Structure and spectroscopic properties of porphyrinato group 14 derivatives: Part I – Phenylacetylido ligands

2.1 General procedures

All reactions were carried out under nitrogen using common Schlenk techniques. Each flask was flame-dried before its use. Nitrogen was dried via a column of molecular sieves (3 Å) and P₄O₁₀. Organic solvents were dried via a solvent drying system from Innovative Technology Inc. The water content was determined using Karl–Fischer titration and was found to be <5 ppm. Phenylacetylene 99% (Ph–C≡CH) was purchased from Sigma Aldrich and was distilled into a small flask. “AcroSeal” n-BuLi, 1.6 molar in hexane, was purchased from Acros organics. TPPH₂ 97% was purchased from ABCR GmbH & CoKG and was used as received. Deuterated solvents (CDCl₃, C₆D₆) were purchased from Deutero GmbH, VWR Int., dried with P₄O₁₀ and distilled and stored in a flask over molecular sieves (3 Å). Elemental mass analyses were carried out using an Elementar Vario instrument by Heraeus Elementar.

NMR-spectroscopic measurements were performed on a Mercury 300 MHz spectrometer (Varian) at room temperature. Measurements of ¹H (300 MHz), ¹³C (75.4 MHz) and ²⁹Si (59.6 MHz) were carried out using tetramethylsilan (δ=0 ppm) as a reference; measurements of ¹¹⁹Sn (111.8 MHz) were related to Me₄Sn (δ=0 ppm).

High-resolution mass spectrometry (HRMS) (matrix-assisted laser desorption/ionization with time-of-flight – MALDI-TOF) spectra were recorded on a MALDI micro MX (Waters). Sample preparation was carried out as quickly as possible in air. The sample was dissolved in CH₂Cl₂ and crystallized on the target either with or without matrix. Used matrix materials were anthralin and DCTB. Ionization was performed using a 337 nm laser. Mass filtration was performed using a TOF analysator incl. repeller. Calibration used a solution of NaTFA (1 mg/mL) with PEG600 (5 mg mL⁻¹) and PEG1000 (5 mg mL⁻¹).

UV/Vis spectra were acquired under inert atmosphere in quartz glass cuvettes of thickness 1 cm at a Cary 60 UV-Vis from Agilent Technologies.

Luminescence spectra were recorded on a Fluorolog-3 fluorescence spectrometer (from Horiba Jobin Yvon Inc.) equipped with a photomultipliertube (PMT) (Hamamatsu, model 2658). The spectra were corrected for the spectral sensitivity of the PMT using the manufacturer’s correction function in the software. Sample preparation was performed under inert atmosphere in one-sided fused glass tubes of 5 mm diameter and 0.75 mm wall thickness. The tubes were sealed with plastic plugs and parafilm. THF–toluene (v/v, 6:4) was used due to its ability to form solvent glass at low temperature. The solvents were degassed by three freeze-pump-thaw cycles before mixing. Low temperature measurements were performed using FL-1013 Liquid Nitrogen Dewar Assembly (Horiba). The quantum yields were determined relative to 1,6,7,12-tetraphenoxy-N,N′-bis(2,6-diisopropylphenyl)-perylene-3,4,9,10-tetracarboxylic acid diimide (Fluoreszenzrot 94720, from Kremerpigmente, Φ_f=0.96) [4]. Sample concentrations were adjusted to achieve an absorption of approx. 0.15 at the excitation wavelength. Luminescence decay times were measured with a time correlated single photon counting (TCSPC) set-up from Horiba, using a pulsed laser diode nanoLED (453 nm, pulse duration 1.1 ns, Horiba) and spectraLED (392 nm full width at half maximum: 17 nm, Horiba) for fluorescence and phosphorescence, respectively. Analysis of decay data was performed using Das 6.6 Software.

