H–N Hydrogen Bond in a Protonated ... - ACS Publications

Article pubs.acs.org/IC

A Relativity Enhanced, Medium-Strong Au(I)···H−N Hydrogen Bond in a Protonated Phenylpyridine-Gold(I) Thiolate Raphael J. F. Berger,*,† Jürgen Schoiber,†,‡ and Uwe Monkowius*,‡ †

Fachbereich für Materialwissenschaften und Physik, Paris−Lodron Universität Salzburg, Jakob-Haringer-Str. 2a, A−5020 Salzburg, Austria ‡ Institute of Inorganic Chemistry, Johannes Kepler University Linz, Altenbergerstr. 69, 4040 Linz, Austria S Supporting Information *

ABSTRACT: Gold is an electron-rich metal with a high electronegativity comparable to that of sulfur. Hence, hydrogen bonds of the Au(I)···H−E (E = electronegative element) type should be possible, but their existence is still under debate. Experimental results are scarce and often contradictory. As guidance for possible preparative work, we have theoretically investigated (ppyH)Au(SPh) (ppy = 2-phenylpyridine) bearing two monoanionic ligands which are not strongly electronegative at the same time to further increase the charge density on the gold(I) atom. The protonated pyridine nitrogen atom in ppy is geometrically ideally suited to place a proton in close proximity to the gold atom in a favorable geometry for a classical hydrogen bond arrangement. Indeed, the results of the calculations indicate that the hydrogen bonded conformation of (ppyH)Au(SPh) represents a minimum geometry with bond metrics in the expected range for medium-strong hydrogen bonds [r(N−H) = 1.043 Å, r(H···Au) = 2.060 Å, a(N−H···Au) = 141.4°]. The energy difference between the conformer containing the H···Au bond and another conformer without a hydrogen bond amounts to 7.8 kcal mol−1, which might serve as an estimate of the hydrogen bond strength. Spectroscopic properties were calculated, yielding further characteristics of such hydrogen bonded gold species.

1. INTRODUCTION Gold paradoxically is an electronegative metal. In fact, on Pauling’s scale, it scores a value of 2.54, qualifying it for the most electronegative among all metallic elements. It closely approaches nonmetal elements like selenium (2.55) or even sulfur (2.58) in electronegativity (EN). Relativistic effects are described to be responsible in part for this peculiar property of gold since relativity causes an increase both in ionization potential and in electron affinity as compared to the hypothetical nonrelativistic case.1−3 Remarkably, in the many discussions of the rich and extensive experimental findings on the chemistry of gold and its compounds, the very basic criterion of EN seems to not have been exploited in abundance. For example, the generally accepted explanation for the phenomenon of short “aurophilic” contacts between formal Au(I) atoms in solid state structures is that they are based on relativistically enhanced van der Waals interactions.4−6 This “mechanism” indeed provides a concise explanation and is fully supported by ample quantitative quantum chemical analysis. However, a central heuristic aspect, which led to the formulation of the term “aurophilicity”, is that it apparently acts against charge polarization.7 It is a useful concept to explain some counterintuitive arrangements of gold complexes in the solid state by extraordinarily strong aurophilic interactions which work against mutual repulsive Coulomb forces. © XXXX American Chemical Society

In this context, it should be noted that, in the late 70s and early 80s of the last century, many gold compounds have been characterized using 179Au Mößbauer (also called “nuclear gamma resonance” (NGR)) spectroscopy, which led to a substantial deepening of our understanding of the electronic structure of chemically bound Au atoms.8−10 It is possible to assign consistently regions of the map from isotopic shifts (IS) to quadrupolar splitting constants of NGR spectra to formal oxidation states of the element under investigation. The IS directly correlates with the electron density at the nucleus. Thus, in agreement with expectations, the observed IS values in Au(I) compounds show a relatively broad range of variation. The lowest values were recorded for Au(I) compounds of the lighter halides (around 1−2 mm s−1) and the largest values for Au(I) compounds with carbanionic ligands or softer and less electronegative donors like phosphines (around 5−6 mm s−1), which lead to the formation of largely covalent bond situations and to strongly negative polarization of the gold atoms. This means that, due to its high EN, the cationic character of the gold atom in its compounds is often annihilated to a large extent by electron-donating ligands. Thus, while aurophilic interactions dominate the aggregation in many gold(I) complexes, dispersive d10−d10 attractions might be required Received: October 27, 2016

