Donor-activated alkali metal dipyridylamides: Co-complexation reactions with zinc alkyls and reactivity studies with benzophenone

Article abstract of DOI:10.1039/c4dt00914b

Previously it was reported that activation of ^tBu₂Zn by [(TMEDA)Na(μ-dpa)]₂led to tert-butylation of benzophenone at the challenging para-position, where the sodium amide functions as a metalloligand towards ^tBu₂Zn manifested in crystalline [{(TMEDA)Na(dpa)}₂Zn^tBu₂] (TMEDA is N,N,N′,N′-tetramethylethylenediamine, dpa is 2,2′-dipyridylamide). Here we find altering the Lewis donor or alkali metal within the metalloligand dictates the reaction outcome, exhibiting a strong influence on alkylation yields and reaction selectivity. Varying the former led to the synthesis of three novel complexes, [(PMDETA)Na(dpa)]₂, [(TMDAE)Na(dpa)]₂, and [(H₆-TREN)Na(dpa)], characterised through combined structural, spectroscopic and theoretical studies [where PMDETA is N,N,N′,N′′,N′′-pentamethyldiethylenetriamine, TMDAE is N,N,N′,N′-tetramethyldiaminoethylether and H₆-TREN is N′,N′-bis(2-aminoethyl)ethane-1,2-diamine]. Each new sodium amide can function as a metalloligand to generate a co-complex with ^tBu₂Zn. Reacting these new co-complexes with benzophenone proved solvent dependent with yields in THF much lower than those in hexane. Most interestingly, sub-stoichiometric amounts of the metalloligands [(TMEDA)Na(dpa)]₂and [(PMEDTA)Na(dpa)]₂with 1 : 1, ^tBu₂Zn-benzophenone mixtures produced good yields of the challenging 1,6-tert-butyl addition product in hexane (52% and 53% respectively). Although exchanging Na for Li gave similar reaction yields, the regioselectivity was significantly compromised; whereas the K system was completely unreactive. Replacing ^tBu₂Zn with (Me₃SiCH₂)₂Zn shut down the alkylation of benzophenone; in contrast, ^tBuLi generates only the reduction product, benzhydrol. Zincation of the parent amine dpa(H) generated the crystalline product [Zn(dpa)₂], as structurally elucidated through X-ray crystallography and theoretical calculations. Although the reaction mechanism for the alkylation of benzophenone remains unclear, incorporation of the radical scavenger TEMPO (2,2,6,6-tetramethylpiperidine-N-oxyl radical) into the reaction system completely inhibits benzophenone alkylation.

Full text of DOI:10.1039/c4dt00914b

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Donor-activated alkali metal dipyridylamides:
co-complexation reactions with zinc alkyls and
reactivity studies with benzophenone†‡
Cite this: Dalton Trans., 2014, 43,
14409
David R. Armstrong, Etienne V. Brouillet, Alan R. Kennedy, Jennifer A. Garden,*
Markus Granitzka, Robert E. Mulvey* and Joshua J. Trivett
Previously it was reported that activation of tBu2Zn by [(TMEDA)Na(μ-dpa)]2 led to tert-butylation of
benzophenone at the challenging para-position, where the sodium amide functions as a metalloligand
t
towards Bu2Zn manifested in crystalline [{(TMEDA)Na(dpa)}2ZntBu2] (TMEDA is N,N,N’,N’-tetramethylethyl-
enediamine, dpa is 2,2’-dipyridylamide). Here we find altering the Lewis donor or alkali metal within the
metalloligand dictates the reaction outcome, exhibiting a strong influence on alkylation yields and reac-
tion selectivity. Varying the former led to the synthesis of three novel complexes, [(PMDETA)Na(dpa)]2,
[(TMDAE)Na(dpa)]2, and [(H6-TREN)Na(dpa)], characterised through combined structural, spectroscopic
and theoretical studies [where PMDETA is N,N,N’,N’’,N’’-pentamethyldiethylenetriamine, TMDAE is N,N,N’,N’-
tetramethyldiaminoethylether and H6-TREN is N’,N’-bis(2-aminoethyl)ethane-1,2-diamine]. Each new
t
sodium amide can function as a metalloligand to generate a co-complex with Bu2Zn. Reacting these
new co-complexes with benzophenone proved solvent dependent with yields in THF much lower than
those in hexane. Most interestingly, sub-stoichiometric amounts of the metalloligands [(TMEDA)Na(dpa)]2
t
and [(PMEDTA)Na(dpa)]2 with 1 : 1, Bu2Zn–benzophenone mixtures produced good yields of the challen-
ging 1,6-tert-butyl addition product in hexane (52% and 53% respectively). Although exchanging Na for Li
gave similar reaction yields, the regioselectivity was significantly compromised; whereas the K system was
completely unreactive. Replacing tBu2Zn with (Me3SiCH2)2Zn shut down the alkylation of benzophenone;
in contrast, tBuLi generates only the reduction product, benzhydrol. Zincation of the parent amine dpa(H)
generated the crystalline product [Zn(dpa)2], as structurally elucidated through X-ray crystallography and
theoretical calculations. Although the reaction mechanism for the alkylation of benzophenone remains
unclear, incorporation of the radical scavenger TEMPO (2,2,6,6-tetramethylpiperidine-N-oxyl radical) into
the reaction system completely inhibits benzophenone alkylation.
Received 27th March 2014,
Accepted 18th June 2014
DOI: 10.1039/c4dt00914b
www.rsc.org/dalton
components of the synthetic chemist’s toolbox, current
research has brought heterobimetallic amides into the lime-
Introduction
Derived from sterically imposing secondary amines (R2NH), light. The synergic combination of two metals within an
polar metal amides have a long successful track record as organoamide-ligand environment has often led to unexpected
important reagents in synthesis, with several recent reviews con- and unprecedented chemistry, which cannot be reproduced by
firming that their popularity continues to be high.1–6 Although either metal acting in isolation. Sophisticated bimetallic bases
in this category monometallic amides are well-established have been developed, where perhaps counter intuitively, it is
sometimes but not always the softer polyvalent metal (such
as Zn,7–9 Mg,10,11 Al,12,13 Cd,14–16 or Mn17,18) that executes
metallo-deprotonation (conversion of an inert C–H bond to a
WestCHEM, Department of Pure and Applied Chemistry, University of Strathclyde,
Glasgow, G1 1XL, UK. E-mail: r.e.mulvey@strath.ac.uk
labile, more synthetically flexible C–metal bond), assisted by
the presence of a reactivity enhancing alkali metal.
Amongst other types of bond forming reactions, alkylselec-
tive addition to ketones has been much studied using several
different organometallic reagents, including polar metal
amides.8 Benzophenone is often used as a benchmark ketone
†Dedicated to the memory of Ken Wade FRS. A brilliant intuitive chemist, gener-
ous in his wisdom and time to all who asked for his advice. REM remembers
with great affection his postdoctoral period with Ken.
‡Electronic supplementary information (ESI) available: NMR spectroscopic data,
computational details and X-ray data in crystallographic file (CIF) format for
compounds 2, 3 and 6. CCDC 992383–992385. For ESI and crystallographic data
in CIF or other electronic format see DOI: 10.1039/c4dt00914b
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 14409–14423 | 14409

View Article Online
Paper
Dalton Transactions
for assessing organometallic reagents as its reactions can
have several potential outcomes.8,19–22 Monometallic reagents
generally produce significantly greater proportions of the
Results and discussion
Modification of the donor ligand: synthesis of a series of
sodium dipyridylamide metalloligands
hydride reduction product benzhydrol [Ph2C(H)OH] (with
n
nBuMgCl; 56%, BuLi; 38%); whereas bimetallic combinations Striving to establish the importance or otherwise of the donor
can offer superior selectivity, with the use of the synergic employed within the [(TMEDA)Na(dpa)]2 metalloligand, we
lithium magnesiate base nBu3MgLi·LiCl leading to exclusive prepared a series of related sodium amide compounds with
carbonyl (1,2-) addition in an impressive 95% yield.19 Selective the general formula [(donor)Na(dpa)]x. The alternative Lewis
alkylation at the more challenging para (1,6-) position has also donors used were PMDETA (N,N,N′,N″,N″-pentamethyldiethyl-
been realised using a bimetallic reagent, in the structurally enetriamine), TMDAE (N,N,N′,N′-tetramethyldiaminoethyl-
defined sodium amidozincate [(TMEDA)Na(μ-TMP)(μ-tBu)Zn- ether) and H6-TREN [N′,N′-bis(2-aminoethyl)ethane-1,2-
(tBu)] (where TMEDA is N,N,N′,N′-tetramethylethylenediamine diamine] (Scheme 1). We began by deprotonating the parent
and TMP is 2,2,6,6-tetramethylpiperidide).8,23 Thus the amine dpa(H) with n-butylsodium in hexane solvent. This reac-
identity of the organometallic reagent exerts a strong influence tion produced a white suspension. The mixture was sub-
upon the reaction outcome.
sequently treated with the appropriate donor amine. For the
In a recent communication we described how sub-stoichio- trifunctional donors PMDETA and TMDAE, two molar equi-
metric quantities of the sodium amide, [(TMEDA)Na(dpa)]2, valents (on a 1 mmol scale reaction) were required to dissolve
can activate tBu2Zn (within the sodium–zinc compound the suspension. Initially made up at ambient temperature, the
[{(TMEDA)Na(dpa)}2ZntBu2], 1, Fig. 1, where the sodium amide reaction solutions were cooled to −30 °C in order to grow poor
functions as a metalloligand) towards more improved alkyl- quality crystals of [(PMDETA)Na(dpa)]2 [2, 81% isolated yield
ation of benzophenone in the para-position [where dpa is 2,2′- (reported yields are based upon the dpa(H) stoichiometry)]
dipyridylamide, (2-NC5H4)2N−].24 This intriguing result and [(TMDAE)Na(dpa)]2, [3, 37% yield] respectively. Upon
prompted us to ask various questions. Would changing the moving to tetrafunctional H6-TREN (one molar equivalent),
constituents of the metalloligand allow us to tune the reactivity addition of toluene to the yellow-green oil and gentle heating
and selectivity towards the alkylation of benzophenone? Is the produced a pale yellow solution, which upon gradual cooling
coordination of TMEDA to sodium within the metalloligand to ambient temperature afforded a crop of colourless crystals
structure important to the outcome of these addition reac- of [(H6-TREN)Na(dpa)] (4) in a high isolated yield of 91%. New
tions? Can we extend this system to sodium’s nearest sodium amides 2–4 were characterised in solution using multi-
neighbours, lithium and potassium? Can the reaction scope nuclear (1H, 13C{1H}) NMR spectroscopy, and the molecular struc-
be broadened to other alkyl groups? The studies reported tures of 2 and 3 were elucidated by X-ray crystallographic studies.
herein aimed to find answers to these important questions.
