Entropies of organolithium aggregation based on measured microsolvation numbers

Article abstract of DOI:10.1021/om3009348

The recent measurement (J. Am. Chem. Soc.2008, 130, 14179-14188) of the microsolvation numbers of monodentate, nonchelating ethereal donor ligands coordinating to the monomers and dimers of two sterically shielded =C(aryl)-Li compounds permits the determination of well-founded dimerization enthalpies (ΔH⁰) and entropies (ΔS⁰) from properly formulated equilibrium constants, which must include the concentrations of the free donor ligands. The monomers are found to dimerize endothermically (ΔH⁰ > 0) in [D₈]toluene solution in the presence of the donor tBuOMe or THF, but only slightly exothermically (ΔH ⁰ = -0.5 kcal per mol of dimer) with the donor Et₂O. The dimerization entropies ΔS⁰ (in cal mol^-1 K ^-1) with the respective equivalents of released donor ligands are 7.2 and 11.0 (with 2 equiv of tBuOMe in the two cases), 6.1 (with 2 Et ₂O), and 34.1 (with 4 THF). It is shown that the improper omission of microsolvation from the equilibrium constant (a usual practice when the ligand numbers are not known) can lead to contaminated aggregation entropies ΔS_ψ, which may deviate considerably from the true entropies ΔS⁰. A method is provided for estimating the required microsolvation numbers from ¹³C/Li NMR coupling constants ¹J_C,Li for less congested organolithium types whose coordinated and free donor ligands cannot be distinguished by NMR integration.

Full text of DOI:10.1021/om3009348

Article
pubs.acs.org/Organometallics
Entropies of Organolithium Aggregation Based on Measured
Microsolvation Numbers
Rudolf Knorr,* Thomas Menke, and Kathrin Ferchland
̈
̈ ̈
Department Chemie, Ludwig-Maximilians-Universitat Munchen, Butenandtstrasse 5−13 (Haus F), 81377 Munchen, Germany
*
S Supporting Information
ABSTRACT: The recent measurement (J. Am. Chem. Soc. 2008, 130, 14179−14188) of the
microsolvation numbers of monodentate, nonchelating ethereal donor ligands coordinating to
the monomers and dimers of two sterically shielded C(aryl)−Li compounds permits the
0
0
determination of well-founded dimerization enthalpies (ΔH ) and entropies (ΔS ) from
properly formulated equilibrium constants, which must include the concentrations of the free
0
donor ligands. The monomers are found to dimerize endothermically (ΔH > 0) in [D ]toluene
8
0
solution in the presence of the donor tBuOMe or THF, but only slightly exothermically (ΔH = −0.5 kcal per mol of dimer)
0
−1 −1
with the donor Et O. The dimerization entropies ΔS (in cal mol K ) with the respective equivalents of released donor ligands
2
are 7.2 and 11.0 (with 2 equiv of tBuOMe in the two cases), 6.1 (with 2 Et O), and 34.1 (with 4 THF). It is shown that the
2
improper omission of microsolvation from the equilibrium constant (a usual practice when the ligand numbers are not known)
0
can lead to “contaminated” aggregation entropies ΔS , which may deviate considerably from the “true” entropies ΔS . A method
ψ
13
1
is provided for estimating the required microsolvation numbers from C/Li NMR coupling constants J for less congested
C,Li
organolithium types whose coordinated and free donor ligands cannot be distinguished by NMR integration.
