Aggregation behaviour of thallium(I) β-diketiminates

Article abstract of DOI:10.1039/b607887g

Modulation of the steric requirements of a number of N-aryl β-diketiminate ligands results in the isolation of a variety of Tl(i) compounds with different stabilities and nuclearities. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b607887g

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Aggregation behaviour of thallium(I) b-diketiminates
Michael S. Hill,*a Ruti Pongtavornpinyoa and Peter B. Hitchcockb
Received (in Cambridge, UK) 2nd June 2006, Accepted 29th June 2006
First published as an Advance Article on the web 31st July 2006
DOI: 10.1039/b607887g
and VII. Although the resolution of this synthetic challenge
Modulation of the steric requirements of a number of N-aryl
b-diketiminate ligands results in the isolation of a variety of Tl(I)
compounds with different stabilities and nuclearities.
has proved problematical, it has highlighted a marked, but
somewhat capricious dependence of complex stability upon ligand
topology.
The chemistry of formally monovalent and charge neutral group
13 species, XLE, I, (X = anionic substituent, L = neutral two
electron donor, E = group 13 element) with isoelectronic
relationships to the much more widely studied group 14 diyls,
X2E9 (X = anionic substituent, E9 = group 14 element) has
advanced enormously since the turn of the century.1 Central to the
development of neutral species of the form, I, has been the
application of chelating N-donor monoanions. Until very
recently,2,3 examples were exclusively confined to the widely
employed b-diketiminate class of ligand, with either bulky silyl or,
more commonly, aryl N-substituents.4–7 Given the increasing
stability of the univalent state as each group is descended (the
classical ‘inert pair’ effect), the discovery of compounds of types
II–V followed a somewhat unexpected chronology in that the most
recently reported contained the heaviest members of the series,
indium and thallium.6,7 Our own research has concentrated upon
the effects of varying steric demands upon the nuclearity of indium
b-diketiminate derivatives, IV. Unambiguous singlet indium diyls
are isolated when Ar = 2,6-iPr2C6H3,6 while smaller ortho-methyl
(2,4,6-Me3C6H2 or 2,6-Me2C6H3) N-substituents permit solid-state
dimerisation and the formation of compounds, VI,8 which may be
viewed as heavier group 13 ‘alkene analogues’. Although the
maintenance of this interaction in solution is doubtful and the
subject of ongoing studies in our group, we have found that
further moderation of the steric demands of the N-aryl substituents
to 3,5-Me2C6H3 results in the isolation of a remarkable linear
catenated hexa-indium complex, VII, which displays behaviour
consistent with a ‘s-delocalized’ manifold of molecular orbitals.9
We have previously reported that the mononuclear Tl(I) complex,
V, may be synthesised and handled without apparent difficulty.7a
Our own work and recent density functional theory (DFT)
calculations have suggested that a satisfactory view of the of the
E–E bonding within dimerised ‘carbenoid’ group 13 species of
indium and thallium is as much a question of semantics (i.e. when
is a double/multiple bond a double/multiple bond?) as the
application of conventional bonding theory.8,10 As part of a more
wide ranging synthetic effort to address these questions, we now
report the use of a suite of less sterically demanding b-diketiminate
ligands with the aim of synthesising Tl(I) complexes similar to VI
Despite numerous attempts, work up of the reaction of Tl(I)
and [CH{(Me)CN-2,4,6-Me3C6H2}2K] (formed in situ from
K{N(SiMe3)2} and the b-iminoamine precursor) provided no
identifiable b-diketiminato thallium complexes in pure isolable
form. Reaction in THF was accompanied by deposition of
elemental thallium. This observation is reminiscent of our previous
work in indium chemistry where lower valent complexes have been
identified by careful work-up and crystallisation from hydrocarbon
solvents. Extraction of the solid dark residue with hexane and
filtration produced pale yellow solutions, which proved to be
highly unstable toward the production of a metallic thallium
mirror on the inner surfaces of the flasks even when stored in the
absence of light and at low temperature. Although the intended
compound most likely forms under these conditions, we were
neither able to crystallise nor identify definitively by spectroscopic
means the Tl(I) b-diketiminate. Prolonged storage and repeated
filtration of one such solution did, however, result in the isolation
of a small number of colourless single crystals. These were
identified as the unusual mixed valence [Tl(I)/Tl(III)] Tl8 cluster,
compound 1, by a single crystal X-ray diffraction analysis.{ The
structure of 1 is illustrated in Fig. 1 and selected bond lengths and
angles are provided in the figure caption. The structure has
crystallographically imposed twofold symmetry and may be
rationalised as comprising two Tl4 subunits connected via a single
oxide (O6) and two hydroxyl (O5, O59) bridging ligands. Each half
of the cluster includes a single trivalent and three univalent centres.
