Titanium and zirconium complexes of a phosphorus-containing p-tert-butylcalix[5]arene ligand: Importance of metal and conformation on ligand/metal binding

Article abstract of DOI:10.1021/ic0601283

Binding of a calix[5]arene containing a single phosphorus ligand and three hydroxyl groups, calix[5]PNMe₂(OH)₃, 1, toward titanium and zirconium is investigated to yield insight into the factors that determine the strength of the phosphorus/metal interaction within the constraint of the calix[5]arene. Treatment of 1 with tetrakis(dimethylamino)-titanium yields three complexes, 4a, 4b, and 4c, each of which shows the loss of 3 mol of dimethylamine in the reaction with the titanium bound to three oxygens. Treatment of 1 with tetrakis(diethylamino)zirconium proceeds similarly, although only two products, 5a and 5b, were isolated. X-ray structures of the products were obtained. Complexes 4a and 5a show similar geometries, with the calix[5]arene in an approximate cone conformation and the phosphorus lone pair directed toward the metal. The P...M distances are, however, markedly different: 3.69 A in 4a and 3.18 A in 5a, the former indicative of no interaction and the latter a weak one. Complexes 4b and 5b each are dimers, featuring a planar four-membered M-O-M-O ring; however, the titanium is five-coordinate in 4b with no phosphorus/metal bond, while the zirconium in 5b is six-coordinate with a P-Zr distance of 2.95 A. Complex 4c is monomeric with the calix[5]arene in an approximate 1,2-alternate conformation with a P...Ti distance of 2.90 A. The two most significant aspects controlling the phosphorus/metal contact are the metal, with the larger zirconium showing stronger interaction, and the calix[5]arene conformation, with the cone conformation showing the weaker interaction.

Full text of DOI:10.1021/ic0601283

Inorg. Chem. 2006, 45, 6490−6496
Titanium and Zirconium Complexes of a Phosphorus-Containing
p-tert-Butylcalix[5]arene Ligand: Importance of Metal and Conformation
on Ligand/Metal Binding
Maomian Fan, Hongming Zhang, and Michael Lattman*
Department of Chemistry, Southern Methodist UniVersity, Dallas, Texas 75275-0314
Received January 21, 2006
Binding of a calix[5]arene containing a single phosphorus ligand and three hydroxyl groups, calix[5]PNMe2(OH)3,
1, toward titanium and zirconium is investigated to yield insight into the factors that determine the strength of the
phosphorus/metal interaction within the constraint of the calix[5]arene. Treatment of 1 with tetrakis(dimethylamino)-
titanium yields three complexes, 4a, 4b, and 4c, each of which shows the loss of 3 mol of dimethylamine in the
reaction with the titanium bound to three oxygens. Treatment of 1 with tetrakis(diethylamino)zirconium proceeds
similarly, although only two products, 5a and 5b, were isolated. X-ray structures of the products were obtained.
Complexes 4a and 5a show similar geometries, with the calix[5]arene in an approximate cone conformation and
the phosphorus lone pair directed toward the metal. The P‚‚‚M distances are, however, markedly different: 3.69
Å in 4a and 3.18 Å in 5a, the former indicative of no interaction and the latter a weak one. Complexes 4b and 5b
each are dimers, featuring a planar four-membered M
−O
−M
−O ring; however, the titanium is five-coordinate in 4b
with no phosphorus/metal bond, while the zirconium in 5b is six-coordinate with a P−Zr distance of 2.95 Å. Complex
4c is monomeric with the calix[5]arene in an approximate 1,2-alternate conformation with a P‚‚‚Ti distance of 2.90
Å. The two most significant aspects controlling the phosphorus/metal contact are the metal, with the larger zirconium
showing stronger interaction, and the calix[5]arene conformation, with the cone conformation showing the weaker
interaction.
Because of their ease of synthesis, most calixarene research
has focused on the even-numbered members of the family,
particularly the calix[4]arenes, and to a lesser extent, the
calix[6]- and calix[8]arenes.1 Our work2 has taken advantage
of the constraint of the small cavity of the calix[4]arene to
stabilize high-coordinate main-group element geometries and
study their transformations within the cavity. However, in
most cases, the cavity is too small to study the interaction
of two atoms within the framework. To that end, we reported3
the control of a phosphorus/metal interaction within the
cavity of a calix[5]arene4 by changing the electronic require-
ments at the metal by ligand substitution. Specifically,
replacing the excellent π-back-bonding amido ligand, NHBut,
in 2 with the much weaker binding triflate, results in a
significantly stronger phosphorus/tungsten interaction (Scheme
1). In addition, we recently reported insertion of silyl5 and
phosphorus6 groups into p-tert-butylcalix[5]arene. We believe
that this calix[5]arene cavity possesses the ideal balance
between constraint and flexibility to study and control ligand/
metal interactions. We herein report the reaction of 1 with
* To whom correspondence should be addressed. E-mail:
mlattman@smu.edu. Tel: 214-768-2467. Fax: 214-768-4089.
