Hyperfine Effects in Ligand NMR: Paramagnetic ... - ACS Publications

However, paramagnetic transition-metal complexes such as those with ruthenium in the +3 oxidation state are troublemakers because their unpaired elect...
0 downloads 0 Views 3MB Size
This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Article Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX

pubs.acs.org/IC

Hyperfine Effects in Ligand NMR: Paramagnetic Ru(III) Complexes with 3‑Substituted Pyridines Jan Novotný,† David Přichystal,†,‡ Martin Sojka,†,‡ Stanislav Komorovsky,§ Marek Nečas,†,‡ and Radek Marek*,†,‡ †

CEITEC − Central European Institute of Technology, Masaryk University, Kamenice 5, CZ − 62500 Brno, Czechia Department of Chemistry, Faculty of Science, Masaryk University, Kamenice 5, CZ − 62500 Brno, Czechia § Institute of Inorganic Chemistry, Slovak Academy of Sciences, Dúbravská cesta 9, SK-84536 Bratislava, Slovakia ‡

S Supporting Information *

ABSTRACT: NMR spectroscopy is an indispensable tool in characterizing molecular systems, including transition-metal complexes. However, paramagnetic transition-metal complexes such as those with ruthenium in the +3 oxidation state are troublemakers because their unpaired electrons induce a fast nuclear spin relaxation that significantly broadens their NMR resonances. We recently demonstrated that the electronic and spin structures of paramagnetic Ru(III) systems can be characterized in unprecedented details by combining experimental NMR results with relativistic density-functional theory (Novotny et al. J. Am. Chem. Soc. 2016, 138, 8432). In this study we focus on paramagnetic analogs of NAMI with the general structure [3-R-pyH]+trans-[RuIIICl4(DMSO)(3-R-py)]−, where 3-R-py stands for a 3-substituted pyridine. The experimental NMR data are interpreted in terms of the contributions of hyperfine (HF) NMR shielding and the distribution of spin density calculated using relativistic DFT. The DFT computational methodology is evaluated, and the effects of substituents, environment, and relativity on the hyperfine shielding are discussed. Particular attention is paid to the analysis of the fundamental Fermi-contact (FC), spin-dipole (SD), and paramagnetic spin−orbit (PSO) terms that contribute to the hyperfine 1H and 13C NMR shifts of the individual atoms in the pyridine ligands and the spin-polarization effects in the ligand system that are linked to the character of the metal−ligand bond. The individual HF shielding terms are systematically discussed as they relate to the traditional, but somewhat mixed, contact and pseudocontact NMR contributions used extensively by experimental spectroscopists in biomolecular NMR and the development of PARACEST magnetic-resonance contrast agents. observation and assignment of the NMR resonances.4 In the past few years the magnetic-response theory for open-shell systems with arbitrary multiplicity and containing heavy elements has matured, and this has stimulated extensive development of paramagnetic NMR (pNMR) spectroscopy and its linking to electronic structure.5−14 Interest in using electronic structure methods to accurately calculate and interpret the NMR chemical shifts of paramagnetic compounds is currently increasing, but the treatment of open-shell systems represents a frontier in modern quantum chemistry. Particularly in the field of transition-metal

1. INTRODUCTION Paramagnetic (open-shell) molecular systems have traditionally been investigated by electron paramagnetic resonance (EPR) spectroscopy and characterized by electronic g-tensors and hyperfine coupling tensors (A-tensors).1,2 Electronic g-shift and hyperfine coupling constants (A) can be obtained for paramagnetic molecules in isotropic solutions. But these open-shell systems can also be characterized by nuclear magnetic resonance (NMR) spectroscopy. Note that the NMR signals in these systems are usually dramatically affected by additional hyperfine (de)shielding caused by the unpaired electron spin density and shifted to unusual spectral regions.3 In addition, fast electron−nuclear spin relaxation can significantly broaden the NMR signals and further hamper © XXXX American Chemical Society

Received: September 25, 2017

A

DOI: 10.1021/acs.inorgchem.7b02440 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

and discussed. The four-component Dirac−Kohn−Sham (DKS) approach is used in the program ReSpect24 with solvent effects included (by means of a recently implemented PCM model)25 to estimate the mechanism of the hyperfine interaction for selected atoms of the ligands. Particular attention is paid to the mechanisms by which the spin density is distributed in the ligand moiety upon formation of the metal−ligand bond and to the analysis of the fundamental Fermi-contact, spin-dipole, and paramagnetic spin−orbit terms that contribute to the HF NMR shifts of the individual atoms in the pyridine ligands. In this work the temperature-dependent part of the ligand NMR shift is denoted as the hyperfine (HF) contribution in order to stress that it originates in the spin−spin interaction between the nucleus and an electron, as does the HF coupling constant of electron paramagnetic resonance (EPR) spectroscopy. Note that in the current literature this type of temperature-dependent NMR contribution is alternatively termed paramagnetic or Curie.10,13,20

complexes, computational methods require large basis sets and treatment of the relativistic effects. Most pNMR studies of these systems therefore utilize methods of density-functional theory (DFT). The hyperfine (HF, or paramagnetic) contributions to the NMR shifts for systems in the doublet electronic state (i.e., with a single unpaired electron) can be reconstructed from the electronic g-tensor and electron− nucleus hyperfine (HF) coupling A-tensor in a straightforward manner (for details, see Section 2). The g represents a global response property of the molecular or supramolecular system, whereas A is specific to any nucleus in the system. The spindipole (SD) contribution to A is given by the interaction of the nuclear dipole with the electron spin density and is generally highly anisotropic. In contrast, the Fermi-contact (FC) contribution is given directly by the imbalance of the α and β electron spin density at the position of the spectator nucleus and is rather isotropic. Therefore, the SD contribution is typically interpreted in terms of the molecular topology and the relative mutual orientations of the molecular or supramolecular subunits,4 and the FC term is governed by the covalence and polarity of the chemical bonds involved. The NMR spectra of some NAMI-A type Ru(III) complexes15−17 have previously been calculated by using DFT methods, with the approaches ranging from nonrelativistic to fully relativistic.18,19,12 We recently demonstrated that the electronic and spin structures of paramagnetic Ru(III) systems can be characterized in unprecedented details by combining experimental NMR results with relativistic density-functional theory (DFT).20 Herein we extend our previous study by investigating Ru(III) compounds with the general structure [3R-pyH]+trans-[RuIIICl4(DMSO)(3-R-py)]−, where 3-R-py stands for a 3-substituted pyridine, Figure 1.

2. METHODS 2.1. Synthesis. Materials. The starting compounds, RuCl3·xH2O, 3-methylpyridine, pyridine-3-carbonitrile, 3-fluoropyridine, and pyridine-3-carboxylic acid, were used as obtained from our supplier without further purification. Solvents of p.a. grade were used unless otherwise stated. The [DMSOH][trans-RuCl4(DMSO)2] was prepared according to a procedure reported in the literature.26 2.2. X-ray Diffraction. Structural data for the Ru(III) complexes were collected on a Rigaku MicroMax-007 HF rotating anode fourcircle diffractometer using Mo Kα radiation. The temperature during data collection was 120(2) K. The structures were solved by direct methods and refined by full-matrix least-squares refinement using the software package ShelXTL. Crystal data and structure refinement parameters are listed in Table 1. 2.3. NMR Spectroscopy. The 1H, 13C, and 2D NMR spectra of Ru(III) complexes 1−5 were measured on a Bruker Avance III HD 700 MHz spectrometer. The NMR samples were prepared by dissolving 5−15 mg of the complex in 0.5 mL of DMF-d7. The signals of the solvent (δ(1H) = 8.03 ppm, δ(13C) = 163.2 ppm) served to reference the temperature-dependent NMR spectra. 1H-coupled 13 C NMR spectra were measured to distinguish between the NMR resonances of nonquaternary C−H and quaternary Cq atoms. The 19F NMR spectra of compound 5 were collected on a Bruker Avance III HD 850 MHz spectrometer with the signals referenced relative to the resonance frequency of deuterium in DMF-d7, the solvent used for the NMR measurements. 2.4. Quantum Chemical Calculations. Geometry and Energy Decomposition Analysis. The structures were optimized using a previously verified DFT protocol20,27−29 (PBE0 functional30,31 and the def2-TZVPP32 basis set), with the corresponding relativistic effective core potentials (def2-ECPs)33 for the metal centers (ECP substituting 28 electrons for Ru), as implemented in the Turbomole 6.03 program.34 The structures were optimized by using the COSMO (Conductor-like Screening Model)35 solvent model with the default solvent parameters (dielectric constant 37, solvent radius 3.13 Å) used for DMF. For the Cartesian coordinates, see Supporting Information. All calculations were performed using an m5 integration grid with the following convergence criteria: 10−6 for the energy change and 10−4 for the geometry gradient. Energy decomposition analysis (EDA) as implemented in the ADF 2014 program (at scalar-relativity level, in the gas phase) was used to analyze the bonding between the nitrogen of the pyridine ligand and the paramagnetic Ru center. The electron deformation density was analyzed using the EDA-NOCV approach.36,37 For more details, see our recent work.20 NMR Chemical Shifts and EPR Parameters. The NMR shielding constants were calculated using the methods of density-functional

Figure 1. Structure and atom numbering scheme for Ru(III)-based compounds 1−5.

