Trimethoxybenzene complexes of pentafluorophenylchlorocarbene

Article abstract of DOI:10.1021/jp2039493

Pentafluorophenylchlorocarbene, generated by laser flash photolysis (LFP) of pentafluorophenylchlorodiazirine, formed π-type complexes with 1,3,5-trimethoxybenzene in pentane. The carbene and carbene complexes were in equilibrium with K = 3.21 × 10^5<

Full text of DOI:10.1021/jp2039493

ARTICLE
pubs.acs.org/JPCA
Trimethoxybenzene Complexes of Pentafluorophenylchlorocarbene
Lei Wang, Robert A. Moss,* and Karsten Krogh-Jespersen*
Department of Chemistry and Chemical Biology, Rutgers, The State University of New Jersey, New Brunswick, New Jersey 08903,
United States
S Supporting Information
b
ABSTRACT: Pentafluorophenylchlorocarbene, generated by laser flash photolysis
(LFP) of pentafluorophenylchlorodiazirine, formed π-type complexes with 1,3,5-
trimethoxybenzene in pentane. The carbene and carbene complexes were in equilib-
rium with K = 3.21 ꢁ 105 Mꢀ1 at 294 K. From the temperature dependence of K,
ΔH° = ꢀ10.2 kcal/mol, ΔS° = ꢀ9.5 eu, and ΔG° = ꢀ7.4 kcal/mol at 298 K. The
carbene complexes were characterized by UVꢀvis spectroscopy and computational
analysis.
well-correlated by the Hammett σp constants for X, with F =
2.48.6 The positive value of F suggests that electron-withdrawing
groups destabilize carbenes 1aꢀ1f, shifting the equilibrium
toward complexes 2aꢀ2f, where TMB donates electron density
to the vacant carbenic p-orbital. Also, stronger charge-transfer
complexes 2aꢀ2f will form when the electron-rich TMB inter-
acts with the more electron-deficient carbene centers and aro-
matic rings induced by an electron-withdrawing substituent.6
The pentafluorophenyl group (C6F5) is more electron-with-
drawing than the phenyl group (C6H5): σI of C6F5 is 0.31
compared to 0.12 for C6H5,7 and σRþ of C6F5 is ꢀ0.078 compared
to ꢀ0.179 or ꢀ0.3010 for C6H5. On the basis of our previous
analysis,6 pentafluorophenylchlorocarbene (F5-PhCCl, 3) should
bind more strongly than PhCCl to TMB. Indeed, our electronic
structure calculations (see below) indicate that F5-PhCCl/TMB
complexes should be approximately 3ꢀ5 kcal/mol more stable in
enthalpy and free energy than PhCCl/TMB complexes.
I. INTRODUCTION
Given the formally vacant p-orbital featured by singlet car-
benes, we expect that electron donation by solvents will lead to at
least transient carbeneꢀsolvent complexes. Indeed, indirect
evidence for such complexation has been offered in the context
of the solvent modulation of carbenic reactivity.1 More recently,
direct evidence for carbene complexation has accrued using such
physical methods as calorimetry,2 IR spectroscopy,3 and UV
spectroscopy,4 coupled with fast laser technology.
We provided the first direct spectroscopic evidence for
carbeneꢀcarbene complex equilibration; see eq 1.5 Phenylchloro-
carbene (PhCCl, 1d) formed a π-type complex with 1,3,5-
trimethoxybenzene (TMB),
Although pentafluorophenylcarbene (C6F5CH, 4) is known,11
F5-PhCCl is not. Carbene 4 is a ground state triplet with (an
experimental) ΔGS-T = ꢀ3.1 kcal/mol at 298 K,11 but pre-
ferential stabilization of singlet 3 by electron donation from its
chlorine substituent should make the singlet the ground state.
