Iminophosphorane neodymium(III) complexes as efficient initiators for lactide polymerization

Article abstract of DOI:10.1021/om1001233

Bis(iminophosphoranyl)methanide ({CH(PPh₂NR)}^-, R = iPr or Ph) neodymium(III) complexes were prepared from NdI₃(THF) _3.5. The steric bulk of the ligand controlled the stoichiometry of the resulting complexes. Thus, three new complexes, bearing one or two ancillary ligands, were prepared and characterized using various spectroscopic techniques and by single-crystal X-ray diffraction. Reaction of the heteroleptic neodymium iodide complexes with amido groups yielded viable initiators for the ring-opening polymerization of lactide. The polymerizations were conducted using either the heteroleptic neodymium amido complexes or the in situ generated alkoxide complexes. Using such conditions, the neodymium complexes showed very fast and well-controlled polymerizations, with complete conversion being obtained in only a few minutes and yielding polylactide with controlled molecular weight and narrow polydispersity index. An initiating system comprising a rare neodymium-alkyl-carbene complex [Nd{C(PPh₂NiPr) ₂}{CH(PPh₂NiPr)₂}] and externally added iPrOH was also an unexpected catalyst for the ROP of lactide.

Full text of DOI:10.1021/om1001233

2892 Organometallics 2010, 29, 2892–2900
DOI: 10.1021/om1001233
Iminophosphorane Neodymium(III) Complexes As Efficient Initiators
for Lactide Polymerization
Antoine Buchard,† Rachel H. Platel,‡ Audrey Auffrant,† Xavier F. Le Goff,†
Pascal Le Floch,† and Charlotte K. Williams*,‡
†
ꢀ ꢀ ꢀ ꢀ
Laboratoire Heteroelements et Coordination, CNRS, Ecole Polytechnique, 91128 Palaiseau Cedex, France,
and ‡Department of Chemistry, Imperial College London, London, SW7 2AZ, United Kingdom
Received February 15, 2010
Bis(iminophosphoranyl)methanide ({CH(PPh2NR)}-, R = iPr or Ph) neodymium(III) complexes
were prepared from NdI3(THF)3.5. The steric bulk of the ligand controlled the stoichiometry of the
resulting complexes. Thus, three new complexes, bearing one or two ancillary ligands, were prepared
and characterized using various spectroscopic techniques and by single-crystal X-ray diffraction.
Reaction of the heteroleptic neodymium iodide complexes with amido groups yielded viable
initiators for the ring-opening polymerization of lactide. The polymerizations were conducted using
either the heteroleptic neodymium amido complexes or the in situ generated alkoxide complexes.
Using such conditions, the neodymium complexes showed very fast and well-controlled polymeriza-
tions, with complete conversion being obtained in only a few minutes and yielding polylactide with
controlled molecular weight and narrow polydispersity index. An initiating system comprising a rare
neodymium-alkyl-carbene complex [Nd{C(PPh2NiPr)2}{CH(PPh2NiPr)2}] and externally added
iPrOH was also an unexpected catalyst for the ROP of lactide.
Introduction
The most widely used synthesis of PLA is the ring-opening
polymerization (ROP) of lactide.5 The ring-opening poly-
merization is thermodynamically favorable (ΔH = 22.1
kJ mol-1 for 1 M solutions of (S,S)-lactide at room temp-
erature), due to the relief of the lactone ring strain.6 When
using metal initiators having a nucleophilic co-ligand
(e.g., an amido or alkoxide group), a coordination insertion
mechanism is proposed (Scheme 1). According to such a
mechanism, the Lewis acidic metal center coordinates and
activates a lactide molecule, enabling it to be attacked by the
metal-bound alkoxide/amide group. The putative tetrahe-
dral intermediate undergoes ring-opening to generate a ring-
opened lactide molecule with an ester/amide chain end and a
new metal alkoxide bond. The metal alkoxide subsequently
goes on to bind, activate, and attack another lactide mole-
cule; this cycle repeats itself until the equilibrium conversion
is reached and polymer is formed. The selection of the ini-
tiator is of the utmost importance, as it governs the poly-
merization rate, control, and any stereoselectivity; therefore
it profoundly influences the physical and chemical properties
of the polymeric material. Homoleptic metal amide/alkoxide
complexes are effective initiators, but they are frequently
rather difficult to characterize, poorly soluble, in some
cases show complex kinetics, and lack any stereocontrol.7
To overcome these limitations, researchers have investigated
With current concerns about the environmental impact of
the plastics industry, in terms of both resource utilization
and plastics’ disposal, bioderived polyesters are a promising
alternative to petrochemicals.1 Among these, polylactide
(PLA) is particularly interesting because it is made from
lactic acid (produced by fermentation of corn or sugar beet),
it has suitable thermal and mechanical properties to displace
polyolefins, and it can be degraded to lactic acid by
hydrolysis.2 PLA is currently manufactured on a large scale
in the United States and by smaller companies in the EU and
Japan3 and has been used in both degradable packaging and
fiber applications. Furthermore, thanks to its biocompat-
ibility and its ability to degrade in vivo, PLA can be applied in
medicine for the fabrication of degradable sutures, bone
screws and pins, as a vector for long-term drug delivery and,
more recently, as a matrix material for tissue engineering.4
*Corresponding author. E-mail: c.k.williams@imperial.ac.uk.
(1) (a) Ragauskas, A. J.; Williams, C. K.; Davison, B. H.; Britovsek,
G.; Cairney, J.; Eckert, C. A.; Fredrick, W. J., Jr.; Hallett, J. P.; Leak,
D. J.; Liotta, C. L.; Mielenz, J. R.; Murphy, R.; Templer, R.;
Tschaplinski, T. Science 2006, 311, 484–489. (b) Mecking, S. Angew.
Chem., Int. Ed. 2004, 43, 1078–1085.
(2) Drumright, R. E.; Gruber, P. R.; Henton, D. E. Adv. Mater. 2000,
12, 1841–1846.
(3) http://www.cargilldow.com; http://www.mitsui.com; http://www.
hycail.com.
(4) (a) Ha, C.-S.; Gardella, J. A. Chem. Rev. 2005, 105, 4205–4232. (b)
Middlenton, J. C.; Tipton, A. J. Biomaterials 2000, 21, 2335–2346. (c) Jain,
R. A. Biomaterials 2000, 21, 2475–2490. (d) Place, E. S.; George, J. H.;
Williams, C. K.; Stevens, M. M. Chem. Soc. Rev. 2009, 38, 1139–1151.
(5) (a) Dechy-Cabaret, O.; Martin-Vaca, B.; Bourissou, D. Chem.
Rev. 2004, 104, 6147–6176. (b) Platel, R. H.; Hodgson, L. M.; Williams,
C. K. Polym. Rev. 2008, 48, 11–63.
(6) Duda, A.; Penczek, S. Thermodynamics, kinetics and mechanisms
of cyclic esters polymerisation. In Polymers from Renewable Resources:
Biopolyesters and Biocatalysis; Scholz, C.; Gross, R. A., Eds.; ACS
Symposium Series; Oxford University Press: Washington, DC, 1996,;
Vol. 5, pp 317-343.
(7) In the case of yttrium, see for example: Stevels, W. M.; Ankone,
M. J. K.; Dijkstra, P. J.; Feijen, J. Macromolecules 1996, 29, 6132–6138.
r
pubs.acs.org/Organometallics
Published on Web 06/15/2010
2010 American Chemical Society

Article
Organometallics, Vol. 29, No. 13, 2010 2893
Scheme 1. Coordination-Insertion Mechanism for the ROP
of Lactide
and some of them are able to exert good polymerization and
tacticity control.9b,h,t,y,12,14b
During the last few decades, neutral and monoanionic
ancillary ligands with sp2-hybridized nitrogen atoms have
become a well-established motif in coordination chemistry,
and their complexes have been extensively explored in
catalysis.13 The phosphorus analogues, i.e., ligands with the
iminosphosphorane (RPdNR0) moiety, have attracted far less
attention, despite their unusual electronic properties.14 In
direct contrast to imines, iminophosphoranes do not exhibit
any π-accepting capacity; the electron density on the nitrogen
is instead stabilized by negative hyperconjugation into the
phosphorus σ* orbitals. Thus iminophosphoranes essentially
behave as strong π and σ donor ligands due to the presence of
two lone pairs on the nitrogen atom. One example of this
ligand class is bis(iminophosphoranyl)methanides, which are
heteroleptic complexes of the form [LnMX], where L is an
ancillary ligand and X is the initiating group.8-11 Hetero-
leptic group 3 and lanthanide complexes have shown parti-
cular promise, as they generally exhibit very favorable rates
(11) Concerning the influence of the ancillary ligand architecture, see
for example: (a) Chamberlain, B. M.; Cheng, M.; Moore, D. R.;
Lobkovsky, E. B.; Ovitt, T. M.; Coates, G. W. J. Am. Chem. Soc.
