Dependence of activation parameters for phenylchlorocarbene-alkene additions on alkane solvent chain length

Article abstract of DOI:10.1016/j.tetlet.2014.08.098

Activation energies, enthalpies, and entropies for the addition of phenylchlorocarbene to tetramethylethylene significantly decrease as the reaction solvent lengthens from n-pentane to n-octane to n-decane; additional decreases are minimal in n-pentadecan

Full text of DOI:10.1016/j.tetlet.2014.08.098

Tetrahedron Letters 55 (2014) 6016–6018
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Dependence of activation parameters for phenylchlorocarbene–
alkene additions on alkane solvent chain length
⇑
⇑
Robert A. Moss , Lei Wang, Karsten Krogh-Jespersen
Department of Chemistry and Chemical Biology, Rutgers, The State University of New Jersey, New Brunswick, NJ 08903, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Activation energies, enthalpies, and entropies for the addition of phenylchlorocarbene to tetramethyleth-
ylene significantly decrease as the reaction solvent lengthens from n-pentane to n-octane to n-decane;
additional decreases are minimal in n-pentadecane and n-heptadecane. Electronic structure calculations,
employing a continuum solvent model, fail to reproduce the observations; instead, a qualitative model
invoking solvent cage effects is proposed.
Received 13 August 2014
Accepted 25 August 2014
Available online 1 September 2014
Keywords:
Carbenes
Ó 2014 Elsevier Ltd. All rights reserved.
Computations
Cyclopropanation
Kinetics
The additions of reactive carbenes such as phenylchlorocarbene
(PhCCl),1,2 dichlorocarbene,3 and chlorotrifluoromethylcarbene4 to
tetramethylethylene (TME) exhibit negative activation energies.
Two rationalizations have been offered for these unusual observa-
tions. (1) The carbene addition is a 2-step process proceeding via a
reversibly formed carbene–alkene complex, in which case kobs is
not a single rate constant and the observed activation energy
reflects the balance of activation energies for the formation/
reversion of the complex and its progression to product.1,2 (2)
The negative activation energy reflects a continuous decrease in
enthalpy all along the reaction coordinate. In this model, a barrier
to addition exists on the free energy surface due to unfavorable
entropic factors.5
TME in a series of alkane solvents of chain lengths 5, 8, 10, 15,
and 17. Our experimental and computational results are presented
here.
PhCCl was generated by laser flash photolysis (LFP) at 351 nm of
phenylchlorodiazirine10 in alkane solutions containing various
concentrations of TME. The disappearance of PhCCl was monitored
at 320 nm, corresponding to the intense
p ? p transition of the
carbene.11 Rate constants (kadd) for additions of PhCCl to TME were
obtained from correlations of kobs for the disappearance of the
carbene versus [TME]. Separate experiments showed the rate
constants to be reproducible to 2%.
In order to insure that enhanced C–H insertion of PhCCl into the
longer alkane solvents did not bias the rate constants, cyclopro-
pane product yields in various solvents were monitored by GC
relative to a tridecane internal standard. The product yields were
reasonably constant and insertion products were not observed.
See the Supplementary data for details.
Activation parameters were obtained from correlations of kadd
versus 1/T at 5 temperatures in each alkane solvent; cf. Table 1.
Precise temperatures ( 0.1 K) were measured at the instant of
LFP by a thermocouple immersed in the reaction solution. All kadd
determinations and Arrhenius correlations are shown in the Sup-
plementary data.
The activation parameters for addition of PhCCl to TME in pen-
tane are very similar to those previously reported for the reaction
in isooctane.2 However, there is a clear dependence of these
parameters on chain length as the solvent is varied from pentane
to octane to decane: the activation enthalpy (energy) decreases
from À2.4(À1.8) to À3.9(À3.3) to À5.8(À5.2) kcal/mol. The values
in decane are the most negative that we have yet observed in
Spectroscopic6 and computational6,7 evidence for carbene–arene
complexes is plentiful, but direct evidence for carbene–alkene com-
plexes is lacking and computational results are equivocal.3,5 Never-
theless, to satisfy the kinetics of model (1), a strongly interacting
‘bonded’ carbene–alkene complex is not required; a less distinct
‘proximity pair’ of reactants in a solvent cage would suffice.1,2
Indeed, this scenario is also compatible with model (2).
