Ligand exchange reactions and hydroamination with tris(oxazolinyl)borato yttrium compounds

Article abstract of DOI:10.1021/ic900689k

Ligand substitution reactions and catalytic hydroamination/cyclization of aminoalkenes have been studied with a new oxazolinylborato yttrium compound, tris(4,4-dimethyl-2-oxazolinyl)phenylborato bis(trimethylsilylmethyl)yttrium ([Y(κ³-To^M<

Full text of DOI:10.1021/ic900689k

8020 Inorg. Chem. 2009, 48, 8020–8029
DOI: 10.1021/ic900689k
Ligand Exchange Reactions and Hydroamination with Tris(oxazolinyl)borato
Yttrium Compounds
Andrew V. Pawlikowski, Arkady Ellern, and Aaron D. Sadow*
Department of Chemistry, Iowa State University, Ames, Iowa 50011
Received April 9, 2009
Ligand substitution reactions and catalytic hydroamination/cyclization of aminoalkenes have been studied with a new
oxazolinylborato yttrium compound, tris(4,4-dimethyl-2-oxazolinyl)phenylborato bis(trimethylsilylmethyl)yttrium ([Y(κ3-
ToM)(CH2SiMe3)2(THF)], 1). THF exchange in 1 is rapid at room temperature, and activation parameters obtained by
simulation of 1H NMR spectra acquired from 190 to 280 K are consistent with a dissociative mechanism (ΔS‡ = 30 ( 1
e.u., ΔG‡ = 11.9 kcal mol-1 at 243 K). The related phosphine oxide adduct [Y(κ3-ToM)(CH2SiMe3)2(OPPh3)] (2) also
undergoes exchange via OPPh3 dissociation with a much higher barrier (ΔG‡ = 15.0 kcal mol-1 at 320 K). Compound
t
1 reacts with the amines BuNH2, para-MeC6H4NH2, and 2,6-iPr2C6H3NH2 to provide six-coordinate [Y(κ3-ToM)-
t
(NHR)2(THF)] (3: R = Bu; 4: R = para-MeC6H4) and five-coordinate [Y(κ3-ToM)(NH-2,6-iPr2C6H3)2] (6). These
oxazolinylborato yttrium compounds are precatalysts for the cyclization of aminoalkenes; the kinetics of catalytic
conversion indicate zero-order substrate dependence and first-order catalyst dependence. Kinetic investigations of
ligand exchange processes and hydroamination reactions indicate that the tris(oxazolinyl)borato-yttrium interaction is
robust even in the presence of excess phosphine oxide and primary and secondary amines.
Introduction
C3v-symmetric ligands, such as the anionic borate-contain-
ing tris(pyrazolyl)borates (Tp),2 tris(phosphino)borates,3
tris(carbene)borates,4 and tris(sulfido)borates,5 have allowed
for the stabilization of reactive moieties bonded to metal
centers. Recently, the related zwitterionic C2-symmetric bis
(oxazolinyl)borate (borabox)6 and neutral tris-oxazoline li-
gands (tris-ox)7 have provided optically active scorpionate-
like ligands. We have recently contributed to this area with
tris(oxazolinyl)borates8 that combine zwitterionic and tri-
dentate features, providing ligands reasonably described as
sterically and electronically modified tris(pyrazolyl)borates.
The zwitterionic nature, tridentate binding motif, and
steric protection afforded by tris(oxazolinyl)borates suggest
that they would be good ancillary ligands for rare earth
centers. Thus, we began to synthesize new compounds
containing the tris(4,4-dimethyl-2-oxazolinyl)borate ligand
Exchange of coordinated product for substrate is an
important part of catalytic chemistry, and thus rare earth
organometallic compounds are valuable in catalysis because
of their facile ligand substitutions.1 However, undesired
substitution of ancillary ligands can diminish yields, lower
selectivity, and provide pathways for decomposition. Addi-
tionally, exchange reactions can be difficult to detect if the
thermodynamically favored product contains kinetically
labile ancillary ligands. A key challenge in organo-rare
earth element chemistry is the discovery of new robust
spectator ligands that control the selectivity of reactions,
such as ligand exchange, at the remaining bonding sites.
In this context, tridentate, fac-coordinating monoanionic
ligands are expected to be useful for trivalent rare earth
chemistry because their charge and the chelate effect in-
hibit ancillary ligand dissociation, whereas two reactive
valences remain available for bonding to additional anionic
ligands.
::
(4) (a) Frankel, R.; Kernbach, U.; Bakola-Christianopoulou, M.; Plaia,
::
U.; Suter, M.; Ponikwar, W.; Noth, H.; Moinet, C.; Fehlhammer, W. P. J.
Organomet. Chem. 2001, 617-618, 530–545. (b) Nieto, I.; Cervantes-Lee, F.;
Smith, J. M. Chem. Commun. 2005, 3811–3813. (c) Cowley, R. E.; Bontchev,
R. P.; Duesler, E. N.; Smith, J. M. Inorg. Chem. 2006, 45, 9771–9779.
(5) Ge, P.; Haggerty, B. S.; Rheingold, A. L.; Riordan, C. G. J. Am.
*Corresponding author. E-mail: sadow@iastate.edu.
(1) (a) Watson, P. L.; Parshall, G. W. Acc. Chem. Res. 1985, 18, 51–56.
Chem. Soc. 1994, 116, 8406–8407.
(6) Mazet, C.; Kohler, V.; Pfaltz, A. Angew. Chem., Int. Ed. 2005, 44,
4888–4891.
::
(b) Anwander, R. Top. Organomet. Chem. 1999, 2, 1–61. (c) Sastri, V. S.;
::
Bunzli, J.-C.; Rao, V. R.; Rayudu, G. V. S.; Perumareddi, J. R. Modern
(7) (a) Bellemin-Laponnaz, S.; Gade, L. H. Angew. Chem., Int. Ed. 2002,
41, 3473–3475. (b) Gade, L. H.; Bellemin-Laponnaz, S. Chem.;Eur. J. 2008,
14, 4142–4152.
(8) (a) Dunne, J. F.; Su, J. C.; Ellern, A.; Sadow, A. D. Organometallics
2008, 27, 2399–2401. (b) Baird, B.; Pawlikowski, A. V.; Su, J.; Wiench, J. W.;
Pruski, M.; Sadow, A. D. Inorg. Chem. 2008, 47, 10208–10210.
Aspects of Rare Earths and Their Complexes; Elsevier: Amsterdam, 2003.
(2) Trofimenko, S. Scorpionates: The Coordination Chemistry of Poly-
pyrazolylborate Ligands; Imperial College Press: London, 1999.
(3) (a) Peters, J. C.; Feldman, J. D.; Tilley, T. D. J. Am. Chem. Soc. 1999,
121, 9871–9872. (b) Betley, T. A.; Peters, J. C. J. Am. Chem. Soc. 2003, 125,
10782–10783.
r
pubs.acs.org/IC
Published on Web 07/08/2009
2009 American Chemical Society

Article
Inorganic Chemistry, Vol. 48, No. 16, 2009 8021
(abbreviated ToM) and investigate their chemistry in sub-
stitution reactions and catalysis.
new catalytic chemistry for scorpionate-type complexes.
In particular, hydroamination/cyclization of aminoole-
fins has recently received significant attention as an
efficient method for preparing nitrogen-containing het-
erocycles.13,16,17 Although most rare earth hydroamina-
tion/cyclization catalysts contain dianionic ancillary
ligand sets,18 fewer catalysts contain one monoanionic
ancillary ligand and two reactive valences.19,20 Although
oxazoline ligands are common ancillary ligands in tran-
sition-metal chemistry, few oxazoline-containing com-
plexes have been studied in rare earth hydroamination
catalyses.20
Here we describe the synthesis and structural character-
ization of the first tris(oxazolinyl)borate yttrium alkyl and
amide compounds, ligand exchange reactions at the reac-
tive yttrium centers, and their catalytic properties in
hydroamination/cyclization of aminoalkenes. We have
investigated the kinetics of ligand exchange and hydro-
amination reactions, and these measurements show that
the yttrium-ToM interaction is robust under catalytic
conditions, whereas the yttrium alkyl, amido, and neutral
ligands react rapidly.