A

DOI: 10.1021/acs.inorgchem.6b02613 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

3. RESULTS AND DISCUSSION We have chosen to closely inspect a simple model system in this work. For obtaining meaningful results, we set the following requirements on it: (1) It should be able to provide us, as far as possible, with a unique answer to the question if hydrogen bonds to gold are favorable or not. (2) It should fulfill all requirements currently known for sustaining as strong as possible hydrogen bonds. (3) It should be a realistic synthetic target. 1H (see Figure 1 for a Lewis formula with formal partial charges including a suggested H-bond) appears to fulfill these

but probably are less dominating than one would expect without taking the high EN of the element into account. It should be noted that, to the best of our current knowledge, there are no systematic investigations on the correlation between 197Au Mößbauer IS and the strength of aurophilic interactions. In this work, we focus on another atomic property, which derives from high EN, namely, the capability to act as a hydrogen bond acceptor. It is interesting to find this aspect of gold chemistry, after many years of intense experimental and theoretical research, still being debated controversially.11,12 Grossly simplifying the state of affairs can be summarized like that: The controversy separates the more theoretically from the more experimentally oriented studies. While certain classes of H-bonds to Au(I) coordination centers have been proposed already early on the basis of quantum chemical calculations13 and still more examples are being proposed,14 evidence from the experimental side is sparse at most and even the principal possibility of effective Au···H hydrogen bonds is set into question.11,15 For this reason, we wanted to investigate the possibilities of Au(I) to act as an H-bond acceptor by theoretical studies able to provide a guideline for synthetic work. Our computational work led to some insights, which might prove useful for further preparative investigations and which we present in the following.

Figure 1. Molecular equilibrium structure of conformer 1Ha at the SCS-MP2/ECP level of theory. Distances are given in units of Å. The N−H1−Au angle and the C1−C2−C3−N torsional angle are shown, and the short Au···H contact is displayed with a dotted line. Atom numbers referred to in the text are shown.

2. QUANTUM CHEMICAL METHODS The geometry optimizations have been performed at the SCS-MP2 level of theory (i.e., Gerenkamp and Grimme’s “spin-componentscaled” variant of Møller−Plesset perturbation theory).16 For 1H and 2H, a frozen core of 24 occupied orbitals and Ahlrichs def2-TZVPP basis sets on all elements was used.17 For the calculations using an electronic core potential (ECP) on gold, the default for the def2TZVPP basis set was used (pseudo potential emulating 60 core electrons), whereas, for the calculations using the “zero order relativistic approximation” (ZORA),18 the SARC-TZVPP basis set from Bühl et al. was used.19 For the comparison of (scalar) relativistic versus nonrelativistic level of theory, again ZORA was used with the inverse fine structure constant set to either c (the vacuum speed of light) or its multiple value 100c. In this way, a consistent nonrelativistic counterpart of the standard ZORA calculations is obtained. These levels of theory are abbreviated as SCS-MP2/ECP, SCS-MP2/ZORA, and SCS-MP2/ZORANR. The potential energy scan shown in Figure 2 was performed at the SCS-MP2/ZORA level of theory using relaxed single point structures. In addition, for each of these structures, single points energies using the DLPNO-CCSD(T) level of theory20 [in the following denoted shorthand with CCSD(T)] were calculated. We have used the SCS-MP2/ECP level of theory for the geometry optimizations to get consistency with the CCSD(T) calculation which was not feasible for the all electron basis set calculation. The SCSMP2/ECP minimum structures show slight deviations from the SCSMP2/ZORA structures but agree in all qualitative features (coordinates can be found in the Supporting Information). The SCS-MP2/ECP calculations were performed with Turbomole (version 6.3)21 and the RI approximation was used. All other calculations have been performed with the ORCA program (version 3.0.3),22 and the RIJONX23 fitting (for the Coulomb integrals only) and the corresponding default fitting basis for the TZVPP basis set were used. The chemical shielding of the H-bonded proton was calculated with the program ReSpect24 at the four-component Kohn−Sham density functional level of theory using a Dirac-Coulomb Hamiltonian, the Perdew−Becke−Ernzerhof hybrid functional (PBE0),25,26 in connection with uncontracted all-electron valence triple-ζ basis sets (VTZ) from Dyall27 for all elements and using the SCS-MP2/ECP geometry.