We imagined that the formation efficiency and robustness of
solvent cages in alkane solvents would be related to chain length,
given that solvent viscosity increases8 and solute diffusion rates
decrease9 as the solvent chain lengthens. Accordingly, we deter-
mined the activation parameters for the addition of PhCCl to
⇑
Corresponding authors. Tel.: +1 848 445 2606; fax: +1 848 445 5312 (R.A.M.);
tel.: +1 848 445 4241; fax: +1 848 445 5312 (K.K.-J.).
E-mail addresses: moss@rutchem.rutgers.edu (R.A. Moss), krogh@rutchem.
rutgers.edu (K. Krogh-Jespersen).
http://dx.doi.org/10.1016/j.tetlet.2014.08.098
0040-4039/Ó 2014 Elsevier Ltd. All rights reserved.

R. A. Moss et al. / Tetrahedron Letters 55 (2014) 6016–6018
6017
Table 1
Experimental activation parameters for additions of PhCCl to TMEa
Solvent
T (K)
TME (mmol)
Ea (kcal/mol)
D
Hà (kcal/mol)
D
Sà (cal/K mol)
D
Gà (kcal/mol)
Pentane
Octane
Decane
Pentadecane
Heptadecane
276–304
283–322
283–324
292–321
299–326
1.5–7.0
0.75–3.75
0.6–3.0
0.6–3.0
0.6–3.5
À1.77 0.12
À3.28 0.23
À5.15 0.42
À5.76 0.37
À5.89 0.24
À2.37 0.12
À3.88 0.23
À5.75 0.42
À6.36 0.37
À6.49 0.24
À28.04 0.42
À33.39 0.77
À39.65 1.40
À41.12 1.20
À41.24 0.77
5.99 0.25
6.07 0.46
6.25 0.84
5.89 0.73
5.80 0.47
a
All solvents are n-alkanes. Errors are derived from least-squares analyses of the Arrhenius correlations. Both Ea
=
D
Hà + RT and
D
Gà
=
D
Hà À T
D
Sà are evaluated at
T = 298 K.
carbene–alkane additions, more negative than those for the addi-
tion of nucleophilic adamantanylidene to methyl acrylate in pen-
(
D
Hà = À6.4 kcal/mol,
D
Sà = À41 e.u., and
D
Gà = 5.9 kcal/mol (cf.
Hà =
Table 1)). In the idealized gas phase, the calculations predict
D
tane (D
Hà = À4.1 kcal/mol,
D
Sà = À38 e.u.,
D
Gà = 7. 1 kcal/mol).12
À8.2 kcal/mol,
D
Sà = À43 e.u., and
D
Gà = 4.5 kcal/mol, that is,
Extension of the solvent chain to pentadecane or heptadecane does
not seem to significantly affect the activation parameters relative
to the values in decane; the chain length effect appears to operate
between pentane, octane, and decane.
2–3 kcal/mol more favorable kinetically than in alkane solvent.
Overall, the computed activation parameters are quite similar for
all model solvents and do not predict the substantial changes in
D
Hà observed as a function of chain length; the perhaps weakly
We also note that, as expected,5 barriers to the addition reac-
tions appear in the free energies as a consequence of the very neg-
ative entropies of activation. Moreover, in solvents pentane,
expressed trend shown by the computed
from experimental observations. The continuum solvent model
clearly fails to reproduce the significant changes observed in
Hà
and
Sà as the solvent chain length increases.