2.2 Crystal structure determinations

Crystals suitable for single-crystal X-ray diffractometry were taken from a vial or a Schlenk tube under N₂ and immediately covered with a layer of silicone oil. A single crystal was selected, mounted on a glass rod on a copper pin, and placed in the cold N₂ stream provided by an Oxford Cryosystems cryostream. XRD data collection was performed on a Bruker Apex II diffractometer with the use of an Incoatec microfocus sealed tube with graphite-monochromatized MoK_α radiation (λ=0.71073 Å) and a CCD area detector. Data collection was performed using ϕ and ω scans. Data reduction and cell refinement were done with Bruker Saint. Empirical absorption corrections were applied using Sadabs or Twinabs [14], [15]. The structures were solved with the use of the intrinsic phasing option in Shelxt [16] and refined with full-matrix least-squares procedures (Shelxl [17], [18]). Non-hydrogen atoms were refined anisotropically. Hydrogen atoms were located in calculated positions using standard bond lengths and angles and refined using a riding model. The space group assignments and structural solutions were evaluated using Platon [19], [20]. Disorder, as observed for compound 5, was treated by modeling the occupancies of the individual orientations using free variables to refine the respective occupancy of the affected fragments (PART [21]). In compound 5, the similar-ADP restraint (SIMU) was used in modeling disorder to make the ADP values of the disordered phenyl group on the phenylacetylide ligand more reasonable. Disordered positions for one of the acetylide phenyl rings were refined using 50/50 split positions with additional restraints to afford optimized geometries (AFIX 66). The anisotropic U_ij values of these atoms were restrained (ISOR) to behave more isotropically. The disordered solvent of crystallization for compound 6 was removed from the refinement by using the SQUEEZE routine available in Platon [22]. Electrostatic non-covalent intermolecular interactions [6], [23], [24], [25] for presented and published compounds were based on a Cambridge Structural Database search and fall within expected ranges. Centroids and planes were determined by features of the programs Mercury [26] and Diamond [27]. All crystal structure representations were made with the program Diamond. Table 1 contains crystallographic data and details of measurements and refinement for compounds 4, 5 and 6.

CCDC 1571785, 1553427 and 1553428 contain the supplementary crystallographic data. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre viawww.ccdc.cam.ac.uk/data_request/cif.

2.3 Density functional theory (DFT) calculations

All calculations were carried out using the Gaussian 09 program package [28] on a computing cluster with blade architecture. The mPW1PW91 hybrid functional [29] was used throughout. For geometry optimizations and vibrational frequency calculations, the SDD denoted the combination of D95 all-electron basis sets [30] on elements up to argon and Stuttgart–Dresden effective core potentials [31], [32] on the heavier elements were employed.

2.4 Synthesis of bis(chlorido)(meso-tetraphenylporphyrinato) E(IV); E=Si (1), Ge (2), Sn (3)

Synthesis of 1 was performed according to [33] in 89% yield. – ¹H NMR (C₆D₆): δ=8.85 (s, 8H, β-H), 7.77 (d, ³J=6.4 Hz, 8H, o-H), 7.31 (m, 12H, m-H, p-H). – ²⁹Si NMR (C₆D₆): δ=–217.3 (s). – ¹³C NMR (CDCl₃): δ=142.73 (s), 138.86 (s), 133.92 (s), 132.05 (s), 128.45 (s), 127.37 (s), 117.21 (s) (all sp²-hybridized). – UV/Vis (C₆H₆): λ_max (lg ε_max)=418 nm (4.43), 437 nm (5.48), 528 nm (3.40), 569 nm (4.08), 610 nm (3.78).

Synthesis of 2 was performed according to [34]. The reaction mixture was cooled to room temperature and centrifuged (2000 rpm, 30 min). The precipitate was transferred into a distillation apparatus and 20 mL toluene was added and distilled off three times. The remaining solid was heated and further dried under vacuum to give a purple solid. Impurities of this solid were washed out with 4 mL CHCl₃, Yield: 67%. – ¹H NMR (CDCl₃): δ=9.11 (s, 8H, β-H), 8.26 (dd, ³J=7.3 Hz, ⁴J=1.4 Hz, 8H, o-H), 7.80 (m, 12H, m-H, p-H). – UV/Vis (C₆H₆): λ_max (rel Int.)=330 nm (4.8), 411 nm (8.7), 431 nm (98), 519 nm (1), 562 nm (3.5), 602 nm (2.0). – HRMS ((+)-MALDI, anthralin): m/z=721.07 (calcd. 721.12 for C₄₄H₂₈N₄ClGe, [M–Cl]⁺), m/z=756.16 (calcd. 756.09 for C₄₄H₂₈N₄Cl₂Ge, [M]⁺).

Synthesis of 3 was performed according to [35].

2.5 Synthesis of bis(phenylacetylido)(meso-tetraphenylporphyrinato) E(IV); E=Si (4), Ge (5), Sn (6)

Synthesis of compound 4 was performed according to [5] in 76% yield. M.p. ≥ 272°C (decomposition). – ¹H NMR (CDCl₃): δ=8.95 (s, 8H, β-H), 8.15 (m, 8H, o-H), 7.74 (m, 12H, m-H, p-H), 6.50 (tr, ³J=7.5 Hz, 2H, p-H′), 6.38 (tr, ³J=7.5 Hz, 4H, m-H′), 5.26 (d, ³J=7.1 Hz, 4H, o-H′). – ²⁹Si NMR (CDCl₃): δ=–240.64. – ¹³C NMR (CDCl₃): δ=142.57, 140.30, 134.13, 132.18, 129.79, 127.86, 127.10, 126.76, 125.24, 123.73, 116.85 (all sp²-hybridized), 102.47, 81.20 (all sp-hybridized). – UV/Vis (C₆H₆): λ_max (lg ε_max)=338 nm (4.44), 432 nm (4.68), 453 nm (5.58), 551 nm (3.45), 594 nm (4.15), 638 nm (4.32). C₆₀H₃₈N₄Si (843.07): calcd. C 85.48, N 6.65, H 4.54; found C 85.97, N 6.30, H 4.78.

