Spectroscopic Study of the Interaction of Nd3+ with Amino Acids: Phenomenological 4f-4f Intensity Parameters

Article

Spectroscopic Study of the Interaction of Nd³⁺ with Amino Acids: Phenomenological 4f-4f Intensity Parameters

Soraya Jericó^a, Célia R. Carubelli^a, Ana M.G. Massabni^a, Elizabeth B.Stucchi^a, Sergio R. de A. Leite^a*, and Oscar Malta^b

^a

Instituto de Química, Universidade Estadual Paulista C.P. 355, 14801-970 Araraquara – SP, Brazil

^bDepartamento de Química Fundamental, Universidade Federal de Pernambuco, 50670-901 Recife – PE, Brazil

Received: June 1, 1998

Estudamos o comportamento dos parâmetros fenomenológicos de intensidade das transições 4f-4f em compostos de Nd³⁺ com glicina, ácido L-aspártico, ácido L-glutâmico, L-histidina, ácido DL-málico e Aspartame®em solução aquosa como função dos valoresde pK e das cargas parciais sobre os átomos de oxigênio dos grupos carboxilatos dessas moléculas. Os resultados são discutidos e interpretados qualitativamente em termos dos mecanismos das intensidades 4f-4f por dipolo elétrico forçado e acoplamento dinâmico, indicando como dominante o mecanismo de dipolo elétrico forçado.

We have studied the bevahior of the phenomenological 4f-4f intensity parameters in compounds of the Nd³⁺ ion with glycine, L-aspartic acid, L-glutamic acid, L-histidine, DL-malic acid and Aspartame^TM in aqueous solution, as a function of the pK values and partial charges on the oxygens of the carboxylate groups of these molecules. The results are discussed and qualitatively interpreted in terms of the forced electric dipole and dynamic coupling mechanisms of the 4f-4f intensities, thus indicating that the forced electric dipole mechanism is dominant.

Keywords: neodymium, amino acids, transition intensity parameters

Introduction

The study of the chemical bonding between trivalent lanthanide ions (Ln³⁺) and amino acids or peptides has its origin in the interest in using these ions as structural probes in biological systems, particularly in those systems which contain Ca²⁺ in their structure^1,2. The Ca²⁺ ion is optically inactive and is, therefore, not suitable for providing information, through optical spectroscopic measurements, about the chemical environment in which it is embedded. On the other hand, almost all Ln³⁺ ions are know to exhibit rich optical spectra, either in absorption or emission, and it occurs that, due to the similarity between ionic radii, they may substitute Ca²⁺ ions in the chemical structures.

There is strong evidence that the bonding between Ln³⁺ ions and amino acids is made with the oxygens of the carboxylate group, and that the bonding via the nitrogen of the amino group is unlikely to occur at least in a range of pH values up to 5.6^3,4. In this paper we examine the relation between the basicity of the carboxylate groups of glycine (Gly), L-aspartic acid (Asp), L-glutamic acid (Glu), L-histidine (His), DL-malic acid (Mal) and Aspartame^TM (APM) (Table 1), and the intensity parameters of 4f-4f transitions in their compounds with Nd³⁺ ion. The intensity parameters are qualitatively interpreted in terms of the forced electric dipole and dynamic coupling mechanisms of 4f-4f intensities^5-10. Both mechanisms are dependent on the chemical environment around the Ln³⁺ ion. The basicity of the carboxylate groups is considered according to pK values and partial atomic charges on the oxygens which were calculated from molecular mechanics and the semi-empirical PM3 quantum chemical method.

Experimental

The samples were prepared from aqueous solutions of Nd(ClO₄)₃ and the amino acids glycine, L-aspartic acid, L-glutamic acid and L-histidine, the dipeptide Aspartame^TM or the DL-malic acid. The absorption spectra were measured in a Carl Zeiss M-40 UV-visible spectrophotometer between 11000 cm^-1 and 30000 cm^-1. Solutions of Nd(ClO₄)₃ with pH 5.0-5.5 were prepared from Nd₂O₃ (Aldrich, 99.99%) and standardized by EDTA / xylenol orange titration. Solutions of the ligands were standardized by titration with NaOH / phenolphtalein or by potentiometric titration. The concentration of all solutions were around 5.00 x 10^-2 mol L^-1.

We have firstly examined the behavior of the ⁴I_{9/2 ®}⁴G_5/2, ²G_7/2 hypersensitive transitions of Nd³⁺, between 16600 cm^-1 and 18200 cm^-1, as a function of the molar ratio Nd³⁺: Ligand and the pH of the solution, which was varied from 1 to 5.5 for each ratio. The solutions of the Nd(ClO₄)₃ and the ligand were mixed in a quartz cell of 1.00 cm optical pathway which was coupled to a quartz bulb with 20 mL capacity. Volumes were measured with calibrated pipetes. The molar ratio Nd³⁺ / Ligand was varied from 1:1 to 1:10. The pH was adjusted by addition of acid or base and measured directly in the cell with a combined glass microelectrode. Concentrations were corrected for the volume. Wavenumber scan was made with variable slits and constant energy beam at the photodetector and the absorbance values were read to the fourth decimal place in a digital display. All the measurements were made at 25 ± 1 °C. Doubly destilated water was used as reference. The samples with the ratio Nd³⁺ / Ligand and pH that gave the highest absorption in the region of hypersensitivity were used to obtain the entire spectra from which the spectroscopic parameters were obtained. The best ratio Nd³⁺ / Ligand is 1/4 except for the Nd³⁺ / Aspartame, in which case the experimental oscilator strength always rises with increasing of the quantity of the ligand. Figure 1 shows the curves of P_expvs. pH for several samples using histidine as ligand. As we see, P_exp decreases as the ratio goes to higher values than 1/4.