Indeed, our CCSD(T) calculations predict ΔGS-T = 4.2 kcal/mol
for 3 but ꢀ5.6 kcal/mol for 4 (see below). An analogous reversal
of ground state multiplicity is found with phenylcarbene
(PhCH) and PhCCl, where prior computational studies
and the carbene and carbene complex were in equilibrium with
K = 1264 Mꢀ1. From the temperature dependence of K, we
extracted the associated thermodynamic parameters: ΔH° =
ꢀ7.1 kcal/mol, ΔS° = ꢀ10.2 eu, and ΔG° = ꢀ4.1 kcal/mol.5
Very recently, we extended this study to the p-substituted
phenylchlorocarbenes 1aꢀf, where X = NO2, CF3, Cl, H,
Me, and MeO.6 Logarithms of the equilibrium constants were
Received: April 28, 2011
Revised:
May 27, 2011
Published: June 22, 2011
r
2011 American Chemical Society
8113
dx.doi.org/10.1021/jp2039493 J. Phys. Chem. A 2011, 115, 8113–8118
|

The Journal of Physical Chemistry A
ARTICLE
Scheme 1. Preparation of Pentafluorophenylchlorodiazirine
Table 1. SingletꢀTriplet Separation (ΔGS-T, kcal/mol)a in
CH2, PhCH, F5-PhCH, PhCCl, and F5-PhCClb
species
B97D
wB97XD
M06-2X
B3LYP
CCSD(T)c
CH2
ꢀ12.0
ꢀ6.6
ꢀ6.7
4.3
ꢀ12.3
ꢀ7.1
ꢀ8.2
3.8
ꢀ14.2
ꢀ8.0
ꢀ9.1
4.8
ꢀ11.8
ꢀ5.8
ꢀ7.0
5.2
ꢀ12.9
ꢀ4.5
ꢀ5.6
7.7
PhCH
F5-PhCH
PhCCl
F5-PhCCl
0.35
ꢀ0.40
0.80
0.70
4.2
a A negative value for this quantity signifies that the triplet state is more
stable than the singlet state. b 6-311þG(d) basis sets used in all
calculations, see Computational Methods. c The electronic to free energy
corrections obtained at the B3LYP level were applied to the electronic
energies obtained from CCSD(T) calculations to arrive at approximate
CCSD(T) free energies.
support a triplet ground state for PhCH (Δ GS-T ∼ ꢀ4 kcal/
mol) 12 but a singlet ground state for PhCCl (ΔGS-T = 7.8 kcal/
mol).13 Our computed CCSD(T) ΔGS-T values for PhCH and
PhCCl are ꢀ4.5 and 7.7 kcal/mol, respectively, and thus agree
fully with the results of the prior studies.
To test our computational prediction that F5-PhCCl will bind
more tightly to TMB than PhCCl, we require a synthesis of
pentafluorophenylchlorodiazirine, which would serve as a photo-
chemical precursor for F5-PhCCl. The desired precursor was
prepared by the sequence of Scheme 1: commercially available
pentafluorobenzonitrile was converted by Garigipati's reaction14
to pentafluorobenzamidine (5), and the latter was converted to
pentafluorophenylchlorodiazirine (6) by Graham oxidation with
aqueous hypochlorite.15 Experimental details for these proce-
dures appear in the Experimental section.
temperature of 100 mL of 12% aqueous NaOCl (commercial
bleach), saturated with NaCl. Stirring was continued for 2 h at
room temperature. The pentane phase was separated and retained,
and the aqueous phase was extracted twice with 50 mL portions of
pentane. The combined pentane solutions were dried over MgSO4,
concentrated to ∼5 mL, and chromatographed over a short column
of silica gel using pentane as the eluent. We thus obtained 1.1 g of
diazirine 6 (in pentane solution), 56% yield.
UV (pentane): 326, 333, 342 nm; see Figure S-1 in the
Supporting Information.
IR (film, cmꢀ1): 1584 (NdN).15
13C NMR (125 MHz, CDCl3, δ): 36.604 (s), 110.18 (m),
1
139.13 and 137.10 (dm, JCꢀF = 254 Hz), 144.60 and 142.54
2. EXPERIMENTAL DETAILS AND COMPUTATIONAL
METHODS
(dm, 1JCꢀF = 258 Hz), 147.11 and 145.07 (dm, 1JCꢀF = 256 Hz).
19F NMR (470 MHz, DMSO-d6, δ): ꢀ139.36 and ꢀ139.40 (d
of m, 2F); ꢀ149.06, ꢀ149.11, and ꢀ149.16 (t of m, 1F);
ꢀ159.81, ꢀ159.85, and ꢀ159.90 (t of m, 2F).
2.1. Experimental Details. Preparation of Pentafluorobenzami-
dine (5).14 A solution of 2 M trimethylaluminum in toluene (100 mL)
was slowly added to a magnetically stirred suspension of 11.7 g
(0.22 mol) of ammonium chloride in 40 mL of dry toluene under
nitrogen at 0 °C. After addition, the mixture was warmed to room
temperature and stirred for 1 h until gas evolution ceased. Then, 29 g
(0.15 mol) of pentafluorobenzonitrile in 50 mL of dry toluene was
slowly added, and the solution was heated to 80 °C for 48 h under
nitrogen. The reaction was then quenched by slow addition of
100 mL of methanol at 0 °C. The reaction mixture was slowly
poured into a slurry of 30 g of silica gel in 80 mL of chloroform and
stirred for 1 h. The silica gel was filtered and washed with methanol.