The experimental 1H and 13C NMR data were measured at several temperatures and analyzed using Curie plots (dependence of the NMR chemical shift on 1/T).21,22 These experimental NMR data are then compared with those calculated using relativistic DFT methods. The two-component ZORA approach, as implemented in the ADF program,23 is used to correlate the theoretical values with the experimental NMR chemical shifts and to investigate the individual factors that govern the propagation of the spin density to the ligand moiety. In particular, the effects of the geometry (the conformation of the aromatic moiety with respect to the trans DMSO ligand) and the solvent on the HF contributions to the NMR chemical shifts calculated by DFT are analyzed B

DOI: 10.1021/acs.inorgchem.7b02440 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Table 1. Selected Crystallographic Data for Compounds 2, 3, and 5a CCDC No Chemical formula Formula weight Crystal system Space group a (Å) b (Å) c (Å) α (deg) β (deg) γ (deg) V (Å3) Z Dcalcd. (g·cm−3) F (000) μ (mm−1) Measured/unique reflections Data/parameters R1/wR2 [I > 2σ(I)] R1/wR2 [all data] GooF Δρmax /Δρmin (e· Å−3) a

2

3

5

1573286 C14H21Cl4N2ORuS 508.26 monoclinic P21/c 9.02740(10) 14.4932(2) 15.4494(2) 90 103.514(2) 90 1965.37(5) 4 1.718 1020 1.451 9227/3721 3721/212 0.0234/0.0580 0.0243/0.0585 1.159 0.975/−0.597

1573287 C14H15Cl4N4ORuS 530.23 triclinic P-1 8.07000(10) 10.5472(2) 11.4555(2) 81.448(2) 77.9690(10) 87.0440(10) 942.81(3) 2 1.868 526 1.520 11667/3565 3565/231 0.0258/0.0700 0.0269/0.0712 1.098 0.739/−0.875

1573288 C12.50H15Cl5F2N2ORuS 557.64 triclinic P-1 7.7036(2) 8.6377(3) 14.7648(5) 89.287(3) 80.541(2) 88.656(3) 968.81(5) 2 1.912 550 1.629 9340/3670 3670/247 0.0230/0.0582 0.0241/0.0591 1.051 0.929/−0.709

For additional data, see Supporting Information. The effect of the corrected exchange-correlation kernel45 on the orbital NMR chemical shift was evaluated for compound 3. The analysis revealed a rather negligible contribution by the XC kernel (δorb XC 1 ∼ 0.1 ppm for 13C; δorb XC ∼ 0.05 ppm for H); see Table S1. The overall performance of the computational setup used in this work was evaluated by calculating the deviations (mean absolute errors − MAE) of the theoretical NMR chemical shifts from the experimentally obtained values (1H and 13C NMR in Figure S2). The spatial distribution of the total spin density46 and the spin populations for individual ligand atoms in the ruthenium complexes were calculated at the scalar-relativistic (ZORA/PBE0/TZ2P) level, 1c, either in vacuum or by using the COSMO solvent model (DMF).35 Four-Component (4c) DFT Approach to EPR Parameters. The EPR parameters were calculated by using the relativistic DFT program ReSpect (version 4.0.0).24 All calculations were performed at the fourcomponent DKS level of theory47−49 using uncontracted Jensen’s pcJ1/pcJ-250 basis sets (termed upcJ-1/upcJ-2) for the light atoms and Dyall’s valence double- and triple-ζ basis sets (termed vdz/vtz)51 for Ru (smaller basis sets were used in some cases to simplify the analysis). The PBE0 functional42,43 and the recently implemented PCM solvent model25 were used to calculate the EPR parameters (gtensors and ligand hyperfine coupling tensors, A-tensors). The spatial distribution of the z-component of the 4c spin density was analyzed to interpret the electronic factors that influence the hyperfine contributions to the NMR chemical shifts. The ReSpect program was also used to investigate the individual contributions of the Fermi-contact (FC), spin-dipole (SD), and paramagnetic nuclear-spin−electron-orbit (PSO) operators to the HF coupling tensor for the individual ligand atoms L (AL). The FC, SD, and PSO terms are defined in the relativistic four-component domain as (using the Hartree system of atomic units)

theory (DFT) mentioned below. The systematic offsets of the DFT methods used in calculating the NMR chemical shifts were reduced by using benzene (in benzene as approximated by the COSMO model)27,28,38 as a secondary reference relative to TMS: δref = 7.15 ppm for 1H and δref = 127.8 ppm for 13C.

δL = σref − σL + δref

(1)

where δL is the ligand NMR chemical shift of interest and σref and σL are the isotropic NMR shielding constants of particular atoms in the secondary reference (benzene) and the ligand being investigated, respectively. The total NMR chemical shifts (δtot)20 can be split into orbital (δorb, very nearly temperature-independent) and hyperfine (δHF, temperature-dependent) contributions (for more detailed discussion of the theory, see refs 5 and 10−12). δ tot = δ orb + δ HF

(2)

Note that in contrast to δpara used in our previous work,20 we now denote the temperature-dependent NMR shift δHF in order to stress its natural dependence on the HF coupling constant. The conventionally analyzed terms that contribute to the HF NMR shifts (known as contact and pseudocontact, see the section Four-Component (4c) DFT Approach to EPR Parameters) were evaluated using the equations that have been shown and described in our previous studies.12,20 The presence of a counterion has been demonstrated to play a marginal role in the NMR shift calculations in our recent study;20 therefore, it is neglected in our production calculations. Two-Component (2c) DFT Approach to EPR and NMR Parameters. The NMR shielding constants along with the EPR parameters were calculated by using the program ADF2016.39 The calculations were performed at the 2c (SO-ZORA) level,40,41 using the PBE0 functional,42,43 the TZ2P basis set,44 and the COSMO solvent model,35 as implemented in the ADF program. As the current twocomponent implementation of the NMR calculations in the ADF program is limited to closed-shell systems, diamagnetic Ru(II) analogs were employed to calculate the orbital contributions (δorb) to the total NMR chemical shifts. In addition, the orbital shift associated with the paramagnetic state Ru(III) was evaluated at the one-component level (ZORA). For a comparison of these approaches, see Section 3.3.

