Stéphanie
Poirier
,
Feriel
Rahmani
and
Christian
Reber
*
Département de chimie, Université de Montréal, Montréal, Québec H3C 3J7, Canada. E-mail: christian.reber@umontreal.ca
First published on 28th March 2017
We present the variable-pressure luminescence spectra of crystals of isostructural palladium(II) and platinum(II) complexes with bis-N-benzyl-N′-3-methylpyridyldithiocarbamate (bmpDTC) ligands. The d–d luminescence band maxima E_{max} for these complexes are compared to others with different peripheral substituents on the dithiocarbamate ligands in the solid state. The comparison reveals significant variations of E_{max} despite very similar metal coordination geometries. E_{max} varies by 3000 cm^{−1} and 1300 cm^{−1} among four dithiocarbamate complexes of platinum(II) and palladium(II), respectively. Variations of E_{max} with pressure reveal the effects of intermolecular M⋯H–C interactions on several complexes. ΔE_{max}/ΔP values are negative for the bmpDTC complexes, unprecedented in the dithiocarbamate family. Static orientation and pressure-induced movement of the C–H bonds involved in intermolecular interactions have a significant effect on E_{max} and ΔE_{max}/ΔP, with a stronger impact on platinum(II) complexes than on their palladium(II) analogs.
The simplest luminescence spectra involve d–d transitions. Ligand-field theory predicts identical energies if ligands, metal–ligand distances and L–M–L angles are identical. Different ligands lead to variations according to the spectrochemical series.^{16,17} This is illustrated by the comparison of luminescence maxima E_{max} for crystalline K_{2}[PtBr_{4}] and K_{2}[PtCl_{4}] at approximately 12500 cm^{−1} (800 nm)^{18} and 12800 cm^{−1} (780 nm),^{18} respectively, with E_{max} higher by 300 cm^{−1} for the stronger-field chloride ligands. The square-planar coordination geometry enables intermolecular interactions due to the empty coordination sites above and below the MX_{4} plane.^{1–4} Molecular crystals can be used to obtain experimental insight on both molecular luminescence properties and weak intermolecular interactions. Such interactions define many properties of chemical systems.^{19–22} In inorganic complexes, M⋯H–C interactions have been reported based on crystal structures,^{20,23,24} identified according to the established distance range of 2.3 to 2.9 Å and specific shifts of ^{1}H NMR peaks^{23} in solution, the latter being limited to intramolecular effects and not applicable to intermolecular interactions in solids. Precise H positions are experimentally difficult to obtain from X-ray diffraction, leading to significant uncertainties. Intermolecular distances at ambient conditions do not lead to quantitative understanding of intermolecular interactions.^{25–27} In addition, these weak interactions are challenging to characterize with electronic structure calculations,^{26} illustrating the need for detailed experimental results.
A promising new approach involves the use of variable pressure on the solid samples to shorten M⋯H–C distances and gain experimental insight on characteristic spectroscopic shifts.^{25} Variable pressure is used to induce small structural variations leading to significant spectroscopic changes.^{28–30} We have recently shown that M⋯H–C interactions can be experimentally characterized by luminescence spectroscopy at variable pressure,^{31} with significantly wider variations occurring than in vibrational spectroscopy. It is advantageous that luminescence is observed only from the lowest excited state, according to Kasha's rule, eliminating problems due to overlapping bands.^{32} For many square-planar complexes, increasing pressure shifts E_{max} to higher energies, a consequence of metal–ligand bond compression. This is easily understood from ligand-field theory and orbital characteristics: shorter metal–ligand bonds lead to stronger destabilization of the LUMO, with M–S σ* character, than of the HOMO with π* character, resulting in positive ΔE_{max}/ΔP values.^{33,34} A series of complexes with chelating ligands have been studied and average trends have been determined for this effect, giving an average ΔE_{max}/ΔP value of +12 ± 2 cm^{−1}/kbar for the shift of d–d luminescence transitions.^{31} Deviations from this average allow us to identify the effects of M⋯H–C interactions, yielding additional quantitative information beyond intermolecular distances from crystal structures at ambient conditions. These interactions have a significant impact on the luminescence energy, even though they are weaker than metal–ligand bonds.