INTRODUCTION
[A Li ]
2 2
[A1Li1]2
■
KMD
=
[free Don](2d
M
−2dD)
Reversible aggregations of polar organometallic compounds are
(2)
(3)
often accompanied by the endothermic release of monodentate
K = [
D]
[M]
(
nonchelating) donor ligands Don (THF, Et O, etc.) from the
= KMD[free Don]−(2dM−2dD)
2
ψ
2
metallic center. This increase of the number of independent
0
particles can result in a positive aggregation entropy (ΔS ),
0
0
0
which may carry information about the number of donor
ligands involved (microsolvation). This entropic bias tends to
favor aggregations at higher temperatures, while the endother-
micity implies that aggregations will become less favored on
ΔG = ΔH − TΔS = −RT ln K
MD
=
−RT ln K − RT(2d − 2d )ln[free Don]
ψ
M
D
(4)
1
ΔG
T
ψ
ΔH
ψ
T
cooling, and the practically important consequences were
R ln K = −
= −
+ ΔSψ
ψ
often emphasized: Working at lower temperatures will increase
the population of the smaller aggregates (which are often the
more reactive ones) and may also suppress unwelcome side-
reactions. Clearly, the detailed knowledge of the thermody-
namics and kinetics of aggregation equilibria can assist
quantitative reactivity studies. It can be of interest, therefore,
to determine reliable thermodynamic parameters (including
0
ΔH
0
=
−
+ ΔS − R(2d − 2d )ln[free Don]
M
D
T
(5)
3
With few exceptions, the degree of such changes (d − d in
M
D
eq 1) of microsolvation was generally unknown, so that the
proper inclusion of [free Don] in the equilibrium constants
(
KMD in eq 2, for example) was not possible. Therefore, and
0
ΔS values) for the quantification of aggregation equilibria, and
because solvation data from X-ray diffraction analyses of
crystalline organolithium compounds are not necessarily valid
for equilibria in solution, there was often no other choice than
we wish to point out the problems that arise, using the simple
example of a dimerizing organolithium monomer A Li in eq 1
1
1
(
where A = carbanion). The concentration [free Don] of the
to ignore the mathematical role of [free Don] in eq 2.
4
uncoordinated (free) portion of the donor ligand will co-
Common practice is to use equilibrium pseudoconstants K
ψ
determine the equilibrium (eq 2) unless d = d , where d and
that depend on [free Don], as formulated in eq 3 for the
M
D
M
dD are the microsolvation numbers per Li center of the
example depicted in eqs 1 and 2. Such a use of K in place of
ψ
KMD would produce erroneous free enthalpies of reaction, ΔG
monomeric (M) and dimeric (D) species A Li and A Li ,
ψ
is
1
1
2
2
−1
−1
2
= −RT ln K = ΔH − TΔS (where R = 1.986 cal mol
K
respectively.
ψ
ψ
ψ
2
A Li &d Don ⇄ A Li &2d Don + (2d − 2d )Don
Received: October 4, 2012
Published: January 9, 2013
1
1
M
2
2
D
M
D
(1)
©
2013 American Chemical Society
468
dx.doi.org/10.1021/om3009348 | Organometallics 2013, 32, 468−472

Organometallics
Article
the universal gas constant), in lieu of the proper results
equilibria of our system (Scheme 1) in [D ]toluene solution as
simple examples with measured microsolvation numbers.
8
0
(
namely, ΔG ) to be expected from eq 4. In particular, the
differential immobilization (that is, d − d ) of portions of
M
D
Don at the monomer and dimer fractions may bring about a
temperature dependence of [free Don] in eqs 2 and 4 in cases
of a temperature-dependent D/M concentration ratio even if
the total [Don] is held constant. Thus, the omission of [free
RESULTS AND DISCUSSION
Syntheses and properties of the two 2-(α-aryl-α-lithiomethyli-
■
10
dene)-1,1,3,3-tetramethylindans 1a and 1b were previously
1
13
described. The H and C NMR chemical shift values δ are
9
Don] and its temperature dependence by using K values may
ψ
characteristic of the monomeric (1aM and 1bM) and dimeric
result in erroneous pseudoenthalpies ΔH and pseudoentropies
ψ
(1aD and 1bD) forms below the M/D coalescence temper-
atures (mostly around −10 °C), so that simple integration of
the pairwise corresponding 1M and 1D signals can provide the
integral fractions y = [1M]/[1] for 1M and y = 2[1D]/[1]
ΔS in eq 5 (which is derived from eq 4 through
ψ
rearrangement). In the special case of a large excess of the
free donor over the organolithium substrate, [free Don] ≈ total
M
D
5
[
Don] would be almost independent of the temperature; it can
for 1D, where the concentration [1] = [1M] + 2[1D] gives the
employed sum of “monomeric” units (“per Li”) as present in
0
then be recognized from eqs 4 and 5 that d ln K /dT = ΔH /
MD
2
2
0
(
RT ) ≈ d ln K /dT = ΔH /(RT ), so that ΔH ≈ ΔH ,
ψ
ψ
ψ
1M and 1D together, because these are the units counted by
0
12
whereby eq 5 simplifies to ΔS ≈ ΔS − R(2d − 2d )ln[free
ψ
M
D
NMR integrations. With yD = 1 − y and dD = 1, the
M
Don]. In other words, ΔS can be a “contaminated” kind of
ψ
dimerization constant KMD of eq 2 takes the form of eqs 6a and
0
entropy and may be more negative or more positive than ΔS ,
6
b, where the known total donor concentration [Don] is
depending on the sign of the contaminating term R ln[free
diminished by the portions coordinated at Li in 1M&d Don
M
6
Don]. In the donor solvent THF, for example, ΔS would
ψ
and 1D&2d Don.