The combined filtrate and wash were stripped of solvent on the rotary
evaporator leaving a residue, which was a mixture of pentafluoroben-
zamidinium chloride and ammonium chloride. The residue was
extracted twice with 400 mL portions of 1:10 methanolꢀchloroform.
The combined extracts were stripped to yield 10 g (27%) of
pentafluorobenzamidinium chloride 5, mp 272ꢀ276 °C (dec.).
IR (film, cmꢀ1): 2978, 2932, 1452, 1448, 1433, 1321, 1317,
1004, 984, 943.
2.2. Computational Methods. Electronic structure calcula-
tions were carried out using methodologies implemented in the
Gaussian 09 suite of programs.16 Density functional theory
(DFT) provided the default methods for the calculation of
ground and excited state structures and energetics.17 Ground
state geometry optimizations of carbene and solvent monomers
and dimeric carbeneꢀsolvent complexes were carried out using
the dispersion-corrected B97D,18 M06-2X,19 or wB97XD20
exchange-correlation functionals and 6-311þG(d)21 basis sets
(B97D/6-311þG(d), etc.). The potential energy surfaces for the
carbeneꢀsolvent complexes are very soft with respect to relative
motion of the aryl rings, and the structural minima possess
several low-frequency vibrational modes. Hence, all geometry
optimizations and normal mode calculations were conducted
with tight convergence criteria (opt=tight) imposed and inte-
gration grid sizes increased beyond their default values
(integral(grid=ultrafine)). Stationary points were characterized
by normal-mode analysis, and the (unscaled) vibrational
frequencies formed the basis for the calculation of vibrational
zero-point energy (ZPE) corrections. Standard thermodynamic
corrections (based on harmonic oscillator/rigid rotor approx-
imations and ideal gas behavior) were then applied to convert
from purely electronic energies (E) to (standard) enthalpies (H°;
T = 298.15 K) and free energies (G°; T = 298.15 K, P = 1 atm).22
Interaction energies of carbeneꢀsolvent complexes were
approximately corrected for basis set superposition errors
(BSSE) through the application of the counterpoise
correction.23 Carbene singletꢀtriplet energy separations were
19F NMR (470 MHz, DMSO-d6, δ): ꢀ138.45 and ꢀ138.50
(d, 2F); ꢀ147.88, ꢀ147.93, and ꢀ147.99 (t, 1F); ꢀ160.23,
ꢀ160.26, ꢀ160.28, and ꢀ160.31 (m, 2F).
13C NMR (125 MHz, DMSO-d6, δ): 155.50 (s, CdN);
145.32 and 143.32 (d of m); 144.07 and 142.37 (d of m);
138.71 and 136.70 (d of m); 106.39 (t).
Preparation of 3-Chloro-3-pentafluorophenyldiazirine (6).15 A
mixture of 5 g of LiCl, 2.0 g (8.1 mmol) of amidine 5, and 50 mL of
DMSO was stirred until most of the LiCl had dissolved. Then,
100 mL of pentane was added, followed by slow addition at room
8114
dx.doi.org/10.1021/jp2039493 |J. Phys. Chem. A 2011, 115, 8113–8118

The Journal of Physical Chemistry A
ARTICLE
Figure 2. Calibrated UVꢀvis spectrum acquired 150 ns after LFP
generation of F5-PhCCl in 0.0128 mM TMB/pentane solution; F5-
PhCCl absorption at 300 nm; absorption by F5-PhCCl/TMB complexes
A and B (see Figure 3) at 420 and 516 nm.
Figure 1. Calibrated UVꢀvis spectrum of F5-PhCCl (3) in pentane
150 ns after the laser flash; π(phenyl) f p(carbene) absorptions at 300
and 372 nm, σ(carbene) f p(carbene) absorption at 596 nm.
evaluated at the B3LYP24 and CCSD(T)25 levels of theory with
6-311þG(d) basis sets (Table 1). Electron population analysis
made use of the conventional Mulliken partitioning scheme.26
Excited state calculations (transition wavelengths (λ) and oscilla-
tor strengths (f)) at optimized B97D ground state geometries utilized
the time-dependent DFT formalism27a and the B3LYP functionals
(TD-B3LYP/6-311þG(d)//B97D/6-311þG(d)). General solvent
effects were incorporated with the polarizable conductor self-
consistent reaction field model (CPCM)28 and default parameters for
n-pentane provided in Gaussian 09. The assignment of a particular
electronic transition (σ f p, π f p, or π f π*) was based on
inspection of the largest transition amplitudes for the excitation and
by visualization of the contributing molecular orbitals (MOs).