Thus, a range of alkali metal dipyridylamide metalloligands
was applied to the alkylation of benzophenone, using Bu2Zn
Structural insights into sodium dipyridylamides 2–4
t
Unfortunately, twinning, disorder and weak diffraction leads
to a sub-optimal structural solution of 2, which prevents any
discussion of its dimensions (see the ESI‡ for a geometry-opti-
mised model, 2calc, using DFT calculations). However, its
atomic connectivity can be discerned (Fig. 2). Adopting a com-
monly observed motif within alkali metal chemistry,25,26 and
as the alkyl transfer reagent. Through the combination of
spectroscopic and structural characterisation data backed up
with theoretical calculations, this study sheds insight upon the
impact of changing parameters such as the Lewis donor, alkali
metal and solvent polarity upon the yields and product
distribution.
Fig. 1 The tert-butylation of benzophenone using tBu2Zn activated by a
t
sodium amide metalloligand (top), in contrast to Bu2Zn alone which is
ineffectual (bottom).
Scheme 1 Synthesis of sodiated dipyridylamine dimers 2 and 3, and
monomer 4.
14410 | Dalton Trans., 2014, 43, 14409–14423
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Fig. 3 Potential ligating modes of dpa(H) and dpa− towards a metal
centre “M”, displaying either syn [N(pyridyl) is directed the same way as
N(amido)] or anti [N(pyridyl) is directed towards the neighbouring
pyridyl ring] bonding conformations.
Fig. 2 Molecular structure of 2 with thermal ellipsoids at 50% prob-
ability level and hydrogen atoms and disordered components omitted
for clarity.
most pertinently for sodium amides27–29 such as [{(TMEDA)-
Na(NPh2)}2]28 and [{(THF)Na(HMDS)}2],30 2 adopts a dimeric,
cyclic arrangement, where each Na centre is coordinated by
two bridging dipyridylamide units and a chelating PMDETA.
An unusual feature is the incomplete chelation by the triamine
donor PMDETA, which engages Na in a bidentate rather than
1
in its common full tridentate capacity. H NMR spectroscopic
analysis of 2 suggests that this bidentate bonding mode is not
retained in C6D6 solution (vide infra). A search of the CSD
revealed only 11 structurally characterised compounds display-
ing this rare bidentate bonding mode, in comparison to 425
that involve tri-coordination of PMDETA to a Lewis acidic
metal centre.31 Most relevant to 2 is the structurally related
phenyl complex [(PMDETA)Na{Ph(2-NC5H4)N}]2:32 both
possess a planar [NaNNaN] ring core; with hexa-coordinated
Na adopting a distorted octahedral geometry; and with two
amido bridges connecting the two Na centres. However, a key
Fig. 4 Molecular structure of 3 with thermal ellipsoids at 50% prob-
ability level. Hydrogen atoms and minor disordered components of
TMDAE have been omitted for clarity.
distinction between the two structures is the bonding mode of adopts a syn–syn conformation equivalent to that observed in
the bridging amide unit. Within [(PMDETA)Na{Ph(2-NC5H4)- 2, where N(amido) bridges between two Na centres with two
N}]2, each amide unit forms both a Na–N(pyridyl)–Na and Na– terminal N(pyridyl)–Na interactions providing further stabilis-
N(amido)–Na bridge. In contrast replacing the phenyl ring ation. In contrast, the second dpa ligand occupies a syn–anti
with a second pyridyl ring in 2 results in loss of the N(pyridyl) conformation, with N(amido) and N(syn-pyridyl) forming an
bridge in exchange for two dative Na–N(pyridyl) interactions unsymmetrical bridge [where the bridging Na1–N8, Na1–N9,
with a terminally attached, syn–syn dpa ligand (Fig. 3). Related Na2–N8 and Na2–N9 bond lengths are 2.680(4), 2.538(3),
studies by Liddle and Clegg have established a series of alkali 2.752(4) and 2.496(3) Å, respectively]. As a result, 3 contains
metal complexes based upon the related parent amines two central four-membered [NaNNaN] rings at its core, each
2-phenylaminopyridine33–35 and 2-trimethylsilylaminopyri- deviating from planarity with torsion angles of 15.281(2)°
+
dine,36,37 including the crown ether solvates [{Na(12C4)2 }(L−)] (Na2–N6–Na1–N9), and 35.359(2)° (Na1–N6–Na2–N8). Indica-
and [{K(L)(12C4)}2].38
tive of a small degree of resonance delocalisation within the
As would be anticipated according to the close similarity dpa units, the dihedral angles between the pyridyl ring planes
between isoelectronic PMDETA and TMDAE, sodium amide 3 are relatively large [48.216° between the N5 and N7 pyridyl ring
exhibits a dimeric arrangement (Fig. 4) akin to that of 2. The planes; 27.787° between the N8 and N10 pyridyl ring planes].
structure of 3 contains two crystallographically distinct Na
The chemistry of the tridentate, mixed N and O donor
centres, each of which engages in a didentate fashion with TMDAE is underdeveloped in comparison to that of its
chelating TMDAE. Each hexa-coordinated Na also interacts triamine counterpart, PMDETA. In this regard, 3 represents
with two bridging dpa units, both in a bidentate fashion, a rare example of a structurally characterised compound
although the coordination mode differs. The first dpa ligand containing a non-substituted TMDAE donor. A search of the
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 14409–14423 | 14411

View Article Online
Paper
Dalton Transactions
Fig. 5 Relative energies of DFT-models 4calcA-C, representative of the
regioisomers of complex 4.
Fig. 6 DFT model of 4calc-A with hydrogen atoms omitted for clarity.
Selected bond lengths (Å) and bond angles (°): Na1–N1, 2.388; Na1⋯N2,
3.720; Na1–N3, 2.539; Na1–N4, 2.516; Na1–N5, 2.756; Na1–N6, 2.526;
Na1–N7, 2.510; N1–C5, 1.353; N1–C6, 1.351; N2–C1, 1.334; N2–C5,
1.368; N3–C6, 1.373; N3–C10, 1.333; C1–C2, 1.389; C2–C3, 1.399; C3–
C4, 1.381; C4–C5, 1.427; C6–C7, 1.428; C7–C8, 1.380; C8–C9, 1.400;
C9–C10, 1.389; N1–Na1–N3, 55.1; N1–Na1–N4, 86.3; N1–Na1–N5,
138.8; N1–Na1–N6, 151.6; N1–Na1–N7, 91.9; N3–Na1–N4, 131.1; N3–
Na1–N5, 159.2; N3–Na1–N6, 99.1; N3–Na1–N7, 99.5; N4–Na1–N5,
69.7; N4–Na1–N6, 107.2; N4–Na1–N7, 111.8; N5–Na1–N6, 69.5; N5–
Na1–N7, 68.4; N6–Na1–N7, 105.3.
CSD31 surprisingly revealed that only one other compound
involving a non-substituted TMDAE ligand has been structu-
rally characterised, namely the solvated copper salt [(TMDAE)-
CuCl2].39 In this example, TMDAE binds to Cu in a tridentate
fashion. Illustrating the capability of TMDAE to act alterna-
tively as
a bidentate donor, the lithium aluminate [Li-
(μ-Me2NCH2CH2OCHCH2NMe2)(μ-TMP)Al(iBu)2] has also been
structurally characterised,13 although in this case, TMDAE is
an anion as it has also been subjected to deprotonation on a
CH2 unit adjacent to the O atom.
between the ‘C–C’ bonds. With steric congestion about the Na
centre blocking a second Na–N(pyr) dative interaction, the
non-coordinated pyridyl N2 is stabilised through the formation
of a hydrogen bond with H6-TREN [N2⋯H(N4), 2.068 Å,
N2⋯N4, 3.071 Å]. A similar delocalisation of the anionic
charge has been observed in alkali metal complexes bearing
the related 2-phenylamidopyridine anion.38
Although the application of either H6- or Me6-TREN as a
highly chelating Lewis donor has been studied extensively
within the context of transition metal chemistry, similar
studies within s-block chemistry are less well evolved. None-
theless, more recently structural studies of alkali metal-TREN
complexes have been forthcoming. A representative selection
is shown in Fig. 7, including 13 monomeric examples.42–46 For
example, the amido lithium anionic crown [{Me6-TREN}Li-
(μ-Cl)Li{Me6-TREN}]+[{Li5(μ-HMDS)5Cl}]−, provides a cationic
example (Fig. 7c), with a linear Li–Cl–Li unit at its core.45
Another unusual example is the tetranuclear lithium com-
pound [Li2(OArMe)2·Me6TREN]2, which demonstrates the good
bridging ability of OArMe ligands.43
Unfortunately, repeated attempts to grow crystals of 4 suit-
able for X-ray diffraction analysis were unsuccessful as all were
found to be badly disordered. We therefore turned again to a
DFT study using the B3LYP method40 and the 6-311G(d,p)
basis set41 to facilitate the modelling of the molecular struc-
ture of 4. Computations were performed to probe the relative
energies of the conformational isomers of 4, where dpa adopts
either a syn–syn (model 4calc-A), syn–anti (model 4calc-B) or
anti–anti (model 4calc-C) arrangement (Fig. 5). As H6-TREN
commonly acts as a monomerisation agent, 4calc-A-C were
modelled as monomeric species.42–46 Significantly, compari-
son of these model isomers revealed only a small difference in
the relative energies of the three conformers, spanning just
9.00 kcal mol−1. This closeness in the relative energy terms
could be a factor in the severe disorder observed in the struc-
ture of crystalline 4. Although model 4calc-A is the energy
minimum structure, the difference between 4calc-A and 4calc-B
is very low at only 3.19 kcal mol−1. However, for brevity, only
the bond parameters of the thermodynamically favourable
model 4calc-A shall be discussed herein (Fig. 6).
Within 4calc-A, sodium engages syn–syn dpa in an asym-
metric fashion (Na1–N1, 2.388 Å; Na1–N3, 2.539 Å, Na1⋯N2,
3.720 Å) and additional chelation by tetradentate H6-TREN
completes the highly congested, distorted octahedral coordi-
nation sphere [N–Na–N bond angles range from 55.1° to
159.2°]. Indicative of resonance delocalisation within the dpa
scaffold, the C–(amido)N bond lengths of dpa are short [1.351
and 1.353 in 4calc-A cf. 1.362 and 1.363 Å in 2calc]. Furthermore,
the bond lengths are indicative of a N–CvC–CvC–C–N pyridyl
pattern (bond lengths 1.334/1.333, 1.389/1.389, 1.399/1.400,
1.381/1.380, 1.427/1.428, 1.368/1.373 Å), although the different
N and H substituents give rise to a significant difference
Solution state characterisation of sodium amides 2–4
Multinuclear (1H, 13C{H}) NMR spectroscopic analysis of
sodium amide complexes 2–4 in d8-THF solution confirmed
that in each case, mono-deprotonation of dpa(H) has occurred
with the loss of the N–H resonance at 8.83 ppm. As a conse-
quence of dpa(H) metallation, the four aromatic resonances
(labelled H1–H4) experience a significant low frequency shift
(Table S2‡). For example, the H4-dpa resonance deviates
upfield from 7.72 ppm in the parent amine dpa(H), by
0.65 ppm in 2, 0.42 ppm in 3 and 0.57 ppm in 4. On the basis
of this information it appears that aggregation (in 2 and 3) has
less bearing on these chemical shifts than the local coordi-
14412 | Dalton Trans., 2014, 43, 14409–14423
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Scheme 2 Synthesis of sodium zincate [{(TMEDA)Na(dpa)}2ZntBu2]
(1).24
t
(dpa)], [(TMDAE)Na(dpa)] or [(H6-TREN)Na(dpa)] with Bu2Zn,
whereupon each reaction mixture was studied via 1H NMR
spectroscopy in C6D6 solution (Table S3‡). Unfortunately,
repeated attempts to grow crystals of these putative co-com-
plexes suitable for X-ray diffraction analysis were unsuccessful.
