Square-Planar Pd(II) Complexes

Probably the most classical studies of the piezochromism of metal complexes concern Ni(II), Pd(II) and Pt(II) complexes with dimethylglyoximate, L¹: dmg⁻. (Ligands appearing in this review and other related ligands are summarised in the Scheme.) These complexes all have square-planar, D_4h, symmetry, and the absorption spectra for each complex in solution and solid form are very different. The crystal structure of [Ni(dmg)₂] was reported by Godychi and Rundle (12). The planar [Ni(dmg)₂] units are stacked one upon another with a rotation of 90° for each alternate layer. The Ni(II) ions line up with an average Ni—Ni distance of 3.233 Å. The Pd(II) and Pt(II) complexes exhibit similar structures with the Pd—Pd and Pt—Pt distances being 3.253 and 3.25 Å, respectively (13, 14). As the M(II)—M(II) distances in these complexes are quite similar, the factor that governs the distances is attributed to the organic groups in the ligand. The complexes show sharp absorption bands near 19,000 cm⁻¹ in the solid, but these rather strong absorption bands are not observed in solution. The bands originate from M(II)—M(II) interactions. Furthermore, the absorption bands are dichroic with the perpendicular component having high intensity.

Zahner and Drickamer examined the pressure dependence of these complexes as solids (15). They used diluted salt pellets for the measurements. For all the complexes, the absorption band near 19,000 cm⁻¹ shifted towards lower energy with pressure. The observed pressure effects may be summarised as follows:

In the solid, the distance between the M(II) ions is very small. The positive charges located on the z axis of the planar complex cause an energy decrease in the vacant p_z and s orbitals, although the energy levels of the p_x and p_y orbitals are little influenced by the adjacent M(II). Therefore, the specific absorption band observed in the solid is assigned to the allowed (z- and x,y-polarised) d-p_z transitions; the d-s transition is Laporte forbidden.

When external pressure is applied to the solid specimen, the effect of the positive charge on the vacant p_z and s orbitals increases: as the distance between two M(II) ions decreases, the energy level of the vacant p_z orbital continuously decreases until a point where the vacant p_z orbital and the occupied d_z² orbital on the adjacent M(II) ions start to overlap. Further compression of the solid then leads to an increase of the p_z energy level and induces a blue shift of the absorption band. As the principal quantum number increases, it is expected that the corresponding s, p and d orbitals gradually expand. However, such a tendency does not apply to the 6s orbital of the third-row transition metals.

The larger relativistic effect for the third-row transition elements causes a contraction of the 6s orbital and the 5d and 6p orbitals become relatively higher in energy than is expected for these elements. Therefore, it seems that for M = Pt, the overlap of the vacant 6p_z orbital and the occupied 5d_z² orbital on adjacent M(II) ions takes place at a rather low pressure. The pressure broadening of the absorption bands was explained by delocalisation of the electron density from A_1g and E_g in the one-dimensional metallic lattice. A similar pressure dependence of the absorption spectra was observed for the Ni(II) and Pd(II) complexes with 1,2-cyclohexanedione dioximate (L²: niox⁻).

A series of studies, including measurements of the electrical conductivity at very high pressures (up to 14 GPa), was carried out for the complex dimethylglyoximatoplatinum(II) (16, 17). Powder X-ray diffraction at elevated pressures revealed that the Pt—Pt distance continuously decreased by ca. 17% up to 14 GPa. A levelling-off in resistivity with pressure was observed at ca. 6.5 GPa. This was attributed to the phase transition from the one-dimensional “metal” to the “semiconductor.” Most interesting results were reported for the shift of the UV-vis bands at very high pressures: the MLCT (metal to ligand charge transfer) band that appeared at the 320 nm region at ambient pressure was observed at ca. 500 nm at 9 GPa. It seems that the large red shift of the MLCT band is related to the “metal” to “semiconductor” phase transition.