criteria. Due to the torsional degree of freedom of the phenylpyridine (ppy) moiety (angle τ), it has enough conformational flexibility so that any formation of close H··· Au contacts in principle can be avoided and it moreover leaves a choice between two alternative forms C−H···Au and N−H··· Au. The possibility of a “classic” (typical) hydrogen bond should lead to a strong preference of the N−H···Au structure. In 1H, the Au atom is bound to two anionic ligands with (group) ENs not significantly exceeding the EN of gold and thus achieving an as high as possible covalent bond character: the carbanionic aryl-substituent ppy and the phenylthioloate. In addition, the thiophilic nature of gold should give rise to covalent bond contribution, and thus, the gold atom should bear a comparably high negative partial charge. For this reason, 1H could be a good candidate to show a strong N−H···Au interaction. Moreover, we find it plausible that this or closely related compounds could be synthesized. The arylation of Au(I) with (ppy)− is already described in the literature,28−31 and given that some LAu(ppy) is accessible (with ligands L like PPh3, thf, PMe3), a reaction with PhSH under mild conditions could potentially lead to the formation of 1H under release of L. 3.1. Molecular Structure. A geometry optimization of 1H at the SCS-MP2/ECP level of theory (see the Quantum Chemical Methods section for details) yields a zero-gradient structure, which could be identified by analytical frequency analysis as a local minimum (1Ha, Figure 1) on the potential energy surface. While the distances Au−S (2.308 Å) and Au−C (2.013 Å) are in the expected range of covalently bound Au/S and Au/C, the most intriguing feature of the structure is the short Au···H contact of 2.060 Å. This is significantly below the sum of the B

DOI: 10.1021/acs.inorgchem.6b02613 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

only the torsional angle at varying fixed values. In addition, for each point, the energy at the CCSD(T)/ECP level of theory was calculated. Figure 2 shows a plot of the resulting potential

van der Waals radii (2.86 Å)32 and moreover only 0.3−0.4 Å above the typical covalent Au−H bond length of about 1.7 Å,33 indicating a considerable stabilizing interaction between Au and H. From a structural point of view, this is confirmed by the elongation of the N−H distance (1.043 Å as compared to 1.013 Å for [ppyH]+) as well as by the absolute size of the torsional angle between the phenyl and the pyridyl moiety, which is lowered going from the [ppyH]+ to complex 1Ha (35.4° versus 25.7°) by almost 10° toward planarization. Also, the difference between the C3−N−H1 and the C4−N−H1 angle is significantly smaller in [ppyH]+ than in 1Ha (1.1° versus 4.0°), further illustrating the distinctive orientation of H1 toward the gold atom. A comparison of the minimum geometries at the HF, MP2 (both SCS and conventional MP2), and CCSD(T) levels shows that the tendency of a planarization of the ppy-Au moiety in the minimum structures is increasing in this order, suggesting that a fully optimized CCSD(T) structure would have an even shorter Au···H1 contact (vide inf ra). A noteworthy structural feature of 1Ha is the short distance between the hydrogen atom bound to C4 and the average plane defined by the phenyl ring of the thiolate moiety (2.313 Å). This suggests the presence of a significant intramolecular C− H···π interaction. Indeed, the question arises if actually this interaction is enforcing or at least strongly supporting a conformation with a short Au···H contact. To address this question, we like to consider three arguments: (1) An optimization (SCS-MP2/ECP) of the methyl- (instead of phenyl-) substituted analogue of 1Ha and 1Hb results again in two different minima structures (1Ha′ and 1Hb′) which are separated by 9.8 kcal mol−1, favoring again the conformation with a short Au···H1 contact (1Ha′; see the Supporting Information for coordinates). In agreement with this, also the structural parameters of 1Ha′ (Au···H1 = 1.966 Å, N−H = 1.062 Å, and Au−H−N = 18.5°) suggest that the Au···H interaction is even stronger in 1Ha′ than in 1Ha. Thus, the inductive effect from the organic rests (methyl: +I versus phenyl: −I) influences the interaction between Au and H via the electron density at the Au atom and exceeds the possible stabilization from a C−H···π interaction. (2) At the nonrelativistic level of theory (SCS-MP2/ZORANR; see section 3.4, Relativistic Effects, for details), we find all structural characteristics for a H-bonded moiety distinctly less pronounced (the Au···H bond is “turned off” by “turning off” relativity), but we find also an even shorter C−H···C6 plane distance of 2.220 Å. (3) In the solid state structure of the isoelectronic mercury compound (ppy)Hg(SPh), intermolecular π-stacking, but no short intramolecular C−H···π contacts, are observed.34 (2) and (3) suggest that the formation of the Au···H contact in 1Ha is rather evoking the C−H···π contact than vice versa. We conclude from (1)−(3) that a possible C−H···π interaction in 1H does neither dictate nor significantly support the short Au··· H contact in 1Ha. From the following, it will become obvious that these structural characteristics hint at a possible significant intramolecular attraction between the hydrogen atom H1 and the gold atom which find their resemblance in all of the further discussed computed molecular properties. 3.2. Conformers, Isomers, and Energies. For a closer inspection of the interaction between H1 and Au in 1H, a potential energy curve by variation of the torsional angle N− C3−C2−C1 was calculated. All points in the scan were optimized at the SCS-MP2/ECP level of theory while keeping

