Studies of Nature of Uncommon Bifurcated I–I···(I–M) Metal-Involving Noncovalent Interaction in Palladium(II) and Platinum(II) Isocyanide Cocrystals

Two isostructural trans-[MI2(CNXyl)2]·I2 (M = Pd or Pt; CNXyl = 2,6-dimethylphenyl isocyanide) metallopolymeric cocrystals containing uncommon bifurcated iodine···(metal–iodide) contact were obtained. In addition to classical halogen bonding, single-crystal X-ray diffraction analysis revealed a rare type of metal-involved stabilizing contact in both cocrystals. The nature of the noncovalent contact was studied computationally (via DFT, electrostatic surface potential, electron localization function, quantum theory of atoms in molecules, and noncovalent interactions plot methods). Studies confirmed that the I···I halogen bond is the strongest noncovalent interaction in the systems, followed by weaker I···M interaction. The electrophilic and nucleophilic nature of atoms participating in I···M interaction was studied with ED/ESP minima analysis. In trans-[PtI2(CNXyl)2]·I2 cocrystal, Pt atoms act as weak nucleophiles in I···Pt interaction. In the case of trans-[PdI2(CNXyl)2]·I2 cocrystal, electrophilic/nucleophilic roles of Pd and I are not clear, and thus the quasimetallophilic nature of the I···Pd interaction was suggested.


Experimental Procedures
For each experiment, single crystals were selected from the sample under microscope, immersed in cryo-oil, and mounted in a MiTeGen loop for the SCXRD data collection.
All data were measured using a dual-source Rigaku SuperNova diffractometer equipped with an Atlas detector and an Oxford Cryostream cooling system using mirrormonochromated Mo-K α radiation (λ = 0.71073 Å). Data collection and reduction for all complexes were performed using the program CrysAlisPro 1 and Gaussian face-index absorption correction method was applied. 1 All structures were solved with Direct Methods or Patterson synthesis (SHELXS) 2 and refined by full-matrix least squares based on F 2 using SHELXL-2013. 3 Anisotropic displacement parameters were assigned for all non-hydrogen atoms unless stated otherwise. In the structure 1•I 2 there is a short contact between I1 and I4 atoms because of the halogen bonding between these two iodines. There is also a similar interaction between I3 and I2. Hydrogen atoms were placed in their idealized positions and refined as riding atoms. Isotropic displacement parameters for all H atoms were constrained to multiples of the equivalent displacement parameters of their parent atoms with U iso (H) = 1.2 U eq (parent atom). Enhanced rigid bond restraints (RIGU) 4,5 with standard uncertainties of 0.001 Å 6 were applied for several atom pairs. The single crystal X-ray data, experimental details and CCDC numbers (2054859-2054862) are given below.

Experimental Procedures
Powder X-ray diffraction measurements were done using PANalytical X´Pert PRO diffractometer in Bragg-Brentano geometry with Cu-K  radiation (λ = 1.5419 Å; 45kV, 40mA). A lightly hand-ground powder sample was prepared on a silicon-made "zerobackground inducing" holder with shallow sample cavity (petrolatum jelly was used as fastener). Diffraction patterns were recorded from a spinning sample by positionsensitive X´Celerator detector using continuous scanning mode in 2 range of 6 -60° with a step size of 0.017° and counting time of 200 s per step. Diffraction data analyses were made using program Malvern Panalytical HighScore Plus (v. 4.8). 7 The room temperature unit cell parameters of 1•I 2 and 2•I 2 cocrystals were indexed by Pawley analysis 8 using the corresponding low temperature single crystal structure parameters (presented in this study) of both phases as the basis of least-squares refinement.
Parameters varied in the whole-pattern fitting were as follows: zero-offset, polynomial background, sample displacement, unit cell and peak profile parameters.

Detailed results of PXRD analysis
According to the obtained results ( Figure S1- Figure S2, Table S3) phase purity of the bulk material was confirmed for both 1•I 2 and 2•I 2 cocrystals.