An extension of these studies appeared in this Journal in 1987 (18). The d_z²(a_1g) → p_z(a_2u) absorption in the Pd^II(niox)₂ complex exhibited a large pressure dependence. Accordingly, the colour of this complex changed from yellow orange (0.8–1.4 GPa) to red (1.4–2.3), to purple (2–3), to blue (2.7–5.5), to yellow green (5.5–6.5), and to pale yellow (above 7 GPa), see Figure 3 (18).

It has been suggested that the use of this piezochromism would enable us to monitor pressure in situ, although further investigations concerning the relationship between the thickness of the specimen and the observed colours under each pressure may have to be carried out.

Mixed-Valent Pd(II)/Pd(IV) Complexes

Pressure effects on the conductivity of crystals and powders of various platinum compounds have been investigated in relation to the electronic interactions between two metal sites through bridging ligands. Pt and Pd complexes with d⁸ electronic configuration normally prefer square-planar coordination geometry, while metal ions with d⁷ and d⁶ electronic configuration prefer octahedral 6-coordination. Although early studies of 1-electron oxidised M(II) complexes, MAX₃ (M = Pt or Pd; A = (NH₃)₂; X = Cl⁻, Br⁻, or I⁻) suggested the existence of M(III) species (19), later structural analyses revealed that these complexes are mixtures of square-planar M(II) and octahedral M(IV) complexes with chains: [—M(II)–X–M(IV)–X—M(II)—] in the crystals (20, 21). These complexes are diamagnetic and have been classified as “Class II compounds”, according to the criteria of Robin and Day (22). The observed absorption bands for these compounds are therefore superpositions of the individual absorption bands for M(II) and M(IV) complexes with a M(II) to M(IV) inter-valence transfer (IT) band in the visible region. The intensity of the latter IT bands is high and these compounds are highly coloured.

Interrante and colleagues investigated the possibility of a spectral shift and a change in conductivity upon compression of the crystals along the —M(II)—X–M(IV)–X— axis for various MAX₃ complexes:

For Pd(NH₃)₂Cl₃, isotropic compression in all three directions of the orthorhombic unit cell was observed, including compression along the intra-chain Pd(II)—Cl–Pd(IV) distance. Observation of the absorption spectra revealed that the IT band was very strong for most of the complexes, even at ambient pressure, and measurement became impossible at elevated pressures.

An exceptional result was obtained for complex Pt(C₂H₅NH₂)₄Cl₃·2H₂O where an absorption change with pressure in the UV-vis region was observed. The absorption maximum at 16,000 cm⁻¹ at ambient pressure for a single crystal (consistent with the previously reported absorption band corresponding to the M(II) to M(IV) intervalence transition) shifted towards longer wavelengths at ca. 30 kbar. This shift in the IT band indicates that the compression along the M(II)—X–M(IV)–X— axis is due to the applied pressure.

The conductivities of these complexes at ambient temperature reversibly increased with pressure. The plot of the logarithmic value of the conductivity (ohm⁻¹ cm⁻¹) against the reciprocal temperature at each pressure was linear. The activation barrier for the change in conductance was plotted against pressure, see Figure 4. The activation barrier was initially found to decrease with pressure up to ca. 106 kbar and then to begin to increase with pressure. However, no phase transition to the metallic state was observed, even at 140 kbar. The inhibition of this phase transition may be caused by the increase in the activation barrier at pressures higher than 106 kbar. Therefore, it was not possible to achieve the Class III state (where a complete electronic delocalisation occurs, according to the Robin and Day definition (22)) for this complex even at very high pressures. The Pd(II)—Cl and Pd(IV)–Cl distances have been reported as 322 pm and 199 pm, respectively, at ambient pressure. As compression of the Pd(IV)–Cl distance is difficult, compression of the Pd(II)—Pd(IV) distance by ca. 35 pm at 60 kbar may be attributed to the decrease in the Pd(II)—Cl distance. However, such a change in the Pd(II)—Cl distance is still too small for the transition to the Class III state, where the Cl atom is expected to sit between a Pd(II) and a Pd(IV) ion.

