Redox-active tripodal aminetris(aryloxide) complexes of titanium(IV)

Article abstract of DOI:10.1021/ic100347h

New sterically encumbered tripodal aminetris(aryloxide) ligands N(CH ₂C₆H₂-3-^tBu-5-X-2-OH)₃ (^tBu,XLH₃) with relatively electron-rich phenols are prepared by Mannich condensation (X = OCH₃) or by a reductive amination/Hartwig-Buchwald amination sequence on the benzyl-protected bromosalicylaldehyde (X = N[C₆H₄-p-OCH₃] ₂), followed by debenzylation using Pearlman's catalyst (Pd(OH) ₂/C). The analogous dianisylamino-substituted compound lacking the tert-butyl group ortho to the phenol (^H,An₂NLH ₃) is also readily prepared. The ligands are metalated by titanium(IV) tert-butoxide to form the five-coordinate alkoxides LTi(O ^tBu). Treatment of the tert-butoxides with aqueous HCl yields the five-coordinate chlorides LTiCl, and with acetylacetone gives the six-coordinate diketonates LTi(acac). The diketonate complexes ^tBu,XLTi(acac) show reversible ligand-based oxidations with first oxidation potentials of +0.57, +0.33, and -0.09 V (vs ferrocene/ferrocenium) for X = ^tBu, MeO, and An₂N, respectively. Both dianisylamine-substituted complexes ^R,An₂NLTi(acac) (R = ^tBu, H) show similar electrochemistry, with three one-electron oxidations closely spaced at ～0 V and three oxidations due to removal of a second electron from each diarylaminoaryloxide arm at ～ + 0.75 V. The new electron-rich tripodal ligands therefore have the capacity to release multiple electrons at unusually low potentials for aryloxide groups.

Full text of DOI:10.1021/ic100347h

Inorg. Chem. 2010, 49, 4687–4697 4687
DOI: 10.1021/ic100347h
Redox-Active Tripodal Aminetris(aryloxide) Complexes of Titanium(IV)
Davide Lionetti, Andrew J. Medvecz, Vesela Ugrinova, Mauricio Quiroz-Guzman, Bruce C. Noll, and Seth N. Brown*
Department of Chemistry and Biochemistry, 251 Nieuwland Science Hall, University of Notre Dame,
Notre Dame, Indiana 46556-5670
Received February 21, 2010
New sterically encumbered tripodal aminetris(aryloxide) ligands N(CH2C6H2-3-tBu-5-X-2-OH)3 (tBu,XLH3) with
relatively electron-rich phenols are prepared by Mannich condensation (X = OCH3) or by a reductive amination/
Hartwig-Buchwald amination sequence on the benzyl-protected bromosalicylaldehyde (X = N[C6H4-p-OCH3]2),
followed by debenzylation using Pearlman’s catalyst (Pd(OH)2/C). The analogous dianisylamino-substituted
compound lacking the tert-butyl group ortho to the phenol (H,An NLH3) is also readily prepared. The ligands are
2
metalated by titanium(IV) tert-butoxide to form the five-coordinate alkoxides LTi(OtBu). Treatment of the tert-butoxides
with aqueous HCl yields the five-coordinate chlorides LTiCl, and with acetylacetone gives the six-coordinate
diketonates LTi(acac). The diketonate complexes tBu,XLTi(acac) show reversible ligand-based oxidations with first
oxidation potentials of þ0.57, þ0.33, and -0.09 V (vs ferrocene/ferrocenium) for X=tBu, MeO, and An2N, respec-
tively. Both dianisylamine-substituted complexes R,An NLTi(acac) (R=tBu, H) show similar electrochemistry, with three
2
one-electron oxidations closely spaced at ∼0 V and three oxidations due to removal of a second electron from each
diarylaminoaryloxide arm at ∼ þ 0.75 V. The new electron-rich tripodal ligands therefore have the capacity to release
multiple electrons at unusually low potentials for aryloxide groups.
Introduction
ability to engage in π bonding, aryloxide complexes of these
metals are difficult to oxidize, and the corresponding aroxyl
radical complexes have been relatively little studied. One
example of such a study is that of scandium(III) complexes of
sterically hindered tris(2-hydroxybenzyl)triazacyclononanes
(Figure 1a), where the hexadentate, trianionic ligand confers
enough stability on the oxidized complexes to observe revers-
ible electrochemical oxidation.6 Titanium(IV) complexes of a
similar ligand have been reported, but their oxidation was not
described.7
While aryloxides have a long history as ligands to transi-
tion and main group metals, the relatively high oxidation
potential of simple aryloxides means that they have only
recently emerged as an important class of redox-active, “non-
innocent” ligands.1,2 In particular, interest in the enzyme
galactose oxidase, which involves a copper-coordinated tyro-
syl radical,3 has spurred the preparation and characterization
of copper complexes of phenoxyl radicals as spectroscopic,
structural, or functional models of the enzyme.4 Other late-
metal aroxyl complexes have also been studied.5 In contrast,
because of the oxophilicity of d0 transition metals and their
€
€
(5) (a) Muller, J.; Kikuchi, A.; Bill, E.; Weyhermuller, T.; Hildebrandt, P.;
Ould-Moussa, L.; Wieghardt, K. Inorg. Chim. Acta 2000, 297, 265–277. (b)
Rotthaus, O.; Thomas, F.; Jarjayes, O.; Philouze, C.; Saint-Aman, E.; Pierre, J.-L.
Chem.;Eur. J. 2006, 12, 6953–6962. (c) Rotthaus, O.; Jarjayes, O.; Thomas, F.;
Philouze, C.; Del Valle, C. P.; Saint-Aman, E.; Pierre, J.-L. Chem.;Eur. J. 2006,
12, 2293–2302. (d) Benisvy, L.; Kannappan, R.; Song, Y.-F.; Milikisyants, S.;
Huber, M.; Mutikainen, I.; Turpeinen, U.; Gamez, P.; Bernasconi, L.; Baerends,
E. J.; Hartl, F.; Reedijk, J. Eur. J. Inorg. Chem. 2007, 637–642. (e) Shimazaki, Y.;
Kabe, R.; Huth, S.; Tani, F.; Naruta, Y.; Yamauchi, O. Inorg. Chem. 2007, 46,
6083–6090. (f) Shimazaki, Y.; Yajima, T.; Tani, F.; Karasawa, S.; Fukui, K.;
Naruta, Y.; Yamauchi, O. J. Am. Chem. Soc. 2007, 129, 2559–2568. (g) Storr, T.;
Wasinger, E. C.; Pratt, R. C.; Stack, T. D. P. Angew. Chem., Int. Ed. 2007, 46,
5198–5201. (h) Shimazaki, Y.; Yajima, T.; Shiraiwa, T.; Yamauchi, O. Inorg.
Chim. Acta 2009, 362, 2467–2474. (i) Shimazaki, Y.; Stack, T. D. P.; Storr, T.
Inorg. Chem. 2009, 48, 8383–8392. ( j) Rotthaus, O.; Jarjayes, O.; Philouze, C.;
Del Valle, C. P.; Thomas, F. Dalton Trans. 2009, 1792–1800.
*To whom correspondence should be addressed. E-mail: seth.n.brown.114@
nd.edu.
(1) Chaudhuri, P.; Wieghardt, K. Prog. Inorg. Chem. 2001, 50, 151–216.
(2) The non-innocent character of catecholates, which are readily oxi-
dized to semiquinones, has been recognized for a long time. See: Pierpont, C. G.;
Lange, C. W. Prog. Inorg. Chem. 1994, 41, 331–442 and references therein.
€
(3) (a) Kruger, H.-J. Angew. Chem., Int. Ed. 1999, 38, 627–631. (b)
Jazdzewski, B. A.; Tolman, W. B. Coord. Chem. Rev. 2000, 200, 633–685. (c)
Mahadevan, V.; Klein Gebbink, R. J. M.; Stack, T. D. P. Curr. Opin. Chem. Biol.
2000, 4, 228–234.
(4) (a) Wang, Y.; DuBois, J. L.; Hedman, B.; Hodgson, K. O.; Stack,
T. D. P. Science 1998, 279, 537–540. (b) Chaudhuri, P.; Hess, M.; Weyherm€uller,
T.; Wieghardt, K. Angew. Chem., Int. Ed. 1999, 38, 1095–1098. (c) Chaudhuri,
P.; Hess, M.; M€uller, J.; Hildenbrand, K.; Bill, E.; Weyherm€uller, T.; Wieghardt,
K. J. Am. Chem. Soc. 1999, 121, 9599–9610. (d) Rotthaus, O.; Jarjayes, O.;
Thomas, F.; Philouze, C.; Saint-Aman, E.; Pierre, J.-L. Dalton Trans. 2007, 889–
895. (e) Mukherjee, A.; Lloret, F.; Mukherjee, R. Inorg. Chem. 2008, 47, 4471–
4480. (f) Storr, T.; Verma, P.; Pratt, R. C.; Wasinger, E. C.; Shimazaki, Y.; Stack,
T. D. P. J. Am. Chem. Soc. 2008, 130, 15448–15459.
(6) Adam, B.; Bill, E.; Bothe, E.; Goerdt, B.; Haselhorst, G.; Hildenbrand,
€
K.; Sokolowski, A.; Steenken, S.; Weyhermuller, T.; Wieghardt, K. Chem.;
Eur. J. 1997, 3, 308–319.
€
(7) Auerbach, U.; Weyhermuller, T.; Wieghardt, K.; Nuber, B.; Bill, E.;
Butzlaff, C.; Trautwein, A. X. Inorg. Chem. 1993, 32, 508–519.
r
2010 American Chemical Society
Published on Web 04/16/2010
pubs.acs.org/IC

4688 Inorganic Chemistry, Vol. 49, No. 10, 2010
Lionetti et al.
N(CH2C6H2-3-tBu-5-OCH3-2-O)3
(
tBu,MeOL3-). We also
3-
describe the preparation of tripodal ligands where a di-(4-
methoxyphenyl)amino group (dianisylamino, An2N) is para
3-
to the phenoxide oxygen, N(CH2C6H2-3-X-5-NAn2-2-O)3
(X = H, H,An NL3-; X = tBu, tBu,An NL3-). Since triaryl-
amines are themselves redox-active and relatively electron-
rich,13 this substituent should introduce a locus of oxidation
more remote from the metal center than in unsubstituted
aryloxides. If this decreases the separation between successive
redox couples of the complexes, as well as lowering their
average oxidation potential, it would make this novel aryl-
oxide an attractive participant in efficient multielectron
reactions.
2
2
Figure 1. Chelating triphenols for investigating the redox properties of
aryloxidescoordinated tod0 transition metals. (a) Hexadentatetris(3-tert-
butyl-2-hydroxybenzyl)triazacyclononanes.6 (b) Tetradentate tris(3-tert-
butyl-2-hydroxybenzyl)amines.
