Metalation in hydrocarbon solvents: The mechanistic aspects of substrate-promoted ortho-metalations

Article abstract of DOI:10.1016/j.tet.2003.07.010

The methoxy-substituted aromatic reagents 1,2- and 1,3-dimethoxybenzene (1,2-DMB and 1,3-DMB) and 1,2,4-trimethoxybenzene (1,2,4-TMB) each undergo directed ortho-metalation in high yield in n-BuLi/hydrocarbon media without the aid of a catalyst. These rea

Full text of DOI:10.1016/j.tet.2003.07.010

Tetrahedron 59 (2003) 8275–8284
Metalation in hydrocarbon solvents: the mechanistic aspects of
substrate-promoted ortho-metalations
*
D. W. Slocum, Seth Dumbris, Scott Brown, Gina Jackson, Roslyn LaMastus, Elwood Mullins,
*
Jonathan Ray, Phillip Shelton, Amy Walstrom, J. Micah Wilcox and R. W. Holman
Laboratory of Covalent Chemistry,† Department of Chemistry, Western Kentucky University, Bowling Green, KY 42101, USA
Received 7 March 2003; revised 31 July 2003; accepted 31 July 2003
Abstract—The methoxy-substituted aromatic reagents 1,2- and 1,3-dimethoxybenzene (1,2-DMB and 1,3-DMB) and 1,2,4-trimethoxy-
benzene (1,2,4-TMB) each undergo directed ortho-metalation in high yield in n-BuLi/hydrocarbon media without the aid of a catalyst. These
reactions, coined ‘substrate-promoted ortho-metalations’, proceed with the methoxy aromatic substrate functioning as both the directing
metalation group (DMG) and as the deoligomerization agent. Evidence that the substrates themselves serve to deoligomerize n-BuLi comes
from 13C NMR. The relative extent of metalated product formed as a function of time for each of the three aromatics directly correlates with
the substrate’s time-dependent ability to coordinate to n-BuLi as measured by 13C NMR. The interpretation of NMR results from experiments
involving 1,2,4-TMB is consistent with the metalation proceeding via the activated complex [(1,2,4-TMB)2·(n-BuLi)2]. Finally, conclusions
from solubility experiments are that for every substrate-promoted metalation investigated, a precipitate forms in the hydrocarbon solvent, and
this precipitate mostly contains the ortho-lithiated aryl intermediate.
q 2003 Elsevier Ltd. All rights reserved.
1. Introduction
issues. Further, in industrial applications, hydrocarbon
solvents do not present as significant environmental or
safety problems as do bulk quantities of ethers. However,
hydrocarbon solvents are nonpolar and therefore cannot
deoligomerize alkyllithium reagents, hence deoligomeriz-
ing agents must be utilized. The main drawback of using
such agents in ether media is that their presence accelerates
media decomposition. Our efforts to prevent such occur-
rences initially involved use of fractional equivalents of
TMEDA in ether solvent2 and, more recently, equivalents of
THF in hydrocarbon solvent3 as well as fractional
equivalents of TMEDA in hydrocarbon solvent.4 These
metalation systems are noteworthy in that they afford
greater regiospecificity of and extents of metalation of the
various arene substrates than have been previously
reported.5
Ethers are the predominant solvents utilized for directed
ortho-metalation (DoM) reactions because they are suffi-
ciently polar to deoligomerize alkyllithium reagents and
also readily dissolve arene substrates. The disadvantages of
ethers as solvents in these applications are, however,
manifold. Ethers tend to react with alkyllithium reagents
under various metalation conditions, particularly when
agents such as tetramethylethylenediamine (TMEDA) are
present.1 Moreover, care must be taken to remove both
peroxides and water contaminants from ether solvents, a
concern that necessitates pre-reaction treatments of such
solvents or the use of excess alkyllithium reagent. As a
result of these instability and safety concerns, ether solvents
are seldom considered for large-scale industrial synthetic
applications.
Our interest here lies in the monometalation of the
dimethoxybenzenes (DMB’s) and the trimethoxybenzenes
(TMB’s), arenes whose metalations are well documented. In
the 1979 review by Gschwend and Rodriguez, some 50
metalations of 1,3-DMB are reported;6 in every case an
ether solvent was utilized. Only three metalation studies of
1,2-DMB are listed in the Gschwend and Rodriguez
review.6 No studies of 1,4-DMB are cited. Relatively
numerous studies of these species have been undertaken
in the ensuing 24 years with the vast majority of these
being undertaken in ether solvents.7–9 Monometalations
of 1,2,3-, 1,2,4-, and 1,3,5-TMB have received some
Hydrocarbons such as n-pentane, n-hexane, cyclohexane,
and toluene can also serve as solvents for metalation
reactions. These solvents are inert towards alkyllithium
reagents and peroxide and water contamination are non-
Keywords: metalation; deoligomerization; hydrocarbon solvents; di- and
tri-methoxybenzenes.
*
Corresponding authors. Tel.: þ1-502-745-3457; fax: þ1-502-745-6471;
e-mail: donald.slocum@wku.edu; robert.holman@wku.edu
Contribution #4 from the Laboratory of Covalent Chemistry at Western
Kentucky University.
†
0040–4020/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2003.07.010

8276
D. W. Slocum et al. / Tetrahedron 59 (2003) 8275–8284
literature attention.10 In two instances, dimetalation of the
trimethoxybenzenes has been reported.10a,d Again, ethers
were the principal component of the reaction media utilized.
2. Discussion and results
2.1. Background
We have recently reported initial observations with 1,2- and
1,3-DMB in which significant metalations in a hydrocarbon
solvent were achieved but with no conventional deoligo-
merization agent present.11 Ostensibly, the substrate
molecules themselves must sufficiently coordinate to the
alkyllithium reagent that deoligomerization occurs. This
assumption is reasonable in that n-BuLi in the hydrocarbon
solvent cannot deoligomerize without a ‘catalyst’ present.
