A prototypical study of NMR chemical shifts in biologically relevant heteroaromatic compounds containing a heavy halogen atom is presented for two isomers of halogen-substituted purines. Complete sets of (1)H-, (13)C- and (15)N-NMR chemical shifts are determined experimentally in solution. Experimental results are complemented by quantum-chemical calculations that provide understanding of the trends in the chemical shifts for the studied compounds and which show how different physical effects influence the NMR parameters. Chemical shifts for isolated molecules are calculated using density-functional theory methods, the role of solvent effects is studied using polarised continuum models, and relativistic corrections are calculated using the leading-order Breit-Pauli perturbation theory. Calculated values are compared with the experimental data and the effects of structure, solvent and relativity are discussed. Overall, we observe a good agreement of theory and experiment. We find out that relativistic effects cannot be neglected even in the chlorine species when aiming at high precision and a good agreement with the experimental data. Relativity plays a crucial role in the bromine and iodine species. Solvent effects are of smaller importance for (13)C shifts but are shown to be substantial for particular (15)N shifts. The test of method performance shows that the BLYP and B3LYP functionals provide the most reliable computational results after inclusion of the solvent and relativistic effects while BHandHLYP may--depending on atom in question--slightly improve but mostly deteriorate the data. Ab initio Hartree-Fock suffers from triplet instability in the Breit-Pauli relativistic part while MP2 provides no clear improvement over DFT in the nonrelativistic region. This work represents the first full application of the Breit-Pauli perturbation theory to an organic chemistry problem.

(1)H, (13)C and (15)N NMR studies of gold(III), palladium(II) and platinum(II) chloride complexes with picolines, [Au(PIC)Cl(3)], trans-[Pd(PIC)(2)Cl(2)], trans/cis-[Pt(PIC)(2)Cl(2)] and [Pt(PIC)(4)]Cl(2), were performed. After complexation, the (1)H and (13)C signals were shifted to higher frequency, whereas the (15)N ones to lower (by ca 80-110 ppm), with respect to the free ligands. The (15)N shielding phenomenon was enhanced in the series [Au(PIC)Cl(3)] < trans-[Pd(PIC)(2)Cl(2)] < cis-[Pt(PIC)(2)Cl(2)] < trans-[Pt(PIC)(2)Cl(2)]; it increased following the Pd(II) --> Pt(II) replacement, but decreased upon the trans --> cis-transition. Experimental (1)H, (13)C and (15)N NMR chemical shifts were compared to those quantum-chemically calculated by B3LYP/LanL2DZ + 6-31G**//B3LYP/LanL2DZ + 6-31G*.

Adenine, an essential building block of nucleic acids present in all living systems, can occur in several tautomeric forms. The phenomenon of tautomerism can be investigated by several experimental methods, including nuclear magnetic resonance. In this study, long-range (1)H-(13)C and (1)H-(15)N coupling constants for N-alkyl derivatives related to four tautomers of adenine are investigated in DMSO and DMF solutions. To investigate the structural dependence of the coupling constants and to understand how polarization propagates in the system, Fermi contact (FC) terms were calculated for the individual isomers and analyzed by using density functional theory (DFT), and the coupling pathways were visualized using real-space functions. The coupling electron deformation densities (CDD) of several (1)H-X (X = (13)C, (15)N) pairs are evaluated and compared. In order to analyze the CDD in more detail, a new approach to break down the CDD into contributions from Boys or Pipek-Mezey localized molecular orbitals (LMOs) has been developed. A similar approach has been applied to split the value of the FC contribution to the J coupling into the LMO contributions. On the basis of chemical concepts, the contributions of sigma-bonds, pi-electrons, and lone pairs of electrons are discussed. The lone pair of electrons at the nitrogen atom contributes significantly to the (1)H-C horizontal line(15)N coupling, whereas the (1)H-C=N-(13)C coupling is affected in a somewhat different way. Surprisingly, the contribution of the intervening C horizontal lineN bond to the FC term for (1)H-C=(15)N coupling originates exclusively in sigma-electrons, with a vanishingly small contribution calculated for the pi-electrons of this fragment. This behavior is rationalized by introducing the concept of "hard and soft J elements" derived from the polarizability of the individual components.