2001, 123, 3229–3238. (b) Amgoune, A.; Thomas, C. M.; Roisnel, T.;
Carpentier, J.-F. Chem.;Eur. J. 2006, 12, 169–179.
(8) Other reviews on lactide ROP promoted by heteroleptic metallic
complexes: (a) O’Keefe, B. J.; Hillmyer, M. A.; Tolman, W. B. J. Chem.
Soc., Dalton Trans. 2001, 2215–2224. (b) Wu, J.; Yu, T.-L.; Chen, C.-T.;
Lin, C.-C. Coord. Chem. Rev. 2006, 250, 602–626. (c) Amgoune, A.;
Thomas, C. M.; Carpentier, J.-F. Pure Appl. Chem. 2007, 79, 2013–2030.
(9) For recent examples of lactide polymerization: (a) Binda, P. I.;
Delbridge, E. E.; Abrahamson, H. B.; Skelton, B. W. Dalton Trans.
(12) (a) Chamberlain, B. M.; Sun, Y.; Hagadorn, J. R.; Hemmesch,
E. W.; Young, V. G., Jr.; Pink, M.; Hillmyer, M. A.; Tolman, W. B.
Macromolecules 1999, 32, 2400–2402. (b) Chamberlain, B. M.; Jazdzewski,
B. A.; Pink, M.; Hillmyer, M. A.; Tolman, W. B. Macromolecules 2000, 33,
3970–3977. (c) Ovitt, T. M.; Coates, G. W. J. Am. Chem. Soc. 1999, 121,
4072–4073. (d) Ovitt, T. M.; Coates, G. W. J. Am. Chem. Soc. 2002, 124,
1316–1326. (e) Ma, H.; Spaniol, T. P.; Okuda, J. Dalton Trans. 2003, 4770–
4780. (f) Ma, H.; Spaniol, T. P.; Okuda, J. Angew. Chem., Int. Ed. 2006, 45,
7818–7821. (g) Cai, C. X.; Amgoune, A.; Lehmann, C. W.; Carpentier, J.-F.
Chem. Commun. 2004, 330–331. (h) Hodgson, L. M.; White, A. J. P.;
Williams, C. K. J. Polym. Sci., Polym. Chem. 2006, 44, 6646–6651. (i)
Westmoreland, I.; Arnold, J. Dalton Trans. 2006, 4155–4163. (j) Alaaed-
dine, A.; Amgoune, A.; Thomas, C. M.; Dagorne, S.; Bellemin-Laponnaz, S.
Eur. J. Inorg. Chem. 2006, 3652–3658. (k) Arnold, P. L.; Buffet, J.-C.;
Blaudeck, R.; Sujecki, S.; Blake, A. J.; Wilson, C. Angew. Chem., Int. Ed.
2008, 47, 6033–603. (l) Ma, H.; Spaniol, T. P.; Okuda, J. Inorg. Chem. 2008,
47, 3328–3339. (m) Konkol, M.; Spaniol, T. P.; Kondracka, M.; Okuda, J.
Dalton Trans. 2007, 4095–4103.
ꢀ
ꢀ
2009, 2777–2787. (b) Otero, A.; Fernandez-Baeza, J.; Lara-Sanchez, A.;
ꢀ
ꢀ
Alonso-Moreno, C.; Marquez-Segovia, I.; Sanchez-Barba, L. F.; Rodríguez,
A. M. Angew. Chem., Int. Ed. 2009, 48, 2176–2179. (c) Mahrova, T. V.;
Fukin, G. K.; Cherkasov, A. V.; Trifonov, A. A.; Ajellal, N.; Carpentier, J.-F.
Inorg. Chem. 2009, 48, 4258–4266. (d) Wheaton, C. A.; Ireland, B. J.;
Hayes, P. G. Organometallics 2009, 28, 1282–1285. (e) Labourdette, G.;
Lee, D. J.; Patrick, B. O.; Ezhova, M. B.; Mehrkhodavandi, P. Organome-
tallics 2009, 28, 1309–1319. (f) Alaaeddine, A.; Thomas, C. M.; Roisnel, T.;
Carpentier, J.-F. Organometallics 2009, 28, 1469–1475. (g) Wang, J.; Yao,
Y.; Zhang, Y.; Shen, Q. Inorg. Chem. 2009, 48, 744–751. (h) Ajellal, N.;
Lyubov, D. M.; Sinenkov, M. A.; Fukin, G. K.; Cherkasov, A. V.; Thomas,
C. M.; Carpentier, J.-F.; Trifonov, A. A. Chem.;Eur. J. 2008, 14, 5440–
5448. (i) Knight, P. D.; White, A. J. P.; Williams, C. K. Inorg. Chem. 2008,
47, 11711–11719. (j) Florczak, M.; Duda, A. Angew. Chem., Int. Ed. 2008,
47, 9088–9091. (k) Douglas, A. F.; Patrick, B. O.; Mehrkodavandi, P. Angew.
Chem., Int. Ed. 2008, 47, 2290–2293. (l) Pang, X.; Du, H. H.; Chen, X.;
Wang, X. M.; Jing, X. Chem.;Eur. J. 2008, 14, 3126–3136. (m) Hodgson,
L. M.; Platel, R. H.; White, A. J. P.; Williams, C. K. Macromolecules 2008,
41, 8603–8607. (n) Platel, R. H.; White, A. J. P.; Williams, C. K. Inorg.
Chem. 2008, 48, 6840–6849. (o) Guofu, Z.; Qiuwen, W.; Li, X.; Haibin, S.
Dalton Trans. 2008, 5930–5944. (p) Chmura, A. J.; Chuck, C. J.; Davidson,
M. G.; Jones, M. D.; Lunn, M. D.; Bull, S. D.; Mahon, M. F. Angew. Chem.,
Int. Ed. 2007, 46, 2280–2283. (q) Platel, R. H.; Hodgson, L. M.; White,
A. J. P.; Williams, C. K. Organometallics 2007, 26, 4955–4963. (r) Heck, R.;
Schulz, E.; Collin, J.; Carpentier, J.-F. J. Mol. Catal. A 2007, 268, 163–168.
(s) Nomura, N.; Ishii, R.; Yamamoto, Y.; Kondo, T. Chem.;Eur. J. 2007,
13, 4433–4451. (t) Du, H.; Pang, X.; Yu, H.; Zhuang, X.; Chen, X.; Cui, D.;
Wang, X.; Jing, X. Macromolecules 2007, 40, 1904–1913. (u) Liu, B.; Cui,
D.; Ma, J.; Chen, X.; Jing, X. Chem.;Eur. J. 2007, 13, 834–845. (v)
Pietrangelo, A.; Hillmyer, M. A.; Tolman, W. B. Chem. Commun. 2009,
2736–2737. (w) Chmura, A. J.; Davidson, M. G.; Frankis, C. J.; Jones, M. D.;
Lunn, M. D. Chem. Commun. 2008, 1293–1295. (x) Platel, R. H.; White,
A. J. P.; Williams, C. K. Chem. Commun. 2009, 4115–4117. (y) Arnold, P. L.;
Buffet, J.-C.; Blaudeck, R.; Sujecki, S.; Wilson, C. Chem.;Eur. J. 2009, 15,
8241–8250. (z) Chmura, A. J.; Cousins, D. M.; Davidson, M. G.; Jones,
M. D.; Lunn, M. D.; Mahon, M. F. Dalton Trans. 2008, 1437–1443. (aa)
Jones, M. D.; Davidson, M. G.; Keir, C. G.; Hughes, L. M.; Mahon, M. F.;
Apperley, D. C. Eur. J. Inorg. Chem. 2009, 435–642. (bb) Johnson, A. L.;
(13) (a) Mindiola, D. J. Angew. Chem., Int. Ed. 2009, 48, 6198–6200.