ALFC =

ASD L = C

γL 4π ⟨S⟩̃ 3c

γL 1 ⟨S⟩̃ 2c

occ

∑ Re{⟨φi J |σδ(L)|ϕi J ⟩}

(3)

i=1

occ

⎧ ⎪

∑ Re⎨ ⎪

i=1



φi J

3(σ · rL)rL − rL2σ rL5

⎫ ⎪ ⎬ ϕi J ⎪ ⎭

(4)

DOI: 10.1021/acs.inorgchem.7b02440 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry ALPSO =

γL 1 ⟨S⟩̃ 2c

occ

⎧ ⎪

∑ Re⎨ ⎪

i=1



φi J 2

rL × p rL3

⎫ ⎪ ⎬ ϕi J ⎪ ⎭

(5)

Here γL is the gyromagnetic ratio of atom L, ⟨S̃⟩ is the effective spin of the system (1/2 in this work), c denotes the speed of light, the superscript J indicates the dependence of the molecular orbitals on the magnetization vector J, rL is the position vector of an electron relative to the nucleus L, rL is the length of the vector rL, p is the momentum operator, δ (L) denotes the delta function centered on the nucleus L, and σ is a vector composed of three Pauli matrices. Finally, φi and ϕi are the large and pseudolarge components of the four-component ith molecular orbital; see ref 52. Note that it is customary to use φLi (ϕLi ) instead of φi (ϕi), but we have dropped the superscript L to avoid confusion with the label for the ligand nucleus L. In addition to the terms in eqs 3−5, there is one additional contribution (termed AREL L in ref 53 which is, however, negligible in most of the cases studied here. For a more detailed discussion of eqs 3−5, see refs 48, 49, and 53. The role of the individual AL terms in the temperature-dependent HF part of the NMR shift (δHF L ) is analyzed by splitting δL into separate contributions arising from the isotropic and anisotropic parts of AL HFi HFa HFi HFa (δHF L = δL + δL ). For more detailed definitions of δL and δL , see eqs 12 and 13 in refs 4 and 12. Because all HF contributions defined by eqs 3−5 in the four-component relativistic regime have nonzero and δHFa can be split into isotropic and anisotropic parts, both δHFi L L HFa three contributions (FC, SD, and PSO). The δHFi L (δL ) corresponds 4 con to the traditional contact (pseudocontact) term δL (δpc L ) only if the distance of the ligand atom L from the metal center M is large enough SD for AFC L (AL ) to exhibit a purely isotropic (anisotropic) form. This observation is consistent with the fact that the structural information in δpc L can be extracted using the point-dipole approximation to the spin distribution on the metal center M only for distances greater than 10−20 Å from the paramagnetic center.54 Keeping this in mind, in , δHF/SD , and Section 3.5 we analyze the individual contributions δHF/FC L L HFi con rather than the somewhat mixed δ = δ and δHFa = δpc δHF/PSO L L L L L sums.

Figure 2. Molecular structure of compound 5 as determined by X-ray diffraction, with atom numbering scheme. For crystallographic data, see Table 1.

Table 2. Ru−N and Ru−S Interatomic Distances for Ru(III) Complexes 2, 3, and 5 as Determined by X-ray Diffraction in This Study and Calculated by DFT (PBE0/def2-TZVPP/ ECP/COSMO)a EXP 2 3 5

3. RESULTS AND DISCUSSION 3.1. Molecular Geometry: X-ray Diffraction and DFT Calculations. The desired Ru(III) complexes were obtained in high yields as pale to bright orange powders stable in air.55,56 The complexes were recrystallized from mixtures of dichloromethane with hexane or diethyl ether, and the crystal structures were determined by X-ray diffraction (see Table 1): the molecular structure of compound 5 is shown as an example in Figure 2. The X-ray data were used as experimental references for the molecular structures optimized by using DFT. The DFT method selected to optimize the geometry (PBE0/def2TZVPP/ECP/COSMO) was calibrated in our previous studies20,27,28 and has been validated here by comparing selected calculated and experimental interatomic distances; see Table 2. To analyze the conformational flexibility of the asymmetric 3-R-pyridine ligand relative to the trans DMSO group, an energy scan of the dihedral angle O−S−Ru−Cl was performed at the PBE0/def2-TZVPP/ECP/COSMO level. The resulting energy profiles (see Figure S3 in the Supporting Information) consist of four minima separated by low barriers to rotation of about 1−3 kcal·mol−1 ; Table S2. The effects of this temperature-accessible conformation flexibility on the EPR and NMR parameters are discussed in Section 3.3. 3.2. Effects of Temperature on the Experimental 1H and 13C NMR Chemical Shifts: Toward Orbital and Hyperfine NMR Contributions. The NMR spectra of the ruthenium complexes 1−5 (for 1H and 13C NMR chemical shifts; see Table 3 and Table 4, respectively) were typically measured in the temperature range 243−323 K, and the

a

DFT

Ru−N

Ru−S

Ru−N

Ru−S

212.7 213.0 211.7

229.8 228.3 227.9

211.7 211.9 212.0

229.2 228.6 228.8

See Section 2. All distances are in picometers (pm).

Table 3. Experimental 1H NMR Chemical Shifts for Ruthenium(III) Complexes 1−5 Measured at 293 Ka 1d

Compound MeDMSO

H2/H6b

H4

H5c

δorb δHF δtot δorb δHF δtot δorb δHF δtot δorb δHF δtot

2

3

4

5

H

Me

CN

COOH

F

+3.9 −16.6 −12.7 +7.8 −14.9 −7.1 +7.4 −1.6 +5.8 +7.6 −9.5 −1.9

+3.7 −16.4 −12.8 +7.3 -14.9 -7.6 +7.2 -1.7 +5.5 +7.5 −9.9 −2.4

+3.3 −16.0 −12.6 +7.7 -12.7 -5.0 +8.8 +0.1 +8.9 +7.5 −8.7 −1.2

+4.1 −16.7 −12.6 +7.5 −13.2 −5.7 +8.1 −1.3 +6.8 +7.9 −9.5 −1.5

+3.2 −15.9 −12.7 +7.0 −13.2 −6.2 +7.1 −0.3 +6.8 +7.4 −9.8 −2.4

a

The separate orbital (δorb) and hyperfine (δHF) contributions according to 1/T and the total chemical shifts (δtot) are given in ppm. bNot resolved because of significant line broadening. cIn compound 1, H3 and H5 are equivalent. dData taken from ref 20.

individual NMR chemical shifts were assigned based on the range of chemical shift, the line-width, the temperature change, and the correlation of the individual chemical shift with the value calculated using DFT; see Section 3.3. The 13C NMR D

DOI: 10.1021/acs.inorgchem.7b02440 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Table 4. Experimental 13C NMR Chemical Shifts for Ruthenium(III) Complexes 1−5 Measured at 293 Ka 1b

Compound MeDMSOc

C2

C3

C4

C5

C6

R

δorb δHF δtot δorb δHF δtot δorb δHF δtot δorb δHF δtot δorb δHF δtot δorb δHF δtot δorb δHF δtot

2

3

4

The approximately temperature-independent orbital shifts (δorb) and temperature-dependent hyperfine shifts (δHF) were extracted from Curie plots (chemical shift versus 1/T). An example of this dependency is shown for compound 3 in Figure 4. All other 1/T plots can be found in the Supporting Information (Figures S4−S6).

5

H

Me

CN

COOH

F

+43c −153c −110 +166 −75 +91 +122 −23 +99 +137 −24 +113 +122 −23 +99 +166 −75 +91

+44c −154c −110 +164 −73 +92 +131 −23 +108 +140 −25 +115 +120 −24 +96 +159 −70 +89 +19 −10 +10

+34c −147c −113 +162 −74 +88 +108 −25 +83 +140 −26 +114 +122 −17 +104 +163 −69 +94 +117 −14 +103

+47c −160c −113 +160 −73 +87 +124 −20 +104 +137 −25 +112 +119 −19 +99 +163 −71 +92 +165 −11 +154

+38c −150c −113 +148 −70 +78 +158 −34 +124 +125 −27 +99 +121 −22 +99 +158 −70 +88 −130d −2d −132d

a

The separate orbital (δorb) and hyperfine (δHF) contributions according to 1/T and the total chemical shifts (δtot) are given in ppm. bData taken from ref 20. cThe differences in the 13C NMR chemical shifts for the significantly broadened MeDMSO signals in compounds 1−5 results from the experimental error (±10 ppm) connected with the low precision of peak picking: the effect of substituent R is expected to be vanishingly small. d19F NMR signals for R in compound 5.

spectrum of compound 3 plotted at three selected temperatures is shown in Figure 3 as an example.

Figure 4. 1/T plots of (a) the 1H and (b) the 13C NMR chemical shifts for compound 3. The orbital contributions (δorb, temperature independent) of the NMR chemical shifts are estimated from an extrapolation to the limit 1/T = 0. The values of the HF contributions (δHF) for individual atoms are shown for a temperature of 293 K. The values obtained for the 1H and 13C NMR chemical shifts are affected by an estimated error of about ±0.2 ppm and ±1 ppm, respectively. (The estimated error for H2/H6 and for C2/C6−which have broad signals because the atoms are located in the near vicinity of the paramagnetic center−amount to up to ±1 ppm and ±5 ppm, respectively).

