Blue-shifted lithium bonds

Abstract

A number of lithium bonding systems (X–Li⋯Y) have been found in which the X–Li bond is shortened due to the lithium bond formation.

---

A characteristic feature of H⋯Y hydrogen bonding in an X–H⋯Y system is X–H bond elongation with a concomitant red shift of the X–H stretching frequency.¹ The latter, readily detected in the IR spectra, is widely regarded as the “signature of hydrogen bonding”. However, some recent studies have reported the existence of blue-shifted hydrogen bonds in which hydrogen bond formation leads to X–H bond shortening and to a blue shift of the X–H stretching frequency.^2,3

Initially blue-shifted hydrogen bonds were reported for C–H bonds only.² Recent studies showed that N–H, O–H, Si–H and P–H bonds could also form blue-shifted hydrogen bonds.⁴ A comprehensive theory for the blue shift has been proposed.⁵ According to it, there is a balance between the X–H elongation effect due to orbital interactions and the X–H contraction effect due to Pauli and nuclei repulsions. If the former effect wins, a red shift will occur. Otherwise, a blue shift will take place.

An interesting interaction analogous to hydrogen bonding is lithium bonding.⁶ Its existence was theoretically predicted by Kollman et al. in 1970.⁷ Experimental evidence for lithium bonding, i.e. a large red shift of the X–Li stretching frequency in some X–Li⋯Y systems, was provided by Pimentel et al. in 1975.⁸ To date lithium bonding has been identified in a variety of systems and the concept of lithium bonding has become important in many fields. However, it remains unknown whether there is any blue-shifted lithium bond.

Herein we wish to report our study on blue-shifted lithium bonds. We will focus on two lithium bond donors, F₃C–Li and F₃Si–Li, because F₃C–H and F₃Si–H have been found to be good hydrogen bond donors for the blue shift. For the lithium bond acceptors, we choose NH₃, H₂O, HF, N₂, He, Ne, Ar, F₂, Cl₂, CF₄, and C₆H₆. For each lithium bonding system, we also compare it with the corresponding hydrogen bond.

Our calculations are performed at the MP2/6-311++G(d,p) level for the complexes of NH₃, H₂O, HF, N₂, He, Ne, Ar, F₂, Cl₂.⁹ For the complexes of CF₄ and C₆H₆ we use the MP2/6-31+G(d) method. Both the zero point energy (ZPE) and basis set superposition error (BSSE)¹⁰ corrections are considered in the calculations. The results are listed in Table 1.

Table 1 Bond lengths, stretching frequencies, and interaction energies of X–H⋯Y and X–Li⋯Y from MP2/6-311++G(d,p) or MP2/6-31+G(d) calculations

X–H or X–Li	Y	d ^a	Δd^b	ν ^c	Δν^d	ΔE^e
a X–H or X–Li bond length (Å). b Change of X–H or X–Li bond length due to the complexation (Å). c X–H or X–Li stretching frequency (cm^−1). d Change of X–H or X–Li stretching frequency due to the complexation (cm^−1). e Binding energy between Y and X–H or X–Li (kJ mol^−1). This energy is corrected with ZPE and BSSE. f MP2/6-31+G(d) results. g Optimization on these complexes fails. h Frequency calculation on this particular complex fails due to the large size and therefore, the binding energy of this complex has not been corrected with ZPE.
F3C–H	—	1.0877 (1.0881^f)	—	3223.3 (3250.0^f)	—	—
	NH3	1.0875	−0.0002	3221.2	−2.1	−12.2
	H2O	1.0854	−0.0023	3260.1	+36.8	−9.8
	HF	1.0851	−0.0026	3267.3	+44.0	−5.6
	N2	1.0865	−0.0012	3243.2	+19.9	−1.6
	He	1.0876	−0.0001	3224.9	+1.6	0.2
	Ne	1.0876	−0.0001	3224.1	+0.8	0.8
	Ar	1.0871	−0.0006	3235.2	+11.9	0.8
	F2	1.0862	−0.0015	3247.7	+24.4	0.6
	Cl2	1.0856	−0.0021	3257.6	+34.3	0.8
	CF4	1.0868^f	−0.0013	3271.9^f	+21.9	−0.6
	C6H6	1.0840^f	−0.0041	3320.7^f	+70.7	−7.2
F3C–Li	—	2.0218 (2.0378^f)	—	501.4 (486.6^f)	—	—
	NH3	2.0434	+0.0216	601.1	+99.7	−85.9
	H2O	2.0391	+0.0173	607.0	+105.6	−73.3
	HF^g	—	—	—	—	—
	N2	2.0278	+0.0060	545.1	+43.7	−21.4
	He	2.0218	+0.0000	510.7	+9.3	−0.7
	Ne	2.0208	−0.0010	519.6	+18.2	−1.2
	Ar	2.0205	−0.0013	538.7	+37.3	−4.3
	F2	2.0199	−0.0019	539.1	+37.7	−5.9
	Cl2	2.0164	−0.0054	567.5	+66.1	−8.6
	CF4	2.0306^f	−0.0072	528.2^f	+41.6	−12.8
	C6H6	2.0211^f	−0.0167	568.2^f	+81.6	−52.6
F3Si–H	—	1.4488	—	2456.8	—	—
	NH3	1.4479	−0.0009	2456.7	−0.1	−3.6
	H2O^g	—	—	—	—	—
	HF	1.4467	−0.0021	2476.9	+20.1	−2.5
	N2	1.4493	+0.0005	2457.6	+0.8	−0.1
	He	1.4492	+0.0004	2455.5	−1.3	0.2
	Ne	1.4485	−0.0003	2462.3	−5.5	0.8
	Ar	1.4487	−0.0001	2461.4	−4.6	0.6
	F2	1.4481	−0.0007	2468.9	−12.1	0.3
	Cl2^g	—	—	—	—	—
	CF4^g	—	—	—	—	—
	C6H6^g	—	—	—	—	—
F3Si–Li	—	2.4822 (2.5080^f)	—	487.2 (474.1^f)	—	—
	NH3	2.5039	+0.0217	594.3	+107.1	−90.0
	H2O	2.4990	+0.0168	603.2	+116.0	−76.9
	HF^g	—	—	—	—	—
	N2	2.4845	+0.0023	528.4	+41.2	−22.3
	He	2.4821	−0.0001	494.4	+7.2	−0.5
	Ne	2.4795	−0.0027	501.2	+14.0	−1.3
	Ar	2.4793	−0.0029	516.8	+29.6	−4.8
	F2	2.4786	−0.0036	518.3	+31.1	−6.6
	Cl2	2.4728	−0.0094	551.7	+64.5	−9.7
	CF4	2.5041^f	−0.0039	517.6^f	+43.5	−14.3
	C6H6	2.4893^f	−0.0187	—	—	−68.7^h