Tris(pyrazolyl)borate rare earth compounds are well-
known,9 but the mechanisms of neutral ligand substitutions
in mono-tris(pyrazolyl)borate complexes still require investi-
gation. For example, evidence favors an associative mechan-
ism for substitution of THF in [Y(κ3-Tp*)Cl2(THF)] (Tp* =
tris(3,5-dimethyl-1-pyrazolyl)borohydride) even though the
yttrium(III) center is six-coordinate.10 In contrast, the bulky
[Y(κ3-TptBuMe)(CH2SiMe3)2] is five-coordinate.11 Establishing
mechanism(s) for ligand exchange pathways in ToMLnR2L
compounds (eq 1, Ln = group 3 or lanthanide element) is
clearly important for understanding substrate coordination
to active sites in catalytic processes.
Additionally, substitutions of alkyl groups via protonoly-
tic ligand elimination reactions have not been well inves-
tigated in the TpLnR2(L)-type systems,10-12 although
these reactions are well-known with other ancillary ligands
in rare earth element chemistry and are steps in the
proposed mechanism for rare earth element catalyzed
hydroamination/cyclization of aminoalkenes.13 Thus, we
were interested to investigate alkyl-group protonolysis
reactions with [Y(κ3-ToM)(CH2SiMe3)2(THF)] (1) to es-
tablish the ToM-Y interaction as inert to cleavage in the
presence of excess amine (eq 2).
Experimental Section
General Procedures. All reactions were performed under a
dry argon atmosphere using standard Schlenk techniques or
under a nitrogen atmosphere in a glovebox, unless otherwise
indicated. Water and oxygen were removed from benzene,
toluene, pentane, diethyl ether, and tetrahydrofuran solvents
using an IT PureSolv system. Benzene-d6, tetrahydrofuran-d8,
and toluene-d8 were heated to reflux over Na/K alloy and
vacuum-transferred. Anhydrous YCl3 was purchased from
Strem and used as received. 2,6-diisopropylaniline and
tert-butylamine were distilled from CaH2 prior to use and
ToMLnR2Lþ2 H-NR20 fToMLnðNR20 Þ2Lþ2 H-R ð2Þ
˚
stored over 4 A mol. sieves in the glovebox, and p-tolui-
dine was purified by sublimation. The aminoalkenes 4-
penten-1-amine,13b N-methyl-2,2-diphenyl-4-penten-1-amine,21 2,2-
diphenyl-4-penten-1-amine, and 2,2-dimethyl-4-penten-1-amine,20
Additional motivation was that few bis(amido) com-
pounds of the type TpLn(NR2)2 have been described,14
although rare earth amido compounds have wide-ranging
chemistry. The reported catalytic chemistry of mono-
tris(pyrazolyl)borate and tris-ox rare earth compounds
has primarily focused on olefin polymerization,10,15 and
our tris(oxazolinyl)borate compounds might provide
(16) (a) Molander, G. A.; Romero, J. A. C. Chem. Rev. 2002, 102, 2161–
2185. (b) Hong, S.; Marks, T. J. Acc. Chem. Res. 2004, 37, 673–686. (c)
Hultzsch, K. C. Adv. Synth. Catal. 2005, 347, 367–391. (d) Muller, T. E.;
Hultzsch, K. C.; Yus, M.; Foubelo, F.; Tada, M. Chem. Rev. 2008, 108,
::
3795–3892.
(17) (a) Molander, G. A.; Dowdy, E. D. J. Org. Chem. 1999, 64, 6515–
6517. (b) Ryu, J.-S.; Marks, T. J.; McDonald, F. E. Org. Lett. 2001, 3, 3091–
3094. (c) Hong, S.; Marks, T. J. J. Am. Chem. Soc. 2002, 124, 7886–7887. (d)
Ryu, J.-S.; Marks, T. J.; McDonald, F. E. J. Org. Chem. 2004, 69, 1038–
1052.
(18) (a) Tian, S.; Arredondo, V. M.; Stern, C. L.; Marks, T. J. Organo-
metallics 1999, 18, 2568–2570. (b) Kim, Y. K.; Livinghouse, T. Angew.
Chem., Int. Ed. 2002, 41, 3645–3647. (c) O’Shaughnessy, P. N.; Knight, P. D.;
Morton, C.; Gillespie, K. M.; Scott, P. Chem. Commun. 2003, 1770–1771. (d)
Gribkov, D. V.; Hampel, F.; Hultzsch, K. C. Eur. J. Inorg. Chem. 2004,
4091–4101. (e) Kim, J. Y.; Livinghouse, T. Org. Lett. 2005, 7, 1737–1739. (f)
Collin, J.; Daran, J.-C.; Jacquet, O.; Schulz, E.; Trifonov, A. Chem.;Eur. J.
2005, 11, 3455–3462. (g) Meyer, N.; Zulys, A.; Roesky, P. W. Organome-
tallics 2006, 25, 4179–4182. (h) Gribkov, D. V.; Hultzsch, K. C.; Hampel, F.
J. Am. Chem. Soc. 2006, 128, 3748–3759. (i) Yu, X.; Marks, T. J. Organo-
metallics 2007, 26, 365–376.
(9) Marques, N.; Sella, A.; Takats, J. Chem. Rev. 2002, 102, 2137–2159.
(10) (a) Long, D. P.; Bianconi, P. A. J. Am. Chem. Soc. 1996, 118, 12453–
12454. (b) Long, D. P.; Chandrasekaran, A.; Day, R. O.; Bianconi, P. A.;
Rheingold, A. L. Inorg. Chem. 2000, 39, 4476–4487.
(11) (a) Cheng, J.; Saliu, K.; Kiel, G. Y.; Ferguson, M. J.; McDonald, R.;
Takats, J. Angew. Chem., Int. Ed. 2008, 47, 4910–4913. (b) Litlabφ, R.;
::
Zimmermann, M.; Saliu, K.; Takats, J.; Turnroos, K. W.; Anwander, R.
Angew. Chem., Int. Ed. 2008, 47, 9560–9564. (c) Zimmermann, M.; Takats,
::
J.; Kiel, G.; Tornroos, K. W.; Anwander, R. Chem. Commun. 2008, 612–614.
(12) Blackwell, J.; Lehr, C.; Sun, Y.; Piers, W. E.; Pearce-Batchilder, S.
D.; Zaworotko, M. J.; Young, V. G., Jr. Can. J. Chem. 1997, 75, 702–711.
ꢀ
(13) (a) Gagne, M. R.; Marks, T. J. J. Am. Chem. Soc. 1989, 111, 4108–
ꢀ
4109. (b) Gagne, M. R.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1992,
114, 275–294.
(14) (a) Roitershtein, D.; Domingos, A.; Pereira, L. C. J.; Ascenso, J. R.;
Marques, N. Inorg. Chem. 2003, 42, 7666–7673. (b) Foley, T. J.; Harrison, B.
S.; Knefely, A. S.; Abboud, K. A.; Reynolds, J. R.; Schanze, K. S.; Boncella,
J. M. Inorg. Chem. 2003, 42, 5023–5032.
(19) (a) Zulys, A.; Panda, T. K.; Gamer, M. T.; Roesky, P. W. Chem.
Commun. 2004, 2584–2585. (b) Lauterwasser, F.; Hayes, P. G.; Brase, S.;
Piers, W. E.; Schafer, L. L. Organometallics 2004, 23, 2234–2237.
(c) Bambirra, S.; Tsurugi, H.; Leusen, D.; Hessen, B. Dalton Trans. 2006,
::
(15) (a) Ward, B. D.; Bellemin-Laponnaz, S.; Gade, L. H. Angew. Chem.,
Int. Ed. 2005, 44, 1668–1671. (b) Lukesova, L.; Ward, B. D.; Bellemin-
Laponnaz, S.; Wadepohl, H.; Gade, L. H. Organometallics 2007, 26, 4652–
4657. (c) Lukesova, L.; Ward, B. D.; Bellemin-Laponnaz, S.; Wadepohl, H.;
Gade, L. H. Dalton Trans. 2007, 920–922.
1157–1161.
(20) Hong, S.; Tian, S.; Metz, M. V.; Marks, T. J. J. Am. Chem. Soc. 2003,
125, 14768–14783.
(21) Stubbert, B. D.; Marks, T. J. J. Am. Chem. Soc. 2007, 129, 4253–
4271.

8022 Inorganic Chemistry, Vol. 48, No. 16, 2009
Pawlikowski et al.
