Meta-metallation of N,N-dimethylaniline: Contrasting direct sodium-mediated zincation with indirect sodiation-dialkylzinc co-complexation

Article abstract of DOI:10.3762/bjoc.7.144

Previously we reported that direct zincation of N,N-dimethylaniline by the mixed-metal zincate reagent 1 ((TMEDA)Na(TMP)(t-Bu)Zn(t-Bu)) surprisingly led to meta-metallation (zincation) of the aniline, as manifested in the crystalline complex 2 ((TMEDA)Na(

Full text of DOI:10.3762/bjoc.7.144

Meta-metallation of N,N-dimethylaniline: Contrasting
direct sodium-mediated zincation with indirect
sodiation-dialkylzinc co-complexation
David R. Armstrong, Liam Balloch*, Eva Hevia, Alan R. Kennedy,
Robert E. Mulvey*, Charles T. O'Hara and Stuart D. Robertson
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2011, 7, 1234–1248.
WestCHEM, Department of Pure and Applied Chemistry, University of
Strathclyde, Glasgow, G1 1XL, United Kingdom
doi:10.3762/bjoc.7.144
Received: 23 February 2011
Accepted: 06 June 2011
Email:
David R. Armstrong - d.r.armstrong@strath.ac.uk;
Liam Balloch* - liam.balloch@strath.ac.uk;
Robert E. Mulvey* - r.e.mulvey@strath.ac.uk
Published: 06 September 2011
This article is part of the Thematic Series "Directed aromatic
functionalization".
*
Corresponding author
Guest Editor: V. Snieckus
Keywords:
alkali metal; crystal structure; isomerisation; metallation; zincation
© 2011 Armstrong et al; licensee Beilstein-Institut.
License and terms: see end of document.
Abstract
Previously we reported that direct zincation of N,N-dimethylaniline by the mixed-metal zincate reagent 1 ((TMEDA)Na(TMP)(t-
Bu)Zn(t-Bu)) surprisingly led to meta-metallation (zincation) of the aniline, as manifested in the crystalline complex 2
(
(TMEDA)Na(TMP)(m-C6H4-NMe2)Zn(t-Bu)), and that iodination of these isolated crystals produced the meta-isomer N,N-
dimethyl-3-iodoaniline quantitatively. Completing the study here we find that treating the reaction solution with iodine produces a
2% conversion and results in a mixture of regioisomers of N,N-dimethyliodoaniline, with the meta-isomer still the major product
ortho:meta:para ratio, 6:73:21), as determined by NMR. In contrast to this bimetallic method, sodiation of N,N-dimethylaniline
7
(
with n-BuNa produced the dimeric, ortho-sodiated complex 3 (((TMEDA)Na(o-C6H4-NMe2))2), as characterised by X-ray crystal-
lography and NMR. No regioisomers were observed in the reaction solution. Introducing t-Bu2Zn to this reaction solution afforded
a cocrystalline product in the solid-state, composed of the bis-anilide 4 ((TMEDA)Na(o-C6H4-NMe2)2Zn(t-Bu)) and the Me2N–C
cleavage product 5 ({(TMEDA)2Na}+{(t-Bu2Zn)2(µ-NMe2)}−), which was characterised by X-ray crystallography. NMR studies of
the reaction mixture that produces 4 and 5 revealed one additional species, but the mixture as a whole contained only ortho-species
and a trace amount of para-species as established by iodine quenching. In an indirect variation of the bimetallic reaction, TMP(H)
was added at room temperature to the reaction mixture that afforded 4 and 5. This gave the crystalline product 6
(
(TMEDA)Na(TMP)(o-C6H4-NMe2)Zn(t-Bu)), the ortho-isomer of the meta-complex 2, as determined from X-ray crystallo-
graphic and NMR data. Monitoring the regioselectivity of the reaction by iodination revealed a 16.6:1.6:1.0 ortho:meta:para ratio.
Interestingly, when the TMP(H) containing solution was heated under reflux for 18 hours more meta-isomer was produced (corres-
ponding ratio 3.7:4.2:1.0). It is likely that this change has its origin in a retro reaction that produces the original base 1 as an inter-
1234

Beilstein J. Org. Chem. 2011, 7, 1234–1248.
mediate. Theoretical calculations at the DFT level using the B3LYP method and the 6-311G** basis set were used to probe the
energetics of both monometallic and bimetallic systems. In accord with the experimental results, it was found that ortho-metalla-
tion was favoured by sodiation; whereas meta- (closely followed by para-) metallation was favoured by direct sodium-mediated
zincation.
Introduction
While the metallation reaction remains an essential tool for sensitive functional groups necessitate the use of sub-ambient
constructing substituted aromatic compounds [1,2], the quest temperatures. These problems can be circumvented by the use
for new improved reagents capable of selectively abstracting of the commercially available, more selective zinc base
hydrogen from organic substrates continues. Routinely, organo- TMPZnCl·LiCl that executes chemo- and regioselective zinca-
lithium reagents have been employed for this purpose with the tion of aryl and heteroaryl substrates containing sensitive
high electropositivity of lithium affording polar, reactive groups typically at ambient temperature [11] and on the multi-
Cδ−–Liδ+ bonds proficient in metallating C–H bonds in organic, gram scale with metallation rates comparable to those obtained
especially aromatic and heteroaromatic, compounds [3-5]. for small scale processes [12]. Alkali metal zincates have also
Although, alkyllithium or lithium amide reagents are regarded been given consideration, principally two-component dialkyl-
as the preferred metallating agents for many transformations, amido zincates MZn(NR2)R′2. The reagent “(LiZn(TMP)(t-
there are major limitations associated with their use, with Bu)2)”, introduced by Kondo and Uchiyama, is capable of
prominent drawbacks including poor functional group toler- directly zincating a range of similar substrates [13,14]. The
ance and low stability of the developing lithio-intermediates sodium zincate 1 ((TMEDA)Na(TMP)(t-Bu)Zn(t-Bu);
necessitating the use of sub-ambient temperatures to effect the TMEDA: N,N,N′,N′-tetramethylethylenediamine), reported by
desired reactions and avoid decomposition or side reactions. In our group, is a potent and versatile zincating agent for a wide
recent years, as improvements on the existing metallating range of aromatic molecules usually inert towards orthodox
agents have been sought, research groups around the world have organozinc reagents including benzene [15] and naphthalene
investigated innovative bimetallic alternatives to the customary [16]. These studies – that have been structurally supported by
homometallic reagents.
X-ray crystallography in tandem with NMR spectroscopy –
have uncovered the chemical synergy that these mixed-metal
As part of these innovations the zinc–hydrogen exchange reac- alternatives can exhibit, which enabled such reagents to perform
tion has undergone a significant revolution in recent times special metallation reactions that cannot be reproduced by either
propelling it into the spotlight. Despite finding utility in various of the single-metal components that constitute the mixed-metal
types of organic reaction [6], simple zinc reagents (alkyls, reagent. While the alkali metal component is essential for the
amides) are kinetically sluggish bases and consequently ineffec- synergic metallation to follow its course, it is the less elec-
tual for deprotonative metallation [7]. Nonetheless, when part of tropositive zinc that replaces the departing hydrogen atom,
an intricate multi-component composition, a striking enhance- prompting metallations of this type to be best regarded as
ment in reactivity can be bestowed upon the zinc reagent as a “Alkali-Metal-Mediated Zincations (AMMZn)” [17].
result of cooperative effects between the different components.
The work of Knochel uncovered a special reactivity and selec- Directed ortho-metallation (DoM) is a special sub-category of
tivity that can be realised with a mixed lithium halide–magne- deprotonative metallation. Defined as a metallation (nearly
sium amide complex sometimes labelled a “turbo-Grignard” always lithiation) of an aromatic ring directed towards a pos-
reagent [8]. It has been proposed that LiCl breaks up the magne- ition adjacent (ortho) to an activating heteroatom-containing
sium amide aggregates allowing more soluble mixed-metal, functional group, it is widely regarded as the number one
mixed-anion reagents such as (TMPMgCl·LiCl) to perform methodology for constructing regiospecifically substituted
regioselective functionalisation of aromatic and heteroaromatic aromatic rings [18,19]. However, by channelling the special
compounds (TMP: 2,2,6,6-tetramethylpiperidide). Modified synergic chemistry of sodium TMP-zincates, it is possible to
from their magnesiating “turbo-Grignard” reagents, Knochel’s overcome DoM effects and direct metallation to more remote
three component system ((TMP)2Zn·2MgCl2·2LiCl) allows positions. Of particular interest to this work, we previously
direct zincation of functionalised arenes and heteroarenes reported that N,N-dimethylaniline can be monozincated by the
[
9,10]. However, common limitations have been noted for the base 1 at the normally inaccessible meta-position [20] (while
use of this reagent with some electron poor and heterocyclic complexation of N,N-dimethylaniline with Cr(CO)3 leads to a
compounds which suffer from the drawback of limited regiose- mixture of ortho-, meta- and para-deprotonation upon lithia-
lectivity, whilst several activated aromatic compounds bearing tion [21], other situations of meta-deprotonation require a
1235

