^{1}, Kristopher J. Ooms

^{1}, Tatyana Polenova

^{1,a)}, Bharat Baruah

^{2}, Debbie C. Crans

^{2,b)}and Jason J. Smee

^{3}

### Abstract

solid-state NMR and density functional theory(DFT) investigations are reported for a series of pentacoordinate dioxovanadium(V)-dipicolinate [-dipicolinate] and heptacoordinate aquahydroxylamidooxovanadium(V)-dipicolinate [V(V)O-dipicolinate] complexes. These compounds are of interest because of their potency as phosphatase inhibitors as well as their insulin enhancing properties and potential for the treatment of diabetes. Experimental solid-state NMR results show that the electric field gradient tensors in the -dipicolinate derivatives are affected significantly by substitution on the dipicolinate ring and range from . The chemical shiftanisotropies show less dramatic variations with respect to the ligand changes and range between and . To gain insights on the origins of the NMR parameters, DFT calculations were conducted for an extensive series of the - and V(V)O-dipicolinate complexes. To assess the level of theory required for the accurate calculation of the NMR parameters, different functionals, basis sets, and structural models were explored in the DFT study. It is shown that the original x-ray crystallographic geometries, including all counterions and solvation water molecules within of the vanadium, lead to the most accurate results. The choice of the functional and the basis set at a high level of theory has a relatively minor impact on the outcome of the chemical shiftanisotropy calculations; however, the use of large basis sets is necessary for accurate calculations of the quadrupole coupling constants for several compounds of the series. These studies demonstrate that even though the vanadiumcompounds under investigations exhibit distorted trigonal bipyramidal coordination geometry, they have a “perfect” trigonal bipyramidal electronic environment. This observation could potentially explain why vanadate and vanadium(V) adducts are often recognized as potent transition state analogs.

The authors thank Dr. Olga Dmitrenko for her generous help with setting up the DFT calculations. T.P. acknowledges financial support of the National Science Foundation (NSF-CAREER No. CHE-0237612) and the National Institutes of Health (No. P20-17716, COBRE individual subproject). D.C.C. acknowledges financial support of the Institute of General Medicine at the National Institutes of Health (No. GM40525) and the National Science Foundation (No. CHE-0314719). J.J.S. thanks The Welch Foundation (Grant No. BP-0037) for partial support. K.J.O. thanks the Natural Sciences and Engineering Research Council of Canada for financial support.

I. INTRODUCTION

II. EXPERIMENT

A. Synthesis of V(V)O- and -dipicolinate complexes

B. NMR spectroscopy

C. Computational methods

III. RESULTS AND DISCUSSION

A. solid-state NMR spectroscopy

B. DFT calculations

C. Relationship between NMR parameters and structure

IV. CONCLUSIONS

V. SUPPORTING INFORMATION

### Key Topics

- Tensor methods
- 46.0
- Vanadium
- 44.0
- Chemical shifts
- 38.0
- Nuclear magnetic resonance
- 34.0
- Density functional theory
- 26.0

## Figures

Molecular structures of the V(V)-dipicolinate derivatives under investigation: [(a)–(e)] derivatives and [(f)–(h)] V(V)O derivatives.

Molecular structures of the V(V)-dipicolinate derivatives under investigation: [(a)–(e)] derivatives and [(f)–(h)] V(V)O derivatives.

MAS NMR spectra of (compound e) acquired at different MAS frequencies: (a) , (b) , (c) , and (d) .

MAS NMR spectra of (compound e) acquired at different MAS frequencies: (a) , (b) , (c) , and (d) .

MAS NMR spectra of the five-coordinate -dipicolinate derivatives shown in Fig. 1(a)–1(e): (a) ; (b) ; (c) ; (d) ; and (e) . The spectra were acquired at a spinning frequency of .

MAS NMR spectra of the five-coordinate -dipicolinate derivatives shown in Fig. 1(a)–1(e): (a) ; (b) ; (c) ; (d) ; and (e) . The spectra were acquired at a spinning frequency of .

Experimental spectrum (a) compared to the spectrum simulated in SIMPSON (b) for [compound a, Fig. 1(a)] at .

Experimental spectrum (a) compared to the spectrum simulated in SIMPSON (b) for [compound a, Fig. 1(a)] at .

Comparison of experimental and calculated quadrupolar coupling constant for the five- and seven-coordinate vanadium(V)-dipicolinate derivatives under investigation computed using different geometries: (a) x-ray geometry with counterions and hydration water molecules included; (b) x-ray geometry without counterions and hydration water molecules; and (c) optimized geometries with no counterions or hydration water molecules. Dotted line is best fit, and solid line represents an ideal agreement between calculated and experimental . Different symbols represent different DFT methods used: open circle, PBE1PBE exchange-correlation functional and basis set; open square, B3LYP exchange-correlation functional and basis set; closed circle, PBE1PBE exchange-correlation functional and TZVP basis set; and closed square: B3LYP exchange-correlation functional and TZVP basis set. The fit parameters are entered in Table II.