We present isostructural palladium(II) and platinum(II) complexes with bis-N-benzyl-N′-3-methylpyridyldithiocarbamate (bmpDTC) ligands, showing M⋯H–C interactions from a CH_{2} group of a neighbouring ligand in the crystal structure. Surprising differences to other dithiocarbamate complexes with variable substituents but similar metal coordination geometry are observed. An important variation of E_{max} by 3000 cm^{−1} at ambient conditions is measured for platinum(II) dithiocarbamates. The variation observed for the palladium(II) analogs is smaller by a factor of two. We use luminescence spectroscopy at variable pressure to understand the differences.
Metal chloride (0.0721 g of PtCl_{2} and 0.0575 g of PdCl_{2}) was dissolved in a minimum (about 125 mL for PtCl_{2} and 70 mL for PdCl_{2}) of dimethylsulfoxide (DMSO). Four equivalents (0.0432 g) of sodium hydroxide (NaOH) were dissolved in a minimum (about 25 mL) of ethanol 95%, with crushing and heating. An excess of carbon disulfide (0.3 mL) was added to the solution of sodium hydroxide. Two equivalents of the amine (0.1 mL, 0.1071 g) in the form of liquid N-nicotinylbenzylamine were mixed to the solution of sodium hydroxide to form the ligand in the form of N-benzyl-N′-3-methylpyridyldithiocarbamate. A powder (yellow for [Pt(bmpDTC)_{2}], orange for [Pd(bmpDTC)_{2}]) precipitates from the combination of N-benzyl-N′-3-methylpyridyldithiocarbamate solution and metal solution. The powder is filtered and washed with diethyl ether. Single crystals were obtained by slow cooling of the solutions in acetone. The crystals were collected with a yield of 58% for [Pt(bmpDTC)_{2}] and 22% for [Pd(bmpDTC)_{2}].
Fig. 1 Crystal structure of [Pt(bmpDTC)_{2}]. The thermal ellipsoids are shown at 50% probability. The Pt⋯H–C interaction is visualized by the green dotted line. |
Temperature (K) | 295 | 150 |
---|---|---|
Empirical formula | C_{28}H_{26}N_{4}PtS_{4} | C_{28}H_{26}N_{4}PtS_{4} |
M (g mol^{−1}) | 741.86 | 741.86 |
Temperature/K | 295 | 150 |
Crystal system | Triclinic | Triclinic |
Space group | P | P |
a/Å | 6.4927(15) | 6.4423(2) |
b/Å | 10.115(3) | 10.0534(3) |
c/Å | 11.568(3) | 11.5378(3) |
α/° | 75.520(14) | 74.9490(10) |
β/° | 79.620(13) | 80.0000(10) |
γ/° | 72.223(13) | 71.4280(10) |
Volume/Å^{3} | 696.0(3) | 680.67(3) |
Z | 1 | 1 |
ρ _{calc}/g cm^{−3} | 1.770 | 1.810 |
μ/mm^{−1} | 8.471 | 8.661 |
F(000) | 364.0 | 364.0 |
Crystal size/mm^{3} | 0.15 × 0.1 × 0.09 | 0.15 × 0.1 × 0.09 |
Radiation | GaKα (λ = 1.34139) | GaKα (λ = 1.34139) |
2Θ range for data collection/° | 6.91 to 122.896 | 6.936 to 121.346 |
Index ranges | −8 ≤ h ≤ 8, −13 ≤ k ≤ 13, −14 ≤ l ≤ 14 | −8 ≤ h ≤ 8, −13 ≤ k ≤ 13, −14 ≤ l ≤ 14 |
Reflections collected | 22836 | 26728 |
Independent reflections | 3198 [R_{int} = 0.0466, R_{sigma} = 0.0226] | 3122 [R_{int} = 0.0443, R_{sigma} = 0.0187] |
Data/restraints/parameters | 3198/0/169 | 3122/0/169 |
Goodness-of-fit on F^{2} | 1.065 | 1.168 |
Final R indexes [I ≥ 2σ(I)] | R _{1} = 0.0407, wR_{2} = 0.1048 | R _{1} = 0.0281, wR_{2} = 0.0672 |
Final R indexes [all data] | R _{1} = 0.0421, wR_{2} = 0.1063 | R _{1} = 0.0283, wR_{2} = 0.0673 |
Largest diff. peak/hole/e Å^{−3} | 1.66/−0.56 | 1.79/−0.68 |
[Pt(bmpDTC)_{2}] | [Pd(bmpDTC)_{2}]^{35} | |
---|---|---|
Bond (Å) | ||
M–S1 | 2.318(1) | 2.339(3) |
M–S2 | 2.318(2) | 2.333(3) |
S1–C | 1.729(5) | 1.749(7) |