Beilstein J. Org. Chem. 2011, 7, 1234–1248.
special combination of substituents on the aromatic ring. Such Recently, we have sought to identify the structures of the reac-
situations, which because of the multiple substitution limit the tion intermediates suggested by the theoretical studies. In addi-
number of sites available for deprotonation, are usually dictated tion to a closely related study for the AMMZn of anisole by the
by steric constraints and therefore should be clearly distin- analogous lithium TMP-zincate ((THF)Li(TMP)(t-Bu)Zn(t-Bu))
guished from monosubstituted aromatic rings in which meta- [27], these mechanistic and structural studies were then
deprotonation is far more difficult to attain [22]). Another extended to the metallation of trifluoromethylbenzene by the
example of a synergic metallation can be found with the alkyl bimetallic base 1 [28]. Providing a greater understanding of the
arene toluene. With conventional metallating reagents such as mechanisms involved, the reaction was structurally traced by
the alkyllithium BuLi·TMEDA, toluene is normally deproto- isolation of the kinetic solvent-separated ortho-deprotonated
nated at the methyl (lateral) site to generate the resonance product ({(TMEDA)2Na}+{Zn(o-C6H4-CF3)(t-Bu)2}−) before
stabilised benzyl “carbanion” (PhCH2−) [23,24]. In contrast, exploring its reactivity towards TMP(H) mirroring the second,
toluene has been directly zincated by reaction with the multipart step of the AMMZn process. Intriguingly, this second
heteroleptic sodium zincate 1 to afford a statistical mixture of step was found to affect strongly the conclusion, swaying not
the meta- and para-regioisomers of (TMEDA)Na(TMP)(C6H4- only the product yield but also the final regioselectivity of the
Me)Zn(t-Bu) with the methyl substituent remaining intact [25]. metallation with a complex mixture of ortho-, meta- and para-
regioisomeric products observed in solution. These findings
The structural insight offered by these findings has been provided the first experimental evidence for a two-step mecha-
augmented by a complementary theoretical study. Filling in the nism in deprotonative metallations with TMP-zincates but addi-
gaps between the crystallographically characterised starting ma- tionally pose the question; “is the unique meta-metallation of
terials and the thermodynamic final products, Uchiyama and N,N-dimethylaniline a consequence of TMP induced isomerisa-
Nobuto have studied and evaluated the possible reaction path- tion?” This study attempts to answer this important question.
ways of the deprotonative metallation of benzene with sodium
TMP-zincate 1 computationally [26]. Their DFT studies Herein we investigate in detail the reaction of 1 with N,N-
propose that these distinctive deprotonations proceed via a step- dimethylaniline, exploring the proposed two-step mechanism
wise mechanism, whereby zincate 1 serves kinetically as an and shedding light on the special synergic facet of the AMMZn
amide base (due to the greater kinetic lability of the Zn–N reaction. We report the synthesis and structural elucidation of
bonds) with concomitant release of TMP(H) with the formation ortho-metallated N,N-dimethylaniline complexes 3
of a bisalkyl(aryl) zincate intermediate (Scheme 1). In the (((TMEDA)Na(o-C6H4-NMe2))2) and 4 ((TMEDA)Na(o-C6H4-
second step, it follows that the co-existent, strongly acidic N–H NMe2)2Zn(t-Bu)) prepared by direct metallation with n-BuNa
bond of TMP(H) is deprotonated by this intermediate to afford followed by (in the case of 4) co-complexation with t-Bu2Zn,
the final thermodynamic product of the reaction identified by before surveying the latter’s reactivity towards TMP(H) repli-
the aforementioned X-ray crystallographic studies, and isobu- cating the second step of the stepwise procedure. Offering
tane.
increased understanding of the possible mechanisms involved in
Scheme 1: Proposed stepwise mechanism for the zincation of benzene.
1236

Beilstein J. Org. Chem. 2011, 7, 1234–1248.
the reactions of TMP-zincates, the solid-state structural features
of an isolated decomposition product 5 ({(TMEDA)2Na}+{(t-
Bu2Zn)2(µ-NMe2)}−) are also discussed, thus overall providing
an illustrative example of the subtle, yet prodigious reactivity of
the bimetallic base 1.
Results and Discussion
The original direct meta-zincation reaction. In comparison to
tertiary amides and O-carbamates, anilines are relatively modest
metallation directing groups [19,29]. Activation of N,N-
dimethylaniline transpires primarily from the acidifying effect
of the N atom; the coordination effect of the N atom is believed
to be less important as a result of the conjugation of its lone pair
with the π-system of the ring, even so N,N-dimethylaniline
undergoes DoM with phenyllithium in poor yield [30] or
n-butyllithium in good yield under forcing conditions [31].
Varying the metallation agent from these mainstream
homometallic species to heterometallic 1 remarkably switched
the orientation of the deprotonation to the meta-site, with the
metallated heterotrianionic product of the reaction 2
Figure 1: Molecular structure of 2 with selective atom labelling.
Hydrogen atoms and minor disorder components are omitted for clarity
(
(TMEDA)Na(TMP)(m-C6H4-NMe2)Zn(t-Bu)) isolated at
[20].
ambient temperature in a stable crystalline form which allowed
its molecular structure to be ascertained by X-ray crystallog-
raphy (Figure 1).
of this surprising synergic meta-deprotonation of N,N-dimethyl-
aniline.
As reported, a preliminary reaction of isolated crystals of 2 in
THF with I2 was carried out to establish whether these meta- The direct meta-zincation reaction: a closer look. Revisiting
zincated dimethylaniline complexes could be intercepted by the reaction, sodium TMP-zincate 1 was prepared in situ in
electrophiles. 1H NMR spectroscopic experiments revealed hexane solution and reacted with one molar equivalent of N,N-
quantitative iodination and the formation of N,N-dimethyl-3- dimethylaniline at room temperature to afford a yellow solution
iodoaniline. Now we are looking to paint the complete picture which was subsequently treated with iodine (Scheme 2). NMR
Scheme 2: Synergic metallation of N,N-dimethylaniline (A) with sodium TMP-zincate 1 to produce 2, which was subsequently quenched with I2 to
produce iodo-anilines.
1237