As expected, the most substituent-affected are the hydrogen atoms H2 and H4 in positions ortho to the substituent R. The difference in total 1H NMR shift (δtot H2/H4) between 2 and 3 amounts to approximately 2.5−3.5 ppm, with a similar difference (∼2 ppm) obtained for their hyperfine contributions (δHF H2/H4), as highlighted in bold in Table 3. This is nicely reproduced by the DFT calculations, vide inf ra. The trend parallels that observed for δorb H2/H4 and seems to be connected with an overall depletion of the electron density in the pyridine ring induced by the strong CN electron acceptor in compound

Figure 3. Portion of the 13C NMR spectrum of compound 3 plotted for temperatures 313, 293, and 273 K. The 13C NMR signals are assigned to individual atoms, and the 13C NMR signals of the countercation are labeled with an accent. In order to unequivocally distinguish the resonances of C5 from those of CN, the 1H coupled (no composite pulse decouplingCPD) 13C NMR spectrum was measured at 293 K, as shown in orange. E

DOI: 10.1021/acs.inorgchem.7b02440 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Table 5. Selected DFT-Calculated (SO-ZORA, PBE0/TZ2P) and Experimental 1H and 13C NMR Chemical Shifts (δ in ppm) for Ruthenium Compounds 2 and 3 in Dimethylformamide (DMF) at 293 Ka Compound 2 2c (SO-ZORA) H2

H4

C2

C3

a

δ δHF δtot δorb δHF δtot δorb δHF δtot δorb δHF δtot orb

Compound 3 Experiment

2c (SO-ZORA)

Experiment

δvac

Δsolv

δsolv

2

δvac

Δsolv

δsolv

3

+9.5 −6.3 +3.2 +6.2 +5.5 +11.7 +160 −90 +70 +119 −5 +114

-0.6 -8.0 -8.6 +1.2 -5.2 -4.0 -5 +12 +7 +14 -19 -5

+8.9 −14.3 −5.4 +7.4 +0.3 +7.7 +155 −78 +77 +133 −24 +109

+7.3 −14.9 −7.6 +7.2 −1.7 +5.5 +164 −73 +92 +131 −23 +108

+9.9 −2.7 +7.2 +7.0 +9.1 +16.1 +163 −100 +63 +98 +3 +101

-0.6 -8.2 -8.8 +0.8 -7.3 -6.5 -4 +10 +6 +7 -28 -21

+9.3 −10.9 −1.6 +7.8 +1.8 +9.6 +159 −90 +69 +105 −25 +80

+7.7 −12.7 −5.0 +8.8 +0.1 +8.9 +162 −74 +89 +108 −25 +83

For computational details, see Section 2: Methods.

the change in the total molecular charge) is counterbalanced by somewhat larger errors for more distant nuclei. The selected NMR data for compounds 2 and 3 summarized in Table 5 demonstrate good agreement between the experimental and theoretical 1H and 13C NMR chemical shifts (for additional data, see Tables S5−S9). The solvent effect on the NMR chemical shift was estimated as the difference between the NMR shift calculated by using the implicit solvent model and that calculated for the identical structure in vacuo. The Δsolv values for individual hydrogen and carbon atoms in compound 3 (Table 5) can be compared qualitatively with the distribution of spin density in this system shown in Figure 5.

3 as compared to that in 2 (see Figure S7 in the Supporting Information). To interpret the hyperfine effects on the 1H NMR resonances induced by the 3-R-substituent, we resort to the DFT calculations of the EPR and NMR parameters found in Section 3.3 and the bonding analysis in Section 3.4. In passing, note that the substituent effects on the 13C NMR chemical shifts (Table 4) are rather marginal except for the “closed-shell-like” orbital shift of the C3 atom, which is wellknown for these substituted aromatic systems (3-substituted pyridines).57 The 19F NMR spectra of compound 5 recorded at several low temperatures are shown in Figure S8. The unequivocal assignment of the two 19F resonances, belonging to the coordination anionic complex [RuIIICl4(DMSO)(3-F-Py)]− and the counterbalancing organic cation [3-F-PyH]+, respectively (cf. Figure 2), was made possible by the two-step addition of the salt [3-F-PyH]+Cl− to a DMF-d7 solution of 5, as demonstrated in Figure S9. Note that the shielding HF contribution (δHF) to the 19F NMR chemical shift of R in compound 5 parallels those of the 13C NMR resonances of the R substituents in compounds 2−4; Table 4. 3.3. DFT Calculations That Reproduce and Help To Interpret the Experimental NMR Data: Orbital Shifts and Solvent Effects on the Hyperfine NMR Shifts. Computational Method and Effects of Solvent. DFT calculations were performed to reproduce and help to interpret the experimental NMR chemical shifts and to investigate the electronic factors that influence the hyperfine contributions (δHF) to the total NMR chemical shifts described in Section 3.2. The PBE0/ TZ2P/SO-ZORA/COSMO approach was employed as calibrated in our previous work.20 The 1H and 13C NMR shielding constants of benzene38,58 were used to convert the shielding constants to NMR chemical shifts by using equations described in Section 2: Methods. For the calculated NMR chemical shifts of closed-shell Rh(III) analogs, see Table S3. The closed-shell Ru(II) model of compound 3 was used to calculate the orbital NMR chemical shifts (δorb) by using the two-component approach (2c, SO-ZORA), and the effect of the closed-shell approximation was evaluated by calculating the open-shell Ru(III) system at the scalar-relativistic (1c, ZORA) level; see Table S4. The substantial improvement for atoms located near the paramagnetic Ru center (which are affected by

Figure 5. Visualization of the 1c spin density (ZORA/PBE0/TZ2P level, isosurfaces at 0.00005 au) calculated for compound 3 in vacuum and by using the implicit model of DMF solvent (COSMO). The excess of the α and the β electronic spin state is shown in blue and red, respectively.

Clearly, the identified solvent effects result from redistribution of the spin density in the system brought about by the effects of the environment. The solvent stabilizes the lone pairs of electrons on the chlorine atoms, which somewhat reduces the spin polarization in the π-space of the pyridine ligands. Generally, solvent effects (simulated by using COSMO) are greater for the hyperfine NMR shifts than for the orbital NMR F

DOI: 10.1021/acs.inorgchem.7b02440 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry contributions. The largest hyperfine 13C NMR solvent effect originating in the reduced spin polarization in the π-space of the pyridine ligand is observed for the C4 atom in compound 3 (ΔHF solv = +33 ppm, Table S8). The solvent effects on the C3 (−30 ppm) and C5 (−23 ppm) HF NMR shifts have similar magnitudes, and the reduced spin polarization in carbon πspace is accompanied here by more delocalized negative spin density in the σC2−C3 and σC6−C5 space (Figure 5). The largest 1 H NMR solvent effects (approximately −10 ppm) are observed for atoms H2 and H6, although the mechanism is not clearly apparent from the visualized spin density. This case is analyzed in detail in Section 3.5. In summary, we particularly highlight the significant solvent effect on δHF H2/H6 and the signinverting solvent effect on δHF C3/C5, with the deshielding effect in vacuo changing to a shielding contribution in solvent.20 This demonstrates the crucial role of the environment in the calculation of NMR parameters for open-shell systems. Effect of the Conformation of the 3-R-Pyridine Ligand Relative to DMSO on the Hyperfine NMR Shift. To evaluate the sensitivity of the hyperfine NMR shielding to the relative orientation of the pyridine and DMSO ligands discussed in Section 3.1, the EPR and NMR parameters (SO-ZORA/PBE0/ TZ2P level) were calculated for compound 3 in the two distinct conformations shown in Figure 6.

character of the metal−ligand bonding, EDA linked to the contributions from natural orbitals for chemical valence (NOCV) was used.36,59,37 The four most important EDANOCV channels separated into α and β parts for the Ru−N bond in the anti conformation of 3 are shown in Figure 7. The

Figure 7. Four most important NOCV channels for the formation of the Ru−N bond in the anti conformation of compound 3 with the corresponding contributions to the EDA orbital term (E) and the charge transfer (ρ) calculated at the ZORA level (1c/PBE0/TZ2P/ vacuum). The channels are ordered according to spatial similarity (note that in the output of the analysis NOCV2α and NOCV3β are switched based on the energy criterion; see Supporting Information). The value 0.0005 au was used to plot the iso-surface of the electron deformation density. The NOCV analysis for compound 2 is shown in Figure S10 and Table S11 in the Supporting Information.