S2–C | 1.726(5) | 1.717(8) |
C–N | 1.314(8) | 1.31(1) |
N–C2 | 1.471(7) | 1.50(1) |
N–C3 | 1.468(7) | 1.48(1) |
Angles (°) | ||
S1–M–S2 | 74.98(5) | 75.69(7) |
S1–C1–S2 | 109.6(3) | 111.6(4) |
Intermolecular distances (Å) | ||
M⋯H_{a}–C_{a} | 2.8142(7) | 2.788(3) |
M⋯C_{a} | 3.692(5) | 3.683(9) |
M⋯H_{b}–C_{b} | 3.3417(8) | 3.473(4) |
M⋯C_{b} | 3.989(5) | 4.118(9) |
M⋯N | 3.912(4) | 3.973(7) |
M⋯C | 4.500(5) | 4.557(8) |
M⋯S1 | 5.165(2) | 5.124(6) |
M⋯S2 | 5.478(2) | 5.584(6) |
M⋯H–C (plane) | 4.047(1) | 4.055(5) |
M⋯C (plane) | 4.528(7) | 4.53(1) |
Metal | Ligand | ΔE_{max} /ΔP (cm^{−1}/kbar) | Shortest intermolecular M⋯H–C (Å) | M–S1 (Å) | M–S2 (Å) | S1–M–S2 (°) | S1–M–S1 (°) | Ref. |
---|---|---|---|---|---|---|---|---|
Pt(II) | EDTC | +15 | 3.0581 | 2.321(2) | 2.314(2) | 75.22(6) | 180.00(6) | 54 |
bmpDTC | −8 ± 3 | 2.8142(7) | 2.318(1) | 2.318(2) | 74.98(5) | 180.00(5) | This work | |
(CH_{3})_{2}DTC | +44 ± 5 | 2.9919(1) | 2.3290(8) | 2.3126(8) | 75.13(3) | 180.00(3) | 12 | |
dopDTC | +11 ± 1 | 3.193(1) | 2.322(4) | 2.330(6) | 74.7(2) | 174.0(2), 176.3(2) | 40 | |
Pd(II) | EDTC | +9 | 41 | |||||
PDTC | +13 | 2.9457(4) | 2.353(2) | 2.343(2) | 75.69(8) | 180.00(8) | 41 | |
bmpDTC | −2 ± 1 | 2.788(3) | 2.339(3) | 2.333(3) | 75.69(7) | 180.00(7) | This work | |
(CH_{3})_{2}DTC | +32 ± 3 | 2.9216(8) | 2.337(1) | 2.309(1) | 75.52(3) | 180.00(3) | 12 |
Luminescence spectra measured on crystals at variable pressure are presented in Fig. 2. Broad bands are observed, which is typical for the d–d transitions reported for dithiocarbamate complexes.^{42} For [Pt(bmpDTC)_{2}], the luminescence maximum at ambient pressure appears at approximately 15600 cm^{−1} and shifts to lower energy as pressure increases. Two ΔE_{max}/ΔP values are measured in two different pressure ranges: −8 ± 1 cm^{−1}/kbar from 0 to 28 kbar and −42 ± 6 cm^{−1}/kbar from 28 to 49 kbar, as seen in Fig. 3. To the best of our knowledge, these are the first negative ΔE_{max}/ΔP values measured for platinum(II) dithiocarbamate complexes, as documented in Table 3, strongly deviating from the average ΔE_{max}/ΔP value of +12 ± 2 cm^{−1}/kbar. The intensity of the luminescence spectra decreases significantly with increasing pressure, as seen in Fig. S1,† leading to the higher noise on the low-energy side of the spectra at pressures above 30 kbar in Fig. 2 where spectra are plotted with matching maximum intensities. This decreasing luminescence intensity at high pressure is a common phenomenon for crystalline samples, caused by increasingly efficient non-radiative relaxation processes. Energy transfer from excited molecules to pressure-induced defects acting as deep traps occurs, with both the number of defects and the efficiency of the energy transfer increasing with pressure. The intensity decrease does not affect the width of the luminescence bands, as illustrated by the constant full-widths at half-maximum (FWHM) in Fig. S2.†
The luminescence maxima of [Pd(bmpDTC)_{2}] show a smaller shift to lower energy than for the platinum(II) analog, with a ΔE_{max}/ΔP value of −2 ± 1 cm^{−1}/kbar as shown in Fig. 3. Its luminescence maximum at ambient pressure is at 14600 cm^{−1}, lower in energy by approximately 1000 cm^{−1} than for the platinum(II) analog. The full-width at half-maximum (FWHM) is also lower than for the platinum(II) complex, with values of 2700 cm^{−1} and 4200 cm^{−1}, respectively. These bandwidths are similar to values reported for other dithiocarbamate and thiocyanate complexes, between 2500 cm^{−1} for palladium(II) and 4000 cm^{−1} for platinum(II).^{41}