Beilstein J. Org. Chem. 2011, 7, 1234–1248.
spectroscopic analysis (1H and 1H–1H COSY NMR spectra in
deuterated benzene) of the crude reaction solution revealed a
complex mixture of iodinated products that contained all three
possible ring regioisomers for the monometallation of N,N-
dimethylaniline. In agreement with the X-ray crystallographic
analysis, the meta-isomer was the principal product, but addi-
tionally in the aromatic region two doublets at 7.42 and 6.13
ppm were observed for the next most abundant component, i.e.,
the para-product and finally four multiplets at 7.75, 6.99, 6.73,
and 6.46 ppm for the minor ortho-product in an overall 11.6:3:1
ratio.
The indirect sequential sodiation-dialkylzinc co-complexa-
tion approach. Closely imitating the constituents of sodium
zincate 1, N,N-dimethylaniline was ortho-sodiated by reaction
with n-BuNa in hexane at 0 °C to give the product as a light
orange precipitate in 64% yield. We employed n-BuNa because
in our hands NaTMP, with or without the addition of TMEDA,
was not sufficiently basic to abstract the ortho-hydrogen of
N,N-dimethylaniline. In an attempt to produce crystals of the
resulting sodium anilide, the red filtrate (after isolation of the
precipitate) was concentrated in vacuo to yield yellow/orange
crystals following storage in a refrigerator. X-ray crystallog-
raphy identified the product as the dimeric, ortho-sodiated 3
Figure 2: Molecular structure of 3 with selective atom labelling and
thermal ellipsoids drawn at the 50% probability level. Hydrogen atoms
are omitted for clarity. The long Na···Cipso and Na···N contacts are
highlighted by dashed lines. Symmetry operation to generate equiva-
lent atoms denoted A: 1−x, −y, 1−z. Selected bond distances (Å) and
angles (°): Na(1)–N(2) 2.5363(14), Na(1)–N(3) 2.5350(14), Na(1)–C(1)
2.5442(15), Na(1)–C(1A) 2.5903(16), Na(1)···C(6A) 2.9231(15),
Na(1)···N(1A) 2.6749(14), N(3)–Na(1)–N(2) 73.37(5), N(3)–Na(1)–C(1)
129.64(5), N(2)–Na(1)–C1) 104.31(5), N(3)–Na(1)–C(1A) 97.45(5),
(
((TMEDA)Na(o-C6H4-NMe2))2) (Figure 2).
N(2)–Na(1)–C(1A) 150.27(5), C(1)–Na(1)–C(1A) 103.12(4),
N(3)–Na(1)–N(1A) 106.68(4), N(2)–Na(1)–N(1A) 99.59(5),
C(1)–Na(1)–N(1A) 122.86(5).
Sodiated aniline 3 adopts a simple dimeric centrosymmetric
molecular structure based on a four membered (Na2(µ-C)2)
ring, which is strictly planar. Occupying a tetrahedral geometry For the purposes of NMR spectroscopic analysis, the optimum
mean angle around Na: 109.76°) the sodium atom in 3 is bound solubility of 3 was achieved in deuterated cyclohexane solution.
(
to a chelating TMEDA ligand and two deprotonated aniline The salient observation from the NMR spectra of the solid ma-
units, the latter primarily through a close contact with the two terial and indeed the mother liquor left following its isolation is
ortho-deprotonated carbon atoms of the aromatic ring that there is no sign of any other metallated products/isomers.
(
Na(1)−C(1) 2.5442(15); Na(1)–C(1A) 2.5903(16) Å). In addi- Therefore the ortho-sodiated N,N-dimethylaniline was the sole
tion to these primary contacts, sodium also engages in longer, metallated product detected during the course of the reaction.
weaker secondary contacts with both the ipso-carbon and 1H NMR spectroscopic analysis of a d12-cyclohexane solution
nitrogen atom of the anilide unit (Na(1)···C(6A) 2.9231(15) and of 3 revealed three multiplets at 7.78, 6.79 and 6.58 ppm repre-
Na(1)···N(1A) 2.6749(14) Å). The Na(1)–C(1A) distance is senting the ortho-deprotonated aromatic ring (overlap at 6.58
comparable to similar distances in the solvent-free sodium aryl ppm between the ortho′- and para-protons) with the NMe2
derivative (NaC6H3–2,6-Mes2)2 prepared by Power (mean singlet resonance of the aniline appearing at 2.83 ppm, a shift of
Na–C bond distance of 2.609 Å) [32], whilst they are naturally 0.12 ppm downfield from that in N,N-dimethylaniline.
longer than those in the lithium dimer ({1-(dimethylamino)-8- Completing the assignment, signals for the TMEDA ligand are
naphthyl}-lithium·THF)2 (2.207(4), 2.223(4) Å) [33]. In this found at 2.22 and 2.04 ppm. Emphasising the synthetically
latter compound the lithium atoms are located ca. 0.37 Å above limiting reactivity of 3, it appears to react with THF, while
and 1.05 Å below the naphthyl planes. By way of comparison, undesired deprotonation of toluene furnished crystals of the
the sodium atoms in 3 are positioned 0.6565(5) Å above and known compound TMEDA solvated benzylsodium as
1
.8248(5) Å below the aryl ring as a result of the increased size confirmed by NMR analysis [34]. Indeed, Slocum et al. have
of sodium; a trend also reflected in the distance between the recently developed a selective ortho-metallation protocol which
averaged main planes of the aromatic rings (1.176(1) Å in 3 and permits the retention of the bromine substituent in para-
a surprisingly small 0.68 Å between the naphthyl rings of the bromoanisole by the use of ortho-lithiodimethylaniline, which
lithium compound).
presumably must exhibit a reduced reactivity compared to its
1238

Beilstein J. Org. Chem. 2011, 7, 1234–1248.
sodium analogue 3, when employed as the metallating agent
35]. Indeed upon storage in a dry box, the shelf life of sodium
[
anilide 3 is no more than a couple of days.
Continuing with the experimental studies, to introduce the
second metal component one molar equivalent of t-Bu2Zn
(
dissolved in hexane solution) was added to a suspension of 3 in
hexane, which was prepared in situ, and the mixture was stirred
at 0 °C for 1 hour and then for a further 2 hours at room
temperature (Scheme 3). Following gentle heating of the mix-
ture, colourless crystals were deposited upon slow cooling of
the Schlenk tube in a Dewar flask of hot water overnight. X-ray
crystallography indicated that this was a 2:1 co-crystalline mix-
ture of two products, 4 ((TMEDA)Na(o-C6H4-NMe2)2Zn(t-
Bu)) and 5 ({(TMEDA)2Na}+{(t-Bu2Zn)2(µ-NMe2)}−).
Figure 3: Molecular structure of 4 with selective atom labelling and
thermal ellipsoids drawn at the 50% probability level. Hydrogen atoms
and disordered component of TMEDA are omitted for clarity. Second-
ary contacts between sodium and the anilide rings are denoted by
dashed lines. Only one of the crystallographically independent mole-
cules is displayed, the parameters for the other are the same within
experimental error. Selected bond distances (Å) and angles (°):
Zn(1)–C(13) 2.025(3), Zn(1)–C(23) 2.042(3), Na(1)–N(1) 2.663(3),
Na(1)···C(12) 2.848(3), Na(1)···C(13) 2.799(3), Na(1)···N(2) 2.851(2),
Na(1)···C(22) 2.884(3), Na(1)–C(23) 2.663(3), Na(1)–N(3) 2.473(2),
Na(1)–N(4) 2.473(3), C(13)–Zn(1)–C(23) 123.30(11), N(3)–Na(1)–N(4)
Bisaryl 4 (Figure 3) can be considered a mixed-metal product of
the co-complexation reaction between 3 and t-Bu2Zn, despite
the incorporation of a second deprotonated organic substrate.
Although, a tert-butyl group has been formally lost, the forma-
tion of 4 may not necessarily be a result of a second deprotona-
tion, with a disproportionation pathway potentially at play. The
molecular structure of 4 can be described as a contacted ion-pair
with the anionic moiety containing a trigonal planar zinc centre
bonded to three carbon atoms, two from the ortho-metallated
aniline fragments and one from the tert-butyl group. In addition
to the nitrogen atoms of the chelating TMEDA, the sodium
74.20(9), N(4)–Na(1)–C(23) 112.74(10), N(3)–Na(1)–C(23) 147.52(9),
N(4)–Na(1)–N(1) 102.85(8), N(3)–Na(1)–N(1) 114.85(9),
C(23)–Na(1)–N(1) 95.05(9).
atoms primary coordination sphere is completed by bonding to distorted tetrahedral geometry (mean angle around Na:
the nitrogen of one of the anilide units (Na(1)–N(1) 2.663(3) Å) 107.87°). Furthermore, the multihapto-sodium centre engages in
and the ortho-deprotonated carbon of the second aromatic ring weaker, secondary contacts to the ipso-carbons of both anilides
(
Na(1)–C(23) 2.663(3) Å), resulting in sodium adopting a (Na(1)···C(12) 2.848(3) and Na(1)···C(22) 2.884(3) Å) and addi-
Scheme 3: Indirect zincation of N,N-dimethylaniline producing 4, 5 and 6, which was then quenched with I2 to produce iodo-anilines.
1239