(1) (a) Calixarenes 2001; Asfari, Z.; Bo¨hmer, V.; Harrowfield, J.; Vicens,
J., Eds.; Kluwer: Dordrecht, 2001. (b) Calixarenes in Action;
Mandolini, L.; Ungaro, R.; Eds.; Imperial College: London, 2000.
(c) Gutsche, C. D. Calixarenes ReVisited; Royal Society of Chemis-
try: Letchworth, 1998. (d) Gutsche, C. D. Calixarenes; Royal Society
of Chemistry: Cambridge, 1989.
(2) (a) Fan, M.; Shevchenko, I. V.; Voorhies, R. H.; Eckert, S. F.; Zhang,
H.; Lattman, M. Inorg. Chem. 2000, 39, 4704. (b) Khasnis, D. V.;
Burton, J. M.; McNeil, J. D.; Santini, C. J.; Zhang, H.; Lattman, M.
Inorg. Chem. 1994, 33, 2657. (c) Khasnis, D. V.; Lattman, M.;
Gutsche, D. J. Am. Chem. Soc. 1990, 112, 9422. (d) Khasnis, D. V.;
Burton, J. M.; Lattman, M.; Zhang, H. J. Chem. Soc., Chem. Commun.
1991, 562. (e) Shang, S.; Khasnis, D. V.; Burton, J. M.; Santini, C.
J.; Fan, M.; Small, A. C.; Lattman, M. Organometallics 1994, 13,
5157. (f) Shang, S.; Khasnis, D. V.; Zhang, H.; Small, A. C.; Fan,
M.; Lattman, M. Inorg. Chem. 1995, 34, 3610. (g) Shevchenko, I. V.;
Zhang, H.; Lattman, M. Inorg. Chem. 1995, 34, 5405.
(3) Fan, M., Zhang, H.; Lattman, M. J. Chem. Soc., Chem. Commun. 1998,
99.
(4) For a recent review of calix[5]arene work, see Notti, A.; Parisi, M.
F.; Pappalardo, S. in ref 1a, Chapter 3.
(5) Sood, P.; Zhang, H.; Lattman, M. Organometallics 2002, 21, 4442.
(6) Sood, P.; Koutha, M.; Fan, M., Klichko, Y.; Zhang, H.; Lattman, M.
Inorg. Chem. 2004, 43, 2975.
6490 Inorganic Chemistry, Vol. 45, No. 16, 2006
10.1021/ic0601283 CCC: $33.50
© 2006 American Chemical Society
Published on Web 07/07/2006

Complexes of a p-tert-Butylcalix[5]arene Ligand
Scheme 1. Preparation and P‚‚‚W Interactions of Tungsen Complexes
of 1
mixture was stirred for 3 days after which the volatiles were pumped
off leaving a red residue. The following procedure led to isolation
of the products. (All crystallizations were performed at room
temperature.) The residue was dissolved in hot toluene and allowed
to sit at room temperature. Analytically pure 4a crystallized from
this solution. Concentration of the supernatant led to a small amount
of precipitate. The precipitate was redissolved and further concen-
trated by boiling the solution. Crystals of 4b came out at room
temperature. Recrystallization led to analytically pure 4b. The two
supernatants were combined and concentrated again. This led to
precipitation of crude 4c, which was recrystallized twice from
toluene. Although a satisfactory elemental analysis was obtained
1
for 4b, the H NMR spectrum showed about a 20% unidentified
impurity remaining in the final product. All three products are red
and air-sensitive, and all yields are based on the final isolated
compounds.
4a: 0.804 g (29%); mp: 128-130 °C (dec). Anal. Calcd for
C59H77N2O5PTi‚2C7H8: C, 75.76; H, 8.10. Found: C, 75.19; H,
8.29. 1H NMR (CDCl3, δ): 1.08 (s, 18H, t-Bu), 1.15 (s, 9H, t-Bu),
1.38 (s, 18H, t-Bu), 2.94 (d, 6H, PNMe2, 3JPH ) 10.4 Hz), 3.32 (d,
2
2
2H, CH2, JHH ) 15.0 Hz), 3.34 (d, 1H, CH2, JHH ) 12.3 Hz),
2
3.38 (d, 2H, CH2, JHH ) 13.6 Hz), 3.71 (s, 6H, TiNMe2), 4.61
(dd, 1H, CH2, JHH ) 12.3 Hz, JPH ) 2.7 Hz), 4.71 (overlapping
d, 4H, CH2, JHH ) 13 Hz, approx), 6.98 (d, 2H, CH, JHH ) 2.4
Hz), 7.01 (overlapping s, 4H, CH), 7.16 (s, 2H, CH), 7.24 (CH,
overlapping with CDCl3). 31P NMR (CDCl3, δ): 127.7.