Raman spectra at variable pressure and temperature are shown in Fig. S3–S6.† The overall peak pattern stays constant, with broader peaks at higher temperature or pressure. We can use the published calculated spectra of [Pd{(CH_{3})_{2}DTC}_{2}]^{31} to assign matching peaks of [Pd(bmpDTC)_{2}], namely those at 344 and 412 cm^{−1} assigned to the Pd–S asymmetric and symmetric stretching modes, respectively, and 892 and 1000 cm^{−1} to the symmetric and asymmetric C–S stretching modes. The stretching modes of the ligands are at approximately the same frequencies for the platinum(II) analog, but Pt–S stretching are expected to appear at higher energies. Peak maxima shift gradually to higher frequency with increasing pressure. The asymmetric C–S stretching frequency is presented in Fig. S7,† with shifts of +0.18 ± 0.01 cm^{−1}/kbar and +0.22 ± 0.01 cm^{−1}/kbar for [Pd(bmpDTC)_{2}] and [Pt(bmpDTC)_{2}], respectively. Broadening of the peaks with increasing pressure is also observed. A typical width changes from 5.6 to 8.1 cm^{−1} for [Pt(bmpDTC)_{2}] and from 4.3 to 8.2 cm^{−1} for [Pd(bmpDTC)_{2}], from ambient pressure to approximately 50 kbar for the same vibrational mode. This broadening is documented in the literature,^{45} attributed to crystal defects, and was previously reported for other dithiocarbamate complexes.^{41} The absence of any discontinuity in the pressure- or temperature-induced shifts of the peak maxima is evidence for the absence of phase transitions for both complexes, in particular for [Pt(bmpDTC)_{2}] in the pressure range where the slope of the variable-pressure luminescence maxima changes in Fig. 3.
Fig. 4 Comparison of ΔE_{max}/ΔP slopes for dithiocarbamate complexes of platinum(II) (left) and palladium(II) (right). Points represent luminescence maxima of Fig. 2. Lines are traced from ΔE_{max}/ΔP and E_{max} at ambient pressure from the literature. |
For platinum(II) complexes, a variation of E_{max} by approximately 3000 cm^{−1} is shown in Fig. 4, corresponding to a variation by 23%, induced only by varying the substituents on the periphery of the dithiocarbamate ligands. This is surprisingly high in view of the E_{max} variation by only 2% between K_{2}[PtBr_{4}] and K_{2}[PtCl_{4}] resulting from the different ligands. Even changing from a Br^{−} ligand to a SCN^{−} ligand induce a lower E_{max} variation of 17% than the difference observed among DTC complexes, as compared with K_{2}[Pt(SCN)_{4}], with a E_{max} of 14650 cm^{−1}.^{46} Modification on the substituent on the periphery of ligands is not expected to induce higher energy changes than variation of the ligands itself, as supported by the spectrochemical series. The palladium(II) dithiocarbamate analogs show a lower E_{max} variation of 1300 cm^{−1} (12%) than their platinum(II) counterparts. This is also a more important variation of E_{max} than the one induced by variation of halide ligands between K_{2}[PdBr_{4}]^{47} and K_{2}[PdCl_{4}],^{47} which induce a 2% variation. The E_{max} variation for palladium(II) in DTC complexes is of the same order of magnitude as the change from K_{2}[PdBr_{4}] to (n-Bu_{4}N)_{2}[Pd(SCN)_{4}],^{48} which is of 14%.