Figure 2. N−C3−C2−C1 potential energy curves at HF, SCS-MP2, and CCSD(T) levels of theory in units of kcal mol−1 relative to the SCS-MP2 minimum torsional angle of 23.7°.

energy functions including the Hartree−Fock reference values. Each curve was thereby shifted to zero at the SCS-MP2/ECP minimum torsional angle of 23.7°. The energy scan reveals a second conformer (1Hb) with a torsional angle N−C3−C2−C1 of 119.1° (fully optimized at the SCS-MP2 level of theory), lying 7.5 kcal mol−1 above the conformer with the short H1−Au contact but which is separated only by a minute barrier of less than 1 kcal mol−1 (Figure 3). At the CCSDT(T)/ECP level of theory, the

Figure 3. Equilibrium structure of conformer 1Hb at the SCS-MP2/ TZVPP level of theory.

difference between the energy minimal scan points at −125.7° and at −15.7° amounts to 7.8 kcal mol−1, confirming the SCSMP2 result. It can be concluded that conformer 1Ha is significantly more stable than conformer 1Hb. From a bond theoretical point of view, the driving force of this stabilization can only be found in an attractive hydrogen bond-like interaction between the gold atom and the N−H1 moiety. It should be noted that, on the other hand, no elongation of the N−H distance (1.030 Å versus 1.031 Å in [ppyH]+) is found in 1Hb. The difference in energy between the two conformers 1Hb and 1Ha of 7.8 kcal mol−1 may serve as an estimate of the hydrogen bond strength. This is in the regime of medium to strong hydrogen bonds, like typical O−H···N hydrogen bonds or like the hydrogen bonds between water molecules in bulk water. In addition, we have considered a constitutional isomer which usually has to be taken into account when regarding the metal organic chemistry of the phenylpyridine or bipyridyl C

DOI: 10.1021/acs.inorgchem.6b02613 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry ligands;35 this is N-thiophenolato-gold(I)-phenylpyridine isomer 2H (see Figure 4).