2
5
2
4
4b: 0.344 g (15%); mp: >238 °C (dec). Anal. Calcd for
(C59H77N2O5PTi)2: C, 72.82; H, 7.98. Found: C, 72.95; H, 8.28.
1H NMR (CDCl3, δ): 0.78 (d, 12H, PNMe2, 3JPH ) 10.0 Hz), 1.17
(s, 18H, t-Bu), 1.18 (s, 36H, t-Bu), 1.25 (s, 36H, t-Bu), 3.03 (s,
12H, TiNMe2), 3.42, 3.45, 3.50 (overlapping, 10H, CH2), 4.40 (d,
4H, 2JHH ) 16.6 Hz), 5.07 (d, 4H, 2JHH ) 12.4 Hz), 5.36 (dd, 2H,
CH2, 2JHH ) 15.7 Hz, 5JPH ) 7.5 Hz), 6.86 (d, 4H, CH, 4JHH ) 2.1
4
4
Hz), 6.89 (d, 4H, CH, J
) 2.2 Hz), 6.98 (d, 4H, CH, JHH
)
HH
2.1 Hz), 7.04 (s, 4H, CH), 7.11 (d, 4H, CH, JHH ) 2.3 Hz). 31P
4
titanium and zirconium and demonstrate that calixarene
conformation, as well as the specific metal, plays a crucial
role in the phosphorus/metal interaction.
NMR (CDCl3, δ): 136.4.
4c: 0.575 g (24%); mp: >295 °C (dec). Anal. Calcd for
C59H77N2O5PTi‚0.5C7H8: C, 73.66; H, 8.01. Found: C, 73.69; H,
7.68. 1H NMR (CDCl3, δ): 1.16 (s, 9H, t-Bu), 1.23 (s, 18H, t-Bu),
1.40 (s, 18H, t-Bu), 2.20 (s, 6H, TiNMe2, 4JPH ) 2.0 Hz), 2.44 (d,
6H, PNMe2, 3JPH ) 10.7 Hz), 3.29 (d, 2H, CH2, 2JHH ) 14.5 Hz),
3.31 (d, 2H, CH2, 2JHH ) 14.8 Hz), 3.46 (d, 1H, CH2, 2JHH ) 13.0
Experimental Section
All reactions and manipulations were carried out under an
atmosphere of nitrogen in a Vacuum Atmospheres drybox or by
using standard Schlenk techniques, unless otherwise indicated.
Solvents were dried using standard procedures and distilled under
a nitrogen atmosphere and either used immediately or stored in
the drybox prior to use. Glassware was oven-dried at 140 °C
overnight prior to use. The reagents tetrakis(dimethylamino)titanium
and tetrakis(diethylamino)zirconium were obtained commercially
and used without further purification. Ligand 1 was synthesized
by the literature procedure.3 All NMR spectra were recorded on a
Bruker AVANCE DRX-400 multinuclear NMR spectrometer
resonating at 400.137 (1H) and 161.979 MHz (31P). 1H NMR
resonances were measured relative to residual proton solvent peaks
and referenced to Me4Si. 31P NMR resonances were measured
relative to external 85% H3PO4. Melting points were obtained in
nitrogen-filled tubes on a Mel-Temp capillary apparatus and are
uncorrected. Elemental analyses were obtained from Complete
Analysis Laboratories, Inc. (E + R Microanalytical Division),
Parsippany, NJ. Elemental analyses calculated with solvent included
were determined by NMR integrations.
2
5
Hz), 4.34 (dd, 1H, CH2, JHH ) 12.9 Hz, JPH ) 4.1 Hz), 4.49 (d,
2
2
2H, CH2, JHH ) 14.4 Hz), 4.57 (d, 2H, CH2, JHH ) 14.7 Hz),
7.00 (s, 2H, CH), 7.14-7.19 (overlapping, 8H, CH). 31P NMR
(CDCl3, δ): 128.6.
Synthesis of Zirconium Complexes. A stirred solution of 1 (2.20
g, 2.49 mmol) in toluene (40 mL) was treated dropwise with
tetrakis(diethylamino)zirconium (0.954 g, 2.51 mmol). The mixture
was stirred at room temperature for 3 days after which the volatiles
were pumped off leaving a light-yellow residue. The residue was
heated in toluene and filtered hot. The toluene filtrate was
concentrated and left for several days to give 5a as colorless crystals.