D
−1
then be too negative by an entropy equivalent of ca. 21 cal K
per mol of dimer in the case of the release of four THF
molecules, as encountered further below with d = 3 and d =
−
2
KMD = 0.5y [1](y [1])
D
M
M
D
(2dM−2dD)
×
([Don] − d y [1] − d y [1])
M D
M
D
(6a)
Scheme 1
−
2
=
0.5(1 − y )[1](y [1])
M M
(2dM−2)
×
{[Don] − d y [1] − (1 − y )[1]}
M M
M
(6b)
(7)
−
1
yM = (δ − δ )(δ − δ )
coal
D
M
D
The integral fractions y and y can no longer be obtained
M
D
by integration when the M/D interconversion rates become
rapid on the NMR time scale above ca. −5 °C: As illustrated in
Figure 1 for three of the 16 M/D pairs of 13C nuclei in 1a
(
Scheme 1), the coalesced shift values (δcoal) depend now on
the equilibrium composition (y ), so that yM can be
M
determined by eq 7, provided that the chemical shift values
δM and δD are known above the coalescence temperatures;
Figure 1 shows that the linear extrapolations of δ and δD
M
13
1
. For comparison, the more precise evaluation of RT ln[free
appear to be sufficiently reliable. By these two analytical
methods, yM could be measured between the temperature
extrema denoted as “low T/high T” in Table 1. The best yM
value at a particular temperature was obtained as an average
over the primary data from suitable signal pairs that were
chosen from the 16 types of 13C and nine types of H nuclei.
The yM values were then used for computing KMD (eq 6b),
whose temperature dependence (eq 4) yielded the thermody-
THF] for eq 5 in ref 5 would correspond to an entropic
−1
−1
equivalent of −20.9 (±0.9) cal mol
K
for the release of four
THF ligands. Understandably, there is a strong demand for
improving our knowledge of microsolvation numbers, as
witnessed by many quantum-chemical calculations and various
1
7
,8
attempts through diffusion-ordered NMR spectroscopy.
9
10
In our experimental model system (1a,b in Scheme 1),
0
0
steric congestion impedes the exchange of coordinated,
monodentate, nonchelating ethereal donor ligands to such a
degree that the microsolvation numbers d could be measured
directly through integration of the NMR signals: The
monomers 1aM and 1bM carry two tBuOMe (tert-butyl
methyl ether, dM = 2), two Et O (d = 2), or three THF
ligands (d = 3), while their dimers 1aD and 1bD are always
disolvated (d = 1). On the basis of the measured d values,
we discovered the more general empirical relationship d =
L( J n) − a, where the sensitivity factor L is ca. 42 Hz for
the Li isotope in Scheme 1, J is the scalar one-bond C−Li
spin−spin coupling constant, n is the number of Li cations
bound directly to the inspected carbanionic center, and a is the
number of carbanionic centers bound directly to a Li cation, so
that n = a = 1 for our monomers 1M in Scheme 1 and n = a = 2
for our dimers 1D. We will now quantify the dimerization
namic parameters ΔH and ΔS with low statistical errors
(Table 1) because of the very broad range of experimental
13
temperatures. Notice that these parameters refer to the
equilibrium formulation of eq 1 and are therefore defined per
dimer (that is, for two “monomeric” units). Back-calculation
0
0
(eq 4) at any other temperatures from these ΔH and ΔS data
2
M
gave K values, which predicted y (eqs 6) and finally (eq 7)
M
MD
M
11
theoretical temperature dependencies of δcoal; these are shown
as short lines between the extrapolated M and D lines on the
high-temperature side of Figure 1 and agree reasonably well
with the observed δcoal values.