The substrate is assumed to function as both the directing
metalation group (DMG) and as the deoligomerization
agent. To our knowledge, our work represents the first
‘substrate-promoted’ metalation recorded in the literature.11
Existing views hold that there are two general limiting
mechanisms for directed ortho-metalation (DoM) reactions
and that most DMG’s direct by a combination of these two
limiting mechanisms (Fig. 1). One of these limiting
mechanisms involves a prior-coordination complex
between the alkyllithium reagent and the DMG of the
aromatic substrate. This prior coordination complex is
paramount to involvement of the ‘complexed-induced
proximity effect’ (CIPE).12 Facilitated by the proximity of
the lithium atom to an ortho-hydrogen on the substrate, a
lithium/ortho-H exchange occurs. The energetics of this
exchange have been computationally determined to be
‘kinetically enhanced’,13 but there is little experimental
evidence to corroborate the predictions. In the second
limiting mechanism,6 the acidity of the ortho-hydrogen(s) is
the controlling factor (Fig. 1). The acidic ortho-hydrogen is
more readily removed by the basic n-BuLi, thereby
promoting a high rate of regiospecific metalation of the
aromatic ring at the ortho-position. In this mechanism,
herein coined the ‘overriding-base mechanism’, no prior-
coordination complex is involved.
The principal goals of this paper are (1) to further the
mechanistic understanding of ‘substrate-promoted’ metala-
tions in hydrocarbon solvents, (2) to extend the range of
arenes investigated over the two reported in our initial
publication11 and (3) to investigate the constitution of the
copious quantities of solids formed during the reactions.
The mechanistic goal is addressed by utilizing 13C NMR
methods for evaluating the extent of complexation between
the substrate and n-BuLi in these metalations. The emphasis
of the NMR work is to test the assumption that substrates
capable of undergoing metalation in hydrocarbon media do
so via coordination with and deoligomerization of n-BuLi.
The second goal is achieved by extending the protocol of
substrate-promoted metalation to include the three isomeric
trimethoxybenzenes. The total array of substrates investi-
gated include anisole, 1,2-dimethoxybenzene, 1,3-dimethoxy-
benzene, 1,2,4-trimethoxybenzene, 1,2,3-trimethoxybenzene,
1,3,5-trimethoxybenzene and 3-methoxydimethylaniline.
Lastly, to attain the third goal, the constitution of the isolated
solid material is investigated resulting in a relationship
between the extent of complexation (as determined by 13C
NMR) and the extent of metalation being proposed.
Consideration of alkyllithium dimer structure(s) must be
taken into account in the understanding of DoM reactivity.
The most commonly used alkyllithium, n-BuLi, has the
ability to equilibrate among three different olgimeric
structures: dimer, tetramer, and hexamer.14 Both the nature
of the solvent, as well as that of the deoligomerization agent,
affect oligomeric structure. In non-polar hydrocarbon
solvents a hexameric structure exists, which permits little,
if any, metalation of aromatic substrates. In diethylether
n-BuLi exists in a tetrameric form, whereas in THF a
mixture of the dimeric and tetrameric structures exist.
Moderate to high metalation yields can be achieved in these
ether solvents. While the tetramer may be reactive in some
Figure 1. The two limiting mechanisms for directed ortho-metalation.

D. W. Slocum et al. / Tetrahedron 59 (2003) 8275–8284
8277
cases, the dimer is considered the most reactive of the
n-BuLi oligomeric forms.15
methoxy groups as in 1,3-DMB, a computationally realized
minimum involving the bridged complex in Figure 3a has
been proposed.20
When catalysts such as THF and TMEDA are added to non-
polar hydrocarbon solvents in low concentration, significant
metalation yields result.3,4 Recently, low temperature NMR
studies have shown formation of a symmetrical dimeric
n-BuLi–TMEDA complex in the presence of stoichiometric
amounts of TMEDA in toluene-d8 under extreme dilution
(Fig. 2).16 Subsequently, Collum has shown that the dimer is
formed in a .20:1 ratio over any other oligomer structure in
toluene-d8 when TMEDA is added.17 Through kinetics
studies, Collum has also found that the transition state
stoichiometry for the metalation of anisole is [anisole·
(n-BuLi)2·(TMEDA)2].18
2.2. Metalation studies
The assertion that such substrate/n-BuLi complexes could
facilitate metalation is unprecedented. In fact, two recent
articles speculate that two methoxy groups ortho to each
other on the aromatic ring will either sterically interfere with
one another and thereby decrease the rate of ortho-
metalation10f or form a bidentate complex similar to that
in Figure 3b which is speculated to be less reactive.8k This is
contrary to what we have observed. Metalation of 1,2-DMB
in n-hexane at 258C without the presence of a catalyst
affords a maximum 78% corrected GC yield of the mono-
trimethylsilyl (TMS) product (Fig. 4). Interestingly, there
are slight solvent effects when the hydrocarbon solvent is
changed from hexane to toluene and then to cyclohexane.
As a function of time, the percent 3-TMS product in
cyclohexane tracks that in toluene, each yielding 72%
product after 6 h. However, in n-hexane solvent the yield of
3-TMS product approaches 80%. In all three hydrocarbon
solvent systems, copious amounts of precipitate are formed
with no trace of the product from dimetalation being
detected.21 The latter may be attributed to the relative
insolubility of the 3-lithio intermediate such that it is
removed from solution before a second metalation, known
to take place at the 6-position,10a can occur.
Figure 2. n-BuLi dimer formation facilitated by THF (monodentate) and
TMEDA (bidentate) coordination.
It can be envisioned that an aromatic species possessing
DMG’s sufficiently able to complex and thereby deoligo-
merize n-BuLi could simultaneously function as both the
substrate and as the ‘catalyst’. That is, the substrate itself
would first deoligomerize the alkyllithium reagent to a
reactive dimer, and then, once in this reactive form, the
substrate can react to generate a lithiated aromatic
intermediate. A potential species to affect such chemistry
is 1,2-dimethoxybenzene (1,2-DMB). Heretofore, this
substrate would be considered to undergo 3-metalation by
a combination of the two limiting mechanisms (Fig. 1). It is
our hypothesis that the 1,2-situation of the methoxy groups
will bis-coordinate to a lithium within a dimeric complex
(Fig. 3b). Calculations at the ab initio density functional
level (B3LYP 6-311þG**) support a 1,2-DMB bis-
coordinated dimer structure (Fig. 3b) in that (1) the through
A competitive metalation of 1 equiv. of a system made up of
1,2-DMB and 1 equiv. of anisole was performed in
cyclohexane at 258C using 1 equiv. of n-BuLi. After 6 h,
54% metalation of the 1,2-DMB occurred along with 15%
metalation of anisole. This demonstrated, in a manner
different from that in our preliminary publication,11 that the
rate of metalation of 1,2-DMB is greater than that of
anisole.10k It is curious, however, that the ratio of time-
dependent yields for this in situ competition is no better than
4:1 whereas comparison of separate experimentally deter-
mined time-dependent metalation yields, as reported in our
preliminary publication, resulted in a ratio of 24:1.11 The
relatively high extent of metalation of anisole (15%)
observed under the in situ conditions is the cause of this
discrepancy since the extent of anisole metalation in
hydrocarbon solvent is recorded as being only 3% in
6 h.3a This suggested that the presence of 1,2-DMB was
accelerating the rate of metalation of anisole. To test this
hypothesis, experiments were set up where 0.2 equiv. of
1,2-DMB and 1.0 equiv. of anisole in cyclohexane were
treated with 1.0 equiv. of n-BuLi. After 6 h at 258C, the
0.2 equiv. of 1,2-DMB was about 90% metalated and the
full equiv. of anisole was over 50% metalated. In other
words, 1,2-DMB promoted the metalation of anisole. This
remarkable observation reinforces our contention that n-
BuLi in the presence of 1,2-DMB can form the dimer
intermediate as in Figure 3b and that this dimer intermediate
can be intercepted by anisole in a manner similar to which
the n-BuLi·TMEDA dimer is intercepted by anisole.
