this
electron-withdrawing group results in a greater localization of
the positive charge on the N atom, resulting in an observed
aromatic CdC stretch for 3 that is nearly identical with that
for 4-chloroaniline.52
With respect to theory, the data in Table 2 show the expected
qualitative trend in singlet-triplet splittings as a function of
para substitution, namely that as the π-donating character of
the substituent increases, the singlet state is increasingly
stabilized over the triplet. A more complete computational study
of this phenomenon has been published elsewhere for several
4-substituted-N-protiophenylnitrenium ions.23 For the purposes
of this work, which focuses on IR stretching frequencies as
reporters of quinoidal character, we simply emphasize that in
all four experimentally characterized cases the singlet state is
predicted to lie substantially below the triplet. Given that these
species were generated by direct photolysis, it is highly unlikely
that the observed IR spectra correspond to a triplet transients
this point is also supported by the computed IR frequencies, as
will be discussed next.
While analytical computation of the normal modes provides
the full 3N - 6 frequencies expected for nonlinear molecules
having N atoms, we focus here on only two modes that are
particularly diagnostic of structural differences between the
differently substituted nitrenium ions. The C-N stretching
frequency is of interest to the extent that it reflects imine
character in the nitrenium ion (increasing imine/quinoidal
character would be expected to move this stretch to a higher
frequency), but the stretch does not appear to be observable in
the experimental spectra. This is consistent with the rather low
Figure 3. Molecular geometries for para-substituted N-methyl-N-
phenylnitrenium ion singlets (above) and triplets (below). Substituents
from left to right are p-methyl, p-chloro, p-phenyl, and p-methoxy.
bond lengths for both states of the four nitrenium ions are
collected in Table 4. Figure 3 illustrates the complete molecular
geometries as ball-and-stick models.
Discussion
The formation of nitrenium ions through the photolysis of
1-aminopyridinium ions is well precedented. Takeuchi and
Abramovitch first demonstrated this route several years
ago.34-36,38,48 Recently we have used it for LFP investigations
of diphenylnitrenium ion and several related species.33,49,50 The
particular 1-aminopyridinium ions studied here follow the
expected behavior. Photolysis of 15-18 in the presence of
nucleophiles results in the formation of the anticipated ring
adducts 22-25. LFP-TRUV experiments also show typical
nitrenium ion UV-vis absorptions. In the case of the 4-phenyl
derivative 2 and the 4-methoxy derivative 4, the TRUV spectra
show good agreement with spectra reported by McClelland et
(48) Abramovitch, R. A.; Beckert, J. M.; Chinnasamy, P.; Xiaohua, H.;
Pennington, W.; Sanjivamurthy, A. R. V. Heterocycles 1989, 28, 623-
628.
(49) Moran, R. J.; Cramer, C. J.; Falvey, D. E. J. Org. Chem. 1996, 61,
3195-3199.
(50) Chiapperino, D.; Anderson, G. B.; Robbins, R. J.; Falvey, D. E. J.
Org. Chem. 1996, 61, 3195-3199.
(51) March, J. AdVanced Organic Chemistry, 2nd ed.; McGraw-Hill:
New York, 1977.
(52) The Aldrich Library of Infrared Spectra; Pouchert, C. J., Ed.; Aldrich
Chemical Co.: Milwaukee, WI, 1981; p 723.

8276 J. Am. Chem. Soc., Vol. 122, No. 34, 2000
SriVastaVa et al.
Table 4. CNC Valence Angle (deg) and Heavy-Atom Bond Lengths (Å) for Para-Substituted Methylphenylnitrenium Ionsa
para
substituent
electronic
state
r1
r2
r3
r4
r5
r6
r7
r8
r9
Meb
1A′
3A
1.487
1.494
1.705
1.715
1.454
1.461
1.316
1.322
1.440
1.430
1.433
1.422
1.446
1.433
1.445
1.432
1.427
1.426
1.424
1.375
1.382
1.376
1.383
1.373
1.382
1.367
1.377
1.380
1.384
1.379
1.459
1.445
1.458
1.444
1.458
1.438
1.463
1.442
1.458
1.444
1.458
1.323
1.333
1.323
1.334
1.321
1.345
1.317
1.342
1.422
1.413
1.421
1.413
1.428
1.422
1.428
1.419
Cl
1A′
3A′′
1A
Ph
1.438
1.433
1.434
1.427
1.374
1.382
1.373
1.383
1.457
1.438
1.462
1.441
3A
MeOc
1A′
3A
a BPW91/cc-pVDZ level. Bond lengths identical to others by symmetry are not reported. See Figure 3 for ball-and-stick figures. b One C-H
bond of the substituent methyl group eclipses the C-C bond labeled r3 in both states. c The O-C bond of the substituent methoxy group eclipses
the C-C bond labeled r3 in both states.