Comparison of experimental and calculated quadrupolar coupling constant for the five- and seven-coordinate vanadium(V)-dipicolinate derivatives under investigation computed using different geometries: (a) x-ray geometry with counterions and hydration water molecules included; (b) x-ray geometry without counterions and hydration water molecules; and (c) optimized geometries with no counterions or hydration water molecules. Dotted line is best fit, and solid line represents an ideal agreement between calculated and experimental . Different symbols represent different DFT methods used: open circle, PBE1PBE exchange-correlation functional and basis set; open square, B3LYP exchange-correlation functional and basis set; closed circle, PBE1PBE exchange-correlation functional and TZVP basis set; and closed square: B3LYP exchange-correlation functional and TZVP basis set. The fit parameters are entered in Table II.

Quadrupolar coupling constants calculated using different basis sets for (a) , (b) , (c) , (d) , (e) , (f) , and (g) . The solid line represents the experimentally determined , and the shaded area represents the experimental error.

Quadrupolar coupling constants calculated using different basis sets for (a) , (b) , (c) , (d) , (e) , (f) , and (g) . The solid line represents the experimentally determined , and the shaded area represents the experimental error.

Comparison of experimental and calculated principal components of the chemical shift anisotropy tensor for different groups of compounds: (a) full set, (b) , and (c) V(V)O. Dotted line is best fit, and solid line represents an ideal agreement between calculated and experimental . Different symbols represent different DFT methods used: open circle, PBE1PBE exchange-correlation functional and basis set; open square, B3LYP exchange-correlation functional and basis set; closed circle, PBE1PBE exchange-correlation functional and TZVP basis set; and closed square: B3LYP exchange-correlation functional and TZVP basis set. The fit parameters are entered in Table III.

Comparison of experimental and calculated principal components of the chemical shift anisotropy tensor for different groups of compounds: (a) full set, (b) , and (c) V(V)O. Dotted line is best fit, and solid line represents an ideal agreement between calculated and experimental . Different symbols represent different DFT methods used: open circle, PBE1PBE exchange-correlation functional and basis set; open square, B3LYP exchange-correlation functional and basis set; closed circle, PBE1PBE exchange-correlation functional and TZVP basis set; and closed square: B3LYP exchange-correlation functional and TZVP basis set. The fit parameters are entered in Table III.

(a) The molecular structure of (compound e) with the unique components of the shielding ( in yellow) and the EFG tensor ( in orange). is perpendicular to the dipicolinate plane, while is perpendicular to the plane. (b) Image of the electrostatic potential surface of compound e created from the total self-consistent field density calculated in GAUSSIAN. The blue indicates regions of negative charge (attracts a positive test charge), while the red represents regions of positive charge (repels a positive test charge). Note the difference in magnitude at the extremes of the scale.

(a) The molecular structure of (compound e) with the unique components of the shielding ( in yellow) and the EFG tensor ( in orange). is perpendicular to the dipicolinate plane, while is perpendicular to the plane. (b) Image of the electrostatic potential surface of compound e created from the total self-consistent field density calculated in GAUSSIAN. The blue indicates regions of negative charge (attracts a positive test charge), while the red represents regions of positive charge (repels a positive test charge). Note the difference in magnitude at the extremes of the scale.

The four molecular orbitals with the greatest vanadium -character and which contribute the most to the vanadium isotropic paramagnetic shielding with the percent contribution, energy level, and the vanadium -orbital contributions are indicated below the image.

The four molecular orbitals with the greatest vanadium -character and which contribute the most to the vanadium isotropic paramagnetic shielding with the percent contribution, energy level, and the vanadium -orbital contributions are indicated below the image.

## Tables

Experimental solid-state NMR parameters for five- and seven-coordinate vanadium(V)-dipicolinate derivatives. (i) The chemical shift parameters are defined such that and (Ref. 64). The components of the chemical shift tensor are , , and . Note that, according to Haeberlen-Mehring-Spiess, the anisotropy of the CSA tensor is defined as . (ii) The EFG parameters are and , where , is the electron charge, is Planck’s constant, and is the nuclear quadrupolar moment of , (Ref. 65).

Experimental solid-state NMR parameters for five- and seven-coordinate vanadium(V)-dipicolinate derivatives. (i) The chemical shift parameters are defined such that and (Ref. 64). The components of the chemical shift tensor are , , and . Note that, according to Haeberlen-Mehring-Spiess, the anisotropy of the CSA tensor is defined as . (ii) The EFG parameters are and , where , is the electron charge, is Planck’s constant, and is the nuclear quadrupolar moment of , (Ref. 65).

Summary of the DFT calculations of the quadrupolar coupling constant : Correlations between calculated and experimental for different starting geometries.

Summary of the DFT calculations of the quadrupolar coupling constant : Correlations between calculated and experimental for different starting geometries.

Summary of the DFT calculations of the principal components of the CSA tensor: Correlations between calculated and experimental for different starting geometries.

Summary of the DFT calculations of the principal components of the CSA tensor: Correlations between calculated and experimental for different starting geometries.

Article metrics loading...

Full text loading...

Commenting has been disabled for this content