D
9
1
−1
C,Li
6
1
13
C,Li
Suitable samples of 1a for equilibrium measurements were
prepared in NMR tubes from solutions of the crystalline,
9
unsolvated trimer in [D ]toluene at ambient temperature
8
under a cover of dry argon gas. Addition of the donor tBuOMe
(only 0.96 equivalent, 0.46 M) generated the disolvated dimer
4
69
dx.doi.org/10.1021/om3009348 | Organometallics 2013, 32, 468−472

Organometallics
Article
Scheme 2
microsolvation numbers d = 2 and d = 1 (entry 3 of Table
M
D
1
), the analogous Δpt = 1 results in the similar dimerization
0 −1 −1
entropy of ΔS = 6.1 cal mol K . The much more positive
0
11
ΔS value in entry 4 is a consequence of the established
microsolvation numbers d = 3 and dD = 1 for THF: The
M
particle balance Δpt in Table 1 is now three because five
particles (Scheme 2b: 1D plus four liberated THF) are formed
0
−1 −1
from two monomers, which results in ΔS = 34.1 cal mol
K
(
entry 4), that is, on average 11.4 entropy units per
independent particle, in reasonable agreement with the above
0
ΔS values for Don = tBuOMe and Et O.
2
Figure 1. Three examples of the temperature dependence of chemical
shifts δ , δ , and (coalesced) δ of 1aM and 1aD coordinating with
The bimolecular processes in Scheme 2a might formally be
envisioned as beginning with the cleavage of two second-
coordinating ligands Don from two molecules of 1M&2Don,
M
D
coal
the donor tBuOMe in [D ]toluene. Concentrations: filled symbols, [1]
16
8
=
[
≤
0.13−0.28 M, [tBuOMe] = 0.80 M; open symbols, [1] = 0.28 M,
tBuOMe] = 1.41 M; “I”, [1] = 0.45 M, [tBuOMe] = 0.70 M; “X”, [1]
0.48 M, [tBuOMe] = 0.46 M.
leaving two hypothetical monosolvated monomers
(
1M&1Don), which stick to the more strongly bound first-
16
coordinating donor ligands; the latter two fragments would
unite to give the dimer 1D&2Don through formation of two
1
3
17
0
1
aD&2tBuOMe quantitatively, as expected from the
new A−Li “bonds”. In a simplifying interpretation of the ΔH
9
measured microsolvation numbers d = 2 and d = 1; the
monomer 1aM&2tBuOMe grew in with additional amounts of
tBuOMe. The fast dimerization reactions of both 1aM and
value, the two new A−Li bonds in the dimer appear to be
M
D
0
−1
weaker by ΔH = 3.5 kcal mol than the two bonds connecting
16 18
two Li with two second-coordinating Don in 1M&2Don.
The inferior donor quality of Et O is illustrated by the very
weakly exothermic dimerization (ΔH = −0.5 kcal mol ),
which results from K values (Table S2 ) that were found to
0
−1
1
bM are endothermic, with ΔH = +3.5 kcal mol in entries 1
2
0
−1
and 2 of Table 1, which implies that the equilibrium constants
1
4
14
14
KMD, listed in Tables S1 (1a) and S4 (1b), grow with
increasing temperatures. This energetic preference of the
monomers 1a,bM&2Don is counterbalanced by the dimeriza-
MD
diminish only slightly with increasing temperatures. Loosely
speaking, the two bonds connecting two Li with two second-
0
16
tion entropy: The positive ΔS values (entries 1 and 2), which
favor the dimers upon heating, result partially through the
formation (Scheme 2a) of three product particles (namely, 1D
coordinating Et O donors (Scheme 2a) in 1aM&2Et O
2
2
0
18
appear weaker by −ΔH than the two new A−Li bonds in
1aD&2Et O. The low solubility of the p-SiMe analogue
2
3
plus 2dM − 2d = 2 released Don ligands) from two
1bD&2Et O in [D ]toluene prevented an analysis of a
corresponding equilibrium.
D
2 8
monomeric particles, which contributes the entropy increments
of 11.0 (entry 1) and 7.2 (entry 2) cal mol−
1
K
−1
on average for
A monomer 1M&3THF must lose the second- and the third-
15
16
a balance of one independent particle (Δpt = 1 in Table 1).