(b) Bourget-Merle, L.; Lappert, M. F.; Severn, J. S. Chem. Rev. 2002, 102,
3031–3066. (c) Holland, P. L. Acc. Chem. Res. 2008, 41, 905–914. (d)
Gibson, V. C.; Redshaw, C.; Solan, G. A. Chem. Rev. 2007, 107, 1745–1776.
(14) (a) Cavell, R. G.; Aparna, K.; Babu, R. P. K.; Wang, Q. Y. J.
€
Mol. Catal. A 2002, 189, 137–143. (b) Gamer, M. T.; Rastatter, M.; Roesky,
P. W.; Steffens, A.; Glanz, M. Chem.;Eur. J. 2005, 11, 3165–3172. (c)
Gamer, M. T.; Roesky, P. W.; Palard, I.; Le Hellaye, M.; Guillaume, S. M.
Organometallics 2007, 26, 651–657. (d) Wiecko, M.; Roesky, P. W.;
Burlakov, V. V.; Spannen, A. Eur. J. Inorg. Chem. 2007, 876–881. (e)
Panda, T. K.; Zulys, A.; Gamer, M. T.; Roesky, P. W. Organometallics 2005,
€
24, 2197–2202. (f) Rastatter, M.; Zulys, A.; Roesky, P. W. Chem. Commun.
€
2006, 874–876. (g) Rastatter, M.; Zulys, A.; Roesky, P. W. Chem.;Eur. J.
2007, 13, 3606–3616. (h) Buchard, A.; Auffrant, A.; Klemps, C.; Vu-Do, L.;
Boubekeur, L.; Le Goff, X. F.; Le Floch, P. Chem. Commun. 2007, 1502–
1504. (i) Buchard, A.; Komly, B.; Auffrant, A.; Le Goff, X. F.; Le Floch, P.
Organometallics 2008, 27, 4380–4385. (j) Buchard, A.; Heuclin, H.;
Auffrant, A.; Le Goff, X. F.; Le Floch, P. Dalton Trans. 2009, 1659–1667.
(k) Picot, A.; Dyer, H.; Buchard, A.; Auffrant, A.; Vendier, L.; Le Floch, P.;
Sabo-Etienne, S. Inorg. Chem. 2010, 49, DOI: 10.1021/ic902339j.
(15) (a) Babu, R. P. K.; Aparna, K.; McDonald, R.; Cavell, R. G.
Inorg. Chem. 2000, 39, 4981–4984. (b) Al-Benna, S.; Sarsfield, M. J.;
ꢁ
Thornton-Pett, M.; Ormsby, D. L.; Maddox, P. J.; Bres, P.; Bochmann, M. J.
Chem. Soc., Dalton Trans. 2000, 4247–4257. (c) Babu, R. P. K.; Aparna, K.;
McDonald, R.; Cavell, R. G. Organometallics 2001, 20, 1451–1455. (d) Wei,
P. R.; Stephan, D. W. Organometallics 2002, 21, 1308–1310. (e) Evans, D. J.;
Hill, M. S.; Hitchcock, P. B. Dalton Trans. 2003, 570–574. (f) Hill, M. S.;
Hitchcock, P. B. Dalton Trans. 2003, 4570–4571. (g) Hill, M. S.; Hitchcock,
P. B. Chem. Commun. 2003, 1758–1759. (h) Sarsfield, M. J.; Steele, H.;
Helliwell, M.; Teat, S. J. Dalton Trans 2003, 3443–3449. (i) Wei, P. R.;
Stephan, D. W. Organometallics 2003, 22, 601–604. (j) Hill, M. S.; Hitch-
cock, P. B. J. Organomet. Chem. 2004, 689, 3163–3167. (k) Bibal, C.; Pink,
M.; Smurnyy, Y. D.; Tomaszewski, J.; Caulton, K. G. J. Am. Chem. Soc.
2004, 126, 2312–2313. (l) Panda, T. K.; Benndorf, P.; Roesky, P. W. Z.
Anorg. Allg. Chem. 2005, 631, 81–84. (m) Ahmed, S. A.; Hill, M. S.;
Hitchcock, P. B. Organometallics 2006, 25, 394–402. (n) Gamer, M. T.;
Dehnen, S.; Roesky, P. W. Organometallics 2001, 20, 4230–4236.
ꢀ
Davidson, M. G.; Perez, Y.; Jones, M. D.; Merle, N.; Raithby, P. R.; Richards,
S. P. Dalton Trans. 2009, 5551–5558. (cc) Peckermann, I.; Kapelski, A.;
Spaniol, T. P.; Okuda, J. Inorg. Chem. 2009, 28, 5526–5534. (dd) Buffet,
J.-C.; Okuda, J.; Arnold, P. L. Inorg. Chem. 2010, 49, 419–426. (ee) Chen,
H.-Y.; Lin, M.-Y.; Sutar, A. K.; Lin, C.-C. Inorg. Chem. 2010, 49, 665–674.
(ff) Bouyahyi, M.; Roisnel, T.; Carpentier, J.-F. Organometallics 2010, 29,
491–500.
(10) About the influence of the initiating group, see for example: (a)
Chisholm, M. H.; Delbridge, E. E. New J. Chem. 2003, 27, 1177–1183. (b)
Ma, H.; Okuda, J. Macromolecules 2005, 38, 2665–2673.

2894 Organometallics, Vol. 29, No. 13, 2010
Scheme 2. Synthesis of Complexes 2a, 2b, and 3a
Buchard et al.
able to donate six electrons to the metal center and are well
adapted to stabilize electron-poor metal centers, including
lanthanides.15 Hill and co-workers have already reported bis-
(iminophosphoranyl)methanide zinc alkoxide complexes for
the ROP of lactide, nicely illustrating the potential for such
ligand systems.16 Yet, despite the good precedence for the use
of neodymium complexes in a range of catalytic reactions,
including for the ROP of lactide,9b,11b iminophosphorane-
based neodymium complexes are still under-exploited.14b
Herein, we wish to report the coordination of anionic bis(imino-
phosphoranyl)methanide ligands to various neodymium pre-
cursors and the use of the resulting complexes as initiators for
lactide ROP.
phy, is presented in Figure 1. Most of the relevant structural
parameters are listed in the corresponding caption.
In complex 3a, the Nd atom is coordinated by two
tridentate bis(iminophosphoranyl)methanide ligands, each
in a twisted boat conformation, and an iodide ligand. The
geometry at the metal center is distorted pentagonal bipyr-
amidal with a N1-Nd1-I1 angle of 172.20(8)°. The steric
congestion around the metal center precludes any solvent
coordination. This suggests that the steric hindrance of the
ligand, in particular the size of the nitrogen substituents, is
key to controlling the formation of this complex. In contrast,
the analogous reaction between the phenyl-substituted li-
gand 1b and 0.5 equiv of [NdI3(THF)3.5] at room tempera-
ture resulted, as shown by 31P{1H} NMR spectroscopy, in a
mixture of complex 2b (broad singlet at -150 ppm) and the
residual ligand. The formation of [NdI{CH(PPh2NPh)2}2]
was not observed, presumably due to the increased steric
hindrance of the ligand. On the other hand, the reaction of 1b
with 1 equiv of [NdI3(THF)3.5] successfully led to 2b, which was
isolated as a pale blue-green solid in good yield (Scheme 2). It
was fully characterized by NMR spectroscopy, elemental
analysis, and X-ray crystallography (see Experimental Section
and Supporting Information, Figure S1(ii)).
Results and Discussion
The ligand syntheses were carried out following the
Kirsanov route,17 which enables the nitrogen substituent to
be easily altered. The two bis(iminophosphoranyl)methanide
anions 1a and 1b, bearing isopropyl and phenyl nitrogen
substituents, respectively, were obtained by in situ deprotona-
tion of the bis(aminophosphonium) salt, using 3 equiv of
potassium hexamethyldisilazane in THF.18 Subsequent coordi-
nation of the ligands to [NdI3(THF)3.5], in ratios of 1:1 and 2:1
1a/1b:Nd, yielded the corresponding neodymium iodide com-
plexes (Scheme 2). This synthetic approach was devised so that
complexes with either one or two initiating groups could be
obtained by metathesis reactions with the halide precursors.