3. RESULTS AND DISCUSSION
3.1. Pentafluorophenylchlorocarbene and Complex For-
mation with Trimethoxybenzene. The laser flash photolysis
(LFP) of diazirine 6 in pentane gave F5-PhCCl (3), whose
calibrated29 UVꢀvis spectrum (Figure 1) displayed strong and
weak π(phenyl) f p(carbene) absorptions at 300 and 372 nm,
together with a weak σ f p absorption at 596 nm. Our electronic
structure calculations show that the 300 and 372 nm bands
represent charge-transfer type transitions from the nearly degen-
erate set of high-lying phenyl π-orbitals (e1g symmetry in
benzene) to the vacant carbenic p-orbital (lowest unoccupied
molecular orbital (LUMO) in 3). The MO with a nonzero
coefficient on C1(phenyl) provides the intense 300 nm band,
whereas the other MO gives rise to the much weaker band at
372 nm. The absorptions are computed30 at 296 nm (f = 0.33)
and 362 nm (f = 0.042), respectively. The long wavelength band
at 596 nm (computed too far to the red at 867 nm, f = 0.003) is
principally associated with electron promotion from the carbenic
carbon's filled σ-orbital (highest occupied molecular orbital
(HOMO) in 3) to its vacant p-orbital. The interaction between
the aryl π-orbitals and the carbene p-orbital leads to B3LYP
computed σꢀp (HOMOꢀLUMO) separations which are too
small, as noted by us previously.4 The general tendency of DFT
to underestimate the HOMOꢀLUMO separation for weakly
Figure 3. Top and side views of computed π-complexes of high stability
formed between F5-PhCCl and TMB (in perspective, carbene on top;
light green, fluorine; dark green, chlorine; red, oxygen).
interacting systems and thus of TD-DFT to underestimate the
electronic excitation energies, when local, time-independent
functionals are employed, has been documented.27b
For PhCCl, analogous transitions are observed at 308 and 578 nm5
and computed at 292 nm (f = 0.44) and 737 nm (f = 0.002). The
weak π f p PhCCl transition is computed at 330 nm (f = 0.032) but
not observed because it is buried beneath the much stronger π f p
band at 308 nm. In F5-PhCCl, the weak π f p absorption is shifted
30ꢀ40 nm to the red and is observed as a separate peak at 372 nm
(Figure 1).
The LFP of diazirine 6 in the presence of 0.0128 mM TMB
affords the calibrated UVꢀvis spectrum shown in Figure 2, with
distinct absorptions at 300, 420, and 516 nm. Computational
analysis (see below) assigns the 420 and 516 nm absorptions to
F5-PhCCl/TMB π-complexes A and B, respectively; see Figure 3.
Computed absorptions30 for A and B are at 443 nm (f = 0.16) and
8115
dx.doi.org/10.1021/jp2039493 |J. Phys. Chem. A 2011, 115, 8113–8118

The Journal of Physical Chemistry A
ARTICLE
Figure 4. Relative absorption intensities at 300 nm/516 nm vs 1/[TMB]
(Mꢀ1) for the LFP generation of F5-PhCCl in TMB/pentane solution at
294 K. The slope of the correlation line is 1.25 ꢁ 10ꢀ5 M (r = 0.993),
leading to K = 3.21 ꢁ 105 Mꢀ1 for eq 2.31
Figure 5. Plot of ln K (Mꢀ1) vs 1/T (Kꢀ1) for the equilibrium of eq 2.
The slope (5126) affords ΔH° = ꢀ10.2 kcal/mol, and the intercept
(ꢀ4.78) gives ΔS° = ꢀ9.5 eu. The correlation coefficient is r = 0.996.
5.17 ꢁ 105 at 288 K; and 1.71 ꢁ 105 at 303 K. A correlation of
ln K vs 1/T appears in Figure 5, from which the slope and intercept
give ΔH° = ꢀ10.2 ( 0.6 kcal/mol and ΔS° = ꢀ9.5 ( 1.9 eu,
respectively, leading to ΔG° = ꢀ7.4 ( 0.8 kcal/mol at 298 K.
From these results, the complexation of F5-PhCCl by TMB is
about 3 kcal/mol more favorable in both enthalpy and free
energy than complexation of PhCCl. Our computational predic-
tion (see above and below) is for energetic advantages favoring
the complexation of F5-PhCCl by ∼4 kcal/mol in ΔG° and
3.7ꢀ5.3 kcal/mol in ΔH°, so that the experimental and compu-
tational results are in quite good agreement.