Li[ToM],8a and [Y(CH2SiMe3)3(THF)2]22 were prepared by pub-
lished procedures. Anhydrous H[ToM] was prepared from
CNCMe2CH2O), 78.62 (br, CNCMe2CH2O), 79.89 (br,
CNCMe2CH2O), 125.63 (para-C6H5), 126.82 (meta-C6H5),
Li[ToM] and triethylammonium chloride in CH2Cl2. 1H, 11B, 13
C
1
4
129.24 (d, JPC =108 Hz, OPPh3), 129.36 (d, JPC =12.8 Hz,
OPPh3), 133.58 (d, 3JPC=10.8 Hz, OPPh3), 133.93 (d, 5JPC=2.3
Hz, OPPh3), 136.90 (ortho-C6H5). 11B NMR (128 MHz, ben-
zene-d6): δ -18.3. 29Si{1H} NMR (79.5 MHz, benzene-d6):
δ -3.9 (d, 2JYSi=4.3 Hz). 31P{1H} NMR (162 MHz, benzene-
{1H}, and 29Si{1H} spectra were collected on Bruker DRX-400,
DRX-500, and AV-600 spectrometers. 15N chemical shifts were
determined by 1H-15N HMBC experiments on a Bruker
Avance II 700 spectrometer with a Bruker Z-gradient inverse
TXI 1H/13C/15N 5 mm cryoprobe; 15N chemical shifts were
originally referenced to liquid NH3 and recalculated to the
CH3NO2 chemical shift scale by adding -381.9 ppm. 11B
2
d6): δ 36.69 (d, JYP = 9.8 Hz). IR (KBr, cm-1): 3055 (w),
2959 (s), 2883 (m), 1575 (s), 1462 (w), 1438 (s), 1367 (w), 1270
(m), 1248 (m), 1195 (m), 1152 (s), 1120 (s), 1091 (w), 997 (m), 973
(w), 862 (s), 809 (w). Anal. Calcd for C47H66BN3O4PSi2Y: C,
61.10; H, 7.20; N, 4.55. Found: C, 61.19; H, 7.35; N, 4.62. mp
205 ꢀC (dec.)
NMR spectra were referenced to an external sample of BF3
3
Et2O. Elemental analyses were performed using a Perkin-Elmer
2400 Series II CHN/S by the Iowa State Chemical Instrumenta-
tion Facility. X-ray diffraction data was collected on a Bruker-
AXS SMART 1000 CCD diffractometer using Bruker-AXS
SHELXTL software.
[Y(K3-ToM)(NHtBu)2(THF)] (3). Tert-butylamine (14.9 μL,
0.142 mmol) was added to a solution of 1 (0.051 g, 0.071 mmol)
in benzene at ambient temperature. After 5 min, the volatiles
were removed from the reaction mixture, and the resulting
residue was washed with pentane and dried under vacuum to
yield 0.035 g of [Y(κ3-ToM)(NHtBu)2(THF)] (3) as a white powder
(0.051 mmol, 73%). 1H NMR (400 MHz, benzene-d6): δ 1.28 (s,
[Y(K3-ToM)(CH2SiMe3)2(THF)] (1). A THF solution of
H[ToM] (0.761 g, 1.99 mmol) was added in a dropwise
fashion to a pentane solution of [Y(CH2SiMe3)3(THF)2]
(0.988 g, 2.0 mmol) at room temperature. Pale yellow
crystals of 1 formed after standing at -30 ꢀC for 2 days.
The crystals were isolated by filtration, washed with
pentane, and dried under vacuum to give 1.09 g of 1 (1.52 mmol,
77%). X-ray quality crystals of 1 were grown from a concen-
trated toluene solution of [Y(κ3-ToM)(CH2SiMe3)2(THF)]
layered with hexane at -30 ꢀC. 1H NMR (400 MHz, benzene-
d6): δ -0.34 (d, 2JYH=2.8 Hz, 4 H, YCH2SiMe3), 0.45 (s, 18 H,
18 H, CNCMe2CH2O), 1.38 (m, 4 H, THF β-CH2), 1.56 (s, 18 H,
NHCMe3), 1.82 (d, 2JYH=2.0 Hz, 2 H, NHCMe3), 3.39, (s, 6 H,
3
CNCMe2CH2O), 3.71 (m, 4 H, THF R-CH2), 7.36 (t, JHH
=
7.6 Hz, 1 H, para-C6H5), 7.55 (t, 3JHH=7.6 Hz, 2 H, meta-C6H5),
8.36 (d, JHH=7.2 Hz, 2 H, ortho-C6H5). 13C{1H} NMR (100
3
MHz, benzene-d6):
δ
25.74 (THF β-CH2), 28.25
CH2SiMe3), 1.20 (s, 18 H, CNCMe2CH2O), 1.35 (m, 4 H, THF
2
(CNCMe2CH2O), 37.37 (NHCMe3), 53.92 (d, JYC = 4.5 Hz,
NHCMe3), 67.78 (CNCMe2CH2O), 69.70 (THF R-CH2), 79.18
β-CH2), 3.33 (s, 6 H, CNCMe2CH2O), 3.89 (m, 4 H, THF
3
R-CH2), 7.35 (t, JHH = 7.2 Hz, 1 H, para-C6H5), 7.53 (t,
3
3JHH=7.2 Hz, 2 H, meta-C6H5), 8.23 (d, JHH=7.2 Hz, 2 H,
ortho-C6H5). 13C{1H} NMR (100 MHz, benzene-d6): δ 5.55
(CNCMe2CH2O), 125.61 (para-C6H5), 126.82 (meta-C6H5),
136.78 (ortho-C6H5). 11B NMR (128 MHz, benzene-d6):
δ -18.3. 15N{1H} NMR (71 MHz, benzene-d6): δ -141.2
(CNCMe2CH2O), -224.6 (NHCMe3). IR (KBr, cm-1): 3044
(w), 2961 (s), 2887 (m), 1576 (s), 1462 (w), 1431 (w), 1366 (m),
1347 (w), 1270 (m), 1247 (w), 1195 (m), 1173 (m), 1146 (m), 1028
(w), 972 (s), 920 (w), 887 (w), 836 (w), 788 (w). Anal. Calcd for
C33H57BN5O4Y: C, 57.65; H, 8.36; N, 10.19. Found: C, 57.23; H,
8.27; N, 9.64. mp 195 ꢀC (dec.)
(CH2SiMe3), 25.29 (THF β-CH2), 28.28 (CNCMe2CH2O),
37.08 (d, 1JYC=35 Hz, YCH2SiMe3), 67.92 (CNCMe2CH2O),
71.09 (THF R-CH2), 79.56 (CNCMe2CH2O), 125.90 (para-
C6H5), 126.93 (meta-C6H5), 136.64 (ortho-C6H5). 11B
NMR (128 MHz, benzene-d6): δ -18.2. 15N{1H} NMR (71
MHz, benzene-d6): δ -141.7. 29Si{1H} NMR (79.5 MHz,
benzene-d6): δ -2.7 (2JYSi = 4.6 Hz). IR (KBr, cm-1): 2961
(m), 2892 (m), 1568 (s), 1463 (w), 1368 (w), 1272 (m), 1248
(m), 1234 (w), 1194 (m), 1149 (m), 1024 (w), 994 (m), 922 (w),
860 (s), 810 (w). Anal. Calcd for C33H59BN3O4Si2Y: C, 55.
22; H, 8.29; N, 5.85. Found: C, 54.98; H, 8.12; N, 5.83. mp
185 ꢀC (dec.)
[Y(K3-ToM)(NH-para-MeC6H4)2(THF)] (4). para-Toluidine
(0.018 g, 0.167 mmol) and 1 (0.060 g, 0.084 mmol) were dis-
solved in toluene (2 mL) at ambient temperature. After 5 min,
the volatiles were removed from the reaction mixture, and
the resulting white powder was washed with pentane and
dried under vacuum to afford 0.055 g of analytically pure
[Y(κ3-ToM)(NH-para-MeC6H4)2(THF)] (0.073 mmol, 87%).