first NOCV channel, NOCV1, represents classical donation of the lone pair of electrons of the nitrogen atom to the metal center (σ-bonding).20,60 This type of σ-donation has been demonstrated for some classes of compounds to be spin selective, thus generating a spin imbalance in the donating ligands.61,62 However, the α and β NOCV1 channels are almost equally important for the Ru−N bonding in compound 3 (see Figure 7) and can hardly rationalize the ligand NMR observations. The π-back bonding related purely to the charge transfer from the d atomic orbitals (AO) of the metal to the p AO of the nitrogen is shown in the second and third channels. In NOCV2, a clear difference between the α and β contributions is identified. As the β-state is more involved in the back-donation from the metal to the ligand moiety, this should result in an overabundance of β spin electron density in the ligand, as is reflected in the predominant hyperfine shielding of the ligand atoms (cf. Tables 3 and 4 and Figure 5). In parallel, the polarization of the SOMO (single-occupied molecular orbital, derived from the metal dxy AO) toward the lone electron pairs of the Cl atoms and the nitrogen atom through the π-space identified in the third channel contributes additionally to the α/β imbalance in the pyridine system. Note in passing that this phenomenon somewhat resembles the propagation of spin−orbit HALA NMR effects63,64 through the σ- and π-spaces of the metal−ligand bond.65 The last channel considered, NOCV4, shows the very small contribution arising from the through-space polarization existing between the Cl ligands and the aromatic moiety, with a vanishingly small difference between α and β. The larger involvement identified for the β electronic state in the formation of the metal−ligand

Figure 6. Two distinct conformations (defined as approximately anti and syn subspaces, respectively, and depicted on opposite sides of the magenta line, which represents the projection of a plane perpendicular to the page) for compound 3 and the δHF values calculated for the hydrogen and carbon atoms nearest to the ruthenium center (SOZORA/PBE0/TZ2P level).

Clearly, the pseudoequivalent atoms near the paramagnetic center (H2−H6 and C2−C6) are influenced by their positions relative to the oxygen atom of the trans DMSO ligand. A switch between the two conformations occurs when the ligand is rotated between the syn and anti subspaces defined by a plane perpendicular to the pyridine ring and referenced relative to the substituent at position 3 (see the schematic magenta line, which represents the projection of a plane perpendicular to the plane of the page, in Figure 6). In passing, note that the orbital shifts are almost independent of the conformation. However, averaging the δHF values of the two conformations has only a marginal effect on the deviations between the calculated and experimental values. The effect of the Ru−N bond vibration, which can be simulated by bond stretching as shown for NAMI in ref 12, seems to be of greater importance. The syn or anti conformation was used for the production calculations as specified in the following sections. 3.4. Role of Back-donation in the Metal−Ligand Bonding for the Spin-Polarization Mechanism. The spin polarization in the pyridine moiety is naturally linked to the character of the metal−ligand bond. To get an insight into the G

DOI: 10.1021/acs.inorgchem.7b02440 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Table 6. Fermi-Contact (FC), Spin-Dipole (SD), and Paramagnetic Nuclear-Spin−Electron-Orbit (PSO) Contributions to Aiso (in kHz) for the Individual Hydrogen and Carbon Atoms of Compound 3, Calculated at the 4c Level (PBE0/dyall-VTZ/upcJ-2) with the COSMO Model of DMF Solventa Compound 3 Aiso [kHz]

δHFi [ppm]

δHFa [ppm]

δHF [ppm]

atom

H2

H4

H5

H6

C2

C3

C4

C5

C6

FC SD PSO Total FC SD PSO Total FC SD PSO Total FC SD PSO Total Experiment

−44 −21 −151 −216 −1.3 −0.6 −4.5 −6.4 −0.1 −4.8 −0.2 −5.0 −1.4 −5.4 −4.7 −11.4 −12.7

+144 −17 −55 +73 +4.3 −0.5 −1.6 +2.1 −0.1 −1.3 0.0 −1.4 +4.2 −1.8 −1.7 +0.7 +0.1

−210 −10 −65 −285 −6.2 −0.3 −1.9 −8.4 0.1 −1.8 −0.1 −1.7 -6.1 −2.1 −2.0 −10.1 −8.7

−28 −25 −162 −214 −0.8 −0.7 −4.8 −6.4 0.0 −5.3 −0.2 −5.6 −0.9 −6.0 −5.0 −11.9 −12.7

−545 −55 +10 −588 −64 −6 +1 −69 2 −11 0 −9 −63 −17 +1 −78 −74

−147 +4 −33 −175 −17 +1 −4 −20 0 −2 0 −2 −17 −1 −4 −22 −25

−186 −21 −32 −239 −22 −2 −4 −28 1 −5 0 −4 −21 −7 −4 −32 −26

−100 +5 −36 −130 −12 +1 −4 −15 0 −2 0 −2 −12 −1 −4 −17 −17

−549 −52 −4 −600 −64 −6 0 −71 2 −10 0 −9 −63 −16 −1 −80 −69

a The δHFi/FC, δHFi/SD, and δHFi/PSO values (in ppm) were obtained by recalculating the Aiso contributions to the ligand HF NMR shifts (δHFi is L HF/FC HF/SD ,δ , and δHF/PSO values represent the total individual contributions to δHF, equivalent to the contact shift, δcon L ; see Section 2) at 293 K. The δ HFi (equivalent to the pseudo-contact shift, δpc including both δHFa L L ; see Section 2) and δL . For the definitions of FC, SD, and PSO, see eqs 3−5 in Section 2. For additional data, see the Supporting Information.

moiety to reach the spectator nucleus. It should be highlighted that the FC is the only mechanism contributing to the isotropic NMR shifts in the absence of SO coupling (nonrelativistic or scalar-relativistic 1c level) as the SD and PSO shift terms arise exclusively in the SO regime (see Table S12). In passing, note that the SO coupling (both one and two electron) is the only source of an angular magnetic moment of electrons. In the presence of SO coupling, the PSO term contributes to the HF shift due to an additional spin polarization (imbalance in the occupied molecular spin−orbitals contributed by atomic orbitals of both the metal and the spectator ligand atom), whereas the contribution to the HF shift made by the SD term is linked to the arising anisotropy of the g-tensor in the SO regime (g is fully isotropic in the absence of spin−orbit coupling). The FC term dominates the Aiso values for all of the atoms of the pyridine ring except H2 and H6 (Table 6, line 1). These two hydrogen atoms are very near the paramagnetic metal center, with greatly concentrated α-spin density in the RuCl4 fragment (predominantly in the metal-based dxy orbital; see Figure 5)20 and a spin structure notably affected by the relativistic SO effects (Figure 8a). As a consequence, the HF NMR shifts of H2 and H6 are dominated by the paramagnetic spin−orbit (PSO) and spin-dipole (SD) mechanisms in the isotropic and anisotropic part, respectively. As mentioned above, the hyperfine isotropic PSO shielding contribution (δHFi/PSO ∼ −4.5 ppm) arises exclusively in the spin−orbit regime (here fully relativistic 4c, see Table 6) and spreads from the transition-metal center to the nearby hydrogen atoms H2 and H6. To demonstrate the link between this HFi/PSO contribution and the spin−orbit-induced redistribution of the spin density, a picture of this phenomenon is provided in Figure 8a. The shielding contribution with similar magnitudes for H2 and H6 is observed for the ASD ani contribution to the HF shielding (δHFa/SD ∼ −5 ppm) which arises in the spin−orbit

bond and its effect on the hyperfine interactions in the pyridine ligand are analyzed in detail in Section 3.5. 3.5. Analyzing the Nature of Hyperfine NMR Shifts. Hyperfine 1H NMR Shifts. Somewhat larger experimental hyperfine shielding effects (δHF in Table 3, low-frequency shift) for the H2 and H4 atoms in compound 2 as compared with those observed in 3 are mentioned in Section 3.2 and highlighted in Table 3. This behavior is correctly reproduced by the DFT calculations as shown in Table 5 and seems to reflect a higher electron density around the atoms neighboring the substituent R in 2 (positive inductive effect and hyperconjugation of the Me group) compared to that in 3 (negative inductive and mesomeric effect of the CN group); see Figure S7. Therefore, the observed higher electron density in 2 is suggested to be more susceptible to spin-polarization effects through π-back-donation in the Ru−N bond as discussed in Section 3.4. To further understand the physical origin of the HF shifts for the individual ligand atoms and to create a link with the chemical structure, the Aiso values (recalculated also to δHFi) were analyzed using the program ReSpect. First, the isotropic contributions from Fermi-contact (FC), spin-dipole (SD), and paramagnetic spin−orbit (PSO) terms to the individual HF coupling constants and HF NMR shifts were calculated for compound 3 (see the Aiso and δHFi values in Table 6). The NMR shifts arising from the anisotropic parts of the HF coupling constants, δHFa, were then added to the isotropic values to obtain the total HF shift contributions of the individual terms: δFC, δSD, and δPSO. These terms are discussed in the following text as they represent contributions from distinct physical mechanisms. The main Fermi-contact mechanism is related to the propagation of spin polarization from the unpaired molecular spin−orbital at the transition-metal center through the metal− ligand bond, with further polarization steps inside the ligand H