Experimental Section
Unless otherwise noted, all procedures were carried out on
the benchtop without precautions to exclude air or moisture.
When dry solvents were needed, chlorinated solvents were
Another important class of chelating triaryloxide ligands
are the tetradentate aminetri(aryloxides) N(CH2C6HnR4-n
-
˚
dried over 4 A molecular sieves, followed by CaH2. Benzene
3-
2-O)3 (Figure 1b), which have been used extensively to
form highly stable complexes with d0 metals such as titanium-
(IV).8 In contrast to the triazacyclononane-derived ligands,
the lower denticity of the aminetriphenoxides means that a
metal complex may have one or two open sites available for
coordination to exogenous ligands. This feature is important
if one wishes to link redox events at the aryloxide ligands
to stoichiometric or catalytic transformations of substrates
coordinated to the metal center. Such reactions, where d0
metal centers with redox-active ligands mediate substrate
oxidation or reduction, are being explored as a new mode of
metal-mediated oxidation.9
In addition to the coordinative properties of the metal, the
electrochemical properties of the ligands are critical if redox-
active ligands are to serve efficiently as electron reservoirs in
multielectron transformations. The ligand must be able to
store enough electrons to accomplish the desired transforma-
tion of the substrate, and must have a redox potential great
enough to allow the substrate reaction to be thermodynami-
cally favorable. For optimal energy efficiency, the overall
redox potential of the ligand should be close to the thermo-
dynamic redox potential of the substrate. In addition to
having a suitable average redox potential, it should ideally
also have closely spaced redox potentials for successive one-
electron oxidations. Disparate redox potentials would mean
that the final stages of ligand oxidation would require more
energy than needed for substrate oxidation or reduction,
resulting in overpotentials and concomitant energy losses.
Here we describe the redox properties of known titanium
and toluene were dried over sodium, and ether and tetra-
hydrofuran over sodium benzophenone ketyl. Dry N,N-dime-
thylformamide (DMF) was anhydrous grade from Acros and
was stored in the drybox until needed. Deuterated solvents
were obtained from Cambridge Isotope Laboratories, dried
using the same procedures as their protio analogues, and
werestoredinthedryboxpriortouse. [N(CH2C6H2-3,5-tBu2-
2-O)3]Ti(OtBu) (tBu,tBuLTi[OtBu]),10 tBu,tBuLTi(acac),12 and
5-bromo-3-tert-butyl-2-hydroxybenzaldehyde14 were pre-
pared as described in the literature. All other reagents were
commerciallyavailableandusedwithoutfurtherpurification.
Routine NMR spectra were measured on a Varian VXR-300
spectrometer. Chemical shifts for 1H and 13C{1H} spectra are
reported in parts per million (ppm) downfield of tetramethyl-
silane (TMS), using the known chemical shifts of the solvent
residuals. Infrared spectra were recorded on NaCl plates on a
Perkin-Elmer PARAGON 1000 FT-IR spectrometer. FAB
mass spectra were obtained on a JEOL LMS-AX505HA
mass spectrometer using the FAB ionization mode and
3-nitrobenzyl alcohol or nitrophenyl octyl ether as a matrix.
Peaks reported are the mass number of the isotopomer
containing all of the most abundant isotope of theconstituent
elements; isotope envelopes showed satisfactory agreement
with calculated distributions. UV-visible spectra were mea-
sured on dichloromethane solutions (unless otherwise noted)
in 1 cm quartz cuvettes using a Beckman DU-7500 diode
array spectrophotometer. Elemental analyses were perfor-
med by M-H-W Laboratories (Phoenix, AZ).
Tris(3-tert-butyl-5-methoxy-2-hydroxybenzyl)amine, tBu,MeOLH3.
Into a 100 mL round-bottom flask were added a stir bar, 2.42 g
of hexamethylenetetramine (Aldrich, 17.2 mmol), 37.17 g of
2-tert-butyl-4-methoxyphenol (Aldrich, 206.2 mmol), and6.2 mL
of 37% aqueous formaldehyde (83 mmol). The flask was fitted
with a reflux condenser and a heating mantle, and was heated to
reflux. After 3 d, reflux was stopped, and the reaction mixture
was allowed to cool to room temperature. The mixture was
dissolved in 200 mL of CHCl3 and transferred to a 1000 mL
beaker. To the mixture was added 200 mL of 1 M hydrochloric
acid and a magnetic stir bar. After several minutes of stirring, a
white precipitate formed. The aqueous layer was decanted off,
and the organic layer was filtered on a glass frit and the solid
washed with 2 ꢀ 20 mL of chloroform. The solid was transferred
to a 1 L Erlenmeyer flask, and 400 mL of CH2Cl2 and 500 mL of
saturated aqueous sodium bicarbonate solution were added.
The mixture was stirred overnight to dissolve all the solid. The
complexes of the di-tert-butyl-substituted tripod N(CH2C6-
H2-3,5-tBu2-2-O)3
(
tBu,tBuL3-),10-12 as well as those of the
3-
new, more electron-rich, methoxy-substituted tripodal ligand
(8) Licini, G.; Mba, M.; Zonta, C. Dalton Trans. 2009, 5265–5277.
(9) (a) Haneline, M. R.; Heyduk, A. F. J. Am. Chem. Soc. 2006, 128, 8410–
8411. (b) Stanciu, C.; Jones, M. E.; Fanwick, P. E.; Abu-Omar, M. M. J. Am.
Chem. Soc. 2007, 129, 12400–12401. (c) Zarkesh, R. A.; Ziller, J. W.; Heyduk,
A. F. Angew. Chem., Int. Ed. 2008, 47, 4715–4718. (d) Blackmore, K. J.; Lal, N.;
Ziller, J. W.; Heyduk, A. F. J. Am. Chem. Soc. 2008, 130, 2728–2729. (e)
Nguyen, A. I.; Blackmore, K. J.; Carter, S. M.; Zarkesh, R. A.; Heyduk, A. F.
J. Am. Chem. Soc. 2009, 131, 3307–3316.
(10) Kol, M.; Shamis, M.; Goldberg, I.; Goldschmit, Z.; Alfi, S.;
Hayut-Salant, E. Inorg. Chem. Commun. 2001, 4, 177–179.
(11) Ugrinova, V.; Ellis, G. A.; Brown, S. N. Chem. Commun. 2004, 468–
469.
(13) Amthor, S.; Noller, B.; Lambert, C. Chem. Phys. 2005, 316, 141–152.
(14) Cavazzini, M.; Manfredi, A.; Montanari, F.; Quici, S.; Pozzi, G. Eur.
J. Org. Chem. 2001, 4639–4649.
(12) Fortner, K. C.; Bigi, J. P.; Brown, S. N. Inorg. Chem. 2005, 44, 2803–
2814.

Article
Inorganic Chemistry, Vol. 49, No. 10, 2010 4689
organic layer was collected, dried over magnesium sulfate,
filtered, and the solvent evaporated on a rotary evaporator
to obtain a light yellow oil. The oil was dissolved in methanol
(100 mL), and the solution allowed to stand for 3 d at room
temperature, at which point large colorless blocks had precipi-
tated. The solid was filtered, washed with methanol, and air-
solution of 28.0 mg of tri-tert-butylphosphine (Strem, 0.138
mmol, 1.8 mol % per Br) in 2 mL of benzene. The vial was then
capped tightly and taken out of the drybox. After 17 h stirring at
room temperature, the vial was opened to the air and the
reaction mixture poured into a 50 mL round-bottom flask con-
taining 15 mL of saturated aqueous ammonium chloride. The
reaction vial was rinsed with a few milliliters of ether, and the
washing combined with the rest of the mixture. The aqueous/
organic mixture was stirred for 5 min and then transferred to a
separatory funnel. The organic layer was collected, dried over
magnesium sulfate, filtered, and evaporated to dryness to obtain
3.17 g of a dark brown oil (95%). 1H NMR (CDCl3): δ 3.64 (s,
18H, OCH3), 3.68 (s, 6H, NCH2Ar), 4.88 (s, 6H, OCH2Ph), 6.61
(d, 9 Hz, 12H, 2,6-NC6H4OCH3), 6.65 (br m, 6H, Ar 4-H, 6-H),
6.81 (d, 9 Hz, 12H, 3,5-NC6H4OCH3), 7.12 (br m, 3H, Ar 3-H),
7.31 (m, 15H, OCH2C6H5). 13C{1H} NMR: δ 53.16 (NCH2),
55.53 (OCH3), 70.21 (OCH2Ph), 112.23, 114.43, 122.64, 124.46,
125.25, 127.28, 127.74, 128.64, 129.40, 137.92, 141.75, 142.17,
152.33, 154.62. IR (evapd film, cm-1): 3041 (w), 2999 (w), 2947
(w), 2906 (w), 2833 (w), 1605 (w), 1588 (w), 1503 (s), 1463 (m),
1372(w), 1280 (w), 1238(s), 1180 (w), 1106(w), 1037 (m), 917 (w),
825 (w), 736 (w), 697 (w). FAB MS: 1287 (MþþH). Anal. Calcd
for C84H78N4O9: C, 78.35; H, 6.11; N, 4.35. Found: C, 78.41; H,
6.16; N, 4.24.
1
dried on a glass frit to yield 4.19 g of tBu,MeOLH3 (10%). H
NMR (CDCl3): δ 1.37 (s, 27H, ArC(CH3)3), 3.61 (s, 6H,
NCH2Ar), 3.75 (s, 9H, OCH3), 6.54 (d, 3 Hz, 3H, ArH), 6.83
(d, 3 Hz, 3H, ArH). 13C{1H} NMR (CDCl3): δ 29.71, 35.15,
55.87, 56.72, 113.32, 114.13, 123.33, 139.30, 147.90, 152.85. IR
(evaporated film, cm-1): 3542 (s, br, νOH), 2999 (m), 2954 (s),
2906 (s), 2835 (m), 1605 (s), 1476 (s), 1456 (s), 1434 (s), 1392 (m),
1361 (m), 1311 (m), 1270 (m), 1213 (s), 1134 (m), 1094 (m), 1059
(s), 1024 (w), 1001 (w), 978 (w), 945 (w), 916 (w), 854 (m), 800
(m), 770 (m), 755 (m), 735 (m), 704 (w), 690 (w). FAB-MS: 593
(Mþ). Anal. Calcd for C36H51NO6: C, 72.82; H, 8.66; N, 2.36.
Found: C, 72.51; H, 8.34; N, 2.46.