1H NMR (400 MHz, benzene-d6): δ 1.12 (br s, 18 H,
[Y(K3-ToM)(CH2SiMe3)2(OPPh3)] (2). Compound 1 (0.116 g,
0.162 mmol) was dissolved in benzene (3 mL) and triphenylpho-
sphine oxide (0.045 g, 0.162 mmol) was added. After 10 min, the
volatiles were removed and the remaining solid was washed with
pentane and dried under vacuum to yield 0.108 g of analytically
pure [Y(κ3-ToM)(CH2SiMe3)2(OPPh3)] (2) as an off-white solid
(0.117 mmol, 73%). 1H NMR (400 MHz, benzene-d6): δ -0.21
(d, 2JYH=2.4 Hz, 4 H, YCH2SiMe3), 0.48 (s, 18 H, CH2SiMe3),
CNCMe2CH2O), 1.34 (m, 4 H, THF β-CH2), 2.31 (s, 6 H, NH-
para-MeC6H4), 3.38 (s, 6 H, CNCMe2CH2O), 3.77 (m, 4 H,
THF R-CH2), 4.66 (s, 2 H, NH-para-MeC6H4), 6.86 (d, 3JHH
=
3
8.2 Hz, 4 H, NH-para-MeC63H4), 7.10 (d, JHH=8.2 Hz, 4 H,
NH-para-MeC6H4), 7.37 (t, JHH =7.6 Hz, 1 H, para-C6H5),
7.57 (t, 3JHH=7.6 Hz, 2 H, meta-C6H5), 8.30 (d, 3JHH=7.6 Hz,
2 H, ortho-C6H5). 13C{1H} NMR (100 MHz, benzene-d6):
21.17 (NH-para-MeC6H4), 25.85 (THF β-CH2), 28.18
0.88 (br s,
6 H, CNCMe2CH2O), 1.14 (br s, 6 H,
CNCMe2CH2O), 1.56 (br s, 6 H, CNCMe2CH2O), 3.17 (br s,
2H, CNCMe2CH2O), 3.42 (br m, 4 H, CNCMe2CH2O), 7.04
(m, 9 H, OPPh3), 7.37 (t, 3JHH=7.2 Hz, 1 H, para-C6H5), 7.58 (t,
3JHH=7.6 Hz, 2 H, meta-C6H5), 7.65 (m, 6 H, OPPh3), 8.49 (d,
3JHH =7.2 Hz, 2 H, ortho-C6H5). 13C{1H} NMR (125 MHz,
(CNCMe2CH2O), 67.85 (CNCMe2CH2O), 70.32 (THF R-CH2),
79.77 (CNCMe2CH2O), 117.44 (NH-para-MeC6H4), 122.40
(NH-para-MeC6H4), 125.98 (para-C6H5), 127.03 (meta-C6H5),
130.54 (NH-para-MeC6H4), 136.62 (para-C6H5), 155.78 (NH-
para-MeC6H4). 11B NMR (128 MHz, benzene-d6):
δ -18.1. 15N{1H} NMR (71 MHz, benzene-d6): δ -142.9
(CNCMe2CH2O), -231.2 (NH-para-MeC6H4). IR (KBr, cm-1):
2963 (m), 2929 (m), 2893 (m), 1608 (m), 1570 (s), 1505 (s), 1462 (w),
1431 (w), 1367 (w), 1352 (w), 1297 (m), 1270 (m), 1251 (w),
1192 (m), 1175 (m), 1150 (m), 1103 (w), 1014 (w), 988 (w),
benzene-d6): δ 5.74 (CH2SiMe3), 27.05 (br, CNCMe2CH2O),
1
29.30 (br, CNCMe2CH2O), 35.07 (d, JYC = 35 Hz,
YCH2SiMe3), 67.96 (br, CNCMe2CH2O), 68.45 (br,
(22) (a) Lappert, M. F.; Pearce, R. J. Chem. Soc., Chem. Commun. 1973,
126. (b) Hultzsch, K. C.; Voth, P.; Beckerle, K.; Spaniol, T. P.; Okuda, J.
Organometallics 2000, 19, 228–243.

Article
Inorganic Chemistry, Vol. 48, No. 16, 2009 8023
973 (w), 922 (w), 854 (w), 816 (m). Anal. Calcd for
C39H53BN5O4Y: C, 61.99; H, 7.07; N, 9.27. Found: C, 62.02; H,
7.02; N, 9.22. mp 132 ꢀC (dec.)
5 mm Wilmad NMR tubes equipped with a J. Young Teflon
valve. Samples were prepared by dissolution of 1 in benzene-d6
containing a known concentration of C6Me6 or C8H16 standard
and aminoolefin immediately before being placed in the NMR
probe, which was preset to the required temperature. The probe
temperature was calibrated before each measurement using an
external thermocouple in an NMR tube lowered into the probe;
once the probe temperature stabilized, it was monitored with the
instrument’s thermocouple. Single scan spectra were acquired
automatically at preset time intervals. The peaks were integrated
relative to C6Me6 or C8H16 as an internal standard. Rate
constants were obtained by nonweighted linear least-squares
fit of the concentrations to the integrated rate law.
t
[Y(K3-ToM)(NHtBu)2(NH2 Bu)] (5). A solution of 1 (0.096 g,
0.134 mmol) in benzene was treated with an excess of tert-
butylamine (0.3 mL, 2.85 mmol) at room temperature. The
volatiles were removed under reduced pressure, and the result-
ing white powder was redissolved in benzene and treated with
another 0.3 mL of tert-butylamine. The volatiles were again
removed under vacuum to yield 0.085 g of [Y(κ3-ToM)-
t
(NHtBu)2(NH2 Bu)] (5) as a white powder (0.123 mmol,
92%). Crystals of 5 suitable for X-ray diffraction were obtained
from a concentrated toluene solution layered with pentane at
-30 ꢀC. 1H NMR (400 MHz, benzene-d6): δ 1.26 (s, 18 H,
CNCMe2CH2O), 1.44 (br s, 27 H, NH2CMe3 and NHCMe3),
1.97 (br s, 4 H, NH2CMe3 and NHCMe3), 3.34, (s, 6 H,
CNCMe2CH2O), 7.35 (t, 3JHH=6.8 Hz, 1 H, para-C6H5), 7.54
(t, 3JHH=7.2 Hz, 2 H, meta-C6H5), 8.29 (d, 3JHH=7.2 Hz, 2 H,
ortho-C6H5). 13C{1H} NMR (125 MHz, benzene-d6): δ 28.74
Results and Discussion
Synthesis and X-ray Structural Characterization of [Y-
(K3-ToM)(CH2SiMe3)2(THF)] (1). Alkane elimination oc-
curs upon treatment of [Y(CH2SiMe3)3(THF)2] with one
equivalent of H[ToM] dissolved in minimal amounts of
pentane and THF, respectively, at ambient temperature
to afford the six-coordinate complex [Y(κ3-ToM)(CH2Si-
Me3)2(THF)] (1) in high isolated yield (77%, eq 3).
(CNCMe2CH2O), 35.88 (br, NH2CMe3 and NHCMe3), 52.65
(br, NH2CMe3 and NHCMe3), 67.64 (CNCMe2CH2O), 79.20
(CNCMe2CH2O), 125.68 (para-C6H5), 126.86 (meta-C6H5),
136.69 (ortho-C6H5). 11B NMR (128 MHz, benzene-d6):
δ -18.2. 15N{1H} NMR (71 MHz, benzene-d6): δ -142.5
(CNCMe2CH2O). IR (KBr, cm-1): 3343 (vw), 3042 (w), 2964
(s), 2875 (m), 1575 (s), 1463 (m), 1431 (w), 1367 (m), 1349 (w),
1270 (m), 1194 (m), 1174 (m), 1147 (m), 1072 (w), 993 (m), 973
(m), 920 (w), 889 (w), 837 (w), 800 (w). Anal. Calcd for
C33H60BN6O3Y: C, 57.56; H, 8.78; N, 12.20. Found: C, 57.91;
H, 8.50; N, 12.14. mp 192 ꢀC (dec.)
A single-crystal X-ray structure confirms the identity of
1 revealing κ3-N,N,N-coordination of the ToM, one THF,
and two trimethylsilylmethyl ligands bonded to yttrium
(Figure 1). Although 1 is fluxional in solution as indicated
by 1H NMR spectroscopy (see below), only one specific
and distorted geometry is observed in the solid state such
that two independent molecules in the unit cell have
similar metrical parameters. The tridentate ToM ligand
is tied-back which results in small —N-Y1-N angles of
80.2(2), 79.9(2), and 75.7(2)ꢀ; the third angle is signifi-
cantly smaller than the other two. The Y1-N bond
[Y(K3-ToM)(NH-2,6-iPr2C6H3)2] (6). 2,6-Diisopropylaniline
(0.032 mL, 0.169 mmol) was added to a benzene solution of
1 (0.060 g, 0.084 mmol) at ambient temperature, and the reaction
mixture was stirred for 5 min. After evaporation of the volatiles,
the resulting residue was washed with cold pentane (5 mL) and
dried under vacuum, affording [Y(κ3-ToM)(2,6-iPr2C6H3NH)2]
as a pale yellow powder (0.055 g, 0.067 mmol, 81% yield).