Although spectra indicated a pure product, we were unable to obtain
a reliable elemental analysis. The insoluble material from the
filtration was recrystallized from chloroform/toluene to give analyti-
cally pure 5b as colorless crystals. (Crystals of 5b for X-ray
crystallography were grown from chloroform.)
5a: 0.804 g (25%); mp: 150-152 °C (dec). Anal. Calcd for
C61H81N2O5PZr‚3C7H8: C, 74.56; H, 8.01. Found: C, 71.53; H,
8.03. 1H NMR (CDCl3, δ): 1.10 (s, 9H, t-Bu), 1.16 (s, 18H, t-Bu),
1.31 (s, 18H, t-Bu), 1.46 (t, 6H, ZrNCH2CH3, 3JHH ) 7.1 Hz), 2.90
Synthesis of Titanium Complexes. A stirred solution of 1 (2.09
g, 2.36 mmol) in toluene (40 mL) was treated dropwise with
tetrakis(dimethylamino)titanium (0.535 g, 2.39 mmol). The reaction
3
2
(d, 6H, PNMe2, JPH ) 10.4 Hz), 3.21 (d, 2H, CH2, JHH ) 13.9
Inorganic Chemistry, Vol. 45, No. 16, 2006 6491

Fan et al.
Table 1. Crystal Dataa
4a
4b
4c
empirical formula
fw
cryst syst
space group
a, Å
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
C59H77N2O5PTi‚2C7H8
1157.36
hexagonal
P63/m
22.874(1)
22.874(1)
23.625(3)
90
90
120
10705.0(15)
6
1.077
C118H154N4O10P2Ti2‚6C7H8
2499.00
triclinic
C59H77N2O5PTi‚C7H8
1065.23
triclinic
P1h
P1h
16.781(1)
17.030(2)
17.376(1)
116.638(7)
118.220(4)
92.656(7)
3694.4(5)
1
12.7577(11)
13.080(2)
20.4886(17)
72.614(10)
84.752(5)
67.018(9)
3002.4(6)
2
Z
F
calcd, g cm-3
1.123
0.187
0.0004(3)
0.103
0.257
1.178
0.219
0.0014(4)
0.106
0.303
µ, mm-1
0.189
0.0030(6)
0.104
extinctn coeffb
R1 [I > 2σ(I)]c
wR2 [all data]c
0.298
5a
5b
C122H162N4O10P2Zr2‚8CHCl3
3043.88
triclinic
P1h
14.605(3)
16.059(2)
17.379(3)
89.49(1)
72.51(2)
72.72(1)
3697.7(11)
1
empirical formula
fw
cryst syst
space group
a, Å
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
C61H81N2O5PZr‚2C7H8
1228.74
hexagonal
P63/m
22.987(1)
22.987(1)
24.080(4)
90
90
120
11019(2)
6
Z
F
calcd, g cm-3
1.111
0.218
n/a
0.102
0.313
1.367
0.650
0.0003(6)
0.103
0.306
µ, mm-1
extinctn coeffb
R1 [I > 2σ(I)]c
wR2 [all data]c
2
2
a Graphite monochromatized Mo KR radiation, λ ) 0.71073 Å. b See ref 7. c R1 ) ∑||Fo| - |Fc|/|∑|Fo|, wR2 ) {∑[w(Fo2 - Fc )2]/∑[w(Fo )2]}1/2, where
2
2
2
w ) 1/[σ2(Fo ) + (aP)2 + bP], P ) [2Fc + Fo ]/3.
2
Hz), 3.28 (d, 2H, CH2, JHH ) 13.2 Hz), 3.57 (q, 4H, ZrNCH2
against F2.7 Solvate toluene molecules were found in each of the
crystal lattices. Most solvent molecules, as well as some methyl
groups of tert-butyls, are positionally disordered. These disordered
atoms were refined with occupancy and distance restraints. All non-
hydrogen atoms were refined anisotropically, while H atoms were
constrained with a riding model. Further details regarding the crystal
data and refinement, as well as full tables of bond lengths and angles
for each structure reported in this paper, are presented in CIF format
in the Supporting Information.
CH3, 3JHH ) 7.1 Hz), 3.52 (d, 1H, CH2, 2JHH ) 14.9 Hz), 4.39 (d,
2
2
1H, CH2, JHH ) 14.9 Hz), 4.50 (d, 2H, CH2, JHH ) 13.8 Hz),
4.72 (d, 2H, CH2, 2JHH ) 13.1 Hz), 6.99 (s, 2H, CH), 7.00 (d, 2H,
4
4
CH, JHH ) 2.3 Hz), 7.13 (d, 2H, CH, JHH ) 2.6 Hz), 7.16 (4H,
CH, overlapping). 31P NMR (CDCl3, δ): 118.0.