As described in a preliminary report from our group,19
using ligand 1a resulted in neodymium complexes with one
(3a) or two (2a) iodide ligands. Complexes [Nd{CH(PPh2-
NiPr)2}(THF)2I2] (2a) and [Nd{CH(PPh2NiPr)2}2I] (3a)
were isolated as pale blue-green, air- and moisture-sensitive
solids. They were unequivocally characterized, despite their
As iodide ligands would not be sufficiently nucleophilic to
initiate lactide ROP, σ-metathesis reactions were performed
in order to replace them with better initiating groups. Initial
attempts using potassium alkoxides (KOEt, KOtBu) led to
only redistribution products. However, the substitution of
the iodide groups with amido ligands could be achieved by
reaction of THF solutions of complexes 2a/2b with potas-
sium hexamethyldisilazide (Scheme 3). The resulting com-
plexes 4a and 4b were characterized by broad singlets in the
31P{1H} NMR spectra (in THF) at -140 and -118 ppm,
1
respectively. Moreover, H NMR spectroscopy confirmed
1
paramagnetism, by multinuclear (31P{1H} and H) NMR
the presence of the N(SiMe3)2 groups. The molecular struc-
tures of both 4a and 4b were confirmed by X-ray crystal-
lographic analyses of single crystals grown either by slow
evaporation of a pentane solution (4a) or by diffusion of
petroleum ether into a THF solution (4b).
spectroscopy and by X-ray crystallography (see Supporting
Information Figure S1(i), and ref 19). A representation of the
molecular structure of 3a, obtained by X-ray crystallogra-
(16) Hill, M. S.; Hitchcock, P. B. J. Chem. Soc., Dalton Trans. 2002,
4694–4702.
(17) Demange, M.; Boubekeur, L.; Auffrant, A.; Mezailles, N.;
Figure 2 presents the molecular structures of complexes 4a
and 4b, along with significant bond distances and angles in
the corresponding caption.
The coordination polyhedra of 4a and 4b are, as predicted,
made up of two bis(trimethylsilyl)amido co-ligands and a
tridentate bis(iminophosphoranyl)methanide ligand. There
is not any THF coordination in either complex, probably as a
ꢀ
Ricard, L.; Le Goff, X.; Le Floch, P. New J. Chem. 2006, 30, 1745–1754.
ꢀ
(18) Klemps, C.; Buchard, A.; Houdard, R.; Auffrant, A.; Mezailles,
N.; Le Goff, X.-F.; Ricard, L.; Saussine, L.; Magna, L.; Le Floch, P.
New J. Chem. 2009, 1748–1752.
(19) Buchard, A.; Auffrant, A.; Ricard, L.; Le Goff, X.-F.; Platel,
R. H.; Williams, C. K.; Le Floch, P. Dalton Trans. 2009, 10219–10222.

Article
Organometallics, Vol. 29, No. 13, 2010 2895
Scheme 3. Synthesis of complexes 4a and 4b
more, no γ-CH agostic interaction could be observed be-
tween the methyl group on the silicon atom and the metal.
In the case of complex 3a, as we have already reported,19
the substitution reaction of the iodide ligand using 1 equiv of
KHMDS did not occur. Instead, a deprotonation reaction
occurred, yielding an unexpected mixed alkyl-carbene neo-
dymium complex, 5a (Scheme 4).
Figure 1. ORTEP view of complex 3a. Thermal ellipsoids are
drawn at the 50% probability level. Hydrogen atoms and solvent
˚
molecules have been omitted for clarity. Selected distances (A) and
Given the ability of lanthanide initiators in the ROP of
lactide,9b,h,t,y,12,14b it was of interest to determine the activity
of complexes 4a and 4b. Although lacking an initiating group,
complex 5a was also tested. The polymerization reactions were
carried out using 1 M concentrations of rac-lactide and 0.005 M
concentrations of complexes 4a/4b in THF (i.e., at a ratio of
200:1, LA:4a/4b). Under these conditions, the solutions became
very viscous after only a few seconds of reaction, indicating
rapid polymerization was occurring. The crude products were
angles (deg): Nd1-I1=3.270(4), Nd1-N1=2.454(4), Nd1-N2=
2.512(3), Nd1-N3 = 2.525(3), Nd1-N4 = 2.456(3), Nd1-C1=
2.898(3), Nd1-C32 = 2.837(3), N1-P1 = 1.609(3), N2-P2 =
1.594(3), N3-P3 = 1.611(3), N4-P4 = 1.607(3), C1-P1 =
1.731(4), C1-P2 = 1.739(4), C32-P3 = 1.716(4), C2-P1 =
1.836(4), N1-Nd1-I1 = 172.20(8), N2-Nd1-I1 = 83.67(8),
N3-Nd1-I1=85.07(7), N4-Nd1-I1=85.70(8), C1-Nd1-I1=
116.86(7), C32-Nd1-I1 = 107.50(8), P1-C1-P2 = 115.9(2),
P4-C32-P3 = 136.2(2), N1-Nd1-N2 = 88.8(1), N4-Nd1-
N3 = 112.0(1), Nd1-P1-P2-C1 = 80.29(15), P1-N1-N2-
Nd1=55.08(15), Nd1-N1-N2-C1=80.34(12), Nd1-P4-P3-
C32=81.65(19), P4-N4-N3-Nd1=20.25(20), Nd1-N3-N4-
C32 = 46.71(17), N1-P1-P2-C1 = 72.96(21), N3-P3-P4-
C32=81.04(1), N1-P1-P2-N2=8.87(20), N4-P4-P3-N3=
11.29(24), N1-Nd1-C1-P2 = 57.74(21), N2-Nd1-C1-P1 =
66.06(18), N3-Nd1-C32-P4=42.20(20), N4-Nd1-C32-P3=
34.23(17).
1
analyzed by H NMR spectroscopy (by integration of the
methine resonances due to lactide and polylactide; see the
Experimental Section); this showed that complete conversion
(>95%) was achieved within 5 min using both initiators. The
GPC analyses showed a linear increase in the Mn of the PLA
produced with the percentage conversion of lactide (e.g., using
4b, see Figure S2a of the Supporting Information), and the
PLA had relatively narrow PDI values (around 1.3), as ex-
pected for a controlled polymerization. In addition, MALDI-
ToF mass spectrometry confirmed that the polymer chains
were systematically end-capped with a bis(trimethylsilyl)amido
group and a hydroxy group, with no cyclic oligomer being
observed (for the spectrum obtained using 4a; see Figure S3 of
the Supporting Information). The MALDI-ToF spectrum also
showed evidence for transesterification reactions (series sepa-
rated by 72 atomic mass units). However, the Mn values for the
result of the steric hindrance of the trimethylsilyl groups.
Like in 2a, in 2b, and in other reported bis(iminopho-
sphoranyl)methanide neodymium(III) complexes,14b a me-
tallacycle (N1-P1-C1-P2-N2-Nd1) was formed upon
chelation of the iminophosphorane moieties to the metal.
In this boat-shaped conformation, the central carbon atom
and the neodymium atom are displaced from the almost
planar (dihedral angle N1-P1-P2-N2 = 5.08(5)° and
6.24(22)° in 4a and 4b, respectively) N2P2 base (N1-Nd1-
C1-P2 = 44.92(8)° and 52.33(11)°, N2-Nd1-C1-P1=
-49.42(8)° and 47.51(10)° in 4a and 4b, respectively). The
steric hindrance around the Nd center also influences the
bonding of the (N,C,N) fragments in both complexes. In-
deed, the Nd-Ccentral bonds are rather longer in both 4a,b
PLA were approximately 10 times those calculated (Mn,calc
=
% conversion ꢀ Mlactide ꢀ 200/number of initiating groups),
throughout the polymerization run (Figure S2a). This anom-
alously high Mn suggested that only one-tenth of the initiators
were active. For 4b, the polymerization kinetics were first-
order20 with respect to the rac-LA concentration (kapp
=
0.013 s-1; see Figure S2b), and no significant initiation period
was observed, discounting slow initiation as the cause of the
unexpectedly high molecular number. The polymerization was
also analyzed in situ by 31P{1H} NMR spectroscopy at com-
plete monomer conversion, and this showed significant quan-
tities of unreacted catalyst 4b (δ = -118 ppm). Comparable
results were also obtained using solutions of 4b in toluene,
precluding the coordination of THF to the Nd as an explana-
tion for the reduced control (further reinforced by findings that
4b is not solvated by THF molecules in the solid state). Finally,
it has already been established that bis(trimethylsilyl)amido
groups are relatively poor initiating groups, due to their steric
˚
compared to those measured in 2a,b (2.886(2) and 2.857(3) A
˚
in 4a and 4b, respectively, vs 2.807(4) and 2.801(4) A in 2a
and 2b, respectively). Likewise, the Nd-N bonds are elon-
gated in complexes 4a,b compared to those in complexes 2a,
˚
b. Yet interestingly in 4b they are both equal to 2.478(3) A (vs
ꢀ
˚
Nd-N = 2.413 A on average in 2b), revealing a symmetrical
binding of the {CH(PPh2NPh)2} ligand. On the contrary, in
4a, the Nd-N bond lengths differ, showing an unsymme-
trical bonding of the two ligands in the complex (Nd1-N1
˚
and Nd1-N2 being 2.490(2) and 2.471(2) A, respectively, vs
2.403(1) and 2.418(4) in 2a). Lastly, the two amido groups
are not equivalent but are orthogonal to one another; this
arrangement may account for the steric crowding. Further-
(20) A slow-down of the reaction rate from 90% conversion was
observed.