---

It is found that for most of the F₃C–H⋯Y and F₃Si–H⋯Y systems the C–H or Si–H bond is shortened due to the formation of the hydrogen bond. This bond contraction leads to a blue shift of the C–H or Si–H stretching frequency.^2–5 However, it is also found that for the lithium bonds an increase in X–Li bond length may cause a blue shift of the X–Li stretching frequency. For example, in F₃C–Li⋯NH₃ the C–Li bond length increases substantially by 0.0216 Å but the C–Li stretching frequency also increases significantly by 99.7 cm⁻¹.

The reason for the inconsistency between X–Li bond length and stretching frequency in the lithium bonds is probably that the X–Li stretching frequency is not much higher than that of other bonds (e.g. C–F). Hence, the coupling between X–Li vibration and the vibration of other bonds can be very strong. The “observed” X–Li frequency does not completely belong to the X–Li vibration and the “observed” blue shift of the X–Li frequency is not fully caused by the change of bonding between X and Li.¹¹

Since we are more interested in the effects of lithium bonding on the properties of the X–Li bond itself, a blue shift of the “observed” X–Li frequency caused by the vibrations of other bonds is not an interesting phenomenon to the present study. Compared to the “observed” X–Li frequency, the X–Li bond length is a property completely belonging to the X–Li bond itself. Therefore, in the following we decide to focus on the blue-shifted lithium bonds where the X–Li bond is shortened due to the lithium bond formation.

It is found that the X–Li stretching frequency is blue shifted in all the F₃C–Li⋯Y complexes. Nevertheless, NH₃, H₂O, and N₂ lead to elongation of the C–Li bond, whereas Ne, Ar, F₂, Cl₂, CF₄, and C₆H₆ lead to contraction of the C–Li bond. Therefore, the blue-shifted and shortened lithium bonds do exist. It is worthy of note that the variation of the C–Li bond length in lithium bonding is much more dramatic than that of the C–H bond length in hydrogen bonding. In F₃C–Li⋯NH₃ the C–Li bond is elongated by 0.0216 Å, whereas in F₃C–Li⋯C₆H₆ the C–Li bond is shortened by 0.0167 Å (See Fig. 1).

Fig. 1 F₃C–Li, F₃Si–Li, and their complexes with C₆H₆.

For F₃Si–Li, NH₃, H₂O, and N₂ lead to elongation of the Si–Li bond whereas He, Ne, Ar, F₂, Cl₂, CF₄, C₆H₆ lead to contraction of the Si–Li bond. The largest contraction is seen for F₃Si–Li⋯C₆H₆ (−0.0187 Å). Thus the blue shifted and shortened lithium bonds also exist in some F₃Si–Li complexes.

In order to understand the mechanism of the shortened lithium bonds, we studied F₃C–Li⋯Ne (shortened) and F₃C–Li⋯OH₂ (elongated) in detail. By fixing the C⋯Y distances in F₃C–Li⋯Y and by optimizing the remaining coordinates of the complexes, we obtained curves of the interaction energy (ΔE, not corrected with BSSE) and the variation of the C–Li bond length (Δd) as functions of the C⋯Y distance (Fig. 2).