A more recent analysis of the complex [Pd(chxn)₂][Pd(chxn)₂Br₂]Br₄ (L³: chxn = 1R,2R-cyclohexanediamine) indicates that Pd(II) and Pd(IV) in the one-dimensional chain are energetically close and produce 10⁻³ of paramagnetic Pd(III) (24). Moreover, this mixed-valence palladium complex, which is dark brown at ambient pressure, turns light green upon exposure to Br₂: an uptake of two Br₂ molecules to each [Pd(chxn)₂][Pd(chxn)₂Br₂] (25). This process induced the oxidation of Pd(III) to Pd(IV).

Electronic conduction in solid samples generally takes place through:

In the former case, the application of external pressures may strengthen the interactions between d_z² orbitals by axial compression: the phase transition to achieve the ultimate electronic delocalisation may be expected to occur. In the latter case, the ultimate state/condition for this type of interaction is as “organic metal”, such as the TCNQ-TTF (tetracyanoquinodimethane-tetrathiafulvalene).

A mixed-valence palladium complex with a conjugated planar macrocyclic ligand, tetrabenzo-[b, f, j, n]-1,5,9,13-tetraazacyclohexadecine (L⁴: TAAB), was examined at elevated pressures (27). The mixed-valence complex was prepared by the partial oxidation of the corresponding Pd(II) complex by I₂. Powder and crystalline samples of the resulting [Pd(TAAB)]^2.7+(I₃⁻)_2.7 was examined by IR and Raman spectroscopic methods. It was revealed that the I₃⁻ anion exists as a linear unit in the crystal, based on the selection rules. Although X-ray analysis was not carried out for this mixed-valence complex, it was concluded that the I₃⁻ ion is located in space with little perturbation from the neighbouring group. Therefore, this mixed-valence complex does not have a one-dimensional chain structure. The mixed-valence complex is dark red and this colour may relate to the ML/LMCT band. The conductivity of the complex increased from 10⁻⁸ ohm⁻¹ cm⁻¹ at ambient pressure to 10⁻⁵ ohm⁻¹ cm⁻¹ at 10 kbar. The conductivity of solid [Pd^II(TAAB)]²⁺ was 10⁻⁸ ohm⁻¹ cm⁻¹ at ambient pressure and was independent of pressure. Moreover, the spectra of the mixed-valence complex hardly changed with pressure. Therefore, the small pressure-dependent conductivity of this mixed-valence complex was attributed to the crystal packing effect.

Pressure Effects on Emission Spectra, Emission Intensities and Lifetimes of Excited Pd(II) Complexes

One of the most intensively studied Pd(II) complexes in the last decade is [Pd^IIL₄]²⁺ (L = SCN⁻ or SeCN⁻). The XCN⁻ ligands are ambidentate and may coordinate through X or N, depending on the polarisability of the central metal ion. Crystal structures of the [Pd(XCN)₄]²⁻ fragments have D_4h symmetry; the average Pd–S and Pd–Se distances are 2.33 and 2.44 Å, respectively, and the M–X–C angles were 109° and 105–107° for the tetra-n-butylammonium salt of the SCN⁻ and SeCN⁻ complexes, respectively (28).