1H NMR (400 MHz, benzene-d6):
δ 1.02 (s, 18 H,
3
CNCMe2CH2O), 1.37 (d, JHH = 6.8 Hz, 24 H, CHMe2), 3.30
(m, 3JHH=6.8 Hz, 4 H, CHMe2), 3.35 (s, 6 H,CNCMe2CH2O), 5.00
˚
distances of 2.443(5), 2.477(5), and 2.571(7) A are also
(s, 2 H, NH), 6.88 (t, 3JHH=7.6 Hz, 2 H, NH-2,6-iPr2C6H3), 7.17 (d,
inequivalent, and the anomalously long bond is trans to
one CH2SiMe3 rather than THF. The two shorter Y1-N
bonds (2.477(5) and 2.443(5) A) trans to the other
3
3JHH = 7.6 Hz, 4 H, NH-2,6-iPr2C6H3), 7.34 (t, JHH
=
7.6 Hz, 1 H, para-C6H5), 7.52 (t, 3JHH=7.6 Hz, 2 H, meta-C6H5),
˚
8.19 (d, JHH = 7.6 Hz, 2 H, ortho-C6H5). 13C{1H} NMR
3
CH2SiMe3 and to THF are statistically equivalent. Be-
cause the two Y1-N bond lengths trans to CH2SiMe3 are
(100 MHz, benzene-d6): δ 24.44 (CHMe2), 28.01 (CNCMe2CH2O),
˚
30.25 (CHMe2), 67.67 (CNCMe2CH2O), 79.93 (CNCMe2CH2O),
116.54 (NH-2,6-iPr2C6H3), 123.65 (NH-2,6-iPr2C6H3), 126.35
(para-C6H5), 127.33 (meta-C6H5), 133.94 (NH-2,6-iPr2C6H3),
different by ca. 0.1 A, it is surprising that the Y1-C
distances 2.405(7) and 2.409(7) A are identical. The angles
˚
between THF and the two CH2SiMe3 ligands are also
very different: —O4-Y1-C26 is 83.5(3) and —O4-Y1-
C22 is 99.7(2)ꢀ. Because both molecules in the unit cell
have similar geometries, these distorted structural
features are best explained by interligand steric interac-
tions between the bulky ToM and CH2SiMe3 ligands and
have little to do with electronic differences between
CH2SiMe3 and THF ligands. The two CH2SiMe3 groups
are oriented to avoid each other, forcing one CH2SiMe3
to be directed toward the THF ligand resulting in the
large —O4-Y1-C22 angle. The other CH2SiMe3 points
toward the CMe2 group of an oxazoline ring, and that
angle is also large (—C26-Y1-N2=112.0(3)ꢀ).
2
136.17 (ortho-C6H5), 152.28 (d, JYC = 3.9 Hz, NH-2,6-
iPr2C6H3). 11B NMR (128 MHz, benzene-d6):
δ -18.2.
15N{1H} NMR (71 MHz, benzene-d6): δ -139.8 (CNCMe2CH2O),
-236.6 (NH-2,6-iPr2C6H3). IR (KBr, cm-1): 3403 (vw), 3049 (w),
2962 (m), 2868 (m), 1621 (w), 1567 (s), 1461 (m), 1426 (s), 1384 (w),
1366 (w), 1329 (w), 1259 (s), 1191 (m), 1153 (m), 1109 (w), 1039 (w),
959 (m), 926 (w), 884 (w), 842 (m), 790 (w), 742 (m). Anal. Calcd. for
C45H65BN5O3Y: C, 65.61; H, 7.95; N, 8.50. Found: C, 65.62; H,
8.22; N, 7.93. mp 155 ꢀC (dec.).
Hydroamination Catalysis. In a typical experiment, an NMR
tube equipped with a J. Young Teflon valve was charged with 1
(10 μmol), the appropriate substrate (0.2 mL of a 1 M solution in
benzene-d6; 200 μmol), and 0.3 mL of benzene-d6. The 1H NMR
spectrum of the reaction mixture was recorded at regular intervals
to accurately determine the time until complete conversion.
Kinetic Measurements. Reactions were monitored by 1H
NMR spectroscopy with a Bruker DRX400 spectrometer using
˚
Because yttrium(III) is relatively large (0.9 A six-co-
ordinate radius)23 and distortions in the complex’s
(23) Shannon, R. D. Acta Crystallogr., Sect. A 1976, 32, 751–767.

8024 Inorganic Chemistry, Vol. 48, No. 16, 2009
Pawlikowski et al.
Figure 1. ORTEP diagram of one molecule of [Y(κ3-ToM)(CH2SiMe3)2(THF)] (1) contained in the asymmetric unit at 50% probability. Hydrogen atoms
and the second structurally related molecule are omitted for clarity.
geometry are due to interligand steric interactions, the
effective steric properties of ToM bonded to yttrium were
evaluated. The solid angle for a ligand, which is deter-
mined by treating the metal center as a point source
of light and evaluating the surface area of the shadow
cast by the ligand on a sphere surrounding the complex,
provides a good method for evaluating the steric pro-
perties of ligands (particularly multidentate ligands).24
In 1, the solid angle of ToM (5.82 steradians) determined
using the program Solid-G25 corresponds to 46.3% of
the surface area of the sphere. The solid angle of a
CH2SiMe3 group is 1.74 steradians (13.9%) and THF is
1.6 steradians (13.0%). Because tridentate N,N,N-coor-
dination is the only accessible bonding mode (see below),
redistribution to give [Y(κ3-ToM)2(CH2SiMe3)] and
[Y(CH2SiMe3)3(THF)] is disfavored by the size of ToM
even on the large yttrium center. Importantly, the solid
angles and the X-ray structure also reveal that a
hemisphere around the yttrium center (corresponding
to three fac-sites of an octahedron) is available for
reactivity.
and 3.33 ppm, integrating to 18 and 6 H, respectively) are
detected in the 1H NMR spectrum of 1, which is consistent
with a C3v symmetric ToM ligand (benzene-d6 at room
temperature). Both of these singlets correlate to a single
oxazoline 15N resonance at -141.7 ppm, as shown by a
1H-15N HMBC experiment; this indicates that the oxazo-
lines are bonded to yttrium (cf. “free” 2H-4,4-dimethyl-2-
oxazoline 15N NMR: -127.5 ppm). Under these con-
ditions, the two CH2SiMe3 groups are also equivalent and
1
2
bonded to yttrium (2JYH = 2.8 Hz, JYC =35 Hz, JYSi
=
4.6 Hz). THF is also coordinated (3.89 and 1.35 ppm, 4 H
each), and only a single set of THF resonances is observed in
the presence of excess THF in benzene-d6 indicating that
coordinated and free THF exchange rapidly at room tem-
perature. Because the mechanism responsible for this ob-
served symmetrization could involve oxazoline and/or THF
dissociation, we investigated the fluxional process in detail to
establish the labile or inert nature of the tris(oxazolinyl)borate
ligand in 1.
The low-temperature 1H NMR spectrum of 1 (toluene-
d8, 220 K) contained three singlet resonances in a 1:1:1
ratio for oxazoline methyl groups as well as a pair of
coupled doublets and a singlet for the methylene groups.
Additionally, resonances corresponding to coordinated
and free THF were observed (in the presence of excess
THF). The low-temperature Cs-symmetric structure of 1
is depicted in Figure 2.8a
For comparison, the X-ray structure of [Y(κ3-Tp*)-
Cl2(3,5-Me2N2C3H)] reveals a similar tied-back geometry
for the Tp* ligand (—N-Y-N angles of 79.4(2), 79.3(2),
and 78.6(2)ꢀ), Y-N bond distances of 2.368(6), 2.418(5),
˚
and 2.420(5) A, and a slightly smaller solid angle for Tp*
of 5.58 steradians (44.4% of the surface area of a sphere
1H NMR spectra of 1 were recorded between 200 and
280 K, and [Y(κ3-ToM)(CH2SiMe3)2(THF)] is the only
species detected over this temperature range. This indicates
that temperature-dependent changes to the equilibria will
not complicate our analysis of the exchange process. The
resonances for free and coordinated THF decoalesce at
243 K, and resonances for Cs-symmetric ToM decoalesce at
the same temperature, strongly suggesting that the flux-
ional process that exchanges the oxazoline groups is
correlated with THF exchange. Additionally, the exchange
process occurs in the absence of excess THF (to 1H NMR
detection limits). Therefore, we postulate that the mechan-
ism for symmetrization of 1 involves a dynamic equilibri-
surrounding the complex is occupied).10b
Exchange Process in Compound 1. There is no crystal-
lographic symmetry imposed upon 1 in the solid state (Pna21
space group), and the ligands’ conformations give C1-sym-
metric structures. However, the solid state and apparent
solution symmetry are incongruent because of fluxional
process(es), as only one set of oxazoline resonances (1.20
(24) (a) White, D.; Taverner, B. C.; Leach, P. G. L.; Coville, N. J. J.
Comput. Chem. 1993, 14, 1042–1049. (b) White, D.; Coville, N. J. Adv.