5b: 0.494 g (17%); mp: >370 °C (dec). Anal. Calcd for
(C61H81N2O5PZr)2‚2C7H8: C, 71.85; H, 7.89. Found: C, 72.21; H,
7.62. 1H NMR (CDCl3, δ): 0.08 (t, 12H, ZrNCH2CH3, 3JHH ) 7.0
Hz), 0.89 (d, 12H, PNMe2, JPH ) 8.6 Hz), 1.15 (s, 18H, t-Bu),
1.19 (s, 36H, t-Bu), 1.27(s, 36H, t-Bu), 3.23 (d, 2H, CH2, JHH
14.9 Hz), 3.34 (d, 4H, CH2, JHH ) 14.9 Hz), 3.38 (d, 4H, CH2,
2JHH ) 12.6 Hz), 3.65 (br, 8H, ZrNCH2CH3), 4.00 (d, 2H, CH2,
2JHH ) 14.8 Hz), 4.94 (d, 4H, CH2, 2JHH ) 14.8 Hz), 5.25 (d, 4H,
3
Results and Discussion
2
)
2
The three hydroxyl oxygens in 1 are ideally arranged to
bind a single transition metal. However, choice of a suitable
reagent is critical. For the Group 4 metals, we were unable
to isolate any compounds with titanium(IV) and zirco-
nium(IV) chlorides. Dialkylamino reagents were used instead.
2
4
CH2, JHH ) 12.5 Hz), 6.94 (d, 4H, CH, JHH ) 2.3 Hz), 6.99 (d,
4H, CH, 4JHH ) 2.5 Hz), 7.17 (s, 4H, CH), 7.18 (d, 4H, CH, JHH
4
) 2.6 Hz), 7.19 (d, 4H, CH, 4JHH ) 2.2 Hz). 31P NMR (CDCl3, δ):
Treatment of 1 with tetrakis(dimethylamino)titanium in a
1:1 molar ratio (Scheme 2) at ambient temperature led to a
mixture of several products, 4a, 4b, and 4c, which were
isolated by fractional crystallization (see Experimental Sec-
tion). Several key features in the H NMR spectra of the
products indicate a similarity among all three. First of all,
121.4.
Single-Crystal X-ray Diffraction Studies. Three titanium crystal
samples are orange colored, while the zirconium ones are colorless.
All crystals were grown from toluene except for 5b. Suitable X-ray
quality crystals of 5b could only be obtained from chloroform. All
data were collected on a Siemens P4 diffractometer. The crystals
were coated with mineral oil under a low-temperature nitrogen
stream. Crystallographic data are summarized in Table 1. All
structures were solved by direct methods and subsequent difference
Fourier syntheses and refined by full-matrix least-squares methods
1
the overall Cs symmetry of the free ligand is retained in the
(7) Sheldrick, G. M. SHELXS97 and SHELXL97; University of Go¨ttin-
gen: Go¨ttingen, Germany, 1997.
6492 Inorganic Chemistry, Vol. 45, No. 16, 2006

Complexes of a p-tert-Butylcalix[5]arene Ligand
Scheme 2. Preparation of Titanium and Zirconium Complexes of 1
Figure 1. Molecular structure of 4a. Thermal ellipsoids are shown at 50%
probability. For clarity, carbon atoms of the calix[5]arene are shown at an
arbitrary size and hydrogen atoms are omitted.
compounds: three resonances are observed for the tert-butyl
groups (in a 2:2:1 ratio). In addition, six doublets (due to
2JHH geminal coupling) are found for the methylene bridging
groups (in a 2:2:2:2:1:1 ratio). Finally, two resonances
assigned to dimethylamino groups are observed; one is the
phosphorus-bound group, easily assigned by its characteristic
doublet (3JPH ca. 10 Hz), and the other, most likely
dimethylamino bound to titanium. These integrate to a 1:1
ratio, signifying the loss of 3 mol of dimethylamine in the
reaction, as expected.
Clues to the differences among the products are found
upon closer examination of the spectra, specifically the
chemical shifts of the dimethylamino groups. In 4a, these
appear at δ 2.94 (PNMe2) and 3.71 (TiNMe2). Both
resonances appear upfield in the two other products: 4b, δ
0.78 (PNMe2) and 3.03 (TiNMe2); 4c, δ 2.44 (PNMe2) and
2.20 (TiNMe2). Such chemical shift differences are common
in calixarene chemistry and usually indicate increased
shielding due to protons positioned “above” an aromatic ring.
One final observation is noted for 4c: a small phosphorus
coupling (2.0 Hz) is clearly observed in the Ti-bound
dimethylamino resonance and absent in 4a and 4b, perhaps
indicative of a stronger phosphorus/metal interaction com-
pared to the other titanium complexes.