intensity predicted for this mode by the computations (Table
3). However, a C-C stretching mode in the aromatic ring
(symmetric combination of bonds r4 and r5 as labeled in Table
4) provides a similar measure of quinoidal character, where
again the frequency is expected to move to higher frequency
with increasing imine character. This absorption is predicted to
be much more intense, suggesting that it is, indeed, the mode
observed in the experimental spectra.
For the 4-methyl (5), -chloro (3), and -methoxy (4)-substituted
cases, we observe nearly quantitative agreement between the
experimentally observed IR frequencies in the region just above
1600 cm-1 and those computed for the symmetric aromatic
stretching mode of the singlet state at the BPW91/cc-pVDZ
level. Indeed, the average deviation between theory and experi-
ment is only 1 cm-1! In the p-phenyl case (2) there are two
observed frequencies against which to compare. Agreement
between theory and experiment is still very good for the higher
frequency mode and somewhat less satisfactory for the lower
frequency mode. Nevertheless, the average deviation between
theory and experiment over all five observed modes is only 6
cm-1, which is remarkably close agreement.
Comparison of the analogous modes for the triplet states of
the four experimentally characterized nitrenium ions gives an
average error (unsigned) of 21 cm-1, again suggesting that the
observed spectra do not arise from triplet transients. As a last
point in this regard, for the biphenylylnitrenium (2) triplet it is
predicted that the aromatic stretching modes for both rings would
overlap in the experimental spectrum. The observed spectrum,
however, shows two distinct peaks in this region, with the lower
wavelength absorption 18 cm-1 below the higher and about
twice as intense. This is in excellent agreement with the
theoretical predictions for both the separation between the peaks
(28 cm-1) and the relative intensities (roughly 2.6:1).
The trend in the IR frequencies is exactly that expected from
qualitative considerations. Improved π-donating character of the
para substituent leads to a shift of the arylnitrenium aromatic
stretch absorption to higher frequency. From the chloro to the
methoxy substituent, the observed shift is 25 cm-1stheory
predicts 24 cm-1. In sum, the agreement between theory and
experiment appears to be definitive with respect to the nature
and behavior of this normal mode and the assignment of the
observed transients to the corresponding nitrenium ion singlet
state.
differently substituted arylnitrenium ions. Bond alternation in
a quinoidal sense should give rise to relatively longer bonds
r2, r3, r6, and r7 and shorter bonds r4 and r5 (Table 4). In
addition, since π-donation is expected to be more stabilizing
for the singlet state than for the triplet (since the conjugated
nitrogen p orbital is formally empty in the former state but half-
filled in the latter), one might expect to see shorter bonds r1
and r8 in the singlet state than in the triplet state for a given
substitution pattern. Finally, since r8 and r9 are the same kinds
of bonds in all four substituted cases, one might predict that
increasing quinoidal character with improved π-donation would
shorten r8 and lengthen r9 (since hyperconjugation by the
methyl group becomes less important as a stabilizing effect)
across the substitution series.
All of these effects are observed in the computed geometries,
although they are not as dramatic as those seen in the IR
frequencies. In the 4-chloro (3) singlet, the average of bond
lengths r4 and r5 is 1.378 Å, and this shortens to 1.370 Å in
the 4-methoxy (4) singlet. For those respective singlets, the
averages for the remaining aromatic bond lengths are 1.443 and
1.451 Å. Thus, the degree of bond alternation (the difference
between these two averages) increases with increasing π-donat-
ing character of the para substituent. This effect does exist in
the triplet as well. However, the difference in average bond
lengths is reduced in magnitude by about 0.02 Å. The triplets
also show longer bonds from the aromatic ring to the substituents
(r1) and to the nitrenium nitrogen (r8), in each case by about
0.01 Å. The latter effect is particularly noteworthy because, in
the absence of conjugation effects, bond lengths to triplet
nitrenium nitrogen atoms are typically shorter than bond lengths
to singlet atoms. This is due to the wider bond angle (and hence
greater s character in σ bonding orbitals) adopted by the triplets
this hybridization effect is observed for r9, which is shorter for
the triplets than for the singlets by about 0.01 Å.53,54 With
increased π-donation, bond lengths r8 and r9 do move in the
expected directions; from 4-chloro to 4-methoxy, r8 decreases
by 0.006 Å while r9 increases by 0.007 Å.