2896 Organometallics, Vol. 29, No. 13, 2010
Buchard et al.
Figure 2. ORTEP views of complex 4a (i) and 4b (ii). Thermal ellipsoids are drawn at the 50% probability level. Hydrogen atoms and
˚
solvent molecules have been omitted for clarity. Selected distances (A) and angles (deg): 4a: Nd1-N3=2.345(2), Nd1-N4=2.384(2),
Nd1-N1=2.490(2), Nd1-N2=2.471(2), Nd1-C1=2.886(2), N1-P1=1.622(2), N2-P2=1.618(2), C1-P1=1.728(2), C1-P2=
1.820(2), N3-Nd1-N4=110.60(7), N3-Nd1-N1=103.8(8), N3-Nd1-N2=123.3(7), N4-Nd1-N1=118.26(7), N4-Nd1-N2=
95.70(7), N3-Nd1-C1=97.85(6), N4-Nd1-C1=150.10(6), P2-C1-P1=131.1(1), N2-Nd-N1=106.02(7), Nd1-N1-N2-C1=
60.93(7), Nd1-P1-P2-C1=82.92(9), P1-N1-N2-Nd1=36.65(10), N1-P1-P2-C1=80.84(13), N1-P1-P2-N2=5.08(15)(1),
N1-Nd1-C1-P2 = 44.92(8), N2-Nd1-C1-P1 = 49.42(8); 4b: Nd1-N3 = 2.325(3), Nd1-N4 = 2.384(2), Nd1-N1 = 2.478(3),
Nd1-N2=2.478(3), Nd1-C1=2.857(3), N1-P1=1.626(3), N2-P2=1.620(3), C1-P1=1.720(3), C1-P2=1.736(3), N3-Nd1-N4=
110.2(1), N3-Nd1-N1=110.6(1), N3-Nd1-N2=118.14(8), N4-Nd1-N1=109.33(8), N4-Nd1-N2=105.27(8), N3-Nd1-C1=
94.7(1), N4-Nd1-C1=155.0(1), P1-C1-P2=131.1(2), N2-Nd-N1=102.7 (1), Nd1-N1-N2-C1=66.46(12), Nd1-P1-P2-C1=
79.83(22), P1-N1-N2-Nd1=46.92(10), N1-P1-P2-C1=75.15(20), N1-P1-P2-N2=6.24(22), N1-Nd1-C1-P2=52.33(11),
N2-Nd1-C1-P1=47.51(10).
Scheme 4. Synthesis of complexes 4a and 4b
Scheme 5. In Situ Formation of Complex 6b
bulk and reduced nucleophilicity, and this is the probable
explanation for the behavior observed here.9p,11b
Complexes with alkoxide initiating groups, which mimic
the putative propagating actives species, are known to be
superior initiators.9f,g,11b,21 However, isolated examples of
such heteroleptic lanthanide alkoxide complexes are not very
common.9c,h,22 Previously, metal alkoxide complexes have
been generated in situ by addition of alcohols to metal amido
complexes and then used directly in the ROP.9h,q,n,11,16 For
lanthanide complexes, these alkoxides can be observed by
NMR spectroscopy but are difficult to isolate, as they are
believed to aggregate.10,23 Nevertheless, the isolation of a
bis(alkoxide) Nd complex, [Nd{CH(PPh2NPh)2}(OR)2],
was attempted. The addition of sodium isopropoxide or
potassium tert-butoxide to the iodide complex 2b did not
yield the expected alkoxide complexes but rather led to
ligand displacement. However, the addition of 2 equiv of
iPrOH to complex 4b led to the formation of a new alkoxide
complex. Indeed, in situ 31P{1H} NMR spectroscopy showed
a broad signal at -122 ppm, consistent with the formation of
a heteroleptic alkoxide complex. 1H NMR spectroscopy was
not conclusive due to the complex paramagnetism, but the
release of bis(trimethylsilyl)amine was apparent.
(21) (a) Williams, C. K.; Brooks, N. R.; Hillmyer, M. A.; Tolman,
W. B. Chem. Commun. 2002, 2132–2133. (b) Cheng, M.; Moore, D. R.;
Reczek, J. J.; Chamberlain, B. M.; Lobkovsky, E. B.; Coates, G. W. J. Am.
Chem. Soc. 2001, 123, 8738–8749. (c) Lobkovsky, E. B.; Coates, G. W. J.
Am. Chem. Soc. 2001, 123, 8738–8749. (d) Simic, V.; Spassky, N.; Hubert-
Pfalzgraf, L. G. Macromolecules 1997, 30, 7338–7340. (e) Stevels, W. M.;
The heteroleptic complex is unstable, and after two hours,
only free ligand is observed (see Scheme 5). Although com-
plex 6b could not be isolated, its lifetime was sufficient to
enable its use in situ in polymerization runs (Table 1). Thus,
addition of 2 equiv of iPrOH to a 0.005 M solution of 4b, in
THF, and subsequent addition of the mixture to a 1 M
solution of LA (THF) led to highly efficient ROP. The
polymerizations were even faster than using solutions of
4b, with 90% conversion being reached in only 90 s and
complete conversion within 6 min (the reaction rate decreas-
ing at the end of the reaction; see Figure S5 of the Supporting
ꢀ
Ankone, M. J. K.; Dijkstra, P. J.; Feijen, J. Macromolecules 1996, 29, 6132–
6138.
(22) Izod, K.; Liddle, S. T.; McFarlane, W.; Clegg, W. Organome-
tallics 2004, 23, 2734–2743.
(23) (a) Barnhart, D. M.; Clark, D. L.; Huffman, J. C.; Vincent, R. L.;
Watkin, J. G. Inorg. Chem. 1993, 32, 4077–4083. (b) Gromada, J.;
Mortreux, A.; Chenal, T.; Ziller, J. W.; Leising, F.; Carpentier, J.-F.
Chem.;Eur. J. 2002, 8, 3773–3778.

Article
Organometallics, Vol. 29, No. 13, 2010 2897
Table 1. Ring-Opening Polymerization of rac-Lactide in THF by in Situ Generated 6a and 6ba
b
c
entry
initiator [I]
[LA]0/[I]0
time [s]
conversion [%]
Mn,exp
Mn,calc
PDIb,d
Pre
1
2
3
4f
5
6
7
6b
6b
6b
6b
200
400
1000
200
200
200
400
360
360
600
300
300
300
300
95
95
94
96
96
95
96
12 600
25 100
109 000
12 700
14 600
17 500
29 000
13 700
27 400
67 700
13 800
13 800
9100
1.05
1.04
1.21
1.07
1.07
1.17
1.18
0.58
0.60
0.57
6a
[Nd(OiPr)3]
[Nd(OiPr)3]
0.57
0.51
0.52
18 400
a All reactions were performed with [rac-LA]0 = 1 M at 298 K, in THF, under inert atmosphere. 6a and 6b are generated in situ by adding 2 equiv of dry
2-propanol to solutions of 4a and 4b, respectively, in THF. Conversions were determined by integration of the methine resonances of LA (1H NMR δ =
4.92 ppm in CDCl3) and PLA (1H NMR δ = 5.00-5.30 ppm in CDCl3). b Experimental Mn and Mw were determined by GPC in THF, using multiangle
laser light scattering (GPC-MALLS). c Mn,calc = 144 ꢀ 200 ꢀ % conversion LA ꢀ 0.5. d PDI = Mw/Mn. e Pr is the probability of racemic linkages
1
1
between monomer units and is determined from the homonuclear decoupled H NMR spectrum: H{1H} NMR δ 5.14 and 5.22 ppm in CDCl3 for
heterotactic PLA.24 f (S,S)-LA was used instead of rac-LA in order to detect epimerization.