Synthesis of 5: 12 mL Et₂O and 0.09 mL C₈H₆ (0.82 mmol) were transferred into a flame-dried flask and cooled to 0°C. n-Buli (0.51 mL) in hexane (0.82 mmol) was added dropwise to give a dark yellow solution. In a second flame-dried flask, 0.157 g of 2 (0.177 mmol) was dissolved in 5 mL toluene and cooled to 10°C. After 30 min half of the yellow solution (0.415 mmol of C₈H₅Li) was added dropwise. The reaction mixture was refluxed for 2 hours and stirred overnight. The resultant reaction mixture was warmed and evaporated to dryness. The remaining solid was washed with 10 mL CHCl₃ and filtered. The solid was washed three times with THF at 0°C and filtered to give a purple/blue crystalline product. Yield: 43%; m.p.>295°C. – ¹H NMR (C₆D₆): δ=9.11 (s, 8H, β-H), 8.06 (dd, ³J=8.1 Hz, ⁴J=1.4 Hz, 8H, o-H), 7.41 (m, 12H, m-H, p-H), 6.10 (tr, ³J=8.9 Hz, 2H, p-H′), 5.95 (tr, ³J=8.0 Hz, 4H, m-H′), 5.20 (d, ³J=7.7 Hz, 4H, o-H′). – UV/Vis (C₆H₆): λ_max (lg ε_max)=337 nm (4.48), 423 nm (4.70), 445 nm (5.60), 545 nm (3.63), 587 nm (4.08), 630 nm (4.30). – HRMS ((+)-MALDI, DCTB): m/z=787.10 (calcd. 787.19 for C₅₂H₃₃N₄Ge, [M–C₈H₅]⁺), m/z=888.13 (calcd. 888.23 for C₆₀H₃₈N₄Ge, [M]⁺).

Synthesis of 6: In a typical reaction, 32 mL Et₂O and 0.14 mL C₈H₆ (1.24 mmol) were transferred into a flame-dried flask and cooled to 0°C. n-Buli (0.75 mL) in hexane (1.24 mmol) was added dropwise to give a dark yellow solution. In a second flame-dried flask, 0.250 g of 3 (0.310 mmol) was dissolved in 130 mL benzene and cooled to 10°C. After 30 min, half of the yellow solution (0.63 mmol of C₈H₅Li) was added dropwise. The reaction mixture was refluxed for 30 min and filtered. The blue/green filtrate was concentrated to half, and 200 mL heptane was added and filtered. The precipitate was recrystallized at –20°C from benzene and CH₂Cl₂. The solid was dried under vacuum to give a purple/blue crystalline product. Yield: 61%; m.p. ≥ 280°C (decomposition). – ¹H NMR (C₆D₆): δ=9.09 (s, ⁴J1_H119_Sn 7.9 Hz, 8H, β-H), 8.05(dd, ³J=8.4 Hz, ⁴J=1.2 Hz, 8H, o-H), 7.42 (m, 12H, m-H, p-H), 6.14 (tr, ³J=7.5 Hz, 2H, p-H′), 6.01 (dd, ³J=7.5 Hz, 7.6 Hz, 4H, m-H′), 5.36 (d, ³J=7.6 Hz, 4H, o-H′). – ¹¹⁹Sn NMR (C₆D₆): δ=–625.70 (s). – ¹³C NMR (C₆D₆/o-C₆H₄Cl₂; 1:5): δ=146.48, 141.97, 135.47, 132.89, 132.40, 130.40, 130.27, 129.84, 126.93, 123.40, 121.80 (all sp²-hybridized), 105.03, 86.66 (all sp-hybridized). – UV/Vis (C₆H₆): λ_max (lg ε_max)=340 nm (4.39), 416 nm (4.76), 438 nm (5.53), 539 nm (3.51), 580 nm (4.08), 624 nm (4.23). – HRMS ((+)-MALDI, DCTB): m/z=732.10 (calcd. 732.13 for C₄₄H₂₈N₄Sn, [M–2 C₈H₅]⁺), m/z=833.14 (calcd. 833.17 for C₅₂H₃₃N₄Sn, [M–C₈H₅]⁺), m/z=934.16 (calcd. 934.21 for C₆₀H₃₈N₄Sn, [M]⁺).