That notwithstanding, detection of resonances attributed to
[(donor)Na(dpa)] (where donor is PMDETA, TMDAE or
t
H6-TREN) and Bu2Zn in a 2 : 1 ratio, infers the formation of a
sodium–zinc co-complex. Since non-coordinated tBu2Zn is
highly volatile under reduced pressure at ambient temperature,
it would have been removed in vacuo on preparing the
t
samples. Examination of the Bu2Zn resonances, which are
different in each case (1.58 ppm in 2; 1.53 ppm in 3, 1.66 ppm
in 4) confirms there is a considerable shift towards higher
frequency in comparison to non-coordinated tBu2Zn (1.08 ppm
Fig. 7 Graphical representation of the molecular structures of s-block-
TREN complexes (a) ref. 42, (b) ref. 44, (c) ref. 45, (d) ref. 46, (e) ref. 43.
t
in C6D6 solution). Incorporation of Bu2Zn also significantly
alters the resonances associated with the sodium amide
moieties 2–4 (Table S3‡). For example, upon addition of
tBu2Zn to 2, H3-dpa experiences an upfield shift of 0.16 ppm,
whilst the PMDETA-N(CH3)2 resonance moves upfield by
0.19 ppm.
Diffusion ordered spectroscopy (DOSY) experiments in
C6D6 solution on sodium amides 2–4 in the presence of
tBu2Zn suggest that all of these amides are co-complexed with
the zinc reagent but that their respective donor ligands,
PMDETA, TMDAE, and H6-TREN compete with C6D6 solution
molecules in coordinating to the sodium (see ESI‡ for full
details).
Although di-tert-butylzinc free 2calc and di-tert-butylzinc
containing 1 have different Lewis donor components, the
didentate coordination mode of trifunctional PMDETA within
2calc mimics that of difunctional TMEDA within 1. Thus the
similarity between the sodium amide component of 2calc and 1
allows for tentative comparisons to be drawn. Within 2calc, the
steric congestion at Na is reflected by the narrow N–Na–N bite
angle of 67.8° for PMDETA, which is considerably smaller than
the corresponding TMEDA bite angles within 1 [N–Na–N bond
nation environment of the Na centre. Within the 13C{H} NMR
spectra, metallation of dpa(H) translates as a downfield shift
of the Cipso resonance, by 11.4 ppm in 2, 7.4 ppm in 3 and
11.2 ppm in 4.
Comparison of the 1H NMR spectroscopic resonances
attributed to PMDETA in 2 with those of non-coordinated
PMDETA give good agreement in d8-THF solvent [NCH2, 2.42
and 2.31 ppm; NCH3, 2.18 ppm; N(CH3)2, 2.14 ppm]. Thus, the
bulk Lewis donor solvent d8-THF is exposed as a non-innocent
solvent medium, which displaces the trifunctional amine from
coordination to the electropositive Na centre in solution. More-
over, this finding is mirrored by the resonances ascribed to
TMDAE and H6-TREN within 3 and 4 respectively, which are in
close agreement with those of non-coordinated TMDAE and
H6-TREN in d8-THF solution.
Evaluating formation of a bimetallic sodium–zinc product
upon co-complexation of tBu2Zn with sodium amides 2, 3
and 4
Evidence that combining metalloligand [(TMEDA)Na(dpa)]2 angles are 74.63(5)° and 74.18(6)°]. It therefore seems logical
t
with Bu2Zn results in the formation of a co-complex comes in that in 1, coordination of the metalloligand {(TMEDA)Na-
the form of the previously reported donor–acceptor sodium– (dpa)}2 towards tBu2Zn relieves the steric strain at the Na
zinc compound [{(TMEDA)Na(dpa)}2ZntBu2] (1) (Scheme 2).24 centre, as one dpa unit welcomes tBu2Zn into its pyridyl
Attempts to ascertain whether 2, 3 and 4 can also behave as pocket by switching its coordination mode from syn–syn to
t
metalloligands towards Bu2Zn were carried out through the anti–anti. Investigation into the dpa bond lengths of the
2 : 1 equimolar combination of sodium amides [(PMDETA)Na- sodium precursor 2calc suggests that the anionic charge of
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 14409–14423 | 14413

View Article Online
Paper
Dalton Transactions
each dpa ligand is coupled into the pyridyl rings through detrimental effect on the reaction because reactive three coor-
resonance delocalisation. This contrasts with the pattern dinate monomers can be transformed into inert four coordi-
observed in 1, where although the anionic charge is deloca- nate species. We therefore decided to investigate the effect of
lised throughout the syn–syn dpa unit, the anionic charge of modifying the Lewis donor within our metalloligand system
the anti–anti dpa unit is predominantly focussed on N(amido). (Table 1). A range of donors was investigated, including the
It seems plausible that this difference is due to the steric mixed O/N donors TEMPO (2,2,6,6-tetramethylpiperidine-
requirements of tetrahedrally coordinated Zn interrupting the N-oxyl radical) and TMDAE, and the all N donors DMAP
planar geometry usually observed with resonance delocalisa- (4-dimethylaminopyridine), TMEDA, PMDETA, H6-TREN and
tion. Structural modelling of the {(TMEDA)Na(dpa)}2 metallo- Me6-TREN.
t
ligand, 8calc
,
through DFT calculations revealed close
Using the homometallic zinc reagent Bu2Zn without a sup-
similarities between 2calc and 8calc (refer to ESI‡ for further porting sodium amide, a miserable 1% yield of the 1,6-(para)-
detail). The gross structural factors remain the same: Na addition product was obtained at ambient temperature, with
adopts a distorted octahedral coordination sphere and both no other addition products detected, but only unreacted benzo-
dpa ligands adopt a syn–syn conformation with the anionic phenone. However, introduction of the sodium dipyridyl-
charge coupled into the pyridyl rings.
amide metalloligand with DMAP, TMEDA or PMDETA as the
Lewis donor boosted the reactivity of tBu2Zn, leading to moder-
ate para-alkylation yields of 24% (entry 2), 40% (entry 4) and
43% (entry 5), respectively. In each case, para-addition was
obtained as the major product hinting at a common mechan-
ism within these donor distinct reactions. In contrast, using
TEMPO, TMDAE, H6-TREN or Me6-TREN as the Lewis donor
gave no significant reactivity enhancement over tBu2Zn (entries
3 and 6–8). It is noteworthy that using the TMDAE metallo-
ligand under reflux conditions led to a sizeable improvement
Reactivity studies: investigating the outcome of varying the
Lewis donor upon benzophenone alkylation
It has long been known that the addition of Lewis donors to
polar organometallic reagents can significantly alter their
structures47–50 leading to changes in their reactivity.51 In the
case of organoalkali metal compounds this change can nor-
mally be an enhancement due to deaggregation effects;
whereas with organoaluminium compounds it can have a
Table 1 Reaction of zinc reagents with benzophenone in hexane solvent
Product yielda (%)
Reaction temp.
(°C)
para-Addition
(1, 6-)
Carbonyl-addition
(1, 2-)
ortho-Addition
(1, 4-)
Benzhydrol
Entry
Donor
(H− addition)
Total
Stoichiometric conditionsb
1
Nonec 24
DMAP
r.t.
r.t.
r.t.
r.t.
r.t.
r.t.
r.t.
r.t.
75
75
75
75
75
1
24
0
40
43
2
0
6
0
6
2
1
1
1
1
14
11
3
9
4
0
0
0
1
1
0
0
0
0
0
2
1
1
0
0
0
5
1
8
11
0
4
1
8
8
14
10
3
7
11
1
35
1
55
57
3
2
3
4
5
TEMPO
TMEDA24
PMDETA
TMDAE
6
7
8
9
10
11
12
13
14
15
H6-TREN
Me6-TREN
Nonec 24
DMAP
2
1
7
3
11
25
44
46
45
5
20
47
71
60
58
16
58
TMEDA24
PMDETA
TMDAE
H6-TREN
75
75
Me6-TRENd
34
13
Sub-stoichiometric conditionse
16
17
18
19
DMAP
75
75
75
75
26
52
53
25
8
12
6
0
0
0
0
7
7
6
3
41
71
65
35
TMEDA24
PMDETA
TMDAE
7
a Yields determined by 1H NMR spectroscopy using hexamethylbenzene (10 mol%) as an internal standard. b Conditions: Nadpa (2 mmol), donor
t
(2 mmol), Bu2Zn (1 mmol), PhC(vO)Ph (1 mmol), hexane (8 mL) for 18 hours. c tBu2Zn alone was used as the reagent. d The reaction was
performed four times and the yields proved variable, with the 1,6-addition product ranging from 26% to 41% (refer to ESI for further details).
e Conditions: Nadpa (1 mmol), donor (1 mmol), tBu2Zn (5 mmol), PhC(vO)Ph (5 mmol), hexane (40 mL) for 18 hours.
14414 | Dalton Trans., 2014, 43, 14409–14423
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
of the para-alkylation yield, from 2% to 45% at 75 °C (compare key distinctions between 2 and 3. The donor chelation in 2 is
entries 6 and 13). This surprising result was mirrored by Me6- via two N atoms, whereas in 3 it is via one O and one N atom.