545 nm (f = 0.083), respectively. The apparent intensity reversal
between the observed and computed A and B absorptions is
precedented.6
The relative intensities of the F5-PhCCl absorption at 300 nm
and the F5-PhCCl/TMB complex absorptions at 420 and 516 nm
vary with TMB concentration in a manner consistent with the
carbene (3)ꢀcarbene complex (7) equilibrium expressed in eq 2;
see Figures S-3 to S-6 in the Supporting Information.
3.3. Computational Studies. We first determined the ground
state spin multiplicity of F5-PhCCl and related carbenes by
evaluating the energies of the singlet (S, σ2 occupancy) and
triplet (T, σ1p1 occupancy) states at their respective optimized
geometries. DFT results for the S-T free energy separations of
CH2, PhCH, F5-PhCH ( 4), PhCCl (1d), and F5-PhCCl (3)
obtained with the 6-311þG(d) basis set and B97D, M06-2X,
wB97XD, or B3LYP functionals are shown in Table 1. We also
present wave function based CCSD(T)/6-311þG(d) results
obtained from single point calculations using the B3LYP/6-
311þG(d) optimized geometries of these carbenes. S-T poten-
tial energy differences are available in Table S-1 in the Supporting
Information.
All five computational methods applied here clearly predict
that PhCH and F5-PhCH possess triplet ground states. Perfluor-
ination of the aryl ring increases the S-T separation uniformly
by ca. 1 kcal/mol, except in the B97D calculations that show
virtually no change (∼0.1 kcal/mol). For PhCCl, the computa-
tional methods again agree and uniformly predict a singlet
ground state for this carbene. The preferential stabilization of
the triplet state by perfluorination of the aryl ring is larger for
F5-PhCCl (∼3.5ꢀ4.5 kcal/mol) than F5-PhCH (∼0.1ꢀ1.2
kcal/mol). The DFT methods predict a S-T separation less
than 1 kcal/mol for F5-PhCCl; the application of the wB97XD
functionals actually predicts, albeit barely, a triplet ground
state for F5-PhCCl. In contrast, our CCSD(T) calculations
arrive conclusively at a singlet ground state for F5-PhCCl with
ΔGS-T = 4.2 kcal/mol.
Indeed, a plot of the quotient of the calibrated absorption
intensities for F5-PhCCl at 300 nm and carbene complex B at
516 nm versus 1/[TMB] at 294 K affords the linear correlation of
Figure 4, where the slope (1.25 ꢁ 10ꢀ5 M) leads to a value K =
3.21 ꢁ 105 Mꢀ1 for the equilibrium constant of eq 2.31
Given that the reaction of F5-PhCCl with TMB produces
complex A in addition to complex B, the actual equilibrium
constant for eq 2 will be somewhat larger than 3.21 ꢁ 105 Mꢀ1
,
which is based solely on the formation of complex B. The
formation of the F5-PhCCl/TMB complexes from carbene 3 and
TMB is not only thermodynamically favorable (K . 1) but is also
∼250 times more favorable than formation of the corresponding
PhCCl/TMB complexes for which K = 1.26 ꢁ 103 Mꢀ1 (ΔΔG° ∼
3.3 kcal/mol).5,32
3.2. Enthalpy and Entropy of Complexation. We deter-
mined K for eq 2 at four additional temperatures (see Figures S-7
to S-10 in the Supporting Information), obtaining the following
values of K (Mꢀ1): 2.39 ꢁ 106 at 263 K; 6.52 ꢁ 105 at 281 K;
8116
dx.doi.org/10.1021/jp2039493 |J. Phys. Chem. A 2011, 115, 8113–8118

The Journal of Physical Chemistry A
ARTICLE
Table 2. Computed Thermodynamic Parameters for F5-
PhCCl/TMB Complexes A and B
Table 3. Electronic Transition Wavelengths (λ, nm) and
Oscillator Strengths (f, Dimensionless) Computed at the TD-
B3LYP/6-311þG(d)//B97D/6-311þG(d) Level with
CPCM Solvent Corrections (n-Pentane)
species/method
ΔH°a
ΔS°a
ΔG°a,b
ΔH(corr)°a,c
A/B97D
ꢀ17.6
ꢀ16.2
ꢀ14.9
ꢀ14.8
ꢀ15.3
ꢀ13.7
ꢀ40.8
ꢀ37.6
ꢀ37.5
ꢀ37.4
ꢀ37.3
ꢀ36.1
ꢀ5.4
ꢀ5.0
ꢀ4.3
ꢀ3.6
ꢀ4.2
ꢀ1.6
ꢀ13.5
ꢀ11.9
ꢀ11.1
ꢀ11.2
ꢀ11.3
ꢀ10.2
λ, f
F5-PhCCl
complex A
complex B
A/M06ꢀ2X
A/wB97XD
B/B97D
λ1, f1
λ2, f2
λ3, f3
λ4, f4
λ5, f5
867, 0.003
362, 0.042
296, 0.332
292, 0.048
279, 0.020
775, 0.018
550, 0.003
443,0.161
338, 0.068
330, 0.061
823, 0.016
609, 0.004
545, 0.083
357, 0.032
346, 0.039
B/M06ꢀ2X
B/wB97XD
a ΔH° and ΔG° in kcal/mol, ΔS° in eu, relative to the separated
reactants; T = 298.15 K. b The free energy differences were computed
using a reference state of 1 M concentration for each species participat-
ing in the reaction. c Counterpoise-corrected enthalpies.