hydrogen bond” was coined by Hobza.36 In 1Ha, the calculated (SCS-MP2/ECP) stretching frequency ν(N−H) is 3006.4 cm−1, which is substantially red-shifted by more than 560 cm−1 relative to the frequency of the ν(N−H) stretching mode in the phenylpyridinium cation [ppyH]+ (3566.8 cm−1), or of 575.2 cm−1 relative to 1Hb (3581.6 cm−1). This large red shift underlines what already has been found for the N−H bond length and the relative conformer energies: The interaction between Au and H1 in 1H is that of a rather strong classical hydrogen bond. Another interesting diagnostics for the presence of hydrogen bonded protons is 1H NMR chemical shifts. Usually, hydrogen-bonded protons appear at low fields in the spectrum, which, in a simplifying manner, can be rationalized by the depletion in electron density of the proton. Fully relativistic four-component calculations (see the Quantum Chemical Methods section) of the chemical shielding of H1 using density functional theory indeed yield an isotropic chemical shift of 14 ppm, relative to the proton signal of Si(CH3)4. This again confirms the hydrogen bond nature of the Au···H1 interaction. 3.4. Relativistic Effects. A comparison between the parameters of optimized geometries at the SCS-MP2/ECP, SCS-MP2/ZORA, and SCS-MP2/ZORANR levels of theory shows that, according to expectations, the agreement between the levels of theory taking into account scalar relativistic effects (SCS-MP2/ECP, SCS-MP2/ZORA) is fairly good, whereas, at the nonrelativistic level of theory (SCS-MP2/ZORANR), the parameters most closely associated with the Au···H hydrogen bond show the strongest deviations. For example, the Au−H distance is increasing from 2.108 Å (SCS-MP2/ZORA) to 2.424 Å (SCS-MP2/ZORANR); also, the N−H−Au angle which should be around 180° in an ideal hydrogen bond is decreasing from 141.4° to 121.5° and the torsional angle between the phenyl and the pyridine moiety deviates stronger from planarity at the nonrelativistic level (−35.5°) than at the relativistic level (27.7° at SCS-MP2/ZORA), which finds its explanation in the tendency toward planarization of the ppy moiety, induced by the intramolecular H···Au interaction. These structural trends are paralleled by the difference in total energy between the isomers 1Ha and 1Hb. At the relativistic levels, 1Ha is more stable by 7.5 (7.3) kcal mol−1 at the SCS-MP2/ECP (SCS-MP2/ZORA) level of theory, whereas, without taking relativistic effects into account (SCSMP2/ZORANR), the energy difference drops to 5.3 kcal mol−1. The difference of 2.0 kcal mol−1 between the nonrelativistic and the relativistic energy difference of 1Ha and 1Hb might be interpreted as a lower bound to the relativistic contribution to the hydrogen bond due to the additional formation of the C− H···π contact in 1Ha. Similarly, the calculated frequency of the ν(N−H) stretching mode is 3006.4 cm−1 at the SCS-MP2/ ECP level of theory, while it is 3516.1 cm−1 at the nonrelativistic level of theory. In summary, according to expectations, the ability of gold to form short stabilizing (attractive) interactions to hydrogen atoms bound to electronegative elements is clearly a relativistic property of gold.

Figure 4. Equilibrium structure of isomer 2H at the SCS-MP2/ECP level of theory.

Remarkably, 2H is more stable in energy by 18.8 kcal mol−1 than hydrogen bonded isomer 1Ha. A simple explanation for this finding is that, in 1Ha, a partial charge separation into different spatial regions of the molecule is operative, which is energetically less favorable. This is suggested already by the formal Lewis formula representation of 1H shown in Scheme 1. Scheme 1

The formal negative partial charge of the gold atom in this representation gains chemical relevance by the high electronegativity of gold and also by its high affinity to form covalently dominated bonds with sulfur. In this simplified picture of charge separation, also the relative stability of 1Ha versus 1Hb finds an explanation. From a more abstract point of view, it becomes clear that systems which can potentially exhibit hydrogen bonds to Au(I) are intrinsically at risk to rearrange, due to the isolobality between the LAu+ and H+. This is because, apparently, any donor D able to form stable coordination compounds [Au(L)D]+ is potentially also able to coordinate H+ instead (to form HD+). As well as any electronegative hydrogen bond donor X which can form X−H···Au(L)D also can coordinate to LAu+ to form LAuX. Thus, the possibility of the formation of Au···H bonds crucially depends on the competition between the two alternative systems: X−H···Au(L)D and XAuL + HD+. 3.3. Spectroscopic Properties. Hydrogen bonding can lead to either a red shift or a blue shift in the vibrational stretching frequency of the respective covalently bound H−X moiety. The first case is usually referred to as a “classical hydrogen bond”, which lowers the bond order of the covalent bond to hydrogen; for the latter case, the term “improper

4. CONCLUSION In conclusion, we have demonstrated that hydrogen bonds of the form Au···H−N should be feasible if the right molecular components are chosen: (i) The gold atom has to be electronrich, e.g., by using sufficiently electron-donating ligands. (ii) The proton has to be offered to the gold atom in a favorable geometric position. Both requirements are fulfilled for (ppyH)D

DOI: 10.1021/acs.inorgchem.6b02613 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

