Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2008 Dec 8.
Published in final edited form as: Org Lett. 2007 Jul 21;9(17):3359–3362. doi: 10.1021/ol0713234

Impact of Na- and K-C π-Interactions on the Structure and Binding of M3(sol)n(BINOLate)3Ln Catalysts

Alfred J Wooten 1, Patrick J Carroll 1, Patrick J Walsh 1,*
PMCID: PMC2597426  NIHMSID: NIHMS63527  PMID: 17658838

Abstract

Shibasaki’s heterobimetallic complexes M3(THF)n(BINOLate)3Ln [M = Li, Na, K, Ln = lanthanide(III)] are among the most successful asymmetric Lewis acid catalysts. Why does M3(THF)n(BINOLate)3Ln readily bind substrates when M = Li but not when M = Na or K? Structural studies herein indicate Na- and K-C cation-π interactions and alkali metal radius may be more important than even lanthanide radius. Also reported is a novel polymeric [K3(THF)2(BINOLate)3Yb]n structure that provides the first evidence of interactions between M3(THF)n(BINOLate)3Ln complexes.

Shibasaki’s M3(THF)n(BINOLate)3Ln complexes (Figure 1) are among the most effective asymmetric Lewis acid catalysts known, exhibiting high enantioselectivities over a broad range of reactions.1–3 Understanding how these heterobimetallic catalysts work, however, has proven challenging.4–9 In particular, seemingly subtle alterations in the catalyst composition result in dramatic changes in selectivity. For example, in the nitro aldol reaction with M3(THF)n(BINOLate)3Ln, 94% ee was obtained when M = Li and 2% ee when M = Na. In contrast, the asymmetric Michael reaction gave 92% ee when M = Na and 29% ee for M = Li.10 The first step in unraveling the factors that are responsible for these striking differences is understanding the impact of the alkali metal on substrate binding to the lanthanide centers.

Figure 1.

Figure 1

Open in a new tab

Shibasaki’s M3(THF)n(BINOLate)3Ln catalysts.

Reported herein are solution and solid state studies of M3(sol)n(BINOLate)3Ln complexes that illuminate dramatic differences in Ln binding ability when M = Li vs. Na and K. Also disclosed is an unprecedented helical polymer, [K3(THF)2(BINOLate)3Yb]n

We recently demonstrated that DMF reversibly binds to paramagnetic lanthanides in Li3(THF)n(BINOLate)3Ln (Ln = Eu, Pr), exhibiting > 2 ppm lanthanide induced shift (LIS) in the formyl C–H resonance in the 1H NMR spectrum.8 Salvadori, on the other hand, reported7 that Na3(THF)6(BINOLate)3Yb does not bind water in solution or the solid state, which was attributed to the small ionic radius of Yb (La = 1.17, Eu = 1.09, Yb = 1.01). This dichotomy prompted us to examine binding of the lithium analog, Li3(THF)n(BINOLate)3Yb, with DMF. In the presence of Li3(THF)n(BINOLate)3Yb, the formyl C–H shifted over 4 ppm, consistent with binding to Yb. Furthermore, crystallization of Li3(THF)n(BINOLate)3Yb from pyridine yielded 7-coordinate Li3(py)5(BINOLate)3Yb•py, the ORTEP of which is shown in Figure 2. We next examined binding of DMF to Na3(THF)6(BINOLate)3Ln [Ln = Yb, Eu]11–13 and K3(THF)6(BINOLate)3Yb14,7 under the same conditions. Surprisingly, no LISs (>0.1 ppm) were observed, indicating that binding to the lanthanide in the Na and K analogs is much less favorable than in the Li series.

Figure 2.

Figure 2

Open in a new tab

Structure of Li3(py)5(BINOLate)3Yb(py).

To probe the influence of Li vs. Na on substrate binding at the lanthanide, Na3(THF)6(BINOLate)3La (Figure 3) and isostructural Na3(THF)6(BINOLate)3Eu (Figure S1) were crystallographically characterized. In contrast to 7- and 8-coordinate Li3(THF)4(BINOLate)3Ln•THF (Ln=Eu, La) and Li3(py)5(BINOLate)3La•py2,8 Na3(THF)6(BINOLate)3Ln are only 6-coordinate.

Figure 3.

Figure 3

Open in a new tab

Structure of Na3(THF)6(BINOLate)3La illustrating the Na-C π-interactions. The triangular face is drawn in black. Two views down the C3-axes with the 3-H protons in blue are shown on the right.