One final note regarding geometry concerns the rotation about
the Cipso-N bond in the arylnitrenium ions. In the singlet state,
the N-CH3 bond is coplanar with the aromatic system, as
(53) Lim, M. H.; Worthington, S. E.; Dulles, F. E.; Cramer, C. J. In
Density Functional Calculations of Radicals and Diradicals; Laird, B. B.,
Ross, R. B., Ziegler, T., Ed.; American Chemical Society: Washington,
DC, 1996.
(54) Cramer, C. J.; Dulles, F. J.; Storer, J. W.; Worthington, S. E. Chem.
Phys. Lett. 1994, 218, 387-394.
The computations permit an examination of the predicted
geometries to assess the degree to which the changes in IR
frequencies reflect changes in the relative geometries of the

Structures of ReactiVe Nitrenium Ions
J. Am. Chem. Soc., Vol. 122, No. 34, 2000 8277
(400 MHz, CDCl3) δ 7.69 (s, 2 H), 7.56 (d, J ) 7.7 Hz, 2 H), 7.52 (d,
J ) 8.8 Hz, 2 H), 7.43 (t, J ) 7.7 Hz, 2 H), 7.33 (t, J ) 7.7 Hz, 1 H),
6.40 (br d, J ) 8.8 Hz, 2 H), 3.67 (s, 3 H), 2.67 (s, 3 H), 2.63 (s, 6 H);
13C NMR (100 MHz, CDCl3) δ 161.3, 158.3, 142.9, 139.7, 134.7, 129.7,
129.1, 128.9, 127.2, 126.6, 110.8, 38.4, 22.2, 19.3; MS (FAB) m/z
(relative intensity) 303 ([M - BF4]+, 46), 182 (100); HRMS (FAB)
calcd for C21H23N2 [M - BF4]+ 303.1861, found 303.1858.
expected since this maximizes conjugation and allows electron
delocalization onto the aggressively π-withdrawing nitrogen
atom. In the triplet state, however, the N-CH3 bond is always
predicted to be rotated 90° out of the aromatic plane. This
presumably relieves some steric congestion associated with the
coplanar geometry. In addition, since the triplet nitrenium ions
have valence bond angles at nitrogen typically in the range of
130-140°, overlap with the formally sp2 hybrid at nitrogen is
only somewhat reduced compared to the pure p orbital on that
atom. As both are half-filled and the former is of lower energy
by virtue of its s character, there is not a strong driving force to
remain coplanar. This bond rotation is computed to have a very
low barrierscoplanar triplets within the systems studied here
are typically only 1 or 2 kcal mol-1 higher in energy than the
equilibrium structures. These results, however, prompted us to
reexamine the unsubstituted parent system, N-methyl-N-phe-
nylnitrenium ion, which was previously predicted by two of us
to have a coplanar structure.16 The coplanar structure, in fact,
is a transition state for rotation about the Cipso-N bond, but the
barrier is a mere 0.02 kcal mol-1; i.e., the situation is that of a
free rotor. The observation that singlet and triplet arylnitrenium
ions possess very different geometries suggests that it might be
possible to engineer a specific system with a long-lived spin
isomerism. In such a system the triplet state would be lowest
in energy at its equilibrium geometry while the singlet state
would be lowest in energy at its equilibrium geometry. Berson
and co-workers have recently proposed similar long-lived spin
isomerism for a substituted non-Kekule´ diradical.55 Likewise,
Bally and McMahon have observed a similar spin isomerism
for 2-naphthyl(carbomethoxy)carbene. In the latter case the
triplet is planar and the singlet is nonplanar.56
1-(N-(4-Chlorophenyl)-N-methylamino)-2,4,6-trimethylpyridin-
ium Tetrafluoroborate (16). In a solvent mixture of 15 mL each of
water, HOAc, and ethanol stirred on an ice bath, 1.23 g (7.21 mmol)
of 4-chloro-N-methyl-N-nitrosoaniline (12, see Supporting Information)
was added followed by 4.5 g of finely powdered Zn. After 30 min of