Beilstein J. Org. Chem. 2011, 7, 1234–1248.
tionally, interacts with not only the second deprotonated carbon potentially leading to the presence of this cleaved amide frag-
(
(
Na(1)···C(13) 2.799(3) Å) but also the second aniline N atom ment [39]. The zinc-nitrogen bond distances in 5 (Zn(3)–N(13)
a long Na(1)···N(2) length of 2.851(2) Å compared to the 2.059(2) and Zn(4)–N(13) 2.062(2) Å) are the same within
Na–NTMEDA bond lengths of Na(1)–N(3) 2.473(2) and experimental error to each other and in good agreement with
Na(1)–N(4) 2.473(3) Å). Notably, the carbanion which was related distances in the zinc bisamide ({(PhCH2)2N}Zn)2·C6H6
initially bound to sodium is now directly attached to the more (2.0083(13) and 2.0620(12) Å) [40] and in dimeric
carbophilic zinc as illustrated by the short, strong zinc–carbon methyl(diphenylamido)zinc (average value of 2.072(8) Å for
σ-bonds (Zn(1)–C(13) 2.025(3), Zn(1)–C(23) 2.042(3)). A the four distinct Zn–N bond lengths) [41]. The best comparison
search of the Cambridge Structural Database [36] found that the for 5 is possibly provided by the dinuclear zinc guanidinate
bisaryl model has precedent in alkali-metal zincate chemistry, complex ({Zn(OAr)}2(µ-{Me2NC(Ni-Pr)2})(µ-NMe2))
for example, the previously prepared biscarboxamide prepared by Coles [42]. In both complexes, the zinc atoms
(
(TMEDA)Li{2-(1-C(O)N(i-Pr)2)C6H4}2Zn(t-Bu)) [37], adopt a distorted trigonal planar geometry (sum of bond angles
however the donor ability of the different directing groups at Zn(3) and Zn(4) in 5 equal to 358.31° and 358.28°, respect-
limits comparison with 4. A salient feature of this structure is ively) but with each zinc atom bound to a bulky aryloxide and
the bond between lithium and the Lewis basic carbonyl oxygen guanidinate ligand, the Zn–(µ-amidoN)–Zn bond angle is
atoms which close a ten-membered ring, resulting in a compressed to 97.34°, in contrast to 103.67(9)° in the more
contracted Ph–Zn–Ph bond angle of 111.76(10)°. In contrast, as open structure of 5. A possible consequence of this increased
a result of the dimethylamino groups modest directing ability, steric hindrance in the guanidinate complex is the shorter
with coordination less important, the sodium atom in 4 engages zinc–nitrogen bond distances (1.990(3) and 1.982(3) Å) in com-
the π-system of the arene ring, in turn increasing the parison to those in 5.
C(13)–Zn(1)–C(23) bond angle to 123.30(11)°.
The 1H NMR spectrum in deuterated benzene solution of the
T h e s o l v e n t - s e p a r a t e d i o n - p a i r c o - p r o d u c t
5
reaction of 3 and t-Bu2Zn, which produced crystals of 4 and 5,
(
{(TMEDA)2Na}+{(t-Bu2Zn)2(µ-NMe2)}−) (Figure 4) adopts a revealed the presence of a third distinct species. Four multiplets
markedly different structural motif from that of 4. The cation at 8.03, 7.17, 7.06, and 6.72 ppm are observed for the major
comprises a nearly square planar sodium centre coordinated by ortho-metallated product but additionally, shadowing these
two TMEDA molecules (near square planar alkali-metals are signals, another different ortho-metallated product is evident
rare) [38], whereas the anion consists of two di-tert-butylzinc from four multiplets at 7.87, 7.01, 6.86, and 6.60 ppm. Poten-
units bridged by a dimethylamino group. The unexpected mani- tially, this additional unidentified product could be the related
festation of this bridging amide is almost certainly a conse- dialkyl(aryl) sodium zincate ((TMEDA)Na(o-C6H4-NMe2)t-
quence of utilising the robust homometallic reagent n-BuNa at BuZnt-Bu), reminiscent of the lithium congeners ((THF)3Li(o-
fairly ambient temperatures, with decomposition of either N,N- C6H4-OMe)t-BuZnt-Bu) and ((PMDETA)Li(o-C6H4-OMe)t-
dimethylaniline or with precedent in the literature, TMEDA BuZnt-Bu) prepared by co-complexation reactions of lithiated
anisole with t-Bu2Zn and the relevant Lewis base [27]. The
metallation regioselectivity exhibited in the crystalline material
is confirmed by treatment of the in situ reaction mixture with
iodine. Electrophilic quenching of the two different ortho-
deprotonated organometallic compounds is depicted in the
1
H NMR of the crude residue, with a 20.2:1.0 mixture of the
ortho- and para-isomers of iodo-N,N-dimethylaniline obtained
and significantly, a negligible quantity of the meta-iodinated
product present.
Figure 4: Solvent-separated ion-pair structure of 5 with selective atom
labelling and thermal ellipsoids drawn at the 50% probability level.
Hydrogen atoms and disordered component of TMEDA are omitted for
clarity. Only one of the crystallographically independent cations is
displayed, the parameters for the other are the same within experi-
mental error. Symmetry operations to generate equivalent atoms
denoted A: 1−x, 1−y, −z. Selected bond distances (Å) and angles (°):
Zn(3)–N(13) 2.059(2), Zn(4)–N(13) 2.062(2), Zn(3)–N(13)–Zn(4)
According to theoretical studies, compound 4 and its related
bisalkyl(aryl) derivative could be putative intermediates in the
first step of the metallation of N,N-dimethylaniline by TMP-
zincates. Thus, the reactivity of this in situ mixture towards the
amine TMP(H) was studied, purposely for the formation of
(
TMEDA)Na(TMP)(C6H4-NMe2)Zn(t-Bu) products. First the
103.67(9), Na(4)–N(11) 2.516(2), Na(4)–N(12) 2.494(3),
N(11)–Na(4)–N(12) 75.21(8), N(11)–Na(4)–N(11A) 180.0,
N(11)–Na(3)–N(12A) 104.79(8), N(12)–Na(3)–N(12A) 180.0.
reactivity towards 1 molar equivalent of TMP(H) at room
temperature was investigated and within 30 minutes a white
1240

Beilstein J. Org. Chem. 2011, 7, 1234–1248.
solid precipitated which could be recrystallised from hexane to
afford a small amount of translucent crystals.
An X-ray crystallographic study showed the molecular struc-
ture of these crystals to be 6 ((TMEDA)Na(TMP)(o-C6H4-
NMe2)Zn(t-Bu)) (Figure 5). Revealing a contact ion-pair struc-
ture, 6 comprises of the Na–TMP–Zn structural backbone of
bimetallic base 1 where TMEDA chelates the sodium atom and
zinc is bound to a terminal tert-butyl group. In keeping with
previous illustrations of the AMMZn reaction, zinc has filled
the position vacated by the departing hydrogen atom and
significantly in 6, zinc resides in the ortho-position of the
aromatic ring. Comparing the regioisomers 2 and 6, a strong
Zn–C σ-bond is formed between the metal and deprotonated
carbon atom in both compounds (2.035(4) and 2.079(2) Å in 2
and 6, respectively), as a result the main difference between the
structures arises from the manner in which the metallated arene
interacts with the alkali metal. In 2, with the NMe2 unit
detached from sodium, the alkali metal interacts more with the
π-system of the aromatic ring, principally with the deproto-
nated meta-carbon (2.691(4) Å) generating a four-membered
Figure 5: Molecular structure of 6 with selective atom labelling and
thermal ellipsoids drawn at the 50% probability level. Hydrogen atoms
are omitted for clarity. Secondary contacts between sodium and the
anilide rings are denoted by dashed lines. Selected bond distances (Å)
and angles (°): Na(1)–N(1) 2.586(2), Na(1)–N(2) 2.568(2), Na(1)–N(3)
2.6991(19), Na(1)–N(4) 2.469(2), Na(1)···C(32) 2.917(2), Na(1)···C(37)
(
NaNZnC) ring. In contrast, for 6 the close proximity of the
2
.768(2), Zn(1)–C(37) 2.079(2), Zn(1)–N(4) 2.0279(17),
NMe2 group enables it to bind with the sodium centre giving a
slightly longer than average Na(1)–N(3) bond distance of
N(1)–Na(1)–N(4) 131.05(7), N(2)–Na(1)–N(4) 122.82(7),
N(1)–Na(1)–N(2) 72.56(6), N(3)–Na(1)–N(4) 105.70(6),
N(2)–Na(1)–N(3) 109.31(6), N(1)–Na(1)–N(3) 111.60(7), N(4)-
Zn(1)–C37) 114.27(8).
2
2
2
.6991(19) Å (compared to a Na(1)–N(1) bond distance of
.663(3) Å in 4 and Na–N bond distances in the range
.448(4)–2.621(3) Å for a series of mononuclear zinc com-
pounds) [36], which closes a larger six-membered highly puck- tion of an in situ reaction mixture, it was treated with iodine, the
ered (NaNZnCCN) central ring. Nevertheless, sodium still crude 1H NMR spectrum (Figure 6) revealing that while much
π-engages with the aromatic ring (transannular distances of of the ortho- and para-substitution pattern is retained, a small
2
.917(2) and 2.768(2) Å) in the characteristic fashion of several portion of these zincated aryl molecules have isomerised to the
AMMZn products.
meta-position (overall ortho:meta:para ratio of 16.6:1.6:1.0
from an ortho:meta:para ratio of 20.2:0:1.0 before TMP(H) ad-
The relatively simple 1H NMR spectrum of the white precipi- dition). For the addition of TMP(H) to the isolated, ortho-metal-
tate in deuterated THF solution revealed the solid to be exclu- lated kinetic intermediate ({(TMEDA)2Na}+{Zn(o-
sively product 6, whereby N,N-dimethylaniline has been ortho- C6H4–CF3)(t-Bu)2}−), we previously reported that the second-
metallated with only four resonances, two doublets (7.48 and ary amine facilitates rearrangement of some ortho-zincated aryl
6
.81 ppm) and two triplets (6.92 and 6.75 ppm) observed in the molecules to meta- and para-isomers (in a 4.3:2.6:1 ratio,
aromatic region. Furthermore, the analysis of the aliphatic resulting from the amination reaction of TMP(H) with one t-Bu
region of both the 1H and 13C{1H} NMR spectra showed reso- group) [28]. The smaller degree of isomerisation in the N,N-
nances for both tert-butyl and TMEDA but notably, resonances dimethylaniline system can be rationalised at least in part by
which can be attributed to coordinated TMP at 1.74, 1.37 and considering the electronic nature of the different directing
1
.20 ppm were observed (note that 1H NMR resonances of groups. With respect to NMe2, CF3 exerts an increased induc-
TMP(H) in d8-THF appear at 1.63, 1.29 and 1.06 ppm).
tive effect thus weakening the Zn–Cortho bond to a greater
extent, potentially leading to an easier cleavage and thus
Regenerating the TMP bridge between sodium and zinc can be isomerisation. However, stimulated by this result, the TMP
considered a copy of the second step of Uchiyama and Nobuto’s induced isomerisation was pursued with the aim of trans-
theoretically proposed stepwise mechanism. However, should forming more ortho-material to meta. In order to meet this aspi-
the AMMZn reaction proceed by the two-step procedure, then ration, the reaction mixture was refluxed overnight, aiding the
the meta-isomer might emerge as the predominant product solubility of 6 and thus presenting greater opportunity for
despite the isolation of 6. To establish the complete constitu- regioisomerisation.
1241