The formally trivalent and six-coordinate Tl1 (Tl19) atoms are
ligated by the bridging oxide and hydroxyl ligands in a cis fashion.
The remaining coordination sites about these atoms are provided
by the geminal oxygens of two bis-imino acetal ligands formed by
apprarent oxidation of the b-diketiminate. The formally univalent
Tl atoms are all two-coordinate and serve to bridge between the
aDepartment of Chemistry, Imperial College London, Exhibition Rd,
South Kensington, London, UK SW7 2AZ.
E-mail: mike.hill@imperial.ac.uk; Fax: +44 (0)20 7594 5804;
Tel: +44 (0)20 7594 5709
bSchool of Chemistry, University of Sussex, Falmer, Brighton, UK
BN1 9QJ
3720 | Chem. Commun., 2006, 3720–3722
This journal is ß The Royal Society of Chemistry 2006

View Article Online
Fig. 1 ORTEP of compound 1 (20% ellipsoids). Selected bond lengths
˚
(A) and angles (u); Tl(1)–O(2) 2.238(6), Tl(1)–O(4) 2.369(6), Tl(1)–O(1)
2.394(6), Tl(2)–O(59) 2.446(6), Tl(2)–O(4) 2.460(6), Tl(3)–O(1) 2.444(6),
Tl(3)–O(6) 2.484(3), Tl(4)–O(3) 2.447(7), Tl(4)–O(2) 2.470(6); O(6)–Tl(1)–
O(1) 81.73(17), O(5)–Tl(1)–O(1) 88.1(2), O(3)–Tl(1)–O(1) 127.1(2), O(2)–
Tl(1)–O(1) 59.2(2), O(4)–Tl(1)–O(1) 166.7(2), O(59)–Tl(2)–O(4) 81.4(2),
O(1)–Tl(3)–O(6) 74.65(17), O(3)–Tl(4)–O(2) 70.53(19), where (9) in the
atom labels indicates that these atoms are at (3/2 2 x,y,3/2 2 z).
Fig. 2 ORTEP of compound 2 (50% ellipsoids). Selected bond lengths
˚
(A) and angles (u); Tl–N(1) 2.401(2), N(1)–C(1) 1.327(4), N(1)–C(4)
1.417(4), N(2)–C(12) 1.336(5), N(2)–C(17) 1.414(5), N(3)–C(14) 1.300(5),
N(3)–C(22) 1.415(5); N(19)–Tl–N(1) 76.84(12), C(1)–N(1)–C(4) 120.8(3),
C(1)–N(1)–Tl 130.9(2), C(4)–N(1)–Tl 108.4(2), where (9) in the atom labels
indicates that these atoms are at (x,1/2 2 y,z).
oxygen atoms about the Tl(III) centres, either ‘intra’ (Tl4, Tl49) or
‘inter’ (Tl2, Tl29/Tl3, Tl39) -molecularly.