Table 2. Selected Interatomic Distances (Å) and Angles (deg) for 4a
and 5a
4a
5a
4a
5a
M-O1
M-O2
M-N1
1.828(8) 2.005(10) P-O3
1.818(5) 1.984(7) P-N2
1.835(10) 1.976(14) P‚‚‚M
1.638(5) 1.638(7)
1.673(10) 1.657(12)
3.690
3.176(4)
O1-M-O2
O1-M-N1
O2-M-N1
102.1(2) 94.3(2)
98.0(5) 95.6(5)
109.4(2) 112.8(2) P-M-O1
O3-P-O3A 99.9(4)
103.0(5)
98.1(4)
176.3(3)
O3-P-N2
96.6(3)
179.7
O2-M-O2A 130.2(3) 132.4(4)
Table 3. Selected Interatomic Distances (Å) and Angles (deg) for 4b
and 5b
4b
5b
4b
5b
M-O1A
M-O1
M-O2
M-O5
M-N1
1.976(5) 2.181(8)
2.144(5) 2.260(8)
1.816(6) 2.010(9)
1.823(5) 2.005(11) P-N2
1.886(6) 2.041(11)
M‚‚‚P
P-O3
P-O4
2.954(4)
1.654(6)
1.638(6)
1.648(7)
1.612(10)
1.620(10)
1.668(13)
O1A-M-O2 114.9(2) 104.6(3)
O1A-M-O5 115.4(2) 100.8(4)
O2-M-P
77.2(3)
99.1(4)
75.7(3)
96.9(3)
105.1(5)
106.8(6)
96.0(6)
94.1(4)
114.8(3)
136.7(4)
O5-M-N1 98.8(3)
O5-M-P
O1A-M-O1 68.9(2)
O1A-M-N1 92.8(2)
O1A-M-P
66.5(4)
89.3(4)
173.2(3)
88.7(3)
86.2(3)
N1-M-P
O3-P-O4
O3-P-N2
O4-P-N2
O3-P-M
O4-P-M
N2-P-M
93.9(3)
99.3(3)
101.2(3)
O1-M-O2
O1-M-O5
O1-M-N1
O1-M-P
87.8(2)
90.7(2)
161.7(2) 155.9(4)
107.2(2)
O2-M-O5
O2-M-N1
125.2(2) 149.5(4)
99.2(3)
97.7(4)
M-O1-MA 111.1(2) 113.5(4)
The reaction of 1 with tetrakis(diethylamino)zirconium
proceeds similarly, except that only two complexes, 5a and
5b, could be isolated. Integration of the dialkylamino
resonances supports the loss of 3 mol of diethylamine in the
products. The PNMe2 resonance in 5a (δ 2.90) appears at
almost the same chemical shift as in 4a, while the methyl
resonance of the diethylamino group appears at δ 1.46. In
5b, these move upfield: δ 0.89 (PNMe2) and 0.08 (CH3 of
ZrNEt2).
X-ray structures of the products were obtained and
illustrate some similarities, as well as several marked
differences, among the titanium and zirconium products.
Selected distances and angles are tabulated in Tables 2-4.
While the structures obtained are not of the highest quality,
key features are clear, and it is these that will be discussed.
The most similar titanium/zirconium structures are 4a and
Table 4. Selected Interatomic Distances (Å) and Angles (deg) for 4c
Ti-O1
Ti-O2
Ti-O5
Ti-N1
1.886(8)
1.871(8)
1.853(7)
1.848(10)
Ti-P
2.897(4)
1.642(8)
1.635(8)
1.664(10)
P-O3
P-O4
P-N2
O1-Ti-O2
O1-Ti-O5
O1-Ti-N1
O1-Ti-P
94.8(4)
100.7(3)
96.0(4)
164.8(3)
141.4(4)
109.1(4)
77.4(3)
O5-Ti-P
N1-Ti-P
O3-P-O4
O3-P-N2
O3-P-Ti
O4-P-N2
O4-P-Ti
N2-P-Ti
78.7(3)
98.9(3)
98.3(4)
101.5(5)
116.1(3)
98.5(5)
120.0(3)
118.6(4)
O2-Ti-O5
O2-Ti-N1
O2-Ti-P
O5-Ti-N1
104.2(4)
5a, Figures 1 and 2, respectively (and Chart 1). Both are
monomeric with the calix[5]arene backbone in an ap-
proximate cone conformation,8 and the phosphorus lone pair
is directed toward the metal atom. However, the P‚‚‚Ti
Inorganic Chemistry, Vol. 45, No. 16, 2006 6493

Fan et al.