DOI: 10.1021/acs.inorgchem.7b02440 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

previously.20,66 However, nonvanishing contributions from the δHF/SD termwhich arises mainly in the anisotropic partfor atoms in the ortho (C2 and C6) and para (C4) positions relative to the metal−ligand bond indicate the role of the relativistic effects, which originate at the metal atom and propagate to the ligand atoms. Similarly to that for the FC mechanism,67,58 the SD pattern for 13C NMR resembles an alternating resonance structure. This is in clear contrast to 1H NMR, where a gradual decrease in the SD contributions as a function of distance from the metal center was observed. The alternating pattern of the SD 13C shielding contributions could indicate a role involving carbon p atomic orbitals for the local magnetization around the spectator atoms. We expect this phenomenon to be revealed by inspecting the molecular currents, which is beyond the scope of this work. Note also that approximately 30% of the HF/SD shielding contribution comes from the Aiso part; the link between it and the magnetization that governs the HF/FC term is therefore apparent (Table 6). Understanding this mechanism in detail requires a somewhat different type of analysis, and such work is underway in our laboratories. In Section 3.3, the significant solvent effects on the δHF values of the carbon atoms in the pyridine ring are demonstrated. These effects are dominated by the HF/FC contribution (Table S13), which is linked directly to the solvent-induced rearrangement of the spin density, as demonstrated in Figure 8b. In contrast to the above-mentioned spin−orbit effects on the FC 13 C NMR shifts, which propagate predominantly via the πspace of the Ru−N bond (Figure 8a, cf. back-bonding in Section 3.4), the solvent effects also involve a significant σ-bond component (Figure 8b). We suggest that the difference in the alteration of the spin-polarization mechanism (σ vs π) caused by spin−orbit and solvent effects results in opposite signs for the corresponding FC contributions (compare hyperfine 13C shifts in Figures 8a and 8b). Note, however, the different magnitudes of these effects in the Ru(III) compounds 2 and 3 investigated.

Figure 8. Visualization of the spin−orbit-induced (top, threshold 10−5 a.u.) and solvent-induced (bottom, threshold 2 × 10−5 a.u.) deformations of the z-component of the z-magnetized 4c spin density (PBE0/vtz/upcJ-2) for the syn-conformation of compounds 2 (left) and 3 (right). The SO-induced and solvent-induced contributions to the hyperfine Fermi-contact (HF/FC) 13C (black) and 1H (orange) NMR shifts (HF/SD and HF/PSO contributions in the SO regime are shown in parentheses) are given in ppm. For full data sets, see Tables S12−S13. Increases in parallel and antiparallel density are shown in blue and red, respectively.

regime as a result of the emerging anisotropy of the g-tensor.2 Whether this SD shift for hydrogen atoms stems exclusively from the finite distribution of magnetization in the RuCl4 core (involving particularly the lone pairs of electrons of the chlorine atoms, cf. Figure 8a) and interacts with the spectator nucleus through space, or contributions from the spin polarization at the neighboring carbon atoms (C−H groups) of the ligand play some role remains to be shown. We have recently demonstrated the crucial influence of solvent effects on the hyperfine 1H and 13C NMR shifts in closely related complexes of Ru(III) with 4-substituted pyridines.20 To further elucidate the nature of these effects, we perform here a detailed analysis of the individual contributions to the hyperfine 1H shifts (Table S13). As shown schematically in Figure 8b, all of the hydrogen atoms analyzedexcept H5 at the “passive” meta position relative to the polarizable RuCl4 coreare solvent-shielded (lowfrequency NMR shift) by about (−7)−(−8) ppm, of which approximately −6 ppm comes from the FC term. This indicates a slight change in the β/α balance of the spin density in the pyridine ring in the solvent environment and reflects the spinpolarization mechanism discussed in Section 3.4.20 Hyperfine 13C NMR Shifts. In contrast to 1H NMR, the hyperfine 13C NMR shifts for all of the atoms investigated in compound 3 are dominated by the δHFi/FC term with the characteristic pattern alternating as a function of the bonddistance (number of bonds) from the Ru−N bond described

4. CONCLUSIONS In this contribution, a systematic study of the 1H and 13C NMR chemical shifts, δ, in a series of newly prepared pyridine-based Ru(III) complexes was performed. The geometry of three compounds was determined by X-ray diffraction analysis, and the structures of all of the compounds were optimized by using relativistic DFT methods. The experimental NMR spectra recorded at several temperatures enabled the individual chemical shifts to be broken down into the orbital (δorb) and hyperfine (δHF) contributions. Several relativistic DFT approaches to calculate the electronic g-tensors, hyperfine coupling tensors (A), and total NMR chemical shifts were evaluated. The four-component DKS approach was used to calculate and analyze the physical contributions to A and, subsequently, δHF for the individual ligand atoms. Although the Fermi-contact term is shown to dominate the hyperfine contributions to the NMR chemical shifts (related to the A coupling constants) for most of the pyridine ligand atoms, the HF shielding contributions for the hydrogen atoms H2 and H6 in the ortho positions are governed by the isotropic PSO and anisotropic SD terms. The FC 1H shift is probably quenched here as a result of opposing contributions from the σ and π coupling pathways. We also show that the alternating pattern for the SD 13C shielding contributions in the pyridine ring could indicate a role played by the local magnetization I

DOI: 10.1021/acs.inorgchem.7b02440 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

