Intermolecular insertion of an N,N-heterocycIic carbene into a nonacidic C-H bond: Kinetics, mechanism and catalysis by (K-HMDS)2 (HMDS = Hexamethyldisilazide)

Article abstract of DOI:10.1002/chem.200600266

The reaction of 2-[¹³C]-1-ethyl-3-isopropyl-3,4,5,6- tetrahydropyrimidin-1-ium hexafluorophosphate ([¹³C ₁]-1-PF₆) with a slight excess (1.03 equiv) of dimeric potassium hexamethyldisilazide ("(K-HMDS)₂

Full text of DOI:10.1002/chem.200600266

FULL PAPER
DOI: 10.1002/chem.200600266
Intermolecular Insertion of an N,N-Heterocyclic Carbene into a Nonacidic
ꢀ
C H Bond: Kinetics, Mechanism and Catalysis by (K-HMDS)2
(HMDS=Hexamethyldisilazide)
Guy C. Lloyd-Jones,* Roger W. Alder, and Gareth J. J. Owen-Smith[a]
Abstract: The reaction of 2-[13C]-1-
ethyl-3-isopropyl-3,4,5,6-tetrahydropyr-
(kH/kD =1.18
A
({x2 kuncat}+{(1ꢀx2)kcat})[2]1
[toluene]1,
ACHTREUNG
effects on the deuteration of the tol-
uene methyl group. The reaction is cat-
alysed by K-HMDS, but proceeds with-
out cross over between toluene methyl
protons and does not involve an
HMDS anion acting as base to gener-
ate a benzyl anion. Detailed analysis of
the reaction kinetics/kinetic isotope ef-
fects demonstrates that a pseudo-first-
order decay in 2 arises from a first-
order dependence on 2, a first-order
dependence on toluene (in large
excess) and, in the catalytic manifold, a
complex noninteger dependence on the
K-HMDS dimer. The rate is not satis-
factorily predicted by equations based
on the Brønsted salt-effect catalysis
law. However, the rate can be satisfac-
torily predicted by a mole-fraction-
weighted net rate constant: ꢀd[2]/dt=
in which x2 is determined by a standard
bimolecular complexation equilibrium
term. The association constant (Ka) for
rapid equilibrium–complexation of 2
imidin-1-ium
([13C1]-1-PF6) with
(1.03 equiv) of dimeric potassium hexa-
hexafluorophosphate
a
slight excess
A
G
with (K-HMDS)2 to form [2ACHTREUNG(K-
toluene generates 2-[13C]-3-ethyl-1-iso-
propyl-3,4,5,6-tetrahydropyrimid-2-yli-
dene ([13C1]-2). The hindered meta-
HMDS)2] is extracted by nonlinear re-
gression of the 13C NMR shift of C(2)
in [13C1]-2 versus [(K-HMDS)2] yield-
stable N,N-heterocyclic carbene [ C1]-2
ing: Ka =62(ꢁ7)mꢀ1
;
dC(2) in 2=
thus generated undergoes a slow but
quantitative reaction with toluene (the
solvent) to generate the aminal 2-[13C]-
2-benzyl-3-ethyl-1-isopropylhexahydro-
pyrimidine ([13C1]-14) through formal
237.0 ppm; dC(2) in [2ACHTRENU(G K-HMDS)2]=
226.8 ppm. It is thus concluded that
there is discrete, albeit inefficient, mo-
lecular catalysis through the 1:1 car-
A
G
complex
[2ACHTREUNG(K-
ꢀ
C H insertion of C(2) (the “carbene
HMDS)2], which is found to react with
toluene more rapidly than free 2 by a
factor of 3.4 (=kcat/kuncat). The greater
carbon”) at the toluene methyl group.
Despite
a significant pKa mismatch
(DpKa 1+ and toluene estimated to be
ca. 16 in DMSO) the reaction shows all
the characteristics of a deprotonation
mechanism, the reaction rate being
strongly dependent on the toluene para
reactivity of the complex [2ACHTREUNG(K-
HMDS)2] over the free carbene (2)
may arise from local Brønsted salt-
effect catalysis by the (K-HMDS)2 lib-
AHCTREUNG
Keywords: carbenes
neous catalysis · isotope effects ·
kinetics · potassium
substituent (1=4.8
ing substantial and rate-limiting pri-
mary (kH/kD =4.2
(ꢁ0.6)) and secondary
(ꢁ0.3)), and display-
AHCTREUNG
ACHTREUNG
Introduction
particular have found widespread application as ligands in
transition-metal catalysis,[3] as catalysts in their own right for
organic transformations[4] and as chemical entities that fasci-
nate and challenge the theoretical and practical chemist
alike. As part of a research programme exploring the mech-
Since the early demonstrations of the surprising kinetic, and
even thermodynamic, stability of heteroatom-stabilised car-
benes towards dimerisation,[1] the area has enjoyed sustained
interest and development.[2] N,N-Heterocyclic carbenes in
[5]
anism of dimerisation of stable N,N-heterocyclic carbenes,
we recently prepared the tetrahydropyrimidinium salt 1-PF6
as a precursor for base-mediated generation of carbene 2,
which under certain conditions dimerises to yield tetraami-
noethylene derivative 3.[6]
[a] Prof. Dr. G. C. Lloyd-Jones, Prof. Dr. R. W. Alder,
G. J. J. Owen-Smith
The Bristol Centre for Organometallic Catalysis
School of Chemistry
University of Bristol, Cantockꢀs Close, Bristol BS81TS (UK)
Fax : (+44)117-929-8611
To aid the analysis of the kinetics of the dimerisation re-
E-mail: guy.lloyd-jones@bris.ac.uk
action by 13C{1H} NMR we prepared [13C1]-1, and thus
Chem. Eur. J. 2006, 12, 5361 – 5375
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5361

has become a general procedure for the generation of N,N-
heterocyclic carbenes.[8] Wanting to generate carbene 2 from
salt 1-PF6 without competing dimerisation[5,6] to form 3, we
treated 1-PF6, as
a vigorously agitated suspension in
[D8]toluene, under strictly anhydrous and anaerobic condi-
tions with a small excess (1.03 equiv) of the hindered dimer-
ic[9] base [K-N
ACHTERNGU(SiMe3)2]2 (“ACHRTE(UGN K-HMDS)2”, commercial
sample, 0.43m in toluene). This procedure exploits the ex-
tremely low solubility of the ionic precursor 1-PF6 and the
co-product, KPF6, in the rather apolar toluene medium,
thereby avoiding extensive dimerisation of carbene 2. Using
this method, we found we could smoothly generate 2 in a
sealed NMR tube, as evidenced by a low intensity singlet at
d=236 ppm arising from C(2) in the 13C{1H} NMR spec-
trum, which shows only small quantities of dimer 3. The
NMR tube was repeatedly agitated (sonication) and then
analysed by 13C{1H} NMR spectrum until >98% of the (K-
HMDS)2 (singlet at d=7.2 ppm) had reacted, thereby gener-
ating 2 in essentially quantitative yield[10] together with
[13C1]-2 and [13C2]-3, in which the carbene carbon C(2) in 2
is 13C-labelled. The resulting increase in sensitivity in the
13C{1H} NMR spectrum facilitated the detection of
a
number of species arising as byproducts (<ca. 2.5%) from
the generation of [13C1]-2 in toluene. Of these species,
aminal 4 was identified as a minor component, but found to
increase with time to become the major product. The identi-
fication of 4 and a detailed investigation of its mechanism of
generation from toluene forms the basis of the work pre-
sented herein. This is the first report of the formal intermo-
lecular insertion of a stable carbene into a “nonacidic”
HNACHTREU(NG SiMe3)2 (“H-HMDS”; singlet at d=2.6 ppm).
[7]
ꢀ
(pKa ꢂ25) C H bond. Indeed, even the reaction of stable
ꢀ
carbenes with “acidic” C H bonds (pKa ꢃ25) is still compa-
Side products arising from generation of carbene 2 from 1
ratively rare, being limited to the examples shown in
Scheme 1.
andK-HMDS : The asymmetry of 2 and the resulting com-
1
plexity of the H NMR spectrum of mixtures of 2, (E)-3 and
(Z)-3, prompted us to prepare [13C1]-2, in which C(2) bears
the 13C-label, so that we could monitor the kinetics of the di-
merisation of 2 by 13C{1H} NMR spectroscopy.[5] In doing so,
the signal at d=236 ppm arising from C(2) in 2 became ap-
proximately 99-fold stronger; we also noted a range of other
minor signals at d=162.1, 161.9, 161.7, 161.5, 106.1, 91.2 and
76.5 ppm in the 13C{1H} NMR spectrum that were not attrib-
utable to 2 (or 1 or 3) but were clearly arising from a 13C-la-
belled carbon, Figure 1. It should be noted that the intensity
of the signal arising from C(2) in 2 is relatively weak due to
poor proton-mediated nOe magnetisation and inefficient
{1H}-induced relaxation. A proton-coupled 13C{1H-gated}
ACHTREUNG
NMR spectrum revealed that the signals at d=162.1, 161.9,
ꢀ
161.7, 161.5, 106.1, 91.2 and 76.5 ppm arise from C H spe-
cies and thus their relative intensity is significantly higher
than C(2) in [13C1]-2. In other words, the 13C-labelling had
revealed the presence of some rather minor side products
(<ca. 2.5%).
Scheme 1. Formal insertion of stable carbenes 5,[7a,d] 7[7c] and 9[7b] into C
ꢀ
ꢀ
H bonds, the precursors for which (H Z) all have pKa ꢃ25, except for 9
for which the reaction 9!10 is intramolecular. Mes=2,4,6-trimethyl-
phenyl; Ad=adamantyl. Note that for the unsaturated analogue of 5,
The signals at d=106.1 and 91.2 ppm were found to
reduce in intensity or even disappear during, or shortly-
after, the reaction with an accompanying increase in the sig-
nals at d=161.5–162.1 ppm and at d=236 ppm ([13C1]-2).
Based on the 13C-chemical shifts, we speculated that the sig-
nals arose from the moieties of the type R2N-13C(=O)H (d=
162.1, 161.9, 161.7, and 161.5 ppm); (R2N)213C(H)-O (d=
which is aromatic, reaction with CH3CNfails, whilst reaction with CHCl
3
gives complex mixtures.[7a]
The reaction of 2 with toluene (pKa ꢄ42–44) is surprising:
there is a significant pKa mismatch between reactants (DpKa
1 and toluene ca. 16–18) and the reaction is found to be cat-
alysed by dimeric potassium hexamethyldisilazide (“
HMDS)2”).
A
106.1 ppm);
(R2N)213C(H)-N(
d=91.2 ppm)
and
(R2N)213C(H)-C (d=76.5 ppm), in which the labelled carbon
atom derives from the precursors [13C1]-1 or [13C1]-2. The
first two assignments were confirmed by reaction of [13C1]-1
Results andDiscussion
with anhydrous KOH/ACHTER(UNG K-HMDS)2 in [D6]benzene
(Scheme 2); this reaction gave a mixture of alkoxide [13C1]-
11 and formamides [13C1]-12H and [13C1]-13H (each display-
ing two rotameric amide isomers in the 13C{1H} NMR spec-
ꢀ
The C H deprotonation reaction of N,N,N’,N’-tetraalkylated
formamidinium salts with sterically hindered strong bases
5362
www.chemeurj.org
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5361 – 5375

Insertion Reactions
FULL PAPER
trum). Addition of further (K-HMDS)2 partially converted
formamides [13C1]-12H and [13C1]-13H to the alkoxide
[13C1]-11, whilst exposure to moisture (air) converted the
alkoxide [13C1]-11 to the neutral formamides [13C1]-12H and
[13C1]-13H, presumably by means of an equilibrium with the
ring-opened nitrogen anions [13C1]-12 and [13C1]-13 (not ob-
served), which can deprotonate water. Interestingly, 12H
1
and 13H, the identities of which were assigned by H nOe/
CH correlation, are initially generated in a 1:1 ratio, but un-
dergo slow interconversion, presumably via 11 (or 11H), to
give an equilibrium mixture that favours 13H (K=9). It is
thus concluded that 11, 12H and 13H arise in the genera-
tion of carbene 2, by reaction of salt 1 with traces of KOH
in the commercial (K-HMDS)2 reagent.
The signal at d=91.2 ppm, assigned as that arising from a
(R2N)213C(H)-Nunit, was suggestive of (K-HMDS) effect-
2
ing nucleophilic attack at C(2) of [13C1]-1, in competition
with deprotonation, to generate the HMDS adduct [13C1]-14
(Scheme 2). A competing nucleophilic attack by Li-HMDS
ꢀ
in the C H deprotonation reaction of related N,N’-dialkylat-
ed tetrahydropyrimidinium salts has been reported by Alder
et al.[11] The decrease in the intensity of the signal of [13C1]-
14 during or shortly after complete reaction suggests that
the addition of the HMDS anion to 1 is reversible, as indeed
is found to be the case, vide infra.
Finally, on close inspection, the signal at d=76.5 ppm,
was found to have a neighbouring multiplet (d=75.8 ppm)
of very low intensity, but with an equipollent triplet struc-
ture characteristic of a deuterium-bearing carbon atom (1J-
Figure 1. The 13C{1H} NMR sub-spectrum (60–240 ppm) of the product
mixture obtained on sonication of
a
suspension of [13C1]-1-PF6 in
[D8]toluene after addition of 1.03 equiv of (K-HMDS)2 (0.43m in tol-
uene). For full assignments see text and Scheme 2. Note that 2 lacks a
proton at C(2) and thus the signal intensity is about one third of that of
ꢀ
C H compounds 4, 11, 12H and 13H. The signal identified by “ ” is un-
*
assigned, but is similar in chemical shift to C(2) of [1]-PF6 (d=155 ppm)
which is only sparingly soluble in toluene.
AHCTRE(UNG C,D)=21.5 Hz). Since the [D8]toluene, used as a co-solvent
with the commercial toluene (K-HMDS)2 solution, was the
only source of deuterium, we repeated the procedure using
freshly prepared (K-HMDS)2 (prepared from KH and H-
HMDS)[12] in [D8]toluene: the intensity of the triplet was
found to be approximately the same, but there was essential-
ly no trace of the singlet at d=76.5 ppm, and the amount of
alkoxide/formamides (11/12H and 13H) generated was re-
duced. The tentative assignment of the signal at d=
76.5 ppm as that of C(2) in the toluene adduct [13C1]-4 was
confirmed by independent, albeit inefficient, preparation of
4 from phenacetaldehyde.[13] When pure [D8]toluene is used
as solvent, the major side product from reaction of [13C1]-1
with (K-HMDS)2 is [2H8]
suspected that this was generated by addition of
[D7]benzyl anion to [13C1]-1-PF6. However, it soon became
[13C1]-4, vide infra, and we initially
ACHTREUNG
a
evident that the concentration of [2H8][13C1]-4 continued to
ACHTREUNG
increase after all of the [13C1]-1-PF6 had been consumed,
thus implicating the involvement of [13C1]-2 rather than
[13C1]-1.
Kinetics of reaction of carbene
2 with [D8]toluene—
K-HMDS catalysis: As noted above, the reaction of [13C1]-1
with (K-HMDS)2 in [D8]toluene to generate [13C1]-2, in-
volves alternating periods of sonication then analysis by
13C{1H} NMR until generation of [13C1]-2 is complete. Al-
though the sonication results in a stochastic profile to the
evolution of [13C1]-2, the profile for the generation of [2H8]-
Scheme 2. Side/co-products arising from generation of [13C1]-2 by reaction
of [13C1]-1-PF6 with commercial solutions of (K-HMDS)2 in toluene
(0.43m) as identified by 13C{1H} NMR spectroscopy in [D8]toluene (see
Figure 1).
Chem. Eur. J. 2006, 12, 5361 – 5375
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5363