A PhCCl/TME complex was located in both idealized gas phase
and model alkane solvent calculations. The complexes are posi-
tioned early on the reaction coordinate (Fig. 1) and strongly resem-
D
Sà values is reversed
D
octane, and decane,
D
Sà becomes more negative as
D
Hà decreases.
D
As a result, the positive change in (ÀT
D
Sà) offsets the decrease in
D
Hà, and
D
Gà is effectively constant (ꢀ6 kcal/mol) over the entire
range of solvents. This apparent compensation between
D
D
Hà and
Sà appears to be a common feature of the additions of poorly
ble TS’s structurally, reflecting both carbene–alkene
p-type
stabilized carbenes to highly nucleophilic alkenes like TME.3,4
Recently, we demonstrated that electronic structure calcula-
tions based on density functional theory (DFT), employing the
MN12-SX functional and 6-311+G(d) basis sets,13,14 successfully
described the energetics associated with alkene cycloaddition of
another highly reactive carbene, adamantanylidene.12 From analo-
gous MN12-SX/6-311+G(d) calculations, applying the detailed
SMD continuum model of Marenich, Cramer, and Truhlar for sol-
ute–solvent interactions,15 we obtain the activation parameters
for PhCCl–TME cycloaddition shown in Table 2 (see the Supple-
mentary data for computational details).
electronic interactions and phenyl-methyl -C–H van der Waals-
p
type interactions. In all cases, the complex is both enthalpically
and entropically favored relative to the adjacent cyclopropanation
TS (Table 2). The complexes are 2–3 kcal/mol below the respective
TS’s in free energy, which may suggest that they possess at least
transient stability and hence finite lifetime; however, potential
complex formation will not directly influence the reaction kinetics.
Also, the standard free energies of complex formation are positive
by several kcal/mol (relative to the separated reactants; Table 2),
so we do not anticipate a significant concentration of PhCCl–TME
complex to accumulate in alkane solution.
We note that the calculations predict a steady but miniscule
decrease in activation enthalpy with increasing alkane chain length,
How do we rationalize the unusual solvent effect described
above? The additions of PhCCl to TME in decane, pentadecane, or
viz. computed
D
Hà = À6.1 kcal/mol in pentane and À6.3 kcal/mol in
heptadecane display
D
Hà of ca. À5 to À6 kcal/mol and
D
Sà ꢀÀ40
pentadecane. The computed activation entropies in the model sol-
e.u. The failure of the electronic structure calculations, employing
an implicit solvent model, to confirm the observed trends suggests
that the measured effects are not manifestations of general
solvent–solute interactions; rather, they must be associated with
specific interactions operating at the microscopic level. We pro-
pose that these reactions occur in fully or partially formed solvent
cages of low energy in which reactant diffusion is constrained, not
only due to the presence of the cage, but likely also from the forma-
tion of PhCCl–TME ‘proximity pairs’ or even complexes. An
increase in accessible solvent surface area should increase cohesive
solute–solvent interactions; hence, a longer aliphatic chain should
be able to form more robust and stabilizing cages than a shorter
vents pentane and octane are nearly identical (
À46 e.u.) and hence the activation free energies (
vents are very similar:
D
Sà = À45 to
D
Gà) for these sol-
D
Gà = 7.6 kcal/mol and 7.3 kcal/mol, respec-
tively. For decane and pentadecane, on the other hand, we calculate
nearly identical activation entropies of À43 e.u. and the computed
activation energies in these solvents (ꢀ6.5 kcal/mol) hence differ
by less than 0.1 kcal/mol; they are slightly less than those computed
for pentane and octane, however. Incidentally, for the specific case
of pentadecane the calculations predict
D
ment with the experimental results for this particular solvent
D
Hà = À6.3 kcal/mol,
Sà = À43 e.u., and
D
Gà = 6.4 kcal/mol, which is in very good agree-
Table 2
Computed (MN12-SX/6-311+G(d), SMD solvent model) activation parameters (
D
Hà,
D D DH, DS, and DG)
Sà, and Gà) for PhCCl–TME cycloaddition and thermodynamic parameters (
for PhCCl–TME complex formation in simulated gas phase or alkane solventa
Solvent
D
Hà (kcal/mol)
D
Sà (cal/K mol)
D
Gà (kcal/mol)
D
H (kcal/mol)
D
S (cal/K mol)
DG (kcal/mol)
Pentane
Octane
Decane
Pentadecane
Gas phase
À6.11
À6.19
À6.26
À6.32
À8.22
À45.9
À45.4
À42.7
À42.8
À42.6
7.58
7.34
6.47
6.44
4.49
À6.75
À6.85
À6.92
À6.99
À8.81
À40.1
À39.0
À36.1
À36.0
À34.7
5.22
4.78
3.84
3.74
1.54
a
All solvents are modeled as n-alkanes. Solvent parameters for experimental solvent heptadecane (cf. Table 1) are not available. Activation and thermodynamic parameters
are evaluated at P = 1 atm, T = 298 K.