between the protonated ligand and the thallium centre, the
presence of the wholly organic solvate undoubtedly exerts a
profound structural influence upon the nuclearity of the isolated
thallium species (vide infra). We have noted previously that the
bond lengths and angles about the N–In–N fragment of
monomeric and dimeric In(I) b-diketiminates are little altered
by the presence of the In–In contact.8 The corresponding bond
lengths and angles within the structure of 2 [N(1)–Tl–N(19),
Although the process that produces the cluster molecule, 1, is ill-
defined, we were surprised that a similar reaction of Tl(I) and the
potassium b-diketiminate [CH{(Me)CN-2,6-Me2C6H3}2K] pro-
duced, after work-up, a reasonable (35%) yield of the pale yellow
complex 2.{ Warren has previously reported the use of the
thallium b-diketiminate [CH{(Me)2CN-2,6-Me2C6H3}2Tl], synthe-
sised by a different route, as a reagent in the synthesis of copper
complexes.11 The isolated crystals of complex 2 displayed notable
stability in the solid state and could be stored in ambient light for
several days with no apparent darkening or indication of
decomposition. Although solutions were somewhat photolabile,
˚
76.84(12)u; Tl–N(1), 2.401(2) A] are little affected by the
presence of the protonated ligand and lie within the ranges
previously established by the unambiguous monomers
[CH{(Me)CN-2,6-iPr2C6H2}2Tl],7a
[CH{(Ph)2CN(SiMe3)}2Tl]
1
and [CPh{(H)CN-2,6-iPr2C6H2}2Tl].7b
promptly collected H NMR spectra of 2 displayed a b-diketimi-
nate environment identical with the data reported for Warren’s
crystalline (but notably light-sensitive) complex. Our spectra,
however, also displayed a second set of resonances in a 1 : 1 ratio
with those ascribed to the Tl(I) b-diketiminate, which, from
comparison to reference spectra, were identified as resulting from
the presence of an equivalent of the (protonated) b-iminoamine
precursor. An explanation for these observations became apparent
from a further X-ray diffraction experiment performed on single
crystals of 2 isolated from hexane solution.{ The results of this
analysis revealed that the compound 2 has crystallographically
imposed twofold symmetry and are illustrated in Fig. 2. Selected
bond lengths and angles are given in the figure caption. The solid-
state structure of 2 is reminiscent of the outcome of the reaction of
TlBr and the potassium amidinate [{ArNC(tBu)NAr}K] (Ar =
2,6-iPr2C6H3) in that it contains a cocrystallised molecule of the
protonated ligand precursor.2 Both the N-mesityl, and the directly
analogous In(I) species, [CH{(Me)CN-2,6-Me2C6H3}2In], are
isolated as dimeric aggregates in the solid-state, which may be
viewed as being formed by the, albeit very weak, interaction of two
singlet carbenoid species (i.e. heavier group 13 analogues of an
alkene). Although there there are no notably shortened contacts
Our work with the similar but less sterically demanding meta-
xylyl b-diketiminate ligand, [CH{(Me)2CN-3,5-Me2C6H3}]2 has
not yet yielded a stable univalent In(I) derivative in either
monomeric or dimeric form. Rather, the striking, but formally
mixed [In(I)/In(II)] valence, species, VII, resulted from the reaction
of InI and [CH{(Me)CN-3,5-Me2C6H3}K] under similar reaction
and work-up conditions to those described above.9 With this result
in mind, we performed the analogous reaction of [CH{(Me)CN-
2,4,6-Me3C6H2}K] with an equimolar quantity of Tl(I). This
reaction proceeded in a similar manner with concurrent produc-
tion of a heavy precipitate of elemental thallium and resulted in a
light-sensitive pale yellow solution after extraction of the reaction
mixture with hexane. Concentration and storage of this solution at
230 uC for one week resulted in the formation of an isolable
quantity (ca. 10% estimated yield) of bright yellow, but highly
light-sensitive, needles, compound 3. Although a promptly
1
acquired H NMR spectrum of 3 again indicated the presence of
a single b-diketiminate environment, it was noted that the sample
darkened notably in the intervening time between sample
preparation and collection of these spectral data and the
isolated crystalline material was subject to unavoidable, but
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 3720–3722 | 3721

View Article Online
Any treatment that presupposes or inadvertently imposes either
view of the bonding within similar lower valent complexes of the
heavier group 13 elements must be treated, therefore, with caution
if any possible ‘experimenter effect’ is to be avoided. We are
continuing to attempt the synthesis and to study the reactivity of
further related compounds in our effort to address these
fundamental questions within heavier main group element
chemistry.
MSH would like to thank the Royal Society for a University
Research Fellowship.