G. C. Lloyd-Jones et al.
Table 1. Pseudo-first-order rate constants[a] (kobs) for reactions of excess
[D8]toluene, toluene, substituted toluenes (4-CF3-toluene (15), 4-Cl-tol-
uene (16), 4-tBu-toluene (17) and ethylbenzene (18); 40–118 equiv), with
carbene [13C1]-2 (generated in situ from [13C1]-1-PF6) in [D8]toluene or
[D6]benzene solvent medium, in the presence of excess (K-HMDS)2 at
238C.
ACHTREUNG
[13C1]-4 versus time was found to be rather smooth—in
other words, the rate appeared to be unrelated to the con-
centration of [13C1]-2. However, we reasoned that since the
generation of [13C1]-2 from [13C1]-1-PF6 is directly coupled to
the consumption of (K-HMDS)2 and thus as the concentra-
tion of [13C1]-2 rises, the concentration of (K-HMDS)2 falls
linearly,[14] a compensatory effect would arise if the K-
HMDS catalysed the reaction of [13C1]-2 with the toluene.
To explore this possibility, we prepared 0.08m [D8]toluene
solutions of [13C1]-2 from [13C1]-1 with larger excesses of (K-
Arene
A
ACHTREUNG
[m]
ACHTREUNG
1
2
3
4
5
6
7
8
9
[D8]toluene
[D8]toluene
[D8]toluene
[D8]toluene
[D8]toluene
[D8]toluene
[D8]toluene
[D8]toluene
[D0]toluene
9.32
9.29
9.22
9.14
9.02
8.61
9.02
9.02
2.81
2.78
2.78
2.77
2.81
79
81
81
82
81
85
163
39
67
69
68
68
69
68
24
5.6
20.0
50.1
83.3
134
308
93.7
156
68
68
1.64
2.40
3.29
3.49
4.07
4.40
3.39
4.10
9.46
ACHTREUNG
ACHTREUNG
ACHTREUNG
ACHTREUNG
[15]
HMDS)2 (1.1, 1.5, 2.2, 3.0, 4.3 and 8.3 formal equivalents
ACHTREUNG
of HMDS anion). The solutions of [13C1]-2 were generated
ACHTREUNG
very cleanly over a period of ca. 10 min, with [2H8]
A
ACHTREUNG
ACHTREUNG
essentially the only side product (ꢃ1%). Over a period of
ACHTREUNG
ACHTREUNG
days, [2H8][13C1]-4 was found to grow smoothly at the ex-
[c]
10 15
11 16
12 17
13 18
–
pense of [13C1]-2.
68
68
68
68
67.8
A
The reactions were found to evolve through a logarithmic
decay in [13C1]-2, giving good linear correlations, Figure 2
(left). An additional two samples with half and double the
initial concentrations of [13C1]-2 behaved analogously, thus
allowing determination of macroscopic first-order rate con-
stants kobs (sꢀ1) over a range of initial concentrations of
[13C1]-2, (K-HMDS)2 and HMDS, Table 1, entries 1–8.
2.14
1.77
6.09
4.32
ACHTREUNG
ACHTREUNG
14 [D1]toluene[d] 2.94
ACHTREUNG
15 [D8]toluene
9.02
163
ACHTREUNG
[a] Determined by linear regression of ln([2]0/[2]t) versus t, as determined
by 1H/13C NMR analysis, see Figure 2. [b] The formal concentration of
excess (K-HMDS)2 based on consumption of 0.5 equivalents (K-HMDS)2
in the deprotonation of [1]-PF6. [c] Reaction complete (>95%) within
29 min. A value of 1.72”10ꢀ3 sꢀ1 is estimated as the lower limit for kobs
.
[d] a-[D1]Toluene (>98.8% 2H1), contains about 1% diethyl ether.
ꢀ
ꢀ
[e] The partitioning is 87.6% C H insertion, 12.4% C D insertion.
[f] Reaction after filtration from precipitated KPF6 (0.45 mm PTFE mem-
brane) then 3.5-fold dilution into 0.175m (K-HMDS)2 in [D8]toluene.
Kinetics of reaction of carbene 2 with substitutedtoluenes
in [D6]benzene—linear free energy relationship andkinetic
ꢀ
isotope effects: During the reaction that generates 4, the C
H bond of the toluene methyl group is cleaved, allowing
formal insertion of C(2) of 2, see scheme in Figure 3. To
ꢀ
probe the polarity of the C H cleavage, we determined the
pseudo-first-order rate constants for reaction of [13C1]-2,
generated in C6D6 from (K-HMDS)2 (2.0 equiv HMDS
anion) with separate samples of excess (ca. 40 equiv) tol-
uene, 4-CF3-toluene (15), 4-Cl-toluene (16), 4-tert-butyltol-
Figure 2. Left-hand graph: typical examples of the linear relationship be-
tween time and ln([2]0/[2]t) at various concentrations of (K-HMDS)2 for
the reaction of 2 with [D8]toluene (solvent) to generate 4 (numbers II,
IV and VI refer to entries 2, 4, and 6, respectively, in Table 1). Right-
hand graph: the complex relationship between (K-HMDS)2 concentra-
AHCTREuUNG ene (17) and a-methyltoluene (ethylbenzene, 18). All five
13
ꢀ
substrates cleanly gave carbene ([ C1]-2) aryl C H insertion
products, ([13C1]-4, [13C1]-19, [13C1]-20, [13C1]-21 and [13C1]-22
(0% de), Figure 3) as evidenced by intense signals in the
d=76.1–81.9 ppm region.
tion and the normalised pseudo-first-order rate constants kobs
/
[[D8]toluene]0 (as determined by linear regression (left-hand graph) of
data from Table 1, entries 1–8). The solid line passing through data points
is a nonlinear regression based on a model involving catalysis by (K-
The rates of reaction of [D0]toluene, 15, 16 and 17
(Table 1, entries 9, 10, 11 and 12) yielded a relatively large
HMDS)2 complexation: Ka =62mꢀ1, kuncat =1.46”10ꢀ8 dm3 molꢀ1 sꢀ1, kcat
5.11”10ꢀ8 dm3 molꢀ1 sꢀ1, [2]=54.3 mm; see text for full discussion.
=
and positive 1 value (1ꢀ =4.8
ACHTREUNG
AHCTREUNG
correlation has a greater dependence on the resonance term
than on the field term (R/F=1.37), Figure 3. Both correla-
tions are indicative of a substantial increase in electron den-
sity at the aryl-bound carbon en route to the rate-limiting
transition state. The greater dependence on the resonance
term (R) in the Swain–Lupton correlation also suggests that
there is significant sp2 hybridisation at the benzylic carbon
atom in the transition state to facilitate p delocalisation. Re-
Although the effect of excess [K-HMDS] on the rate re-
sults in a complex relationship (Figure 2 right), in which
ꢀd[2]/dt/(fACHTREUNG[K-HMDS]) in which f=non-linear function,
vide infra, the large excess of [D8]toluene (ca. 55–231 equiv)
and constant nature of the (K-HMDS)2 and H-HMDS con-
centrations are consistent with a simple pseudo-first-order
relationship of the rate to carbene concentration, [[13C1]-2].
5364
www.chemeurj.org
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5361 – 5375

Insertion Reactions
FULL PAPER
conditions in toluene) we exploited the net kinetic isotope
ꢀ
effects (KIEs) associated with the C H(D) cleavage process,
vide supra, by conducting a series of reactions of 2 (0.08m)
with mixtures of [D0]toluene and [D8]toluene (0.0, 15.0,
30.0, 45.0 and 60.0% [D0]toluene; Table 2, entries 1–5).
Table 2. Pseudo-first-order rate constants (kobs),[a] net primary/secondary
kinetic isotope effects (kH/kD)[b] and partitioning ratios[c,d] for reaction of
carbene [13C1]-2 (82–84 mm) with [D0]toluene/[D8]toluene mixtures (9.13–
9.15m), in the presence of excess (K-HMDS)2 (82–83 mm) at 238C.
Mol ratio
[D]0/[D8]toluene
dC(2)
2
kobs
[sꢀ1”107]
kH/kD
Partitioning (P)
([D0]4/[D8]4)
A
[c]
1
2
3
4
5
–
229.3
229.4
228.9 14.9
229.8 21.2
(ꢁ0.2)
229.8 25.2
(ꢁ0.4)
3.49
9.26
(ꢁ0.02)
–
–
0.176
0.429
0.818
1.50
N
A
A
9.3
9.5
8.2
G
ACHTREUNG
G
U
ACHTREUNG
C
G
ACHTREUNG
[a] Determined by linear regression of ln([2]0/[2]t) versus t, as determined
by 1H/13C NMR analysis. [b] Determined by the relationship (k[D ]tol
/
0
k[D ]tol)ACHTREUNG
8
A
G
A
ACHTREUNG
AHCTREUNG
1
troscopy. [e] Determined by H NMR spectroscopy.
The rate of reaction of [13C1]-2 in combination with the
partitioning ratio of adducts 4 derived from [D0]toluene
versus [D8]toluene ([13C1]-4/[2H8][13C1]-4=P) can be em-
ACHTERNGU ACHTREUGN
ployed to extract the net deuterium KIEs on both the parti-
tioning[17] and the absolute rate. As with a-[D1]toluene, sig-
nificant partitioning (P) was found to occur, with the net
effect determined by the mole fractions [D0]toluene/
[D8]toluene, in conjunction with a net KIE that is essentially
constant (error-weighted average k[D ]tol/k[D ]tol =9.5ꢁ0.8).[18]
Figure 3. Linear free energy relationship of the relative rates of reaction
of [13C1]-2 with toluene, 4-CF3-toluene (15), 4-Cl-toluene (16) and 4-tert-
butyl-toluene (17) in [D6]benzene in the presence of (K-HMDS)2
(0.068m). y axis: rate data from Table 1, entries 9–12; x axis: substituent
constant sSL derived from Swain–Lupton parameters, sSL =(1.0”R)+
(0.73”F). A value of 1=4.8 (ꢁ0.3) is obtained in a standard Hammett
correlation employing s.
0
8
The mole fractions of [D0]toluene/[D8]toluene are also
found to affect the overall reaction rate (kobs in sꢀ1) in a
linear manner, Figure 4.
The gradient of the correlation kobs =xH(k(H)ꢀk(D))+k(D)
,
in which xH =mole fraction [D0]toluene, yields the net KIE
on the rate of reaction, k(H)/k(D) =10.9ꢁ1.9, which is in rea-
sonable agreement with the isotope effect attending parti-
tioning (k[D ]tol/k[D ]tol =9.5ꢁ0.8) and thus indicative of a uni-
action of a-[D1]toluene (Ph-CH2D) proceeded at approxi-
mately 0.64-fold the rate of [D0]toluene (compare entries 9
and 14, Table 1) indicating a substantial net kinetic isotope
effect, which also results in partitioning of the a-[D1]toluene
0
8
AHCTREmUNG olecular involvement of toluene in rate-limiting adduct 4
formation.[19] In all of the reactions studied, the product 4
was found to derive from a single molecule of toluene, with
no evidence for cross-over. Thus, [D8]toluene generates ex-
ꢀ
to give products arising from insertion into C H (0% de)
ꢀ
versus C D in an 87.6:12.4 ratio. From these values, the pri-
mary and secondary effects are estimated as kH/kD =4.2-
clusively [2H8]
[13C1]-4 (see Figure 5, right-hand spectrum),[20]
ACHTREUNG
A
G
[D1]toluene generates [2H1][13C1]-4 (as an 87.6:12.4 mixture
ACHTREUNG
ꢀ
ꢀ
tion for details. The reaction of 18 (Table 1, entry 13) was
found to be more than fivefold slower than that of
[D0]toluene.
of C(2) H/C(2) D) and [D0]toluene generates exclusively
[13C1]-4.
Mechanism of reaction of carbene 2 with toluene: The
mechanism of “insertion” of singlet carbenes (R-C-R; in
which one or both of “R” are “stabilising”: aryl, halogen,
Reaction of carbene 2 with [D0]toluene and[D 8]toluene
mixtures—toluene molecularity: The large excess of
ꢀ
[D8]toluene used to explore the effect of (K-HMDS)2 con-
alkoxy, amino etc.) into X H bonds (X=O, N, C) has long
2
ꢀ
centration on the rate of C H insertion (Table 1, entries 1–
been of interest. Three general classes of mechanism have
been proposed,[20] which can be summarised as: 1) a path-
way dominated by the electrophilic nature of the vacant p
orbital on carbon (“ylidic”; pathway I, Scheme 3); 2) a path-
8) results in a pseudo-zero-order rate dependency on the
[D8]toluene concentration under these conditions. To ex-
plore the toluene molecularity (under pseudo-zero-order
Chem. Eur. J. 2006, 12, 5361 – 5375
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5365