slightly more than 3 kcal/mol larger than the analogously calculated
binding enthalpies of PhCCl/TMB type B complexes (ꢀ7.5, ꢀ8.0,
and ꢀ7.0 kcal/mol, respectively). The computed entropies are
numerically much larger than the experimentally determined ones,
most likely due to the difference in physical phase serving as reference
for the calculations (gas) and experiments (condensed). It seems
likely that some molecular degrees of freedom have effectively been
frozen out in the solution phase, before carbeneꢀsolvent complex
formation occurs. Generally, the computed ΔG° values are more
favorable for type B F5-PhCCl/TMB than PhCCl/TMB complexes
by approximately 4 kcal/mol, corresponding to an increase in
equilibrium constant by a factor of ∼103; the experimental estimate
is ∼250 (see above).
It is well-accepted that single-determinant methods (HF or
DFT) tend to overestimate the stability of high spin states and
that extensive recovery of electron correlation energy through,
for example, configuration interaction must be made to correct
this bias.33 The CCSD(T) derived values for the singletꢀtriplet
(S-T) separations in PhCH (ꢀ4.5 kcal/mol) and F5-PhCH
(ꢀ5.6 kcal/mol) match the experimentally derived values fairly
well (ꢀ2 to ꢀ3 kcal/mol for PhCH, ∼ ꢀ4 kcal/mol for F5-
PhCH),11,34,35 although they still lean toward an exaggerated
preference for the triplet state. A comparison of our computed
DFT and CCSD(T) results for ΔGS-T in the aryl carbenes
(Table 1) shows a further systematic overestimation of relative
triplet state stability by DFT of ca. 2ꢀ4 kcal/mol. The experi-
mental S-T separation in methylene (CH2) is ꢀ9.1 kcal/mol
(triplet ground state);36 the computational methods used here,
including CCSD(T), predict a methylene S-T separation larger
by 3ꢀ5 kcal/mol (Table 1). Thus, the S-T separation in F5-
PhCCl (3) should be even larger than the CCSD(T) predicted
value of 4.6 kcal/mol and could approach 7ꢀ9 kcal/mol.
Computed UVꢀvis transitions for F5-PhCCl and complexes A
and B are presented in Table 3. The lowest energy transition (λ1)
retains the σ f p label in all three species, although the
“carbenic” σ and p-orbitals attain some Cl and phenyl π-
character, respectively, from orbital mixing. The intense
296 nm F5-PhCCl transition as well as the F5-PhCCl/TMB
complex signature transitions at 443 nm (complex A) and
545 nm (complex B) may all be assigned as π f p transitions;
hence they carry considerable charge transfer character. Finally,
we note that it is the intensities of these latter transitions (f3,
Table 3) that are employed when determining the value of the
experimental equilibrium constant (see the Supporting
Information).
We have previously documented that arylhalocarbenes and
TMB preferentially form sandwich-type π-complexes (viz., pro-
totype complexes A and B in Figure 3) and that more spatially
extended complexes of the O-ylidic type are slightly higher in
energy (2ꢀ3 kcal/mol above A or B) and not competitive. In the
A π-complex, the carbene center interacts strongly with an
unsubstituted carbon of TMB (C2); for example, the computed
C(carbene)ꢀC2(TMB) distance is 2.72 Å (B97D/6-311þG-
(d)), while the C(carbene)ꢀC1(TMB) separation is 3.09 Å. In
the B complex, the carbene center interacts with the C1ꢀC2
bond of TMB; the C(carbene)ꢀC1 and C2 (TMB) distances are
3.07 and 3.04 Å, respectively. Substantial stabilizing face-to-face
overlap originates between the aromatic ring moieties, in parti-
cular for complexes of type A. The net charge transferred from
TMB to F5-PhCCl is 0.21 e in A and 0.12 e in B.