TREN, as moving to reflux conditions increased the para-alkyl- Also, within 2, both dpa units adopt a syn–syn conformation,
ation yield to 34% (entries 8 and 15). As aforementioned, whilst in contrast the two dpa units within 3 differ in their
using sub-stoichiometric quantities (10 mol%) of 1 gives a conformation. The first mirrors the syn–syn bonding observed
52% yield of the para-addition product.24 This enhanced within 2, whilst the second adopts a syn–anti conformation,
tBu2Zn reactivity is successfully reflected when using sub- connecting the two Na centres via two Na–N(amido) and two
stoichiometric quantities of PMDETA-containing metallo- Na–N(syn-pyridyl) interactions.
ligand 2, culminating in a 53% yield of para-addition (entry
Functioning as a protective ligand shield to partly cover an
18) and suggesting that the metalloligand can be recycled to a alkali metal cation (Fig. 7), tetradentate H6-TREN and Me6-
modest degree within this reaction system. Again the clear TREN have the capacity to act as monomerisation agents
preference for para-addition in nearly all of these reactions toward organometallic species including alkali metal benzyl44
suggests a similar mechanism though the activation barriers and 3,5-dimethylbenzyl46 salts, where the alkali metal is
appear to be higher for TMDAE and Me6-TREN as they require lithium, sodium or potassium. This ability is again reflected in
higher temperature to achieve greater para-addition.
the molecular structure of crystalline 4 (refer to ESI‡). The det-
Selection of PMDETA as the Lewis donor gives competitive rimental effect of H6-TREN (and Me6-TREN) upon reactivity
para-addition yields to those obtained using TMEDA (respect- could therefore be an artefact of too much steric shielding of
ive yields of 43% and 40% at ambient temperature, entries 4 the Na centre. As aforementioned, the solution NMR data
and 5). The fact that in the molecular structure of 2, PMDETA point to a co-complex between Nadpa and tBu2Zn with possibly
mimicks TMEDA by acting as a didentate donor towards the TREN ligands going on and off the sodium atoms so the
sodium, seems likely to have a bearing on these similar steric shielding could be manifested in monomeric structures
results. It seems unlikely that this asymmetric bonding mode in contrast to the dimeric structures seen with the other
is retained in C6D6 solution, as spectroscopic analysis of 2 in donors. These findings show that careful selection of the
this aromatic hydrocarbon medium reveals only one set of donor ligand is crucial to ensure that the reaction yields are
resonances corresponding to PMDETA. Alternatively, the tri- not compromised.
amine could be undergoing a fast exchange process. Further-
more, DOSY analysis of 2 with tBu2Zn in C6D6 solvent suggests
Evaluating the reaction dependency of the alkali metal
that C6D6, which can act as a Lewis donor, is in competitive
equilibrium with PMDETA. However, as 2 is crystallised from
the non-polar solvent hexane, it seems plausible that the
hexane solution state structure of 2 mirrors that of the solid
state structure.
Despite their close similarity, exchanging PMDETA for iso-
electronic TMDAE markedly reduces the yield of the para-
product (most noticeably from 43% to 2% at ambient tempera-
ture, entries 5 and 6). As could be expected based on the like-
ness of PMDETA and TMDAE, precursor complexes 2 and 3
adopt similar solid state structures: both are dimeric, each
potentially tridentate Lewis donor bonds to sodium in a diden-
tate fashion, and the two Na centres of each dimer are bridged
via the amido N of deprotonated dpa. However, there are two
As the best para-alkylation yields were obtained using PMDETA
as the donor, this system was chosen to probe the effect of
changing the alkali metal component of the metalloligand.
Remarkably, simply substituting Na by Li destroyed the regio-
selectivity of the reaction, giving a mixture of para- and
carbonyl-addition and benzhydrol in a relative ratio of approxi-
mately 1 : 1 : 1 (Table 2, entry 1). Surprisingly, the reaction was
almost completely suppressed when K was employed as the
alkali metal, with an insignificant 2% yield of para product
obtained (Table 2, entry 3) along with unreacted benzo-
phenone. Thus the size and Lewis acidity of the alkali metal
significantly impacts upon the yield and regioselectivity of the
reaction, with the optimum results obtained using the inter-
mediately sized Na. These findings bear some comparison to
Table 2 Reactions of alkali metal dipyridylamide, PMDETA, tBu2Zn and benzophenone in hexane solvent for 18 hours at ambient temperature
Product yielda,b (%)
Entry
Metal M
para-Addition (1, 6-)
Carbonyl-addition (1, 2-)
ortho-Addition (1, 4-)
Benzhydrol (H− addition)
Total
1
2
3
Li
Na
K
21
43
2
19
2
2
0
1
0
18
11
0
58
57
4
a Yields determined by 1H NMR spectroscopy using hexamethylbenzene (10 mol%) as an internal standard. b Conditions: Mdpa (2 mmol, M = Li,
Na or K), PMDETA (2 mmol), tBu2Zn (1 mmol), PhC(vO)Ph (1 mmol), hexane (8 mL) for 18 hours at ambient temperature.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 14409–14423 | 14415

View Article Online
Paper
Dalton Transactions
Table 3 Reaction of zincate [{(TMEDA)Na(dpa)}2ZntBu2] (1) with benzophenone for 18 hours at ambient temperature
Product yielda,b (%)
Entry
Solvent
para-Addition (1, 6-)
Carbonyl-addition (1, 2-)
ortho-Addition (1, 4-)
Benzhydrol (H− addition)
Total
1
2
3
Hexane
40
7
11
6
3
3
1
0
0
8
1
1
55
11
15
Hexane–THFc
THF
a Yields determined by 1H NMR spectroscopy using hexamethylbenzene (10 mol%) as an internal standard. b Conditions: Nadpa (2 mmol),
t
TMEDA (2 mmol), Bu2Zn, (1 mmol), PhC(vO)Ph (1 mmol), solvent (8 mL) for 18 hours at ambient temperature. c A 1 : 1 ratio of hexane–THF
solvent was used.
previous studies by Ishihara, where moving from the organo-
1H NMR spectroscopic analysis of 1 in d8-THF solution exhi-
t
lithium reagent nBuLi, to the related Grignard reagent bits a resonance attributed to the Bu groups at 0.87 ppm, an
nBuMgCl, changed the major product from that of carbonyl- upfield deviation of just 0.11 ppm in comparison to free
addition (in a 58% yield) to benzhydrol (in a 56% yield), but tBu2Zn (at 0.98 ppm in d8-THF solution). Upon moving to C6D6
significantly no para-addition was evident in either case.22
solvent, a more significant difference of 0.49 ppm is observed,
t
where this time the Bu resonance of 1 (at 1.57 ppm) lies
t
further downfield than that of free Bu2Zn (at 1.08 ppm). It
Exploring the reaction dependency of the solvent
therefore seems plausible that bulk THF solvent extrudes, or
t
mostly extrudes Bu2Zn from the [{(TMEDA)Na(dpa)}2ZntBu2]
Curious to find out whether or not increasing the solvent
polarity would improve the reaction yield, we next investigated
Lewis basic THF as the bulk solvent in the reaction of 1 with
benzophenone. However, a significant diminution of the para-
addition yield was observed, from 40% in hexane solvent to
11% in neat THF solution (Table 3, respective entries 1 and 3).
Analysis of crystalline 1 in C6D6 solvent suggests that TMEDA
coordinates to Na in this medium, as a significant upfield
shift is observed for the CH2-TMEDA and CH3-TMEDA reso-
co-complex; whereas in C6D6 solvent it remains fully attached.
t
t
The effect of substituting Bu2Zn by BuLi as the tert-butyl
source was also evaluated. It was found that substituting
t
tBu2Zn by BuLi as the tert-butyl ligand source in combination
with [(PMDETA)Na(dpa)] shuts down any para-addition, with
the yield falling from 48% to 0% (see ESI‡ for full details).
Turning from tertiary alkyl tBu to primary alkyl CH2SiMe3
nances (by 0.66 and 0.38 ppm respectively from their values in Previous studies of the alkylation of benzophenone using
free TMEDA, Table S4, entries 1 and 2‡). In contrast, when polar organometallic reagents have shown the product distri-
lone pair coordinating d8-THF was used as the solvent, the bution to be dependent on the nature of the alkyl group.57 For
n
TMEDA resonances attributed to TMEDA within 1 give close example, the reaction of PrMgBr with a range of para-substi-
agreement with those of non-coordinated TMEDA (Table S4, tuted benzophenone substrates gave only carbonyl-addition
entries 3 and 4‡).
and benzhydrol products; whereas ortho-addition was also
i
It has long been recognised in organoalkali metal chemistry observed using the branched isomer, PrMgBr. Intrigued by
that the solution state structure does not always reflect the these findings, we decided to investigate R2Zn as an alkyl
solid state structure as the former is often more complicated transfer reagent (where R is CH2SiMe3). On account of the sta-
involving multiple species and dynamic processes.1,52–56 Far bilising α-silyl group upon the trimethylsilylmethyl “carb-
t
from being an idle spectator, d8-THF participates through the anions”, R2Zn behaves as a more gentle reagent than Bu2Zn
displacement of TMEDA within 1, which is accompanied by when combined with
a sodium amide in a molecular
low reaction yields. It therefore seems logical that the micro- environment.58
scopic coordination at the sodium centre is key, and that the
In this experiment, freshly prepared R2Zn was added to
coordination of stoichiometric TMEDA in comparison to the metalloligand [(TMEDA)Na(dpa)]2 in a 1 : 1 ratio in hexane
effect of using bulk THF plays a pivotal role in the extent and solvent. Multinuclear (1H, 13C{H}) NMR spectroscopic analysis
selectivity of the alkylation reaction. Indeed, it could in fact be of the resultant white solid dissolved in C6D6 solvent revealed
the decoordination of TMEDA, which is paramount to the resonances attributable to dpa, TMEDA and R2Zn. Although
success of the nucleophilic addition, through creating a vacant the solubility of this white solid in C6D6 solvent is limited, its
site at the Lewis acidic sodium centre. It follows that this 1H NMR spectrum shows a 2 : 2 : 1 ratio of dpa–TMEDA–R2Zn
could facilitate the coordination of the Lewis donor benzo- resonances, which hints that the soluble product is contami-
phenone towards sodium, hence bringing it into close nated with a second, insoluble product still visible in the NMR
proximity with the activated tBu2Zn fragment.