6018
R. A. Moss et al. / Tetrahedron Letters 55 (2014) 6016–6018
Figure 1. Illustrative views (front, side on, and top) of the PhCCl–TME complex (MN12-SX/6-311+G(d), pentadecane model solvent). Qualitatively, the transition state for
cycloaddition looks similar to the computed complexes/TS’s for other model solvents and the gas phase. Some specific geometric parameters obtained with pentadecane
model solvent for the complex are C1–C19 = 2.902 Å, C2–C19 = 3.091 Å, <C2C1C19 = 84.9°, and <C3C1C19C21 = À161.7°; and for the TS, C1–C19 = 2.322 Å, C2–C19 = 2.640 Å,
<C2C1C19 = 87.3°, and <C3C1C19C21 = À164.8°.
one. However, there ought to be an imprecise upper limit to how
many methyl/methylene groups can make substantial contact with
the solute. Decane appears to be long enough to provide efficient
caging and van der Waals stabilization of the PhCCl–TME reactant
pair or complex, as well as the transition state for addition.
Solvents longer than decane confer little additional stabilization
even if the cages they provide might be more robust.
References and notes
1. Turro, N. J.; Lehr, G. F.; Butcher, J. A., Jr.; Moss, R. A.; Guo, W. J. Am. Chem. Soc.
1982, 104, 1754–1756.
2. Moss, R. A.; Lawrynowicz, W.; Turro, N. J.; Gould, I. R.; Cha, Y. J. Am. Chem. Soc.
1986, 108, 7028–7032.
3. Moss, R. A.; Wang, L.; Zhang, M.; Skalit, C.; Krogh-Jespersen, K. J. Am. Chem. Soc.
2008, 130, 5634–5635.
4. Moss, R. A.; Wang, L.; Krogh-Jespersen, K. J. Org. Chem. 2013, 78, 11040–11044.
5. (a) Houk, K. N.; Rondan, N. G.; Mareda, J. Tetrahedron 1985, 41, 1555–1563; (b)
Houk, K. N.; Rondan, N. G.; Mareda, J. J. Am. Chem. Soc. 1984, 106, 4291–4293;
(c) Houk, K. N.; Rondan, N. G. J. Am. Chem. Soc. 1984, 106, 4293–4294.
6. (a) Moss, R. A.; Wang, L.; Odorisio, C. M.; Krogh-Jespersen, K. J. Am. Chem. Soc.
2010, 132, 10677–10679; (b) Moss, R. A.; Tian, J.; Sauers, R. R.; Krogh-Jespersen,
K. J. Am. Chem. Soc. 2007, 129, 10019–10028; (c) Moss, R. A.; Wang, L.;
Weintraub, E.; Krogh-Jespersen, K. J. Phys. Chem. A 2008, 112, 4651–4659; (d)
Moss, R. A.; Wang, L.; Odorisio, C. M.; Krogh-Jespersen, K. J. Phys. Chem. A 2010,
114, 209–217.