stirring, the solution was filtered to remove precipitated salts and then
neutralized with aqueous NaHCO3 and solid Na2CO3. The solution was
extracted three times with CH2Cl2, the combined organic extracts were
dried over MgSO4 and filtered, and the solvent was evaporated under
reduced pressure. To this crude product mixture, 10.0 mL of 100%
EtOH was added, followed by 0.670 g (3.19 mmol) of freshly prepared
20. The solution was stirred at room temperature for 40 min, cooled
on an ice bath, and added to 200 mL of cold diethyl ether. Light yellow
crystals were isolated by filtration, giving 0.930 g (2.66 mmol, 84%)
1
of 16: mp (EtOH) 178-180 °C; H NMR (400 MHz, CDCl3) δ 7.69
(s, 2 H), 7.29 (d, J ) 8.9 Hz, 2 H), 7.30 (br, 2 H), 3.61 (s, 3 H), 2.65
(s, 3 H), 2.58 (s, 6 H); 13C NMR (100 MHz, CDCl3) δ 161.6, 158.1,
142.3, 130.4, 129.9, 126.9, 111.7, 38.6, 22.2, 19.1; MS (FAB) m/z
(relative intensity) 261 ([M - BF4]+, 18), 142 (30), 140 (100), 122
(95), 121 (30); HRMS (FAB) calcd for C15H18ClN [M - BF4]+
261.11584, found 261.11512.
1-(N-(4-Methoxyphenyl)-N-methylamino)-2,4,6-trimethylpyri-
dinium Tetrafluoroborate (17). Compound 17 was prepared in the
same manner described for the synthesis of 16. Starting with 1.83 g
(11.0 mmol) of nitroso compound 8, the crude mixture from Zn
reduction was combined with 0.840 g (4.00 mmol) of pyrillium salt
20 in EtOH, producing 1.18 g (3.44 mmol) of 17 as a yellow solid,
1
representing an 86% yield: H NMR (400 MHz, CDCl3) δ 7.70 (s, 2
Conclusions
H), 6.87 (d, J ) 9.1 Hz, 2 H), 6.31 (br d, 2 H), 3.75 (s, 3 H), 3.54 (s,
3 H), 2.62 (s, 3 H), 2.58 (s, 6 H); 13C NMR (100 MHz, CDCl3) δ
161.0, 158.3, 154.7, 137.7, 129.8, 115.7, 112.1, 55.7, 38.7, 22.0, 19.2;
MS (FAB) m/z (relative intensity) 257 ([M + H]+, 11), 136 (100), 122
(34), 121 (14); HRMS (FAB) calcd for C16H21N2O [M - BF4]
257.1654, found 257.1665.
The excellent quantitative agreement between the theoretically
and experimentally derived IR frequencies for singlet nitrenium
ions 2-5 leads to several important findings. First, it verifies
previous conclusions that arylnitrenium ions are the primary
products from the photolysis of 1-(N-arylamino)pyridinium ions.
Second, it demonstrates that the DFT calculations at the levels
presented here reliably reproduce key structural features of this
family of reactive intermediates. Third, it supports the theoretical
prediction that 2-5 have singlet ground states. Finally, both
the experimental IR frequencies and the DFT calculations
indicate that these arylnitrenium ions possess structures that are
well described as 4-iminocyclohexa-2,5-dienyl cations.
1-(N-Methyl-N-4-tolylamino)-2,4,6-trimethylpyridinium Tetraflu-
oroborate (18). Using the same procedures described above for 16,
4.40 g of unpurified N-methyl-N-nitroso-4-toluidine 14 (see Supporting
Information) was used in the Zn reduction, along with 2.61 g (12.4
mmol) of pyryllium salt 20. Filtration to collect the precipitate yielded
4.01 g (12.2 mmol, 98%) of 18 as bright yellow crystals: 1H NMR
(400 MHz, CDCl3) δ 7.69 (s, 2 H), 7.12 (d, J ) 8.5 Hz, 2 H), 6.22 (br
d, J ) 8.5 Hz, 2 H), 3.58 (d, J ) 1.1 Hz, 3 H), 2.65 (s, 3 H), 2.58 (s,
6 H); 13C NMR (100 MHz, CDCl3) δ 161.1, 158.4, 141.5, 131.2, 131.0,
129.7, 110.5, 38.4, 22.1, 20.3, 19.2; UV-vis (MeCN) 235, 272, 365
nm; MS (FAB) m/z (relative intensity) 241 ([M - BF4]+, 72), 122
(64), 121 (19), 120 (100), 91 (15); HRMS (FAB) calcd for C16H21N2
[M - BF4]+ 241.1705, found 241.1706.