Beilstein J. Org. Chem. 2011, 7, 1234–1248.
Figure 6: Aromatic region of 1H NMR spectra for deuterated benzene solutions of (a) the crude mixture obtained from the reaction of 1
(TMEDA)Na(TMP)(t-Bu)Zn(t-Bu)) with 1 equivalent of N,N-dimethylaniline at room temperature following iodine quenching; (b) the crude mixture
(
obtained from the reaction of BuNa·TMEDA, N,N-dimethylaniline and t-Bu2Zn at room temperature following iodine quenching; (c) the crude mixture
obtained from the reaction of BuNa·TMEDA, N,N-dimethylaniline, t-Bu2Zn and TMP(H) at room temperature following iodine quenching; (d) the crude
mixture obtained from the reaction of BuNa·TMEDA, N,N-dimethylaniline, t-Bu2Zn and TMP(H) following an overnight reflux and iodine quenching; (e)
a standard of N,N-dimethylaniline.
Following the addition of TMP(H) and an 18 hour reflux, the TMP(H) and monitoring by 1H NMR spectroscopy – high-
yellow solution was treated with iodine. 1H NMR spectro- lighted the significance of this second step in determining not
scopic analysis revealed that the make-up of the solution had only the final regioselectivity of the metallation but also the
dramatically changed, with each of the three regioisomers for reaction yield with the amine found to partially target the metal-
the mono-iodination of N,N-dimethylaniline present. Notably, lated aryl anion regenerating sodium TMP-zincate 1 and trifluo-
the predominant isomer of iodo-N,N-dimethylaniline is now the romethylbenzene. The existence of such a parallel competitive
meta-substituted product denoted by three multiplets at 7.05, reaction path could promote the re-formation of N,N-dimethyl-
6
.70, and 6.34 and a singlet at 6.98 ppm (Figure 6). The ortho- aniline and when in the presence of concomitant zincate 1, be
product is still well represented denoted by four multiplets at responsible for the increase observed in the meta-deprotonated
7
6
.75, 7.02, 6.74, and 6.46 ppm, while two doublets at 7.42 and substrate. Further evidence for the regeneration of bimetallic
.13 ppm denote the minor para-product with the three regio- base 1 comes in the identity of two unknown products in the
isomers present in an overall ortho:meta:para 3.7:4.2:1.0 ratio. crude mixture, although evidently N,N-dimethylaniline deriva-
However, the spectrum also revealed a substantial amount of tives, positive identification from NMR analysis was made
free N,N-dimethylaniline (multiplets at 7.22, 6.77, and 6.60 ppm difficult by the complexity of the aromatic region (two distinct
and a singlet at 2.55 ppm for the NMe2 group). Simulating the singlets visible in NMe2 region at 2.35 and 2.29 ppm). GC-MS
multipart-step of the AMMZn reaction in the aforementioned (CI mode) analysis verified the presence of the three regio-
trifluoromethyl benzene study [28] – by reacting isolated crys- isomers of iodo-N,N-dimethylaniline but also shed light on the
tals of ({(TMEDA)2Na}+{Zn(C6H4-CF3)(t-Bu)2}−) with identity of the two unknown species with the MH+ peaks at m/z
1242

Beilstein J. Org. Chem. 2011, 7, 1234–1248.
3
73.8 consistent with regioisomers of diiodo-N,N-dimethylani- On modelling the introduction of TMEDA, the ortho-isomer
line. Thus following the addition of TMP(H), the presence of remains the energy minimum structure at 0.00 kcal mol−1,
sodium TMP-zincate 1 in conjunction with the harsher reflux however, the energy gap to the other three regioisomers
conditions is likely to be the cause for a degree of di-deprotona- decreases slightly (methyl at 1.35 kcal mol−1; meta at 6.93 kcal
tion of the aromatic molecule.
mol−1; and para at 7.15 kcal mol−1). Inspecting the dimensions
of the ortho-product, TMEDA chelation of the Na atom results
in a longer and thus weaker Na–NMe2 secondary interaction
DFT calculations
In an attempt to rationalise the different metallation regioselec- (increase from 2.289 to 2.501 Å), while the meta- and para-
tivities afforded by the homo and heterobimetallic bases, theo- isomers benefit from TMEDA’s participation with the energies
retical calculations at the DFT level using the B3LYP method of TMEDA coordination equal to −20.86 and −20.74 kcal
and the 6-311G** basis set were employed to compute the rela- mol−1, respectively. Tellingly, although the overall energy of
tive stabilities of the four possible regioisomers for the reaction for these isomers is exothermic by −5.81 and −5.59
monometallation of N,N-dimethylaniline.
kcal mol−1, by way of comparison the product of ortho-deproto-
nation is significantly more exothermic by −12.74 kcal mol−1.
The relative stabilities of the four model regioisomers (ortho-,
meta-, para- and methyl-positions), in which the aniline ring is With respect to dimerisation the most disfavoured is the methyl
sodium substituted were calculated and in support of the experi- isomer (relative energy of 8.60 kcal mol−1) but in contrast the
mental studies, the ortho-isomer is the energetically preferred meta (at 4.65 kcal mol−1) and para (at 5.48 kcal mol−1) regio-
product (relative energy: 0.00 kcal mol−1) with the presence of isomers benefit in relation to the ortho-product (Figure 7). The
a Na–N interaction helping its stability. Closely following is the −22.08 kcal mol−1 energy of dimerisation gained by the ortho-
methyl isomer (at 1.61 kcal mol−1) but destabilised are the isomer is relatively low (compared to −31.29 kcal mol−1 for the
meta- and para-isomers (by 8.60 and 8.71 kcal mol−1, respect- meta-isomer) due to a combination of increased steric repul-
ively).
sions and loss of stabilising interactions. On dimerisation, the
Figure 7: Relative energy sequence of the four theoretical regioisomers of the experimentally observed product 3.
1243

Beilstein J. Org. Chem. 2011, 7, 1234–1248.
Na–N interaction in the ortho-product increases in length from product observed experimentally. A comparison of the dimen-
2
.501 to 2.662 Å with the Na atom now involved in bonding sions in these regioisomers, reveal that the Na contacts to the
with two ortho-deprotonated C atoms, while there is also a aryl C atoms are systematically shorter in the energetically
noticeable increase in the Na–N(TMEDA) bond lengths from preferred meta-isomer 2A (bond distances: 2.594, 3.070, 3.212,
2
.521 and 2.533Å to 2.604 to 2.674 Å. Through these model- 3.959, 4.097, 4.403 Å) than in the analogous para-product 7
ling studies, ortho-deprotonation and dimerisation is found to (bond distances: 2.596, 3.178, 3.239, 4.135, 4.185, 4.596 Å),
be the energetically preferred metallation pathway by the whereas the Zn–C(aryl) bonds are within 0.001 Å of each other
homometallic reagent, the overall reaction found to be (2.084 and 2.083 Å, respectively) (Table 1). In the ortho-isomer
exothermic by −47.55 kcal mol−1. Incidentally, on addition of 6A, the product found previously to be expected from conven-
t-Bu2Zn to each of the three monomeric sodiated regioisomers tional (non-synergic, single metal) metallation, the Na–C(ortho)
there is very little difference in the energies of the three prod- bond is decidedly longer (2.747 Å) and thus, weaker than the
ucts (ortho at 0.00 kcal mol−1; meta at 0.32 kcal mol−1; and Na–C(meta) bond in the favoured meta-isomer 2A as a result of
para at 0.41 kcal mol−1).
increased steric constraints. Additionally, with the sodium atom
bonding to the NMe2 group in 6A, lone pair–lone pair repul-
Moving to calculations on the sodium TMP-zincate 1, concur- sion between the lone pair of the nitrogen and the nearby ortho-
ring with the experimental findings and in sharp contrast to the carbanion may be a factor in the relative instability of 6A.
product obtained via monometallic n-BuNa/TMEDA, the meta- Consequently, the principal distinction lies in maximising the
isomer 2A is found to be the minimum energy structure (rela- strength of the Na···Cπ contacts, and although they may be
tive energy: 0.00 kcal mol−1) but is closely followed by the weak individually, collectively they must contribute appre-
para-isomer 7 (at 0.66 kcal mol−1). Notably, the ortho-isomer ciably to the overall stability of the unexpected meta-deproto-
6
A is destabilised further (at 4.32 kcal mol−1); and the least nated product.
stable of all is metallation at the methyl group (at 9.01 kcal
mol−1). Hence in the synergic twofold metal system, the meta- Conclusion
and para-regioisomers are the thermodynamic products. Model- Closing remarks on direct AMMZn versus indirect sodia-
ling studies reveal all four deprotonations to be exothermic, tion-zinc co-complexation approaches. The results of this
with the meta-isomer most exothermic by −21.46 kcal mol−1. study highlight the inherent complexity of these metallation
reactions even at the metal stage, with a diverse array of
The modest difference in the relative energies for 2A and 7 organometallic complexes isolated, prior to the recognised intri-
could explain the formation of the sizeable proportion of para- cacies of subsequent electrophilic interception.
Table 1: Selected bond distances calculated for the three regioisomers 2A, 6A and 7.
Bond distance (Å)
2A
6A
7
Zn–Cmetallated
Na–Cmetallated
Na···Caryl
2.084
2.594
2.107
2.747
2.937
2.083
2.596
3.070, 3.212
3.178, 3.239
1244