2-Benzyloxy-5-bromobenzaldehyde. In a 100 mL round-
bottom flask were placed 5.00 g of 5-bromosalicylaldehyde
(Alfa Aesar, 24.9 mmol), 13.12 g of potassium carbonate
(95.0 mmol, 3.8 equiv), and a magnetic stirbar. The flask was
filled with nitrogen and capped with a septum, and 23 mL of
dry DMF were added, followed by 5.08 g of benzyl bromide
(29.7 mmol, 1.2 equiv) dissolved in 10 mL of dry DMF. After
stirring 18 h, the reaction mixture was poured into 107 mL of
distilled water, forming a light yellow precipitate. The solution
was filtered on a glass frit, and the solid washed with H2O
to obtain 7.19 g of 2-benzyloxy-5-bromobenzaldehyde (99%).
Spectroscopic data were in agreement with published values.15
Tris(2-benzyloxy-5-bromobenzyl)amine, N(CH2C6H3-5-Br-2-
OCH2Ph)3. In the drybox, ammonium acetate (0.359 g, 4.66
mmol), 2-benzyloxy-5-bromobenzaldehyde (4.07 g, 13.98 mmol,
3.0 equiv), and sodiumtriacetoxyborohydride(4.44g, 21.0 mmol,
4.5 equiv) were placed into a 50 mL round-bottom flask. A
stirbar and dry tetrahydrofuran (THF) were added, and the
flask was stoppered with a rubber septum. After stirring for 22 h,
the volatiles were removed on a rotary evaporator. The oily
residue was dissolved in ethyl acetate, and the resulting solution
was washed with 2 ꢀ 35 mL of 5% aqueous KOH and 35 mL of
brine. The organic layer was collected, dried over magnesium
sulfate, and evaporated to dryness to yield a light yellow oil
which solidified into a white solid. The solid was washed with
methanol and air-dried on a glass frit to obtain 2.19 g of tris(2-
benzyloxy-5-bromobenzyl)amine (26.0 mmol, 56%). 1H NMR
(CDCl3): δ 3.73 (s, 6H, NCH2Ar), 5.05 (s, 6H, OCH2Ph), 6.69
(d, 9 Hz, 3H, Ar 3-H), 7.20 (dd, 9 Hz, 2.5 Hz, 3H, Ar 4-H), 7.34
(m, 15H, OCH2C6H5), 7.65 (d, 2.5 Hz, 3H, Ar 6-H). 13C{1H}
NMR (CDCl3): δ 52.77 (NCH2), 70.39 (OCH2Ph), 113.35,
113.57, 127.26, 128.05, 128.81, 130.54, 130.62, 132.92, 137.11,
155.97. IR (evapd film, cm-1): 3093 (w), 3065 (w), 3032 (w), 2926
(w), 2867 (w), 1591 (m), 1485 (s), 1454 (s), 1403 (m), 1371 (m),
1240 (s), 1179 (m), 1135 (m), 1105 (m), 1025 (m), 970 (w), 890
(w), 848 (w), 802 (m), 735 (s), 696 (s), 652 (m). FAB-MS: 840
(MþþH, 79Br3 isotopomer). Anal. Calcd for C42H36Br3NO3: C,
59.88; H, 4.31; 1.66. Found: C, 60.16; H, 4.54; N, 1.67.
Tris(2-hydroxy-5-[(di-p-methoxyphenyl)amino]benzyl)amine,
H,An NLH3. To a solution of the benzyl-protected ligand
2
N(CH2C6H3-5-[N(C6H4OCH3)2]-2-OCH2Ph)3 (3.07 g, 2.39 mmol)
in 20 mL of EtOAc were added 0.31 g of palladium, 10 wt % on
activated carbon (Aldrich). The flask was flushed with hydrogen
and stirred 5 d under balloon pressure of H2. The mixture was
gravity filtered, and the solvent evaporated from the filtrate to
obtain a dark brown oil. Trituration with methanol/water gave a
light brown precipitate, which was isolated by suction filtra-
tion and dried under vacuum overnight to obtain 2.23 g (92%)
of N(CH2C6H3-5-[N(C6H4OCH3)2]-2-OH)3, H,An NLH3. 1H
2
NMR (CDCl3): δ 3.62 (br s, 6H, NCH2Ar), 3.75 (s, 18H,
OCH3), 6.75 (m, 21H, ArH þ 2,6-NC6H4OCH3), 6.94 (br d, 9
Hz, 12H, 3,5-NC6H4OCH3), 7.46 (v br s, 3H, OH). 13C{1H}
NMR (CDCl3): δ 55.68 (NCH2), 57.23 (OCH3), 114.69, 117.25,
119.75, 123.25, 124.91, 126.35, 141.25, 142.14, 151.31, 155.02.
IR (evapd film, cm-1): 3310 (w, br, νOH), 2998 (w), 2926 (w),
2849 (w), 2906 (w), 2833, (w), 1662 (w), 1606 (w), 1586 (w),
1504 (s), 1464 (m), 1440 (m), 1322 (w), 1239 (s), 1178 (m),
1107 (w), 1036 (m), 953 (w), 877 (w), 826 (m), 779 (w), 718
(w), 580 (w). FAB-MS: 1017 (M þ H)þ. Anal. Calcd for
C63H60N4O9: C, 74.39; H, 5.95; N, 5.51. Found: C, 74.67; H,
6.17; N, 5.31.
2-Benzyloxy-3-tert-butyl-5-bromobenzaldehyde. 5-Bromo-
3-tert-butyl-2-hydroxybenzaldehyde (11.22 g, 43.6 mmol) was
benzylated in DMF (40 mL) using PhCH2Br (8.95 g, 52.3 mmol,
1.2 equiv, dissolved in 30 mL of DMF) and K2CO3 (22.91 g, 64.9
mmol, 3.8 equiv) as described above for 5-bromo-2-hydroxy-
benzaldehyde. After quenching the reaction with water
(300 mL), the mixture was extracted with 100 mL of hexane.
The organic layer was collected, dried over magnesium sulfate,
and the solvent evaporated to give a yellow oil. Purification
by column chromatography on silica gel (15:1 hexanes/ether)
yielded 10.83 g (71%) of the benzylated bromosalicylaldehyde
as a pale yellow solid. 1H NMR (CDCl3): δ 1.44 (s, 9H,
C(CH3)3), 5.04 (s, 2H, OCH2Ph), 7.49 (m, 5H, OCH2C6H5),
7.70 (d, 2.4 Hz, 1H, ArH), 7.85 (d, 2.4 Hz, 1H, ArH). 13C{1H}
NMR (CDCl3): δ 30.89 (C(CH3)3), 35.69 (C(CH3)3), 81.05
(OCH2), 117.82 (CBr), 127.26, 128.64, 128.98, 130.31, 131.61,
136.10, 136.58, 146.82, 161.07 (COCH2Ph), 189.06 (CHO). IR
(evapd film, cm-1): 3093 (w), 3061 (w), 3030 (w), 2963 (w), 2916
(w), 2871 (w), 2740 (w), 1689 (s, νCdO), 1607 (w), 1569 (m), 1498
(w), 1455 (m), 1433 (s), 1370 (s), 1332 (w), 1311 (w), 1236 (s),
1206 (s), 1156 (s), 1104 (w), 1078 (w), 1000 (m), 969 (m), 918 (w),
882 (m), 865 (w), 842 (w), 803 (m), 765 (w), 735 (m), 696 (m), 661
Tris(2-benzyloxy-5-[(di-p-methoxyphenyl)amino]benzyl)amine,
N(CH2C6H3-5-[N(C6H4OCH3)2]-2-OCH2Ph)3. In the drybox,
tris(2-benzyloxy-5-bromobenzyl)amine (2.19 g, 2.60 mmol) was
dissolved in 7 mL of dry benzene. Into a 20 mL screw-cap scin-
tillation vial were added 1.79 g of 4,40-dimethoxydiphenylamine
(Alfa Aesar, 7.81 mmol, 3.0 equiv), 1.13 g of sodium tert-but-
oxide (Aldrich, 11.8 mmol, 4.5 equiv), 80.0 mg of tris(dibenzylid-
eneacetone)dipalladium (Strem, 0.0874 mmol, 2.2 mol % Pd per
Br), and a stirbar. The tris(2-benzyloxy-5-bromobenzyl)amine
solution was added to the solid reagents in the vial, followed by a
(15) Fresneda, P. M.; Molina, P.; Bleda, J. A. Tetrahedron 2001, 57, 2355–
2363.

4690 Inorganic Chemistry, Vol. 49, No. 10, 2010
Lionetti et al.
(m). FAB MS: 346 (Mþ, 79Br isotopomer). Anal. Calcd for
C18H19BrO2: C, 62.26; H, 5.51. Found: C, 62.68; H, 5.11.
Tris(2-benzyloxy-3-tert-butyl-5-bromobenzyl)amine. In a 50
mL round-bottom flask in the drybox were combined 0.95 g
of ammonium acetate (12.3 mmol), 12.83 g of 2-benzyloxy-
5-bromo-3-tert-butylbenzaldehyde (36.96 mmol, 3.0 equiv), and
11.75 g of sodium triacetoxyborohydride (55.44 mmol, 4.5
equiv). A stirbar was added, and the flask sealed with a Teflon
needle valve. Dry 1,2-dichloroethane (100 mL) was added to the
reaction flask by vacuum transfer, and the reaction mixture was
stirred for 3 d. The volatiles were removed on a rotary evapora-
tor, and the residue was dissolved in EtOAc (2 ꢀ 100 mL). The
solution was washed with 2 ꢀ 50 mL of 5% aqueous KOH solu-
tion and 50 mL of brine. The organic layer was collected, dried
over magnesium sulfate, and the solvent was evaporated to
obtain the crude product as a light yellow oil. After standing 2 d
at room temperature, the mixture deposited clear crystals which
were washed with hexane. The hexane supernatant was then
evaporated and purified by column chromatography on silica
gel, eluting with 5% ethyl acetate in hexanes to give a light
yellow solid that was recrystallized from hot hexanes. The
combined products yielded 7.31 g of tris(2-benzyloxy-5-bromo-
gravity filtration, and the solvent was evaporated to yield a
crude brown oil. The oil was dissolved in 30 mL of CH2Cl2 and
washed with 100 mL of 0.01 M aqueous sodium dithionite to
give a clear, yellow solution. The biphasic mixture was allowed
to stir under nitrogen for 30 min. The organic layer was dried
over MgSO4 under N2, then filtered under N2 to give a yellow
solution. The solvent was removed in vacuo to give a pale yellow
solid, which was isolated and stored in the drybox (5.59 g, 73%).
