Theoretical studies on hyperpolarizabilities and UV-vis-IR spectra of a diamminecobalt(III) tetripeptide transition-metal complex

Abstract

The second-order polarizabilities and the UV-vis-IR spectra of a transition-metal complex Co(NH₃)₂(L-ala–gly–gly) have been studied by using the MP2 and TDHF methods. The complex has a maximum β component in the direction from the N(alanyl) group to the N(glycyl) groups. A transparent optical spectrum region from 0.55 to 5.5 µm was found, which offers potential applications as an optical material. The alkyl substitution of the glycyl group only slightly affected the β value and retained the IR transparent region but may cause the molecules to have a favorite packing fashion in the bulk crystal that leads to larger second-order nonlinear optical coefficients.

---

1 Introduction

As a result of a series of circular dichroism, X-ray diffraction and UV-vis spectrum studies, the manner in which metal ions form complexes with peptides is well understood.^1–8 The cobalt(III) tetripeptide, Co(NH₃)₂(L-ala–gly–gly) (Fig. 1) is a near-planar complex. The UV-vis spectrum shows that the absorption of the molecule tends to be weak in the near infrared region. The crystal structure is triclinic with non-centric space group P₁. These two factors exhibit a possibility that it may be a candidate as a IR nonlinear optical (NLO) material. However, to our knowledge, there are no experimental or theoretical studies concerning the NLO properties on this complex. In this paper we carried out an ab initio calculation to study the electron structure, UV-vis and IR spectra and the second-order polarizability β of the cobalt(III) tetripeptide complex.

Fig. 1 Molecular structure and orientation of Co(NH₃)₂(L-ala–gly–gly).

2 Method

The second-order polarizability β was calculated by using the finite-field approach and Møller–Plesset perturbation correction to the second-order (MP2) method.⁹ The time-dependent Hartree–Fock (TDHF) method is used to calculate the UV-vis spectrum.

The choice of the basis is as following: for metal Co, the Los Alamos ECP plus DZ basis set (Lanl2dz¹⁰) is used. For atom C, N, O and H, the Lanl2dz extended by diffuse and polarization function (p, d function for C, H and O; s, p for H) are used: C (P, 0.0311; D, 0.587), N (P, 0.0533; D, 0.736), O (P, 0.0673; D, 0.961), H (S, 0.0498; P, 0.356) (denoted as Lanl2dz+pd).¹¹

3 Results and discussion

3.1 The second-order polarizability

As shown in Fig. 1, the tripeptide is coordinated as a quadridentate chelate through the terminal NH₂, two peptide N⁻ and terminal CO₂⁻ groups. The tripeptide is almost planar. The other two NH₃ groups are perpendicular to this plane. We let the xy plane coincide with the tripeptide plane and used the crystal structure data such as bond lengths and angles in the calculations of the β values. Although the polarizability is a geometry-dependent property, using the X-ray structure is reasonable because the structure in the crystalline environment makes the calculated results be comparable to the experimental result. Owing to the presence of the L-alanyl group, the molecular symmetry is C₁ and there is no symmetric operation that may lead some of the β components vanish. For the sake of comparison, all the ten β components obtained with different basis sets and theoretical methods were listed in Table 1.

Table 1 The values of β components with various methods and basis sets (10⁻³⁰ esu)

Method/basis set	βxxx	βxxy	βxyy	βyyy	βxxz	βxyz	βyyz	βxzz	βyzz	βzzz
a Co: Lanl2dz, C, H, O, N: Lanl2dz+pd ECP polarization. b The result of MP2/Lanl2dz+pd, but the molecule has been rotated anticlockwise 38, 3 and 2° around the z, x and y axes, respectively.
HF/Lanl2dz	−1.86	0.58	−0.87	0.33	−0.00	0.05	0.01	−0.13	0.11	−0.04
HF/Lanl2dz+pd^a	−1.50	0.46	−0.77	0.14	0.06	0.05	−0.00	−0.09	−0.03	−0.05
MP2/Lanl2dz	−2.58	0.75	−1.42	0.75	−0.05	0.10	−0.04	−0.15	0.12	−0.06
MP2/Lanl2dz+pd	−2.27	0.65	−1.36	0.46	0.09	0.09	−0.05	−0.12	−0.01	−0.07
New orientation^b	−3.19	−0.00	−0.42	−1.26	−0.00	0.07	0.03	−0.09	−0.07	−0.07

---

It is well known that in order to obtain reliable β values, the electron correlation effect and the more diffuse basis function are critical. Table 1 shows that the β component values of MP2 calculations are larger than the correspondent HF result, which are the general phenomena in treating the organic molecules.^12–14 At both HF and MP2 levels, the implemented diffuse functions led most of β components to decrease, especially for the β_xxx, β_xxy, β_xyy, β_yyy components.