When comparing M3(THF)6(BINOLate)3Ln structures, we found three interdependent factors to be important. The first and most informative aspect of these structures is the presence of Na-C cation-π interactions.15,16 Positioned directly above the sodium atoms are BINOLate naphthyl rings, three carbons of which are well within the expected range for Na-C cation-π interactions (Table 1, entries 1 and 2).17 Our analysis of the Yb homologue12,7 reveals similar interactions (entry 3). Particularly noteworthy is that the Na-C interactions (2.91–3.23 Å) in the La, Eu, and Yb complexes are identical and are independent of the lanthanide radius.

Table 1.

Structural Data for M3(THF)6(BINOLate)3Ln complexes.

entry M Ln M3-Ln (Å) M-C π-dist (Å) triangular face (Å)
1 Na3 La 1.111(7) 2.94–3.23 2.84 5.02
2 Na3 Eu 0.950(1) 2.91–3.20 2.78 4.76
3 Na3 Yb 0.762(7) 2.92–3.20 2.78 4.42
4 Na3 La•OH2 1.130(2) 2.96–3.24 2.83 5.05
5 Li3 La•THF 0.767(4) 2.71–4.82 3.36a 5.35 a
6 Li3 Eu•THF 0.737(3) 2.65–4.77 3.28 a 5.30 a
7 Li3 Yb•py 0.479(2) 2.73–4.93 3.58 a 5.10 a
8 Li3 La•py2 0.042(2) 2.79–5.10 5.02 a 4.93 a
9 K3 Yb 0.354(6) 3.22–3.62 3.29 a 3.69 a
Open in a new tab
[a]

Average value

The second factor is the Ln displacement from the Na3 plane (Table 1) of 1.111(7) Å (La) and 0.950(1) Å (Eu). We calculated a similar displacement for known Na3(THF)6(BINOLate)3Yb [0.762(7) Å].12 Although the displacement had been noted, its origin was unknown.12,7 The displacement distances parallel the Ln radii (La > Eu > Yb). Salvadori observed that Na3(THF)6(BINOLate)3Yb exhibits D3 symmetry by NMR down to −100 °C7 suggesting that in solution the Yb sits in the Na3 plane. On binding substrate to form a 7-coordinate adduct, however, the lanthanide must be displaced from the M3 plane (Table 1, entries 4–7).

The Na-C cation-π interactions cause the Ln-Na3 displacement and result in the inequivalent C3-symmetric faces in Na3(THF)6(BINOLate)3Ln in the solid state (Figure 3, right). To quantify this inequivalency, we define the third structural factor, the triangular faces, as the distances between the three 3-H protons of the BINOLate ligands (Figure 3, blue hydrogens). The 3-H protons interact most directly with incoming or Ln-bound substrates. The equilateral triangular faces of Na3(THF)6(BINOLate)3Ln highlight these differences (La: 2.84 vs. 5.02 Å and Eu: 2.78 vs. 4.76 Å, Table 1). It is also telling that the triangular faces of Na3(THF)6(BINOLate)3La and Na3(THF)6(BINOLate)3La•OH2 are almost identical,12 as are the La-Na3 displacements and Na-C distances (Table 1, entry 1 vs. 4). Thus, almost no structural change is observed in the solid state between Na3(THF)6(BINOLate)3La and Na3(THF)6(BINOLate)3La•OH2. It follows that Na3(THF)6(BINOLate)3Yb, with the much smaller Yb radius, would require significant structural reorganization to bind water or larger Lewis bases (La = 1.17 vs. Yb = 1.01 Å radii).7

Comparison of the lanthanide displacements in Na3(THF)6(BINOLate)3La and 7-coordinate Li3(THF)4(BINOLate)3La•THF is also informative. Despite the greater displacement of the La in 6-coordinate Na3(THF)6(BINOLate)3La than in 7-coordinate Li3(THF)4(BINOLate)3La•THF, the triangular face (or binding pocket) of the Na analog is smaller. Butting of the sodium atoms with the naphthyl π systems positions the BINOLate 3-H’s into the substrate-binding site. In contrast, the smaller lithium does not exhibit Li-C π-interactions17 in Li3(THF)4(BINOLate)3Ln complexes, allowing greater distortion of the (BINOLate)3Ln core to accommodate substrates larger than water.