G. C. Lloyd-Jones et al.
Scheme 3. Three general mechanisms for the “insertion” of singlet car-
ꢀ
benes (R-C-R; one or both R=aryl, halogen, alkoxy, amino etc.) into X
[21]
H bonds (X=O, N, C). Note that the “ylide” mechanism (I) may in-
I
ꢀ
volve oligomeric X H species to facilitate proton transfer (k2).
certed”; pathway II) and 3) a pathway dominated by the nu-
cleophilic (basic) nature of the filled sp2 orbital on carbon
(“deprotonation”; pathway III). Essentially all of the mecha-
nistic investigations reported to date have centred around
the insertion of singlet carbenes (generated thermally or
photochemically from diazoalkanes[21a–d] or photochemically
from diazirenes)[21e,f] into the O H bond of H2O or ROH.
ꢀ
Distinction between pathways I, II and III for the inser-
ꢀ
tion of singlet carbenes into the O H bond of H2O or ROH
Figure 4. The linear relationship between the pseudo-first-order rate con-
stant (kobs) for reaction of [13C1]-2 with [D0]toluene/[D8]toluene (ca.
110 equiv) and the initial mol fraction [D0]toluene xH. The line passing
has predominantly been based on the magnitude of the pri-
mary KIEs (2H or H) for the hydrogen transfer and Brøn-
sted correlations (pKa of RO H versus logkobs). For path-
3
through data points is
xH(k(H)ꢀk(D))+k(D), in which (k(H)ꢀk(D))=36.9
3.74
(ꢁ0.5)”10ꢀ7 sꢀ1
a
linear regression of the equation kobs
=
ꢀ
A
AHCTREUNG
=
way I, a negligible primary KIE is expected,[21d] providing
that step 1 (kI1) is rate-limiting. The use of mixtures of D2O
(or T2O) with H2O allows the isotope effect in the second
step (kI2) to be determined through partitioning.[21a] For path-
way II, primary KIEs of an intermediate magnitude are ex-
pected due to the nonlinearity
A
.
way in which both the vacant p orbital and filled sp2 orbital
on carbon interact simultaneously with X and H (“con-
of H transfer in kI1I;[21f] a low
degree of charge separation
ꢀ
means that the pKa of RO H
will be of less influence. For
pathway III, larger primary
KIEs are expected, provided
that step 1 (kI1II) is rate-limit-
ing; equilibrium KIEs close to
unity would be expected if
step 2 (kI2II) is rate-limiting. A
Brønsted correlation would be
expected in both cases (unless
KI1II is very large).[21e] The
Figure 5. 13C{1H} NMR experiments probing for the involvement of the HMDS anion in the (K-HMDS)2-cata-
lysed reaction of 2 with toluene. Left-hand sub-spectrum: 50:50 v/v [D0]toluene/[D8]toluene containing 83 mm
(K-HMDS)2 +237 mm H-HMDS, 2 weeks at 238C. Right-hand sub-spectrum: (K-HMDS)2-catalysed reaction
of [13C1]-2 with [D8]toluene (99.1% ArCD3/0.9% ArCD2H) in the presence of 1 equiv H-HMDS, which gives
mechanism(s) involved in the
ꢀ
C H insertion reactions of
stable carbenes (Scheme 2)
have not been studied,[7] al-
though on the basis of the
96.2:3.8 [2H8]
discussion.
ACHTREUNG ACHTREUNG ACHTREUGN
[13C1]-4/[2H7][13C1]-4 in which >98% of the benzylic carbons are dideuterated; see text for full
5366
www.chemeurj.org
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5361 – 5375

Insertion Reactions
FULL PAPER
ꢀ
“acidic” nature of the C H components, namely, CH3CN,
CHCl3, PhSO2CH3 and acetylene, it has been proposed that
these all involve the “deprotonation”, pathway III.[7a,c]
The data obtained for the kinetics of the reaction of 2
with toluene, deuterated toluenes and substituted toluenes
versible deprotonation of the toluene methyl group, even
after 2 weeks at room temperature, Figure 5.
Although (K-HMDS)2 alone does not deprotonate tol-
uene to any significant extent, the presence of 2 may in-
crease the basicity of the HMDS anion, through complexa-
ꢀ
to generate the formal C H insertion products 4 and 19!22
tion of 2 with (K-HMDS)2. The complexation of N,N-het-
+
should, in principle, allow a more confident interpretation of
the mechanism. The reaction has the following characteris-
tics: 1) the rate of reaction depends on the ability of the
aryl ring to support negative charge development at the tol-
uene methyl group (1=4.8ꢁ0.3); 2) there are substantial
A
N
was first reported by Alder et al., and has subsequently
been explored in considerable detail.[25] The complexation
can be readily detected in solution by an upfield shift of the
signal arising from the carbene carbon in the 13C NMR spec-
trum. Analysis of the signal arising from C(2) in [13C1]-2 and
the Me signals in the (K-HMDS)2 at the various values of
excess (K-HMDS)2 reveals that complexation does occur
and is dynamic: a single time-average chemical shift of C(2)
in [13C1]-2 is observed that is weakly dependent on [[13C1]-2]
and strongly dependent on [(K-HMDS)2],[26] vide infra,
Figure 6 (left). The involvement of the HMDS anion de-
ꢀ
KIEs attending the toluene methyl C H cleavage: namely, a
net KIE, kH/kD =9.5ꢁ0.8 for [D8]toluene, and a primary kH/
kD =4.2
N
N
[D1]toluene; 3) the reaction rate has first-order dependency
on toluene, a first-order dependency on carbene (2), and a
complex dependence in (K-HMDS)2; 4) there is no evidence
for cross-over: each molecule of the product 4 is entirely de-
rived from a single molecule of 2 and a single molecule of
toluene and 5) the reaction is catalysed by (K-HMDS)2.
The large magnitude of the secondary KIEs and the de-
ꢀ
pendence of reaction rate on toluene methyl C H acidity
(1=4.8ꢁ0.3) and charge delocalisation into the aryl p
system (R/F=1.37 in the Swain–Lupton correlation,
ꢀ
Figure 3) are strongly suggestive of a C H deprotonation re-
action (cf. pathway III, Scheme 3) with a late transition state
in which substantial sp2 character and partial negative
charge has been established at the benzylic carbon.[18] How-
ever, the generation of a full benzyl anion by means of de-
protonation of the toluene methyl group would require a
very strong base: the pKa of toluene is about 42 in
DMSO.[22] Addition of a large excess (26 equiv) of H-
HMDS to [13C1]-2 in [D8]toluene, containing less than 1%
of the HMDS adduct [13C1]-8, resulted in the reasonably
rapid generation of an equilibrium mixture (ca. 83% [13C1]-
2/17% [13C1]-8), which on fourfold dilution into pure
[D8]toluene, underwent substantial reversion (>50%) to
[13C1]-2 and H-HMDS. This result ([2]+H-HMDS$[8]; K=
0.4ꢁ0.2mꢀ1) shows that the difference in thermodynamic
basicity of 2 and the HMDS anion is small. Analogously, the
acidities of H-HMDS (pKa 26 in DMSO)[22] and 1 (pKa of 1-
Cl estimated as 26 in DMSO)[23a] are rather similar. The
quantitative reaction of 0.5 equivalents (K-HMDS)2 with 1-
PF6 in [D8]toluene to generate 2, H-HMDS and KPF6, sug-
gests that the co-generation of KPF6 is a significant driving
force; in other words, the “acidity” of 1 is counterion de-
pendent.[23]
Figure 6. Left-hand graph: nonlinear regression of a 1:1 binding isotherm
accounting for variation in time-average 13C NMR chemical shift of C(2)
in [13C1]-2 with [(K-HMDS)2] in [D8]toluene (data points); solid line:
with freedom in all four coefficients, yielding Ka =62(ꢁ7)mꢀ1, [2]=53
A
(ꢁ0.2) ppm; dC(2) in [2
ACHTRE(UNG KHMDS)2]=226.8
A
graph: mole fraction of carbene 2 (54.3 mm) complexed to K-HMDS
dimer, calculated from [(K-HMDS)2] and Ka, versus apparent second-
order-rate constant for reaction of 2 with [D8]toluene. Ka =62mꢀ1 as de-
rived from 13C{1H} NMR shifts (left-hand graph).
rived from such a complex was probed by detailed 13C{1H}
NMR analysis of the C(2) signals in the product from reac-
tion of [13C1]-2/[D8]toluene/
ACHTREU(NG K-HMDS)2 (4.2 equiv)/H-
HMDS (1.0 equiv formed by deprotonation of salt [13C1]-1
by HMDS anion); such an analysis allows the deduction of
the presence of various isotopologues of 4 due to the multi-
plicity and isotope shifts induced by deuterium at 13C(2). By
reference to products from [D0]toluene and [D1]toluene,
vide supra, it is evident that the process generates [2Hn]-
Mulvey et al. recently reported that (K-HMDS)2 cleanly
deprotonates a methyl group in toluene, m-xylene and mesi-
tylene, when reaction is conducted in the presence of Zn-
ACHTREUNG
(HMDS)2.[24] Such a process could conceivably proceed
through ZnACHTREUNG(HMDS)2-trapping of PhCH2K generated in
trace concentration by equilibrium of PhCH3 with K-
HMDS. However, in a control experiment involving a mix-
ture of (K-HMDS)2, H-HMDS, [D0]toluene and [D8]toluene,
we found no evidence whatsoever for cross over through re-
ACHTREUNG
AHCTREUNG
the isotopic purity of the [D8]toluene solution (99.1% CD3,
0.9% CD2H; H NMR spectroscopy).[20] The involvement of
1
Chem. Eur. J. 2006, 12, 5361 – 5375
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5367

G. C. Lloyd-Jones et al.
the free HMDS anion as a base would generate D-HMDS/
H-HMDS, the latter liberating [2H7][13C1]-4, in which C(2)
bears a hydrogen atom. From this analysis it is evident that
a free HMDS anion, derived from either K-HMDS or a
complex thereof, is not acting as the base in the process gen-
erating 4.
From the preceding analysis, it can be concluded that it is
the carbene carbon C(2) of 2, and not an HMDS anion, that
acts to deprotonate the toluene. However, (K-HMDS)2
whilst not actively involved as the base, demonstrably catal-
yses the reaction. Three limiting mechanisms can be envis-
aged. Firstly, in the direct reaction of carbene 2 with the tol-
uene to generate 4 by means of pathway III in Scheme 3,
the involvement of a benzyl/tetrahydropyrimidinium (1) ion-
pair would cause significant charge generation from the neu-
tral precursor pairing (2 and toluene) on approach to the
transition state. In an apolar aromatic solvent, for example,
toluene or benzene, the ionogenic nature of the process (2+
stant kobs according to initial toluene concentration) with the
G
mole fraction [2
A
HMDS)2] and Ka =62(ꢁ7)mꢀ1 at [2]t =53 mm,[28] yields a
good linear correlation, Figure 6 (right). An important rami-
fication of such a correlation is that the observed rate is a
mole-fraction-weighted average of two separate reactions of
toluene: reaction with 2 (mole fraction x2) and reaction with
[2
A
+
(1ꢀx2)kcat)[2]1
[toluene]1. Nonlinear regression thus yields
ACHTREUNG
second-order-rate constants for reaction proceeding via [2-
A
ACHTREUNG
2
k
A
10ꢀ8 dm3 molꢀ1 sꢀ1. The latter rate constant is not easy to de-
termine directly because the generation of 2 from 1 in the
absence of a slight excess of (K-HMDS)2 results in dimerisa-
tion to give 3, at a rate that is upwards of one order of mag-
nitude faster than toluene insertion at the concentrations
considered herein.
toluene!
C
To address the possibility that the inherent rate (kuncat) for
direct reaction of 2 with toluene arises by catalysis from
complexation of 2 with the KPF6 (surface or solution), we
conducted the following experiment. A sample of [13C1]-2 in
[D8]toluene containing 0.135m excess (K-HMDS)2 was fil-
tered (0.45 mm PTFE membrane) into 0.175m (K-HMDS)2
in [D8]toluene, giving rise to a 0.165m solution of (K-
HMDS)2 in which the KPF6-saturated toluene has thus un-
dergone 3.5-fold dilution and been separated from particu-
late KPF6. The pseudo-first-order rate constant for the reac-
tion of 2 to give 4 was found to be kobs =4.32ꢁ0.18”10ꢀ7 sꢀ1
(Table 1, entry 15), which is identical to the value predicted
The second and third mechanisms considered involve
Brønsted salt-effect catalysis (medium effect) or discrete
molecular catalysis of the reaction of 2 with toluene by its
potassium complex [2ACHTRE(UNG K-HMDS)2]. The nonlinear regres-
sion of a simple 1:1 binding isotherm[28] to the 13C N MR
signal arising from C(2) in [13C1]-2 during conversion to
[2H8]
A
(4.3”10ꢀ7 sꢀ1) from the relationship ꢀd[2]/dt=(x2 kuncat
+
tion constant for the equilibrium of 2 and (K-HMDS)2 to
form [2(K-HMDS)2]. The goodness of fit (Figure 6, left,
(1ꢀx2)kcat)[2]1
[toluene]1, using the appropriate values of [2]0,
ACHTREUNG
T
[[D8]toluene]0 and binding constant Ka =62mꢀ1. It can thus
be concluded that trace KPF6 in the toluene does not cata-
lyse reaction to any significant extent and, therefore, it ap-
pears that there is a direct reaction between the carbene 2
and the toluene to generate 4.
solid line) together with low sensitivity to the concentration
of [13C1]-2 on the 13C NMR shift of C(2),[25,27] is consistent
with a low association constant for the equilibrium of [2ACHTREUNG
HMDS)2] with a further molecule of 2 to form [(2)2ACHTREUNG
HMDS)2].[29]
The association constant for the 1:1 equilibrium, Ka =
62(ꢁ7)mꢀ1, as determined by 13C{1H} NMR spectroscopy in-
dicates that the proportion of 2 that is complexed under the
reaction conditions at the [(K-HMDS)2] concentrations con-
sidered herein varies from about 8 to 94%. As a conse-
quence, a mechanism involving Brønsted salt-effect catalysis
Conclusions
A study of the generation of the 13C-labelled N,N-heterocy-
clic carbene [13C1]-2 by reaction of a suspension of tetrahy-
dropyrimidinium precursor [13C1]-1-PF6 with excess (K-
HMDS)2 in toluene reveals that although carbene 2 is kinet-
ically stable towards dimerisation, it undergoes formal inser-
tion into the toluene methyl group to generate an aminal
(4). The reaction is slow but quantitative and is mildly cata-
lysed by the excess (K-HMDS)2. The rate of reaction is
strongly dependent on the toluene para substituent, yielding
a Hammett 1ꢀ value of 4.8ꢁ0.3 and, in a separate analysis,
a resonance-dominated Swain–Lupton term (sSL =0.73F+
R). Deuterium labelling at the toluene methyl group ([D8]-,
[D1]- and [D0]toluene) results in moderate primary and
large secondary KIEs for [D1]toluene in C6D6 (primary (kH/
by the complex [2ACHTREUNG(K-HMDS)2] on the direct reaction of 2
with toluene can be ruled out on the basis that although the
mole fraction of free 2 (x2) relative to that which is com-
plexed with (K-HMDS)2 (1ꢀx2) is essentially constant
through the evolution of the reaction,[28] the concentrations
of both species would decrease linearly with conversion and
thus the relationship: logkobs /[2]1[2
(K-HMDS)2]0.5 would
ACHTREUNG
not give rise to the simple pseudo-first-order kinetics ob-
served under all conditions.
Considering then a mechanism involving discrete reaction
of the potassium-complexed carbene [2ACTHER(UNG K-HMDS)2] with
the toluene, analysis of the second-order-rate constant (ob-
tained by normalisation of the pseudo-first-order rate con-
kD)=4.2
N
G
for [D0]toluene/[D8]toluene as reactant and solvent (kH/kD =
5368
www.chemeurj.org
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5361 – 5375