LFP-TRUV. Data from laser flash photolysis with time-resolved
ultraviolet-visible absorption detection experiments were collected
using either an excimer laser (Questek 2120 with Xe/HCl gas providing
5-12 ns, 308 nm) or a Nd:YAG laser (Continuum Surelite II-10, 4-6
ns, 355 or 266 nm) as the pulsed excitation source. Excitation energies
were attenuated such that 5-20 mJ/pulse was used. During a given
experiment, the pulse energy varied by approximately 5%. The transient
absorptions were monitored using a probe beam from an Oriel 350-W
Xe arc lamp passed through the sample cuvette perpendicular to the
excitation beam. Transient waveforms were recorded with a LeCroy
9420 digital oscilloscope which digitizes at a rate of 1 point/10 ns with
a bandwidth of 350 MHz. Samples for pulsed irradiation were prepared
such that the OD of photolabile substrates was approximately 2.0 at
the excitation wavelength employed. Minimal sample depletion was
confirmed by steady-state UV spectrum measurement after the LFP
experiments.
Experimental and Computational Methods
1-(N-(4-Biphenylyl)-N-methylamino)-2,4,6-trimethylpyridini-
um Tetrafluoroborate (15). 1-Methyl-1-(4-biphenylyl)hydazine (19,
1.33 g, 7.01 mmol, see Supporting Information for its preparation and
characterization) was dissolved in 70 mL of 100% EtOH, and then
1.20 g (5.71 mmol) of freshly prepared 2,4,6-trimethylpyrylium
tetrafluoroborate 2057 was added to the solution. The reaction mixture
was stirred for 30 min, by which time a bright yellow precipitate had
formed. The mixture was cooled on an ice bath and then poured into
500 mL of cold diethyl ether to maximize precipitate formation. The
product was collected by filtration, and crystals were washed with cold
diethyl ether and then dried under reduced pressure. Based on the
limiting amount of 20, a 96% isolated yield (2.14 g, 5.48 mmol) of
bright yellow crystals of 15 was obtained: mp 190-192 °C; 1H NMR
(55) Bush, L. C.; Maksimovic, L.; Feng, X. W.; Lu, H. S. M.; Berson,
J. A. J. Am. Chem. Soc. 1997, 119, 1416.
(56) Zhu, Z.; Bally, T.; Stracener, L. L.; McMahon, R. J. J. Am. Chem.
Soc. 1999, 121, 2863-2874.
(57) Balaban, A. T.; Paraschiv, M. ReV. Roum. Chim. 1982, 27, 513-
521.

8278 J. Am. Chem. Soc., Vol. 122, No. 34, 2000
SriVastaVa et al.
Photoproduct Analysis. Photolyses for GC/MS analysis were carried
out on N2-purged solutions containing 6-11 mg of the stated 1-ami-
nopyridinium salt derivative (15-18) and 10-15 mM tetra-n-buty-
lammonium chloride in 3.0 mL of dry CH3CN. Photolyses were carried
out using the filtered (>295 nm) output from a 300-W Xe arc lamp.
Photolysis times ranged from 10 to 20 min. Sample were then directly
analyzed by (GC/MS) using an XTI-5 GC column (30 m, 5% phenyl)
with a 0.25-mm internal diameter and 0.25-µm film thickness. Argon
was used as the mobile phase. The method begins with 4 min at 50
°C, and then the temperature is increased from 50 to 300 °C over 15
min and held at 300 °C for an additional 10 min. Photoproducts 21-
25 and 31 were identified by comparison of their retention times and
EI-MS with those of authentically synthesized compounds (see Sup-
porting Information).