For Et O as the donor ligand with the same measured
coordinating of its three THF ligands (Scheme 2b) on
9
dimerization: Four of these Li−oxygen coordinative bonds
2
a
0
−1
0
−1 −1
Table 1. Microsolvation Numbers d and d , Thermodynamic Parameters ΔH (kcal mol ) and ΔS (cal mol K ), and
M
D
Auxiliary Data for the Dimerization of Monomers 1aM and 1bM in [D ]Toluene
8
b
b
0
0
c
d
e
entry
1
p-R
Don
2dM
2dD
ΔH
ΔS
low T/high T
ΔDon
Δpt
1
2
3
4
a
b
a
a
H
tBuOMe
tBuOMe
4
4
4
6
2
2
2
2
3.5 (±0.4)
3.4 (±0.2)
−0.5 (±0.2)
8.8 (±0.6)
11.0 (±1.6)
7.2 (±0.9)
6.1 (±0.8)
34.1 (±2.2)
−92/+70
−96/−20
−85/+25
−94/+75
2
2
2
4
1
1
1
3
SiMe3
H
Et O
2
H
THF
a
b
10
c
Defined per dimer (or per two monomers). Measured microsolvation numbers of 1a and 1b from ref 9. Low and high temperature limits (°C) of
d
e
the measurements. Number of free donor molecules released on dimerization. Balance of independent particles (starting particles subtracted from
products).
4
70
dx.doi.org/10.1021/om3009348 | Organometallics 2013, 32, 468−472

Organometallics
together are stronger by ΔH = 8.8 kcal mol than the two
Article
0
−1
scarce and the usually employed solvent mixtures of Me O,
2
1
8
new A−Li bonds in 1D&2THF. Subtraction of the
Et O, and THF with lower freezing points would be unsuited
for simple analyses of donor coordination. (Me O alone as a
2
14
thermodynamic data and the microsolvation interactions for
the ligand tBuOMe in entry 1 from those for THF in entry 4
can provide an assessment of the thermodynamic role of the
2
4
,26
27
solvent appears to be suitable
below −141 °C.) In this
situation, a habit of ignoring microsolvation and using
pseudoconstants Kψ (easily recognized if their units are
reported) to quantify aggregation equilibria can furnish
16
third-coordinating THF ligand in 1M&3THF: With the
proviso of some positive or negative (perhaps partially
14
canceling) correction terms, the entropic effect should be
erroneous thermodynamic parameters ΔHψ and ΔS , of
ψ
−1
−1
roughly (34 − 11)/2 = +11 cal mol K , and the enthalpic
which especially the “contaminated” entropy ΔS may differ
ψ
−1
0
15
effect roughly (8.8 − 3.5)/2 = +2.7 kcal mol . Similar
considerably from the “true” ΔS value. Nevertheless, K
ψ
−1
−1
19,20
magnitudes of ca. 10 cal mol
K
are known
for the
values encode product/source ratios that may be utilized under
closely equal conditions (such as the same [free Don] in eq 3),
entropy changes of melting crystals of moderately sized
molecules (relative mass 100−300), which suggests that the
while ΔH can provide the temperature dependence of a series
of such Kψ data under those conditions. However, the
respective ΔS values are contaminated by the temperature-
independent part of R ln K (eq 5) and hence may be used only
in this sense rather than for a direct interpretation of entropy
values, which is reserved for ΔS . Corresponding considerations
apply to organometallic ion-pair equilibria or to the reversible
ψ
+
mobility and flexibility of donor ligands released from Li may
on the whole resemble those properties of the particles in a
liquefying solid. In cyclopentane as the solvent in place of
ψ
ψ
[
D ]toluene, the dimerization of 1aM&3THF was found to
8
3
−3
have an equilibrium constant of KMD = 4.1 (±2.1) mol L at
2
0
5 °C, which resembles that in [D ]toluene (Table S3);
8
however, analyses at lower temperatures were not possible
because of the lower solubility.
The interconversion rates of our 1M/1D models are slow on
28
recombinations of S 1 ion pairs to their covalent isomers,
N
although the deviations are probably less spectacular.