˚
space distance between the oxygens of 1,2-DMB is 2.63 A,
˚
which compares to 2.98 A for the N–N distance in
TMEDA, (2) a favorable 5-membered ring chelation
interaction is formed in both cases and (3) a stable local
minimum structure corresponding to Figure 3b is found
whose geometry is reasonable.19 For the 1,3-arrangement of
Several reactions were examined where ether was used
with or without added TMEDA as the media for metalating
1,2-DMB so that a direct comparison could be made with
Figure 3. (a) The bidentate complexation of 1,3-DMB proposed by Saa20
and (b) an envisioned complex between 1,2-DMB and the n-BuLi dimer.
The second 1,3-DMB ligand in (a) has been omitted for clarity.

8278
D. W. Slocum et al. / Tetrahedron 59 (2003) 8275–8284
Figure 4. Substrate-promoted metalation of 1,2-DMB in common hydrocarbon solvents.
Table 1. % Complex formation between methoxyarenes and n-BuLi via
13C NMR
our metalation results in hydrocarbon media. No conditions
were found where the metalation of 1,2-DMB proceeded as
well in ether solvent as it did in hydrocarbon solvent. The
best reported results for the metalation of 1,2-DMB are from
the laboratories of Crowther and Sundberg10a who report an
ultimate yield of 86% monometalated product in n-BuLi/
ether with TMEDA as catalyst. However, a 10% excess of
the 1:1 n-BuLi/TMEDA reagent and a two day reaction
period were utilized. When run for shorter periods under
similar conditions, but with no excess of n-BuLi and
utilizing fractional equivalents of TMEDA,4 lower extents
of metalation were observed (Fig. 5).
Substrate
Time Solid Complex in solution Total complexed
(%)a
(h)
(%)
(%)
Anisole 1 M
0
1
3
24
0
1
3
0
1
3
0
1
3
0
1
3
0
0
0
0
8
49
51
87
95
.98
100
100
100
.98
.98
.98
0
0
,1
7
26
43
68
29
52
–
–
–
–
–
–
–
0
0
,1
7
32
71
84
91
1,3-DMB 1 M
1,2-DMB 1 M
1,2,4-TMB 1 M
1,2,4-TMB 0.1 M
97
.98
100
100
100
.98
.98
.98
a
It is assumed that the lithiated material that has precipitated has passed
through the soluble complex stage. The % total complexed¼%
solidþ(1002% solid)£% complex in solution.
Figure 5. Metalation of 1,2-DMB in diethylether.
A second potential candidate for substrate-promoted
metalation is 1,3-dimethoxybenzene (1,3-DMB). NMR
data mandate that 1,3-DMB forms a complex with n-BuLi
(Table 1), the nature of which is still open to investi-
gation.19,20 Nevertheless, an analogous CIPE mechanism
can be envisioned in which the 1,3-DMB serves as both the
substrate and as the catalyst (Fig. 3a). When we conducted
the reaction of 1,3-DMB in pure hydrocarbon solvent at
258C at various concentrations, greater than 50% yield of
the 2-metalated product was achieved (Fig. 6). Significantly
less precipitate formation for this system was observed than
for any of the other substrate systems. The range of
concentration of reactants that exhibited the highest yields
Figure 6. Substrate-promoted metalation of 1,3-DMB in n-hexane at
various concentrations.
was from 0.9 to 2.0 M. To test for temperature effects on the
overall yield of derivatized product, the reaction in
cyclohexane was performed at 25, 45, and 608C (Fig. 7).

D. W. Slocum et al. / Tetrahedron 59 (2003) 8275–8284
8279
and NMR data. Nucleophilic displacement of a methoxy
group from an aromatic ring is not unprecedented. In
appropriate arenes, methoxy group displacement has been
used in synthesis.23 However, this secondary reaction
pathway has gone unrecognized in previous reports of the
metalation of the 1,2,3-TMB system.10a,c
Lastly, 3-methoxydimethylaniline (3-MDMA) has been
briefly examined. At 608C in cyclohexane about 4%
metalation was achieved. As 3-MDMA is quite soluble in
cyclohexane, unlike 1,3,5-TMB, and is relatively polar as
compared to 1,3,5-TMB, the reason(s) for the lack of
reactivity of 3-MDMA is puzzling. These systems, as well
as related analogs, are still under investigation.
Figure 7. Substrate-promoted metalation of 1,3-DMB in cyclohexane at
various temperatures.
Examination of the precipitates produced by metalations
under substrate-promoted conditions revealed that they are
chiefly comprised of the appropriate ortho-lithio intermedi-
ates. The data presented in Figures 4–8 arise from
experiments in which the ortho-lithiated aryl ion of the
substrate was quenched by reaction with ClTMS. The extent
of TMS derivative formed was then plotted against time. In
the generation of such data, a wide bore syringe was utilized
to remove, from a rapidly stirred suspension, a representa-
tive aliquot containing both solid and liquid, which is added
to a second vial with the quenching agent, ClTMS. We
modified this approach for the purpose of determining the
identity of the solid material. In one type of experiment, the
reaction media was allowed to settle without stirring and a
syringe needle with a filter was utilized so as to ensure that
the liquid phase aliquot taken contained no solid. When this
liquid phase aliquot was quenched with ClTMS in the usual
manner and the resulting mixture was analyzed, no
derivitized product was observed for any of the DMG
substrates investigated. In each case, unreacted starting
material was detected. In a second type of experiment, the
majority of the liquid from the unstirred reaction vessel was
removed so that what remained was almost pure solid
material. This solid was then transferred via a wide bore
syringe and added to a second vessel containing ClTMS.