1H NMR (CDCl3): δ 1.31 (s, 27H, C(CH3)3), 3.45 (s, 6H,
NCH2), 3.77 (s, 18H, OCH3), 6.27 (br s, 3H, OH), 6.60 (d, 2.5
Hz, 3H, ArH), 6.75 (d, 9 Hz, 12H, NC6H4OCH3 2,6-H), 6.94 (d,
9 Hz, 12H, NC6H4OCH3 3,5-H), 6.97 (d, 2.5 Hz, 3H, ArH).
13C{1H} NMR (CDCl3): 29.77 (C(CH3)3), 35.04 (C(CH3)3),
55.64 (NCH2), 56.26 (OCH3), 114.66, 123.27, 123.38, 124.12,
124.91, 138.64, 141.07, 142.14, 149.20, 154.99. IR (nujol mull,
cm-1): 2951 (m), 2852 (m), 1603 (w), 1502 (m), 1364 (w), 1332
(w), 1297 (w), 1270 (w), 1237 (m), 1178 (w), 1134 (w), 1105 (w),
1035 (w), 1009 (w), 872 (w), 856 (w), 824 (w), 793 (w), 770 (w),
724 (w). FAB-MS: 1185 (Mþ). Anal. Calcd for C75H84N4O9: C,
75.99; H, 7.14; N, 4.73. Found: C, 75.60; H, 7.30; N, 4.71.
tBu,MeOLTi(OtBu). In the drybox, 0.807 g of tBu,MeOLH3 (1.36
mmol) and 0.464 g of titanium tert-butoxide (Strem, 1.36 mmol)
were dissolved in 30 mL of CH2Cl2 in a 100 mL round-bottom
flask. The reaction was stirred for 5 min under N2. Evaporation
of the solvent on the rotary evaporator yielded 0.963 g (99%) of
bright yellow air-stable tBu,MeOLTi(OtBu). 1H NMR (CDCl3): δ
1.45 (s, 27H, ArC(CH3)3), 1.63 (s, 9H, OC(CH3)3), 2.84 (d, 14
Hz, 3H, CHH0), 3.74 (s, 9H, OCH3), 3.95 (d, 14 Hz, 3H, CHH0),
6.50 (d, 3 Hz, 3H, ArH), 6.77 (d, 3 Hz, 3H, ArH). 13C{1H} NMR
(CDCl3): δ 29.72 (ArC(CH3)3), 32.62 (OC(CH3)3), 35.34
(ArC(CH3)3), 55.88 (NCH2), 58.76 (OCH3), 85.76 (OC-
(CH3)3), 111.80, 112.63, 125.50, 137.55, 152.66, 157.50. IR
(evapd film, cm-1): 2951 (m), 1602 (w), 1464 (s), 1429 (s), 1358
(w), 1325 (w), 1235 (s), 1208 (s), 1178 (m), 1060 (m), 1024 (s), 921
(w), 848 (s), 795 (w), 762 (w). FAB-MS: 711 (Mþ). Anal. Calcd
for C40H57NO7Ti: C, 67.50; H, 8.07; N, 1.97. Found: C, 66.05;
H, 7.36; N, 2.00.
1
3-tert-butylbenzyl)amine (59%) as a colorless solid. H NMR
(CDCl3): δ 1.34 (s, 27H, C(CH3)3), 3.57 (s, 6H, NCH2), 4.69 (s,
6H, OCH2Ph), 7.30 (m, 15H, OCH2C6H5), 7.34 (d, 2.4 Hz, 3H,
ArH), 7.59 (d, 2.4 Hz, 3H, ArH). 13C{1H} NMR (CDCl3): δ
31.06 (C(CH3)3), 35.57 (C(CH3)3), 52.00 (NCH2), 75.93
(OCH2), 117.46 (CBr), 126.88, 127.84, 128.70, 129.65, 131.59,
134.84, 137.17, 145.41, 156.29. IR (evapd film, cm-1): 2962 (s),
2909 (m), 2876 (m), 1570 (w), 1498 (w), 1454 (m), 1430 (s), 1404
(w), 1367 (s), 1263 (w), 1211 (s), 1156 (s), 1106 (w), 1014 (w), 979
(m), 871 (w), 843 (w), 733 (m), 694 (m). FAB MS: 1007 (Mþ,
79Br3 isotopomer). Anal. Calcd for C54H60Br3NO3: C, 64.17; H,
5.98; N, 1.39. Found: C, 64.09; H, 6.18; N, 1.47.
Tris(2-benzyloxy-3-tert-butyl-5-[(di-p-methoxyphenyl)amino]-
benzyl)amine. Tris(2-benzyloxy-3-tert-butyl-5-bromobenzyl)-
amine (7.31 g, 7.23 mmol) was aminated with 4,40-dimethoxy-
diphenylamine (4.97 g, 21.7 mmol, 3 equiv) as described above
for tris(2-benzyloxy-5-bromobenzyl)amine, except that the re-
action was allowed to stir for 3 d. The crude material was puri-
fied by recrystallization from 2:1 hexanes/ether at -20 °C,
yielding a yellow crystalline solid. Three crops gave a total yield
of 9.32 g (88%). This compound is slightly sensitive to air and
tBu,An NLTi(OtBu). The compound was prepared ana-
2
logously using 0.4202 g of tBu,An NLH3 (0.3544 mmol) and
2
0.1246 g of titanium tert-butoxide (0.3624 mmol) to yield
0.4656 g (99%) of orange tBu,An NLTi(OtBu), which is stable
2
for brief periods in the air but should be stored in the drybox. 1H
NMR (CDCl3): δ 1.38 (s, 27H, ArC(CH3)3), 1.64 (s, 9H, OC-
(CH3)3), 2.63 (d, 13 Hz, 3H, CHH0), 3.77 (s, 18H, OCH3), 3.92
(d, 14 Hz, 3H, CHH0), 6.55 (d, 2.5 Hz, 3H, ArH), 6.75 (d, 9 Hz,
12H, NC6H4OCH3 2,6-H), 6.90 (d, 9 Hz, 12H, NC6H4OCH3
3,5-H), 6.93 (d, 2 Hz, 3H, ArH). 13C{1H} NMR (CDCl3): δ
29.88 (ArC(CH3)3), 32.59 (OC(CH3)3), 35.22 (ArC(CH3)3),
55.69 (OCH3), 58.49 (NCH2), 77.43 (OC(CH3)3), 114.64,
121.69, 122.47, 125.05, 125.76, 136.95, 140.87, 142.18, 155.02,
158.89. IR (evapd film, cm-1): 2953 (m), 2833 (w), 1599 (m),
1505 (s), 1463 (m), 1441 (m), 1389 (w), 1359 (w), 1335 (w), 1297
(w), 1240 (s), 1180 (m), 1144 (w), 1106 (w), 1037 (m), 1021 (m),
882 (m), 828 (m), 819 (m), 739 (m), 686 (w), 637 (w), 599 (w).
Anal. Calcd for C79H90N4O10Ti: C, 72.80; H, 6.96; N, 4.30.
Found: C, 72.65; H, 7.01; N, 4.08.
1
light and was stored wrapped in foil in the drybox. H NMR
(CDCl3): δ 1.22 (s, 27H, C(CH3)3), 3.58 (s, 6H, NCH2), 3.66 (s,
18H, OCH3), 4.67 (s, 6H, OCH2Ph), 6.64 (d, 9 Hz, 12H,
NC6H4OCH3 2,6-H), 6.81 (d, 2.5 Hz, 3H, ArH), 6.82 (d, 9 Hz,
12H, NC6H4OCH3 3,5-H), 6.97 (d, 2.5 Hz, 3H, ArH), 7.22 (m,
15H, OCH2C6H5). 13C{1H} NMR (CDCl3): δ 31.39 (C(CH3)3),
35.46 (C(CH3)3), 52.77 (NCH2), 55.65 (OCH3), 75.33 (OCH2-
Ph), 114.62, 121.41, 123.09, 125.18, 127.16, 127.57, 128.60,
133.45, 138.06, 142.02, 143.47, 143.81, 151.85, 155.01. IR
(evapd film, cm-1): 3036 (s), 2999 (s), 2955 (s), 2834 (s), 1681
(w), 1654 (w), 1595 (s), 1510 (s), 1463 (s), 1441 (s), 1391 (w), 1370
(s), 1328 (m), 1296 (m), 1242 (s), 1209 (s), 1180 (m), 1135 (w),
1106 (m), 1078 (w), 1038 (s), 1017 (m), 927 (w), 914 (w), 873 (w),
862 (w), 827 (s), 792 (w), 780 (m), 764 (m), 736 (s), 699 (s), 640
(m), 601 (s). FAB MS: 1455 (Mþþ H). Anal. Calcd for
C96H102N4O9: C, 79.20; H, 7.06; N, 3.85. Found: C, 79.40; H,
7.02; N, 3.59.
tBu,tBuLTiCl. A yellow solution of tBu,tBuLTi(OtBu) (0.5000 g,
0.665 mmol) in 20 mL of C6H6 was placed in a 125 mL sepa-
ratory funnel and shaken with 50 mL of 1 M HCl for 10 min. The
red C6H6 layer was collected and dried over MgSO4, and the
solvent was removed on a rotary evaporator to give 0.4551 g of
red tBu,tBuLTiCl (91%). 1H NMR (CDCl3): δ 1.28 (s, 27H,
C(CH3)3), 1.45 (s, 27H, C(CH3)3), 3.09 (d, 14 Hz, 3H, CHH0),
4.07 (d, 14 Hz, 3H, CHH0), 7.02 (d, 2 Hz, 3H, ArH), 7.25 (d, 2
Hz, 3H, ArH). 13C{1H} NMR (CDCl3): δ 29.77 (C(CH3)3),
31.79 (C(CH3)3), 34.70 (C(CH3)3), 35.13 (C(CH3)3), 58.99
(NCH2), 123.53, 123.76, 124.16, 135.98, 144.51, 161.06.