9
The novel alternative possibility of estimating micro-
solvation numbers d from C/ Li (better than from
our NMR time scales below the coalescence temperatures T of
C
13
6
roughly −5 °C for 1a with tBuOMe (Figure 1), +5 °C for 1a
13
7
29
1
C/ Li) NMR spin−spin coupling constants J
via d =
C,Li
with Et O, −40 °C for 1a with THF, and −20 °C for 1b with
2
L( J n)−1 − a (as explained in the Introduction) provides a
1
C,Li
tBuOMe. These data correspond to crudely estimated free
‡
tool for determining d values indirectly along the following
activation enthalpies ΔG (2 M → D) in the range of 11 to 14
6
−1
lines. The coordination with n Li nuclei can be read at
kcal mol . Inspection of Scheme 1 leads one to expect that the
sufficiently low temperatures from the multiplicity and intensity
deaggregating rupture of two of the four A−Li “bonds” in 1D
of the 13C NMR splitting pattern of the carbanionic center,
(
Schemes 2a,b) implies loss of magnetic coherence between the
1
namely, a 1:1:1 triplet for n = 1 or a 1:2:3:2:1 quintet for n = 2,
separating nuclei, so that the J NMR splitting should vanish
C,Li
1
and so on. The magnitude of J
serves to identify the
with rates similar to those of the M/D interconversion (or
C,Li
9
aggregational and “fluxional” states of the equilibrium
components; at this stage, it is normally easy to deduce a
correct a value. The requisite sensitivity parameter L may be
faster, that is, at lower T , if more efficient mechanisms of Li
C
1
4
exchange operate). Our experimental observations are
consistent with these expectations.
3
0
chosen from a suitable (structurally related) example in the
3
1
32
CONCLUSIONS
published collection. As confirmed in that collection,
microsolvation numbers for monomeric tBuCH Li (dM = 3)
■
2
The availability and application of reliable microsolvation
numbers for monodentate ethereal donor ligands, as
exemplified here with 1a and 1b, can raise the quality of
thermodynamic analyses of C−Li aggregation to a level
comparable with that of N−Li compounds with properly
33
and its dimer (d = 2) in the solvent THF had been guessed
D
33
0
correctly; the reported dimerization entropy ΔS = +11.4 cal
−1
−1
mol
K
per dimer contains contributions of the release of
two-third-coordinating THF ligands from two monomers and
of dimerization of the two formally remaining fragments.
2
1−23
formulated
dimerization equilibria. We have attained
reasonably accurate thermodynamic parameters through
extension of the analytical scope (column “low T/high T” in
Table 1) from direct integration of the separate monomeric and
dimeric components to far above the NMR coalescence
temperatures. Our dimerization entropy of 1a with THF in
EXPERIMENTAL SECTION
■
Equipment and Data Collection. Reference 9 provides all details
10
1
of the preparation of 1a and 1b, the NMR equipment ( H NMR at
13
4
00 MHz, C NMR at 100.6 MHz), and the spectroscopic techniques
0
−1 −1
[
D ]toluene (ΔS = 34.1 cal mol
K
in entry 4 of Table 1) is
8
used here. Temperatures were determined before and after measure-
ments of the sample tubes, using calibrated NMR tubes (5 mm) filled
with methanol or ethylene glycol which have temperature-dependent
proton NMR shift differences. To circumvent C NMR relaxation
problems, dimerization equilibria were analyzed through measure-
ments of separate integral ratios of pairwise corresponding CH
signals for M and D on one hand and of pairwise corresponding signals
of nonprotonated 13C nuclei on the other hand. When these ratios
0
−1
−1
practically equal to ΔS = 33.8 cal mol
dimerization of monomeric 2-(lithiomethylidene)-1,1,3,3-tetra-
methylindan
in THF as the solvent. Approximately 11 cal mol
total entropy change might be attributed to the release of the
K
for the
2
4,25
13
(that is, 1a without the α-phenyl substituent)
−1
−1
K
of the
1
3
16
third-coordinating THF ligand.
We have used here the simple example of an organolithium
dimerization to draw attention to the general importance of
knowing microsolvation numbers for the determination and
interpretation of entropies of aggregation for mechanistic
purposes. At present, however, microsolvation by monodentate
1
were similar to those determined from H NMR signals, the reliable
values were averaged and then used to calculate the yM values and to
estimate their errors (Tables S5−S12). These errors propagate
0
0
through K (eq 2 and Tables S1−S4) into ΔH and ΔS (eq 4
MD
and Table 1). Our thermodynamic results are based on the usual
standard state concentrations (1 M). Natural logarithms (ln) are
understood to be calculated with the dimensionless magnitudes of the
employed quantities.