The resulting mixture, when analyzed, contained a high
yield of TMS derivative and little or no unreacted starting
material. The conclusion from these experiments is that for
every substrate investigated, the solid formed is constituted
primarily of the ortho-lithiated aryl intermediate. Therefore,
since the ortho-lithiated species are relatively insoluble, no
direct evidence for such species in our NMR data can be
anticipated.
The yield improved with increasing temperature to provide
over 82% derivatized product, which equals the best
reported yields for metalation of this substrate in the
literature.22
Extension of the substrate-promoted concept to the TMB’s
revealed further surprises. 1,3,5-TMB did not undergo
significant metalation in any of the hydrocarbon solvents.
This may be attributable to a relative lack of solubility of
1,3,5-TMB in hydrocarbon solvents, but it also could be
caused by some subtle electronic factor.10k In direct
contrast, 1,2,4-TMB underwent immediate substrate-
promoted metalation in cyclohexane and toluene with an
84% yield being achieved in cyclohexane in just 10 min
(Fig. 8). Rapid solidification of the reaction mixture was
also observed. That 1,2,4-TMB undergoes metalation in the
3-position has been documented in the literature,10a but the
extraordinary rapid rate of metalation under these con-
ditions is unprecedented. A combination of the bridging
intermediates proposed for the 1,2- and 1,3-DMB system
(Fig. 3) can be envisioned. The extraordinary rapid rate of
metalation of 1,2,4-TMB may be attributable to the relative
greater acidity of the 3-position proton in this arene.
2.3. NMR studies
Having established that the solids produced during these
metalations were the respective ortho-lithio intermediates,
the proposed bis-coordinated intermediates (Fig. 3) were
inferred to remain in solution. To demonstrate the presence
of the proposed soluble bis-coordinated species, NMR
examination of reaction solutions was instigated. 7Li
spectra, insightful in other of our investigations, pro-
vided no information regarding substrate complexation. In
contrast, 13C spectra proved particularly rewarding as
separate signals for complexed and uncomplexed methoxy
groups could be discerned in all spectra. Specifically, we
analyzed anisole, 1,2-DMB, 1,3-DMB, and two isomeric
Figure 8. Substrate-promoted metalation of 1,2,4-TMB in common
hydrocarbon solvents.
1,2,3-TMB undergoes substrate-promoted metalation, but
with a complicating feature. In both hydrocarbon and ether
media a second product was detected. Precipitate formation
was observed in both media, with that in hydrocarbon media
being quite significant. This second product, produced to the
extent of 5–20% in both media, has now been isolated and
identified as 2-n-butyl-1,3-DMB by a combination of MS

8280
D. W. Slocum et al. / Tetrahedron 59 (2003) 8275–8284
trimethoxybenzenes, 1,2,4-TMB and 1,2,3-TMB. As stated
previously, the 1,2,3-TMB exhibited a complicating
secondary metalation which confused interpretation of the
13C spectra. In each case, the data was analyzed with an
emphasis on measuring the extent of substrate/n-BuLi
complex formed as a function of time and then comparing
that evidence against the degree of substrate-promoted
metalation that occurred for the same substrate and n-BuLi
under reaction conditions.
signal to a methoxy group from a 1,3-DMB that is
complexed to an n-BuLi oligomer. Because there is only
one symmetry type for the methoxy group, it is certain that
both methoxy groups of the 1,3-DMB must be complexed in
a bidentate fashion. Saa20 would suggest it to be the
bidentate structure incorporating 1,3-DMB shown in
Figure 3a. It should be noted that new aromatic signals
also emerge with time, consistent with condensed-phase
complexed species. The amount of the complexed form of
1,3-DMB versus uncomplexed 1,3-DMB increases dramati-
cally faster than that for anisole (Table 1). At time zero,
defined as the time in which the n-BuLi was added to
substrate plus the analysis time for the NMR spectrum (ca.
45 min), the amount of total complexed 1,3-DMB corre-
sponds to 32%. After 1 h reaction time, this mixture
constitutes 71% and increases to roughly 84% after 3 h.
Interestingly, after 3 h, the extent of solid formation is 51%,
closely approximating the yield of 2-TMS product after
the same period (Fig. 6). Therefore, direct evidence for
1,3-DMB substrate-promoted complexation to n-BuLi is
provided with the amount of complex formed as a function
of time tracking the time-dependant yield of metalated
product.
In order for any species to be observed by the NMR methods
employed here, it is necessary for that species to be soluble
in the solvent media utilized. If the complex or the aryl
metalated (ortho-lithiated) ion that comes from the complex
is insoluble and falls out of solution, direct evidence for that
species is lost, in that the line widths of the solid material
broaden into the baseline and, therefore, their signals do not
appear in the spectrum. Since some solid forms in every
substrate-promoted reaction performed, we endeavored to
determine the identity of the insoluble solid material prior to
the NMR investigations enabling proper assignments in the
13C spectra once that data was collected.
2.3.1. Anisole. Anisole is not efficiently metalated in hydro-
carbon media. Without addition of THF or TMEDA,3,4 the
extent of metalation of anisole in hydrocarbon solvents at
258C never exceeds 10%, even after 24 h. The reason that
anisole cannot undergo substrate-promoted metalation is its
relative inability to complex to and deoligomerize n-BuLi.
Because anisole possesses only a single methoxy group, it
cannot undergo what seemingly is required for high effi-
ciency substrate-promoted deoligomerization—bidentate
complexation.