Organomet. Chem. 1994, 36, 95–158.
(25) (a) Guzei, I. A.; Wendt, M. Dalton Trans. 2006, 3991–3999. (b)
Guzei, I. A.; Wendt, M. Program Solid-G; University of Wisconsin:
Madison, WI, 2004.

Article
Inorganic Chemistry, Vol. 48, No. 16, 2009 8025
served [THF]-independent exchange rates and the
observation of 1 as the only ToMY-containing species
present. Yttrium inversion and oxazoline-arm dissocia-
tion are also unlikely since ΔS‡ is large, and this mechan-
ism suggests that kinv is fast with respect to THF
dissociation. The entropic and enthalpic contributions
to the activation barrier are counterbalanced and give a
small overall ΔG‡ (50.0 kJ mol-1 or 11.9 kcal mol-1) at
the coalescence temperature of 243 K.
For comparison, the compound [Y(κ3-Tp*)Cl2(THF)] is
fluxional to at least 200 K in THF-d8, but in 3,5-dimethyl-
tetrahydrofuran a static Cs-symmetric structure for [Y(κ3-
Tp*)Cl2(3,5-Me2OC4H6)] is observed (i.e., slow exchange
occurs with a bulkier ligand) suggesting that the ligand
exchange mechanism in this pyrazolylborate system is
associative.10b Additional evidence for associative substi-
tution in the Tp-system is that the parent [Y(κ3-Tp)-
X2(THF)2] (X = Cl, Br) compounds are seven-coordi-
nate.10b The different spatial distribution of steric bulk in
ToM versus Tp* may be responsible for this mechanistic
change, although the difference in chloride versus alkyl
ligand may also affect the mechanism. [Y(κ3-Tp*)(CH2Si-
Me3)2(THF)] is fluxional at room temperature, but no
mechanism is proposed for exchange.10,11 While 3,5-di-
methyl pyrazole in Tp* is planar and provides a sterically
opencoordination sitefor a second THF tobind, oxazoline
groups are nonplanar and their 4,4-dimethyl groups are
directed toward the THF ligand’s coordination site (See
Figure 2 above), and this may facilitate THF dissociation.
Reactions of Compound 1: Substitutions and Bis(Amido)
Syntheses. Although 1 is unchanged after weeks at room
temperature in the solid state, its benzene-d6 solutions
Figure 2. Newman projection of 1 (L=THF) or 2 (L=OPPh3) looking
along the Y-B axis. The groups attached to N represent CH2 or CMe2 of
an oxazoline ring.
Figure 3. Proposed mechanism for symmetrization of [Y(κ3-ToM)-
(CH2SiMe3)2(THF)] (1). The process involves a THF dissociation/co-
ordination at equilibrium, with kdiss, kassoc, followed by rapid inversion
of the five-coordinate yttrium(III) species. The open coordination sites on
the five-coordinate species are illustrated to clarify the overall site
exchange and do not imply a particular geometry (square pyramidal or
trigonal bipyramidal) or mechanism (turnstile or pseudorotation) for site
exchange.
1
produce Me4Si and THF as the H NMR resonances
corresponding to 1 disappear over one day at room
temperature. New resonances corresponding to yttrium-
ToM containing products are broad and the species
formed have not been identified; however the ToM ligand
itself has not decomposed since exposure of this solution
to air gives H[ToM]. The extrusion of SiMe4 from 1 in
solution is inhibited by excess THF, and after rapid
displacement of the labile THF by THF-d8, spectra of 1
are unchanged over two days in THF-d8 at room tem-
perature. These observations suggest that the pathway for
decomposition of 1 involves THF dissociation as a first
step. Because ToM is ancillary in the symmetrization and
decomposition processes, substitution reactions of the
THF and alkyl ligands were investigated.
um that favors 1 and proceeds via THF dissociation,
inversion of a five-coordinate yttrium species, followed
by THF coordination (Figure 3).
The dynamic behavior of both the methyl and methy-
lene regions was simulated using gNMR,26 and this
simulation provides rate constants for site exchange of
three oxazoline groups. The rate constants, measured
at concentrations of [1] = 35.7 and 71.4 mmol (e.g.,
k2ex2c5hKange =7 s-1), are independent of [1] and unaffected
by added equivalents of THF. An Eyring plot, using the
rate constants obtained from dynamic NMR simulations,
provides values of ΔH‡=80.1(1.5 kJ mol-1 (19.1 ( 0.4
kcal mol-1) and ΔS‡=124 ( 6 J mol-1 K-1 (30 ( 1 eu).
These activation parameters are clearly consistent with a
dissociative exchange process because the large ΔH‡ is
consistent with bond breaking and the large positive ΔS‡
indicates increased disorder in the transition state. These
parameters indicate that the rate determining step in the
overall exchange process is THF dissociation rather than
coordination (where the opposite sign for ΔS‡ is ex-
pected); this dissociation is also consistent with the ob-
Addition of OPPh3 to a benzene-d6 solution of 1 pro-
duces THF and [Y(κ3-ToM)(CH2SiMe3)2(OPPh3)] (2); the
formation of 2 was easily verified by a 31P{1H} NMR
spectrum that contained a doublet resonance at 36.69 ppm
coupled to yttrium (d, 2JYP=9.8 Hz). The phosphine oxide
ligand in 2 is labile, but its exchange is slower than THF
exchange in 1. Thus at room temperature in toluene-d8,
1
broad oxazoline resonances in the H NMR spectrum of
2 were consistent with a Cs-symmetric structure slowly ex-
changing on the NMR time scale (see Figure 2, L=OPPh3).
Due to this exchange process, crosspeaks were not observed
in a 1H-15N HMBC spectrum, thus nitrogen chemical shifts
could not be determined for 2.27 At 270 K, sharp peaks in
(27) Variable-temperature NMR spectra could not be acquired with a
(26) Budzelaar, P. H. M. gNMR, v. 5.0; IvorySoft.
Bruker Z-gradient inverse TXI 1H/13C/15N 5 mm cryoprobe.

8026 Inorganic Chemistry, Vol. 48, No. 16, 2009
Pawlikowski et al.
the 1H NMR spectrum of the static Cs-symmetric structure
were observed. The coalescence temperature for oxazoline
exchange is 320 K, as compared to 243 K for 1, and
this TC is unaffected by added OPPh3. An Eyring plot
based on line shape analysis provides the following acti-
vation parameters: ΔH‡ = 71 ( 1.1 kJ mol-1 (17 (
0.3 kcal mol-1) and ΔS‡=26 ( 1 J mol-1 K-1 (6.2 (
0.2 e.u.). The positive ΔS‡ and zero-order dependence
on [OPPh3] are again consistent with a dissociative mechan-
ism. However, we were surprised to find similar ΔH‡ but a
significantly smaller ΔS‡ value for 2 versus 1.
Reactions of 1 and two equivalents of the small pri-
mary amines tBuNH2 and para-MeC6H4NH2 afford
six-coordinate bis(amido)yttrium compounds [Y(κ3-ToM)-
(NHtBu)2(THF)] (3) and [Y(κ3-ToM)(NH-para-Me-
C6H4)2(THF)] (4) along with two equivalents of Me4Si
1
within 5 min at room temperature, as determined by H
NMR spectroscopy (eq 4). As in bis(alkyl) 1, bis(amido)
THF adducts 3 and 4 are fluxional, so the ToM ligand
appears C3v symmetric in their 1H NMR spectra.
One possible explanation for the low activation entropy
in 2 is that oxazoline-arm dissociation, which is expected
to have a small dependence on entropy, is occurring
rather than OPPh3 dissociation. However, several other
experiments rule out this interpretation. In the presence
of 3 equiv. of OPPh3, 2 is the only ToM-containing species
observed, and there is no evidence for a bis(phosphine
oxide) species that might be formed upon oxazoline
dissociation. Phosphine oxide exchange is occurring,
and support for this comes from the reaction of OPPh2Cy
and 2, which gives a mixture containing 2, free OPPh3,
[Y(κ3-ToM)(CH2SiMe3)2(OPPh2Cy)], and OPPh2Cy
(assigned by 31P NMR spectroscopy) after 10 min at
room temperature in benzene-d6. Variable-temperature
31P NMR spectra of a mixture of 2 and OPPh3, acquired
from 270 to 360 K in toluene-d8, show resonances for 2
and free OPPh3 that broaden as the temperature in-
creases. Broad resonances in the high-temperature spec-
tra prove that free and coordinated phosphine oxide
exchange rapidly on the 31P NMR time scale. If oxazo-
line-arm dissociation rather than OPPh3 dissociation was
involved in the symmetrization process (observed in the
1H NMR spectra), then the resonance due to free OPPh3
would not be broadened. These 31P NMR spectra were
simulated with gNMR26 to obtain rate constants for
exchange of free and coordinated phosphine oxide. From
these data, activation parameters for exchange of coordi-
nated phosphine oxide in 2 with free phosphine oxide
(ΔH‡=80 ( 4 kJ mol-1; ΔS‡=23 ( 1 kJ mol-1K-1) are
comparable to the parameters determined from the
1H NMR spectra for the symmetrization process in 2,
and this suggests that both processes have a common
mechanism. Taken together, these observations strongly
favor OPPh3 dissociation as the rate determining step for
the symmetrization of 2 observed at elevated temperatures.