sample (vide infra). These resonances mostly deviate upfield
14416 | Dalton Trans., 2014, 43, 14409–14423
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Table 4 Comparison of 1H NMR data (C6D6, 400.03 MHz, 300 K) for
TMEDA, R2Zn, dpa(H) and 5 (R = CH2SiMe3)
is likely to be a key factor towards the significantly diminished
alkylation yields.61
Chemical shift (ppm)
Evaluating the effect of adding TEMPO to the reaction system
Atom assignment
TMEDA
R2Zn
dpa(H)
5
Described recently as a “chameleonic ligand” due to its
different forms,62,63 the nitroxide TEMPO can perform as a
Lewis base either in its native radical form, or its reduced,
anionic form (TEMPO*). Often exploited as a radical trapping
reagent,64 TEMPO can also be used to provide insight into a
postulated reaction mechanism.65 For instance, [(TMEDA)Na-
(μ-TMP)(μ-tBu)ZntBu], which can regioselectively tert-butylate
benzophenone at the para-position,8 undergoes a redox reac-
tion with TEMPO to afford the crystalline triheteroleptic
sodium zincate [(TMEDA)Na(μ-TMP)(μ-TEMPO*)ZntBu]. In this
TMEDA (NCH3)
TMEDA (NCH2)
R [CH2Si(CH3)3]
R [CH2Si(CH3)3]
dpa (H1)
dpa (H2)
dpa (H3)
2.12
2.35
—
—
—
—
—
—
—
—
−0.62
0.08
—
—
—
—
—
—
—
—
8.16
6.75
7.53
7.72
1.86
1.82
−1.02
0.39
8.01
6.27
7.10
7.10
dpa (H4)
from those of the parent amine dpa(H), non-coordinated example, TEMPO has been reduced to its anion TEMPO* with
t
TMEDA and non-coordinated R2Zn (Table 4). For example, the concomitant formation of a Bu radical, which is thought to
t
respective ZnCH2Si(Me)3 and ZnCH2Si(CH3)3 resonances shift react with another Bu radical to produce iso-butane and iso-
by 0.40 ppm upfield and 0.31 ppm downfield when combined butene. Thus this study provides evidence that [(TMEDA)Na-
with the metalloligand [(TMEDA)Na(dpa)]2. This suggests that (μ-TMP)(μ-tBu)ZntBu], normally thought to be an anionic base
a co-complex of empirical formula [{(TMEDA)Na(dpa)}2ZnR2] source, can also combine through a SET mechanism.65 To
(5) is formed, indicating that it belongs to the same family compare with these findings, we incorporated TEMPO into our
as 1. However, this system departs from 1, as 1 is fully soluble alkylation system, to probe its effect upon the reaction
in C6D6 solvent and thus no insoluble by-products are outcome. It transpired that the addition of TEMPO to metallo-
t
observed.
ligand [(PMDETA)Na(dpa)]2, Bu2Zn and benzophenone com-
Upon the ambient temperature reaction of 5 with benzo- pletely destroyed the nucleophilic releasing ability of
phenone in hexane, only a trace (<1%) of the para-addition [(PMDETA)Na(dpa)]2·tBu2Zn. Accordingly, only an insignificant
product was observed, providing a marked reduction from the trace of the para product was observed (1% yield compared to
40% yield of para-addition obtained using the tert-butyl 48% in the absence of TEMPO). Multinuclear (1H, 13C{1H})
variant 1 under identical reaction conditions (Table 5, entries NMR spectroscopic analysis of the crude reaction mixture con-
1 and 2). Using 5 under reflux conditions did not improve the firmed that the known nitroxide compound TEMPO-tBu was
yields of para-59 and carbonyl-alkylation,60 although a low present, alongside unreacted benzophenone.66 Diagnostic of
yield of benzhydrol was obtained (8%, Table 5, entry 3). Pre- TEMPO-tBu, two distinctive singlets arising from the two
vious studies have suggested that the alkylation of benzo- inequivalent methyl groups of the nitroxide are observed at
t
phenone occurs through a single electron transfer (SET) 1.15 and 1.10 ppm. A sharp singlet attributed to the Bu group
reaction mechanism,57 and the product distribution has there- is also seen at 1.28 ppm. Completing the assignment, multi-
fore been linked to radical stability. If this alkylation reaction plets corresponding to β-TEMPO and γ-TEMPO hydrogen
were to proceed through a SET pathway, the lesser stability of atoms were observed at 1.48 and 1.30 ppm respectively in a
t
the primary R radical in comparison to the tertiary Bu radical 1H NMR spectrum recorded in CDCl3 solvent.
Table 5 Reaction of zinc reagents with benzophenone in hexane solvent for 18 hours
Product yielda,b (%)
Entry
R
Reaction temp. (°C)
para-Addition (1,6-)
Carbonyl- addition (1,2-)
Benzhydrol (H− addition)
Total
1
2
3
4
CH2SiMe3
tBu24
r.t.
r.t.
75
0
40
0
0
6
0
0
8
8
0
54
8
CH2SiMe3
tBu24
75
33
11
14
58
1
a Yields were determined by H NMR spectroscopy using hexamethylbenzene (10 mol%) as an internal standard. b Conditions: Nadpa (2 mmol),
TMEDA (2 mmol), R2Zn, (1 mmol), PhC(vO)Ph (1 mmol), hexane (8 mL) for 18 hours at ambient temperature.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 14409–14423 | 14417

View Article Online
Paper
Dalton Transactions
Mechanistic insights
Although TEMPO-tBu was detected in the reaction mixture,
this could either be produced from the interception of a Bu
t
The nature of the mechanism for the addition of polar organo- radical prior to its reaction with benzophenone, or generated
metallic reagents to ketones has posed a long standing ques- as a side product from reaction of TEMPO with tBu2Zn.
tion, where the two major contending mechanisms involve a
Although the reaction mechanism (or mechanisms)
single electron transfer (SET) reaction pathway, or a polar reac- appears convoluted and is not yet well understood, it is clear
tion mechanism. Evidence to support a SET pathway for the that the choice of ligand has a considerable impact upon the
alkylation of benzophenone has been accrued,67,68 including reaction yield. In spite of this, observation of para-addition as
the observation of the homo-coupled side-product benzo- the major product in each reaction system indicates that there
pinacol [Ph2C(OH)C(OH)Ph2].69 Furthermore, Ashby and is a common mechanism within these reactions. Previously,
t
Bowers observed the cyclisation of a radical probe incorporated tBu2Zn·TMEDA and Bu2Zn·(pyridine)2 were shown to be in-
into a Grignard reagent, which infers that an alkyl radical active towards the alkylation of benzophenone using our reaction
species was formed prior to addition to benzophenone.70
conditions, whilst the sodium zinc donor–acceptor complex
The inhibition of our alkylation reaction by TEMPO, com- [{(TMEDA)Na(dpa)}2ZntBu2] generates 40% of the para-
bined with the interception of a tBu radical in the formation of addition product at ambient temperature.24 In contrast, highly
TEMPO-tBu, initially implied that this reaction could follow a coordinating ligands such as H6-TREN have an adverse effect
SET reaction pathway. To gain further insight into the reaction upon the reaction yield, which could be an artefact of too
t
system, TEMPO was added to a hexane solution of Bu2Zn, much steric shielding of the Na centre, or that monomeric
upon which the characteristic red-orange colour associated complexes are inactive compared to dimeric complexes. Chan-
with the TEMPO radical disappeared. Cooling the resultant ging the bulk solvent medium from non-polar hexane to Lewis
colourless solution to −30 °C afforded a crystalline solid after donor THF, which displaces the donor ligands TMEDA,
24 hours. Unfortunately, severe disorder in the crystallographic PMDETA, TMDAE and H6-TREN in the solution state, leads to
structure made it impossible to refine the molecular structure a significant decrease in the alkylation yields. Furthermore,
satisfactorily; hence no structural data can be presented. spectroscopic evidence points towards the displacement of
However, NMR spectroscopic analysis of the crystalline tBu2Zn from the [(TMEDA)Na(dpa)]2 metalloligand in d8-THF
material in C6D6 solution confirmed the presence of anionic solvent. Preservation of the synergic sodium–zinc donor–
TEMPO* in addition to tert-butyl ligands in a 1 : 1 ratio and so acceptor complex therefore appears to be a dominant factor in
its formula therefore appears to be [(TEMPO*)Zn(tBu)] (yield the reactivity enhancement of the attacked tBu2Zn moiety.
0.08 g; 29%). Carmona has previously reported the structural
characterisation of the related heteroleptic dimer, [(TEMPO*)-
Zn(Et)]2.71 Present in the filtrate, TEMPO-tBu was also pro-
duced from the reaction of Bu2Zn with TEMPO, as evidenced
Conclusions
t
through NMR spectroscopy.
A synergic reactivity is clearly operative in these alkylation reac-
Given the propensity of TEMPO to act as a Lewis base tions. Co-complexation of sodium dipyridylamide metallo-
ligand towards a metal centre, it cannot be ruled out that ligands [(TMEDA)Na(dpa)]2, [(PMDETA)Na(dpa)]2, [(TMDAE)Na-
TEMPO shuts down the alkylation reaction by sequestering Zn (dpa)]2, or [(H6-TREN)Na(dpa)] promotes the reactivity of
from the sodium zinc co-complex. Henceforth, [(PMDETA)Na- tBu2Zn towards the para-alkylation of benzophenone. On its
t
(dpa)]2, tBu2Zn and TEMPO were combined in a 1 : 1 : 1 stoichio- own Bu2Zn is incapable of such reactivity. In terms of reactiv-
t
metric ratio in hexane solvent and the reaction mixture ity and selectivity sodium is the best partner for Bu2Zn, with
was analysed by NMR spectroscopy, which revealed resonances lithium less selective and potassium unreactive. Different
corresponding to TEMPO-tBu.66 This suggests that TEMPO donor ligands also produce para-alkylation (1,6-addition)
breaks up the synergic, sodium metalloligand/tert-butylzinc albeit to different extents. This shared selectivity implies a
partnership by extracting a butyl radical from the zinc centre, commonality in reaction mechanism though the precise
which could be a factor in the inhibition of the benzophenone nature of it remains to be established. With both
t
alkylation reaction. As a control comparison, the failure of the (Me3SiCH2)2Zn and BuLi failing to achieve similar selective
neutral heteroleptic zinc complex [(TEMPO*)ZntBu] to tert- alkylations of benzophenone, the scope of the reaction with
butylate benzophenone, at any position, was tested and con- respect to different alkyl transfer agents would appear limited,
firmed. It is also possible that TEMPO sequesters the alkali though extensions may be possible with other conjugated
metal from the metalloligand complex. Most TEMPO coordi- ketones and other unsaturated substrates. This will be
nation chemistry has been studied with d block,71–76
p
explored in future studies.
block77–80 and f block elements,81 however more recently,
structural studies of s block element-TEMPO structures have
been carried out,62,63,65 including the alkali metal complexes
(THF)2·[Li(TEMPO*)]4 and [(THF)Na(TEMPO*)]4.63 With the
Experimental
data accumulated thus far, it is therefore impossible to All reactions were performed under a protective argon atmos-
unequivocally state that our reaction follows a radical pathway. phere using standard Schlenk techniques. Hexane, toluene
14418 | Dalton Trans., 2014, 43, 14409–14423
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
and THF were dried by heating to reflux over sodium benzo- syringe, which produced a yellow solution. Cooling the resul-
phenone ketyl and distilled under nitrogen prior to use. tant solution to −30 °C for 48 hours afforded a crop of colour-
nBuNa82 and tBu2Zn7 were prepared according to literature less crystals [yield 0.13 g; 37% yield based upon the dpa(H)
methods. NMR spectroscopy was recorded on a Bruker DPX stoichiometry].