Information). Moreover, the molecular number of the poly-
mer showed an excellent correlation with the calculated
values. The polydispersity index was narrow (1.05), indicat-
ing a high level of polymerization control. End-group anal-
ysis by MALDI-ToF mass spectrometry showed that the
polymer chains were systematically end-capped with an
isopropyl ester and a hydroxy group; no cyclic oligomer
was detected (see Figure S6), although, once again, there
was evidence for transesterification (series separated by 72
atomic mass units). The use of lower initiator loadings was
also possible (up to 1000:1), with complete conversion being
achieved within 15 min (Table 1, entries 2 and 3). Some
molecular weight control is, however, lost for high monomer:
initiator ratios, as already observed.25 The controlled nature
of the polymerization was also confirmed by a sequential
monomer addition experiment. The experiment involved
addition of a further 100 equiv of lactide to a typical poly-
merization run (100:1, LA:6b, Mn = 6800, PDI = 1.05) after
5 min (93% conversion); these further equivalents were
polymerized to yield PLA of increased molecular number
(Mn = 14 200, PDI = 1.14, 95% conversion overall). As 6b
was known to slowly degrade in THF (or toluene), further
investigation was required to rule out the involvement of any
homoleptic tris-alkoxide neodymium complexes. Thus, Nd
alkoxide species were formed in situ (by reaction of 3 equiv of
isopropanol with [Nd(N(SiMe3)2)3])26 and were then used to
initiate a LA ROP (entries 6 and 7 in Table 1). Previously,
in situ generated rare-earth homoleptic tris-isopropoxide
species (e.g., Y or La) were reported to be very fast initiators
for the LA ROP, with excellent control over molecular
number on the basis of three polymer chains growing per
metal center (i.e., all three alkoxide groups initiate).27 In our
case, the homoleptic neodymium species was poorly soluble
in THF, which makes any discussion of its control of the
molecular weight more difficult, although there was not a
good agreement between the experimental values for Mn and
those calculated assuming initiation from three sites. On the
exclusive basis of the excellent correlation between experi-
mental and calculated Mn using 6a,b, it was not possible to
unambiguously rule out the presence of significant quantities
of homoleptic Nd alkoxide complex. However, the in situ
generated heteroleptic complexes 6a,b were soluble in THF
and, more importantly, on the time scale of the polymeriza-
tion runs, exhibited a single signal in the 31P{1H} NMR
spectra, without any traces of free ligand, which would be
expected to result from the formation of homoleptic species.
For these reasons, when using 4b and 2-propanol for poly-
merizations, we do not expect there to be significant quan-
tities of homoleptic alkoxides contaminating.
No stereocontrol was observed using 4b/iPrOH (Pr =
0.58, probability of racemic linkage between the LA units).
Furthermore, a controlled polymerization of (S,S)-lactide by
the 4b/iPrOH system (entry 4 in Table 1) resulted in only
isotactic PLA, evidenced by the single sharp resonance in the
methine region in the homonuclear decoupled 1H NMR
spectrum (1H NMR δ 5.16 ppm in CDCl3).23 Some polym-
erization reactions were also conducted in THF by addition
of 2 equiv of iPrOH to a solution of 4a (Table 1, entry 5) to
generate in situ the bis-alkoxide complex 6a. This 4a/iPrOH
system also led to highly efficient ROP. With a monomer:
initiator ratio of 200:1, polymerization was even faster than
when using in situ generated 6b, with 91% conversion being
reached in only 30 s and 96% conversion being achieved
within 5 min. GPC analysis revealed good control of the PLA
molecular number and a narrow polydispersity index. The
influence of the steric hindrance at the nitrogen atom of the
iminophosphorane groups on the rates was, however, un-
remarkable.
Polymerization experiments conducted using complex 5a
were unsuccessful, as expected from the lack of an initiating
group. Addition of a single equivalent of alcohol led to
protonation of the methanide or methanediide moieties,
and the iminophosphorane ligand displacement occurred
immediately. Moreover, polymerizations initiated using a
solution of 5a and 2-propanol indicated that the active
species was a homoleptic Nd alkoxide species (as evidenced
by very fast conversion and poor control of Mn). However
completely different behavior was observed if the monomer
and then the alcohol were added to a solution of 5a. In this
case, the polymerization occurred relatively quickly (com-
plete conversion within 90 min) and showed good control of
the molecular number and a narrow PDI (<1.1) over the
entire run (Table 2, entries 1-3). MALDI-ToF analysis, at
low conversion, revealed PLA end-capped with isopropoxy
ester groups. The high degree of polymerization control was
confirmed by a sequential monomer addition experiment. By
adding 100 extra equivalents of monomer after 60 min of a
typical run (LA:6b:iPrOH = 100:1:1, Mn = 11 000, PDI =
1.1 at 88%), complete consumption of the lactide and an
(24) Coudane, J.; Ustariz-Peyret, C.; Schwach, G.; Vert, M. J. Polym.
Sci., Polym. Chem. 1997, 35, 1651–1658.
(25) Williams, C. K.; Breyfogle, L. E.; Choi, S. K.; Nam, W. W.;
Young, V. G., Jr.; Hillmyer, M. A.; Tolman, W. B. J. Am. Chem. Soc.
2003, 123, 11350–11359.
(26) Kritikos, M.; Moustiakimov, M.; Wijk, M.; Westin, G. J. Chem.
Soc., Dalton Trans., 2001, 1931-1938, and references therein.
ꢀ
(27) (a) Stevels, W. M.; Ankone, M. J. K.; Dijkstra, P. J.; Feijen, J.
Macromolecules 1996, 29, 3332–3333. (b) Vesna Simic, V.; Spassky, N.;
Hubert-Pfalzgraf, L. G. Macromolecules 1997, 30, 7338–7340.

2898 Organometallics, Vol. 29, No. 13, 2010
Table 2. Ring-Opening Polymerization of rac-Lactide in THF Using 5aa
Buchard et al.
b
c
entry
[LA]0:[5a]0:[iPrOH]
t [min]
conversion [%]
Mn,exp
Mn,calc
PDIb,d
Pre
1
2
200:1:1
200:1:1
200:1:1
200:1:1
5
30
90
60
5
57
80
100
81
93
85
16 000
24 200
31 000
21 100
19 200
20 200
16 400
23 000
28 800
23 300
13 400g
24 400
1.05
1.08
1.11
1.04
1.66
1.05
3
0.60
4f
5
200:0:0 ([KOEt]0 = 0.01)g
200:0:0 ([KOEt]0/[5a]0= 1)h
6f
60
a All reactions were performed with [LA]0 = 1 M and [LA]0:[I]0:[iPrOH] = 200:1:1 at 298 K, in THF, under inert atmosphere. Conversions were
determined by integration of the methine resonances of LA (1H NMR δ = 4.92 ppm in CDCl3) and PLA (1H NMR δ = 5.00-5.30 ppm in CDCl3).
b Experimental Mn and Mw were determined by GPC, in THF, using multiangle laser light scattering (GPC-MALLS). c Mn,calc = 144 ꢀ ([LA]0/[I]0) ꢀ
conversion LA. d PDI = Mw/Mn. e Pr is the probability of racemic linkages between monomer units and is determined from the homonuclear decoupled 1H
NMR spectrum: δ 5.14 and 5.22 ppm in CDCl3 forheterotactic PLA.24 f (S,S)-LA was used instead of rac-LA todetectepimerization. g KOEtwas usedas the
only initiator for the polymerization, Mn,calc = 144 ꢀ 100 ꢀ % conversion LA. h KOEt was added to the reaction mixture instead of 2-propanol.