Density functionals, which explicitly include dispersion correc-
tions, provide appropriate binding energies of arylhalocarbene/TMB
complexes.5,6 The computed energetic parameters for the F5-
PhCCl/TMB complexes A and B are shown in Table 2. The binding
enthalpies and free energies are systematically slightly more favorable
for type A than type B complexes by 1ꢀ2 kcal/mol. According to our
interpretation of the spectroscopic data, the measured equilibrium
constant is associated with B type complexes for which our B97D,
M06-2X, and wB97XD calculations predict binding enthalpies
ΔH° = ꢀ11.2, ꢀ11.3, and ꢀ10.2 kcal/mol, respectively
(Table 2). These computed enthalpies compare very favorably with
the measured value (ΔH° = ꢀ10.2 ( 0.6 kcal/mol), and they are
4. CONCLUSIONS
Pentafluorophenylchlorocarbene, generated by laser flash
photolysis of pentafluorophenylchlorodiazirine, formed π-type
complexes with 1,3,5-trimethoxybenzene in pentane. The car-
bene and carbene complexes were in equilibrium with K = 3.21 ꢁ
105 Mꢀ1 at 294 K. The thermodynamic parameters for the
equilibrium were determined from the temperature dependence
of K: ΔH° = ꢀ10.2 kcal/mol, ΔS° = ꢀ9.5 eu, and ΔG° = ꢀ7.4
kcal/mol at 298 K. The carbene complexes were characterized by
UVꢀvis spectroscopy and computational analysis. The com-
puted spectra and energies of the carbene and carbene complexes
were in good agreement with the experimental data.
’ ASSOCIATED CONTENT
S
Supporting Information. Figures S-1 to S-10; calculation
b
of the equilibrium constant; complete reference to Gaussian 09;
Table S-1; optimized geometries and energies for F5-PhCCl,
TMB, and F5-PhCCl/TMB complexes A and B at the B97D/6-
311þG(d) level. This material is available free of charge via the
Internet at http://pubs.acs.org.
8117
dx.doi.org/10.1021/jp2039493 |J. Phys. Chem. A 2011, 115, 8113–8118

The Journal of Physical Chemistry A
ARTICLE
’ AUTHOR INFORMATION
(25) Pople, J. A.; Head-Gordon, M.; Raghavachari, K. J. Chem. Phys.
1987, 87, 5968.
(26) Mulliken, R. S. J. Chem. Phys. 1955, 23, 1833.
(27) (a) Casida, M. E.; Jamorski, C.; Casida, K. C.; Salahub, D. R.
J. Chem. Phys. 1998, 108, 4439. (b) See, e.g., Dreuw, A.; Weisman, J. L.;
Head-Gordon, M. J. Chem. Phys. 2003, 119, 2943.
Corresponding Author
*E-mail: moss@rutchem.rutgers.edu (R.A.M.); krogh@rutchem.
rutgers.edu (K.K.-J.).
(28) Barone, V.; Cossi, M. J. Phys. Chem. A 1998, 102, 1995.
(29) Calibration corrects the raw UVꢀvis absorptions for wave-
length-dependent variations in sample absorptivity, xenon monitoring
lamp emission, and detector sensitivity.5 The calibration spectrum is
shown in the Supporting Information (Figure S-2).
’ ACKNOWLEDGMENT
We are grateful to the National Science Foundation for
financial support.
(30) Results obtained at the TD-B3LYP/6-311þG(d)//B97D/6-
311þG(d) level with SCRF = CPCM, solvent = pentane. See Section
2.2, above.
’ REFERENCES
(31) K = (1/slope) (f3/f1), where f3 (0.332) is the computed
oscillator strength of F5-PhCCl at 296 nm and f1 (0.0829) is the
computed oscillator strength of carbene complex B at 545 nm. See the
Supporting Information for details.
(1) (a) Tomioka, H.; Ozaki, Y.; Izawa, Y. Tetrahedron 1985, 41,
4987. (b) Ruck, R. T.; Jones, M., Jr. Tetrahedron Lett. 1998, 39, 2277.
(c) Moss, R. A.; Yan, S.; Krogh-Jespersen, K. J. Am. Chem. Soc. 1988,
120, 1088. (d) Krogh-Jespersen, K.; Yan, S.; Moss, R. A. J. Am. Chem. Soc.
1999, 121, 6269.
(2) Kahn, M. I.; Goodman, J. L. J. Am. Chem. Soc. 1995, 117, 6635.
(3) (a) Tippmann, E. M.; Platz, M. S.; Svir, I. B.; Klymenko, O. V.
J. Am. Chem. Soc. 2004, 126, 5750. (b) Wang, J.; Kubicki, J.; Peng, H.;
Platz, M. S. J. Am. Chem. Soc. 2008, 130, 6604.
(4) (a) Moss, R. A.; Tian, J.; Sauers, R. R.; Krogh-Jespersen, K. J. Am.