As we observed nine groups of transitions in the entire spectral region under investigation, we obtained nine equations for P (Eq. 2), where the following values were introduced for the calculation of t_l: experimental P values as obtained in the Eq. 1, experimental s values as the baricenter of a group of transitions obtained from the area measurements under the absorption curve, U^(l) values as obtained of the medium value for the U^(l) of the transitions in each group¹¹. A program in BASIC was developed to treat these nine equations by a least square method to obtain the t_l parameters. The least square method reduces a system of n equations with 3 unknown quantities to a system of 3 equations with 3 unknowns, which are the phenomenological t_l parameters.

Results and Discussion

The experimental oscillator strengths are obtained through the expression

were e(s) is the molar extintion coefficient at wavenumber s (cm^-1) and the integral in Eq. 1 is directly proportional to the area under the absorption curve.

According to the theory of 4f-4f intensities⁵, the oscillator strength of a transition between two manifolds, with respective total angular momenta J and J’, of a given 4f^N electronic configuration is given by:

where s is the baricenter of the transition energy (in wavenumbers), U^(l) is a unit tensor operator of rank l and the t_l are the so-called intensity parameters which depend on the chemical environment, radial integrals and interconfigurational energy differences in the lanthanide ion. The reduced matrix elements of U^(l) in Eq. 2 have been calculated, in the intermediate coupling scheme, for the whole series of the trivalent lanthanides¹².

An alternative way of expressing the theoretical oscillator strength has been of common use in the literature, in terms of the W_l intensity parameters which are related to the t_l parameters by W_l = t_l/1.085 x 10¹¹ c cm^-1, where c = (h² + 2)²/9h, h being the index of refraction of the medium¹³. For the sake of comparison with the results of previous studies on 4f-4f intensities in compounds of trivalent lanthanides with amino acids^14,15, the expression of the theoretical oscillator strength in terms of the t_l intensity parameters as in Eq. 2 is used in the present work.

In principle these parameters can be calculated from theoretical models provided structural data around the lanthanide ion are available^5,9. However, a common procedure is to treat them as adjustable parameters to reproduce the observed oscillator strengths. The t_l thus obtained are refered to as phenomenological intensity parameters. In this procedure the least square method is commonly used in which the input data are the values of the measured oscillator strengths, the squared reduced matrix elements of U^(l) and the transition energies s. Table 2 presents the values of the oscilator strengths and transition energies, corresponding to the transitions observed in the absorption spectrum of the complex with L-aspartic acid, as a function of the pH. It may be noted that the intensity of the hypersensitive ⁴I_9/2®⁴G_5/2,2G_7/2 transitions increases as the pH increases up to aproximately 5.4 value. For pH values above 5.5 the Nd³⁺ ion hydrolyses.

In the fitting procedure to obtain the phenomenological t_l parameters, the squared reduced matrix elements of U^(l) for the transitions separated by groups, as indicated in Table 2, were summed together. The results are presented in Table 3.

In the case of the glycine ligand, the t_l values presented in Table 3 in the range of pH above the pK₁ agree with the values obtained by Legendziewicz et al.^14,15 for the compounds in the crystalline phase. The same agreement is not observed when the ligand is glutamic acid, where t₄ and t₆ are discrepant from the solid state values. This fact can be explained if we consider that the two carboxylate groups of the glutamic acid may or may not be involved in the coordination at the pH value used in these measurements.

Among the t_l parameters, in general t₂ is the most sensitive to the coordination geometry and the characteristics of the ligands¹³. We have examined the behavior of t₂ with the ligand’s pK; t₂ has varied linearly with pK₁ provided the monocarboxylic and dicarboxylic species were considered separately (Fig. 3). We have also examined the behavior of t₂ with the average value <pk> = (pK₁ + pK₂) / 2, since at pH ~5 both carboxylic groups are expected to be equally deprotonated. In this case, t₂ has increased with <pk>, but not linearly (Fig. 4). Note that in this plot, <pk> = pK₁ for the monocarboxylic species.

An interesting correlation is also obtained between t₂ and the average partial charges on the carboxylate oxygens. The molecular amino acid modelling was performed by empirical calculation methods. Firstly, the geometry optimization was made by a method of molecular mechanics, using a modified MM2 force field^16,17,18 named MM+, and the Polak-Ribiere minimum energy search procedure¹⁹. Secondly, a single point calculation was performed by the quantum mechanical semi-empirical PM3 method²⁰. This leads to the partial atomic charges in the ligands. The idea is not to get the most reliable set of partial atomic charges, but rather to follow the trends involving partial charges on the oxygens, the pK values, and consequently the basicity of the oxygens, and t₂. The results are summarized in Table 4. Figure 5 shows a plot of t₂vs. the average charges on the carboxylate oxygens. As in the case of Fig. 4 an increasing behavior of t₂ is also observed.