Scheme 6. Activated Monomer Mechanism Proposed to Rationalize the Controlled Activity of 5a/KOEt (a) and 5a/iPrOH (b) in the
ROP of Lactide
increase in the Mn of all the polymer chains were observed
(Mn = 21 700, PDI = 1.18, overall conversion of 96% after 2
extra hours). Finally, using this protocol, no epimerization
of (S,S)-lactide was detected (only isotactic PLA was
obtained).
either externally added alcohol or alkoxide groups.29 Ac-
cording to such a mechanism, the Nd complex activates the
monomer, which is subsequently attacked by the externally
added nucleophiles (either alcohol or alkoxide). If an alk-
oxide is added, the tetrahedral intermediate undergoes ring-
opening and proton exchange to liberate complex 5a and a
new alkoxide (a ring-opened species with potassium R-ester-
ω-alkoxide end groups). The cycle of reactions (i.e., activa-
tion, nucleophilic attack, and release of an alkoxide) repeats
until all the monomer is consumed and a polymer is formed
(see Scheme 6a). When adding alcohol, the tetrahedral
intermediate undergoes ring-opening and proton exchange
to liberate complex 5a and a new alcohol (a ring-opened
species with R-ester-ω-alcohol end groups). The activation,
nucleophilic attack, and release of an alcohol series of
reactions repeat until all the monomer is consumed and a
polymer is formed (see Scheme 6b). Once all the monomer is
consumed, the polymeric alcohol is free to attack 5a and
decomposition occurs (presumably generating a polymeric
alkoxide, which is protonated during the aqueous workup).
The use of excess alcohol was not feasible, as it led only to the
decomposition of 5a to the protonated ligand.
31P{1H} NMR spectroscopy was used to analyze the
mixture; no 5a-monomer adduct was detected, and in situ
31P{1H} NMR analysis of polymerization runs revealed only
the unchanged signal of complex 5a. Furthermore, by using
in situ ATR FTIR spectroscopy, no stretching of the carbo-
nyl bond of the lactide monomer could be evidenced in the
presence of 5a. Crystallization attempts to isolate a 5a-
monomer adduct28 were also unsuccessful. However, after
several hours and the completion of the reaction, the decom-
position of the complex and the release of the protonated
ligands were observed. Successful polymerization was also
possible using KOEt as the additive instead of 2-propanol
(Table 2, entries 5 and 6). While KOEt proved to be a very
active but chaotic initiator for the polymerization when used
alone (Table 2, entry 5), control of the molecular weight and
narrow polydispersity index was obtained when it was
combined with 5a (Table 2 entry 6).
In summary, we have developed a series of new neody-
mium complexes, bearing either one or two iminophosphor-
ane ligands and either iodide or amido co-ligands, with the
complex stoichiometry depending on the nitrogen substitu-
ent (either phenyl or isopropyl). All the Nd complexes were
fully characterized, including in all cases by single-crystal
X-ray diffraction. The amido complexes (4a/b) were rapid
An activated-monomer mechanism is proposed to explain
the activity of complex 5a (see Scheme 6) in the presence of
(28) Koeller, S.; Joji Kadota, J.; Alain Deffieux, A.; Peruch, F.;
ꢀ
Massip, S.; Leger, J.-M.; Desvergne, J.-P.; Bibal, B. J. Am. Chem. Soc.
2009, 131, 15088–15089.
(29) Nomura, N.; Taika, A.; Tomioka, T.; Okada, M. Macromole-
cules 2000, 33, 1497–1499.

Article
Organometallics, Vol. 29, No. 13, 2010 2899
initiators for the ring-opening polymerization of rac-lactide
but showed reduced polymerization control. However,
in situ addition of alcohol to complexes 4a and 4b, generated
isopropoxide complexes, which were even faster initiators
and still showed excellent polymerization control. An initi-
ating system comprising a rare neodymium-alkyl-carbene
complex [Nd{C(PPh2NiPr)2}{CH(PPh2NiPr)2}] (5a) and ex-
ternally added iPrOH was also an unexpected catalyst for the
ROP of lactide. This may be due to an activated-monomer
mechanism, in contrast to the traditional coordination-
insertion mechanism invoked for initiator 4a or 4b.
(t, 2JHP = 3.5 Hz, 1H, PCHP), 6.33 (t, 3JHH = 7.0 Hz, 2H, p-CH
(NPh)), 6.53 (m, JHH = 8.0 Hz, 4H, m-CH (NPh)), 6.75 (t,
3
3JHH = 7.0 Hz, 4H, m-CH (NPh)), 7.20 (m, JHH = 7.5 Hz,
3
12H, m,p-CH (PPh2)), 7.86 (m, 8H, o-CH (PPh2)) ppm. 13C{1H}
NMR (75.5 MHz, [D8]-THF, 20 °C): δ 14.3 (t, 1JCP = 131.0 Hz,
PCHP), 113.0 (s, p-CH (NPh)), 121.2 (t, 3JCP = 9.5 Hz, o-CH
(NPh)), 126.0 (t, 2JCP = 5.0 Hz, m-CH (PPh2)), 126.8 (s, m-CH
(NPh)), 127.4 (s, p-CH (PPh2)), 130.8 (t, 3JCP = 4.5 Hz, o-CH
(PPh2)), 138.7 (d, 1JCP = 98.5 Hz, ipso-C (PPh2)), 153.5 (s, ipso-
C (NPh)) ppm.
Synthesis of [Nd{CH(PPh2NPh)2}I2(THF)2], 2b. In THF (5 mL),
[NdI3THF3.5] (49 mg, 0.063 mmol) was added to a solution of
bis(iminophosphoranyl)methanide (1b) (39 mg, 0.063 mmol),
resulting in an immediate color change of the reaction mixture.
After 2 h of stirring, the reaction mixture was centrifuged and the
supernatant solution was dried in vacuo. The light green solid
residue was finally washed with petroleum ether (2 ꢀ 5 mL).
Crystals suitable for X-ray diffraction were obtained by slow
diffusion of hexanes into a solution of 2b in THF at room
temperature (298 K). 2b: yield 51mg (0.046mmol, 73%). 31P{1H}
NMR (121.5 MHz, [D8]-THF, 20 °C): δ -150 (bs, P) ppm. 1H
NMR (300 MHz, [D8]-THF, 20 °C): δ -50.7 (bs, 2H, p-CH
NPh), -25.0 (bs, 1H, PCHP), -14.9 (bs, 4H, CH NPh), -10.5
(bs, 4H, CH NPh), 4.5 (bs, 8H, OCH2CH2 THF), 6.5 (bs, 8H,
OCH2CH2 THF) 12.1 (bs, 4H, p-CH (PPh2)), 14.5 (bs, 8H, CH
(PPh2)), 19.8 (bs, 8H, CH (PPh2)) ppm. Anal. Calcd (%) for
C45H47N2O2P2NdI2: C 48.79, H 4.28 N 2.53. found: C 49.17, H
4.37 N 2.45.
Experimental Section
General Conditions. All experiments were performed under an
atmosphere of dry nitrogen or argon using standard Schlenk
and glovebox techniques. Solvents were freshly distilled under
nitrogen from Na/benzophenone (THF, diethyl ether, petroleum
ether), from P2O5 (dichloromethane), or from CaH2 (2-propanol).
Unless stated, reagents were purchased from commercial sources
and used without further purification. [Nd(BH4)3(THF)3], [Nd(N-
(SiMe3)2)3], [NdI3THF3.5],30 isopropyl-substituted bis(aminopho-
sphonium)methane salts, and complexes 2a, 3a, 4a, and 5a were
prepared according to literature procedures.19 rac-Lactide was
generously donated by Purac Plc. and was recrystallized from
anhydrous toluene and sublimed three times under vacuum prior
to use.
Synthesis of [Nd{CH(PPh2NPh)2}(N(SiMe3)2)2], 4b. In THF
(5 mL), KHMDS (36 mg, 0.180 mmol) was added to a solution
of complex 2b (100 mg, 0.090 mmol), resulting in an immediate
precipitation of KI. After 2 h of stirring, the reaction mixture
was centrifuged and the supernatant solution was dried in vacuo.
The light green solid residue was finally washed with petroleum
ether (2 ꢀ 5 mL). Crystals suitable for X-ray diffraction were
obtained by slow diffusion of hexanes into a solution of 4b in
THF at room temperature (298 K). 4b: yield 83 mg (0.081 mmol,
90%). 31P{1H} NMR (121.5 MHz, [D8]-THF, 20 °C): δ -118 (s,
P) ppm. 1H NMR (300 MHz, [D8]-THF, 20 °C): δ -14.7 (bs, 8H,
CH (PPh2)), -10.3 (bs, 36H, Si(CH3)3), 8.3-10.5 (m, 10H, CH
NPh), 15.2 (bs, 8H, CH (PPh2)), 53.7 (bs, 4H, CH (PPh2)) ppm.