2.3.3. 1,2-DMB. There exists one methoxy carbon signal
at 54.8 ppm in the 13C NMR spectrum of 1,2-DMB in
cyclohexane. When n-BuLi in cyclohexane is added to the
1,2-DMB to generate a mixture with a 1:1 molar ratio
of n-BuLi:1,2-DMB, a second methoxy carbon signal at
53.9 ppm begins to emerge with increasing time. The new
signal represents the methoxy groups from a 1,2-DMB that
is complexed to some n-BuLi oligomer. Again, there is only
one symmetry type for the methoxy group, and therefore
both methoxy groups of the 1,2-DMB are asserted to be
complexed in a bidentate fashion, presumably as in
Figure 3b. The amount of the complexed form of 1,2-DMB
versus uncomplexed 1,2-DMB increases with time more
markedly than that for 1,3-DMB. At time zero, 91% total
complexation is observed (Table 1). After 1 h reaction time,
the total complexed form is approximately 97%. After 3 h,
the NMR sample tube containing the 1,2-DMB/n-BuLi/
cyclohexane is nearly entirely solid, corresponding to better
than 98% total complexation. The NMR spectrum after 3 h
showed almost no uncomplexed species remaining, and the
soluble complexed species was barely above the noise level
(though much larger than the signal for the uncomplexed
species). Here it is assumed that, as the concentration of
the complex increases, it becomes insoluble and, as such,
becomes transparent to the NMR observation methods
employed. Interestingly, after 1 h reaction time, only 32% of
1,3-DMB had complexed to n-BuLi whereas, after the same
period, 1,2-DMB had almost completely complexed. This is
quite consistent with the observed relative metalation
chemistry, where the time-dependent yield of metalated
product for 1,2-DMB (Fig. 4), is over twice that for
1,3-DMB (Fig. 6). Therefore, the increased extent of
metalation product formed as a function of time for
1,2-DMB over 1,3-DMB directly correlates with the
increased ability that 1,2-DMB has over 1,3-DMB to
complex to n-BuLi.
When the 13C NMR spectra of anisole in cyclohexane is
compared to a similar spectrum of a 1:1 mixture of
anisole:n-BuLi that has been allowed to equilibrate at
room temperature for 1, 3, and 24 h, several observations
are noted (Table 1). The NMR signals for the anisole
carbons do not change in the first 3 h of equilibration time.
After 24 h, a set of minor signals have emerged. The signals
correspond to a new form of anisole, the most tractable
signal being the one at ca. 53.9 ppm which is assigned to the
complexed methoxy group. This signal, after 24 h, has
grown to roughly 7% of the intensity of the original
methoxy signal at 54.9 ppm. Because we have demonstrated
that these ortho-lithiated aryl intermediates fall out of
solution and become transparent to NMR observation, the
new set of signals are assigned to a soluble anisole-
complexed n-BuLi oligomer. Hence, as expected, and
consistent with metalation results, anisole does not complex
to an n-BuLi oligomer (tetramer or dimer) until after 24 h of
room temperature equilibration, and then only to a small
degree.
2.3.2. 1,3-DMB. There exists one methoxy carbon signal at
54.7 ppm in the 13C NMR spectrum of 1,3-DMB in
cyclohexane. When sufficient n-BuLi in cyclohexane is
added to the 1,3-DMB to generate a mixture with a 1:1
molar ratio of n-BuLi to 1,3-DMB, a second methoxy
carbon signal at 53.5 ppm begins to emerge with increasing
time. The new signal represents an additional methoxy
group environment with a single symmetry. We assign this
In view of these findings, a reasonable question arises: why
does 1,2-DMB complex to, and therefore deoligomerize,

D. W. Slocum et al. / Tetrahedron 59 (2003) 8275–8284
8281
n-BuLi better than does 1,3-DMB? If it is asserted that the
complex involves an n-BuLi dimer, then it could be argued
that 1,2-bidentate complexation is more favorable than
1,3-bidentate complexation. The investigations with the
trimethoxybenzenes (TMB’s) discussed below shed some
light on this issue.
return to a 1:1 ratio, once more evidence for the substrate
and for n-BuLi vanishes from the spectrum (and additional
solid falls out of solution). It is possible to see NMR
evidence for the 1,2,4-TMB substrate only when it exists
in greater than a 1:1 molar ratio with respect to n-BuLi. The
n-BuLi signals are present only when excess alkyllithium
exists. Whenever a 1:1 molar ratio of 1,2,4-TMB:n-BuLi is
achieved, neither species survives in solution and hence the
solution is transparent to the NMR measurements made
under the conditions of the experiment.
2.3.4. 1,2,4-TMB. The 13C NMR spectrum of 1,2,4-TMB
possesses three distinct signals between 54 and 56 ppm that
correspond to the three different methoxy groups of the
unsymmetric substrate (Fig. 9). When n-BuLi in cyclo-
hexane is added to the 1,2,4-TMB to generate a mixture
with a 1:1 molar ratio of n-BuLi to 1,2,4-TMB (1 M
concentration of n-BuLi), complete solidification of the
system in the NMR tube occurs and solution NMR
spectroscopy is not possible (Table 1). In an effort to gain
more insight, the sample tube was again prepared at a 1:1
molar ratio of n-BuLi to 1,2,4-TMB, but this time at 0.1 M
in n-BuLi. The lower concentration still resulted in some
solid formation, but an NMR spectrum could be obtained
(Fig. 9, middle spectrum). No evidence for either 1,2,4-TMB
or n-BuLi exists in the spectrum (the only observable
signals were from cyclohexane and toluene-d8). The two
species apparently formed a complex and/or reacted and
fell out of solution. To test this, excess 1,2,4-TMB was
added and a new spectrum was obtained (Fig. 9, top
spectrum). This spectrum clearly shows the added, uncom-
plexed 1,2,4-TMB. If more n-BuLi is added so as to again
The most reasonable conclusion that can be drawn from
these observations is that about 75% lithiation of the
1,2,4-TMB instantaneously occurs (Fig. 8) generating
the 3-lithio intermediate which immediately precipitates
from solution. In so doing, it includes the remaining 25% of
substrate and n-BuLi. It may be fortuitous that the ratio of
the n-BuLi to substrate that disappears from solution is 1:1,
but we infer that rapid complex formation has taken place
to form a bidentate complex analogous to those in Figure 3,
i.e. [(1,2,4-TMB)2·(n-BuLi)2]. Moreover, we contend,
this complex is occluded and/or coprecipitated with the
rapidly formed 3-lithio-1,2,4-TMB intermediate. This
conclusion is reached as follows: since the time-dependent
extent of substrate complexation by NMR (Table 1) and the
time-dependent extent of metalation (Figs. 4–8) track each
other in the order anisole ,1,3-DMB,1,2-DMB for both
variables, the observation that the metalation of 1,2,4-TMB
Figure 9. The room temperature 13C NMR spectra of 1,2,4-TMB in cyclohexane (bottom), a 1:1 mixture of 1,2,4-TMB/n-BuLi (middle), and a mixture of
1,2,4-TMB/n-BuLi with a slight molar excess of 1,2,4-TMB.