Beilstein J. Org. Chem. 2011, 7, 1234–1248.
Collectively, these results emphasise the synthetic utility of the by the synergic base, and the harsh reflux conditions required to
AMMZn reaction with the employment of a synergic bimetallic bring about the presence of the meta-zincated derivative offers
base providing entry to previously inaccessible metallation less control, ultimately leading to competing di-metallation of
regioselectivities. The AMMZn of N,N-dimethylaniline by the aromatic substrate.
sodium zincate 1 ((TMEDA)Na(TMP)(t-Bu)Zn(t-Bu)) has been
shown to afford a mixture of ortho-, meta- and para-regio- In addition DFT studies probing the different metallation
isomers in solution, however, the predominant meta-substituted regioselectivities obtained from homometallic and hetero-
derivative 2 ((TMEDA)Na(TMP)(m-C6H4-NMe2)Zn(t-Bu)) – a bimetallic bases highlight the importance of the interactions
regioselectivity normally closed to synthetic aromatic chem- between the alkali metal and the aromatic substrate. With the
istry – can be isolated in reasonable yields in a clean, pure crys- conventional monometallic reagent n-BuNa/TMEDA, metalla-
talline form.
tion is directed to the ortho-position by the secondary inter-
action between Na and the N atom of the NMe2 group. In
In an attempt to try and advance our knowledge of the synergic contrast, in the reaction of 1 with N,N-dimethylaniline, the
facet of the AMMZn, some control reactions with N,N- meta-isomer is energetically the most preferred and analysis of
dimethylaniline were performed. With the toothless base the dimensions of the model structures suggest that the strength
t-Bu2Zn, incapable of directly zincating aromatics, N,N- of the combined Na···Cπ contacts contribute at least in part to its
dimethylaniline was treated with n-BuNa·TMEDA to afford stability. This favourable enthalpic effect, along with the com-
dimeric 3 (((TMEDA)Na(o-C6H4-NMe2))2) , which in keeping munication between the metals provided by the TMP bridge
with conventional metallation chemistry exhibits ortho-deproto- play a defining role in offering the unique regioselectivity
nated anilines. The instability of 3 at ambient temperatures is observed in the AMMZn of N,N-dimethylaniline.
reflected in its deprotonation of toluene under mild conditions
and following co-complexation with the dialkyl zinc, a mixture Finally, looking at the bigger picture, this study drives home the
of two products were isolated from the reaction, namely 4 fact that bimetallic reagents can offer chemistry distinct to that
(
(
(TMEDA)Na(TMP)(o-C6H4-NMe2)2Zn(t-Bu)) and 5 when two monometallic reagents are added sequentially to the
{(TMEDA)2Na}+{(t-Bu2Zn)2(µ-NMe2)}−). The ion-pair pro- same substrate. The secret to unlocking special synergic reactiv-
duct 5 contains a cleaved NMe2 group, exemplifying nicely ity may be found in coupling both metals within the same
how the reactivity of an alkali-metal can be attenuated when bimetallic reagent. We plan to capitalise on this advantage in
incorporated in a mixed-metal system but to a point where it is ongoing studies, investigating the generality of the concept.
still able to play its part in conducting synthetically useful
Experimental
smooth metallations.
General
The final component of bimetallic base 1, TMP(H) was intro- All reactions were performed under a protective argon atmos-
duced at this stage, in what could be considered a copy of the phere using standard Schlenk techniques. Hexane, THF and
second, multipart-step of the AMMZn process and led to the toluene purchased from Sigma-Aldrich, were dried by heating
isolation of 6 ((TMEDA)Na(TMP)(o-C6H4-NMe2)Zn(t-Bu)), a to reflux over sodium benzophenone ketyl and distilled under
regioisomer of 2 where the aromatic ring remains ortho- nitrogen prior to use. n-BuNa [43], t-Bu2Zn [15] and subse-
zincated. In search of the elusive meta-deprotonated derivative quently 1 ((TMEDA)Na(TMP)(t-Bu)Zn(t-Bu)) [15] were
2
, the reaction mixture was refluxed overnight and as a result of prepared according to literature procedures. The 1H NMR spec-
the harsher conditions and potentially, regeneration of sodium troscopic experiments were performed on a Bruker DPX400
zincate 1, the mono-deprotonated meta-derivative emerged as spectrometer at an operating frequency of 400.13 MHz. The 13C
the predominant product.
NMR spectra were obtained on the same instrument at an oper-
ating frequency of 100.62 MHz. Elemental analyses were
Significantly, these results underline the importance of the acid/ carried out on a Perkin-Elmer 2400 elemental analyser. Due to
base character of TMP(H)/TMP− and the synergic bridge it the extreme air and moisture sensitivity of 3, 4/5 and 6, ideal
provides between the two metals, for only when it is present, analyses could not be obtained. GC–MS analysis was
does the unique meta-regioselectivity materialise. Although this performed on an Agilent 7890A apparatus, GC with 5975C
product has been engineered from the homometallic compo- triple axes detector, run under CI mode (methane).
nents, there are pitfalls in this indirect approach. The low
kinetic stability of the ortho-sodiated intermediate leads to X-ray crystallography
undesired side reactions such as deprotonation of solvent and All data were collected at 123(2) K on an Oxford Diffraction
decomposition of co-present reactants at temperatures tolerated Gemini S Diffractometer with Mo Kα radiation (λ = 0.71073
1245