EI MS: 751 (Mþ). IR (evapd film, cm-1): 3091 (w), 3034 (w),
2959 (s), 2906 (s), 2869 (s), 1773 (w), 1601 (m), 1476 (s), 1446 (s),
Tris(2-hydroxy-3-tert-butyl-5-[(di-p-methoxyphenyl)amino]-
benzyl)amine, tBu,An NLH3. In a 250 mL round-bottom flask,
2
9.32 g of tris(2-benzyloxy-3-tert-butyl-5-[(di-p-methoxyphenyl)-
amino]benzyl)amine (6.40 mmol) was dissolved in 100 mL of
EtOAc. Pearlman’s catalyst (Pd(OH)2, 20% on activated car-
bon, wet [50% water], 3.73 g, 40% loading by mass) was added
to the solution. The flask was flushed with hydrogen, and the
reaction mixture was stirred for 5 d under a hydrogen-filled
balloon. The catalyst was removed from the reaction mixture by

Article
Inorganic Chemistry, Vol. 49, No. 10, 2010 4691
H,An NLTi(acac) (60%). 1H NMR (CDCl3): δ 1.97 (br s, 6H, acac
CH3), 3.70 (v br, 6 H, NCH2), 3.78 (s, 18H, OCH3), 5.68 (s, 1H,
acac CH), 6.46 (br d, 9 Hz, 3H, ArH H-3), 6.70 (d, 2.5 Hz, 3H,
ArH H-6), 6.74 (d, 9 Hz, 12H, NC6H4OCH3 2,6-H), 6.79 (dd, 9,
2.5 Hz, 3H, ArH H-4), 6.93 (d, 9 Hz, 12H, NC6H4OCH3 3,5-H).
13C{1H} NMR (CDCl3): 26.44 (acac CH3), 55.69 (OCH3), 61.50
(NCH2), 105.47 (acac CH), 114.67, 116.55, 124.30, 124.42,
125.08, 125.78, 140.94, 142.16, 155.04, 157.65; acac CO too broad
to be observed. IR (evapd film, cm-1): 3038 (m), 2997 (m), 2950
(m), 2932 (m), 2906 (m), 2834 (s), 1585 (s), 1503 (s), 1485 (s), 1441
(s), 1362 (m), 1267 (s), 1241 (s), 1180 (m), 1148 (w), 1108 (m), 1036
(s), 931 (m), 887 (s), 825 (s), 808 (s), 780 (m), 735 (m), 703 (w), 683
(w), 770 (m), 639 (w), 598 (m). FAB MS: 1160 (Mþ). UV-vis
(CH2Cl2) λmax 302 nm (ε=83400 M-1 cm-1), 358 (sh, 17900), 428
(17000). Anal. Calcd for C68H64N4O11Ti: C, 70.34; H, 5.56; N,
4.83. Found: C, 69.69; H, 5.64; N, 4.80.
2
1414 (m), 1393 (m), 1362 (s), 1306 (m), 1290 (m), 1254 (s), 1238
(s), 1204 (s), 1171 (s), 1127 (s), 1071 (m), 1028 (w), 981 (w), 962
(w), 918 (s), 866 (s), 814 (m), 764 (s), 740 (m), 693 (m), 677 (m),
650 (m), 609 (s), 590 (m), 570 (w). Anal. Calcd for C45H66-
ClNO3Ti: C, 71.84; H, 8.84; N, 1.86. Found: C, 72.03; H, 8.69;
N, 1.82.
tBu,MeOLTiCl. The compound was prepared from tBu,MeOLTi-
(OtBu) (0.373 g, 0.523 mmol) by the same procedure to give
1
0.334 g of tBu,MeOLTiCl (95%). H NMR (CDCl3): δ 1.44 (s,
27H, C(CH3)3), 3.04 (br d, 14 Hz, 3H, CHH0), 3.76 (s, 9H,
OCH3), 4.03 (br d, 14 Hz, 3H, CHH0), 6.55 (d, 3 Hz, 3H, ArH),
6.81 (d, 3 Hz, 3H, ArH). 13C{1H} NMR (CDCl3): δ 29.57
(C(CH3)3), 35.15 (C(CH3)3), 55.84 (OCH3), 58.63 (NCH2),
111.78, 112.90, 124.64, 138.40, 153.99, 157.84. IR (evapd film,
cm-1): 2993 (w), 2956 (s), 2906 (m), 2869 (m), 2835 (w), 1601 (s),
1467 (s), 1431 (s), 1392 (w), 1360 (w), 1327 (m), 1288 (w), 1231
(s), 1208 (s), 1177 (w), 1143 (m), 1058 (s), 1025 (w), 984 (w), 920
(m), 865 (s), 797 (w), 756 (w). FAB-MS: 673 (Mþ). Anal. Calcd
for C36H48ClNO6Ti: C, 64.14; H, 7.18; N, 2.08. Found: C, 63.94;
H, 6.93; N, 2.20.
tBu,An NLTi(acac). The compound was prepared analogously
2
to tBu,MeOLTi(acac) using 0.2016 g of tBu,An NLTi(OtBu) (0.1547
2
mmol) and 15.9 μL of 2,4-pentanedione (0.155 mmol, 1.0 equiv)
to yield 0.1837 g of tBu,An NLTi(acac) as a dark orange solid
2
(91%). 1H NMR (CDCl3): 1.24 (s, 27H, C(CH3)3), 1.93 (s, 6H,
acac CH3), 3.64 (s, 6 H, NCH2), 3.76 (s, 18H, OCH3), 5.58 (s,
1H, acac CH), 6.55 (d, 2.5 Hz, 3H, ArH), 6.73 (d, 9 Hz, 12H,
NC6H4OCH3 2,6-H), 6.88 (d, 2.5 Hz, 3H, ArH), 6.92 (d, 9 Hz,
12H, NC6H4OCH3 3,5-H). 13C{1H} NMR (CDCl3): 26.18 (acac
CH3), 29.84 (C(CH3)3), 35.06 (C(CH3)3), 55.90 (OCH3), 61.56
(NCH2), 104.91 (acac CH), 114.57, 122.75, 122.86, 124.73,
126.08, 136.47, 139.60, 142.42, 154.75, 157.57, 190.30 (acac
CO). IR (evapd film, cm-1): 2998 (w), 2952 (m), 2907 (w),
2833 (w), 1655 (m), 1638 (m), 1596 (m), 1503 (s), 1464 (m),
1441 (m), 1360 (w), 1265 (w), 1239 (s), 1180 (m), 1141 (w), 1037
(m), 933 (w), 875 (m), 825 (m), 813 (m), 779 (w), 736 (w), 709 (w),
668 (w). FAB-MS: 1330 (MþH)þ. UV-vis (CH2Cl2) λmax 300
nm (ε=80800 M-1 cm-1), 365 (sh, 13600), 432 (13300). Anal.
Calcd for C80H88N4O11Ti: C, 72.28; H, 6.67; N, 4.21. Found: C,
71.60; H, 7.05; N, 3.86.
tBu,An NLTiCl. The compound was prepared from tBu,An NLTi-
2
2
(OtBu) (0.2094 g, 0.1607 mmol) by the same procedure to give
0.1518 g of tBu,An NLTiCl (75%). 1H NMR (CDCl3): δ 1.26 (s,
2
27H, C(CH3)3), 2.67 (d, 14 Hz, 3H, CHH0), 3.68 (s, 18H, OCH3),
3.90 (d, 13 Hz, 3H, CHH0), 6.43 (d, 2.5 Hz, 3H, ArH), 6.68 (d, 9
Hz, 12H, NC6H4OCH3 2,6-H), 6.80 (d, 2.5 Hz, 3H, ArH), 6.85
(d, 9 Hz, 12H, NC6H4OCH3 3,5-H). 13C{1H} NMR (CDCl3): δ
29.71 (C(CH3)3), 35.07 (C(CH3)3), 55.68 (OCH3), 58.47 (NC-
H2), 114.79, 120.69, 120.88, 124.78, 125.76, 137.68, 141.65,
142.75, 155.52, 158.80. IR (evapd film, cm-1): 2998 (w), 2956
(m), 2908 (w), 2870 (w), 2834 (w), 2068 (w), 1638 (m), 1597 (m),
1507 (s), 1464 (m), 1441 (m), 1392 (w), 1359 (w), 1336 (w), 1298
(w), 1240 (s), 1217 (m), 1181 (w), 1168 (w), 1143 (w), 1105 (w),
1038 (m), 866 (w), 892 (m), 828 (m), 818 (m), 781 (w), 746 (m),
712 (w), 607 (w). Anal. Calcd for C75H81ClN4O9Ti: C, 71.17; H,
6.45; N, 4.43. Found: C, 71.31; H, 6.65; N, 4.34.
tBu,tBuLTi(acac) was prepared as described previously.12 UV-
vis (CH2Cl2) λmax 259 nm (ε=34000 M-1 cm-1), 340 (sh, 12000),
366 (12900).
Electrochemistry. Electrochemical measurements were per-
formed in the drybox using a BAS Epsilon potentiostat. A
standard three-electrode setup was used, with a glassy carbon
working electrode, Pt or glassy carbon counter electrode, and a
silver/silver chloride pseudoreference electrode. The electrodes
were connected to the potentiostat through electrical conduits in
the drybox wall. Samples were approximately 1 mM in CH2Cl2,
using 0.1 M Bu4NPF6 as the electrolyte. Potentials were refer-
enced to ferrocene/ferrocenium at 0 V,16 with the reference
potential established by spiking the test solution with a small
amount of ferrocene or decamethylferrocene (E°=-0.565 V).
Cyclic voltammograms were recorded with a scan rate of
60 mV s-1. Square wave voltammograms were acquired with
a step size of 3 mV, a pulse height of 30 mV, and a frequency of
15 Hz.
tBu,MeOLTi(acac). A solution of 0.300 g of tBu,MeOLTi(OtBu)
(0.421 mmol) in 15 mL of CH2Cl2 was treated with 0.041 mL of
2,4-pentanedione (0.42 mmol, 1.0 equiv) and stirred for 2 h.
Evaporation of the solvent yielded 0.296 g of tBu,MeOLTi(acac)
as a dark orange solid (95%). 1H NMR (CDCl3): δ 1.30 (s, 27H,
C(CH3)3), 1.78 (s, 6H, acac CH3), 3.73 (s, 9H, OCH3), 3.77 (s,
6H, NCH2), 5.46 (s, 1H, acac CH), 6.51 (d, 3 Hz, 3H, ArH), 6.74
(d, 3 Hz, 3H, ArH). 13C{1H} NMR (CDCl3): 26.32 (acac CH3),
30.16 (C(CH3)3), 34.95 (C(CH3)3), 55.69 (NCH2), 61.74 (OC-
H3), 105.20 (acac CH), 111.56, 112.88, 125.66, 137.40, 151.81,
155.52, 190.30 (acac CdO). IR (evapd film, cm-1): 2951 (m),
2909 (m), 2832 (w), 1739 (w), 1597 (m), 1520 (m), 1464 (s), 1428
(s), 1360 (m), 1320 (w), 1265 (m), 1239 (s), 1208 (s), 1142 (w),
1057 (m), 1025 (w), 979 (w), 919 (w), 837 (m), 795 (w), 767 (w),
665 (w). FAB MS: 737 (Mþ). UV-vis (CH2Cl2) λmax 295 nm (ε=
25300 M-1 cm-1), 345 (13500), 384 (15400). Anal. Calcd for
C41H55NO8Ti: C, 66.75; H, 7.51; N, 1.90. Found: C, 66.87; H,
7.34; N, 1.87.