From Table 1, the β components with the z subscript, i.e.β_xxz, β_xyz, β_yyz, β_xzz, β_yzz and β_zzz were about ten times smaller than the components that with only x and y subscripts. This indicates that the hyperpolarizability was mainly originated from the quadratic chelate tripeptide. Also, it is well known that the β components are coordinate-dependent. If the coordinate was rotated, the magnitudes of the β components will vary so that we could try to find such a coordinate system in which a β component has a maximum. This value would be the largest one in all possible coordinate system. We found that if the molecular orientation was rotated anticlockwise in the following sequence: first rotated 38° around the z axis, then 3° around the x axis and then 2° around the y axis (Fig. 2), the β_xxx component had a maximum value of −3.19 × 10⁻³⁰ esu and became the largest component at MP2/Lanl2dz+pd level of theory. In this new orientation, β_yyy became the second largest component, while other components were very small. It was interesting to note that the new direction of the maximum β_xxx component (x direction in Fig. 2) is almost in the N(alanyl)–Co–N(glyclyl) direction. This indicated that the charge transfer along N(alanyl)–Co–N(glyclyl) dominated in the contribution of molecular hyperpolarizability.

Fig. 2 The structure orientation after special rotation that has a maximum β_xxx component.

3.2. UV-visible and IR spectra

The UV-visible of a molecule and the two molecules in the unit cell (supermolecular dimer) were presented in Fig. 3. The monomer has several intense absorptions around 170, 200, 235 and 327 nm. While for the dimer, the corresponding adsorption peaks were red-shifted to 194, 216, 235 and 336 nm, which were in better agreement with the experimental values of 190, 208, 270 and 356 nm. The molecular orbital compositions showed that the intensive adsorption band near 200 nm are corresponding to the electronic transfer from the tetripeptide to the Co atom. The energy gap between the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) was 547 nm, which agreed well to experimental value of 540 nm. These absorptions make the crystal show a red–brown color. However, the IR spectra (Fig. 4, scaled by a factor of 0.8929 ¹⁵) showed that this molecule had a transparent region from 1800 cm⁻¹ (5.5 µm) to 2800 cm⁻¹. In the region from 2800 to 3400 cm⁻¹ (2.95 µm) there was only the weak absorptions of C–H and N–H stretching. Within the region from 0.55 to 2.95 µm, there are no obvious absorptions in the UV-vis-IR spectra. These results indicated that this complex might be used as a near IR NLO material, especially to be used as a second harmonic generation (SHG) material within this favorite spectroscopic region.

Fig. 3 UV-visible spectra of a Co(NH₃)₂(L-ala–gly–gly) molecule and supermolecular dimer simulated by a Lorentzian line-shape modification with 10 nm half-width (line: monomer, dash: dimer).

Fig. 4 IR spectrum of Co(NH₃)₂(L-ala–gly–gly).