400 MHz spectrometer, operating at 400.03 MHz for 1H and
δH (400.03 MHz, d8-THF, 300 K) 2.17 ppm [12H, s, N(CH3)2-
3
100.59 MHz for 13C, or a Bruker DPX 500 MHz spectrometer, TMDAE], 2.39 (4H, t, JHH 6.1 Hz, NCH2-TMDAE), 3.46 (4H, t,
operating at 500.13 MHZ for 1H and 125.76 MHz for 13C. 3JHH 6.1 Hz, OCH2-TMDAE), 6.38 [2H, ddd, JHH 6.8, 5.0 Hz,
3
Dipyridylamine and benzophenone were purchased from 4JHH 1.2 Hz, H2-dpa], 7.30 [2H, br. d, 3JHH 8.4 Hz, H4-dpa] and
Sigma-Aldrich and were used as received. Single crystal X-ray 7.30 [2H, ddd, 3JHH 8.4, 6.8 Hz, 4JHH 2.0 Hz, H3-dpa], 7.99 [2H,
3
4
5
diffraction measurements were made with Oxford Diffraction ddd, JHH 5.0 Hz, JHH 2.0 Hz, JHH 1.0 Hz, H1-dpa] ppm. δC{H}
instruments and graphite monochromated radiation at 123(2) (100.59 MHz, C6D6, 300 K) 46.3 [N(CH3)2-TMDAE], 59.9 (NCH2-
K. All refinements were against F2 and to convergence and TMDAE), 70.6 (OCH2-TMDAE), 112.3 (C2-dpa), 112.6 (C4-dpa),
used programs from the SHELX family.83 Selected crystallo- 136.9 (C3-dpa), 148.8 (C1-dpa) and 163.2 (C5-dpa) ppm.
graphic and refinement details are given in the ESI.‡
Crystal data for 3: C36H56N10Na2O2, Mr = 706.89, ortho-
rhombic, space group Pbc21, a = 15.8743(8), b = 12.4146(7), c =
19.8597(11) Å, V = 3913.8(3) Å3, Z = 4, μ = 0.096 mm−1, 36 891
Preparation of [(PMDETA)Na(dpa)]2, 2
Mr = 732.98 g. nBuNa (0.08 g, 1 mmol) was suspended in reflections, 9425 unique, Rint 0.0617. With the aid of restraints
hexane (15 ml). To this white suspension, dpa(H) (0.17 g, on bond lengths and relative displacement parameters, one
1 mmol) was added and the resultant suspension was left to TMDAE ligand was modelled as disordered over two sites.
stir for 45 minutes. Subsequent addition of PMDETA (0.42 ml, Final refinement by full-matrix least squares on F2 gave R =
2 mmol) generated a pale grey solution which, upon transferral 0.0769 (F, 6569 obs. data only) and Rw = 0.2266 (F2, all data),
to the refrigerator at −30 °C, deposited colourless crystalline GOF = 1.014.
material after 24 hours [yield 0.30 g; 81% yield based upon the
dpa(H) stoichiometry].
Preparation of [(H6-TREN)Na(dpa)], 4
δH (400.03 MHz, d8-THF, 300 K) 2.14 [12H, s, N(CH3)2- Mr = 339.41 g. To a freshly prepared suspension of nBuNa
PMDETA], 2.18 (3H, s, NCH3-PMDETA), 2.42 and 2.31 [4H, t, (0.08 g, 1 mmol) in hexane (15 ml), dpa(H) (0.17 g, 1 mmol)
3
3JHH 7.5 Hz, NCH2-PMDETA], 6.20 [2H, ddd, JHH 6.3, 5.0 Hz, was added and the reaction mixture was allowed to stir for
3
4
4JHH 1.5 Hz, H2-dpa], 7.07 [2H, ddd, JHH 8.4 Hz, JHH 1.5 Hz, 1 hour. H6-TREN (0.15 ml, 1 mmol) was added via syringe, fol-
3
4
5JHH 1.0 Hz, H4-dpa], 7.14 [2H, ddd, JHH 8.4, 6.3 Hz, JHH 2.0 lowed by toluene (15 ml) with gentle heating to produce a
3
4
5
Hz, H3-dpa], 7.92 [2H, ddd, JHH 5.0 Hz, JHH 2.0 Hz, JHH 1.0 yellow-green solution. Gradual cooling to ambient temperature
Hz, H1-dpa] ppm. δC{H} (100.59 MHz, C6D6, 300 K) 43.3 (NCH3- yielded colourless crystals [yield 0.31 g, 91% yield based
PMDETA), 46.2 [N(CH3)2-PMDETA] and, 58.9 and 57.4 (NCH2- upon the dpa(H) stoichiometry]. Despite repeated attempts,
PMDETA), 110.3 (C2-dpa), 112.5 (C4-dpa), 136.5 (C3-dpa), efforts to grow X-ray quality crystals of the complex were
149.1 (C1-dpa), 167.2 (C5-dpa) ppm.
unsuccessful.
Due to the extreme air- and moisture-sensitivity of this com-
pound and compounds 3, 4, and 6, satisfactory elemental TREN), 2.40 [6H, t, JHH 6.0 Hz, NCH2-H6-TREN], 2.65 [6H, t,
δH (400.03 MHz, d8-THF, 300 K) 1.13 (6H, br. s, NH2-H6-
3
3
microanalysis data could not be obtained.
3JHH 6.0 Hz, H2NCH2-TREN], 6.22 [2H, ddd, JHH 8.1, 5.0 Hz,
Crystal data for 2: C38H62N12Na2, Mr = 732.98, triclinic, 4JHH 1.2 Hz, H2-dpa], 7.15 (4H, br. m, H4-dpa and H3-dpa),
3
4
5
ˉ
space group P1, a = 10.3495(13), b = 11.0583(11), c = 18.5966(18) 7.92 [2H, ddd, JHH 5.0 Hz, JHH 1.8 Hz, JHH 1.3 Hz, H1-dpa]
Å, α = 101.113(8), β = 96.726(9), γ = 90.084(9)°, V = 2073.4(4) ppm. δC{H} (100.59 MHz, C6D6, 300 K) 41.0 (NCH2-H6-TREN),
Å3, Z = 2, μ = 0.754 mm−1. The crystals were treated as 59.1 (NCH2-H6-TREN), 109.9 (C2-dpa), 113.0 (C4-dpa), 136.2
twinned with the two parts related by the matrix −0.9813 (C3-dpa), 149.0 (C1-dpa), 167.0 (C5-dpa) ppm.
0.0624 0.0535 0.0681 1.0029 0.0203 −0.1850 −0.5925 −1.0129.
Preparation of [Zn(dpa)2], 6
Processing the data gave a hklf 5 formatted reflection file with
17 153 reflections. The twin ratio was refined to 0.532(3) : Mr = 405.76 g. Dpa(H) (0.34 g, 2 mmol) was dissolved in a
0.468(3). With the aid of restraints on bond lengths and mixed hexane (10 mL)–toluene (20 mL) solvent system. This
relative displacement parameters, one PMDETA ligand was colourless solution was added to a solution of freshly prepared
modelled as disordered over two sites. Final refinement by Zn(CH2SiMe3)2 (10 mL of a 0.2 M solution in hexane, 2 mmol).
full-matrix least squares on F2 gave R = 0.1510 (F, 8270 obs. The resultant pale yellow solution deposited a crop of colour-
data only) and Rw = 0.4457 (F2, all data), GOF = 1.167.
less crystals within 24 h [yield 0.08 g; 10% yield based upon
the dpa(H) stoichiometry i.e., with a maximum yield of 50%].
Preparation of [(TMDAE)Na(dpa)]2, 3
Rational synthesis
Mr = 706.89 g. Dpa(H) (0.17 g, 1 mmol) was introduced to a
freshly prepared suspension of nBuNa (0.08 g, 1 mmol) in Dpa(H) (0.34 g, 2 mmol) was dissolved in a mixed toluene
hexane (20 ml) and the reaction mixture was stirred for 1 hour. (20 mL)–hexane (10 mL) solvent system. This colourless solu-
TMDAE (0.38 ml, 2 mmol) was subsequently added via tion was slowly added to a hexane solution of Zn(CH2SiMe3)2
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 14409–14423 | 14419

View Article Online
Paper
Dalton Transactions
(10 mL of a 0.1 M solution in hexane, 1 mmol) via syringe. ambient temperature or at 75 °C (see Table 1) prior to work up
A colourless powder was deposited after 24 h (crude yield according to Method A.
0.33 g, 81%).
Variation of donor ligands – catalytic conditions
Crystal data for 6: C20H16N6Zn, Mr = 405.76, monoclinic,
space group C2/c, a = 22.5759(4), b = 9.5830(2), c = 16.0729(3)
nBuNa (0.08 g, 1 mmol) was suspended in hexane (30 mL) and
dpa(H) (0.17 mL, 1 mmol) was added. The reaction mixture
was allowed to stir for 45 minutes. Subsequently, the donor
ligand was added [DMAP (0.12 g, 1 mmol); TMEDA (0.15 mL,
1 mmol); PMDETA (0.21 mL, 1 mmol); TMDAE (0.14 mL,
Å, β = 99.734(2)°, V = 3427.23(12) Å3, Z = 8, μ = 2.133 mm−1
,
10 393 reflections, 3351 unique, Rint 0.0186, final refinement
by full-matrix least squares on F2 gave R = 0.0344 (F, 3102 obs.
data only) and Rw = 0.1000 (F2, all data), GOF = 1.074.
Unfortunately, the lack of solubility of 6 in common NMR
spectroscopic solvents (C6D6, d8-THF, d5-pyridine and
d6-DMSO) prevented satisfactory NMR spectroscopic analysis
from being obtained.
t
1 mmol)]. Following this, Bu2Zn (10 mL of a 0.5 M solution
t
of Bu2Zn in hexane, 5 mmol), and benzophenone (0.90 g,
5 mmol) were added. The reaction mixture was stirred under
reflux conditions for 18 h, prior to work-up according to
Method A.
Method A: work-up procedure
To the reaction mixture, deionised water (10 mL), 2 M HCl
(20 mL) and diethyl ether (20 mL) were added. The organic
layer was separated from the aqueous layer and the aqueous
layer was washed with diethyl ether–hexane (3 × 20 mL).
Magnesium sulfate was used to dry the combined organic
layers. Solvent was removed in vacuo and the crude residue was
spiked with 10 mol% hexamethylbenzene [0.0162 g, 0.1 mmol
for 1 mmol scale (stoichiometric) reactions; 0.081 g, 0.5 mmol
Variation of alkali metal – reaction with Lidpa, Nadpa or Kdpa
Following the introduction of an organometallic reagent R′M
[R′M = nBuLi (1.25 mL, 1.6 M in hexanes solution, 2 mmol);
nBuNa (0.16 g, 2 mmol); or KR (0.25 g, 2 mmol) (where R is
CH2SiMe3)] to hexane solvent (6 mL), dpa(H) (0.34 g, 2 mmol)
was added via a solid addition tube. The resultant beige sus-
pension was then stirred for 45 minutes at ambient tempera-
ture. To this, PMDETA was injected (0.42 mL, 2 mmol),
1
for 5 mmol scale (catalytic) reactions]. H NMR spectroscopic
t
followed by a hexane solution of Bu2Zn (2 mL of a 0.5 M solu-
analysis was performed in CDCl3 solvent and the relative yields
of 2- and 4-tert-butylbenzophenone, benzhydrol83 and
diphenyl-tert-butylmethanol84 were determined by relative
integration. Spectroscopic data show resonances that are in
good agreement with the reference standard of commercially
available samples of 2-tert-butylbenzophenone and 4-tert-
butylbenzophenone.
tion, 1 mmol). Benzophenone (0.18 g, 1 mmol) was sub-
sequently added and the reaction was stirred for 18 hours at
ambient temperature. The reaction was worked up following
the procedure outlined in Method A.