Notes and references
{ Crystal data: 173 K, Nonius Kappa CCD diffractometer, l(Mo Ka) =
˚
0.71073 A. 1: C92H114N8O11Tl8?2(C6H14), M = 3315.22, monoclinic, P2/n
˚
(No. 13), a = 12.4127(3), b = 15.1224(3), c = 29.4433(2) A, b = 95.743(1)u,
V = 5499.1(2) A , Z = 2, m = 11.73 mm21, 29498 collected reflections, 9534
3
˚
independent reflections [R(int) = 0.054], R indices [I . 2s(I)] R1 = 0.045,
wR2 = 0.083, [all data] R1 = 0.076, wR2 = 0.092. 2: C42H51N4Tl, M =
816.24, orthorhombic, Pnma (No. 62), a = 22.3307(3), b = 16.6863(3), c =
3
10.2310(1) A, V = 3812.24(9) A , Z = 4, m = 4.27 mm21, 61777 collected
reflections, 3883 independent reflections [R(int) = 0.070], R indices [I .
2s(I)] R1 = 0.027, wR2 = 0.051, [all data] R1 = 0.042, wR2 = 0.056. 3:
˚
˚
Fig. 3 ORTEP of compound 3 (20% ellipsoids). Selected bond lengths
˚
(A) and angles (u); Tl(1)–N(2) 2.462(6), Tl(1)–N(1) 2.472(6), Tl(1)–Tl(2)
3.5794(4), Tl(2)–N(4) 2.504(6), Tl(2)–N(3) 2.520(6), Tl(2)–Tl(3) 3.7977(4),
¯
C63H75N6Tl3, M = 1529.40, triclinic, P1 (No. 2), a = 10.0548(2), b =
˚
12.3991(3), c = 24.0307(5) A, a = 92.243(2), b = 98.399(1) c = 97.355u, V =
2934.29(11) A , Z = 2, m = 8.26 mm21, 39677 collected reflections, 11528
3
…
Tl(3)–N(5) 2.482(6), Tl(3)–N(6) 2.515(6), Tl(3) Tl(19) 4.1318(4); N(2)–
˚
independent reflections [R(int) = 0.070], R indices [I . 2s(I)] R1 = 0.045,
wR2 = 0.097, [all data] R1 = 0.071, wR2 = 0.109. CCDC 609883–609885.
For crystallographic data in CIF or other electronic format see DOI:
10.1039/b607887g
Tl(1)–N(1) 75.94(19), N(2)–Tl(1)–Tl(2) 133.51(14), N(1)–Tl(1)–Tl(2)
86.13(14), N(4)–Tl(2)–N(3) 75.77(18), N(4)–Tl(2)–Tl(1) 59.77(13), N(3)–
Tl(2)–Tl(1) 68.80(13), N(4)–Tl(2)–Tl(3) 63.57(13), N(3)–Tl(2)–Tl(3)
56.99(12), Tl(1)–Tl(2)–Tl(3) 107.91(1), N(5)–Tl(3)–N(6) 76.81(19), N(5)–
Tl(3)–Tl(2) 128.78(13), N(6)–Tl(3)–Tl(2) 79.23(13).
{ Selected analytical data; 2: Elemental analysis: C42H51N4Tl: calcd C
61.76, H 6.86, N 6.25; found C 59.79, H 7.00, N 8.35% (partially
decomposed in transit); 1H NMR (400 MHz, C6D6, 25 uC): d = 1.65 (s, 6H,
CMe, LH), 1.80 (s, 6H, CMe), 2.24 (s, 12H, o-Me, LH), 2.26 (s, 12H,
o-Me), 4.93 (s, 1H, CH, LH), 4.95 (s, 1H, CH), 7.06–7.21 (m, 12H,
m.p-ArH and m.p-ArH, LH). 13C{1H} (100.7 MHz, C6D6) d = 18.3, 20.0
(o-Me), 24.4 (CMe), 94.1 (c-CH), 123.4 (Ar), 124.5 (Ar), 130.7 (Ar), 132.1
(Ar), 144.1 (i-Ar), 160.6, 161.6 (CN). 3: Elemental analysis: sealed sample
decomposed in transit. 1H NMR (400 MHz, C6D6, 25 uC): d = 2.01 (s, 6H,
CMe), 2.21 (s, 12H, m-Me), 4.96 (s, 1H, CH), 6.69 (s, 2H, p-ArH), 6.74 (s,
4H, o-Ar). 13C{1H} (100.7 MHz, C6D6) d = 20.6 (CMe), 21.2 (m-Me), 97.6
(c-CH), 120.5 (Ar), 124.9 (Ar), 138.2 (Ar), 146.2 (i-Ar), 158.9 (CN).
solution-accelerated, decomposition. This is borne out by the
results of a further X-ray diffraction analysis illustrated in Fig. 3,{
which revealed that 3 crystallises as a trimer with intermolecular
˚
˚
Tl(1)–Tl(2) [3.5794(4) A] and Tl(2)–Tl(3) [3.7977(4) A] interactions.