We propose that the low activation entropies measured
for symmetrization and exchange processes in 2 result
from an early transition state for OPPh3 dissociation
where bond cleavage is less compensated by entropic
factors (in comparison to THF dissociation from 1). As
expected from qualitative analysis of variable-tempera-
ture NMR spectra, the ΔG‡ value at the coalescence
temperature (62.7 kJ mol-1; 15.0 kcal mol-1 at 320 K)
for OPPh3 dissociation is significantly higher than
Addition of 3 equiv. of tBuNH2 to 1 produces a mixture
of THF adduct 3 and amine adduct [Y(κ3-ToM)-
(NHtBu)2(tBuNH2)] (5). These two species undergo rapid
exchange on the 1H NMR time scale, and only one set of
ToM resonances is observed (see below). Pure samples of
compound 5 are isolated by addition of excess tBuNH2 to
a benzene solution of 1 (or the mixture of 3 and 5)
followed by evaporation of the volatile materials. Bis
(amido)amine 5 crystallizes from a concentrated toluene
solution layered with pentane at -30 ꢀC. A low-resolu-
tion X-ray structure verifies its identity as a six-coordi-
nate yttrium center bonded to three oxazoline nitrogens
and three NtBu-containing ligands (see the Supporting
1
Information). The H NMR spectrum of 5 (benzene-d6,
room temperature) contains one set of sharp ToM reso-
nances (the oxazolines are equivalent), one broad tert-
butyl resonance (27 H), and a broad NH resonance (4 H).
As in compounds 1-4, fluxional exchange processes
result in the symmetrization of the ToM ligand which
appears C3v symmetric. However, in 5, the symmetriza-
tion process(es) could involve amine ligand dissociation/
coordination and/or proton transfer between amine and
amido ligands (see Figure 4).
Therefore, low-temperature 1H NMR spectra of 5 are
very interesting. When a solution of 5 in toluene-d8 is
cooled to 250 K, the oxazoline resonances remain sharp.
However, the broad tBu peak reaches its coalescence tem-
perature at 260 K. At 240 K, two distinct tert-butyl reso-
nances appear in a 2:1 ratio that correspond to the two
tert-butylamides and one coordinated tert-butylamine,
respectively. The amide peak is relatively sharp, whereas
the amine resonance is broad, and the decoalescence of
these two signals corresponds to slowing of intramolecu-
lar proton transfer from the amine to an amido ligand.
Desymmetrization of the oxazoline groups begins
to occur below 250 K, with a TC of 220 K. At
190 K, the spectrum is broad but its straightforward
interpretation is consistent with a Cs-symmetric structure
for [Y(κ3-ToM)(NHtBu)2(tBuNH2)], where amine disso-
ciation/coordination and proton exchange are slow on
the 1H NMR time scale. Importantly, signals correspond-
ing to both free and bound tert-butylamine could be
observed at this temperature in the presence of excess
tBuNH2. These data support symmetrization mechanisms
the value measured for THF dissociation (50.0 kJ mol-1
;
11.9 kcal mol-1 at 240 K). Significantly, these measure-
ments set a lower limit for the Gibbs Free Energy barrier
for oxazoline-arm dissociation to be >15 kcal mol-1 (at
least for compound 2 at 320 K), and this limit likely
applies to other tris(oxazolinyl)borate yttrium com-
pounds as well.

Article
Inorganic Chemistry, Vol. 48, No. 16, 2009 8027
Figure 4. Exchange processes occurring in [Y(κ3-ToM)(NHtBu)2(tBuNH2)] (5). Variable-temperature 1H NMR spectroscopic data indicate that proton
transfer is slow relative to amine dissociation.
for 5 thatinvolve both amine dissociation/coordination and
proton exchange between coordinated amine and amide
ligands (see Figure 4). In fact, we conclude that intramole-
cular proton transfer from a coordinated amine to an
amido ligand is slower than amine dissociation/coordina-
tion (for this system). Unfortunately, a detailed kinetic
study of this system is precluded by the relatively similar
rates for the two processes; qualitative comparisons of the
spectra indicate that amine dissociation in 5 is faster than
THF dissociation in bis(alkyl) 1. In the bent lanthanocene
Cp*2LaNHR(NRH2) (R = Me, Et), intramolecular pro-
ton transfer is slow at 173 K.13b In that case, the observed
low-temperature structure is Cs-symmetric due to either a
mirror plane in the static structure or a second exchange
process that is fast on the NMR time scale at 173 K.13b
Additionally, we compared the relative binding affinity
HMBC experiment, are far downfield of the yttrium
amide resonance (-236.6 ppm, detected by a one-bond
1H-15N HMQC experiment). Unlike six-coordinate
compounds 1-5, the apparent symmetry of the ToM
ligand in five-coordinate 6 is C3v even to 190 K. Thus,
inversion of the 5-coordinate yttrium center in 6 (pre-
sumably via a pseudo-rotation or turnstile mechanism)
is faster than the NMR time scale. Importantly, this
observation rules out yttrium inversion as the rate deter-
mining step in the symmetrization process for 1 and
provides additional support for THF dissociation as the
slowest step.
t
of THF versus BuNH2 as neutral ligands (L) for the
metal center in [Y(κ3-ToM)(NHtBu)2(L)]. Addition of
three equivalents of tBuNH2 to 1 in benzene-d6 produces
a single set of oxazoline resonances in the H NMR
1
spectrum at room temperature, and neither free THF
nor free BuNH2 are observed because of their involve-
t
Catalytic Hydroamination. The catalytic competence of
tris(oxazolinyl)borate-containing rare earth compounds
was established using hydroamination/cyclization of
aminopentenes as a test reaction. Since tris(pyrazolyl)-
borate and tris-ox rare earth compounds have not been
reported to catalyze hydroamination reactions, this study
demonstrates new catalytic chemistry for compounds
containing monoanionic fac-coordinating ligands. The
substituted substrates 2,2-diphenyl-4-penten-1-amine (7)
and 2,2-dimethyl-4-penten-1-amine (8) are cyclized by
5 mol % 1 in benzene-d6 to form the corresponding
2-methylpyrrolidine products in quantitative yield.
Cyclizations of diphenyl-substituted 7 proceed readily
at room temperature, whereas reactions of 8 require
50 ꢀC for efficient conversion. In addition, the OPPh3
adduct 2 and the five-coordinate bis(anilido) complex
6 were also tested as precatalysts for the cyclization
of diphenyl substrate 7. Compared to 1, conversion is
only slightly faster with 5-coordinate precatalyst 6, and
much slower in the presence of the strongly coor-
dinating OPPh3 ligand in 2. Additionally, the secon-
dary amine substrate 9 is also readily cyclized by 1 to
form the N-methylpyrrolidine product, but the tris-
(oxazolinyl)borate yttrium catalysts are unable to cyclize
the unsubstituted 4-penten-1-amine (10). The results of
these studies are summarized in Table 1.
ment in rapid dissociation/coordination processes. The
spectra are broad even at 190 K in toluene-d8, although
oxazoline resonances consistent with Cs-symmetric struc-
tures of 3 and 5 are observed at that temperature, as
well as two resonances due to tert-butylamide ligands
in 3 and 5. Resonances for free and bound THF are
also evident, as well as free tBuNH2. Apparently, 3 and 5
have similar formation constants and in the presence
of equal quantities of THF and tBuNH2 both are
present. Integration of free tBuNH2 and THF resonances
(or the tert-butylamide groups of 3 and 5) in the 1H NMR
spectrum recorded at 190 K indicates that 3 and 5 are
present in a 1.2:1 ratio. Thus, THF is a slightly better
ligand for the putative five-coordinate [Y(κ3-ToM)
(NHtBu)2] than tBuNH2.