S-9
Summary of computational studies on noncovalent interactions in 1•I 2 and 2•I 2 cocrystals

General Computational details
All models were calculated with the Gaussian09 (revision C.01) program package 9 at the DFT level of theory. M06-L 10 density functional was utilized together with the def2-TZVP 11-13 basis set. Strength, topology and nature of noncovalent interactions in the cocrystals were studied using the wave functions obtained from the DFT calculations.

ESP analysis
All ESP calculations were done in AIMALL program at 0.001 a.u. contour of molecular surface, utilizing wavefunctions generated at the M06-L/def2-TZVP level of theory.

QTAIM analysis
All attempts to optimize a structure corresponding to the solid-state structure of (1) 2 •I 2 (or (2) 2 •I 2 ) by using only the components of the simplest structural unit resulted in twisted geometries that retained the interaction between I 2 and 1 (or 2) but made the two 1 (or 2) units move almost perpendicular to each other. To retain the structural motif observed in the solid-state, two additional 1 (or 2) units that take part in the - stack formation between Xyl rings had to be included in the optimization as shown in Figure   S6. This indicates that - interaction between Xyl rings locks the complexes together in the solid-state and directs the crystal structure formation.  91 both 1 and 2) is very close to that found for HCN (1.79). 21 In addition to describing the strong covalent bonds, the QTAIM analysis also reveals weak interactions from I 2 atoms to both metal center and iodo ligand. In the optimized structures the contact between iodine atoms I1 and I6 is calculated to be stronger than the interaction of metal center and I6, whereas in the crystal structures the contacts are more alike. Calculated atomic charges (See Table C3) suggest that I 2 in (1) 2 I 2 (and (2) 2 I 2 ) gains a weak negative charge. Furthermore, the atomic charges indicate that the origin of the transferred electron density are the iodo substituents of 1 (or 2) complex rather than the metal center. This can be further seen as a sign that halogen bonding between neutral iodine as donor and negatively charged iodo substituents as acceptors is the driving force for the attachment of I 2 to (1) 2 I 2 (and (2) 2 I 2 ) structure while the M1-I6 interaction has more of a supporting role in the structure.

Contact
Distance  Table S5. Comparison of selected experimental (SP) and M06-L/def2-TZVP optimized (OPT) structural parameters [Å] of (1) 2 •I 2 and their QTAIM bond critical point parameters (in a.u.): electron density  b , Laplacian of electron density  2 , local electronic potential energy density V b , local electronic kinetic energy density G b .

Contact
Distance structures of (2) 2 •I 2 and I 2 : electron density  b , Laplacian of electron density  2 , local electronic potential energy density V b , local electronic kinetic energy density G b .  Figure S7. Fragments of (1) 2 •I 2 (and (2) 2 •I 2 ) structure used in the local energy decomposition analysis.
-69 -288 -29 -7 -392 a Electronic preparation energies resulting from intra-fragment changes in electron density and deformation energies due to geometrical differences of fragments in interacting structure compared to their separated equilibrium geometries that are required to derive the dissociation energies corresponding to the analyzed interactions have not been included in the analysis. Table S9. Energy components (exchange interaction, E exch , electrostatic and polarization energy, E elstat , dispersion interaction, E DISP and contribution from triples correction, E (T) ) of the inter-fragment interaction energies (kJ mol -1 ) in (1) -82 -295 -26 -8 -411 a Electronic preparation energies resulting from intra-fragment changes in electron density and deformation energies due to geometrical differences of fragments in interacting structure compared to their separated equilibrium geometries that are required to derive the dissociation energies corresponding to the analyzed interactions have not been included in the analysis.

NCI-plot analysis. Experimental Procedures
All calculations were done in Critic2 program 15

NCI-plot analysis. Results
The isosurfaces on the 3D plots are colored such that deep blue represent very strong Sign(λ 2 )ρ units are a. u.