8282
D. W. Slocum et al. / Tetrahedron 59 (2003) 8275–8284
is virtually instantaneous suggests that the rate and extent of
complexation are both instantaneous. It is not unreasonable
to conclude that the unreacted substrate and unreacted
n-BuLi lost from the solution and therefore present in the
solid, is there as the bidentately complexed n-BuLi dimer.24
reported here are comparable to the best yields reported for
the metalation of these species in ether solvent with added
catalyst.6,8–10
4. Experimental
3. Conclusions
4.1. Metalation studies
1,2- and 1,3-DMB as well as 1,2,4-TMB undergo ortho-
metalation in high yield in an n-BuLi/hydrocarbon solvent
media without the aid of a catalyst. Since n-BuLi is in its
least reactive oligomeric form in hydrocarbon solvents and,
we believe, does not directly undergo lithium/ortho-H
exchange from this oligomeric form, the substrate must aid
in the deoligomerization of n-BuLi by means of complexa-
tion. The interpretation of NMR data is consistent with this
premise and enables the quantitation of percent complexa-
tion for each substrate with n-BuLi. This has proven
possible because, on the NMR time scale, exchange of the
environments between complexed and uncomplexed
methoxy groups in solution is slow. In contrast, the
TMEDA-catalyzed metalations of similar substrates involve
rapid exchange between complexed and uncomplexed
TMEDA.25 The results of the percent complex formation
as a function of time correlate directly to the extents of
product formation in actual metalation experiments (com-
pare Table 1 and Figures 4, 6 and 8). These complexes are
partially soluble in hydrocarbon media, but substantial
amounts of the complexes can, and do, fall out of solution.
Solubility studies indicate that for each substrate the ortho-
lithiated intermediates that come from the substrate/n-BuLi
complex are themselves insoluble in the media utilized.
Therefore, under our conditions, when metalation reactions
take place, the electrophilic species (in our case, ClTMS)
used to quench the reaction reacts in solution with a solid
form of the nucleophilic ortho-lithiated aryl intermediate.
All reagents were purchased from Aldrich Chemical
Company and used without further purification unless
otherwise mentioned. 0.04 mol of 1,2-DMB or 1,3-DMB
was slurried in 10 mL of dry cyclohexane or other hydro-
˚
carbon solvent (dried over molecular sieves, 4 A, 4–8
mesh) in a one-neck flask sealed with a septum and purged
with N2. For the TMB’s, 0.02 mol of the reactants in 30 mL
of solvent was utilized. The flask was placed in a water bath
maintained at the appropriate temperature. To this solution
was added 20 mL (0.04 mol) of 2.0 M n-BuLi in cyclo-
hexane to bring the volume of cyclohexane to 30 mL. Runs
in hexanes utilized either 1.6 M n-BuLi in hexanes (FMC
Corp) or 2.5 M n-BuLi in hexanes with suitable adjustment
to make the final volume 30 mL. Runs in toluene utilized
10 M n-BuLi. For 1,2-DMB, most hydrocarbon solutions
became pastes upon addition of n-BuLi. After 2 h, stirring of
the solutions could be effected so that representative
samples could be taken with a wide-bore syringe inserted
through the septum resulting in particulate matter as well as
solution being taken up. The plot of percentage metalation
versus time was constructed by extracting 1 mL samples
from the slurry after 2, 10, 30 min and 1, 2, 4 and 6 h, or
other periods as necessary, and quenching with a solution
containing excess ClTMS in n-hexane. After aqueous
workup each sample was subjected to GC analysis
(HP 5890 A with OV-17 packed glass column). Identities
of products were checked by GC–MS analysis (Agilent
5973 MSD with 6890N Network GC system). Analysis of
corrected spectra of TMS derivative(s) provided an estimate
of the extent of metalation as measured by the lithio-
intermediate’s capture of the TMS moiety. Each plot
represents an average of at least three runs, unless otherwise
designated.
The interpretation of the reactivity results from experiments
involving 1,2,4-TMB suggests that the n-BuLi within the
complex is in dimeric form. Further, the stoichiometry of
this complex supports our contention that all three multi-
methoxy substrates engage in bidentate complexation. The
ability of the substrate to form a bidentate complex, then, is
a prerequisite for substrate-promoted ortho-metalation.
Anisole, which possesses a single methoxy group, is not
metalated well without the aid of a catalyst, substantiating
the claim that the key to substrate-promoted ortho-
metalation lies within a substrate’s ability to form a
bidentate complex. Two forms of bidentate complexa-
tion are proposed. When two methoxy groups are in the
1,2-positions, bidentate complexation as depicted in
Figure 3b takes place. Complexation of two methoxy
groups situated 1,3- with respect to one another can also
occur in a fashion perhaps best represented by Saa20 as the
bridged structure in Figure 3a. The former mode of
complexation is apparently more efficient than is the latter.
4.2. NMR studies
All NMR experiments were performed on a 270 MHz JEOL
CPF-270 instrument. The 13C NMR spectra, collected at
room temperature, were an average of 512 scans, taken with
a three second pulse delay, equating to a ca. 43-min
acquisition time for each experiment. Thus, for a spectra
reported as a 3 h point, the sample was equilibrated at room
temperature for 3 h and then the spectra collected between
times 3:00 and ca. 3:43 h. In the NMR investigations, each
NMR tube contained 8£1024 mol of n-BuLi. For the 1 M
samples, the total volume of each NMR tube was 800 mL,
consisting of 200 mL of toluene-d8, 400 mL of 2 M n-BuLi
in cyclohexane and a combination of substrate and
cyclohexane equaling 200 mL volume. The particular
substrate in each experiment was added in a 1:1 molar
ratio with n-BuLi. In the investigation of some substrates,
minor solid formation was observed in the NMR tubes.
Solid formation tended to increase as a function of time.
However, no problem was encountered in the successful
Finally, the percent yields of ortho-substituted aromatic
product in the metalation of 1,2-DMB, 1,3-DMB, and
1,2,4-TMB in hydrocarbon solvent without the aid of
catalyst (78, 82 and 84%, respectively) exceeds that for the
same substrates reacted in pure ether solvent. The yields

D. W. Slocum et al. / Tetrahedron 59 (2003) 8275–8284
8283
generation of valid 13C NMR spectra. For other substrates,
large quantities of solid were produced and excessive line
broadening resulted. In order to reduce solid formation and
study these systems, the molarity of the n-BuLi in the tube
was reduced to 0.1 M. In these cases, we used 8£1025 mol
of n-BuLi in a total volume of 800 mL (with 200 mL of
toluene-d8, 40 mL of 2 M n-BuLi in cyclohexane and a
combination of substrate and cyclohexane of 560 mL).