Prior studies16,18,53,54,64-7064-70 have demonstrated that this level of theory
predicts state splittings to within about 2 kcal/mol as compared to either
experiment or better-converged quantum mechanical calculations for
systems analogous to those studied here. Density functional calculations
were carried out using the Gaussian 94 program suite.71
Acknowledgment. We thank the Chemistry Division of the
National Science Foundation (M.B.S., C.J.C., D.E.F., D.C.,
J.P.T., P.H.R., and S.S.) for funding. C.J.C. is grateful to the
Alfred P. Sloan Foundation for support, and E.C.R. thanks the
Howard Hughes Foundation for an Undergraduate Research
Fellowship.
LFP-TRIR. We conducted LFP-TRIR experiments following the
method of Hamaguchi and co-workers58,59 as described previously.60
Briefly, the broadband output of a MoSi2 IR source (JASCO) is crossed
with excitation pulses from a Nd:YAG laser. Changes in IR intensity
are monitored by an MCT photovoltaic IR detector (Kolmar Technolo-
gies, KMPV11-1-J1), amplified, and digitized with a Tektronix
TDS520A oscilloscope. The experiment is conducted in the dispersive
mode with a JASCO TRIR-1000 spectrometer. TRIR difference spectra
were collected at 16 cm-1 resolution using either a Continuum HPO-
300 diode-pumped Nd:YAG laser (266 nm, 10 ns, 0.4 mJ) or a
Quantronix Q-switched Nd:YAG laser (266 nm, 90 ns, 0.4 mJ)
operating at 200 Hz. Kinetic traces were collected using a Continuum
Minilite II Nd:YAG laser (266 nm, 5 ns, 1-4 mJ) operating at 20 Hz.
DFT Calculations. All geometries were fully optimized at the
density functional (DFT) level using the gradient-corrected functionals
of Becke61 for exchange and of Perdew and Wang62 for correlation.
The cc-pVDZ basis set63 was used for all calculations. Analytic
frequency calculations were carried out at the same level of theory. In
addition to providing harmonic frequencies, these calculations were
also examined to verify the nature of all stationary points and to
calculate zero-point vibrational energies for the singlet-triplet splittings.
Supporting Information Available: Descriptions of the
preparation and characterization of synthetic intermediates
7-14, 19, and 20 as well as photoproducts 22-25 and 31
(PDF). This material is available free of charge via the Internet
at http://pubs.acs.org. Geometries, energies, and animations of
relevant normal modes for 2-5 are also available at http://
pollux.chem.umn.edu/∼sullivan/phNfreq2.
JA001184K
(64) Worthington, S. E.; Cramer, C. J. J. Phys. Org. Chem. 1997, 10,
755.
(65) Schreiner, P. R.; Karney, W. L.; Schleyer, P. v. R.; Borden, W. T.;
Hamilton, T. P.; Schaefer, H. F., III. J. Org. Chem. 1996, 61, 7030.
(66) Wong, M. W.; Wentrup, C. J. Org. Chem. 1996, 61, 7022.
(67) Matzinger, S.; Bally, T.; Patterson, E. V.; McMahon, R. J. J. Am.
Chem. Soc. 1996, 118, 1535.
(68) Patterson, E. V.; McMahon, R. J. J. Org. Chem. 1997, 62, 4398.
(69) Sulzbach, H. M.; Platz, M. S.; Schaefer, H. F., III.; Haddad, C. M.
J. Am. Chem. Soc. 1997, 119, 5682.
(70) Shustov, G. V.; Liu, M. T. H.; Rauk, A. J. Phys. Chem. A 1997,
101, 2509.
(71) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G.
A.; Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski,
V. G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.;
Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.;
Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-
Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian 94, Revision D.1;
Gaussian, Inc.: Pittsburgh, PA, 1995.
(58) Iwata, K.; Hamaguchi, H. Appl. Spectrosc. 1990, 44, 1431.
(59) Yuzawa, T.; Kato, C.; George, M. W.; Hamaguchi, H. Appl.
Spectrosc. 1994, 48, 684.
(60) Wang, Y.; Yuzawa, T.; Hamaguchi, H.; Toscano, J. P. J. Am. Chem.
Soc. 1999, 121, 2875.
(61) Becke, A. J. Chem. Phys. 1986, 84, 4524.
(62) Perdew, J. P.; Burke, K.; Wang, Y. Phys. ReV. B 1996, 54, 6533.
(63) Dunning, T. H. J. Chem. Phys. 1989, 90, 1007.