Figure 6 indicates the partial charges on the concerned atoms of the ligand molecules.

The oxygen charges in Table 4 and Fig. 5 correspond in each case to the average value between the charges on the oxygens of the carboxylate groups.

From the point of view of the ligand field theory, the larger the negative charge on the oxygens, the greater is the ionic interaction between the ligand and the lanthanide ion. Theoretically, it has been accepted that there are two dominating mechanisms contributing to the t_l parameters. These are the forced electric dipole and dynamic coupling mechanisms^8,9. The t_l can be expressed as:

where C is a constant and the quantities B_l,t,p are the so-called intensity parameters for 4f-4f transitions between individual Stark levels. The B_l,t,p are expressed as a sum of the two contributions:

It has been shown that these two contributions have opposite signs¹⁰. A strong ligand field tends to favor the forced electric dipole mechanism. Further, the dynamic coupling Hamiltonian is directly proportional to the oxygen polarizability, which decreases as the localized charge on the oxygens increases. Thus, the observed increasing behavior of t₂ with the pK and with the partial charges on the carboxylate oxygens suggests that, in the present compounds with the Nd³⁺ ion, the forced electric dipole mechanism is dominant. On the other hand, it is not obvious why this increasing behavior of t₂ is approximately linear. This is a point which deserves a more detailed theoretical analysis and is beyond the scope of this paper.

Finally, we notice that in the range of pH with acceptable t_l values, there is a predominant complex species that allows a comparative analysis of the results in the scope of this work.

Acknowledgments

The authors acknowledge the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), the Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) and the Fundação para o Desenvolvimento da UNESP (FUNDUNESP) for financial support. FUNDUNESP also helped in meeting the publication costs of this article.

We are also very grateful to Prof. Romeu Magnani (IQ-UNESP) for the development of the computational program to calculate the t_l parameters.

FAPESP helped in meeting the publication costs of this article

• 1. Nieboer, E. Structure and Bonding, 1975
• 2. Martin, R.B.; Richardson, F.S. Quart. Rev. Biophys 1979, 12, 181.
• 3. Prados, R.; Stadtherr, L.G.; Donato Jr., H.; Martin, R.B. J. Inorg. Nucl. Chem 1974, 36, 689.
• 4. Legendziewicz, J.; Huskowska, E.; Kozlowski, H.; Jezowska-Trzebiatowska, B. Inorg. Nucl. Chem. Letters 1979, 15, 349.
• 5. Judd, B.R. Phys. Rev. 1962, 127, 750.
• 6. Ofelt, G.S. J. Chem. Phys 1962, 37, 511.
• 7. Mason, S.F.; Peacock, R.D.; Stewart, B. Chem. Phys. Letters 1974, 29, 149.
• 8. Judd, B.R. J. Chem. Phys 1979, 70, 4830.
• 9. Malta, O.L.; Ribeiro, S.J.L.; Faucher, M.; Porcher, P. J. Phys. Chem. Solids 1991, 52, 587.
• 10. Xia, S.-D.; Reid, M.F. J. Phys. Chem. Solids 1993, 54, 777.
• 11. Carnall, W.T.; Fields, P.R.; Rajnak, K. J. Chem. Phys. 1968, 49, 4424.
• 12. Carnall, W.T.; Crosswhite, H.; Crosswhite, H.M. In Energy Level Structure and Transition Probabilities of the Trivalent Lanthanides in LaF₃, Argonne National Laboratory, Report, 1977.
• 13. Peacock, R.D. Structure and Bonding 1975, 22, 83.
• 14. Legendziewicz, J.; Huskowka, E.; Strek, W.; Jezowsk-Trzebiatowska, B. J. Luminescence 1981, 24/25, 819.
• 15. Legendziewicz, J.; Huskowska, E.; Argay, Gy.; Waskowska, A. Inorg. Chim. Acta 1984, 95, 57.
• 16. Allinger, N.L. J. Am. Chem. Chem. Soc. 1977, 99, 8127.
• 17. Allinger, N.L.; Yuh, Y.H. In Quantum Chemistry Program Exchange, Bloomington, IN Program # 395.
• 18. Burkert, U.; Allinger, N.L. In Molecular Mechanics, ACS Monograph 177, 1982.
• 19. HyperChem^TM In Computational Chemistry, Autodesk, Sausalito, CA, 1993.
• 20. Stewart, J.J.P. J. Comp. Chem 1991, 12, 320.
• 21. Lide, D.R., editor; "CRC Handbook of Chemistry and Physics", 72nd ed., 1991-1992.

^*

Correspondence should be addressed at: Instituto de Química - UNESP, C.P. 355, 14801-970 Araraquara – SP, Brazil.

Publication Dates

• Publication in this collection
  06 Nov 2002
• Date of issue
  Sept 1998

History

• Received
  01 June 1998