Fig. 2 Interaction energy (ΔE) and variation of C–Li bond length (Δd) as a function of the distance between F₃C–Li and Y: (a) Y = Ne, (b) Y = OH₂ (equilibrium C⋯Y distances indicated by the line labeled “eq”).

The potential energy curves of the two complexes are very similar in shape. At long distance, ΔE becomes more negative as the C⋯Y distance decreases. This behavior is clearly caused by the electrostatic interaction between F₃C–Li and Y. On the other hand, at short distance ΔE becomes less negative as the C⋯Y distance decreases. This behavior is undoubtedly due to the Pauli and nuclei-nuclei repulsions between F₃C–Li and Y.

The curves for the variation of C–Li bond length are also very similar in shape for the two complexes. At long distance, the C–Li bond is elongated for both F₃C–Li⋯Ne and F₃C–Li⋯OH₂. This elongation can only be explained by either the electrostatic attractions or the orbital interactions (e.g. charge transfer). On the other hand, the C–Li bond is shortened for both F₃C–Li⋯Ne and F₃C–Li⋯OH₂ at short C⋯Y distance. This contraction can only be explained as a result of Pauli and nuclei–nuclei repulsions.

The equilibrium position for F₃C–Li⋯Ne is in the contraction region of the curve so that F₃C–H⋯Ne has a shortened C–Li bond. In comparison, the equilibrium position for F₃C–Li⋯OH₂ is in the elongation region so that F₃C–Li⋯OH₂ has an elongated C–Li bond. Thus the difference between shortened and elongated lithium bonds is very simple. For the shortened ones, the bond shortening is greater than bond lengthening when the energy reaches the minimum. On the other hand, for the elongated lithium bonds, there is an additional bond lengthening due to orbital interactions that is not overcome by the modest bond compression resulting from the repulsive interactions.

The above analyses suggest that the mechanism for the blue-shifted and shortened lithium bonds should be the same as that for the blue-shifted hydrogen bonds.⁵ There is a balance between the X–Li elongation effect due to orbital interactions and the X–Li contraction effect due to Pauli and nuclei repulsions. If the former effect wins, the X–Li bond will elongate. Otherwise, the X–Li bond will contract.

We thank NSFC for the financial support.

Notes and references

1. S. Scheiner, Hydrogen bonding, Oxford University Press, New York, 1997.
2. Review: P. Hobza and Z. Havlas, Chem. Rev., 2000, 100, 4253.
3. Recent studies: W. Qian and S. Krimm, J. Phys. Chem. A, 2002, 106, 6628; E. S. Kryachko and T. Zeegers-Huyskens, J. Phys. Chem. A, 2002, 106, 6832; J.-M. Fan, K.-C. Zhang, L. Liu, X.-S. Li and Q.-X. Guo, Chin. J. Chem., 2002, 20, 247; S. N. Delanoye, W. A. Herrebout and B. J. van der Veken, J. Am. Chem. Soc., 2002, 124, 11854; J.-M. Fan, L. Liu and Q.-X. Guo, Chem. Phys. Lett., 2002, 365, 464; E. Mrazkova and P. Hobza, J. Phys. Chem. A, 2003, 107, 1032; J. Li, D.-Q. Xie and G.-S. Yan, Sci. China B, 2003, 33, 21.
4. Y. Fang, J.-M. Fan, L. Liu, X.-S. Li and Q.-X. Guo, Chem. Lett., 2002, 116; P. Hobza, Int. J. Quantum Chem., 2002, 90, 1071; P. Hobza and V. Spirko, Phys. Chem. Chem. Phys., 2003, 5, 1290; I. V. Alabugin, M. Manoharan, S. Peabody and F. Weinhold, J. Am. Chem. Soc., 2003, 125, 5973; J.-T. Wang, Y. Feng, L. Liu, X.-S. Li and Q.-X. Guo, Chem. Lett., 2003, 32, 746.
5. X. Li, L. Liu and H. B. Schlegel, J. Am. Chem. Soc., 2002, 124, 9639.
6. Recent studies: S. Scheiner, in Lithium chemistry: A theoretical and experimental overview, eds. A. M. Sapse and P. von R. Schleyer, John Wiley & Sons Inc., New York, 1995.
7. P. A. Kollman, J. F. Liebman and L. C. Allen, J. Am. Chem. Soc., 1970, 92, 1142.
8. B. S. Ault and G. C. Pimentel, J. Phys. Chem., 1975, 79, 621.
9. All the calculations were done using G98 programs: M. J. Frisch, G. W. Trucks, H. B. Schlegelet al., GAUSSIAN 98 (Revision A.7), Gaussian, Inc., Pittsburgh, PA, 1998.
10. S. F. Boys and F. Bernardi, Mol. Phys., 1970, 19, 553.
11. Compared to the lithium cases, the X–H stretching frequency is much higher than that of any other bond so that the coupling between X–H vibration and the vibration of other bonds is usually very weak. Therefore, in the hydrogen bonding systems the X–H bond length is usually consistent with the X–H stretching frequency.

---

Footnote

† Current address: Department of Chemistry, Columbia University, New York, NY 10027, USA.

---