Beilstein J. Org. Chem. 2011, 7, 1234–1248.
Å). Structures were solved using SHELXS-97 [44] while refine- (m, 2H, Hpara and Hortho′), 2.83 (s, 6H, N(CH3)2), 2.22 (s, 4H,
ments were carried out on F2 against all independent reflec- CH2 (TMEDA)), 2.04 (s, 12H, CH3 (TMEDA)); 13C NMR
tions by the full-matrix least-squares method using the (100.62 MHz, d12-cyclohexane, 300 K) δ 182.3 (Na–Cortho),
SHELXL-97 program [44]. With the exception of the carbon 164.1 (Cipso), 142.7 (Cmeta), 124.6 (Cmeta′), 120.2 (Cpara), 111.5
atoms of the disordered components of TMEDA present in 4 (Cortho′), 58.6 (CH2 (TMEDA)), 46.0 (CH3 (TMEDA)), 45.0
and 5 all non-hydrogen atoms were refined using anisotropic (N(CH3)2).
thermal parameters. CCDC 813771 (3), 813772 (4/5) and
8
13773 (6) contain the full supplementary crystallographic data Synthesis of 4 ((TMEDA)Na(TMP)(C6H4-NMe2)2Zn(t-Bu))
for this paper. These data can be obtained free of charge from and 5 ({(TMEDA)2Na}+{(t-Bu2Zn)2(µ-NMe2)}−): A hexane
The Cambridge Crystallographic Data Centre via http:// solution of t-Bu2Zn (2 mmol, 0.36 g) was added via cannula to
www.ccdc.cam.ac.uk/data_request/cif.
a suspension of 3, prepared as described above. The mixture
was allowed to stir for 1 hour at 0 °C, and then slowly warmed
Crystal data for 3: C28H52N6Na2, M = 518.74, monoclinic, to room temperature and stirred for a further two hours at this
P21/c, a = 8.8060(9), b = 17.5089(12), c = 11.1635(11) Å, temperature. The suspension was gently heated for a couple of
β = 111.693(12)°, V = 1599.3(3) Å3, Z = 2. 8277 reflections minutes to obtain (as close as possible) a yellow homogeneous
collected, 3797 were unique, Rint = 0.0454, R = 0.0471, solution: The latter was left in a Dewar flask filled with hot
Rw = 0.1133, GOF = 0.905, 169 refined parameters, water overnight to afford 0.6 g of colourless crystals. 1H NMR
max and min residual electron density = 0.233 and −0.161 (400.13 MHz, d6-benzene, 300 K) δ 8.03 (d, 1H, Hmeta), 7.87
e·Å−3.
(d, 0.27H, Hmeta), 7.17 (overlap with solvent, 2.5H, Hmeta′),
.06 (t, 1H, Hpara), 7.01 (t, 0.29H, Hpara), 6.86 (t, 0.29H,
7
Crystal data for 4/5: C82H164N13Na3Zn4, M = 1662.71, triclinic, Hmeta′), 6.72 (d, 1H, Hortho′), 6.60 (d, 0.28H, Hortho′); 13C NMR
P−1, a = 9.3229(2), b = 20.0052(4), c = 26.2813(6) Å, α = (100.62 MHz, d6-benzene, 300 K) δ 159.4 (Na–Cortho (major)),
8
6.732(2)°, β = 82.721(2)°, γ = 87.247(2)°, V = 4850.10(18) Å3, 157.0 (Cipso (major)), 140.2 (Cmeta (major)), 139.5 (Cmeta
Z = 2. 57508 reflections collected, 19054 were unique, Rint = (minor)), 126.0 (Cmeta′ (major)), 125.4 (Cpara (minor)), 120.9
.0630, R = 0.0413, Rw = 0.0680, GOF = 0.839, 956 refined (Cpara (major)), 119.5 (Cmeta′ (minor)), 113.8 (Cortho′ (major)),
0
parameters, max and min residual electron density = 0.969 112.2 (Cortho′ (minor)). The relevant resonances for the
and −0.682 e·Å−3. Disorder in the TMEDA groups was treated remaining quaternary carbons in the minor ortho-deprotonated
as being over two sites, appropriate restraints on atom- product, Cipso and Na–Caryl could not be detected. Due to the
atom distances and temperature factors in these groups were presence of three species and resulting complexity, no correla-
applied.
tion could be drawn between these signals and those in the ali-
phatic region.
Crystal data for 6: C27H53N4NaZn, M = 522.09, orthorhombic,
Pbca, a = 16.1371(5), b = 17.2078(5), c = 21.6306(5) Å, V = Synthesis of 6 ((TMEDA)Na(TMP)(C6H4-NMe2)Zn(t-Bu)):
006.5(3) Å3, Z = 8. 32010 reflections collected, 7240 were The above-mentioned procedure was repeated and TMP(H) (2
unique, Rint = 0.0540, R = 0.0371, Rw = 0.0853, GOF = 0.881, mmol, 0.34 mL) was introduced to the mixture. The resulting
6
3
0
11 refined parameters, max and min residual electron density = yellow suspension was allowed to stir overnight, after which
.804 and −0.289 e·Å−3. time the resulting white precipitate was collected by filtration
0.32 g, 31%). The precipitate was re-dissolved in warm hexane
(
Synthesis of 3 (((TMEDA)Na(C6H4-NMe2))2) : n-BuNa (2 and allowed to cool to ambient temperature to afford a small
mmol, 0.16 g) was suspended in hexane (10 mL) and sonicated amount of colourless crystals (recrystallised yield: 0.07 g, 7% –
for 10 min to form a fine dispersion. The Schlenk tube was then not optimised). 1H NMR (400.13 MHz, d8-THF, 300 K) δ 7.48
cooled to 0 °C in an ice bath before the dropwise introduction of (d, 1H, Hmeta), 6.92 (t, 1H, Hmeta′), 6.81 (d, 1H, Hortho′), 6.75 (t,
TMEDA (2 mmol, 0.3 mL). N,N-dimethylaniline (2 mmol, 0.25 1H, Hpara), 2.70 (s, 6H, N(CH3)2), 2.31 (s, 4H, CH2
mL) was then added dropwise to the clear yellow solution to (TMEDA)), 2.15 (s, 12H, CH3 (TMEDA)), 1.74 (m, 2H,
give an orange solution which was stirred at 0 °C for 2 hours γ-TMP), 1.37 (br, 4H, β-TMP), 1.20 (s, 12H, CH3 (TMP)), 0.98
and the resulting light orange precipitate was removed by filtra- (s, 9H, CH3 (t-Bu)); 13C NMR (100.62 MHz, d8-THF, 300 K) δ
tion (0.33 g, 64% – based on monomeric unit). The deep red 165.0 (Zn–Cortho), 160.1 (Cipso), 141.2 (Cmeta), 125.5 (Cmeta′),
filtrate solution was concentrated in vacuo and stored in a 122.2 (Cpara), 114.9 (Cortho′), 58.9 (CH2 (TMEDA)), 53.1
refrigerator (5 °C) to yield a small amount of X-ray quality, (α-TMP), 46.8 (N(CH3)2), 46.2 (CH3 (TMEDA)), 40.7
orange crystalline material. 1H NMR (400.13 MHz, d12-cyclo- (β-TMP), 36.1 (CH3 (TMP)), 35.7 (CH3 (t-Bu)), 20.6 (γ-TMP),
hexane, 300 K) δ 7.78 (d, 1H, Hmeta), 6.79 (t, 1H, Hmeta′), 6.58 19.8 (Cq (t-Bu)).
1246

Beilstein J. Org. Chem. 2011, 7, 1234–1248.
Electrophilic quenching reactions: The solutions prepared as Acknowledgements
described above, were treated with a freshly prepared solution This work was generously sponsored by the UK Engineering
of 1 M iodine in THF (7 mmol, 7 mL) and allowed to stir and Physical Science Research Council (award no. EP/
overnight. A solution of NH4Cl (5 mL) was added along with a F063733/1 and a DTA award to L.B.), and the Royal Society/
saturated Na2S2O3 solution until bleaching occurred (5 mL). Wolfson Foundation (research merit award to R.E.M.).
The organic layer was separated from the aqueous layer and
dried over MgSO4. After filtration, the solvent was removed References
under reduced pressure to give a yellow oil. The NMR spec-
trum of the crude material was recorded to determine the yield
of the iodo-product relative to N,N-dimethylaniline.
1. Schlosser, M., Ed. Organometallics in Synthesis, 2nd ed.; Wiley:
Chichester, U.K., 2002; Chapter 1.
2
.
.
Schlosser, M. Angew. Chem., Int. Ed. 2005, 44, 376–393.
doi:10.1002/anie.200300645
3
Gilman, H.; Bebb, R. L. J. Am. Chem. Soc. 1939, 61, 109–112.
doi:10.1021/ja01870a037
2
-Iodo-N,N-dimethylaniline: 1H NMR (400.13 MHz,
d6-benzene, 300 K) δ 7.76 (d, 1H, Hmeta), 6.98 (t, 1H, Hmeta′), 4. Wittig, G.; Pockels, U.; Dröge, H. Ber. Dtsch. Chem. Ges. 1938, 71,
903–1912. doi:10.1002/cber.19380710922
1
6
.72 (d, 1H, Hortho′), 6.46 (t, 1H, Hpara), 2.45 (s, 6H, N(CH3)2);
5
.
.
Gschwend, H. W.; Rodriguez, H. R. Org. React. 1979, 26, 1.
Knochel, P.; Leuser, H.; Cong, L.-Z.; Perrone, S.; Kneisel, F. F.
Polyfunctional Zinc Organometallics for Organic Synthesis. In
Handbook of Functionalized Organometallics: Application in Sythesis;
Knochel, P., Ed.; Wiley-VCH: Weinheim, Germany, 2005; Chapter 7.
doi:10.1002/9783527619467.ch7
1
3C NMR (100.62 MHz, d6-benzene, 300 K) δ 155.4 (Cipso),
6
1
9
40.5 (Cmeta), 129.1 (Cmeta′), 125.2 (Cpara), 120.9 (Cortho′),
7.9 (I–Cortho), 44.7 (N(CH3)2).
3
-Iodo-N,N-dimethylaniline: 1H NMR (400.13 MHz,
7
8
9
1
.
.
.
Rappoport, Z.; Marek, I., Eds. The Chemistry of Organozinc
Compounds; Patai Series; Wiley: Chichester, U.K., 2006.
Krasovskiy, A.; Krasovskaya, V.; Knochel, P. Angew. Chem., Int. Ed.
d6-benzene, 300 K) δ 7.05 (d, 1H, Hpara), 6.98 (s, 1H, Hortho),
.70 (t, 1H, Hmeta′), 6.34 (d, 1H, Hortho′), 2.32 (s, 6H, N(CH3)2);
3C NMR (100.62 MHz, d6-benzene, 300 K) δ 151.8 (Cipso),
6
1
2006, 45, 2958–2961. doi:10.1002/anie.200504024
1
9
30.6 (Cmeta′), 125.5 (Cpara), 121.4 (Cortho), 111.7 (Cortho′),
6.0 (I–Cmeta), 39.7 (N(CH3)2).
Wunderlich, S. H.; Knochel, P. Angew. Chem., Int. Ed. 2007, 46,
7685–7688. doi:10.1002/anie.200701984
0.Dong, Z.; Clososki, G. C.; Wunderlich, S. H.; Unsinn, A.; Li, J.;
Knochel, P. Chem.–Eur. J. 2009, 15, 457–468.
doi:10.1002/chem.200801558
4
-Iodo-N,N-dimethylaniline: 1H NMR (400.13 MHz,
d6-benzene, 300 K) δ 7.42 (d, 2H, Hmeta), 6.13 (d, 2H, Hortho),
11.Mosrin, M.; Knochel, P. Org. Lett. 2009, 11, 1837–1840.
2
3
.36 (s, 6H, N(CH3)2); 13C NMR (100.62 MHz, d6-benzene,
00 K) δ 150.9 (Cipso), 137.8 (Cmeta), 115.0 (Cortho), 77.7
doi:10.1021/ol900342a
12.Bresser, T.; Monzon, G.; Mosrin, M.; Knochel, P.
(
I–Cpara), 39.8 (N(CH3)2).
Org. Process Res. Dev. 2010, 14, 1299–1303. doi:10.1021/op1001935
1
3.Kondo, Y.; Shilai, M.; Uchiyama, M.; Sakamoto, T. J. Am. Chem. Soc.
999, 121, 3539–3540. doi:10.1021/ja984263t
1
GC–MS (CI) m/z: [M + H]+ calcd for C8H10IN, 246.9; found,
47.9.
1
4.Uchiyama, M.; Matsumoto, Y.; Nobuto, D.; Furuyama, T.;
2
Yamaguchi, K.; Morokuma, K. J. Am. Chem. Soc. 2006, 128,
8748–8750. doi:10.1021/ja060489h
For unidentified species, GC–MS (CI) m/z: [M + H]+ calcd for
15.Andrikopoulos, P. C.; Armstrong, D. R.; Barley, H. R. L.; Clegg, W.;
Dale, S. H.; Hevia, E.; Honeyman, G. W.; Kennedy, A. R.;
Mulvey, R. E. J. Am. Chem. Soc. 2005, 127, 6184–6185.
doi:10.1021/ja050860l
C8H9I2N, 372.8; found, 373.8.
Supporting Information
16.Clegg, W.; Dale, S. H.; Hevia, E.; Hogg, L. M.; Honeyman, G. W.;
Mulvey, R. E.; O’Hara, C. T. Angew. Chem., Int. Ed. 2006, 45,
6548–6550. doi:10.1002/anie.200602288
Supporting Information File 1
1
7.Balloch, L.; Kennedy, A. R.; Mulvey, R. E.; Rantanen, T.;
Robertson, S. D.; Snieckus, V. Organometallics 2011, 30, 145–152.
doi:10.1021/om1009654
Computational details and NMR spectra for compounds 3,
4
, 5 and 6.
[http://www.beilstein-journals.org/bjoc/content/
1
8.Hartung, G. C.; Snieckus, V. The Directed ortho Metalation Reaction –
A Point of Departure for New Synthetic Aromatic Chemistry. In Modern
Arene Chemistry; Astruc, D., Ed.; Wiley-VCH: New York, 2002;
pp 330–367. doi:10.1002/3527601767.ch10
supplementary/1860-5397-7-144-S1.pdf]
Supporting Information File 2
CIF files giving crystallographic data for compounds 3, 4, 5
and 6.
1
9.Clayden, J. Organolithiums: Selectivity for Synthesis; Tetrahedron
Organic Chemistry Series, Vol. 23; Pergamon, Elsevier Science:
Oxford, U.K., 2002; Chapter 2.
[http://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-7-144-S2.txt]
1247