8£1024 mol of substrate was utilized in these experiments
at 0.1 M in n-BuLi, maintaining a 1:1 molar ratio of
n-BuLi/substrate.
from the literature to ensure proper methodology and results
and were able to reproduce reported geometries and
energies to within 1% of relative reported values.13b
Acknowledgements
Support of this research by the Research Corporation,
Grants CC4216 and CC5027 and by NSF CHE 010143 are
greatly appreciated. P. Shelton was the recipient of a
Council on Undergraduate Research Academic-Industrial
Undergraduate Partnership Fellowship, Summer 1996.
Faculty Research Grants (WKU) and travel support
(WKU) have been helpful as have contributions of n-BuLi
from FMC, Lithium Division, Gastonia, NC. Chris
Harrison, Martha J. Watson, Michael Timmons and Wesley
Smith provided computational and experimental assistance.
The quantitation of the percent complex formed between
the substrate and n-BuLi as reported in Table 1 was
determined by the following procedure: the 200 mL of
toluene-d8 in each NMR tube served as both the deuterated
lock solvent and as an internal standard by which to
compare relative changes in peak height. By comparing the
combined integrations of assigned aromatic peaks within
the substrate to that for an internal standard of toluene-d8,
the percentage of solid formation was determined.
Specifically, the relative ratio of the signals for the
aromatic toluene-d8 and the aromatic peaks from the
substrate were determined for a standard solution which
consisted of substrate, solvent, internal standard (toluene-d8)
but no n-BuLi. The substrate cannot complex to anything
in these standard experiments and so the ratio of
aromatic signals for the substrate to the toluene-d8
serve as a benchmark for 0% solid formation (and zero
complexation). Then, in separate experiments, NMR
tubes containing substrate, solvent, internal standard
and n-BuLi were analyzed in a similar fashion. In
these experiments, the ratio of the aromatic signals were
measured at time equals 0, 1, and 3 h. The intensity of
the sample signals with respect to the toluene-d8
progressively decrease as a function of time. Next, the
0, 1 and 3 h ratios were divided by the standard ratio.
These numbers, expressed as percentages, represent the
amounts of substrate remaining in solution in both
complexed and uncomplexed form. Subtractions of these
respective percentages from 100 yields the percent solids
recorded in Table 1. Next, the integrations of the
uncomplexed and the complexed peak were measured
in the methoxy region of the NMR spectrum. The height
of the complexed peak was divided by the sum of the
complexed and uncomplexed peak heights. These
numbers, expressed as percentages, represent the amount
of complexed substrate remaining in solution. The total
complexed substrate, the solid putatively having passed
through the soluble complex state, is determined by
adding the percent solid to the normalized percent
complex remaining in solution.
References
1. Stanetty, P.; Mihavilovic, M. D. J. Org. Chem. 1997, 62,
1514–1515.
2. (a) Slocum, D. W.; Moon, R.; Thompson, J.; Coffey, D. S.; Li,
J. D.; Slocum, M. G.; Gayton-Garcia, R.; Siegel, A.
Tetrahedron Lett. 1994, 35, 385–388. (b) Slocum, D. W.;
Thompson, J.; Friesen, C. Tetrahedron Lett. 1995, 36,
8171–8174.
3. It is our contention that measured equivalents of THF in
hydrocarbon solvents constitute catalytic systems for metala-
tion. The metalations are greatly accelerated by the presence
of small quantities of THF. Moreover, bulk (neat) THF leads
to lesser extents of regiospecific metalation and the advent of
secondary metalations, cf. (a) Slocum, D. W.; Reed, D.;
Jackson., F., III; Friesen, C. J. Organomet. Chem. 1996, 512,
265–267. (b) Slocum, D. W.; Dietzel, P. Tetrahedron Lett.
1999, 40, 1823–1826.
4. Slocum, D. W.; Timmons, M.; Watson, A.; Rucker,
J. Abstracts 222nd National Meeting of the American
Chemical Society, Chicago, IL, 2001, Section ORGN, paper
117.
5. (a) These constitute our efforts to effect atom economical
metalations. cf. (a) Chemistry and Engineering News, June 19,
1995, p. 32. (b) Chemistry and Engineering News, May 5,
1997, p. 47.
6. Gschwend, H. R.; Rodriguez, H. R. Org. React. 1979, 26,
1–360.
7. No studies of 1,4-DMB are cited in Ref. 6. 1,4-DMB lies
outside the scope of this investigation.
8. Recent metalations of 1,2-DMB (a) Majetich, G.; Liu, S.
Synth. Commun. 1993, 23, 2331–2335. (b) Stevens, R. V.;
Bisacchi, G. S. J. Org. Chem. 1982, 47, 2393–2396. (c) Tani,
K.; Lukin, K.; Eaton, P. E. J. Am. Chem. Soc. 1997, 119,
1476–1477. (d) McMurry, J. E.; Farina, V.; Scott, W. J.;
Davidson, A. H.; Summers, D. R.; Ashok, S. J. Org. Chem.
1984, 49, 3803–3812. (e) Albrecht, M.; Frohlich, R. J. Am.
Chem. Soc. 1997, 119, 1656–1661. (f) Prevot-Halter, I.;
Smith, T. J.; Weiss, J. J. Org. Chem. 1997, 62, 2186–2192. (g)
Ng, G. P.; Dawson, C. R. J. Org. Chem. 1978, 43, 3205–3208.
(h) Nerdinger, S.; Kendall, C.; Marchhart, R.; Riebel, P.;
Johnson, M. R.; Yin, C.-F.; Eltis, L. D.; Snieckus, V. Chem.