3.3. Alkyl substitution

Although Co(NH₃)₂(L-ala–gly–gly) had a relative large β values and a wide transparent region in the near IR region, the molecular spatial orientation in the bulk crystal was not favorite to fulfill its NLO potential. In a crystal unit cell, the two constituent molecules are nearly in opposite direction to each other in the tripeptide plane (Fig. 5). According to the oriented-gas model and its improved work,^16–18 the maximum component β_xxx would almost cancel each other in the unit cell, which led a small second-order NLO susceptibility tensor d. The calculated results of the dimer in the unit cell show the dominant component β_xxx has only a value −0.567 × 10⁻³⁰ esu at the MP2/Lanl2dz level. In order to avoid this unfavorable molecular packing pattern, the molecule could be modified by –(CH₂)_n– substitution in the glycyl (arrows shown in Fig. 5). The calculated results show that the β values only slightly change when the number of –(CH₂)– increasing from 1 to 10. For example, when n was 2, 6 and 10, the corresponding dominant components β_xxx were 3.32, 3.69 and 3.80 × 10⁻³⁰ esu, respectively. The substitution of alkyl would cause more absorption in the region from 3000 to 3200 cm⁻¹ due to the C–H stretching vibration, but the regions from 1800 to 2800 cm⁻¹ and 2.95 to 5.5 µm would retain transparency. These results strongly showed that the alkyl substitution in the glycyl would not only retain the favorite molecular NLO properties and near-IR transparent but also improve the molecular packing style for larger NLO coefficients.

Fig. 5 The packing in the unit cell of Co(NH₃)₂(L-ala–gly–gly).

4 Conclusion

Co(NH₃)₂(L-ala–gly–gly) has a relative large second-order polarizability β value (about twice larger than the value of the nitrobenzene¹⁹) and a perfectly transparent region from 0.55 to 5.5 µm. Although the molecular packing in the unit cell is almost inverted, which leads to the small second-order NLO susceptibility, the alkyl substitution in the glycyl moiety may change this situation because the alkyl substitution just slightly changes the β value while retaining the IR transparent region. These results show that with suitable chemical modification, a complex with Co(NH₃)₂(L-ala–gly–gly) molecules as building blocks may be a SHG material within the near IR region.

Acknowledgements

The authors are grateful for the financial support from the National Science Foundation of China NSFC (69978021, 20173064 and 90203017) and FJNSFC (E9910030).

References

1. E. J. Evans, C. J. Hawkins, J. Rodgers and M. R. Snow, Inorg. Chem., 1983, 22, 34.
2. R. B. Martin, J. M. Tsangaris and J. W. Chang, J. Am. Chem. Soc., 1968, 90, 821.
3. J. W. Chang and R. B. Martin, J. Phys. Chem., 1969, 73, 4277.
4. J. M. Tsangaris and R. B. Martin, J. Am. Chem. Soc., 1970, 72, 4255.
5. J. M. Tsangaris, J. W. Chang and R. B. Martin, J. Am. Chem. Soc., 1969, 91, 726.
6. E. W. Wilson and R. B. Martin, Inorg. Chem., 1970, 9, 528.
7. T. P. Pitner, E. W. Wilson and R. B. Martin, Inorg. Chem., 1972, 11, 738.
8. E. J. Evans, J. E. Grice, C. J. Hawkins and M. R. Heard, Inorg. Chem., 1980, 19, 3496.
9. C. Møller and M. S. Plesset, Phys. Rev., 1934, 46, 618.
10. P. J. Hay and W. R. Wadt, J. Chem. Phys., 1985, 82, 270.
11. C. E. Check, T. O. Faust, J. M. Bailey, B. J. Wright, T. M. Gilbert and L. S. Sunderlin, J. Phys. Chem. A, 2001, 105, 8111.
12. B. Champagne and B. Kirtman, in Handbook of Advanced Electronic and Photonic Materials and Devices, ed. H. S. Nalwa, Academic Press, San Diego, CA, 2000, p. 63.
13. C. Lin and K. Wu, Chem. Phys. Lett., 2000, 321, 83.
14. C. Lin, K. Wu, R. Sa, C. Mang, P. Liu and B. Zhuang, Chem. Phys. Lett., 2002, 363, 343.
15. J. B. Foresman and A. Frisch, in Exploring Chemistry with Electronic Structure Methods, Gaussian, Inc., Pittsburgh, PA, 1993, p. 78.
16. D. S. Chemla and J. Zyss, in Nonlinear Optical Properties of Organic Molecules and Crystals, ed. D. S. Chemla and J. Zyss, Orlando, FL, 1987, p. 99.
17. T. Hamada, J. Phys. Chem. A, 1996, 100, 19344.
18. F. Castet and B. Champagne, J. Phys. Chem. A, 2001, 105, 1366.
19. J. D. Singer and A. F. Garito, J. Chem. Phys., 1981, 75, 3572.

---