Rhode and coworkers (28) also reported normal coordinate analyses for these complexes. The force constants for Pd–S and Pd–Se were ca. 1.17 mdyn Å⁻¹, which is significantly smaller than those for the corresponding Pt(II) complexes: Pt–S is 1.44 mdyn Å⁻¹ and Pt–Se is 1.42 mdyn Å⁻¹. The a_1g totally symmetric vibrations of the Pd–S and Pd–Se bonds were 274–303 cm⁻¹ and 180–195 cm⁻¹, respectively, while b_1g non-totally symmetric stretching and b_2g non-totally symmetric bending modes were observed at 260–290 and 170–187 cm⁻¹ for Pd–SCN, and at 140–150 and 100–110 cm⁻¹ for Pd–SeCN complexes, respectively. The ground-state electronic configuration for these complexes is: a_1g(d_z²)²b_2g(d_xy)²e_g(d_xz,d_yz)⁴b_1g(d_x²−y²)⁰ (29) and one-electron excitation from the e_g orbital to the σ antibonding b_1g orbital creates degenerate ³E_g and ¹E_g excited states. The electron occupancy of the σ antibonding b_1g orbital induces totally symmetric (a_1g) elongation of all Pd–X bonds, while the doubly degenerate electronic excited states cause Jahn-Teller distortion – that is, elongation/contraction along one diagonal X–Pd–X axis takes place.

The luminescence bands for these complexes are broad, indicating that the potential energy minimum of the lowest excited ³E_g state is coupled strongly with the Pd–X vibration modes. At ambient temperature, the emission centred at ca. 12,500 cm⁻¹ is almost unobservable, however the intensities as well as the emission lifetime increase with decreasing temperature. Therefore, effective participation of the non-radiative decay mechanism exists at relatively high temperatures. Furthermore, the low temperature emission spectra of these complexes exhibit well-resolved vibronic structures (< 50 K).

Reber and coworkers examined the pressure effects on the d-d luminescence spectra of these Pd(II)–XCN and Pt(II)–XCN complexes (X = S and Se) up to 40 kbar (30). The intensity of the luminescence increased with pressure for all the complexes. The peak position steadily blue-shifted with pressure: 24, 12, 29 and 25 cm⁻¹ kbar⁻¹ for Pt–SCN, Pd–SCN, Pt–SeCN and Pd–SeCN, respectively. However, such a blue-shift may not be explained by the axial metal—metal interactions that [Pt(CN)₄]²⁻ exhibits. Indeed, the metal—metal distance in the crystals of these complexes (ca. 13 Å) are much longer than in the tetracyanoplatinate complex (ca. 3.1–3.7 Å) (28).

In [Pd(XCN)₄]²⁻ and [Pt(XCN)₄]²⁻ the pressure dependence of the luminescence intensities and lifetimes was attributed to the loss of the inversion centres caused by the applied pressure: the probability of the d-d transition increases with the decreasing centrosymmetry of the complex (11). It was shown that the participation of the slow b_2g bending motion coupled with the intermolecular forces induced by the external pressure also contributes to the enhanced intensities of the emission spectra. Such a distortion of the excited state explains the spectral shifts: the bending motion inherent in these complexes may reduce the pπ-dπ interactions between the HOMO (e_g orbitals) and ligand nπ orbital and cause a blue shift of the luminescence bands with pressure. This distortion also enhances the luminescence lifetimes by contributing to the loss of centrosymmetry in the chromophore. No enhancement of the luminescence intensity was observed for K₂[PtBr₄] in which pressure-induced coupling of the b_2g bending mode is not expected.

The luminescence lifetimes of the [Pd(XCN)₄]²⁻ complexes were significantly enhanced with pressure: 62 µs and 48 µs at ca. 30 kbar, and 4 µs and 541 ns at 7 kbar, for X = Se and S, respectively, see Figure 5. By contrast, the lifetimes for the corresponding Pt complexes were not enhanced much: 6 µs and 13 µs at ca. 30 kbar, and 2 µs and 750 ns at 5 kbar, for X = Se and S, respectively. The radiation lifetime is expected to decrease when the d-d transition becomes more allowed by the loss of centrosymmetry with pressure. Therefore, the enhanced luminescence lifetimes observed for these complexes indicate the significant decrease in the rate of the non-radiative process with pressure. The validity of this was verified by the use of Englman-Jortner's radiationless decay theory (31).

Conclusions

In this short review topics concerning piezochromism and related phenomena in solid palladium complexes have been summarised. The data may help towards finding use for these metal complexes, such as for in situ pressure sensors or as conductors in extreme environments.