Insertion Reactions
FULL PAPER
10.9ꢁ1.9 on rate and k[D ]tol/k[D ]tol =9.5ꢁ0.8 for partition-
whilst the formal deprotonation of toluene by 2 is thermody-
namically very unfavourable, the reaction between 2 and
toluene may be closer to a concerted process (pathway II,
Scheme 3) than a fully ionogenic process (pathway III) to
facilitate a sufficiently low activation barrier (116 kJmolꢀ1)
for reaction to proceed at a reasonable rate. The magnitude
of the primary and secondary deuterium isotope effects[18,31]
are consistent with this in that the transition state for proton
transfer would be “late”. Indeed the primary KIE (kH/kD =
0
8
ing).[30] These are consistent with rate-limiting C H cleavage
(kI1I or kI1II in Scheme 3) proceeding via a late transition
state,[18] with an accompanying decrease in p character at
the benzylic carbon and little competing internal return,
which would reduce the primary KIE to an equilibrium iso-
tope effect (K1III in Scheme 3).[31,32] Analysis of the chemical
shift of the 13C{1H} NMR signal arising from C(2) in [13C1]-2
indicates that there is rapid and reversible complexation of
carbene 2 to the (K-HMDS)2 with an association constant of
Ka =62(ꢁ6)mꢀ1. Using this value to determine[28] the mole
fractions of 2 that are complexed (1ꢀx2) and noncomplexed
(x2), the rate of reaction can be satisfactorily predicted by
ꢀ
4.2
et al. for the insertion of dimethoxy carbene into the O H
bond of methanol (kH/kD =3.3
(ꢁ0.5),[21e] for which a path-
A
ꢀ
way of type III (with some characteristics of II) was sugges-
ted.[21f] For the generation of 4, the lowering of both the ac-
tivation barrier and the energy of the intermediate (route
III) or transition state (route II), may arise from a stabilising
secondary orbital interaction involving the vacant p orbital
at C(2) and the ipso-carbon atom on toluene. Analogous in-
teractions of aryl p systems with carbenes have been sug-
gested on the basis of computational studies to account for
rate accelerations of carbene reactions in aromatic sol-
vents.[33] Such a process might be expected to perturb the
electronic effects of the para-tolyl substituents and thus lead
to a nonlinear Hammett correlation. However, a compensa-
tory effect could lead to a simple attenuation, as would be
consistent with the 1 value (1=4.8ꢁ0.3, Figure 3), which is
smaller than would be expected for full carbanion genera-
tion. In effect, this mechanism would have characteristics of
both pathways I and III (the latter intramolecular) in
Scheme 3.
the relationship ꢀd[2]/dt=(x2 kuncat +(1ꢀx2)kcat)[2]1
[toluene]1
ACHTREUNG
(Figure 6, right; Figure 2, right). The correlation leads to the
conclusion that there are two processes for the generation
of 4 from 2: 1) direct reaction of 2 with toluene (kuncat
,
Scheme 2) and 2) reaction of the complex [2(K-HMDS)2]
AHCTREUNG
with toluene (kcat, Scheme 2).[32] Careful study of the 13C{1H}
NMR spectra of the products from 2 and methyl-deuterated
toluenes ([D8]- and [D1]toluene) demonstrates that the
HMDS anion is not the active base in either process as
there is no evidence for cross-over exchange through bulk
phase H-HMDS/D-HMDS.
The smooth generation of 2 from 1-PF6/ACHTRE(UNG K-HMDS)2, and
the equilibrium between 2, H-HMDS and 8, suggests that
the HMDS anion is more basic than 2, but not exceptionally
so. On the basis of lack of protium–deuterium exchange at
the [D8]toluene methyl group in the presence of (K-
HMDS)2/H-HMDS, we find no evidence for reaction 2!4
proceeding via toluene deprotonation by a free HMDS
anion. Yet, 2 reacts smoothly with toluene by what appears
The increased reactivity of the (K-HMDS)2 complex [2-
AHCTREU(NG K-HMDS)2] towards toluene (ca. 3.4-fold over that of the
to be
a
rate-limiting deprotonation mechanism (kuncat
,
free carbene 2) may appear contra-intuitive: K+ complexa-
tion by the occupied sp2 orbital[34] should render C(2) less
basic than in free 2. However, as is evident from the single-
Scheme 4) and despite what appears to be substantial pKa
mismatch between 2 (pKa of [1]ꢄ26 in DMSO)[23] and tol-
uene (pKa ca. 42 in DMSO).[22] It is thus concluded that
(K-HMDS)2],[25a] in which 2’
crystal X-ray structure of [(2’)2ACHTREUNG
is the N,N’-diisopropyl analogue of 2, the carbene–potassium
interaction has a predominantly electrostatic ion–dipole
(K-HMDS)2],[25a] is long
ꢀ
character. The C(2) K bond in [(2’)
(3.00 Š) relative other N-heterocyclic-carbene–metal com-
plexes (ꢃ2.30 Š).[2a] A possible explanation for the in-
creased reactivity of [2ACHTRE(UNG K-HMDS)2] towards toluene may lie
in the localised polarisation and electrostriction of the nas-
cent tight-ion pair {1-CH2Ph}° (pathway III) or polarised
nonsynchronous concerted process (pathway II) in which
(K-HMDS)2 is liberated from 2 (DGassoc =ca. 10 kJmolꢀ1 at
238C) within the solvent shell (kcat; Scheme 4). This com-
pensatory effect of the (K-HMDS)2 must just outweigh the
reduced availability of the filled sp2 orbital at C(2), resulting
in a slightly increased reaction rate relative to 2 (DDG°,
based on kcat/kuncat =3 kJmolꢀ1 at 238C). In effect, the reac-
tion of the toluene with the potassium-complexed carbene
experiences localised Brønsted salt-effect catalysis within
the solvent shell surrounding the partially ionogenic pro-
Scheme 4. A dual-manifold mechanism ((K-HMDS)2-catalysed and un-
ACHTREUNG
AHCTREUNG
[D8]toluene at 238C,
k
cat =5.1
.
A
,
kuncat =
ꢀ
1.5
A
ACHTREeUNG rocyclic carbene 2 into the “nonacidic” C H bond of the
Chem. Eur. J. 2006, 12, 5361 – 5375
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5369