(5) Moon, S.; Patchkovskii, S. First-Principles Calculations of Paramagnetic NMR Shifts. In Calculation of NMR and EPR Parameters; Kaupp, M., Bühl, M., Malkin, V., Eds.; Wiley-VCH Verlag, 2004; pp 325−338. (6) Autschbach, J. Chapter One - NMR Calculations for Paramagnetic Molecules and Metal Complexes. In Annual Reports in Computational Chemistry; Dixon, D. A., Ed.; Elsevier, 2015; Vol. 11, pp 3−36. (7) Pennanen, T. O.; Vaara, J. Density-Functional Calculations of Relativistic Spin-Orbit Effects on Nuclear Magnetic Shielding in Paramagnetic Molecules. J. Chem. Phys. 2005, 123, 174102. (8) Hrobárik, P.; Reviakine, R.; Arbuznikov, A. V.; Malkina, O. L.; Malkin, V. G.; Kö hler, F. H.; Kaupp, M. Density Functional Calculations of NMR Shielding Tensors for Paramagnetic Systems with Arbitrary Spin Multiplicity: Validation on 3d Metallocenes. J. Chem. Phys. 2007, 126, 024107. (9) Pennanen, T. O.; Vaara, J. Nuclear Magnetic Resonance Chemical Shift in an Arbitrary Electronic Spin State. Phys. Rev. Lett. 2008, 100, 133002. (10) Van den Heuvel, W.; Soncini, A. NMR Chemical Shift as Analytical Derivative of the Helmholtz Free Energy. J. Chem. Phys. 2013, 138, 054113. (11) Van den Heuvel, W.; Soncini, A. NMR Chemical Shift in an Electronic State with Arbitrary Degeneracy. Phys. Rev. Lett. 2012, 109, 073001. (12) Komorovsky, S.; Repisky, M.; Ruud, K.; Malkina, O. L.; Malkin, V. G. Four-Component Relativistic Density Functional Theory Calculations of NMR Shielding Tensors for Paramagnetic Systems. J. Phys. Chem. A 2013, 117, 14209−14219. (13) Gendron, F.; Sharkas, K.; Autschbach, J. Calculating NMR Chemical Shifts for Paramagnetic Metal Complexes from FirstPrinciples. J. Phys. Chem. Lett. 2015, 6, 2183−2188. (14) Vaara, J.; Rouf, S. A.; Mareš, J. Magnetic Couplings in the Chemical Shift of Paramagnetic NMR. J. Chem. Theory Comput. 2015, 11, 4840−4849. (15) Alessio, E. Thirty Years of the Drug Candidate NAMI-A and the Myths in the Field of Ruthenium Anticancer Compounds: A Personal Perspective. Eur. J. Inorg. Chem. 2017, 2017, 1549−1560. (16) Bergamo, A.; Sava, G. Linking the Future of Anticancer MetalComplexes to the Therapy of Tumour Metastases. Chem. Soc. Rev. 2015, 44, 8818−8835. (17) Allardyce, C. S.; Dyson, P. J. Metal-Based Drugs That Break the Rules. Dalton Trans. 2016, 45, 3201−3209. (18) Rastrelli, F.; Bagno, A. Predicting the NMR Spectra of Paramagnetic Molecules by DFT: Application to Organic Free Radicals and Transition-Metal Complexes. Chem. - Eur. J. 2009, 15, 7990−8004. (19) Autschbach, J.; Patchkovskii, S.; Pritchard, B. Calculation of Hyperfine Tensors and Paramagnetic NMR Shifts Using the Relativistic Zeroth-Order Regular Approximation and Density Functional Theory. J. Chem. Theory Comput. 2011, 7, 2175−2188. (20) Novotný, J.; Sojka, M.; Komorovsky, S.; Nečas, M.; Marek, R. Interpreting the Paramagnetic NMR Spectra of Potential Ru(III) Metallodrugs: Synergy between Experiment and Relativistic DFT Calculations. J. Am. Chem. Soc. 2016, 138, 8432−8445. (21) Shokhirev, N. V.; Walker, F. A. Analysis of the Temperature Dependence of the 1H Contact Shifts in Low-Spin Fe(III) Model Hemes and Heme Proteins: Explanation of “Curie” and “Anti-Curie” Behavior within the Same Molecule. J. Phys. Chem. 1995, 99, 17795− 17804. (22) Banci, L.; Bertini, I.; Luchinat, C.; Pierattelli, R.; Shokhirev, N. V.; Walker, F. A. Analysis of the Temperature Dependence of the 1H and 13C Isotropic Shifts of Horse Heart Ferricytochrome c: Explanation of Curie and Anti-Curie Temperature Dependence and Nonlinear Pseudocontact Shifts in a Common Two-Level Framework. J. Am. Chem. Soc. 1998, 120, 8472−8479. (23) te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Fonseca Guerra, C.; van Gisbergen, S. J. A.; Snijders, J. G.; Ziegler, T. Chemistry with ADF. J. Comput. Chem. 2001, 22, 931−967.

around the spectator atoms and involving the p atomic orbitals of carbon. It remains to be shown that this phenomenon can be deciphered from the molecular currents. In order to understand the experimental hyperfine 1H and 13 C NMR shifts, the relativistic spin−orbit and solvent effects on the distribution of spin density and the hyperfine contributions to the NMR shifts are demonstrated. Further investigations of long-range substituent effects and spin−orbit relativistic contributions to the hyperfine NMR shifts are underway in our laboratories.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.7b02440. Methods, experimental and calculated NMR data, additional figures, and Cartesian coordinates (PDF) Accession Codes

CCDC 1573286−1573288 contain the supplementary crystallographic data for this paper. These data can be obtained free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by emailing [email protected], or by contacting The Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: +44 1223 336033.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Jan Novotný: 0000-0002-1203-9549 David Přichystal: 0000-0001-7682-5461 Stanislav Komorovsky: 0000-0002-5317-7200 Radek Marek: 0000-0002-3668-3523 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by the Czech Science Foundation (15-09381S) and partly by the Ministry of Education of the Czech Republic (LQ1601) and the SASPRO Program (1563/ 03/02), cofinanced by the European Union and the Slovak Academy of Sciences. The CIISB research structure project LM2015043 funded by MEYS CR is gratefully acknowledged for the financial support of the measurements at the CF Josef Dadok National NMR Center and CF X-ray Diffraction and Bio-SAXS. Computational resources were provided by the CESNET (LM2015042) and the CERIT Scientific Cloud (LM2015085).



REFERENCES

(1) Symons, M. C. R. Chemical and Biochemical Aspects of Electron Spin Resonance Spectroscopy; Van Nostrand Reinhold Inc.,U.S.: New York, 1978. (2) Mabbs, F. E.; Collison, D. Electron Paramagnetic Resonance of d Transition Metal Compounds; Elsevier, 2013. (3) Mar, G. N. L.; Horrocks, W. D.; Holm, R. H. NMR of Paramagnetic Molecules: Principles and Applications; Elsevier: New York, 2013. (4) Bertini, I.; Luchinat, C.; Parigi, G. Solution NMR of Paramagnetic Molecules: Applications to Metallobiomolecules and Models; Elsevier, 2001. J

DOI: 10.1021/acs.inorgchem.7b02440 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry (24) Repisky, M.; Komorovsky, S.; Malkin, V. G.; Malkina, O. L.; Kaupp, M.; Ruud, K.; Bast, R.; Ekstrom, U.; Kadek, M.; Knecht, S.; Konečný, L.; Malkin, E.; Malkin Ondík, I. Relativistic Spectroscopy DFT Program ReSpect, Developer Version 4.0.0; http://rel-qchem.sav.sk. (25) Remigio, R. D.; Repisky, M.; Komorovsky, S.; Hrobarik, P.; Frediani, L.; Ruud, K. Four-Component Relativistic Density Functional Theory with the Polarisable Continuum Model: Application to EPR Parameters and Paramagnetic NMR Shifts. Mol. Phys. 2017, 115, 214−227. (26) Alessio, E.; Balducci, G.; Calligaris, M.; Costa, G.; Attia, W. M.; Mestroni, G. Synthesis, Molecular Structure, and Chemical Behavior of Hydrogen Trans-Bis(Dimethyl Sulfoxide)Tetrachlororuthenate(III) and Mer-Trichlorotris(Dimethyl Sulfoxide)Ruthenium(III): The First Fully Characterized Chloride-Dimethyl Sulfoxide-Ruthenium(III) Complexes. Inorg. Chem. 1991, 30, 609−618. (27) Vicha, J.; Patzschke, M.; Marek, R. A Relativistic DFT Methodology for Calculating the Structures and NMR Chemical Shifts of Octahedral Platinum and Iridium Complexes. Phys. Chem. Chem. Phys. 2013, 15, 7740−7754. (28) Vicha, J.; Novotny, J.; Straka, M.; Repisky, M.; Ruud, K.; Komorovsky, S.; Marek, R. Structure, Solvent, and Relativistic Effects on the NMR Chemical Shifts in Square-Planar Transition-Metal Complexes: Assessment of DFT Approaches. Phys. Chem. Chem. Phys. 2015, 17, 24944−24955. (29) Bühl, M.; Kabrede, H. Geometries of Transition-Metal Complexes from Density-Functional Theory. J. Chem. Theory Comput. 2006, 2, 1282−1290. (30) Adamo, C.; Barone, V. Toward Reliable Density Functional Methods without Adjustable Parameters: The PBE0Model. J. Chem. Phys. 1999, 110, 6158−6170. (31) Adamo, C.; Scuseria, G. E.; Barone, V. Accurate Excitation Energies from Time-Dependent Density Functional Theory: Assessing the PBE0Model. J. Chem. Phys. 1999, 111, 2889−2899. (32) Schäfer, A.; Huber, C.; Ahlrichs, R. Fully Optimized Contracted Gaussian Basis Sets of Triple Zeta Valence Quality for Atoms Li to Kr. J. Chem. Phys. 1994, 100, 5829−5835. (33) Andrae, D.; Häußermann, U.; Dolg, M.; Stoll, H.; Preuß, H. Energy-Adjustedab Initio Pseudopotentials for the Second and Third Row Transition Elements. Theor. Chim. Acta 1990, 77, 123−141. (34) TURBOMOLE V6.2 2010, a Development of University of Karlsruhe and Forschungszentrum Karlsruhe GmbH, 1989−2007, TURBOMOLE GmbH, since 2007; http://www.turbomole.com. (35) Klamt, A.; Schüürmann, G. COSMO: A New Approach to Dielectric Screening in Solvents with Explicit Expressions for the Screening Energy and Its Gradient. J. Chem. Soc., Perkin Trans. 2 1993, 799−805. (36) Mitoraj, M.; Michalak, A. Natural Orbitals for Chemical Valence as Descriptors of Chemical Bonding in Transition Metal Complexes. J. Mol. Model. 2007, 13, 347−355. (37) Frenking, G.; Matthias Bickelhaupt, F. The EDA Perspective of Chemical Bonding. In The Chemical Bond; Frenking, G., Shaik, S., Eds.; Wiley-VCH Verlag GmbH & Co. KGaA, 2014; pp 121−157. (38) Standara, S.; Bouzková, K.; Straka, M.; Zacharová, Z.; Hocek, M.; Marek, J.; Marek, R. Interpretation of Substituent Effects on 13C and 15N NMR Chemical Shifts in 6-Substituted Purines. Phys. Chem. Chem. Phys. 2011, 13, 15854−15864. (39) ADF2014, SCM, Theoretical Chemistry; Vrije Universiteit: Amsterdam, The Netherlands, http://www.scm.com. (40) van Lenthe, E.; Snijders, J. G.; Baerends, E. J. The Zero-order Regular Approximation for Relativistic Effects: The Effect of Spin− orbit Coupling in Closed Shell Molecules. J. Chem. Phys. 1996, 105, 6505−6516. (41) Saue, T. Relativistic Hamiltonians for Chemistry: A Primer. ChemPhysChem 2011, 12, 3077−3094. (42) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865−3868. (43) Adamo, C.; Barone, V. Toward Chemical Accuracy in the Computation of NMR Shieldings: The PBE0Model. Chem. Phys. Lett. 1998, 298, 113−119.