DFT calculations of the lowest-energy absorption maximum on single molecules of square-planar complexes show quantitatively comparable variations, indicating that the spectrochemical series theory is applicable to square-planar complexes.^{49} For instance, the calculated differences between absorption maxima are 900 cm^{−1} (10%) between [PtBr_{4}]^{2−} and [PtCl_{4}]^{2−} and 1500 cm^{−1} (16%) between [PtBr_{4}]^{2−} and [Pt(SCN)_{4}]^{2−}. For palladium(II), a 9% variation is calculated between [PdBr_{4}]^{2−} and [Pd(SCN)_{4}]^{2−}.^{49} In contrast, the surprising variation of E_{max} observed within the dithiocarbamate family can not be explained by the spectrochemical series. Clearly, other factors than the nature of the metal and ligator atoms have to be taken in account to understand the variation occurring in these complexes.
An important factor involved in the d–d luminescence energy of a single complex is the local symmetry at the metal site. The dithiocarbamate platinum(II) complex with the lowest E_{max} value (13000 cm^{−1}) in Fig. 4 is [Pt(dopDTC)_{2}]. The substituents on the ligand of this complex induce a small distortion of the square-plane resulting in a τ_{4} value^{50} of 0.04, in contrast to all the other square-planar complexes analyzed here which have a τ_{4} value of 0 and perfect inversion symmetry. The τ_{4} value is the geometry index for ML_{4} complexes varying from square-planar (τ_{4} = 0) to tetrahedral (τ_{4} = 1), and is defined by:
(1) |
This structural distortion contributes to the variation of E_{max} values, but is not applicable to DTC complexes with perfect inversion symmetry. Their differences in E_{max} are due to external effects, such as M⋯H–C interactions. The highest E_{max} value reported in Fig. 4 for a square-planar platinum(II) DTC complex is 16000 cm^{−1} for [Pt(EDTC)_{2}].^{41} In its crystal packing, there is no interaction with the metal since the closest C–H bond is poorly aligned, with a M⋯H distance of 3.06 Å. The HOMO–LUMO energy difference is therefore entirely due to the molecular structure, without additional contributions from intermolecular interactions. For the [Pt(bmpDTC)_{2}] complex, the E_{max} value is 15600 cm^{−1}, lower in energy than for [Pt(EDTC)_{2}]. A weak M⋯H–C interaction perpendicular to the MS_{4} plane is assumed (Fig. 5), as reported for the palladium(II) analog.^{35} Interaction with the d_{z2} orbital usually induces a destabilization of the HOMO level, decreasing the luminescence energy, in agreement with the lower E_{max} than for [Pt(EDTC)_{2}]. The [Pt{(CH_{3})_{2}DTC}_{2}]^{12} complex shows an even lower E_{max} value of approximately 14200 cm^{−1}, indicating a more destabilized HOMO and, thus, a stronger M⋯H–C interaction. Even if the intermolecular Pt–H distance of 2.99 Å for [Pt{(CH_{3})_{2}DTC}_{2}] is longer than for [Pt(bmpDTC)_{2}] with 2.81 Å, the alignment is more favorable to overlap with the d_{z2} electron density, as seen in Fig. 5d and f, rationalizing the different E_{max} values. For the palladium(II) complexes, E_{max} values at ambient pressure for [Pd(bmpDTC)_{2}], [Pd(EDTC)_{2}],^{41} and [Pd(PDTC)_{2}]^{41} are very similar, within 300 cm^{−1}, as shown in Fig. 4 and Table S1.† Both [Pd(EDTC)_{2}] and [Pd(PDTC)_{2}] show no short M⋯H–C distances in their crystal structure nor adequate alignment. [Pd(bmpDTC)_{2}] shows a very similar E_{max} value, indicating that the reported^{35} Pd⋯H–C interaction is not strong enough at ambient pressure to significantly influence the HOMO energy of this complex. The E_{max} value alone is therefore not sufficient to identify the presence of M⋯H–C interactions in this system even though there is evidence from the crystal structure. The [Pd{(CH_{3})_{2}DTC}_{2}]^{12} complex shows E_{max} at 13000 cm^{−1}, lower in energy by approximately 1400 cm^{−1} than for complexes with no interaction, indicative of stronger Pd⋯H–C interactions in this complex. For the platinum(II) complexes, the corresponding difference is 1800 cm^{−1} between [Pt(EDTC)_{2}] and [Pt{(CH_{3})_{2}DTC}_{2}]. The platinum(II) dithiocarbamate complexes show bigger variations of E_{max} due to M⋯H–C interactions than their palladium(II) analogs.