X-ray Crystallography of tBu,MeOLH3, tBu,tBuLTiCl 0.5C6H6,
3
and tBu,MeOLTiCl C6H6. Very large crystals of tBu,MeOLH3
3
deposited from a solution of the oily crude material from the
preparation in methanol; a fragment was cut from a larger
crystal for diffraction. Irregular orange crystals of tBu,tBuLTiCl
0.5C6H6 were grown by layering a solution in benzene with
3
acetonitrile (1:5 v/v). Crystals of tBu,MeOLTiCl C6H6 were
H,An NLTi(acac). In the drybox, 0.356 g of H,An NLH3
(0.350 mmol) was added to a 25 mL round-bottom flask
containing a Teflon-coated stir bar. The ligand was dissolved
in CH2Cl2, and 100 μL of titanium isopropoxide (0.339 mmol,
0.97 equiv) was added via digital pipet. The flask was capped
with a rubber septum, taken out of the drybox, and stirred
for 5 min. Through the septum 2,4-pentanedione (35.0 μL,
0.340 mmol, 0.97 equiv) was then added via syringe. After
stirring under N2 for 30 min, the flask was opened to the air
and the solvent was removed on a rotary evaporator. The dark
brown solid was crystallized by vapor diffusion of ether into
a CH2Cl2 solution of the crude material to obtain 0.238 g of
2
2
3
grown by slow evaporation from benzene. All crystals were
placed in inert oil before transferring to the cold N2 stream
of a Bruker Apex II CCD diffractometer. Data were reduced,
correcting for absorption, using the program SADABS. The
structures were solved using direct methods, except for
tBu,MeOLTiCl C6H6, which was solved using a Patterson synthe-
3
sis. All nonhydrogen atoms not apparent from the initial solu-
tions were found on difference Fourier maps, and all heavy
atoms were refined anisotropically.
(16) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877–910.

4692 Inorganic Chemistry, Vol. 49, No. 10, 2010
Lionetti et al.
Table 1. Crystal Data for tBu,MeOLH3, tBu,tBuLTiCl 0.5C6H6, and tBu,MeOLTiCl C6H6
3
3
tBu,MeOLH3
tBu,tBuLTiCl 0.5C6H6
tBu,MeOLTiCl C6H6
3
3
empirical formula
temperature (K)
λ
C36H51NO6
120(2)
0.71073 (Mo KR)
P1
28678
8436
0.0317
6385
10.4142(8)
12.9974(10)
14.0615(10)
73.7099(11)
69.0650(11)
88.2183(12)
1701.2(2)
C48H69ClNO3Ti
100(2)
0.71073 (Mo KR)
P21/n
49294
11529
0.0308
9394
15.6030(8)
16.1332(8)
18.5363(9)
90
95.5600(10)
90
4644.1(4)
4
C42H54ClNO6Ti
100(2)
0.71073 (Mo KR)
P212121
47204
9703
0.0256
9091
8.9631(4)
16.7443(6)
26.4720(10)
90
space group
total data collected
no. of indep reflns.
Rint
obsd. refls. [I > 2σ(I)]
˚
a (A)
˚
b (A)
˚
c (A)
R (deg)
β (deg)
γ (deg)
90
90
3972.9(3)
4
3
˚
V (A )
Z
2
cryst size (mm)
no. refined params.
R indices [I > 2σ(I )]a
R indices (all data)a
goodness of fit
0.29 ꢀ 0.26 ꢀ 0.13
0.4 ꢀ 0.3 ꢀ 0.3
0.61 ꢀ 0.35 ꢀ 0.34
592
735
676
R1 = 0.0392, wR2 = 0.0926
R1 = 0.0559, wR2 = 0.1021
1.049
R1 = 0.0442, wR2 = 0.1173
R1 = 0.0553, wR2 = 0.1262
1.038
R1 = 0.0272, wR2 = 0.0666
R1 = 0.0307, wR2 = 0.0683
1.022
P
P
P
P
a R1 = ||Fo| - |Fc||/ |Fo|; wR2 = [ w(Fo2 - Fc2)2/ w(Fo2)2]1/2
.
In the structure of tBu,tBuLTiCl 0.5C6H6, two of the tert-butyl
Table 2. Selected Bond Distances (A) and Angles (deg) in tBu,MeOLH3
˚
3
groups (those attached to C15 and C35) were disordered and
were refined in two different orientations (the major orienta-
tions refined to 60.4(3)% and 57.6(3)% occupied, respectively).
All hydrogens in the three structures, except those on the disor-
dered tert-butyl groups and those on the lattice benzene in
O1-C12
O2-C22
O3-C32
N-C10
N-C20
N-C30
1.3754(14)
1.3806(13)
1.3849(13)
1.4839(14)
1.4804(13)
1.4809(13)
tBu,tBuLTiCl 0.5C6H6, were found on difference Fourier maps
3
and were refined isotropically. All calculations used SHELXTL
(Bruker AXS),17 with scattering factors and anomalous disper-
sion terms taken from the literature.18 Further details about the
individual structures are given in Table 1.
C10-N-C20
C10-N-C30
C20-N-C30
N-C10-C11
N-C20-C21
N-C30-C31
111.26(8)
111.70(8)
108.66(8)
111.31(9)
111.56(9)
113.81(9)
Results and Discussion
Preparation of Electron-Rich Tripodal Aminetris-
(phenol) Ligands. The most common method for prepar-
ing symmetrical aminetris(phenol) ligands N(CH2C6H2-
3-R1-5-R2-2-OH)3 is Mannich condensation of the
appropriate 2,4-dialkylphenol with formaldehyde and
ammonia.19 In this vein, reaction of 2-tert-butyl-4-methoxy-
phenol with hexamethylenetetramine and aqueous for-
maldehyde does give the tripodal ligand N(CH2C6H2-
3-tBu-5-OMe-2-OH)3 (tBu,MeOLH3) (eq 1). Yields are low,
but the starting materials are inexpensive and the reaction
can be run at a large scale, so this represents a simple route
to prepare multigram quantities of the previously un-
known methoxy-substituted ligand.
the central nitrogen (Figure 2). This conformation, which is
preorganized for metal binding, has been observed with
other tripodal triphenols with tert-butyl groups in the ortho
position,20 while less bulky triphenols adopt a conformation
with one of the OH groups turned away from the amine.21 A
similar preorganized conformation has also been seen with
other amine-centered tripodal ligands.22
We then sought to make an even more electron-rich
tripodal triphenol, with a diarylamino group in place of
the para-methoxy group in tBu,MeOLH3. Installation of the
diarylamino group was envisaged to proceed via palladium-
catalyzed amination of the appropriate (possibly protected)
bromophenol, since this is known to be a highly efficient
method for the formation of triarylamines.23 Since literature
precedent suggested that p-diarylaminophenols would be
(19) (a) Hultzsch, K. Chem. Ber. 1949, 82, 16–25. (b) Dargaville, T. R.;
De Bruyn, P. J.; Lim, A. S. C.; Looney, M. G.; Potter, A. C.; Solomon, D. H.;
Zhang, X. J. Polym. Sci. A: Polym. Chem. 1997, 35, 1389–1398. (c) Cortes,
ꢀ
S. A.; Hernandez, M. A. M.; Nakai, H.; Castro-Rodriguez, I.; Meyer, K.; Fout,
A. R.; Miller, D. L.; Huffman, J. C.; Mindiola, D. J. Inorg. Chem. Commun. 2005,
8, 903–907.
(20) Kelly, B. V.; Weintrob, E. C.; Buccella, D.; Tanski, J. M.; Parkin, G.
Inorg. Chem. Commun. 2007, 10, 699–704.
(21) (a) Kubono, K.; Hirayama, N.; Kokusen, H.; Yokoi, K. Anal. Sci.
2001, 17, 913–914. (b) Chartres, J. D.; Dahir, A.; Tasker, P. A.; White, F. J. Inorg.
Chem. Commun. 2007, 10, 1154–1158.
The free triphenol is revealed by X-ray crystallography
(Tables 1-2) to have, at least in the solid state, an
approximately C3-symmetric conformation with the
three phenols pointed inward toward the lone pair of
(17) Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, A64, 112–122.
(18) International Tables for Crystallography; Kluwer Academic Publish-
ers: Dordrecht, The Netherlands, 1992; Vol C.
(22) Boon, J. M.; Lambert, T. N.; Smith, B. D.; Beatty, A. M.; Ugrinova,
V.; Brown, S. N. J. Org. Chem. 2002, 67, 2168–2174.
(23) Hartwig, J. F. Angew. Chem., Int. Ed. 1998, 37, 2046–2067.

Article
Inorganic Chemistry, Vol. 49, No. 10, 2010 4693
appropriate benzyl-protected brominated tribenzylamine.
t
This was then subjected to Bu3P/Pd(0)-catalyzed amina-
tion using 4,40-dimethoxydiphenylamine in the presence of
sodium tert-butoxide,28 which provided the tris(triaryl-
amine) in good yield. This material was somewhat light-
sensitive and was stored in the dark before use.
Deprotection by hydrogenolysis using the standard
catalyst, 10% palladium on carbon,26,29,30 was unsuccess-
ful, with reactions even at elevated hydrogen pressures
and high catalyst loadings proceeding at impractica-
bly slow rates. Attempted debenzylation using sodium
iodide/trimethylsilyl chloride31 was also unsuccessful.
However, Pearlman’s catalyst (wet Pd(OH)2/C) proved
more active than Pd/C, and its use allowed hydrogen-
olysis to take place in high yield and in a reasonable
amount of time. There was no sign of cleavage of the
benzylic C-N bonds in these reactions, which contrasts
sharply with reports that this catalyst (when used in
ethanol) is actually selective for N-debenzylation in the
presence of O-benzyl groups.32 The final deprotected
ligand, tBu,An NLH3, was moderately air-sensitive as
2
a solid or in solution; it was decolorized with sodium
dithionite after deprotection and then stored under
nitrogen.
Figure 2. Thermal ellipsoid plot (50% ellipsoids) of tBu,MeOLH3.
Scheme 1. Preparation of Diarylamino-Substituted Ligands tBu,An NLH3
2
Unlike the Mannich condensation, which relies on
blocking one of the positions ortho to the phenol to pre-
vent polymer formation, the reductive amination se-
quence is compatible with a ligand that has no substituent
ortho to the phenol. This ligand is prepared analogously
(Scheme 1, R=H), starting with commercially available
5-bromosalicylaldehyde and proceeding through the
previously prepared O-benzyl derivative.15 Debenzyla-
tion in the final step to yield the ortho-unsubstituted
and H,An NLH3 (An = anisyl, p-C6H4OCH3)
2
ligand H,An NLH3 proceeds smoothly with palladium on
2
carbon as catalyst. Apparently either a para-diarylamino
or an ortho-tert-butyl substituent by itself does not pre-
vent hydrogenolysis over 10% Pd/C, but the combination
of both renders the hydrogenolysis very sluggish with this
catalyst.
Preparation and Characterization of Tripodal Titanium
Complexes. The metalation of tripodal aminetriphenol
ligands with titanium alkoxides is a general route to five-
coordinate titanium alkoxides.10,11,29,30,33 The new tert-
butyl substituted ligands tBu,MeOLH3 and tBu,An NLH3
2
significantly air-sensitive,24 we hoped to carry out the amina-
tion at a late stage of the synthesis. However, the appropri-
ate phenolic starting material, 4-bromo-2-tert-butylphenol,25
failed to undergo Mannich condensation with ammonia/
formaldehyde, possibly because it is too electron-poor.