(44) Güell, M.; Luis, J. M.; Solà, M.; Swart, M. Importance of the Basis Set for the Spin-State Energetics of Iron Complexes. J. Phys. Chem. A 2008, 112, 6384−6391. (45) Autschbach, J. The Role of the Exchange-Correlation Response Kernel and Scaling Corrections in Relativistic Density Functional Nuclear Magnetic Shielding Calculations with the Zeroth-Order Regular Approximation. Mol. Phys. 2013, 111, 2544−2554. (46) Kaupp, M.; Köhler, F. H. Combining NMR Spectroscopy and Quantum Chemistry as Tools to Quantify Spin Density Distributions in Molecular Magnetic Compounds. Coord. Chem. Rev. 2009, 253, 2376−2386. (47) Repisky, M.; Komorovsky, S.; Malkin, E.; Malkina, O. L.; Malkin, V. G. Relativistic Four-Component Calculations of Electronic g-Tensors in the Matrix Dirac−Kohn−Sham Framework. Chem. Phys. Lett. 2010, 488, 94−97. (48) Malkin, E.; Repisky, M.; Komorovsky, S.; Mach, P.; Malkina, O. L.; Malkin, V. G. Effects of Finite Size Nuclei in Relativistic FourComponent Calculations of Hyperfine Structure. J. Chem. Phys. 2011, 134, 044111. (49) Gohr, S.; Hrobárik, P.; Repisky, M.; Komorovsky, S.; Ruud, K.; Kaupp, M. Four-Component Relativistic Density Functional Theory Calculations of EPR g- and Hyperfine-Coupling Tensors Using Hybrid Functionals: Validation on Transition-Metal Complexes with Large Tensor Anisotropies and Higher-Order Spin−Orbit Effects. J. Phys. Chem. A 2015, 119, 12892−12905. (50) Jensen, F. The Basis Set Convergence of Spin-Spin Coupling Constants Calculated by Density Functional Methods. J. Chem. Theory Comput. 2006, 2, 1360−1369. (51) Dyall, K. G. Relativistic Double-Zeta, Triple-Zeta, and Quadruple-Zeta Basis Sets for the 4d Elements Y−Cd. Theor. Chem. Acc. 2007, 117, 483−489. (52) Dyall, K. G.; Faegri, K., Jr Introduction to Relativistic Quantum Chemistry; Oxford University Press: Oxford, NY, 2007. (53) Haase, P. A. B.; Repisky, M.; Komorovsky, S.; Bendix, J.; Sauer, S. P. A. Relativistic DFT Calculations of Hyperfine Coupling Constants in the 5d Hexafluorido Complexes: [ReF6]2- and [IrF6]2-. Chem.Eur. J. 2017, DOI: 10.1002/chem.201704653. (54) Suturina, E. A.; Kuprov, I. Pseudocontact Shifts from Mobile Spin Labels. Phys. Chem. Chem. Phys. 2016, 18, 26412−26422. (55) Webb, M. I.; Chard, R. A.; Al-Jobory, Y. M.; Jones, M. R.; Wong, E. W. Y.; Walsby, C. J. Pyridine Analogues of the Antimetastatic Ru(III) Complex NAMI-A Targeting Non-Covalent Interactions with Albumin. Inorg. Chem. 2012, 51, 954−966. (56) Webb, M. I.; Wu, B.; Jang, T.; Chard, R. A.; Wong, E. W. Y.; Wong, M. Q.; Yapp, D. T. T.; Walsby, C. J. Increasing the Bioavailability of Ru(III) Anticancer Complexes through Hydrophobic Albumin Interactions. Chem. - Eur. J. 2013, 19, 17031−17042. (57) Retcofsky, H. L.; Friedel, R. A. Carbon-13 Nuclear Magnetic Resonance Studies of 3-Substituted Pyridines. J. Phys. Chem. 1968, 72, 290−293. (58) Pawlak, T.; Munzarová, M. L.; Pazderski, L.; Marek, R. Validation of Relativistic DFT Approaches to the Calculation of NMR Chemical Shifts in Square-Planar Pt2+ and Au3+ Complexes. J. Chem. Theory Comput. 2011, 7, 3909−3923. (59) Mitoraj, M. P.; Michalak, A.; Ziegler, T. A Combined Charge and Energy Decomposition Scheme for Bond Analysis. J. Chem. Theory Comput. 2009, 5, 962−975. (60) Mitoraj, M. P.; Zhu, H.; Michalak, A.; Ziegler, T. On the Origin of the Trans-Influence in Square Planar D8-Complexes: A Theoretical Study. Int. J. Quantum Chem. 2009, 109, 3379−3386. (61) Pritchard, B.; Autschbach, J. Theoretical Investigation of Paramagnetic NMR Shifts in Transition Metal Acetylacetonato Complexes: Analysis of Signs, Magnitudes, and the Role of the Covalency of Ligand−Metal Bonding. Inorg. Chem. 2012, 51, 8340− 8351. (62) Martin, B.; Autschbach, J. Kohn−Sham Calculations of NMR Shifts for Paramagnetic 3d Metal Complexes: Protocols, Delocalization Error, and the Curious Amide Proton Shifts of a High-Spin Iron(II) K

DOI: 10.1021/acs.inorgchem.7b02440 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Macrocycle Complex. Phys. Chem. Chem. Phys. 2016, 18, 21051− 21068. (63) Pyykko, P. Relativistic Effects in Structural Chemistry. Chem. Rev. 1988, 88, 563−594. (64) Pyykkö, P.; Görling, A.; Rösch, N. A Transparent Interpretation of the Relativistic Contribution to the N.M.R. “Heavy Atom Chemical Shift. Mol. Phys. 1987, 61, 195−205. (65) Novotný, J.; Vícha, J.; Bora, P. L.; Repisky, M.; Straka, M.; Komorovsky, S.; Marek, R. Linking the Character of the Metal−Ligand Bond to the Ligand NMR Shielding in Transition-Metal Complexes: NMR Contributions from Spin−Orbit Coupling. J. Chem. Theory Comput. 2017, 13, 3586−3601. (66) Cano, J.; Ruiz, E.; Alvarez, S.; Verdaguer, M. Spin Density Distribution in Transition Metal Complexes: Some Thoughts and Hints. Comments Inorg. Chem. 1998, 20, 27−56. (67) Standara, S.; Maliňaḱ ová, K.; Marek, R.; Marek, J.; Hocek, M.; Vaara, J.; Straka, M. Understanding the NMR Chemical Shifts for 6Halopurines: Role of Structure, Solvent and Relativistic Effects. Phys. Chem. Chem. Phys. 2010, 12, 5126−5139.

L

DOI: 10.1021/acs.inorgchem.7b02440 Inorg. Chem. XXXX, XXX, XXX−XXX