G. C. Lloyd-Jones et al.
6H; NCH
A
G
ACHTREUNG(H,H)=
toluene methyl group to give aminal 4 has been investigated
by kinetic and isotopic labelling strategies. The reaction ap-
pears to follow a mechanism that has characteristics of both
3.0 Hz, 1H; C(5)Heq), 1.82 (qt, 2J
E
ACHTREUNG
(H,H)=5.2 Hz, 1H; C(5)Hax), 2.44 (ddq, 2J
G
ACHTREUNG
R
ACHTREUNG
3
the classic “concerted” and “deprotonation” pathways pro-
(H,H)=7.0 Hz, J
A
[20]
ꢀ
posed for O H insertions (pathways II and III, Scheme 3)
(m, 1H; C(4)H(H)), 2.69–2.81 (m, 3H; C(6)H2 and C(4)H(H)), 2.83 ppm
(double septet, 3J(H,H)=6.4 Hz, 3J
(H,C)=3.1 Hz, 1H; NCH(CH3)2);
13C{1H} NMR (125 MHz, [D8]toluene, 24.58C): d=13.8 (s; NCH2CH3),
21.8 (s; NCH(CH3)CH3), 22.1 (s; NCH(CH3)CH3), 22.7 (s; C(5)H2), 41.4
(s; C(6)H2), 45.2 (s; C(4)H2), 47.6 (s; NCH2CH3), 50.7 (s; NCH(CH3)2),
G
G
ACHTREUNG
and, whilst this may be unexpected on the basis of pKa, it is
unlikely be restricted either to 2[36] or to toluenes. The possi-
ble use of simple analogues of 2 as reagents for the prepara-
tion of aldehydes through homologation/aminal hydrolysis is
of interest, particularly if more efficient catalysts can be de-
veloped, and is being actively pursued in these laboratories.
U
ACHTREUNG
ACHTREUNG
(C,D)=21.5 Hz; 13C(2)).
75.8 ppm (t, JACHTREUNG
1
Preparation of [13C1]-2 in mixed[D 0]toluene/[D8]toluene solution and
analysis of the kinetic isotope effects in generation of [13C1]-4 and[ 2H8]-
AHCTREUNG A
[13C1]-4: typical procedure was as follows: [13C1]-1-PF6 (13.9 mg,
0.0462 mmol) was transferred into an NMR tube (with Young valve)
under N2. (K-HMDS)2 (0.35m solution (HMDS anion) in [D8]toluene,
0.390 cm3, 0.07 mmol (K-HMDS)2) was added to the salt. The tube was
sealed and the tip of the tube was submerged in liquid nitrogen until the
contents were frozen. The tube was removed from the liquid nitrogen
and the solvent was carefully removed under high vacuum. Under an at-
mosphere of nitrogen, distilled [D0]toluene (0.330 cm3, 3.10 mmol) and
distilled [D8]toluene (0.220 cm3, 2.07 mmol) were then added and the
NMR tube was sealed. The tip of the NMR tube was submerged in an ul-
trasonic bath and the contents were sonicated until all of the colourless,
crystalline salt [13C1]-1-PF6 had dissolved and been replaced with a fine,
white precipitate of KPF6 (10 min). The reaction was monitored by NMR
spectroscopy. The isotopic partitioning in the products was calculated by
comparison of the intensity of the double triplet signal at d=3.74 ppm
(arising from the nondeuterated product only) with intensities of signals
arising from a combination of both nondeuterated and fully deuterated
products.
Experimental Section
General: [D6]benzene, toluene, [D8]toluene, [D1]toluene, ethylbenzene
and 4-tert-butyltoluene were distilled from sodium under a nitrogen at-
mosphere immediately before use. 4-Trifluoromethyltoluene and 4-chlo-
ACHTREUNG
ACHTREUNG
A
ACHTREUNG
A
A
ACHTREUNG
tion (125 C, 15 mmHg). N-Ethyl-N’-isopropyl-1,3-diaminopropane was
prepared by adaptation[37] of the method of Yamamoto and Maruoka.[38]
The salts 1-PF6 and [13C1]-1-PF6 were prepared from N-ethyl-N’-isopro-
ACHTREUNGpyl-1,3-diaminopropane and by adaptation of the method of Alder
Data for adduct [13C1]-4: 1H NMR (400 MHz, [D8]toluene, 22.28C): d=
et al.[39] All manipulations were conducted under an inert (N2) atmos-
phere by using Schlenk line techniques. NMR samples were prepared in
5 mm tubes, sealed by means of a Young valve. NMR spectra were ac-
quired on JEOL instruments (ECP300, ECP400 and Alpha 500) using
0.91 (t, 3J
NCH
1H; C(5)Heq), 1.82 (qt, 2J
5.2 Hz, 1H; C(5)Hax), 2.44 (ddq, 2J
(H,C)=4.0 Hz, 1H; NCH(H)CH3), 2.51 (ddq, 2J
(H,H)=7.0 Hz, 3J
(H,C)=4.0 Hz, 1H; NCH(H)CH3), 2.50–2.62 (m, 1H;
C(4)H(H)), 2.69–2.81 (m, 3H ([13C1]-4&[2H8][13C1]-4)+2H ([13C1]-4);
C(6)H2, C(4)H(H), and PhCH213C(2)H), 2.83 (double septet, 3J
(H,H)=
(H,H)=7.0 Hz, 3H; NCH2CH3), 0.92 (d, 3J
ACHRTEUNG ACHTREUNG
A
G
ACHTREUNG
A
N
ACHTREUNG
the H-signal from the solvent as a frequency lock. 1H and 13C{1H} NMR
2
A
ACHTREUNG
spectra were referenced internally to [D6]benzene (1H d=7.15 ppm; 13C
d=128.0 ppm) or the toluene methyl group (1H d=2.09 ppm (CD2H);
13C d=20.4 ppm (CD3)). Assignments are based on HH COSY, HMQC
A
ACHTREUNG
A
ACHTREUNG
A
ACHTREUNG
(1J(C,H)), HMBC (>1J
ACHTREUNG ACHTREUNG(C,H)) and DEPT. Reaction kinetics were ana-
ACHTREUNG
3
1
3
6.4 Hz, JCAHTRE(UGN H,C)=3.1 Hz, 1H; NCHAHCERT(UNG CH3)2), 3.74 (dt, JACHERT(UNG H,C)=145 Hz, J-
lysed by application of standard first-order (or pseudo-first-order) linear
relationships, rate constants, binding constants and chemical shifts and as-
sociated errors were estimated by linear or nonlinear regression using
MacCurveFit software.[40]
AHCTREUNG
(H,H)=6.5 Hz, 1H ([13C1]-4); PhCH213C(2)H), 6.92–7.19 ppm (m, 5H
([13C1]-4, obscured by toluene signals); 5”Ar-H); 13C{1H} NMR
(100 MHz,[D8]toluene, 23.28C): d=13.8 (s; NCH2CH3), 21.8 (s; NCH-
(CH3)CH3), 22.7 (s; C(5)H2), 29.4 (d, 1J
ACHRTUNEG ACHTREUNG(C,C)=
Preparation of 2-[13C]-1-ethyl-3-isopropyl-3,4,5,6-tetrahydropyrimid-2-yli-
U
dene [13C1]-2 in [D8]toluene andanalysis of the kinetics of generation of
NCH2CH3), 50.7 (s; NCHACHTERNGU ACHTREUGN
(CH3)2), 75.8 (t, 1J(C,D)=21.5 Hz; 13C(2)D),
ACHTREUNG
[2H8][13C1]-4: A typical procedure was as follows: [13C1]-1-PF6 (13.5 mg,
76.5 (s; 13C(2)H), 141.7 (s; ipso-Ar (other aromatic signals obscured by
0.0449 mmol) was transferred into an NMR tube (with Young valve)
under N2. (K-HMDS)2 (0.35m solution (HMDS anion) in [D8]toluene,
0.390 cm3, 0.07 mmol (K-HMDS)2) was added to the salt. Distilled
[D8]toluene (0.160 cm3) was added to the mixture and the NMR tube was
sealed. The tip of the NMR tube was submerged in an ultrasonic bath
and the contents were sonicated until all of the colourless, crystalline salt
[13C1]-1-PF6 had dissolved and been replaced with a fine, white precipi-
tate of KPF6 (10 min). The reaction was monitored by NMR spectrosco-
py.
toluene signals)).
Preparation of [13C1]-2 in [D6]benzene/Ar-R solution andanalysis of the
kinetics of generation of [13C1]-4, [13C1]-19, [13C1]-20, [13C1]-21 and[ 13C1]-
22:
A
typical procedure was as follows: [13C1]-1-PF6 (13.9 mg,
0.0462 mmol) was transferred into an NMR tube (with Young valve)
under N2. (K-HMDS)2 (0.26m solution (HMDS anion) in [D6]benzene,
0.350 cm3, 0.091 mmol) was added to the salt. The tip of the NMR tube
was submerged in an ultrasonic bath and the contents were sonicated for
5 min. Distilled [D6]benzene (0.060 cm3) and distilled 4-trifluoromethyl-
Data for carbene [13C1]-2: 1H NMR (400 MHz, [D8]toluene, 20.88C): d=
AHCTREUNG
toluene (0.260 cm3, 1.87 mmol) were added to the mixture (causing a
1.05 (t, 3J
NCH
(H,H)=6.0 Hz, 2H; C(4)H2), 2.39 (t, 3J
3.23 (dq, 3J(H,H)=7.1 Hz, 3J
(H,C)=5.2 Hz, 2H; NCH2CH3), 3.78 ppm
(double septet, 3J(H,H)=6.9 Hz, 3J
(H,C)=3.9 Hz, 1H; NCH(CH3)2);
13C{1H} NMR (100 MHz, [D8]toluene, 22.08C): d=14.8 (s; NCH2CH3),
21.0 (s; NCH(CH3)2), 21.7 (s; C(5)H2), 34.7 (broad unresolved doublet;
C(6)H2), 40.4 (d, 2J(C,C)=5.7 Hz; C(4)H2), 54.1 (d, 2J
(C,C)=23.0 Hz;
NCH2CH3), 58.6 (d, 2J (CH3)2), 229.6 ppm (13C(2)).
(C,C)=24.0 Hz; NCH
[13C1]-4: 1H NMR (500 MHz, [D8]toluene, 24.58C):
(H,H)=7.0 Hz, 3H; NCH2CH3), 0.92 (d, 3J
(H,H)=6.4 Hz,
ACHTREUNG ACHTRENUG
(H,H)=7.1 Hz, 3H; NCH2CH3), 1.07 (d, 3J
(H,H)=6.9 Hz, 6H;
(H,H)=6.0 Hz, 2H; C(5)H2), 2.31 (t, 3J-
(H,H)=6.0 Hz, 2H; C(6)H2),
ACHTREUNG ACHTRENUG
(CH3)2), 1.30 (quintet, 3J
colour-change from colourless to bright yellow initially, then to orange
over 10 min) and the NMR tube was sealed. The contents were sonicated
for a further 5 min until all of the colourless, crystalline salt [13C1]-1-PF6
had dissolved and been replaced with a fine, white precipitate of KPF6.
A
ACHTREUNG
A
ACHTREUNG
G
N
ACHTREUNG
Generation of the toluene adduct [13C1]-4, 4-CF3-toluene adduct [13C1]-19
1
and 4-Cl-toluene adduct [13C1]-20 was monitored by H NMR spectrosco-
ACHTREUNG
py. The 13C{H} NMR shifts for 13C(2)H in [13C1]-4, [13C1]-19 and [13C1]-20
were d=76.4, 76.1 and 76.2 ppm respectively. In the case of [13C1]-19 and
[13C1]-20 a smaller signal at d=76.5 ppm (ca. 10% of the intensity of the
signals at d=76.1 and 76.2 ppm, respectively) was also apparent. This
E
ACHTREUNG
A
ACHTREUNG
Data for adduct [2H8]
ACHTREUNG
d=0.91 (t, 3J
N
ACHTREUNG
5370
www.chemeurj.org
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5361 – 5375

Insertion Reactions
FULL PAPER
may indicate the presence of a product formed from reaction of a second
molecule of carbene [13C1]-2 with [13C1]-19 and [13C1]-20. The 4-tert-butyl-
toluene adduct [13C1]-21 and ethylbenzene adducts (two diastereomers)
[13C1]-22 were characterized by the appearance of the 13C(2)H signal d=
76.5 ppm for [13C1]-21, and d=80.1/81.9 ppm for [13C1]-22 in the 13C{1H}
spectrum, and rates of formation were monitored by measuring the dis-
appearance of the 13C(2) signal of [13C1]-2 at d=231 ppm, by using the
combined intensities of the (K-HMDS)2 (d=7.2 ppm) and H-HMDS
(d=2.6 ppm) signals as an internal standard.
C(4)H(H)), 2.68-2.81 (m, 3H (C(2)-protonated and C(2)-deuterated)+
2H (C(2)-deuterated)+1H (C(2)-protonated); C(6)H2, C(4)H(H),
PhCH213C(2)D, and PhCH (D)13C(2)H), 2.85 (double septet, 3J
ACHTREUNG
6.4 Hz, 3J
AHCTREUNG
(CH3)2), 3.76 (dd, 1J
ACHTRE(UNG H,C)=2.9 Hz, 1H; NCHACHERTNUG ACHTREUNG
3J(H,H)=6.4 Hz, 1H (C(2)-protonated); PhCH(D)13C(2)H), 6.95–
7.22 ppm (m, 5H (Ph-) obscured by toluene signals); 5”Ar-H); 13C{1H}
NMR (100 MHz, [D6]benzene, 23.28C): d=13.8 (s; NCH2CH3), 21.7 (s;
NCH
stereomer), 22.1 (s; NCH
39.0 Hz, 1J(C,D)=21 Hz; PhCH(D)13C(2)H), 29.7 (d, 1J
PhCH213C(2)D), 41.5 (s; C(6)H2), 45.2 (s; C(4)H2), 47.6 (s; NCH2CH3),
A
ACHTRE(UNG CH3)CH3 of one dia-
1
A
ACHTREUNG(C,C)=
Data for 4-CF3-adduct [13C1]-19: 1H NMR (400 MHz, [D6]benzene,
A
ACHTREUNG
3
3
22.38C): d=0.86 (t, J
6.1 Hz, 3H; NCH
(CH3)CH3), 0.88 (d, 3J
(CH3)CH3), 1.04 (double quintet, 2J(H,H)=12.7 Hz, 3J
1H; C(5)Heq), 1.70–1.90 (m (obscured by CH3 of 15), 1H; C(5)Hax), 2.38
(ddq, 2J(H,H)=11.6 Hz, 3J(H,H)=7.4 Hz, 3J
(H,C)=4.2 Hz, 1H;
NCH(H)CH3), 2.48 (ddq, 2J(H,H)=12.2 Hz, 3J(H,H)=7.1 Hz, 3J
(H,C)=
G
ACHTRE(UNG H,H)=
50.6 (s; NCHACTHRENUG ACHTREUGN
(CH3)2), 75.9 (t, 1J(C,D)=22.0 Hz; 13C(2)D), 76.4 (s;
N
ACHTREUNG
13C(2)H), 141.7 ppm (s; ipso-Ar); other aromatic signals obscured by tol-
A
G
ACHTREUNG
uene signals.
U
G
N
Calculation of primary andseconadry kinetic isotope effects
Table 1, entry 9: obs-D0 =3kH =9.46 From Table 1,
(ꢁ0.31)”10ꢀ7 sꢀ1
entry 14: kobs-D1 =(2kH/sec)+(1”kD/prim)=6.09
: From
G
A
A
k
A
.
3.9 Hz, 1H; NCH(H)CH3), 2.49–2.59 (m, 1H; C(4)H(H)), 2.61–2.74 (m,
5H; C(6)H2, C(4)H(H), and PhCH213C(2)H), 2.77 (double septet, 3J-
ACHTREUNG
ꢀ
ꢀ
tioning between C H and C D insertion=P=7.06ꢁ0.7, thus kH/sec (for
(H,H)=6.4 Hz, 3J
145 Hz, 3J
G
(CH3)2), 3.63 (dt, 1J
U
PhCHD H)=k
obs-D1 A
(ꢁ0.08).
ꢀ
(H,H)=6.6 Hz, 1H; ArCH213C(2)H), 6.59–7.48 ppm (m, 5H
2Pꢀ1)/3kobs-D1 =1.18
A
Primary
KIE=kH/kD/prim =kobs-D0ACHTREUNG(1+P)/
3kobs-D1 =4.2
(ꢁ0.6). Errors are calculated by maximum propagation of
ꢀ
ꢀ
ꢁ1% error in C H/C D and the sum-squared error from linear regres-
sion of ln([2]0/[2]t) versus t.
E
Independent preparation of 4 from phenacetaldehyde: Phenacetaldehyde
(0.020 cm3, 0.171 mmol) was added to N-ethyl-N’-isopropyl-1,3-diamino-
propane (0.0253 g, 0.176 mmol) in [D8]toluene (0.500 cm3) in an N MR
tube containing two pellets of 4 Š “molecular sieves”. The contents of
the tube were shaken to allow proper mixing. NMR analysis showed that
a mixture of products had been formed, of which adduct 4 was the major
component. 1H NMR (500 MHz, [D8]toluene, 24.58C): d=0.91 (t, 3J-
AHCTREUNG
obscured by [D6]benzene and 4-CF3-toluene signals.
Data for 4-Cl-adduct [13C1]-20: 1H NMR (400 MHz, [D6]benzene,
3
3
22.18C): d=0.87 (t, J
6.4 Hz, 6H; NCH
(CH3)2), 1.03 (double quintet, 2J
(H,H)=2.7 Hz, 1H; C(5)Heq), 1.75–2.00 (m (obscured by CH3 of 17),
1H; C(5)Hax), 2.39 (ddq, 2J(H,H)=11.5 Hz, 3J(H,H)=7.6 Hz, 3J
(H,C)=
3.9 Hz, 1H; NCH(H)CH3), 2.48 (ddq, 2J(H,H)=11.5 Hz, 3J
(H,H)=
7.1 Hz, 3J
(H,C)=3.7 Hz, 1H; NCH(H)CH3), 2.45–2.60 (m, 1H;
C(4)H(H)), 2.59–2.73 (m, 5H; C(6)H2, C(4)H(H), and ArCH213C(2)H),
2.78 (double septet, 3J(H,H)=6.4 Hz, 3J
(H,C)=2.4 Hz, 1H; NCH-
(CH3)2), 3.61 (dt, (H,C)=145 Hz, (H,H)=6.6 Hz, 1H;
1J 3J
G
ACHTRE(UNG H,H)=
E
ACHTREUNG
ACHTREUNG
G
G
A
A
ACHTREUNG
N
C
A
N
ACHTREUNG
N
A
N
ACHTREUNG
A
ACHTREUNG
E
A
A
ACHTREUNG
A
A
ACHTREUNG
ArCH213C(2)H), 6.47–7.23 ppm (m, 5H (obscured by signals aryl carbons
ACHTREUNG
of 17); 5”Ar-H); 13C{1H} NMR (100 MHz, [D6]benzene, 23.28C): d=
A
ACHTREUNG
13.9 (s; NCH2CH3), 21.8 (s; NCH
ACHRTUG(NE CH3)CH3), 22.0 (s; NCHAHCRTE(UGN CH3)CH3),
22.5 (s; C(5)H2), 29.7 (d, 1J(C,C)=37.0 Hz; ArCH213C(2)H), 41.3 (s;
A
A
ACHTREUNG
C(6)H2), 45.1 (s; C(4)H2), 47.6 (s; NCH2CH3), 50.6 (s; NCHACHTRE(UNG CH3)2),
76.2 ppm (s; 13C(2)H); aromatic signals obscured by [D6]benzene and 4-
Cl-toluene aromatic signals.
ACHTREUNG
Ar); other aromatic signals obscured by [D8]toluene signals.
Preparation of [13C1]-2 in [D6]benzene/a-[D1]toluene andanalysis of the
kinetic isotope effects: [13C1]-1-PF6 (13.7 mg, 0.0455 mmol) was transfer-
red into an NMR tube (with Young valve) under N2. (K-HMDS)2 (0.26m
solution (HMDS anion) in [D6]benzene, 0.350 cm3, 0.091 mmol) was
added to the salt. Distilled [D6]benzene (0.120 cm3) and distilled a-
[D1]toluene (0.200 cm3, 1.97 mmol) were added to the mixture and the
NMR tube was sealed. The tip of the NMR tube was immersed in an ul-
trasonic bath and the contents were sonicated until all of the colourless,
crystalline salt [13C1]-1-PF6 had dissolved and been replaced with a fine,
white precipitate of KPF6 (10 min). The reaction was monitored by NMR
spectroscopy. The isotopic partitioning in the products was calculated by
comparison of the intensity of the dd signal at d=3.77 ppm (arising from
the C(2)-protonated product only) with intensities of signals arising from
a combination of both C(2)-protonated and C(2)-deuterated products.
The major regioisomer produced in this reaction was the C(2)-protonated
adduct. This consists of two diastereomers, which are just distinguishable
Preparation of a-[D1]toluene: Benzylmagnesium chloride (1.0m solution
in Et2O, 100 cm3, 100 mmol) was added by cannula to a cooled (ice bath)
stirred mixture of D2O (6 cm3, ca. 300 mmol) and Et2O (20 cm3). The re-
action mixture was allowed to warm slowly to room temperature over
1.5 h then filtered and the diethyl ether removed from the filtrate at
508C–708C, 760 mmHg on a rotary evaporator. The residue was distilled
(15 cm Vigreux column) at 760 mmHg, then redistilled to yield the title
compound (6.002 g. 64.5 mmol, 64.5%) as a colourless liquid. 1H N MR
analysis indicated that the isotopic purity was >98.8% and that ca. 1%
Et2O remained. 1H NMR (400 MHz, a-[D1]toluene, 22.58C): d=2.10 (t,
2J
1H; para-Ar-H), 7.09 (m, 2H; 2 ” meta-Ar-H); 13C{1H} NMR (100 MHz,
a-[D1]toluene, 23.58C): d=21.1 (t, 1J
(C,D)=19.2 Hz; Ph-CH2D), 125.6
ACHTRE(UNG H,D)=2.6 Hz, 2H; PhCH2D), 6.97 (m, 2H; 2 ” ortho-Ar-H), 7.01 (m,
AHCTREUNG
(s; para-Ar), 128.4 (s; meta-Ar), 129.3 (s; ortho-Ar), 137.5 ppm (s; ipso-
Ar).
Equilibrium of H-HMDS/carbene [13C1]-2 with [13C1]-14: [13C1]-1-PF6
(13.5 mg, 0.0449 mmol) was transferred into an NMR tube (with Young
valve) under N2. (K-HMDS)2 (0.175m solution in [D8]toluene, 0.300 cm3,
0.11 mmol) was added to the salt and the NMR tube was sealed. The tip
of the NMR tube was submerged in an ultrasonic bath and the contents
were sonicated until all of the colourless, crystalline salt [13C1]-1-PF6 had
dissolved and been replaced with a fine, white precipitate of KPF6
(10 min). Distilled H-HMDS (0.150 cm3, 1.19 mmol) was added to the
mixture. The mixture was analysed by 13C{1H} NMR spectroscopy which
by 1H NMR spectroscopy (maximum Dd of 0.002 ppm at 400 MHz).
1
Data for adduct [2H1]
d=0.92 (t, 3J
3H; NCH
(CH3)CH3), 0.94 (d, 3J
1.03 (double quintet, 2J(H,H)=12.5 Hz, 3J
1.82 (qt, 2J(H,H)=12.4 Hz, 3J(H,H)=12.4 Hz, 3J
C(5)Hax), 2.45 (ddq, 2J(H,H)=12.0 Hz, 3J(H,H)=7.1 Hz, 3J
4.0 Hz, 1H; NCH(H)CH3), 2.52 (ddq, 2J(H,H)=12.0 Hz, 3J
(H,H)=
7.1 Hz, 3J
(H,C)=4.0 Hz, 1H; NCH(H)CH3), 2.52–2.61 (m, 1H;
A
A
ACHTREUNG
A
N
ACHTREUNG
A
ACHTREUNG
A
N
ACHTREUNG
A
N
ACHTREUNG
A
ACHTREUNG
A
indicated a ratio of 5.0:1.0 [13C1]-2 (d=233 ppm)/[13C1]-14 (d=91.2 ppm).
ACHTREUNG
Chem. Eur. J. 2006, 12, 5361 – 5375
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5371