Recent DFT studies performed upon the N-phenyl analogue of 3
[HC{C(Me)NPh}2Tl] at the B3LYP level of theory have predicted
that dimerisation is endothermic by ca. 10 Kcal mol21 while our
own similar calculations upon the crystallographically identified
dimeric indium species indicated that negligible energy may be
expected to be involved in dissociation to the monomeric singlet
‘carbenoids’.8,10 These results and the metrical data provided by
the solid-state structure shown in Fig. 3 lead us to conclude that
the Tl–Tl interactions within compound 3 are better interpreted as
examples of dispersive attractive forces between the closed shell
(5d106s2–5d106s2) Tl(I) centres. While compounds of monovalent
gold have provided the most spectacular objects of study in this
context,12 the presence of weak interactions has been accepted as a
key factor to explain the supramolecular structures of Tl(I) amide
derivatives.13 Indeed the Tl–Tl distances within 3 are entirely
commensurate with many of the intermetallic contacts observed
1 Review: P. P. Power, Chem. Rev., 1999, 99, 3463.
2 C. Jones, P. C. Junk, J. A. Platts, D. Rathmann and A. Stasch, Dalton
Trans., 2005, 2497.
3 C. Jones, P. C. Junk, J. A. Platts, D. Rathmann and A. Stasch, J. Am.
Chem. Soc., 2006, 128, 2206.
4 C. Cui, H. W. Roesky, H.-G. Schmidt, M. Noltemeyer, H. Hao and
F. Cimpoesu, Angew. Chem., Int. Ed., 2000, 39, 4274.
5 N. J. Hardman, B. E. Eichler and P. P. Power, Chem. Commun., 2000,
1991.
6 M. S. Hill and P. B. Hitchcock, Chem. Commun., 2004, 1818.
7 (a) M. S. Hill, P. B. Hitchcock and R. Pongtavornpinyo, Dalton Trans.,
2005, 1433; (b) Y. Cheng, P. B. Hitchcock, M. F. Lappert and M. Zhou,
Chem. Commun., 2005, 752.
8 M. S. Hill, P. B. Hitchcock and R. Pongtavornpinyo, Angew. Chem.,
Int. Ed., 2005, 44, 1433.
9 M. S. Hill, P. B. Hitchcock and R. Pongtavornpinyo, Science, 2006, 311,
1904.
10 C.-H. Chen, M.-L. Tsai and M.-D. Su, Organometallics, 2006, 25, 2766.
11 X. L. Dai and T. H. Warren, Chem. Commun., 2001, 1998.
12 H. Schmidbaur, S. Cronje, B. Djordjevic and O. Schuster, Chem. Phys.,
2005, 311, 151.
13 L. H. Gade, Dalton Trans., 2993, 267 and references therein.
14 K. W. Hellmann, L. H. Gade, R. Fleischer and D. Stalke, Chem.
Commun., 1996, 2515.
15 J. C. Peters, S. B. Harkins, S. D. Brown and M. W. Day, Inorg. Chem.,
2001, 40, 5083.
in compounds such as the tripodal thallium amide
t
[MeSi{SiMe2N(Tl) Bu}3] [3.653(2), 3.673(2) A] and the polymeric
˚
thallium bis(8-quinolinyl)amide [3.5336(5) A].
˚
14,15
On this basis any argument that presents the possibility of a
thallium(I) ‘alkene analogue’ (i.e. providing some element of Tl–Tl
multiple bonding comparable to that readily identified in the
chemistry of the lighter p-block elements) is most likely specious
and raises the possibility that the In–In bonding within IV may
also be a reflection of similar (i.e. 4d105s2–4d105s2) phenomena.
3722 | Chem. Commun., 2006, 3720–3722
This journal is ß The Royal Society of Chemistry 2006