We thus turned to reductive amination to prepare a
suitable aminotriphenol (Scheme 1), a strategy described
by Licini.26 Benzylation of 3-tert-butyl-5-bromo-2-hydroxy-
benzaldehyde (prepared by bromination of commercially
available 3-tert-butylsalicylaldehyde14) proceeded smoothly
and was followed by reductive amination using sodium
triacetoxyborohydride in 1,2-dichloroethane27 to give the
both react with titanium(IV) tert-butoxide to give the
corresponding LTi(OtBu) complexes (Scheme 2). 1H NMR
spectroscopy at room temperature is consistent with
C3-symmetry, with the appearance of the methylene hydro-
gens as a pair of doublets indicating a substantial barrier
to racemization in the five-coordinate complex.10 As pre-
viously reported for tBu,tBuLTi complexes,11 the pre-
sence of the ortho tert-butyl substituents renders the
(28) Hartwig, J. F.; Kawatsura, M.; Hauck, S. I.; Shaughnessy, K. H.;
Alcazar-Roman, L. M. J. Org. Chem. 1999, 64, 5575–5580.
(29) Axe, P.; Bull, S. D.; Davidson, M. G.; Gilfillan, C. J.; Jones, M. D.;
Robinson, D. E. J. E.; Turner, L. E.; Mitchell, W. L. Org. Lett. 2007, 9, 223–
226.
€
(24) Linkletter, S. J. G.; Pearson, G. A.; Walter, R. I. J. Am. Chem. Soc.
1977, 99, 5269–5272.
(30) Bernardinelli, G.; Seidel, T. M.; Kundig, E. P.; Prins, L. J.; Kolarovic,
A.; Mba, M.; Pontini, M.; Licini, G. Dalton Trans. 2007, 1573–1576.
(31) Olah, G. A.; Narang, S. C.; Gupta, B. G. B.; Malhotra, R. J. Org.
Chem. 1979, 44, 1247–1251.
(32) (a) Marzi, M.; Misiti, D. Tetrahedron Lett. 1989, 30, 6075–6076. (b)
Bernotas, R. C.; Cube, R. V. Synth. Commun. 1990, 20, 1209–1212.
(33) (a) Kim, Y.; Verkade, J. G. Organometallics 2002, 21, 2395–2399. (b)
Mba, M.; Prins, L. J.; Licini, G. Org. Lett. 2007, 9, 21–24.
(25) Berthelot, J.; Guette, C.; Desbene, P. L.; Basselier, J. J.; Chaquin, P.;
Masure, D. Can. J. Chem. 1989, 67, 2061–2066.
ꢀ
ꢀ
(26) Prins, L. J.; Blazquez, M. M.; Kolarovic, A.; Licini, G. Tetrahedron
Lett. 2006, 47, 2735–2738.
(27) Abdel-Magid, A. F.; Carson, K. G.; Harris, B. D.; Maryanoff, C. A.;
Shah, R. D. J. Org. Chem. 1996, 61, 3849–3862.

4694 Inorganic Chemistry, Vol. 49, No. 10, 2010
Lionetti et al.
Scheme 2. Synthesis and Reactivity of tBu,RLTi Complexes (R = tBu,
CH3O, An2N)
diketonate complexes) show room temperature NMR
spectra with apparent C3v symmetry of the tripod and
equivalent ends of the acetylacetonate because of rapid
fluxional processes.12
Optical spectra of the complexes (e.g., of LTi(acac),
Figure 4) show intense (ε ∼ 104 M-1 cm-1) absorptions in
the near-UV and blue regions because of aryloxide-to-
titanium charge transfer transitions.35 Consistent with
this assignment, changing the apical group on titanium
from the more electron-donating tert-butoxide to the
less electron-donating chloride results in a red shift of
the absorption (for example, tBu,tBuLTi(OtBu) is yellow,
λ
max=329 nm, while tBu,tBuLTiCl is orange, λmax=388 nm).
The variation of λmax with the para substituent on the
aryloxide provides a gauge for the relative electron-
donating power of the aryloxides (Table 4), with the peak
charge-transfer absorption of tBu,tBuLTi(acac) at 0.16 eV
higher energy than that of tBu,MeOLTi(acac), which is in
turn blue-shifted 0.36 eV relative to tBu,An NLTi(acac).
2
The charge-transfer bands, especially those of tBu,tBuLTi-
(acac) and tBu,MeOLTi(acac), appear to have some shoulders
on the high-energy side, possibly indicating the presence
of overlapping bands of similar energy. This might result
from slight differences between the aryloxide trans to the
diketonate and those cis to it; in tBu,tBuLTi(acac) the
complexes resistant to hydrolysis, and both new tert-
butoxide complexes can be handled in the air for moder-
ate (tBu,An NLTi(OtBu)) or extended (tBu,MeOLTi(OtBu))
2
periods without degradation. The ortho-unsubstituted
ligand H,An NLH3 reacts with Ti(OtBu)4 as well, but the
2
product H,An NLTi(OtBu) is quite moisture-sensitive and
2
˚
trans aryloxide forms a 0.038 A shorter Ti-O bond than
do the cis aryloxides.12 The very intense peaks at about
300 nm and the less intense shoulders at about 360 nm in
the diarylamino-substituted complexes are ligand-based
transitions, as they correspond almost exactly to bands
observed in N(C6H4OCH3)3.13
was not isolated. Tripods with small ortho substituents
such as methyl34 or phenyl30 hydrolyze to give simple
μ-oxo dimers, but hydrolysis of H,An NLTi(OtBu) appears to
2
be more complicated, with the compound reacting rapidly
on exposure to air to give a mixture of products.
Electrochemistry of Tripodal Titanium Complexes.
Another way to gauge the relative electron-donating abil-
ity of the different tripodal complexes is through electro-
chemistry. The five-coordinate complexes LTi(OtBu) and
LTiCl of the tert-butyl- and methoxy-substituted ligands
show electrochemically irreversible voltammograms
(Supporting Information, Figures S1-S2). One possible
explanation for the observed irreversibility is that the
coordinatively unsaturated species might bind electrolyte
or aggregate in their oxidized forms. Consistent with this
hypothesis, the six-coordinate acetylacetonate complexes
all show well-defined, largely reversible oxidations in
their cyclic voltammograms (Figure 5). Even the dialkyl-
substituted complex tBu,tBuLTi(acac) shows a single oxi-
dation wave at E°=þ0.57 V (vs ferrocene/ferrocenium in
dichloromethane), and the methoxy-substituted complex
tBu,MeOLTi(acac) shows two reversible oxidations, at E°=
þ0.33 V and þ0.68 V vs Fcþ/Fc. For comparison, the
neutral six-coordinate d0 tacn-tris(aryloxide) (Figure 1a)
scandium complexes studied by Wieghardt et al. show
first oxidations at E°=þ0.52 V and þ0.27 V (measured
in CH3CN) for the tert-butyl and methoxy-substituted
aryloxides, respectively.6 The similarity of the redox
potentials of the scandium and titanium complexes sup-
ports the assignment of the oxidation as aryloxide-based
(rather than, say, acac-centered) in the latter. The ∼250
mV shift to lower potential on going from tert-butyl to
methoxy has been observed consistently in a wide variety
Apical group substitution in the tert-butyl compounds
is straightforward, with the compounds reacting with
aqueous HCl to provide the five-coordinate complexes
LTiCl. The preparation of tBu,tBuLTiCl by reaction of
tBu,tBuLH3 with TiCl4 has been described by Nielson and
co-workers,34 and 1-Ad,tBuLTiCl has similarly been pre-
pared from TiCl4(THF)2 and the corresponding free triphe-
nol.19c The coordination geometry of the compounds is
confirmed by X-ray crystallography of both tBu,tBuLTiCl
and tBu,MeOLTiCl (Figure 3, Tables 1 and 3), which
crystallize as benzene solvates of isolated five-coordinate
monomers. (tBu,tBuLTiCl has also been crystallized as an
etherate,34 and data from this structure are included in
Table 3 for comparison.) The three structures are extre-
mely similar, with indistinguishable Ti-Cl and Ti-O
distances, for example. The only hint that the para-meth-
oxy group increases the electron-donating power of the
aryloxides as expected is the slight elongation of the
˚ ˚
LTiCl (0.036 A and 0.056 A
tBu,MeO
Ti-N distance in
longer than in the benzene and ether solvates, respec-
tively, of tBu,tBuLTiCl).
As reported previously for tBu,tBuLTi(OtBu),12 all of the
tert-butoxide complexes tBu,RLTi(OtBu) react quantita-
tively with 2,4-pentanedione to form the six-coordinate
tBu,RLTi(acac) complexes. The sterically unhindered com-
plex H,An NLTi(OiPr), generated in situ from H,An NLH3
and Ti(OiPr)4, reacts analogously to give air-stable
2
2
H,An NLTi(acac). While tBu,tBuLTi(acac) is six-coordinate
2
and has C1 symmetry in the crystal, it (and all of the other
(35) (a) Tinoco, A. D.; Valentine, A. M. J. Am. Chem. Soc. 2005, 127,
11218–11219. (b) Tinoco, A. D.; Incarvito, C. D.; Valentine, A. M. J. Am. Chem.
Soc. 2007, 129, 3444–3454.
(34) Nielson, A. J.; Shen, C.; Waters, J. M. Polyhedron 2006, 25, 2039–
2054.

Article
Inorganic Chemistry, Vol. 49, No. 10, 2010 4695
Figure 3. Thermal ellipsoid plots (50% ellipsoids) for (a) tBu,tBuLTiCl 0.5 C6H6 and (b) tBu,MeOLTiCl C6H6. Hydrogen atoms and solvents of
crystallization are omitted for clarity.