Figure 2. Molecular structure of 5a. Thermal ellipsoids are shown at 50%
probability. For clarity, carbon atoms of the calix[5]arene are shown at an
arbitrary size and hydrogen atoms are omitted.
Figure 3. Molecular structure of 4b. Thermal ellipsoids are shown at 40%
probability. For clarity, carbon atoms of the calix[5]arene are shown at an
arbitrary size and hydrogen atoms are omitted.
Chart 1
Figure 4. Molecular structure of 5b. Thermal ellipsoids are shown at 40%
probability. For clarity, carbon atoms of the calix[5]arene are shown at an
arbitrary size and hydrogen atoms are omitted.
suggesting a weak interaction. These values can be compared
to typical P-Ti and P-Zr bond distances of 2.5-2.6 and
2.7-2.8 Å, respectively.9 Complexes 4b and 5b are dimeric,
featuring a planar four-membered M-O-M-O ring (Figures
3 and 4, respectively), a very common structure in titanium
and zirconium chemistry.10
distance is 3.69 Å in 4a, well outside the range of any
conceivable interaction. In 5a, the phosphorus atom is
significantly closer to the metal (P‚‚‚Zr distance of 3.18 Å),
(9) See, for example, (a) Harvey, B. G.; Basta, R.; Arif, A. M.; Ernst, R.
D. Dalton Trans. 2004, 1221. (b) Ernst, R. D.; Freeman, J. W.; Stahl,
L.; Wilson, D. R.; Arif, A. M.; Nuber, B.; Ziegler, M. L. J. Am. Chem.
Soc. 1995, 117, 5075. (c) Wengrovius, J. H.; Schrock, R. R.; Day, C.
S. Inorg. Chem. 1981, 20, 1844.
(10) For recent examples, see Fric, H.; Schubert, U. New J. Chem. 2005,
29, 232, and references therein.
(8) For more accurate descriptions of calix[5]arene conformations, see
(a) Thondorf, I.; Brenn, J. J. Chem. Soc., Perkin Trans. 2 1997, 2293.
(b) Harada, T.; Shinkai, S. J. Chem. Soc., Perkin Trans. 2 1995, 2231.
(c) Stewart, D. R.; Krawiec, M.; Kashyap, R. P.; Watson, W. H.;
Gutsche, C. D. J. Am. Chem. Soc. 1995, 117, 586.
6494 Inorganic Chemistry, Vol. 45, No. 16, 2006

Complexes of a p-tert-Butylcalix[5]arene Ligand
Table 5. Comparison of “Equatorial” Angles (deg) around the Metal in
4a, 4c, and 5aa
4a
4c
5a
Oa-M-Ob
Oa-M-N
Ob-M-N
sum
130.1
109.5
109.5
349.1
141.4
109.1
104.2
354.7
132.4
112.7
112.7
357.8
a Oa and Ob are O2 and O2A for 4a and 5a; they are O2 and O5 for 4c.
Oc is O1 in all.
1
While the H NMR chemical shifts give insight into the
detailed structures, the 31P NMR shifts do not. For example,
4a and 4c have almost identical chemical shifts (δ 127.7
and 128.6, respectively), even though the P‚‚‚Ti distance is
substantially shorter in the latter. This similarity could be
due to the fact that the P‚‚‚Ti distance in 4c is still about 0.3
Å longer than a normal P-Ti bond. Both are shifted slightly
upfield compared to the free ligand 1 (δ 130.6).3 We observed
a similar upfield shift in complex 3 (δ 125),3 a compound
with a long P-W bond (see below).
In an effort to ascertain the relative stabilities of the
titanium and zirconium complexes, we attempted to inter-
convert them thermally. Isolated samples of each complex
were heated overnight in refluxing toluene, but no evidence
of interconversion was found. A small amount of decom-
position was observed in most cases, but otherwise, the
spectra remained unchanged. Since there are many more low-
energy conformations of calix[5]arenes compared to calix-
[4]arenes,8 it is likely that enthalpy differences among the
products are not significant.
An examination of selected angles about the metal in the
monomeric structures provides insight into the constraint
imposed by the calixarene and the ability of the phosphorus
to interact with the metal. Table 5 lists the angles at the metal
involving nitrogen, O2, and O2A (or O5 for 4c). The
approach of the phosphorus should alter the angles at the
metal from tetrahedral to trigonal bipyramidal. While this is
the case to some extent (the sum of the “equatorial” angles
in 4a increases by 6° in 4c), it is not that dramatic. Moreover,
the sum of these angles in all three complexes is much closer
to planar than tetrahedral. A closer examination of these
angles shows the two O-M-N angles are close to tetrahedral
in all three complexes. It is the Oa-M-Ob angle that is
different, 130-141°. The constraint of three adjacent phe-
nolic calixarene groups bound to a single metal appears to
impose this large angle. It is important to point out that it is
the three adjacent groups that impart this constraint, rather
than the overall calix[5]arene backbone, since the first calix-
[4]arene titanium complex ever synthesized (6)12 shows a
relatively large value (127°) for the corresponding angle. In
Figure 5. Molecular structure of 4c. Thermal ellipsoids are shown at 30%
probability. For clarity, hydrogen atoms are omitted.