G. C. Lloyd-Jones et al.
A 0.150 cm3 portion of the solution was removed by syringe and added to
distilled [D8]toluene (0.450 cm3) in a separate NMR tube (with Young
valve) under N2. The mixture was analysed again by 13C{1H} NMR spec-
Harvey, F. Paolini, J. Schütz, Angew. Chem. 2004, 116, 6020–6036;
Angew. Chem. Int. Ed. 2004, 43, 5896–5911.
[2] a) W. A. Herrmann, C. Kçcher, Angew. Chem. 1997, 109, 2256–
2282; Angew. Chem. Int. Ed. Engl. 1997, 36, 2162–2187; b) D. Bour-
issou, O. Guerret, F. P. Gabbaï, G. Bertrand, Chem. Rev. 2000, 100,
39–92; c) J. K. W. Chui, T. Ramnial, J. A. C. Clyburne, Comments
Inorg. Chem. 2003, 24, 165–187; d) W. Kirmse, Angew. Chem. 2004,
116, 1799–1801; Angew. Chem. Int. Ed. 2004, 43, 1767–1769; e) V.
Nair, S. Bindu, V. Sreekumar, Angew. Chem. 2004, 116, 5240–5245;
Angew. Chem. Int. Ed. 2004, 43, 5130–5135; f) Y. Canac, M. Soleil-
havoup, S. Conejero, G. Bertrand, J. Organomet. Chem. 2004, 689,
3857–3865; g) Y. Cheng, O. Meth-Cohn, Chem. Rev. 2004, 104,
2507–2530; h) V. Nair, R. S. Menon, V. Sreekumar, Pure Appl.
Chem. 2005, 77, 1191–1198; i) W. Kirmse, Eur. J. Org. Chem. 2005,
237–260; j) W. Kirmse, Angew. Chem. 2005, 117, 2530–2533;
Angew. Chem. Int. Ed. 2005, 44, 2476–2479; k) M. Dawid, D. L.
Reid, J. Warkentin, G. Mloston, J. Phys. Org. Chem. 2005, 18, 86–
89; l) N. I. Korotkikh, O. P. Shvaika, G. F. Rayenko, A. V. Kiselyov,
A. V. Knishevitsky, A. H. Cowley, J. N. Jones, C. L. B. MacDonald,
ARKIVOC 2005, 8, 10–46; m) V. Nair, V. Sreekumar, S. Bindu, E.
Suresh, Org. Lett. 2005, 7, 2297–2300; n) N. Kuhn, A. Al-Sheikh,
Coord. Chem. Rev. 2005, 249, 829–857; o) X. Catton, K. Miqueu,
H. Gornitzka, D. Bourissou, G. Bertrand, J. Am. Chem. Soc. 2005,
127, 3292–3293; p) Y. Canac, S. Conejero, B. Donnadieu, W. W.
Schoeller, G. Bertrand, J. Am. Chem. Soc. 2005, 127, 7312–7313;
q) D. Martin, O. Illa, A. Baceiredo, G. Bertrand, R. M. OrtuÇo, V.
Branchadell, J. Org. Chem. 2005, 70, 5671–5677; r) C. Präsang, B.
Donnadieu, G. Bertrand, J. Am. Chem. Soc. 2005, 127, 10182–
10183; s) F. E. Hahn, M. Paas, D. Le Van, R. Frçlich, Chem. Eur. J.
2005, 11, 5080–5085; t) Y. Canac, S. Conejero, M. Soleilhavoup, B.
Donnadieu, G. Bertrand, J. Am. Chem. Soc. 2006, 128, 459–464.
[3] a) W. A. Herrmann, Angew. Chem. 2002, 114, 1342–1363; Angew.
Chem. Int. Ed. 2002, 41, 1290–1309; b) A. F. Littke, G. C. Fu,
Angew. Chem. 2002, 114, 4350–4386; Angew. Chem. Int. Ed. 2002,
41, 4176–4211; c) B. Cornils, W. A. Herrmann, J. Catal. 2003, 216,
23–31; d) W. A. Herrmann, K. Öfele, D. v. Preysing, S. K. Schneider,
J. Organomet. Chem. 2003, 687, 229–248; e) K. Fagnou, M. Lautens,
Chem. Rev. 2003, 103, 169–196; f) R. H. Crabtree, Pure Appl.
Chem. 2003, 75, 435–443; g) K. J. Cavell, D. S. McGuiness, Coord.
Chem. Rev. 2004, 248, 671–681; h) E. Peris, R. H. Crabtree, Coord.
Chem. Rev. 2004, 248, 2239–2246; i) C. M. Crudden, D. P. Allen,
Coord. Chem. Rev. 2004, 248, 2247–2273; j) R. B. Bedford, C. S. J.
Cazin, D. Holder, Coord. Chem. Rev. 2004, 248, 2283–2321; k) P. Es-
pinet, A. M. Echavarren, Angew. Chem. 2004, 116, 4808–4839;
Angew. Chem. Int. Ed. 2004, 43, 4704–4734; l) N. M. Scott, S. P.
Nolan, Eur. J. Inorg. Chem. 2005, 1815–1828; m) N. E. Leadbeater,
Chem. Commun. 2005, 2881–2902; n) X. Cui, K. Burgess, Chem.
Rev. 2005, 105, 3272–3296; o) M. Basato, R. A. Michelin, M.
Mozzon, P. Sgarbossa, A. Tassan, J. Organomet. Chem. 2005, 690,
5414–5420; p) R. H. Crabtree, J. Organomet. Chem. 2005, 690,
5451–5457; q) D. Astruc, New J. Chem. 2005, 29, 42–56; r) I. Dragu-
tan, V. Dragutan, L. Delaude, A. Demonceau, ARKIVOC 2005, 10,
206–253; s) U. Christmann, R. Vilar, Angew. Chem. 2005, 117, 370–
378; Angew. Chem. Int. Ed. 2005, 44, 366–374; t) R. Drozdzak, B.
Allaert, N. Ledoux, I. Dragutan, V. Dragutan, F. Verpoort, Coord.
Chem. Rev. 2005, 249, 3055–3074; u) J. Garrison, W. J. Youngs,
Chem. Rev. 2005, 105, 3978–4008.
troscopy which indicated a ratio of 7.2:1.0 [13C1]-2/
ACHTREUNG
Lack of isotope exchange in
a
[D0]toluene and[D 8]toluene: (K-HMDS)2 (0.175m solution in
[D8]toluene, 0.300 cm3, 0.11 mmol) was added to distilled H-HMDS
(0.032 cm3, 0.15 mmol) in distilled [D0]toluene (0.300 cm3) in an N MR
tube (with Young valve) under N2. The tube was sealed and the contents
were shaken to ensure proper mixing. 13C{1H} NMR spectroscopy, imme-
diately after reaction, and two weeks later, showed no change in the ab-
solute or relative intensities of the Ar-CH3 singlet and Ar-CD3 septet,
and no new signals appeared in this region of the spectrum.
Determination of the association constant Ka
between (K-HMDS)2/2 and[2
(K-HMDS)2] by 13C{1H} NMR spectrosco-
(mꢀ1) for the equilibrium
ACHTREUNG
AHCTREUNG
Table 3. 13C{1H}-Chemical shift of C(2) in [13C1]-2 at various concentra-
tions of (K-HMDS)2 in [D8]toluene at 238C.
A
dC(2)
2
No. of data
points
SD[a]
[m]
average [ppm]
1
2
3
4
5
6
0.0056
0.0515
0.0833
0.134
0.1555
0.3078
234.23
231.45
229.60
228.37
228.01
227.41
7
6
41
20
1
0.117
0.188
0.367
0.075
–
11
0.038
[a] Standard deviation.
py (Figure 6, left): The entire set of 13C{1H} chemical shift data (Table 3)
for C(2) in [13C1]-2 obtained from spectra acquired during determination
of the kinetics of its reaction with [D8]toluene (Table 1, entries 1–6) were
plotted (y axis) against the [(K-HMDS)2] concentration (x axis). A non-
linear regression based on 1:1 association of 2 with the dimer and derived
from the quadratic equation for a standard A+B=C binding isotherm:
[(Ka[A]2 +((Ka[B]0ꢀKa[A]0 +1)[A])ꢀ[A]0 =0] was applied: y=[{(([{((a”
(xꢀb))+1)2 +(4”a”b)}0.5 +((a”(bꢀx))ꢀ1)]/
C
in
.
This binding isotherm requires “b”, the concentration of 2, to be defined.
As this value varies through reaction, albeit with only a low impact on
the 13C{1H} chemical shift data,[26] we included this as a fourth coefficient
in the regression. The value obtained on regression of the data (53.4 mm)
corresponds well with the mid-point for most runs (80!30 mm; 1.5 half-
lives). The data was solved for the four unknowns: Ka, [2], dcarbene and
dcomplex
,
which were found to be 62(ꢁ7) mꢀ1
,
53.2
(ꢁ0.5) mm, 237.0-
A
N
(237.0
A
dependently as 236.92 ppm from experiments in which substoichiometric
quantities of (K-HMDS)2 were added to 1 (dimerisation of the resultant
2 proceeded rapidly and thus the (K-HMDS)2 concentration must be neg-
ligible or zero).
Acknowledgements
[4] a) D. Enders, T. Balensiefer, Acc. Chem. Res. 2004, 37, 534–541;
b) R. Singh, S. P. Nolan, Chem. Commun. 2005, 5456–5458; c) P. I.
Dalko, L. Moisan, Angew. Chem. 2004, 116, 5248–5286; Angew.
Chem. Int. Ed. 2004, 43, 5138–5175; d) M. Christmann, Angew.
Chem. 2005, 117, 2688–2690; Angew. Chem. Int. Ed. 2005, 44, 2632–
2634; e) O. Coulembier, A. P. Dove, R. C. Pratt, A. C. Sentman,
D. A. Culkin, L. Mespouille, P. Dubois, R. M. Waymouth, J. L. He-
drick, Angew. Chem. 2005, 117, 5044–5048; Angew. Chem. Int. Ed.
2005, 44, 4964–4968; f) J. M. D. Storey, C. Williamson, Tetrahedron
Lett. 2005, 46, 7337–7339; g) J. J. Song, Z. Tan, J. T. Reeves, F.
Gallou, N. K. Yee, C. H. Senanayake, Org. Lett. 2005, 7, 2193–2196;
h) W. Ye, G. Cai, Z. Zhuang, X. Jia, H. Zhai, Org. Lett. 2005, 7,
Dr. A. C. OꢀDonoghue, University of Durham, UK, kindly provided pre-
publication aqueous pKa data for 1,3-diisopropyl-3,4,5,6-tetrahydropyri-
midin-1-ium and 1,3-diethyl-3,4,5,6-tetrahydropyrimidin-1-ium chlorides.
We thank Butt Park Chemicals and the EPSRC (through the South-West
Regional Development Agency, Bristol) for a postgraduate studentship
(G.J.J.O.-S.).
[1] a) A. J. Arduengo III, R. L. Harlow, M. Kline, J. Am. Chem. Soc.
1991, 113, 361–363; b) R. W. Alder, M. E. Blake, L. Chaker, J. N.
5372
www.chemeurj.org
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5361 – 5375