Variation of solvent – reaction in THF or THF–hexane
nBuNa (0.16 g, 2 mmol) was suspended in hexane (6 mL) and
dpa(H) (0.34 g, 2 mmol) was added. The resultant beige sus-
pension was allowed to stir for 45 minutes at ambient temp-
erature, following which, all solvent was removed in vacuo.
THF (6 mL), TMEDA (0.30 mL, 2 mmol), and tBu2Zn (2 mL of a
0.5 M solution in THF, 1 mmol) were then added.a Following
the addition of benzophenone (0.18 g, 1 mmol), the reaction
was allowed to stir at ambient temperature for 18 h, over
which time a colour change from orange to green was
observed. The reaction was worked up as per Method
Control reactions
Stoichiometric conditions: a standard solution of tBu2Zn
(2 mL of a 0.5 M solution in hexane, 1 mmol) was transferred
to a Schlenk flask under argon atmosphere. A further 6 mL of
hexane was added, followed by benzophenone (0.18 g,
1 mmol). The reaction mixture was allowed to stir at ambient
temperature for 18 h prior to work-up as per Method A.
t
Catalytic conditions: a standard solution of Bu2Zn (10 mL
of a 0.5 M solution in hexane, 5 mmol) was diluted with a
further 30 mL of hexane. Benzophenone (0.90 g, 5 mmol) was
added and the reaction mixture was allowed to stir for 18 h
under reflux conditions prior to work-up as per Method A.
a
A. When the reaction was performed in a mixed THF–hexane
solvent system, hexane (4 mL), TMEDA (0.30 mL, 2 mmol),
t
THF (2 mL) and Bu2Zn (2 mL of a 0.5 M solution in THF,
1 mmol) were injected.
Variation of donor ligands – stoichiometric conditions
Variation of tert-butyl anion source – reaction with tBuLi or
nBuNa (0.16 g, 2 mmol) was suspended in hexane (6 mL) and
dpa(H) (0.34 g, 2 mmol) was added. The resultant beige sus-
tBu2Zn
pension was allowed to stir for 45 minutes at ambient temp- Dpa(H) (0.34 g, 2 mmol) was added to a freshly prepared sus-
erature. To this, the donor ligand was injected [DMAP (0.24 g, pension of nBuNa (0.16 g, 2 mmol) in hexane (6 mL). PMDETA
2 mmol); TEMPO (0.31 g 2 mmol); TMEDA (0.30 mL, 2 mmol); (0.42 mL, 2 mmol) was subsequently introduced, followed by
PMDETA (0.42 mL, 2 mmol); TMDAE (0.38 mL, 2 mmol); H6- either tBu2Zn (2 mL of a 0.5 M solution in hexane, 1 mmol), or
TREN (0.30 mL, 2 mmol); Me6-TREN (0.42 mL, 2 mmol)]. A tBuLi (1.18 mL of a 1.7 M solution in pentane, 2 mmol). Fol-
t
hexane solution of Bu2Zn (2 mL of a 0.5 M solution, 1 mmol) lowing the addition of benzophenone (0.18 g, 1 mmol), the
was subsequently added, followed by benzophenone (0.18 g, reaction was stirred at ambient temperature for 18 hours prior
1 mmol) and the reaction was stirred for 18 h, either at to work up according to Method A.
14420 | Dalton Trans., 2014, 43, 14409–14423
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Variation of R2Zn
(0.18 g, 1 mmol) was added. The reaction mixture was stirred
for 18 h at ambient temperature prior to work up as per
Method A.
nBuNa (0.16 g, 2 mmol) was suspended in hexane (6 mL) and
dpa(H) (0.34 g, 2 mmol) was added. After the reaction mixture
was stirred for 1 h, TMEDA (0.30 mL, 2 mmol) was added, fol-
lowed by R2Zn (2 mL of a 0.5 M solution in hexane, 1 mmol).
Following the addition of benzophenone (0.18 g, 1 mmol), the
reaction was allowed to stir for 18 h, either at ambient temp- Acknowledgements
erature or under reflux conditions (Table 5) prior to work up
according to Method A.
This work was generously supported by the UK Engineering
and Physical Science Research Council (award no. EP/K001183/
1), George Fraser (scholarship award to J.A.G.) and the Royal
Society/Wolfson Foundation (research merit award to R.E.M.).
We would also like to thank Eva Hevia, Charles T. O’Hara and
Stuart D. Robertson for stimulating discussions.
Addition of TEMPO
nBuNa (0.16 g, 2 mmol) was suspended in hexane (6 mL) and
dpa(H) (0.34 g, 2 mmol) was added. PMDETA (0.42 mL,
t
2 mmol) was then injected, followed by Bu2Zn (2 mL of a
0.5 M solution in hexane, 1 mmol). After the addition of
benzophenone (0.18 g, 1 mmol) and TEMPO (0.31 g, 2 mmol),
the reaction was allowed to stir at ambient temperature for
18 hours prior to work up following the procedure outlined in
Method A. 1H NMR spectroscopic resonances corresponding
to TEMPO-tBu were observed.64
Notes and references
1 H. J. Reich, Chem. Rev., 2013, 113, 7130–7178.
2 A. Harrison-Marchand and F. Mongin, Chem. Rev., 2013,
113, 7470–7562.
3 A. Harrison-Marchand and F. Mongin, Chem. Rev., 2013,
113, 7563–7727.
δH (400.03 MHz, CDCl3, 300 K) 1.10 and 1.15 (6H, s,
t
Me-TEMPO), 1.28 (9H, s, Bu), 1.30 (2H, m, γ-TEMPO), 1.48
(4H, m, β-TEMPO) ppm. δC{H} (100.60 MHz, CDCl3, 300 K) 17.2
(γ-TEMPO), 20.4 (Me-TEMPO), 29.7 (tBu), 34.8 (Me-TEMPO),
40.9 (β-TEMPO), 59.1 (α-TEMPO), 77.3 (quat. tBu) ppm.
4 R. E. Mulvey and S. D. Robertson, Angew. Chem., Int. Ed.,
2013, 52, 11470–11487.
5 B. Haag, M. Mosrin, H. Ila, V. Malakhov and P. Knochel,
Angew. Chem., Int. Ed., 2011, 50, 9794–9824.
Reaction of tBu2Zn and TEMPO
TEMPO (0.16 g, 1 mmol) was added to a hexane (5 mL) solu-
6 R. E. Mulvey, Dalton Trans., 2013, 42, 6676–6693.
7 P. C. Andrikopoulos, D. R. Armstrong, H. R. L. Barley,
W. Clegg, S. H. Dale, E. Hevia, G. W. Honeyman,
A. R. Kennedy and R. E. Mulvey, J. Am. Chem. Soc., 2005,
127, 6184–6185.
8 E. Hevia, G. W. Honeyman, A. R. Kennedy and
R. E. Mulvey, J. Am. Chem. Soc., 2005, 127, 13106–13107.
9 D. R. Armstrong, J. A. Garden, A. R. Kennedy,
S. M. Leenhouts, R. E. Mulvey, P. O’Keefe, C. T. O’Hara and
A. Steven, Chem. – Eur. J., 2013, 19, 13492–13503.
10 A. R. Kennedy, R. E. Mulvey and R. B. Rowlings, J. Am.
Chem. Soc., 1998, 120, 7816–7824.
t
tion of Bu2Zn (0.18 g, 1 mmol), upon which the characteristic
red-orange colour associated with TEMPO disappeared. After
cooling the resultant colourless solution to −30 °C, a crystal-
line solid, [(TEMPO*)Zn(tBu)], was deposited after 24 hours
(yield 0.08 g; 29%).
δH (500.13 MHz, C6D6, 300 K) 1.18 (12H, s, Me-TEMPO),
1.28 (2H, m, γ-TEMPO), 1.39 (4H, m, β-TEMPO), 1.41 (9H, s,
tBu) ppm. δC{H} (125.76 MHz, C6D6, 300 K) 17.7 (γ-TEMPO),
25.7 (Me-TEMPO), 34.0 (tBu), and 40.4 (β-TEMPO) ppm.
Due to the extreme air- and moisture-sensitivity of this
compound, satisfactory elemental microanalysis data could
not be obtained.
11 P. García-Álvarez, D. V. Graham, E. Hevia, A. R. Kennedy,
J. Klett, R. E. Mulvey, C. T. O’Hara and S. Weatherstone,
Angew. Chem., Int. Ed., 2008, 47, 8079–8081.
Reaction of 2, tBu2Zn and TEMPO
Dpa(H) (0.34 g, 2 mmol) was added to a freshly prepared sus-
pension of nBuNa (0.16 g, 2 mmol) in hexane (8 mL). Sub-
sequently, PMDETA (0.42 mL, 2 mmol), Bu2Zn (2 mL of a
12 E. Crosbie, P. Garcia-Alvarez, A. R. Kennedy, J. Klett,
R. E. Mulvey and S. D. Robertson, Angew. Chem., Int. Ed.,
2010, 49, 9388–9391.
t
0.5 M solution in hexane, 1 mmol), and TEMPO (0.31 g,
2 mmol) were introduced to the reaction mixture. 1H and
13C{H} NMR spectroscopic data were consistent with those
observed for [(TEMPO*)Zn(tBu)] (vide supra).
13 R. Campbell, E. Crosbie, A. R. Kennedy, R. E. Mulvey,
R. A. Naismith and S. D. Robertson, Aust. J. Chem., 2013,
66, 1189–1201.
14 J. M. L’Helgoual’ch, G. Bentabed-Ababsa, F. Chevallier,
M. Yonehara, M. Uchiyama, A. Derdour and F. Mongin,
Chem. Commun., 2008, 5375–5377.
Reaction of [(TEMPO)Zn(tBu)] and benzophenone
Following the procedure outlined above on a 3 mmol scale, the 15 K. Snegaroff, J. M. L’Helgoual’ch, G. Bentabed-Ababsa,
t
stoichiometric combination of Bu2Zn and TEMPO was per-
formed. Crystalline [(TEMPO)Zn(tBu)] was isolated (0.25 g,
0.9 mmol), dissolved in hexane (8 mL), and benzophenone
T. T. Nguyen, F. Chevallier, M. Yonehara, M. Uchiyama,
A. Derdour and F. Mongin, Chem. – Eur. J., 2009, 15,
10280–10290.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 14409–14423 | 14421

View Article Online
Paper
Dalton Transactions
16 D. R. Armstrong, A. R. Kennedy, R. E. Mulvey, 41 A. D. McLean and G. S. Chandler, J. Chem. Phys., 1980, 72,
J. A. Parkinson and S. D. Robertson, Chem. Sci., 2012, 3,
5639–5648.