(12) Scherf, L. M.; Baer, S. A.; Kraus, F.; Bawaked, S. M.; Schmidbaur, H. Implications of the crystal structure of the ammonia solvate [Au(NH3)2]Cl·4NH3. Inorg. Chem. 2013, 52, 2157. (13) Kryachko, E. S. Where gold meets a hydrogen bond? J. Mol. Struct. 2008, 880, 23. (14) Groenewald, F.; Dillen, J.; Raubenheimer, H. G.; Esterhuysen, C. Preparing gold(I) for interactions with proton donors: The elusive [Au]···HO hydrogen bond. Angew. Chem., Int. Ed. 2016, 55, 1694. (15) Schmidbaur, H.; Raubenheimer, H. G.; Dobrzańska, L. The gold−hydrogen bond, Au−H, and the hydrogen bond to gold, Au··· H−X. Chem. Soc. Rev. 2014, 43, 345. (16) Gerenkamp, M.; Grimme, S. Spin-component scaled secondorder Møller−Plesset perturbation theory for the calculation of molecular geometries and harmonic vibrational frequencies. Chem. Phys. Lett. 2004, 392, 229. (17) Weigend, F.; Häser, M.; Patzelt, H.; Ahlrichs, R. RI-MP2: optimized auxiliary basis sets and demonstration of efficiency. Chem. Phys. Lett. 1998, 294, 143. (18) 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. (19) Bühl, M.; Reimann, C.; Pantazis, D. A.; Bredow, T.; Neese, F. Geometries of third-row transition-metal complexes from densityfunctional theory. J. Chem. Theory Comput. 2008, 4, 1449. (20) Riplinger, C.; Neese, F. An efficient and near linear scaling pair natural orbital based local coupled cluster method. J. Chem. Phys. 2013, 138, 034106. (21) Ahlrichs, R.; Bär, M.; Häser, M.; Horn, H.; Kölmel, C. Electronic structure calculations on workstation computers: The program system turbomole. Chem. Phys. Lett. 1989, 162, 165. (22) Neese, F.; Schwabe, T.; Kossmann, S.; Schirmer, B.; Grimme, S. Assessment of orbital-optimized, spin-component scaled second-order many-body perturbation theory for thermochemistry and kinetics. J. Chem. Theory Comput. 2009, 5, 3060. (23) Neese, F.; Olbrich, G. Efficient use of the resolution of the identity approximation in time-dependent density functional calculations with hybrid density functionals. Chem. Phys. Lett. 2002, 362, 170. (24) Komorovsky, S.; Repisky, M.; Malkin, V. G.; Malkina, O. L.; Kaupp, M.; Ruud, K. with contributions from Bast, R.; Ekström, U.; Knecht, S.; Ondik, I. M.; Malkin, E. ReSpect, version 3.4.0: Relativistic Spectroscopy DFT Program; 2014.http://rel-qchem.sav.sk. (25) Perdew, J. P.; Ernzerhof, M.; Burke, K. Rationale for mixing exact exchange with density functional approximations. J. Chem. Phys. 1996, 105, 9982. (26) Adamo, C.; Barone, V. Toward reliable density functional methods without adjustable parameters: The PBE0 model. J. Chem. Phys. 1999, 110, 6158. (27) Dyall, K. G. Personal communication to S. Komorovsky and M. Repisky, 2014. (28) Gao, L.; Peay, M. A.; Partyka, D. V.; Updegraff, J. B., III; Teets, T. S.; Esswein, A. J.; Zeller, M.; Hunter, A. D.; Gray, T. G. Mono-and di-gold(I) naphthalenes and pyrenes: syntheses, crystal structures, and photophysics. Organometallics 2009, 28, 5669. (29) Hashmi, A. S. K.; Ramamurthi, T. D.; Rominger, F. Synthesis, structure and reactivity of organogold compounds of relevance to homogeneous gold catalysis. J. Organomet. Chem. 2009, 694, 592. (30) Heinrich, A. (Triphenylphosphine) gold(I)chloride. Synlett 2015, 26, 1135. (31) Molteni, R.; Edkins, K.; Haehnel, M.; Steffen, A. C−H Activation of fluoroarenes: Synthesis, structure, and luminescence properties of copper(I) and gold(I) complexes bearing 2-phenylpyridine ligands. Organometallics 2016, 35, 629. (32) Bondi, A. Van der Waals volumes and radii. J. Phys. Chem. 1964, 68, 441. (33) Berger, R. J. F. The smallest ’aurophilic species’. Z. Naturforsch., B: J. Chem. Sci. 2009, 64, 388. (34) Hg(ppy)(SPh) crystallizes in the space group symmetry P21/c. Its molecular structure is very similar to the calculated geometry of the