3769–3771; i) S. Csihony, D. A. Culkin, A. C. Sentman, A. P. Dove,
R. M. Waymouth, J. L. Hedrick, J. Am. Chem. Soc. 2005, 127, 9079–
9084; j) M. S. Kerr, J. Read deAlaniz, T. Rovis, J. Org. Chem. 2005,
70, 5725–5728.
[9] K-NACHTERUG(N SiMe3)2]2 (“ACHTRE(UNG K-HMDS)2”) is dimeric in the solid state: a) K. F.
Insertion Reactions
FULL PAPER
Tesh, T. P. Hanusa, J. C. Huffman, Inorg. Chem. 1990, 29, 1584–
1586; b) P. G. Williard, Acta. Crystallogr. Sect. C 1988, 44, 270–272;
c) in benzene solution: d) H. Buerger, H. Seyffert, Angew. Chem.
1964, 76, 646; Angew. Chem. Int. Ed. Engl. 1964, 3, 577; and in
THF: e) M. A. Nichols, D. Waldmueller, P. G. Williard, J. Am.
Chem. Soc. 1994, 116, 1153–1154.
[5] For leading references on the mechanisms of dimerisation of het-
ACHTREUNGeroatom stabilised carbenes see: a) H. W. Wanzlick, Angew. Chem.
1962, 74, 129–134; Angew. Chem. Int. Ed. Engl. 1962, 1, 75–80;
b) D. M. Lemal, R. A. Lovald, K. I. Kawano, J. Am. Chem. Soc.
1964, 86, 2518–2519; c) H. E. Winberg, J. E. Carnahan, D. D. Coff-
man, M. Brown, J. Am. Chem. Soc. 1965, 87, 2055–2056; d) M. F.
Lappert, J. Organomet. Chem. 1988, 358, 185–214; e) Y.-T. Chen, F.
Jordan, J. Org. Chem. 1991, 56, 5029–5038; f) E. C¸ etinkaya, P. B.
Hitchcock, H. A. Jasim, M. F. Lappert, K. Spyropoulos, J. Chem.
Soc. Perkin Trans. 1 1992, 561–568; g) J. A. Chamizo, P. B. Hitch-
cock, H. A. Jasim, M. F. Lappert, J. Organomet. Chem. 1993, 451,
89–96; h) B. C¸ etinkaya, P. B. Hitchcock, M. F. Lappert, D. B. Shaw,
K. Spyropoulos, N. J. W. Warhurst, J. Organomet. Chem. 1993, 459,
311–317; i) E. C¸ etinkaya, P. B. Hitchcock, H. KüÅükbay, M. F. Lap-
pert, S. Al-Juaid, J. Organomet. Chem. 1994, 481, 89–95; j) G. V.
Tormos, M. G. Bakker, P. Wang, M. V. Lakshmikantham, M. P.
Cava, R. M. Metzger, J. Am. Chem. Soc. 1995, 117, 8528–8535;
k) T. A. Taton, P. Chen, Angew. Chem. 1996, 108, 1098–1100;
Angew. Chem. Int. Ed. Engl. 1996, 35, 1011–1013; l) Z. Shi, V.
Goulle, R. P. Thummel, Tetrahedron Lett. 1996, 37, 2357–2360;
m) M. K. Denk, A. Thadani, K. Hatano, A. J. Lough, Angew. Chem.
1997, 109, 2719–2721; Angew. Chem. Int. Ed. Engl. 1997, 36, 2607–
2609; n) R. W. Alder, M. E. Blake, Chem. Commun. 1997, 1513–
1514; o) B. C¸ etinkaya, E. C¸ etinkaya, J. A. Chamizo, P. B. Hitchcock,
H. A. Jasim, H. KüÅükbay, M. F. Lappert, J. Chem. Soc. Perkin
Trans. 1 1998, 2047–2054; p) M. K. Denk, K. Hatano, M. Ma, Tetra-
hedron Lett. 1999, 40, 2057–2060; q) R. W. Alder, M. E. Blake, J.
Phys. Chem. A 1999, 103, 11200–11211; r) Y. Liu, P. E. Lindner,
D. M. Lemal, J. Am. Chem. Soc. 1999, 121, 10626–10627; s) F. E.
Hahn, L. Wittenbecher, R. Boese, D. Bläser, Chem. Eur. J. 1999, 5,
1931–1935; t) F. E. Hahn, L. Wittenbecher, D. Le Van, R. Frçlich,
Angew. Chem. 2000, 112, 551–554; Angew. Chem. Int. Ed. 2000, 39,
541–544; u) R. W. Alder, L. Chaker, F. P. V. Paolini, Chem.
Commun. 2004, 2172–2173; v) Y. Gok, E. C¸ etinkaya, I. Ozdemir, B.
C¸ etinkaya, M. F. Lappert, Acta Chim. Slov. 2004, 51, 437–446;
w) F. E. Hahn, M. Paas, D. Le Van, R. Frçlich, Chem. Eur. J. 2005,
11, 5080–5085; x) D. C. Graham, K. J. Cavell, B. F. Yates, J. Phys.
Org. Chem. 2005, 18, 298–309; see also reference [1b].
[10] The separation of the supernatent toluene solution of 2/H-HMDS
from the precipitated KPF6 and any residual 1, by filtration (0.45 mm
PTFE filter), was found to have no effect on the chemistry descri-
bed herein, other than to effect trace <5% hydrolysis of 2 to gener-
ate 5, 6H and 7H.
[11] R. W. Alder, unpublished, but discussed in reference [5u].
[12] U. Wannagat, H. Niederprüim, Chem. Ber. 1961, 94, 1540–1547.
[13] Reaction of phenacetaldehyde with N-ethyl-N’-isopropyl-propane-
1,3-diamine gave 4 along with a mixture of other products, including
enamides and imines. Moreover, hydrolysis of a sample of [13C1]-4,
generated in greater quantity by reaction of [13C1]-1/K-HMDS/tol-
uene gave [13C1]-9 (dt at d=9.22 (1H) and 197.2 ppm (13C),
1J
ACHTRE(UNG C,H)=175 Hz). All attempts to prepare 4 by reaction of 1 with
PhCH2MgX, analogous to the preparation of aminals, as reported
by Perrin (C. L. Perrin, D. B. Young, J. Am. Chem. Soc. 2001, 123,
4451–4458) were unsuccessful.
[14] The slight excess (in terms of HMDS anion) of (K-HMDS)2 over 2,
results in a slightly shallower gradient for ꢀd
ACHTRE[UNG K-HMDS]/c than for
d[2]/c (c=conversion in the reaction 1-PF6 +K-HMDS=2+KPF6).
[15] K-HMDS is predominantly or exclusively dimeric in toluene see ref-
erence [9] for leading references; however, the stoichiometry of the
reaction with [13C1]-1 is 1:1 based on HMDS anion.
[16] C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165–195, and
references therein.
[17] The large excess of the toluene over carbene 2 means that the [D0]/
[D8]toluene ratio remains essentially constant throughout reaction.
Kinetic simulation, which takes the small changes in ratio into ac-
count, was also employed and gives the same values. The error in
the isotope effect determination caused by the change in [D0]/
[D8]toluene ratio is far outweighed by the error in determination of
the partitioning ratios ([D0]/[D8] products). The molarities of [D8]-
and [D0]toluene in the neat liquid state (9.43 and 9.40 and moldmꢀ3
,
respectively) are very similar—[D8]toluene is ca. 0.3% more con-
centrated.
[18] Based on the results from reaction of 2 with [D0]- and [D1]toluene
(Table 1, entries 9 and 14) the net KIE arises from a medium pri-
mary KIE (kH/kD ca. 4.2), which is augmented by large secondary
[6] The dimerisation of carbene 2 is catalysed by 1+ to give 3, and
occurs readily when reaction is conducted in THF - even with re-
verse addition of a solution of 1-PF6 to the base: R. W. Alder, G. C.
Lloyd-Jones, G. J. J. Owen-Smith, unpublished results.
ꢀ
ꢀ
KIEs (5”CAr D: 1.62 (average 1.1 per CAr D) and 2”a-D (=
1.182): net secondary kH/kD: 2.26), giving a net KIE of 2.26”4.2=
9.5. For leading references see: a) N. Isaacs in Physical Organic
Chemistry, 2nd. ed., Longman, Harlow, England, 1995, pp. 296–302;
b) E. V. Ansyln, D. A. Dougherty in Modern Physical Organic
Chemistry, University Science Books, Sausilito, California, USA,
2006, pp. 428–430; c) T. H. Lowry, K. Schueller Richardson in
Mechanism and Theory in Organic Chemistry, 3rd ed., Harper Col-
lins, New York, 1987, pp. 232–240.
[7] For publications reporting the discrete reaction of stable carbenes
ꢀ
with C H units, giving the product shown in Scheme 1, see: a) A. J.
Arduengo III, J. C. Calbrese, F. Davidson, H. V. R. Dias, J. R. Goer-
lich, R. Krafczyk, W. J. Marshall, M. Tamm, R. Schmutzler, Helv.
Chim. Acta. 1999, 82, 2348–2364; b) S. Sole, H. Gornitzka, W. W.
Schoeller, D. Bourissou, G. Bertrand, Science 2001, 292, 1901–1903;
c) N. I. Korotikh, G. F. Rayenko, O. P. Shvaika, T. M. Pekhtereva,
A. H. Cowley, J. N. Jones, C. L. B. Macdonald, J. Org. Chem. 1993,
58, 5762–5765; d) G. W. Nyce, S. Csihony, R. M. Waymouth, J. L.
Hedrick, Chem. Eur. J. 2004, 10, 4073–4079.
[19] A dual toluene, six-membered-ring transition state mechanism
ꢀ
(analogous to that proposed for RO H insertion by carbenes, see:
J. R. Pliego, Jr., W. B. DeAlmeida, J. Phys. Chem. A 1999, 103,
3904–3909) is eliminated on this basis—which would anyway be
highly disfavoured on entropic grounds. The pseudo-second-order
rate constant for reaction of 2 with [D0]toluene (9.13m) in the pres-
[8] a) A. J. Arduengo III, H. V. R. Dias, R. L. Harlow, M. Kline, J. Am.
Chem. Soc. 1992, 114, 5530–5534; b) A. J. Arduengo III, J. R. Goer-
lich, W. J. Marshall, J. Am. Chem. Soc. 1995, 117, 11027–11028;
c) W. A. Herrmann, C. Kçcher, L. J. Goossen, G. R. J. Artus, Chem.
Eur. J. 1996, 2, 1627–1636; d) R. W. Alder, P. R. Allen, M. Murray,
A. G. Orpen, Angew. Chem. 1996, 108, 1211–1213; Angew. Chem.
Int. Ed. Engl. 1996, 35, 1121–1123; e) A. J. Arduengo III, J. R.
Goerlich, W. J. Marshall, Liebigs Ann. 1997, 365–374; f) R. W.
Alder, M. E. Blake, Chem. Commun. 1997, 1513–1514; g) R. W.
Alder, M. E. Blake, C. Bortolotti, S. Buffali, C. P. Butts, E. Lineham,
J. M. Oliva, A. G. Orpen, M. J. Quayle, Chem. Commun. 1999, 241–
242.
ence of 82 mm (K-HMDS)2 is predicted to be kobs =4.4
10ꢀ7 dm3 molꢀ1 sꢀ1
by extrapolation of
obs =xH(k(H)ꢀk(D))+k(D)
(Figure 4) to xH =1. This is in reasonable agreement with the value
of obs =3.3
(ꢁ0.1)”10ꢀ7 dm3 molꢀ1 sꢀ1 for reaction of [D0]toluene
(2.81m) in [D6]benzene in the presence of 68 mm (K-HMDS)2.
A
,
k
k
ACHTREUNG
13
[20] The generation of approximately 3.8% of [2H7]
ACHTRE[UNG C1]-4, as the C(2)
ꢀ
H isotopomer, is consistent with the approximate 0.9% [D7]toluene
(CHD2 isotopomer) present in [D1]toluene (1H NMR analysis),
when one takes into account the predicted primary kH/kD =4.2
A
Chem. Eur. J. 2006, 12, 5361 – 5375
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5373