Variable-pressure luminescence spectra allow us to characterize the impact of intermolecular interactions by comparing the ΔE_{max}/ΔP values for the compounds discussed here by determining deviations from the +12 ± 2 cm^{−1}/kbar average. For [M(bmpDTC)_{2}] complexes, shifts to lower energy are observed, which is unusual for dithiocarbamate complexes, as documented in Table 3. The only other dithiocarbamate complex with such a shift to lower energy is [Pd{(CH_{3})_{2}DTC}_{2}] in the 50 to 85 kbar range with a ΔE_{max}/ΔP value of −31 ± 8 cm^{−1}/kbar.^{31} Only one other palladium(II) complex with a negative ΔE_{max}/ΔP value is known: [Pd{PyCHC(C_{3}F_{7})O}_{2}],^{51} a complex with chelating ligands with O and N ligator atoms, with a ΔE_{max}/ΔP value of −15 ± 7 cm^{−1}/kbar, again due to an Pd⋯H–C interaction. Several negative ΔE_{max}/ΔP values have been reported for platinum(II) complexes, namely for M_{x}[Pt(CN)_{4}]·mH_{2}O,^{5} showing values from −320 to −140 cm^{−1}/kbar caused by efficient overlap of d_{z2} orbitals of neighbouring platinum(II) ions. A relevant example is [Pt(SCN)_{2}{(μ-SCN)Mn(NCS)(bipy)_{2}}_{2}],^{52} with a ΔE_{max}/ΔP value of −99 cm^{−1}/kbar due to Pt⋯H–C interactions illustrated in Fig. 5a and b. Negative values of ΔE_{max}/ΔP arise from extended destabilization of the HOMO as pressure increases by shortening of the intermolecular distances. This comparison of ΔE_{max}/ΔP values places [M(bmpDTC)_{2}] complexes intermediate between complexes without interaction (ΔE_{max}/ΔP of +12 ± 2 cm^{−1}/kbar) and complexes with strong M⋯H–C interactions (−99 cm^{−1}/kbar). In the crystal structures of the [M(bmpDTC)_{2}] complexes, the hydrogen atom is slightly misaligned with the metal, as seen in Fig. 5c and d, leading to a weaker M⋯H–C interaction. The difference is striking compared to the much better M⋯H alignment for [Pt(SCN)_{2}{(μ-SCN)Mn(NCS)(bipy)_{2}}_{2}], illustrated in Fig. 5a and b.