With solvents shorter than decane, containment of the reactant
pair or complex is less efficient and diffusion increases. The general
reduction in solution reaction entropies, relative to the (hypothet-
ical) gas phase value (cf. Tables 1 and 2), is presumably related at
least partially to translational and rotational motions becoming
more restricted in the condensed phase.16 Hence, in shorter chain
alkane solvents, more energy resides in translational and rotational
modes of the reactants and transition states; consequently, less
energy is released upon incomplete reactant pair formation, and
7. Krogh-Jespersen, K.; Yan, S.; Moss, R. A. J. Am. Chem. Soc. 1999, 121, 6269–6274.
8. The viscosities (
g
) of the alkane solvents used here are (units of cP, T = 303 K):
(pentane) = 0.215; (octane) = 0.485; (decane) = 0.805;
(pentadecane) = 2.16; and (heptadecane) = 3.98. Lu, H. In Shi you hua gong ji
g
g
g
g
D
Hà is less negative in the shorter alkane chain solvents. At the
g
same time, there is an entropic cost to be paid for the decrease
in freedom of the reactant pair or complex in the cages formed
chu shu ju shou ceTranslation: Handbook of Petrochemical Elementary Data;
Chemical Industry Press: Peking, China, 1982; pp 150–176. and its sequel
edition, Ma, P.S. Chemical Industry Press: Peking, China, 1993; pp 234–238.
9. The Stokes–Einstein equation, for example, relates diffusion constants to the
inverse of the viscosity. Einstein, A. Ann. der Physik 1905, 17, 549–560.
10. Graham, W. H. J. Am. Chem. Soc. 1965, 87, 4396–4397. The diazirine absorbance
was 0.5.
11. (a) Turro, N. J.; Butcher, J. A., Jr.; Moss, R. A.; Guo, W.; Munjal, R. C.; Fedorynski,
M. J. Am. Chem. Soc. 1980, 102, 7576–7578; (b) Pliego, J. R., Jr.; DeAlmeida, W.
B.; Celebi, S.; Zhu, Z.; Platz, M. S. J. Phys. Chem. A 1999, 103, 7481–7486.
12. Moss, R. A.; Wang, L.; Krogh-Jespersen, K. J. Am. Chem. Soc. 2014, 136, 4885–
4888.
by the longer alkane chains, and
cases. Qualitatively, in alkane solution the enthalpic descent is
steeper in longer chain solvents; as a result,
Hà becomes more
negative, as does
Sà, with increasing alkane chain length. The
D
Sà is more negative in these
D
D
PhCCl–TME contact pair/complex and transition state for cycload-
dition are positioned slightly later along the reaction coordinate in
longer chain alkane solvents.
13. Peverati, R.; Truhlar, D. G. Phys. Chem. Chem. Phys. 2012, 14, 16187–16191.
14. (a) Ditchfield, R.; Hehre, W. J.; Pople, J. A. J. Chem. Phys. 1971, 54, 724–728; (b)
Hariharan, P. C.; Pople, J. A. Mol. Phys. 1974, 27, 209–214; (c) Raghavachari, K.;
Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys. 1980, 72, 650–654; (d) Clark,
T.; Chandrasekhar, J.; Spitznagel, G. W.; Schleyer, P. v. R. J. Comp. Chem. 1983, 4,
294–301; (e) McLean, A. D.; Chandler, G. S. J. Chem. Phys. 1980, 72, 5639–5648.
15. Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B 2009, 113, 6378–
6396.
Acknowledgments
We are grateful to the National Science Foundation (United
States) for financial support.
Supplementary data
16. Mammen, M.; Shakhnovich, E. I.; Deutch, J. M.; Whitesides, G. M. J. Org. Chem.
1998, 63, 3821–3830.
Supplementary data associated with this article can be found, in
theonlineversion, at http://dx.doi.org/10.1016/j.tetlet.2014.08.098.