Beilstein J. Org. Chem. 2011, 7, 1234–1248.
2
0.Armstrong, D. R.; Clegg, W.; Dale, S. H.; Hevia, E.; Hogg, L. M.;
43.Schade, C.; Bauer, W.; v. R. Schleyer, P. J. Organomet. Chem. 1985,
295, c25–c28. doi:10.1016/0022-328X(85)80326-0
Honeyman, G. W.; Mulvey, R. E. Angew. Chem., Int. Ed. 2006, 45,
3
775–3778. doi:10.1002/anie.200600720
44.Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, 64, 112–122.
doi:10.1107/S0108767307043930
2
1.Card, R. J.; Trahanovsky, W. S. J. Org. Chem. 1980, 45, 2560–2566.
doi:10.1021/jo01301a002
22.Baston, E.; Maggi, R.; Friedrich, K.; Schlosser, M. Eur. J. Org. Chem.
2001, 3985–3989.
doi:10.1002/1099-0690(200111)2001:21<3985::AID-EJOC3985>3.0.C
O;2-Y
License and Terms
2
2
2
3.Chalk, A. J.; Hoogeboom, T. J. J. Organomet. Chem. 1968, 11,
This is an Open Access article under the terms of the
Creative Commons Attribution License
615–618. doi:10.1016/0022-328X(68)80091-9
4.Broaddus, C. D. J. Am. Chem. Soc. 1966, 88, 4174–4178.
doi:10.1021/ja00970a012
(http://creativecommons.org/licenses/by/2.0), which
permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
5.Armstrong, D. R.; García-Álvarez, J.; Graham, D. V.;
Honeyman, G. W.; Hevia, E.; Kennedy, A. R.; Mulvey, R. E.
Chem.–Eur. J. 2009, 15, 3800–3807. doi:10.1002/chem.200801928
6.Nobuto, D.; Uchiyama, M. J. Org. Chem. 2008, 73, 1117–1120.
doi:10.1021/jo701895z
2
2
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
7.Clegg, W.; Conway, B.; Hevia, E.; McCall, M. D.; Russo, L.;
Mulvey, R. E. J. Am. Chem. Soc. 2009, 131, 2375–2384.
doi:10.1021/ja8087168
(http://www.beilstein-journals.org/bjoc)
The definitive version of this article is the electronic one
which can be found at:
2
8.Armstrong, D. R.; Blair, V. L.; Clegg, W.; Dale, S. H.;
García-Álvarez, J.; Honeyman, G. W.; Hevia, E.; Mulvey, R. E.;
Russo, L. J. Am. Chem. Soc. 2010, 132, 9480–9487.
doi:10.1021/ja1038598
doi:10.3762/bjoc.7.144
2
3
3
3
3
3
3
9.Snieckus, V. Chem. Rev. 1990, 90, 879–933.
doi:10.1021/cr00104a001
0.Wittig, G.; Merkle, H. Ber. Dtsch. Chem. Ges. B 1942, 75, 1491–1500.
doi:10.1002/cber.19420751212
1.Lepley, A. R.; Khan, W. A.; Giumanini, A. B.; Giumanini, A. G.
J. Org. Chem. 1966, 31, 2047–2051. doi:10.1021/jo01345a001
2.Niemeyer, M.; Power, P. P. Organometallics 1997, 16, 3258–3260.
doi:10.1021/om970247d
3.Betz, J.; Hampel, F.; Bauer, W. J. Chem. Soc., Dalton Trans. 2001,
1876–1879. doi:10.1039/B009077H
4.Schade, C.; v. R. Schleyer, P.; Dietrich, H.; Mahdi, W.
J. Am. Chem. Soc. 1986, 108, 2484–2485. doi:10.1021/ja00269a078
5.Slocum, D. W.; Reece, T. L.; Sandlin, R. D.; Reinscheld, T. K.;
Whitley, P. E. Tetrahedron Lett. 2009, 50, 1593–1595.
doi:10.1016/j.tetlet.2009.01.104
3
3
3
3
4
6.Allen, F. H. Acta Crystallogr., Sect. B 2002, 58, 380–388.
doi:10.1107/S0108768102003890
For CSD version 5.31.
7.Clegg, W.; Dale, S. H.; Harrington, R. W.; Hevia, E.; Honeyman, G. W.;
Mulvey, R. E. Angew. Chem., Int. Ed. 2006, 45, 2374–2377.
doi:10.1002/anie.200503213
8.Carrella, L. M.; Förster, C.; Kennedy, A. R.; Klett, J.; Mulvey, R. E.;
Rentschler, E. Organometallics 2010, 29, 4756–4758.
doi:10.1021/om100327w
9.Lorenzen, N. P.; Kopf, J.; Olbrich, F.; Schümann, U.; Weiss, E.
Angew. Chem., Int. Ed. Engl. 1990, 29, 1441–1444.
doi:10.1002/anie.199014411
0.Armstrong, D. R.; Forbes, G. C.; Mulvey, R. E.; Clegg, W.;
Tooke, D. M. J. Chem. Soc., Dalton Trans. 2002, 1656–1661.
doi:10.1039/B110696A
41.Bell, N. A.; Shearer, H. M. M.; Spencer, C. B. Acta Crystallogr., Sect. C
1983, 39, 1182–1185. doi:10.1107/S0108270183007842
42.Coles, M. P.; Hitchcock, P. B. Eur. J. Inorg. Chem. 2004, 2662–2672.
doi:10.1002/ejic.200400031
1248