Although the metal cores are similar in both complexes,
the metal coordination numbers are not: titanium is five-
coordinate with no phosphorus interaction, while zirconium
is six-coordinate with a P‚‚‚Zr distance of 2.95 Å. The
calixarene backbone in 4b is in an approximate 1,2-alternate
conformation11 with the phosphorus-bound oxygens on one
“side” and the titanium-bound oxygens on the opposite; the
phosphorus lone pair points away from the titanium. Perhaps
the best way to describe the overall geometry of the
zirconium analogue 5b is that it is similar to 4b, except that
the phosphorus is inverted, allowing the lone pair to bind to
the zirconium. Complex 4c is again monomeric (Figure 5),
but the calix[5]arene conformation is no longer in the cone
conformation as in 4a. The conformation can be ap-
proximated as 1,2-alternate. The most important observation
in this complex is the P‚‚‚Ti distance; it is 2.90 Å, 0.8 Å
shorter than in 4a. These structures are consistent with the
1H NMR observations: overall Cs symmetry and, of the three
titanium complexes, the only one that shows phosphorus
coupling to the titanium-bound dimethylamino group cor-
responds to the structure with the shortest P‚‚‚Ti distance.
Upfield movement of the chemical shifts is also reasonable.
In the cone conformation (4a and 5a), the methyl protons
of the dialkylamino groups cannot interact with the π systems
of the aromatic rings, and these spectra show the most
downfield chemical shifts for these resonances. In all of the
other products, these protons can, at least to some extent,
interact with the rings. For example, in 4c, the C1‚‚‚aromatic
carbon (C41-C46) distances range from 3.5 to 3.7 Å. The
C3‚‚‚(C11-C16) range is significantly longer (4.9-5.3 Å).
This long distance is the likely reason for the modest upfield
shift of the PNMe2 resonance (∆δ ) -0.5) compared to
TiNMe2 (∆δ ) -1.5). Similar arguments explain the other
upfield resonances.
(11) The four idealized conformations of calix[5]arenes are the same as
calix[4]arenes: cone; partial cone; 1,2-alternate; 1,3-alternate.
(12) Olmstead, M. M.; Sigel, G.; Hope, H.; Xu, X.; Power, P. P. J. Am.
Chem. Soc. 1985, 107, 8087.
Inorganic Chemistry, Vol. 45, No. 16, 2006 6495

Fan et al.
the present case, this constraint should stabilize the addition
of a fifth ligand to the metal coordination. The observation
that this interaction is quite variable suggests that other
factors are important.
P-W bonds. The metal is also important, most clearly seen
from a comparison of 4a and 5a. In the cone conformation,
the phosphorus lone pair is ideally arranged to bind to the
metal, yet only in the case of zirconium can the observed
distance be described as an interaction, albeit weak. The
larger zirconium, accompanied by larger M-O distances, is
better able to accommodate the extra ligand within the calix-
[5]arene. Comparison of 4a and 4c demonstrates the impor-
tance of calix[5]arene conformation, with the 1,2-alternate
conformation yielding a complex with a markedly shorter
P‚‚‚Ti distance. Part of the reason for this decreased distance
in 4c may be the reduced steric repulsion of the dimethyl-
amino groups in this conformation; however, this is not the
only reason, since the tungsten complex 3 exhibits the cone
conformation for the calix[5]arene, and it has the shorter P‚
‚‚W distance compared to 2 in which the calix[5]arene adopts
an approximate 1,2-alternate conformation. We are currently
investigating the binding of a calix[5]arene with two
phosphorus ligands in order to compare the constraints in
that ligand with 1.
Acknowledgment is made to the National Science Foun-
dation (CHE-9522606), Robert A. Welch Foundation, and
the SMU University Research Council.
The P‚‚‚M interaction in all complexes is weaker (or
nonexistent) compared to normal bonds. The calix[5]arene
backbone apparently restricts the approach of the phosphorus.
This was previously observed3 in the tungsten complex 3
that exhibits a P-W distance at the longer end of known
Supporting Information Available: A single file containing
CIF tables for structures. This material is available free of charge
via the Internet at http://pubs.acs.org.
IC0601283
6496 Inorganic Chemistry, Vol. 45, No. 16, 2006