We next desired to explore how the larger potassium would impact the structure of M3(THF)6(BINOLate)3Ln complexes. Thus, K3(THF)6(BINOLate)3Yb was crystallized from pyridine. Like the sodium derivatives, 6-coordinate K3(py)6(BINOLate)3Yb contains K-C cation-π interactions in the solid state17 (3.22–3.62 Å) as well as an unusual η3-pyridine with only K-C π-interactions (Figure 4). The larger potassium radius results in longer K-C π-distances, a smaller Yb-K3 displacement [0.354(6) Å, entry 9 in Table 1], and a smaller difference in the triangular faces [3.16, 3.35, 3.38 vs. 3.55, 3.56, 3.96 Å]. Potassium π-interactions were also observed in coordinatively saturated K3(toluene)6(BINOLate)3In (K-C 3.35–3.42 Å).18 Following the logic outlined above for the sodium analogs, we hypothesize that the propensity of M3(THF)n(BINOLate)3Ln complexes to bind Lewis bases at the lanthanide center is Li >> Na > K due to the alkali metal ionic radii.

Figure 4.

Figure 4

Open in a new tab

One of 6 independent molecules of K3(py)6(BINOLate)3Yb.

Similarly, crystallization of K3(THF)6(BINOLate)3Yb from THF/pentane initially provided poor quality block-like crystals, the structure of which was most consistent with 6-coordinate K3(THF)6(BINOLate)3Yb.19 On standing, solutions of K3(THF)6(BINOLate)3Yb formed X-ray quality needles. Surprisingly, an unprecedented helical polymer, [K3(THF)2(BINOLate)3Yb]n, was observed (Figure 5).

Figure 5.

Figure 5

Open in a new tab

Structure of the [K3(THF)2(BINOLate)3Yb]n helical polymer.

Polymeric [K3(THF)2(BINOLate)3Yb]n exhibits intramonomer (3.27–3.66 Å) and intermonomer (3.02–3.37 Å) K-C π-interactions (Figure 6). K1 binds two BINOLate oxygens, two THF’s, and exhibits intramonomer π-interactions. The other two potassium atoms are disordered over three positions with occupancy 40.5, 45.5 and 14%, the later of which is not discussed. K2 (40.5% occupancy) binds two BINOLate oxygens, as usual, and contains both intra- and intermonomer cation-π interactions. Interestingly, K2’ (45.5% occupancy) binds one BINOLate oxygen in two different monomeric units. The disordered potassiums participate in intra- and intermonomer cation-π interactions. In comparison of K1, which binds four oxygens, to K2 and K2’, which each bind two oxygens but have more contacts with the naphthyl carbons, it is useful to recall that the cation-π interaction, K+•••(benzene), and the binding of water, K+•OH2, are 19 and 18 kcal/mol, respectively, in the gas phase.15 The extensive K-C cation-π interactions in the monomeric and polymeric structures attest to the importance of these interactions.

Figure 6.

Figure 6

Open in a new tab

Structure of the monomeric unit in [K3(THF)2(BINOLate)3Yb]n illustrating the disordered K atoms.

In summary, this study provides insight into the structures and binding preferences of one of the most successful classes of asymmetric catalysts, Shibasaki’s M3(THF)n(BINOLate)3Ln complexes. The solid state structure of Li3(py)5(BINOLate)3Yb•py and solution binding of DMF by Li3(THF)n(BINOLate)3Yb stand in sharp contrast to the sodium and potassium analogs, M3(THF)6(BINOLate)3Yb, which exhibit little or no binding of DMF or water. These results are the first evidence that lanthanide size is not the primary determinant for substrate binding to the lanthanide.

We have also characterized the unique helical polymer [K3(THF)2(BINOLate)3Yb]n which provides the first evidence of interactions between M3(THF)n(BINOLate)3Ln complexes. The additional cation-π interactions in this species compensate for the reduced number of K–O bonds.

Finally, our data supports earlier proposals that reactions promoted by these bifunctional catalysts (M = Na, K) likely involve a BINOLate oxygen first acting as a Brønsted base and accepting a proton from the substrate.7,10 The protonated BINOLate oxygen likely detaches from the lanthanide to open a coordination site before substrate binding at the Ln-center can take place.

Supplementary Material

1File001
2File002
3File003
4File004
5File005
6File007. Supporting Information Available.

Procedures, full characterization of new compounds, and all CIF data (PDF) are available free of charge via the internet at http://pubs.acs.org.

Acknowledgment

This work was supported by the NIH (National Institute of General Medical Sciences, GM58101).