Commun. 1999, 2259–2260. (i) Trauner, D.; Bats, J. W.;
4.3. Computational studies
All ab initio calculations were carried out by using the
Linux Spartane ’02 suite of programs running on a 1.9 GHz
AMD processor with 1 Gb of DDR RAM. Geometry
optimizations were performed in stepwise fashion from
MM2, to PM3 semi empirical, to Hartree–Fock 6-31G* and
then ultimately to density functional theory at the
B3LYP 6-311þG** level. We replicated structures for
alkyllithium complexes with nitrogen and oxygen species

8284
D. W. Slocum et al. / Tetrahedron 59 (2003) 8275–8284
Werner, A.; Mulzer, J. J. Org. Chem. 1998, 63, 5908–5918. (j)
16. Waldmuller, D.; Kotsatos, B. J.; Nichols, M. A.; Williard, P.
J. Am. Chem. Soc. 1997, 119, 5479–5480.
Marti, T.; Peterson, B. R.; Furer, A.; Mordasini-Denti, T.;
Zarske, J.; Jaun, B.; Diederich, F. Helv. Chem. Acta 1998, 81,
109–144. (l) Napolitano, E.; Fiashi, R. Tetrahedron Lett.
2000, 41, 4663–4666.
17. Collum, D. B.; Hoffman, D. J. Am. Chem. Soc. 1998, 120,
5810–5811.
18. Chadwick, S. T.; Rennels, R. A.; Rutherford, J. L.; Collum,
9. Metalations of 1,3-DMB in addition to those in Ref. 6. (a)
Bennetau, B.; Rajarison, F.; Dunogues, J.; Babin, P. Tetra-
hedron 1993, 47, 10843–10854. (b) Von den Bussche-
Hunnefeld, C.; Buhring, D.; Knobler, C. B.; Cram, D. J.
Chem. Commun. 1995, 1085–1087. (c) Starling, S. M.;
Vonwiller, S. C. Tetrahedron Lett. 1997, 38, 2159–2162.
10. Studies of the mono- and di-metalations of the TMB’s. (a)
Crowther, G. P.; Sundberg, R. J.; Sarpshkar, A. M. J. Org.
Chem. 1984, 49, 4657–4663. (b) Carreno, M. C.; Ruano, J. L.
G.; Toledo, M. A.; Urbano, A. Tetrahedron: Asymmetry 1997,
8, 913–921. (c) Schill, G.; Logemann, E. Chem. Ber. 1973,
106, 2910–2917. (d) Cabiddu, S.; Contini, L.; Fattuoni, C.;
Floris, C.; Gelli, G. Tetrahedron 1991, 47, 9279–9288. (e)
Andrus, M. B.; Meredith, E. L.; Soma Shekhar, B. B. V. Org.
Lett. 2001, 3, 259–262. (f) Maggi, R.; Schlosser, M.
Tetrahedron Lett. 1999, 40, 8797–8800.
D. B. J. Am. Chem. Soc. 2000, 122, 8640–8647.
˚
˚
˚
19. (a) (Li–Li¼2.37 A; Li–O¼2.16 A, Li–C¼2.22 A) in direct
comparison to the agreed upon (TMEDA)2. (n-BuLi)2
structure16 (Li–Li¼2.42 A; Li–N¼2.27 A; Li–C¼2.26 A).
(b) Attempts to computationally construct a 1,3-DMB analog
to structure 3b failed (at any level up to and including B3LYP
6-311þG**) in that the 1,3-DMB either was expelled from the
complex entirely, adopted a monodentate complexed form, or
the 1,3-DMB moieties adopted a bidentate complexation
to n-BuLi where the two heteroatoms in the dimethoxybenzene
are attached to a different lithium atom of the n-BuLi dimer
(Structure 3a). This very structure was initially computed by
Saa et al.20 to possess an energy minimum in what was termed
the tweezer effect. Alternative complexed forms of n-BuLi/1,3-
DMB are under current investigation.
˚
˚
˚
20. Saa, J. M.; Deya, P. M.; Suner, G. A.; Frontera, A. J. Am.
Chem. Soc. 1992, 114, 9093–9100.
11. Slocum, D. W.; Ray, J.; Shelton, P. Tetrahedron Lett. 2002,
43, 6071–6073, This reference constitutes contribution # 1
from the Laboratory of Covalent Chemistry at Western
Kentucky University.
21. Often, as with the DMB’s and the TMB’s, a second metalation
to form various amounts of a dilithio intermediate is difficult to
avoid. We suggest that a metalation that affords a single lithio
intermediate for such systems as one that is ‘ion-multiple
specific’.
12. (a) Bauer, W.; Schleyer, P. v. R. J. Am. Chem. Soc. 1989, 111,
7191–7198. (b) Beak, P.; Meyers, A. I. Acc. Chem. Res. 1986,
19, 356–363. (c) Roberts, J. D.; Curtin, D. Y. J. Am. Chem.
Soc. 1946, 68, 1658–1660.
22. For the other substrates elevation of the temperature of the
reaction media brought evidence of decomposition.
13. (a) van Eikema Hommes, N. J. R.; Schleyer, P. v. R. Angew.
Chem., Int. Ed. Engl. 1992, 31, 755–758. (b) van Eikema
Hommes, N. J. R.; Schleyer, P. v. R. Tetrahedron 1994, 50,
5903–5916. (c) Kremer, T.; Junge, M.; Schleyer, P. v. R.
Organometallics 1996, 15, 3345–3359.
23. cf., inter alia (a) Meyers, A. I.; Willemsen, J. Chem. Commun.
1977, 1573–1574. (b) Tanaka, T.; Wakayama, R.; Maeda, S.;
Mikamiyama, H.; Maezaki, N.; Ohno, H. Chem. Commun.
2000, 1287–1288. (c) Shindo, M.; Koga, K.; Tomioka, K.
J. Am. Chem. Soc. 1992, 114, 8732–8733. (d) Norman, D. P.
G.; Bunnell, A. E.; Stabler, S. R.; Flippin, L. A. J. Org. Chem.
1999, 64, 9301–9306. (e) Hutchings, R. H.; Meyers, A. I.
J. Org. Chem. 1996, 61, 1004–1013.
14. For a concise review of this topic, cf. Eischenbroich, Ch.;
Salzer, A. Organometallics: a Concise Introduction; VCH:
New York, 1992.
15. (a) Lewis, H. L.; Brown, T. L. J. Am. Chem. Soc. 1970, 92,
4664–4670. (b) McGarrity, J. F.; Ogle, C. A. J. Am. Chem.
Soc. 1985, 107, 1805–1810. (c) McGarrity, J. F.; Ogle, C. A.;
Birch, Z.; Loosli, H. R. J. Am. Chem. Soc. 1985, 107,
1810–1815.
24. The complete disappearance of all signals was only observed
for the 1,2,4-TMB system.
25. Betz, J.; Bauer, W. J. Am. Chem. Soc. 2002, 124, 8699–8706.