Chem. Soc. 2007, 129, 10019. (b) Moss, R. A.; Wang, L.; Weintraub, E.;
Krogh-Jespersen, K. J. Phys. Chem. A 2008, 112, 4651. (c) Moss, R. A.;
Wang, L.; Odorisio, C. M.; Krogh-Jespersen, K. J. Phys. Chem. A 2010,
114, 209.
(32) Complexation of F5-PhCCl by TMB (K = 3.21 ꢁ 105 Mꢀ1
)
is slightly superior to complexation of p-O2NPhCCl by TMB (K =
1.34 ꢁ 105 Mꢀ1).6
(33) Nguyen, T. L.; Kim, G. S.; Mebel, A. M.; Nguyen, M. T. Chem.
Phys. Lett. 2001, 349, 571–577.
(34) Woodcock, H. L.; Moran, D.; Brooks, B. R.; Schleyer, P. v. R.;
Schaefer, H. F. J. Am. Chem. Soc. 2007, 129, 3763.
(35) Koch, W.; Holthausen, M. C. A Chemist's Guide to Density
Functional Theory; Wiley-VCH: Weinheim, 2000.
(36) McKellar, A. R. W.; Bunker, P. R.; Sears, T. J.; Evenson, K. M.;
Saykally, R. J.; Langhoff, S. R. J. Chem. Phys. 1983, 79, 5251.
(5) Moss, R. A.; Wang, L.; Odorisio, C. M.; Krogh-Jespersen, K.
J. Am. Chem. Soc. 2010, 132, 10677.
(6) Wang, L.; Moss, R. A.; Thompson, J.; Krogh-Jespersen, K. Org.
Lett. 2011, 13, 1198.
(7) Charton, M. Prog. Phys. Org. Chem. 1981, 13, 119; see also pp 143
and 149.
(8) Charton, M. Private communication, Nov 22, 2010.
(9) See ref 7, p 211.
(10) Ehrenson, S.; Brownlee, R. T. C.; Taft, R. W. Prog. Phys. Org.
Chem. 1973, 10, 1; see also p 13.
(11) Admasu, A.; Gudmundsdottir, A. D.; Platz, M. S. J. Phys. Chem.
A 1997, 101, 3832.
(12) Matzinger, S.; Bally, T.; Patterson, E. V.; McMahon, R. J. J. Am.
Chem. Soc. 1996, 118, 1535.
(13) Pliego, J. R., Jr.; De Almeida, W. B.; Celebi, S.; Zhu, Z.; Platz,
M. S. J. Phys. Chem. A 1999, 103, 7481.
(14) (a) Garigipati, R. S. Tetrahedron Lett. 1990, 31, 1969. (b) Moss,
R. A.; Ma, W.; Merrer, D. C.; Xue, S. Tetrahedron Lett. 1995, 36, 8761.
(15) Graham, W. H. J. Am. Chem. Soc. 1965, 87, 4396.
(16) Frisch, M. J., et al. Gaussian 09, Revision A.02; Gaussian, Inc.:
Wallingford, CT, 2009. See the Supporting Information for the com-
plete reference to Gaussian 09.
(17) Parr, R. G.; Yang, W. Density-Functional Theory of Atoms and
Molecules; University Press: Oxford, 1989.
(18) Grimme, S. J. Comput. Chem. 2006, 27, 1787.
(19) Zhao, Y.; Truhlar, D. G. J. Phys. Chem. A 2006, 110, 5121.
(20) Chai, J.-D.; Head-Gordon, M. Phys. Chem. Chem. Phys. 2008,
10, 6615.
(21) (a) Ditchfield, R.; Hehre, W. J.; Pople, J. A. J. Chem. Phys. 1971,
54, 721. (b) Hariharan, P. C.; Pople, J. A. Mol. Phys. 1974, 27, 209.
(c) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys.
1980, 72, 650. (d) McLean, A. D.; Chandler, G. S. J. Chem. Phys. 1980,
72, 5639. (e) Clark, T.; Chandrasekhar, J.; Spitznagel, G. W.; Schleyer,
P. v. R. J. Comput. Chem. 1983, 4, 294.
(22) McQuarrie, D. A. Statistical Thermodynamics; Harper and Row:
New York, 1973.
(23) (a) Liu, B.; McLean, A. D. J. Chem. Phys. 1973, 59, 4557.
(b) Boys, S. F.; Bernardi, F. Mol. Phys. 1970, 19, 553.
(24) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5468. (b) Lee, C.;
Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
8118
dx.doi.org/10.1021/jp2039493 |J. Phys. Chem. A 2011, 115, 8113–8118