The [M{(CH_{3})_{2}DTC}_{2}] complexes in Fig. 5e and f show ΔE_{max}/ΔP values of +44 ± 5 cm^{−1}/kbar (ref. 31 and 40) for the platinum(II) complex and +32 ± 3 cm^{−1}/kbar (ref. 31) for its palladium(II) analog from ambient pressure to 25 kbar, higher values of ΔE_{max}/ΔP than the average +12 ± 2 cm^{−1}/kbar. These high positive shifts are due to a loss of a M⋯H–C interaction as pressure increases. In these compounds, the CH_{3} substituent can rotate around the N–C bond, leading to a likely reorientation of the C–H bonds as pressure increases. This reorientation can strengthen or weaken the M⋯H–C interaction with pressure. In a structure where the carbon is part of a bulky cyclic ligand substituent, such as in [Pt(SCN)_{2}{(μ-SCN)Mn(NCS)(bipy)_{2}}_{2}], changes in the orientation of the C–H bond are restricted unless the whole molecule moves, making this an unlikely rearrangement. As pressure increases, the distance between interacting atoms decreases, resulting in a stronger M⋯H–C interaction and a negative ΔE_{max}/ΔP value of −99 cm^{−1}/kbar. In the [M(bmpDTC)_{2}] compounds, the interaction originates from a CH_{2} group, an intermediate situation in terms of possible reorientation of C–H bonds. Rotation of the C–N bond is formally possible, but strongly limited by the bulky substituents, allowing only small reorientations for the CH_{2} group. The small negative ΔE_{max}/ΔP values of −8 ± 1 cm^{−1}/kbar for [Pt(bmpDTC)_{2}] and −2 ± 1 cm^{−1}/kbar for [Pd(bmpDTC)_{2}] indicate that the M⋯H–C interaction is still present in these complexes as pressure increases, as ΔE_{max}/ΔP is lower than the +12 ± 2 cm^{−1}/kbar, but the position of the C–H bonds is not optimal for the interaction, leading to the low absolute ΔE_{max}/ΔP values. M⋯H–C interactions lead to lower E_{max} values at ambient conditions, but pressure-induced shifts of E_{max} towards lower energy can not be reliably predicted based only on the maxima at ambient conditions. These shifts depend on the orientation and mobility of the C–H bonds involved. Luminescence measurements at variable pressure reveal additional information, as shown by the comparison of M⋯H–C interactions involving mobile CH_{3} groups, less mobile CH_{2} groups in the bmpDTC complexes and immobile C–H bonds on bulky aromatic ligand fragments.
Since the crystal structures of [Pd(bmpDTC)_{2}] and [Pt(bmpDTC)_{2}] complexes are identical, with similar densities of 1.527 g cm^{−3} and 1.770 g cm^{−3}, respectively, comparable decreases of intermolecular distances as well as identical variations of E_{max} with pressure are expected. Significantly different ΔE_{max}/ΔP values are observed for the palladium(II) and platinum(II) complexes, indicating effects of different magnitude for the two metal centers. [Pt(bmpDTC)_{2}] shows two different ΔE_{max}/ΔP slopes with a more negative trend at pressures above 28 kbar, typical for a stronger interaction. This is not the case for the palladium(II) analog, where ΔE_{max}/ΔP stays close to zero across the full pressure range studied, evidence for a weak interaction. These differences indicate that, over the same pressure range, stronger intermolecular interactions occur in the crystalline platinum(II) complex than in its palladium(II) analog. The same qualitative trend is observed for the [M{(CH_{3})_{2}DTC}_{2}] complexes, illustrating the influence of the metal ion on M⋯H–C interactions. Palladium(II) and platinum(II) have similar ionic radii^{53} of 0.64 Å and 0.60 Å, respectively. In square-planar complexes, they also show similar metal–ligand bond lengths. It is, however, important to keep in mind that the 4d and 5d radial wave functions are different. In the direction perpendicular to the coordination plane, the main electron density arises from the d orbitals only, with the 4d function for palladium(II) extending significantly less than the 5d function for platinum(II). This difference qualitatively rationalizes the stronger negative shifts for platinum(II) complexes observed in the same pressure range as for palladium(II).
Footnote |
† Electronic supplementary information (ESI) available: Table of published luminescence band maxima E_{max} for square-planar platinum(II) and palladium(II) complexes, variable-pressure luminescence spectra, variable-pressure full-width at half-maximum (FWHM), variable-pressure and variable-temperature Raman spectra, cif file for [Pt(bmpDTC)_{2}] at 150 and 295 K. CCDC 1527919 and 1527920. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c7dt00545h |
This journal is © The Royal Society of Chemistry 2017 |