Good support for the intermediacy of five-coordinate
yttrium(III) centers in the ToMY-system is provided by
sterically hindered amines. In particular, the larger
primary aniline 2,6-iPr2C6H3NH2 reacts to form the
THF-free five-coordinate species [Y(κ3-ToM)(NH-
2,6-iPr2C6H3)2] (6, eq 5). The ToM ligand appears C3v-
symmetric in 6, which is incommensurate with a static
five-coordinate species, and thus this compound is also
fluxional. The 15N NMR resonance for the oxazolinyl
1
groups in ToM (-139.8 ppm), observed with a H-15N

8028 Inorganic Chemistry, Vol. 48, No. 16, 2009
Pawlikowski et al.
Table 1. Hydroamination Cyclization Catalyzed by Tris(oxazolinyl)boratoyt-
trium(III) Complexes; Typical Conditions: Benzene-d6, 400 mM [substrate], 20-
25 mM [catalyst]
Zero-order substrate dependence is not a result of satura-
tion kinetics, since high catalyst loadings also provide
linear plots of [substrate] versus time for three half-lives.
Additionally, curves obtained by plotting product con-
centration versus time also are linear over the observed
reaction time. A plot of reaction rate versus precatalyst
concentration for a constant [substrate]0 is linear over the
measured concentration range (17.1-40.8 mM), indicating
a first-order dependence on [1]. The resulting rate law, -d
[substrate]/dt=kobs[1]1[substrate]0, is consistent with the
commonly postulated mechanism for rare earth element-
catalyzed intramolecular hydroamination reactions in
which the turnover limiting step is an alkene insertion.13
A nonzero (negative) intercept and slope ¼ 1 for a plot of
kobs versus[catalyst]indicates that a portion ofthecatalytic
sites are deactivated during the reaction, presumably by
reversible binding of a ligand to the sixth coordination site
on yttrium(III) inhibiting conversion. The primary amine
substrate, secondary amine product, and THF are all
likely inhibitors. In fact, the formation of [Y(κ3-ToM)-
(NHtBu)2(tBuNH2)] (5) in the presence of excess tBuNH2
suggests that the resting state of the catalyst during hydro-
amination catalysis is an amine adduct since excess amine
is present under catalytic conditions. The X-ray diffraction
study of 5 provides a model for the structure of the
catalyst’s resting state (see above). Interestingly, in-
creasing [1] increases the observed rate (and presumably
the percentage of active sites) in a linear but greater-
than-unity fashion. As the concentration of 1 is in-
creased at higher loadings, [yttrium] and [THF] increase
in a strictly 1:1 fashion, whereas the yttrium:amine ratio
changes as catalyst concentration is increased. Since
greater inhibition is observed at low catalyst loadings,
we suggest that amine coordination is more significant
to catalyst inhibition than THF binding. Because the
rate is essentially invariant over the reaction time,
product and substrate coordination to yttrium contri-
bute approximately equally to the inhibition.13b,18h
Similar catalytic activities are observed for 1 and the
THF-free precatalyst 6 in the cyclization of diphenyl
substrate 7 (Table 1, entries 2 and 4) which is consistent
with the proposal that amine coordination is more
significant than THF binding in catalysis inhibition.
However, the coordination of OPPh3 to the active
catalyst formed from 2 apparently becomes competitive
with amine coordination and the turnover frequency is
decreased by a factor of 3 (entry 3).
aNt = rate/[catalyst]. Rate was calculated from the slope of linear plots of
[substrate]/time through three half-lives. bClean zero-order kinetics were not
observed for 3 half-lives; Nt calculated from conversion of substrate through two
half-lives.
The catalytic activity of complexes 1, 2, and 6 for the
hydroamination/cyclization of substrates 7 and 8 is rather
modest with respect to highly active lanthanocene cata-
lysts. For example, cyclization of the gem-dimethyl sub-
strate 8 with catalyst 1 proceeded with a turnover
frequency of 1.9 h-1 at 50 ꢀC, compared to 95 h-1 at
25 ꢀC for [Cp*2LaCH(SiMe3)2].13b Marks has reported
that lanthanide bis(oxazoline) catalysts are extremely
efficient in asymmetric hydroamination/cyclization
reactions; a catalyst system derived from (4R,5S)-
Ph2BoxH (Box = 2,20-bis(2-oxazoline)methylenyl) and
[La{N(SiMe3)2}3] cyclizes gem-diphenyl substrate 7 at
23 ꢀC with a turnover rate exceeding that of catalyst 1
by several orders of magnitude (660 h-1 vs 2.4 h-1).20
The activity of catalyst 1 for aminoalkene 8 is more
comparable to other nonmetallocene yttrium-based
hydroamination catalysts. The simple homoleptic tris
(amide) complex [Y{N(SiMe3)2}3] cyclizes 8 with a turn-
over frequency of 11.6 h-1 at 25 ꢀC,28 and well-defined
yttrium alkyl and amido complexes supported by diami-
doamine ligands catalyze the same reaction with turnover
Conclusion
We have shown here that the tridentate monoanionic
tris(oxazolinyl)borate ToM ligand supports six-coordi-
nate bis(alkyl) and five- and six-coordinate bis(amido)
yttrium complexes. The bowl-like ToM ligand occupies
almost a hemisphere of space around the yttrium center
as quantified by solid angles, and this disfavors dispro-
portionation to mixtures of YR3 and bis(ToM)YR-type
complexes. However, the bulky ToM ligand does not
impede substitution reactions in [Y(κ3-ToM)(CH2Si-
Me3)2(THF)] since THF is labile and easily replaced by
OPPh3 and amines. Furthermore, THF substitution
occurs dissociatively through a five-coordinate inter-
mediate, and this is contrasted with associative THF
rates between 1.2 and 10 h-1 28b
.
Kinetics of the cyclization of 2,2-diphenyl-4-penten-1-
amine substrate (0.4 M) catalyzed by 1 (4-10 mol %)
were evaluated at 310 K; the rate is independent of
[substrate] for 3 half-lives (see the Supporting Information).
(28) (a) Kim, Y. K.; Livinghouse, T.; Bercaw, J. E. Tetrahedron Lett.
2001, 42, 2933–2935. (b) Hultzsch, K. C.; Hampel, F.; Wagner, T. Organo-
metallics 2004, 23, 2601–2612.

Article
Inorganic Chemistry, Vol. 48, No. 16, 2009 8029
substitution in [Y(κ3-Tp*)Cl2(THF)]10 as well as the
isolated seven-coordinate Tp*YX2(bipyridine) com-
pounds.14 The alkyl ligands in 1 react rapidly with
amines to form bis(amides), and this reaction occurs
even for relatively hindered amines such as 2,6-diisopro-
pylaniline. We have also found that the amido ligands in
complex 5 rapidly exchange with the amine ligands, and
the coordinated amine is rapidly substituted by free
amine. Qualitative inspection of variable temperature
such as [Y(κ3-ToM)(NHR)2(THF)] or [Y(κ3-ToM)(NHR)2-
(NH2R)] (R=CH2CR’2CH2CHdCH2). However, the isola-
tion and characterization of [Y(κ3-ToM)(NH-2,6-iPr2C6H3)2]
indicates that sufficient steric pressure favors five-coordinate
species in this class of compounds. Thus, we are currently
preparing more hindered oxazolinylborate ligands to destabi-
lize six-coordinate geometries and enhance catalytic activity in
hydroamination, as well as synthesizing optically active tris
(oxazolinyl)borate rare earth compounds for stereoselective
catalytic hydroamination reactions.
t
1H NMR spectra indicates that BuNH2 dissociation
from 5 is faster than THF dissociation from 1. Intramo-
lecular proton transfer between amine and amido li-
gands is slower than both dissociative processes in
these systems. Intramolecular proton transfer may be
slow since some geometric distortion is necessary for the
Acknowledgment. Support was provided by the Roy J.
Carver Charitable Trust and the ACS Green Chem-
istry Institute-Petroleum Research Fund. We thank
Bruce Fulton for assistance with 1H-15N HMBC
experiments.
proton to bridge between two nitrogen ligands (N
distance=3.3-3.5 A).
N
3 3 3
˚
1
Supporting Information Available: Variable-temperature H
The robust ToM-Y interactions and labile Y-L coordina-
tion are important for the catalytic activity of these com-
plexes. We have found that 1 is an active precatalyst for
hydroamination/cyclization of aminoolefins even though
much of the yttrium is tied up in dormant six-coordinate states
and 31P NMR spectra for 1 and 2, Eyring plots from line sha-
pe analysis, plots of conversion versus time for hydroamination
catalysis, and X-ray crystallographic data for 1 and 5 (PDF).
This material is available free of charge via the Internet at
http://pubs.acs.org.