References

  • 1.Shibasaki M, Sasai H, Arai T. Angew. Chem., Int. Ed. 1997;36:1236–1256. [Google Scholar]
  • 2.Shibasaki M, Yoshikawa N. Chem. Rev. 2002;102:2187–2219. doi: 10.1021/cr010297z. [DOI] [PubMed] [Google Scholar]
  • 3.Aspinall HC. Chem. Rev. 2002;102:1807–1850. doi: 10.1021/cr010288q. [DOI] [PubMed] [Google Scholar]
  • 4.Yamagiwa N, Qiin H, Matsunaga M, Shibasaki M. J. Am. Chem. Soc. 2005;127:13419–13427. doi: 10.1021/ja054066b. [DOI] [PubMed] [Google Scholar]
  • 5.Horiuchi T, Gananadesikan V, Ohshima T, Masu H, Katagiri K, Sei Y, Yamaguchi K, Shibasaki M. Chem. Eur. J. 2005;11:5195–5204. doi: 10.1002/chem.200500442. [DOI] [PubMed] [Google Scholar]
  • 6.Yamagiwa N, Matsunaga S, Shibasaki M. Angew. Chem., Int. Ed. 2004;43:4493–4497. doi: 10.1002/anie.200454202. [DOI] [PubMed] [Google Scholar]
  • 7.Di Bari L, Lelli M, Pintacuda G, Pescitelli G, Marchetti F, Salvadori P. J. Am. Chem. Soc. 2003;125:5549–5558. doi: 10.1021/ja0297640. [DOI] [PubMed] [Google Scholar]
  • 8.Wooten AJ, Carroll PJ, Walsh PJ. Angew. Chem. Int. Ed. 2006;45:2549–2552. doi: 10.1002/anie.200504275. [DOI] [PubMed] [Google Scholar]
  • 9.Yamagiwa N, Tian J, Matsunaga S, Shibasaki M. J. Am. Chem. Soc. 2005;127:3413–3422. doi: 10.1021/ja042887v. [DOI] [PubMed] [Google Scholar]
  • 10.Shibasaki M, Sasai H, Arai T, Iida T. Pure Appl. Chem. 1998;70:1027–1034. [Google Scholar]
  • 11.Sasai H, Suzuki T, Itoh N, Tanaka K, Date T, Okamura K, Shibasaki M. J. Am. Chem. Soc. 1993;115:10372–10373. [Google Scholar]
  • 12.Aspinall HC, Bickley JF, Dwyer JLM, Greeves N, Kelly RV, Steiner A. Organometallics. 2000;19:5416–5423. [Google Scholar]
  • 13.Sasai H, Arai T, Satow Y, Houk KN, Shibasaki M. J. Am. Chem. Soc. 1995;117:6194–6198. [Google Scholar]
  • 14.Gröger H, Saida Y, Sasai H, Yamaguchi K, Martens J, Shibasaki M. J. Am. Chem. Soc. 1998;120:3089–3103. [Google Scholar]
  • 15.Ma JC, Dougherty DA. Chem. Rev. 1997;97:1303–1324. doi: 10.1021/cr9603744. [DOI] [PubMed] [Google Scholar]
  • 16.Meyer EA, Castellano RK, Diederich F. Angew. Chem., Int. Ed. 2003;42:1210–1250. doi: 10.1002/anie.200390319. [DOI] [PubMed] [Google Scholar]
  • 17.Weiss E. Angew. Chem., Int. Ed.Engl. 1993;32:1501–1523. [Google Scholar]
  • 18.Chitsaz S, Neumüller B. Organometallics. 2001;20:2338–2343. [Google Scholar]
  • 19.Wooten AJ, Salvi L, Carroll PJ, Walsh PJ. Adv. Synth. Catal. 2007;349:561–565. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1File001
NIHMS63527-supplement-1File001.cif (30.3KB, cif)
2File002
NIHMS63527-supplement-2File002.cif (37.2KB, cif)
3File003
NIHMS63527-supplement-3File003.cif (222.3KB, cif)
4File004
NIHMS63527-supplement-4File004.cif (16.8KB, cif)
5File005
NIHMS63527-supplement-5File005.cif (17.6KB, cif)
6File007. Supporting Information Available.

Procedures, full characterization of new compounds, and all CIF data (PDF) are available free of charge via the internet at http://pubs.acs.org.

NIHMS63527-supplement-6File007.pdf (724.6KB, pdf)

RESOURCES