3
3
Table 3. Selected Bond Distances (A) and Angles (deg) in tBu,tBuLTiCl 0.5C6H6
˚
3
and tBu,MeOLTiCl C6H6
a
3
tBu,tBuLTiCl
0.5C6H6
tBu,tBuLTiCl
Et2O34
tBu,MeOLTiCl
C6H6
3
3
3
Ti-Cl
Ti-O1
Ti-O2
Ti-O3
Ti-N
2.3130(4)
2.3074(7)
1.809(2)
1.817(2)
1.811(2)
2.216(2)
2.3118(4)
1.8132(10)
1.8126(9)
1.8031(9)
2.2717(12)
1.8052(11)
1.8116(11)
1.8160(11)
2.2361(11)
N-Ti-Cl
179.87(4)
83.04(4)
83.45(4)
82.15(4)
96.99(3)
96.65(4)
97.73(4)
116.05(5)
117.65(5)
121.74(5)
144.16(9)
142.92(9)
144.91(10)
179.51(5)
82.76(7)
82.66(7)
83.30(7)
97.10(5)
177.98(3)
83.26(4)
82.61(4)
81.94(4)
98.17(3)
95.46(3)
98.53(3)
116.71(4)
119.76(4)
118.62(4)
143.36(9)
144.17(9)
146.47(9)
N-Ti-O1
N-Ti-O2
N-Ti-O3
Cl-Ti-O1
Cl-Ti-O2
Cl-Ti-O3
O1-Ti-O2
O1-Ti-O3
O2-Ti-O3
Ti-O1-C12
Ti-O2-C22
Ti-O3-C32
97.01(5)
97.18(5)
119.29(7)
115.48(7)
120.72(7)
143.54(15)
143.66(14)
143.97(14)
a Corresponding data from tBu,tBuLTiCl Et2O (ref 34) are included
for comparison.
Figure 4. UV-visible spectra of LTi(acac) in CH2Cl2.
Table 4. Optical and Electrochemical Properties of X,YLTi(acac)
3
of metal complexes.1 The splitting between the first and
second redox waves in tBu,MeOLTi(acac) is larger than that
reported for its tacn-tris(aryloxide) scandium analogue
(350 vs 200 mV), perhaps because of greater electrostatic
effects in the lower dielectric solvent CH2Cl2 compared to
CH3CN.
E°, V vs Cp2Feþ/
Cp2Fe (CH2Cl2,
0.1 M [Bu4N]PF6)
X (ortho Y ( para
substituent) substituent)
λmax
(LMCT)
tBu
tBu
tBu
tBu
OMe
NAn2
366 nm (3.39 eV) þ0.57
384 nm (3.23 eV) þ0.33, þ0.68
432 nm (2.87 eV) -0.09, þ0.06, þ0.12,
þ0.58, þ0.78, þ0.90
The diarylamino-substituted complex tBu,An NLTi-
2
H
NAn2
428 nm (2.90 eV) -0.02, þ0.11, þ0.16,
þ0.65, þ0.84, þ0.89
(acac) shows two noteworthy differences from its tert-
butyl and methoxy-substituted analogues. First, its initial
oxidation taþkes place at a substantially lower potential
(∼ 0 V vs Fc /Fc). Second, while oxidation of the second
arm of tBu,MeOLTi(acac) is 350 mV more difficult than
oxidation of the first arm, all three aryloxides of
These are assigned to removal of a second electron from
each of the dianisylaminophenoxide arms of the ligand.
Such double oxidation of methoxy-substituted triaryl-
amines has been documented to form diamagnetic quino-
noid dications,36 and the cyclic voltammetry of symmetri-
cal (CH3OC6H4)3N in CH2Cl2 shows reversible first and
second oxidations at þ109 and þ860 mV (vs Fc/Fcþ),
tBu,An NLTi(acac) lose an electron within about 200 mV
2
of each other. While the three redox waves are very close
in the cyclic voltammogram, they can be resolved in the
square-wave voltammogram (Figure 6) at -0.09, þ0.06,
and þ0.12 V. There is a second set of three closely spaced
waves in the CV and SWV between þ0.6 and þ0.9 V.
€
(36) Hagopian, L.; Kohler, G.; Walter, R. I. J. Phys. Chem. 1967, 71,
2290–2296.

4696 Inorganic Chemistry, Vol. 49, No. 10, 2010
Lionetti et al.
The electrochemistry of the ortho-unsubstituted
H,An NLTi(acac) is strikingly similar to that of the more
2
hindered tBu,An NLTi(acac) (Figures 5-6), except for a
2
shift of the redox potentials by about þ50 mV, consistent
with the tert-butyl group being slightly more electron-
donating than hydrogen. Notably, the reversibility of the
electrochemical redox reactions remains excellent even in
the absence of steric protection of the phenoxide center.
This contrasts with the results of oxidation of the tacn-
tris(dialkylphenoxide) complexes, where the dimethyl-
phenoxides show irreversible electrochemistry, while the
di-tert-butylphenoxides can be oxidized reversibly.6
Qualitatively, the ∼400 mV shift to lower potential
in tBu,An NLTi(acac) compared to tBu,MeOLTi(acac) is
2
expected, given the greater donating ability of amino
compared to alkoxy substituents.37 For example, a dime-
thylamino-substituted salen nickel complex has been
observed to undergo ligand-centered oxidation at 400
mV lower potential than its methoxy-substituted ana-
logue.38 A second explanation for the lower redox poten-
tials is that the locus of oxidation shifts, at least in part, to
the diarylamino groups in R,An NLTi(acac). Consistent
2
Figure 5. Cyclic voltammetry (CH2Cl2, 0.1 M Bu4NPF6, 60 mV/s) of 1
mM solutions of (from top to bottom) tBu,tBuLTi(acac), tBu,MeOLTi(acac),
with this view is the similarity in redox potentials between
tBu,An NLTi(acac), and H,An NLTi(acac).
(CH3OC6H4)3N (E° = þ0.11 V)13 and R,An NLTi(acac)
2
2
2
(E° = -0.09 and -0.02 V for R = tBu and R = H,
respectively). Analysis of a series of p-substituted radical
cations (XC6H4)3N•þ by EPR indicates that 14N and
ortho-1H hyperfine constants are fairly constant for
X=alkyl, H, or electron-withdrawing substituents, but
are modestly (X = OCH3) or substantially (X = NH2)
reduced for electron-donating substituents,39 suggesting
partial delocalization of the radical onto these substitu-
tents. Combined with the electrochemical results, this
suggests that oxidation occurs principally from the dia-
rylamino groups in R,An NLTi(acac), but with significant
2
interactions with the aryloxide group. This view is also
consistent with the small, but detectable, electrochemical
splittings in the first and second oxidation waves. The
fact that the splittings between the oxidation potential of
the three arms are small (ΔE°= 50-150 mV) suggests
that the loci of oxidation are relatively far from each
other and so there is little sensitivity of the redox
potential to partial oxidation of the ligand. This con-
trasts with the much larger splitting between waves seen
in the more aryloxide-localized tBu,MeOLTi(acac) (ΔE°=
350 mV). However, the fact that there is any discernible
splitting at all indicates that the oxidation of one arm has
some effect on the electron density at the other arms. In
contrast, a recent study of bis(ferrocenyldiketonate)
complexes of titanium(IV), where DFT calculations
showed the redox-active orbitals to be essentially local-
ized on iron, showed no discernible splitting in the
ferrocene-based redox waves.40 Both the narrow spread
of oxidation potentials (to minimize overpotentials),
and the existence of some coupling (as an indication of
interactions with the metal center), validate the ligand
Figure 6. Square wave voltammetry (CH2Cl2, 0.1 M (Bu4N)PF6) of
tBu,An NLTi(acac) (top) and H,An NLTi(acac) (bottom).
2
2
respectively.13 We have prepared the unsymmetrical dimeth-
oxytriarylamine (CH3OC6H4)2NC6H4COCH3 (see Support-
ing Information for details of preparation and characteriza-
tion), and find that it also shows two reversible oxidations in
its CV, shifted as expected to higher potential because of the
replacement of one methoxy group by an acetyl group
(Supporting Information, Figure S3, E°=þ0.38, þ1.03 V).
The 650-750 mV difference between the first and second
oxidations in these triarylamines compares favorably to the
(37) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165–195.
(38) Rotthaus, O.; Jarjayes, O.; Del Valle, C. P.; Philouze, C.; Thomas, F.
Chem. Commun. 2007, 4462–4464.
(39) Pearson, G. A.; Rocek, M.; Walter, R. I. J. Phys. Chem. 1978, 82,
1185–1192.
(40) Dulatas, L. T.; Brown, S. N.; Ojomo, E.; Noll, B. C.; Cavo, M. J.;
Holt, P. B.; Wopperer, M. M. Inorg. Chem. 2009, 48, 10789–10799.
670-780 mV shifts seen in tBu,An NLTi(acac) and strongly
2
supports the proposed assignments.

Article
Inorganic Chemistry, Vol. 49, No. 10, 2010 4697
design of the (diarylamino)aryloxides as promising can-
didates for supporting redox reactions.
The translation of this potential into actual reactivity is
currently being explored.
Conclusions
Acknowledgment. We thank Dr. Allen G. Oliver for
his assistance with the crystal structure of tBu,MeOLH3,
Dr. Alicia M. Beatty for her assistance with the crystal
Electron-rich tripodal aminetriaryloxide ligands can be
prepared by Mannich condensation (tBu,MeOLH3) or by a
synthetic sequence involving reductive amination of a pro-
tected bromosalicylaldehyde and Hartwig-Buchwald ami-
structure of tBu,tBuLTiCl 0.5 C6H6, and Ms. Nancy E.
3
Roback for her assistance with the crystal structure of
(MeOC6H4)2NC6H4COCH3. This work was supported
by the donors of the Petroleum Research Fund, adminis-
tered by the American Chemical Society, and by the
National Science Foundation (CHE-0518243). D.L.
gratefully acknowledges the Notre Dame College of
Science for a summer undergraduate research fellowship,
and A.J.M. thanks the Notre Dame College of Science
and the Glynn Family Honors Program for summer
fellowship support. Partial support for the X-ray diffrac-
tion facility was provided by the NSF (CHE-0443233).
nation as key steps (tBu,An NLH3 and H,An NLH3). Both five-
coordinate (tBu,XLTi(OtBu) and tBu,XLTiCl) and six-coordi-
nate (LTi(acac)) complexes of titanium(IV) can be prepared.
Electrochemistry of the diarylamino-substituted tripods
shows that each can be reversibly oxidized by up to six elec-
trons (two per arm). As possible redox reservoirs for reac-
tions where substrate oxidation is mediated by a redox-inert
metal, the diarylamino-substituted tripods display three
critical features: (1) They support low coordination numbers,
potentially allowing substrate access to the metal center. (2)
The redox potentials for oxidation of successive arms are
closely spaced (ΔE°=50-150 mV), reducing the possibility
of large overpotentials. (3) The redox potentials for oxidation
of successive arms do show some variation, indicating that
there is modest interaction between the redox centers and the
metal center, opening the possibility of communicating the
redox state of the ligand to a substrate bound to the metal.
2
2
Supporting Information Available: Crystallographic informa-
tion on tBu,MeOLH3, tBu,tBuLTiCl 0.5C6H6, tBu,MeOLTiCl C6H6,
3
3
and (MeOC6H4)2NC6H4COCH3 in CIF format; cyclic voltamme-
try of LTi(OtBu) and LTiCl and synthetic details and spectro-
scopic, structural, and electrochemical characterization of (MeO-
C6H4)2NC6H4COCH3 (PDF format). This material is available
free of charge via the Internet at http://pubs.acs.org.