(
nonchelating) ethereal donor ligands cannot yet be
determined directly for most C−Li species because suitable
liquid media for NMR measurements at far below −110 °C are
4
71
dx.doi.org/10.1021/om3009348 | Organometallics 2013, 32, 468−472

Organometallics
Article
(
15) As a further example of the consequences of omitting
ASSOCIATED CONTENT
Supporting Information
■
microsolvation by setting dM = dD in eq 6a, the experimental yM
and y ) values (Table S4) for 1bM (and 1bD), as obtained with
1b] = 0.056 M and [tBuOMe] = 0.17 M in [D ]toluene, would
erroneously furnish ΔH = 2.8 (±0.1) kcal mol and ΔS = 15.2
(±0.6) cal mol
*
S
14
(
[
D
Tables S1−S4 of temperature-dependent equilibrium constants
8
0
and appertaining concentrations; delineation of the ΔH and
−1
ψ
ψ
0
−1
−1
0
ΔS contributions of the third-coordinating THF ligand; Tables
K
per dimer, to be compared with the correct ΔH
0
S5−S12 of temperature-dependent integral fractions y and
= 3.4 and ΔS = 7.2 in entry 2 of Table 1. (Due to [tBuOMe] < 1,
M
chemical shifts δ. This material is available free of charge via the
ln[free tBuOMe] in eq 5 is negative here, so that ΔS is much too
ψ
positive.)
Internet at http://pubs.acs.org.
(16) An impression of successively decreasing Li−ligand bond
*
R
Related Articles
Sterically Congested Molecules, 26. For Part 25, see: Knorr, R.;
Bohrer, G.; Schubert, B.; Bo
506−7515.
energies may be gained from calculations of the sequential
microsolvation steps at the (hypothetical) monomeric methyllithium,
as reported by: Abbotto, A.; Streitwieser, A.; Schleyer, P. von R. J. Am.
Chem. Soc. 1997, 119, 11255−11268, last paragraph on p 11257
therein.
̈
̈
hrer, P. Chem.−Eur. J. 2012, 18,
7
AUTHOR INFORMATION
Corresponding Author
E-mail: rhk@cup.uni-muenchen.de.
(17) The strong electric C−Li dipole moment of a monomer will be
practically cancelled in the corresponding dimer (see Scheme 1 or 2);
this was considered to be the major (thermodynamic) driving force
promoting aggregation in a weakly polar ethereal solvent, as remarked
■
*
Notes
̈
by: Reich, H. J.; Goldenberg, W. S.; Gudmundsson, B. O.; Sanders, A.
The authors declare no competing financial interest.
W.; Kulicke, K. J.; Simon, K.; Guzei, I. A. J. Am. Chem. Soc. 2001, 123,
8
067−8079, on p 8077.
ACKNOWLEDGMENTS
(18) Including adjustments concerning the changing energetic quality
■
of two Li−Don bonds, repulsive interactions, changing molecular
This work is dedicated to Professor Hendrik Zipse on the
occasion of his 50th birthday and in recognition of his
successful efforts to introduce the first-named author to the
techniques of ab initio computation. We thank Professor
Herbert Mayr for his support, the Deutsche Forschungsge-
meinschaft for the initial resources, and a reviewer for his very
thoughtful comments.
17
electric dipole moments, and electrostriction of the solvent. Notice
that the A−Li “bonds” drawn in the dimers are mainly electrostatic
interactions rather than two-electron bonds.
(19) Page, M. I.; Jencks, W. P. Proc. Natl. Acad. Sci. U. S. A. 1971, 68,
1678−1683.
(20) Searle, M. S.; Williams, D. H. J. Am. Chem. Soc. 1992, 114,
10690−10697.
(21) Lucht, B. L.; Collum, D. B. J. Am. Chem. Soc. 1995, 117, 9863−
9
(
7
(
1
(
874, Table 3 therein.
22) Hilmersson, G.; Davidsson, O. J. Org. Chem. 1995, 60, 7660−
669, on p 7665.
23) Lucht, B. L.; Collum, D. B. Acc. Chem. Res. 1999, 32, 1035−
042.
24) This is compound S3 in Table S1 of the Supporting Information
REFERENCES
■
̈
(
1) As explained for the dimerization of dimeric n-butyllithium by:
Seebach, D.; Has
̈
sig, R.; Gabriel, J. Helv. Chim. Acta 1983, 66, 308−
3
(
37, on p 320 therein.
2) “&” in eq 1 symbolizes coordinative bonds between the oxygen
atom of the donor and the lithium cation.
3) For a short list of pioneering discoveries of discrete NMR signals
of ref 9; it has also dM = 3 and d = 1 according to entries 63 and 64
D
(
therein.
for lithium-coordinated donor ligands, see: Reich, H. J.; Kulicke, K. J. J.