Anal. Calcd (%) for C49H67N2P2Si4Nd: C 57.10, H 6.55 N 5.44.
Found: C 57.35, H 6.61 N 5.41.
In Situ Generation of Bis(iminophosphoranyl)methanide Alk-
oxide Neodymium Complex 6a or 6b. In the glovebox, 0.5 mL of a
0.08 M solution of 2-propanol in THF was added at room tem-
perature in a vial to a stirred solution of 4a (19.2 mg, 0.02 mmol)
or 4b (20.6 mg, 0.02 mmol) in THF (0.5 mL). This 0.02 M solu-
tion of putative 6a or 6b was used immediately in a polymeri-
zation experiment.
Polymerization of rac-Lactide Using 4a, 4b, 6b, or 6a. In a
typical experiment, in the glovebox, a freshly prepared solution
of 4a (9.6 mg, 0.01 mmol), 4b (10.3 mg, 0.01 mmol), 6a, or 6b in
THF (0.5 mL) was added via a syringe to a rapidly stirred
solution of rac-lactide (0.288 g, 2 mmol) in THF (1.5 mL) in a
vial, so as to make up a solution of overall concentration in LA
of 1 M. Aliquots of the reaction mixture were taken at various
times and precipitated in hexanes. Solvents were removed under
vacuum to leave the samples of lactide/polylactide as colorless
residues. Conversions were determined by the compared inte-
gration of the methine resonances of lactide (quintuplet δ =
4.92 ppm) and polylactide (multiplet δ = 5.00-5.30 ppm) in the
1H NMR spectrum in CDCl3.
Instruments and Measurements. Nuclear magnetic resonance
spectra were recorded on a Bruker Avance 300 spectrometer
operating at 300 MHz for 1H, 75.5 MHz for 13C, and 121.5 MHz
for 31P. 1H and 13C chemical shifts are reported in ppm relative
to Me4Si as external standard. 31P are relative to a 85% H3PO4
external reference. Coupling constants are expressed in hertz.
The following abbreviations are used: b, broad; s, singlet; d,
doublet; dd, doublet of doublets; t, triplet; m, multiplet; v,
virtual. Elemental analyses were performed by the “Service
ꢀ
d’Analyse de l’Universite de Dijon”, at Dijon, France. MAL-
DI-TOF mass spectra were determined at the EPSRC NMSSC,
Swansea, UK. Mass spectra were performed on a Micromass
Autospec Premier machine, using LSIMS (FAB) ion sources.
The MALDI-TOF conditions were using a dithranol matrix in
THF at a loading of 1:5 with sodium trifluoroacetate as the
cationizing agent. The polymer’s molecular number and poly-
dispersity index were determined from gel permeation chroma-
tography with multiangle laser light scattering (GPC-MALLS).
Two Polymer Laboratories mixed D columns were used in
series, with THF as the eluent, at a flow rate of 1 mL min-1
,
on a Polymer Laboratories PL GPC-50 instrument. The light-
scattering detector was a triple-angle detector (Dawn 8þ, Wyatt
Technology), and the data were analyzed using Astra V version
5.3.1.4. The refractive angle increment for polylactide in THF
was taken as 0.042 mL g-1 31
.
Synthesis of Bis(iminophosphoranyl)methanide, 1b. KHMDS
(38 mg, 0.19 mmol) was added to a suspension of 0.063 mmol of
the corresponding bis(aminophosphonium) chloride 1 in THF
(5 mL). The reaction mixture immediately became clear, taking
a pale yellow color. The potassium chloride precipitate was
removed by centrifugation and the supernatant solution dried
in vacuo to yield a pale yellow powder. 1b: yield 38 mg (0.062
mmol, 99%). 31P{1H} NMR (121.5 MHz, [D8]-THF, 20 °C): δ
10.2 (s, P) ppm. 1H NMR (300 MHz, [D8]-THF, 20 °C): δ 1.65
Polymerization of rac-Lactide Using 5a. In a typical experi-
ment (Table 2, entry 1), in the glovebox, a freshly prepared
solution of 5a (11.3 mg, 0.01 mmol) in THF (0.5 mL) was added
via a syringe to a rapidly stirred solution of rac-lactide (0.288 g,
2 mmol) in THF (1.25 mL), in a vial. After 5 min, 0.25 mL of a
0.08 M solution of 2-propanol in THF was added. The reaction
mixture was stirred vigorously. Aliquots of the reaction mixture
(30) (a) Cendrowski-Guillaume, S. M.; Le Gland, G.; Nierlich, M.;
Ephritikhine, M. Organometallics 2000, 19, 5654–5660. (b) Zhou, S.-L.;
Wang, S.-W.; Yang, G.-S.; Liu, X.-Y.; Sheng, E.-H.; Zhang, K.-H.; Cheng, L.;
Huang, Z.-X. Polyhedron 2003, 1019–1024. (c) Izod, K.; Liddle, S. T.;
Clegg, W. Inorg. Chem. 2004, 43, 214–218.
(31) Dorgan, J. R.; Janzen, J.; Knauss, D. M.; Hait, S. B.; Limoges,
B. R.; Hutchinson, M. H. J. Polym. Sci., Part B: Polym. Phys. 2005, 43,
3100–3111.

2900 Organometallics, Vol. 29, No. 13, 2010
Buchard et al.
were taken at various times and precipitated in hexanes. Sol-
vents were removed under vacuum to leave the samples of
lactide/polylactide as colorless residues. Conversions were de-
termined by the compared integration of the methine resonances
of lactide and polylactide in the H NMR spectrum in CDCl3
(see above).
Acknowledgment. We thank the EPSRC (EP/C544846/
1 and EP/C544838/1), the CNRS, and the Ecole Polytech-
nique, for financial support. rac-Lactide was donated by
Purac Plc. The EPSRC National Mass Spectrometry Ser-
vice Centre, Swansea, is thanked for collecting the MALDI-
ToF spectra. Dr. Florian Jaroschick is thanked for his help
with neodymium precursors and helpful discussions.
1
X-ray Crystallography. Data were collected on a Nonius
˚
Kappa CCD diffractometer using a Mo KR (λ = 0.71073 A)
X-ray source and a graphite monochromator. Experimental
details are described in Tables S1 and S2. CCDC-751373 (2a),
CCDC-751374 (2b), CCDC-751375 (3a), CCDC-751376 (4a),
and CCDC-751377 (4b) contain the supplementary crystallo-
graphic data for this paper. These data can be obtained
free of charge at www.ccdc.cam.ac.uk/conts/retrieving.html or
from the Cambridge Crystallographic Data Centre, 12 Union
Road, Cambridge CB2 1EZ, UK [fax: þ44-1223/336-033;
e-mail: deposit@ccdc.cam.ac.uk]. The crystal structures were
solved using SIR 9732 and Shelxl-97.33 Molecular views of X-ray
structures were made using ORTEP-334 for Windows.
Supporting Information Available: Crystallographic data for
2a, 2b, 3a, 4a, and 4b as CIF files; tables giving crystallographic
data for 2a, 2b, 3a, 4a, and 4b (including refinement details for
the X-ray structure determination); ORTEP plots (50% thermal
ellipsoids) of the molecular structures of complexes 2a and 2b
with relevant bond distances and angles; plot of PLA Mn vs %
rac-LA conversion for 4b; semilogarithmic plot of relative
rac-LA concentration vs time for 4b; MALDI-TOF mass spec-
trometry spectrum of PLA synthesized by ROP using 4a; plot of
Mn vs % rac-LA conversion for 4b/iPrOH system; plot of
rac-LA conversion vs % time for 4b/iPrOH system; semiloga-
rithmic plot of relative rac-LA concentration vs time for 4b/
iPrOH system; MALDI-TOF mass spectrometry spectra of
PLA synthesized by ROP using the 4b/iPrOH system. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(32) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G.;
Giacovazzo, C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.;
Spagna, R. J. Appl. Crystallogr. 1999, 32, 115–119.
(33) Sheldrick, G. M. SHELXS-96, Program for the Solution of Crystal
€
€
Structures; University of Gottingen: Gottingen (Germany), 1996.
(34) Farrugia, L. J. ORTEP-3. J. Appl. Crystallogr. 1997, 32, 565.