G. C. Lloyd-Jones et al.
ꢀ
ꢀ
of [D7]- and [D8]toluene by C H and C D insertion respectively
(4ab)}0.5 +((a
(bꢀx))ꢀ1)]/
(2a))/b)
(cꢀd)}+d]; in which y=dobs, x=
ꢀ
ꢀ
ꢀ
being equivalent (C6D5CD2 D vs. C6D5CD2 H). The C D insertion
product from [D7]toluene, whilst undoubtedly present at about a
0.2% level was not detected by 13C{1H} NMR spectroscopy due to
the low nOe and slow relaxation at the C(2) carbon atom.
[(K-HMDS)2], a=Ka, b=[2], c=dcarbene and d=dcomplex (Figure 6,
left). Analogously, the equation can be applied to analyse rates in a
non-linear manner (Figure 2, right) in which y=kobs
HMDS)2], a=Ka, b=[2], c=kuncat and d=kcat The equation
(1ꢀx2)=[1ꢀ{(([{((a (bꢀx))ꢀ1)]/(2a))/
(xꢀb))+1)2 +(4ab)}0.5 +((a
b)}], in which a=Ka, b=[2], c=[(K-HMDS)2] can be used to calcu-
late the mol-fraction of [2(K-HMDS)2] and thus analyse rates in a
linear correlation (Figure 6, right). These equations are used in pref-
erence to simplified equations based on the assumption that
[(KHMDS)2]=[(KHMDS)2]0 which yields a poorer correlation, Ka =
29 (ꢁ22)mꢀ1. For the quadratic equation for a standard 1:1 binding
isotherm A+B=C binding isotherm: [(Ka[A]2 +((Ka[B]0ꢀKa[A]0 +
1)[A])ꢀ[A]0 =0], the simplified binding isotherm and the associated
simplified NMR-shift relationship, see: See E. V. Ansyln, D. A.
Dougherty in Modern Physical Organic Chemistry, University Sci-
ence Books, Sausilito, California, USA, 2006, pp. 216–221.
, x=[(K-
.
[21] See for example: a) D. Bethell, A. R. Newall, G. Stevens, D. Whit-
taker, J. Chem. Soc. B 1969, 749–754; b) W. Kirmse, K. Loosen, H.-
D. Sluma, J. Am. Chem. Soc. 1981, 103, 5935–5937; c) D. Bethell,
R. D. Howard, J. Chem. Soc. B 1969, 745–748; d) P. M. Warner, I.-S.
Chu, J. Am. Chem. Soc. 1984, 106, 5366–5367; e) X.-M. Du, H. Fan,
J. L. Goodman, M. A. Kesselmayer, K. Krogh-Jespersen, J. A. La-
G
A
ACHTREUNG
AHCTREUNG
ACHTREUNGVilla, R. A. Moss, S. Shen, R. S. Sheridan, J. Am. Chem. Soc. 1990,
112, 1920–1926; f) R. A. Moss, S. Shen, M. Wlostowski, Tetrahedron
Lett. 1988, 29, 6417–6420.
[22] The pKa of toluene has been estimated as: a) 44.4ꢁ0.4 (in 70:30
THF/HMPA): B. Jaun, J. Schwarz, R. Brelow, J. Am. Chem. Soc.
1980, 102, 5741–5748; b) 40.91 (in toluene/cyclohexylamine/lithium
cyclohexylamide): A. Streiweiser, Jr., M. R. Granger, F. Mares,
R. A. Wolf, J. Am. Chem. Soc. 1973, 102, 4257–4261; and c) 42 (in
DMSO): F. G. Bordwell, D. Algrim, N. R. Vanier, J. Org. Chem.
1977, 42, 1817–1819; d) The pKa of H-HMDS has been determined
as 26 (in DMSO): F. G. Bordwell personal communication to J. E.
Bartmess, cited in D. T. Grimm, J. E. Bartmess, J. Am. Chem. Soc.
1992, 114, 1227–1231; and e) 29.5 (in THF): R. R. Fraser, T. S. Man-
sour, J. Org. Chem. 1984, 49, 3442–3444.
[29] It should be noted that [(2’)2ACHTRE(UNG K-HMDS)2], in which 2’ is the N,N’-
diisopropyl analogue of 2, has been isolated and its structure deter-
mined by X-ray diffraction–-see reference [25a]. This structure may
of course arise through the favourable symmetry/packing/packing
forces in the bis-carbene KHMDS dimer complex allowing crystalli-
sation of the minor component of the three-stage equilibrium.
[30] a) The observation that the net KIE on the rate of reaction for [D8]-
versus [D0]toluene as reactant and solvent, k(H)/k(D) =10.9ꢁ1.9 is in
reasonable but not complete agreement with the observed partition-
ing P=k[D ]tol/k[D ]tol =9.5ꢁ0.8 is suggestive of a solvent KIE of
[23] a) The pKa of the related formamidinium species 1,3-diisopropyl-
3,4,5,6-tetrahydropyrimidin-1-ium hexafluorophosphate and 1,3-di-
0
8
ethyl-3,4,5,6-tetrahydropyrimidin-1-ium
in
D2O/KCl
(ionic
about 1.2. Whilst such effects are common with protic solvents that
can effect deuteration at exchangeable reaction sites, and are thus
accountable by for example, proton inventory (see for example,
R. L. Schowen, Prog. Phys. Org. Chem. 1972, 9, 275–332) substan-
tial solvent KIEs in aprotic media are rather rare—see for example
b) L. R. Khundkar, J. W. Perry, J. E. Hanson, P. B. Dervan, J. Am.
Chem. Soc. 1994, 116, 9700–9709; c) Y. Zong, J. L. McHale, J.
Chem. Phys. 1997, 106, 4963–4973; d) R. Paur-Afshari, J. Lin, R. H.
Schultz, Organometallics 2000, 19, 1682–1691. Analysis of the chem-
ical shift of the 13C{1H} NMR signal arising from C(2) in [13C1]-2 in
the various [D8]/[D0]toluene mixtures indicates that Ka for (K-
HMDS)2 complexation is not affected by the [D8]/[D0]toluene ratio.
[31] See for example, A. Streitweiser, Jr., P. H. Owens, G. Sonnichsen,
W. K. Smith, G. R. Ziegler, H. M. Niemeyer, T. L. Kruger, J. Am.
Chem. Soc. 1973, 95, 4254–4257.
strength=1) at 258C have been determined as 28.2 and 27.4, respec-
tively, by modification of the method of Aymes et al. (ref. [23d]) to
take into account background hydrolysis of the pyrimidin-1-ium
cation. An analogous determination for N,N’-ditertbutylimidazolium
chloride gave a pKa value of 24.7 which can be compared to the
value of 22.7 obtained for the same acid cation in DMSO
(ref. [23c])—A. C. OꢀDonoghue, University of Durham, UK, person-
al communication to G. C. Lloyd-Jones. For other determinations/
computed estimations of carbene pKa values see: b) R. W. Alder,
P. R. Allen, S. J. Williams, J. Chem. Soc. Chem. Commun. 1995,
1267–1268; c) Y.-J. Kim, A. Streitwieser, J. Am. Chem. Soc. 2002
124, 5757–5761; d) T. L. Aymes, S. T. Diver, J. P. Richard, F. M.
Rivas, K. Toth, J. Am. Chem. Soc. 2004, 126, 4366–4374; e) A. M.
Magill, K. J. Cavell, B. F. Yates, J. Am. Chem. Soc. 2004, 126, 8717–
8724; f) D. Martin, O. Illa, A. Baceirdo, G. Bertrand, R. M. Ortuno,
V. Branchadell, J. Org. Chem. 2005, 70, 5671–5677.
[32] It should be noted that the KIE values calculated for (K-HMDS)2-
catalysed reactions of 2 with deuterated toluenes in toluene or
[D6]benzene are based on measurements in which the excess [(K-
HMDS)2] amounts to 60–80 mm and thus about 75% of 2 is com-
plexed. We have not determined KIEs for the individual reactions
of free 2 and complexed 2.
[33] For example, the methylene–benzene singlet!ipso-p complex is
computed to have a dissociation energy of 7.2 kcalmolꢀ1 and experi-
mentally (photoacoustic calorimetry) this is found to be 8.7ꢁ
3.1 kcalmolꢀ1: M. I. Khan, J. L. Goodman, J. Am. Chem. Soc. 1995,
117, 6635–6636; for a review of the area see W. Kirmse, Eur. J. Org.
Chem. 2005, 237–260.
[24] W. Clegg, G. C, Forbes, A. R. Kennedy, R. E. Mulvey, S. T. Liddle,
Chem. Commun. 2003, 406–407.
[25] a) R. W. Alder, M. E. Blake, C. Bortolotti, S. Bufali, C. P. Butts, E.
Linehan, J. M. Oliva, A. G. Orpen, M. J. Quayle, Chem. Commun.
1999, 241–242; b) P. L. Arnold, M. Rodden, C. Wilson, Chem.
Commun. 2005, 1743–1745.
[26] The chemical shift of C(2) in [13C1]-2 is found to decrease fractional-
ly with conversion on reaction with toluene. The gradient decreases
with increasing concentration of excess (K-HMDS)2 and most likely
arises from the depletion of excess (K-HMDS)2 at high [[13C1]-2]
and low [(K-HMDS)2]. Linear regression of plots of [[13C1]-2]
(across the 0 to 80 mm range of concentrations studied) versus
chemical shift of C(2) were found to have gradients orders of magni-
tude smaller than the Dd values observed across the range of [(K-
HMDS)2] studied.
[34] For an overview of heterocyclic-carbene–metal complexation and re-
activity of the resulting complexes see reference [2a].
[35] Alternative explanations for the increased reactivity of the complex
(K-HMDS)2] towards toluene, such as toluene–K+ p or s com-
[2ACHTREUNG
plexation (pre-organisation of the toluene proximal to C(2) in the
nascent carbene base generated on decomplexation) or deprotona-
tion of toluene by the K-coordinated HMDS nitrogen to give a
{Sz2c}0.5; in
[27] The Brønsted salt effect catalysis rate law: logkrel =mACHTREUNG
which z=charge=1; c=[K-HMDS]; m=a product of the Debye–
Hückel parameter, as applied to the monomer or mono-ionising
[2{K2ACHTRE(UNG HMDS)ACHERTU(NG H-HMDS)}(tetrahydropyrimidinium)]ACHERT[UNG PHCH2] ion
dimer should yield
a
linear correlation of logkrel versus [(K-
pair (if collapse of the ion pair intracomplex H/K exchange are fast,
then no cross over of H/D with the bulk-medium would occur) seem
less likely but cannot be ruled out at this stage. For an example of
p-tolyl–K+ complexation of (K-HMDS)2 detected in the solid state
by X-ray crystallography, see reference [9b]. For other examples of
p-aryl–K+ and s-aryl–K+ complexation see, a) C. J. Schaverlen, J. B.
van Mecherlen, Organometallics 1991, 10, 1704–1709; b) G. K.
HMDS)2]. See N. Isaacs in Physical Organic Chemistry, 2nd ed.,
Longman Press, Harlow, England, 1995, pp. 213–214. It is unlikely
that the dimer would undergo double ionisation.
[28] For Ka ={ [2
ed binding isotherm was derived by solving the quadratic equation
for standard 1:1 binding isotherm: y=[{(([{((a
(xꢀb))+1)2 +
ACHTREUNG(K-HMDS)2]/([2]ACHTRE[UGN (K-HMDS)2])}, an NMR-shift-detect-
a
ACHTREUNG
5374
www.chemeurj.org
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5361 – 5375

Insertion Reactions
FULL PAPER
Fukin, S. V. Lindeman, J. K. Kochi, J. Am. Chem. Soc. 2002, 124,
8329–8336, and references therein.
[38] H. Yamamoto, K. Maruoka, J. Am. Chem. Soc. 1981, 103, 4186–
4194.
[36] The N,N’-diethyl and N,N’-diisopropyl analogues of 2 undergo anal-
ogous reaction with toluene; G. C. Lloyd-Jones, G. J. J. Owen-Smith,
unpublished results.
[39] R. W. Alder, M. E. Blake, S. Bufali, C. P. Butts, A. G. Orpen, J.
Schütz, S. J. Williams, J. Chem. Soc. Perkin Trans. 1 2001, 1586–
1593.
[37] Triethyl orthoacetate was treated with N-isopropyl-1,3-propanedi-
[40] MacCurveFit V1.3 1996, Kevin Raner Software, Victoria, 3149, Aus-
tralia.
ACHTREUNGamine in the presence of catalytic p-TsOH.H2O at 1408C. The re-
sulting tetrahydropyrimidine intermediate was reduced with
DIBAL-H in toluene to afford N-ethyl-N’-isopropyl-1,3-diaminopro-
pane.
Received: February 24, 2006
Published online: May 4, 2006
Chem. Eur. J. 2006, 12, 5361 – 5375
ꢁ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5375