(
25) To compare the values per mol of dimer, we have doubled the
Am. Chem. Soc. 1996, 118, 273−274.
0
ΔS value found by: Knorr, R.; Freudenreich, J.; Polborn, K.; No
H.; Linti, G. Tetrahedron 1994, 50, 5845−5860, on p 5852.
26) Goldstein, M. J.; Wenzel, T. T. Helv. Chim. Acta 1984, 67,
029−2036.
̈
th,
(
4) (a) Heinzer, J.; Oth, J. F. M.; Seebach, D. Helv. Chim. Acta 1985,
6
8, 1848−1862, Tables 4−6 and Appendix E therein. (b) Jackman, L.
(
2
(
M.; Smith, B. D. J. Am. Chem. Soc. 1988, 110, 3829−3835, footnote 21
therein.
15
27) N-Labeled nitrogen donor ligands are appropriate for
(
5) Pratt, L. M.; Truhlar, D. G.; Cramer, C. J.; Kass, S. R.;
Thompson, J. D.; Xidos, J. D. J. Org. Chem. 2007, 72, 2962−2966, on p
964.
6) The concentration of THF as the solvent is ca. 14 M at ca. −90
C according to: (a) Metz, D. J.; Glines, A. J. Phys. Chem. 1967, 71,
6
determining microsolvation numbers through the Li NMR multi-
plicity pattern of J( N, Li) coupling. Because of their high ligand
1
15
6
2
(
exchange rates, however, such monodentate nitrogen ligands are
23
apparently unsuited for an analysis of the role of free donors in
°
1
1
aggregation equilibria. For an example and further references, see:
158. (b) Bauer, W.; Seebach, D. Helv. Chim. Acta 1984, 67, 1972−
Waldmu
Chem. Soc. 1997, 119, 5479−5480.
28) (a) Feigel, M.; Kessler, H. Chem. Ber. 1978, 111, 1659−1669.
b) Kessler, H.; Feigel, M. Acc. Chem. Res. 1982, 15, 2−8.
29) Fraenkel, G.; Fraenkel, A. M.; Geckle, M. J.; Schloss, F. J. Am.
Chem. Soc. 1979, 101, 4745−4747.
30) Listed together with the pertinent citations in Tables 2, S1, and
S2 of ref 9.
31) A further alternative for identifying the states of aggregation
̈
ller, D.; Kotsatos, B. J.; Nichols, M. A.; Williard, P. G. J. Am.
988, footnote 10 on p 1978 therein.
́
(
7) García-Alvarez, P.; Mulvey, R. E.; Parkinson, J. A. Angew. Chem.
(
2
(
011, 123, 9842−9845; Angew. Chem., Int. Ed. 2011, 50, 9668−9671.
(
8) Li, D.; Keresztes, I.; Hopson, R.; Williard, P. G. Acc. Chem. Res.
(
2
(
009, 42, 270−280.
9) Knorr, R.; Menke, T.; Ferchland, K.; Mehlstaubl, J.; Stephenson,
̈
(
D. S. J. Am. Chem. Soc. 2008, 130, 14179−14188.
(
(
(
10) These compounds correspond to 8 and 12 in ref 9.
(
11) See Table 1 in ref 9.
9
(
but not the “fluxional” state and probably not the d value) of
aryllithium species would make deliberate use of the C-ipso chemical
12) A partial 1 C NMR spectrum, recorded at −82 °C, was
3
13
displayed in Figure 7 of ref 9.
shifts of phenyllithium reported by: Reich, H. J.; Green, D. P.;
(
13) Given the presence of at least one equivalent of the donor
̈
Medina, M. A.; Goldenberg, W. S.; Gudmundsson, B. O.; Dykstra, R.
ligands, the chemical shift data δ of 1aM and 1aD provided no
evidence for perceptible amounts of other species according to pp
S27−S33 in the Supporting Information for ref 9. Therefore, the
microsolvation numbers did not change perceptibly over the
investigated temperature ranges.
R.; Phillips, N. H. J. Am. Chem. Soc. 1998, 120, 7201−7210, Figure 13
therein.
(32) Entries 38 and 43 in Table S1 of ref 9.
(33) Fraenkel, G.; Chow, A.; Winchester, W. R. J. Am. Chem. Soc.
1
990, 112, 6190−6198.
(
14) See the Supporting Information.
4
72
dx.doi.org/10.1021/om3009348 | Organometallics 2013, 32, 468−472