2700–2707.
42 J. H. N. Buttery, N. C. Plackett, B. W. Skelton,
C. R. Whitaker and A. H. White, Z. Anorg. Allg. Chem., 2006,
632, 1856–1869.
17 B. D. Murray and P. P. Power, Inorg. Chem., 1984, 23, 4584–
4588.
18 J. Garcia-Álvarez, A. R. Kennedy, J. Klett and R. E. Mulvey, 43 D. M. Cousins, M. G. Davidson, C. J. Frankis, D. Garciá-
Angew. Chem., Int. Ed., 2007, 46, 1105–1108. Vivó and M. F. Mahon, Dalton Trans., 2010, 39, 8278–8280.
19 M. Hatano, T. Matsumura and K. Ishihara, Org. Lett., 2005, 44 M. G. Davidson, D. García-Vivó, A. R. Kennedy, R. E. Mulvey
7, 573–576. and S. D. Robertson, Chem. – Eur. J., 2011, 17, 3364–3369.
20 M. Hatano, S. Suzuki and K. Ishihara, J. Am. Chem. Soc., 45 A. R. Kennedy, R. E. Mulvey, C. T. O’Hara, G. M. Robertson
2006, 128, 9998–9999.
and S. D. Robertson, Angew. Chem., Int. Ed., 2011, 50, 8525–
21 M. Hatano and K. Ishihara, Synthesis, 2008, 1647–1675.
8528.
22 M. Hatano, S. Suzuki and K. Ishihara, Synlett, 2010, 321– 46 D. R. Armstrong, M. G. Davidson, D. García-Vivó,
324.
A. R. Kennedy, R. E. Mulvey and S. D. Robertson, Inorg.
Chem., 2013, 52, 12023–12032.
23 J. J. Crawford, B. J. Fleming, A. R. Kennedy, J. Klett,
C. T. O’Hara and S. A. Orr, Chem. Commun., 2011, 47, 3772– 47 N. D. R. Barnett, R. E. Mulvey, W. Clegg and P. A. O’Neil,
3774. J. Am. Chem. Soc., 1993, 115, 1573–1574.
24 D. R. Armstrong, J. A. Garden, A. R. Kennedy, R. E. Mulvey 48 M. A. Nichols and P. G. Williard, J. Am. Chem. Soc., 1993,
and S. D. Robertson, Angew. Chem., Int. Ed., 2013, 52, 7190–
7193.
25 K. Gregory, P. v. R. Schleyer and R. Snaith, Adv. Inorg.
Chem., 1991, 37, 47–142.
115, 1568–1572.
49 T. Stey and D. Stalke, The Chemistry of Organolithium Com-
pounds, John Wiley and Sons, Ltd, Chichester, 2004.
50 V. H. Gessner, C. Daschlein and C. Strohmann, Chem. –
Eur. J., 2009, 15, 3320–3334.
26 D. Barr, W. Clegg, L. Cowton, L. Horsburgh,
F. M. Mackenzie and R. E. Mulvey, J. Chem. Soc., Chem. 51 M. D. Rausch and D. J. Ciappenelli, J. Organomet. Chem.,
Commun., 1995, 891–892. 1967, 10, 127–136.
27 I. Cragg-Hine, M. G. Davidson, A. J. Edwards, P. R. Raithby 52 D. B. Collum, Acc. Chem. Res., 1992, 25, 448–454.
and R. Snaith, J. Chem. Soc., Dalton Trans., 1994, 2901– 53 D. B. Collum, Acc. Chem. Res., 1993, 26, 227–234.
2902.
54 J. F. Remenar, B. L. Lucht, D. Kruglyak, F. E. Romesburg,
J. H. Gilchirst and D. B. Collum, J. Org. Chem., 1997, 62,
5748–5754.
28 A. R. Kennedy, J. Klett, C. T. O’Hara, R. E. Mulvey and
G. M. Robertson, Eur. J. Inorg. Chem., 2009, 5029–5035.
29 J. A. Garden, A. R. Kennedy, R. E. Mulvey and 55 H. Oulyadi, C. Fressigné, Y. Yuan, J. Maddaluno and
S. D. Robertson, Dalton Trans., 2011, 40, 11945–11954.
A. Harrison-Marchand, Organometallics, 2012, 31, 4801–
30 M. Karl, G. Seybert, W. Massa, K. Harms, S. Agarwal,
4809.
R. Maleika, W. Stelter, A. Greiner, W. H. B. Neumüller and 56 A. Corruble, J.-Y. Valnot, J. Maddaluno, Y. Prigent,
K. Dehnicke, Z. Anorg. Allg. Chem., 1999, 625, 1301–1309.
31 F. H. Allen, Search performed 06/01/14, Version 1.15, Acta
Crystallogr. Sect. B: Struct. Sci., 2002, 58, 380–388.
32 P. C. Andrews, W. Clegg and R. E. Mulvey, Angew. Chem.,
Int. Ed. Engl., 1990, 29, 1440–1441.
D. Davoust and P. Duhamel, J. Am. Chem. Soc., 1997, 119,
10042–10048.
57 K. Maruyama, Y. Matano and T. Katagiri, J. Phys. Org.
Chem., 1991, 4, 501–519.
58 A. R. Kennedy, J. Klett, R. E. Mulvey and D. S. Wright,
Science, 2009, 326, 706–708.
33 D. R. Armstrong, W. Clegg, R. P. Davies, S. T. Liddle,
D. J. Linton, P. R. Raithby, R. Snaith and A. E. H. Wheatley, 59 A. Fürstner and G. Seidel, Tetrahedron, 1995, 51, 11165–11176.
Angew. Chem., Int. Ed., 1999, 38, 3367–3370.
60 H. Takahashi, K. M. Hossain, Y. Nishihara, T. Shibata and
34 S. T. Liddle and W. Clegg, Polyhedron, 2003, 22, 3507–3513.
K. Takagi, J. Org. Chem., 2006, 71, 671–675.
35 W. Clegg and S. T. Liddle, Acta Crystallogr., Sect. E: Struct. 61 D. C. Nonhebel and J. C. Walton, Free Radical Chemistry,
Rep. Online, 2004, 60, 1492–1494. Cambridge University Press, New York, 1974.
36 S. T. Liddle and W. Clegg, J. Chem. Soc., Dalton Trans., 62 G. C. Forbes, A. R. Kennedy, R. E. Mulvey and
2001, 402–408. P. J. A. Rodger, Chem. Commun., 2001, 1400–1401.
37 S. T. Liddle and W. Clegg, Polyhedron, 2002, 21, 2451– 63 L. Balloch, A. M. Drummond, P. García-Alvarez,
2455.
D. V. Graham, A. R. Kennedy, J. Klett, R. E. Mulvey,
C. T. O’Hara, P. J. A. Rodger and I. D. Rushworth, Inorg.
Chem., 2009, 48, 6934–6944.
38 S. T. Liddle, W. Clegg and C. A. Morrison, Dalton Trans.,
2004, 2514–2525.
39 J. R. Allan, G. M. Baillie, L. A. Macindoe, A. J. Blake, 64 T. Vogler and A. Studer, Synthesis, 2008, 1979–1993.
H. J. Bowley and D. L. Gerrard, Acta Crystallogr., Sect. C: 65 D. R. Armstrong, L. Balloch, J. J. Crawford, B. J. Fleming,
Cryst. Struct. Commun., 1988, 44, 1833–1834.
40 W. Kohn, A. D. Becke and R. G. Parr, J. Phys. Chem., 1996,
100, 12974–12980.
L. M. Hogg, A. R. Kennedy, J. Klett, R. E. Mulvey,
C. T. O’Hara, S. A. Orr and S. D. Robertson, Chem.
Commun., 2012, 48, 1541–1543.
14422 | Dalton Trans., 2014, 43, 14409–14423
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
66 G. Sorin, R. M. Mallorquin, Y. Contie, A. Baralle, 75 M. K. Mahanthappa, K. W. Huang, A. P. Cole and
M. Malacria, J. P. Goddard and L. Fensterbank, Angew.
Chem., Int. Ed., 2010, 49, 8721–8723.
67 T. Holm and I. Crossland, Acta Chem. Scand., 1971, 25,
59–69.
68 C. C. L. Thum, G. N. Khairallah and R. A. J. O’Hair, Angew.
Chem., Int. Ed., 2008, 47, 9118–9121.
69 C. Blomberg, R. M. Salinger and H. S. Mosher, J. Org.
Chem., 1969, 34, 2385–2388.
R. M. Waymouth, Chem. Commun., 2002, 502–503.
76 D. J. Mindiola, R. Waterman, D. M. Jenkins and
G. L. Hillhouse, Inorg. Chim. Acta, 2003, 345, 299–308.
77 T. Iwamoto, H. Masuda, S. Ishida, C. Kabuto and M. Kira,
J. Am. Chem. Soc., 2003, 125, 9300–9301.
78 A. Naka, N. J. Hill and R. West, Organometallics, 2004, 23,
6330–6332.
79 G. H. Spikes, Y. Peng, J. C. Fettinger, J. Steiner and
P. P. Power, Chem. Commun., 2005, 6041–6043.
70 E. C. Ashby and J. J. R. Bowers, J. Am. Chem. Soc., 1981,
103, 2242–2250.
80 C. Jones and R. P. Rose, New J. Chem., 2007, 31, 1484–1487.
71 A. Grirrane, I. Resa, A. Rodríguez, E. Carmona, E. Alvarez, 81 W. J. Evans, J. M. Perotti, R. J. Doedens and J. W. Ziller,
E. Gutiérrez-Puebla, A. Monge, A. Galindo, D. del Río and
R. A. Andersen, J. Am. Chem. Soc., 2007, 129, 693–703.
72 M. H. Dickman and R. J. Doedens, Inorg. Chem., 1982, 21,
682–684.
73 P. Jaitner, W. Huber, G. Hunter and O. Scheidsteger,
J. Organomet. Chem., 1983, 259, C1–C5.
Chem. Commun., 2001, 2326–2327.
82 C. Schade, W. Bauer and P. v. R. Schleyer, J. Organomet.
Chem., 1985, 295, 25–28.
83 G. M. Sheldrick, Acta Crystallogr., Sect. A: Fundam. Crystal-
logr., 2008, 64, 112–122.
84 N. J. Findlay, S. R. Park, F. Schoenebeck, E. Cahard,
S. Zhou, L. E. A. Berlouis, M. D. Spicer, T. Tuttle and
J. A. Murphy, J. Am. Chem. Soc., 2010, 132, 15462–15464.
74 J. Laugier, J. M. Latour, A. Caneschi and P. Rey, Inorg.
Chem., 1991, 30, 4474–4477.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 14409–14423 | 14423