AuSPh. The 2-phenylpyridine which is bound via the phenyl group renders the nitrogen atom of the pyridine moiety free for a protonation. We could show that the hydrogen bonded isomer is the most stable conformer in this molecular setting. We are convinced that this combination of ligands is just one under many possible molecules to realize strong hydrogen bonds with a gold(I) atom as the hydrogen bond acceptor, and we encourage synthetic chemists to keep on searching for hydrogen bonds to gold(I). To find other potential candidate ligand systems, one could also use the experimental data available on the 197Au Mö ßbauer isotopic shifts as an orientation, qualifying those ligands leading to the highest IS values as most promising candidates.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.6b02613. Cartesian coordinates for the calculated minimum geometries in units of Å of 1Ha SCS-MP2/ECP, 1Ha SCS-MP2/ZORA, 1Ha SCS-MP2/ZORANR, 1Hb SCSMP2/ECP, 1Hb SCS-MP2/ZORA, 1Hb SCS-MP2/ ZORANR, 1Ha′ SCS-MP2/ECP, 1Hb′ SCS-MP2/ECP, and 2H SCS-MP2/ECP; (PhS)Hg(ppy) synthesis; and (PhS)Hg(ppy) crystallographic data (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (R.J.F.B.). *E-mail: [email protected] (U.M.). ORCID

Raphael J. F. Berger: 0000-0002-2284-0540 Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Pyykkö, P. Relativistic quantum chemistry. Adv. Quantum Chem. 1978, 11, 353. (2) Pitzer, K. S. Relativistic effects on chemical properties. Acc. Chem. Res. 1979, 12, 271. (3) Pyykkö, P.; Desclaux, J. P. Relativity and the periodic system of elements. Acc. Chem. Res. 1979, 12, 276. (4) Pyykkö, P.; Zhao, Y. Ab initio calculations on the (ClAuPH3)2 dimer with relativistic pseudopotential: Is the “Aurophilic Attraction” a correlation effect? Angew. Chem., Int. Ed. Engl. 1991, 30, 604. (5) Pyykkö, P. Theoretical chemistry of gold. Angew. Chem., Int. Ed. 2004, 43, 4412. (6) Barysz, M.; Pyykkö, P. Au22+ has bound excited states. Chem. Phys. Lett. 2000, 325, 225. (7) Schmidbaur, H. The fascinating implications of new results in gold chemistry. Gold. Bull. 1990, 23, 11. (8) McAuliffe, C. A.; Parish, R. V.; Randall, P. D. Studies in Mossbauer spectroscopy. Part 9. Gold-197 spectra of gold(I) compounds. J. Chem. Soc., Dalton Trans. 1977, 15, 1426. (9) Parish, R. V. Gold and Mössbauer Spectroscopy. Gold Bull. 1982, 15, 51. (10) Braunstein, P.; Schubert, U.; Burgard, M. Synthesis and Spectroscopic Studies on Metal-Metal Bonded Linear Heterotrimetallic Gold(I) Complexes. Crystal Structure of [n-Bu4N][Au[Cr(CO)3-ηC5H5]2]. Inorg. Chem. 1984, 23, 4057. (11) Kraus, F.; Schmidbaur, H.; Al-juaid, S. S. Tracing hydrogen bonding Au··· H−C at gold atoms: A case study. Inorg. Chem. 2013, 52, 9669. E

DOI: 10.1021/acs.inorgchem.6b02613 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry gold compound 1Ha. See the Supporting Information for further experimental and structural details (CCDC 1521789). (35) Constable, E. C.; Housecroft, C. E. The electronic structure of some ruthenium(II) complexes related to [Ru(bipy)3]2+: An investigation of the stepwise replacement of N,N-donors by C, Ndonors. Polyhedron 1990, 9, 1939. (36) Hobza, P.; Špirko, V.; Selzle, H. L.; Schlag, E. W. Anti-hydrogen bond in the benzene dimer and other carbon proton donor complexes. J. Phys. Chem. A 1998, 102, 2501.

F

DOI: 10.1021/acs.inorgchem.6b02613 Inorg. Chem. XXXX, XXX, XXX−XXX