3.1 Synthesis and chemical stability

The synthesis of compounds 5 and 6 was performed as shown in Fig. 1 by the reaction of the corresponding dichlorides 2 and 3 with two equivalents of lithium phenylacetylide in a mixture of Et₂O and benzene or toluene to yield purple/blue crystals. The synthesis is known in the literature for (Ph–C≡C)₂SiTPP 4 and (Ph–C≡C)₂SnTTP [5], [10] (for denotation of porphyrin ring systems see Fig. 2). Synthesis of (Ph–C≡C)₂GeTPP 5 was performed analogously, but due to low solubility of the product in common solvents, a different workup was done. A convenient way to yield pure compound 5 was to wash the precipitate of the reaction mixture in dry THF and filter this suspension. After three repetitions, LiCl and other impurities were extracted and a purple/blue precipitate of 5 remained. Compound 6 was recrystallized from a mixture of benzene and heptane.

Fig. 1:

Synthesis of compounds 4, 5 and 6.

Fig. 2:

Denotation of porphyrin ring systems.

The solid compounds 4, 5 and 6 are stable under inert atmosphere and may even be exposed to air for a short period of time. Solutions of compounds 5 and 6 are stable in anhydrous solvents under inert atmosphere, but hydrolyze rapidly in the presence of water to the corresponding dihydroxides and phenylacetylene (Fig. 3a). This can be evidenced by very high field shifted signals of the OH⁻ groups in the ¹H NMR at around –7.0 ppm [35], [36]. Furthermore, compounds 5 and 6 are unstable in chloroform and react predominantly via replacement of the Ph–C≡C moiety by a chloro substituent (Fig. 3b). A related chlorination is known in the literature for Ph₂SnTPP [9].

Fig. 3:

Reactivity of compounds 5 and 6.

3.2 Crystal and molecular structures

All compounds were crystallized as purple-blue plates by slow evaporation of a benzene (or benzene–CH₂Cl₂) solution of the corresponding porphyrin. The molecular structures of 5 and 6 are shown in Figs. 4 and 5. Despite the fact that compound 4 is known in the literature [5], it was synthesized again and crystallized under similar conditions like the derivatives 5 and 6. The single crystal XRD structure of 4 was re-evaluated at low temperature (100 K) to give a better comparability. Selected bond lengths and angles of 4–6 are listed in Table 2. Only slightly different bond lengths and bond angles were found for compound 4 at 100 K compared to the one reported at room temperature. For example, the differences in C≡C and Si–C bond lengths of 4 at room temperature compared to 100 K are 1.27 vs. 1.21 Å and 1.82 vs 1.87 Å and likely reflect temperature effects in the solid-state molecular structures. The coordination sphere of the metal center in 4–6 is a slightly distorted octahedron with all four nitrogen atoms and E(IV) sitting in an equatorial plane. At 100 K the trans-coordinated axial phenylacetylide ligands do not display linear geometry as expected, but show an angle along E(IV)–C≡C of about 165° (163° at room temp. [5]). Such striking deviations from the valence shell electron pair repulsion (VSEPR) model are known in the literature for alkaline earth metal acetylides [12], [13].

Fig. 4:

Molecular structure of compound 5 and important packing interactions highlighted as a dashed line. All non-carbon atoms are shown as 30% ellipsoids. Hydrogen atoms not involved in intermolecular interactions are removed for clarity. Adjacent molecules involved in interactions are brightened for clarity. (A) C–H/π-type interaction, (B) edge-to-face-type interaction.

Fig. 5:

Molecular structure of compound 6 and important packing interactions highlighted as a dashed line. All non-carbon atoms are shown as 30% ellipsoids. Hydrogen atoms not involved in intermolecular interactions are removed for clarity. Adjacent molecules involved in interactions are brightened for clarity. (A) and (B) C–H/π-type interaction, (C) weak, slightly parallel displaced π–π-type interactions.

Efforts to assign these deviations to influences of the metal center upon going down group 14 (as is also discussed in the literature for alkaline earth metals) seem to fail, as there is almost no significant difference between compounds 5 and 6. Further attempts to attribute these effects to π*_C≡C back bonding also seem to fail, because of the obtained short C≡C bond lengths (all 1.21 Å at 100 K) and typical wavenumber of ϑ(C≡C) stretching vibrations (4: 2145 cm⁻¹; 5: 2142 cm⁻¹; 6: 2122 cm⁻¹). Donation into the empty π*_C≡C orbital may play a minor role only for compound 4 at room temperature due to a relatively long C≡C bond length of 1.27 Å but a relatively moderate wavenumber of the ϑ(C≡C) stretching vibration of 2.145 cm⁻¹.

A more probable reason for these geometrical deviations can be elucidated by a closer look at the solid-state packing interactions in the crystal structures of compounds 4 to 6. Several intermolecular interactions between the phenylacetylide ligand and a surrounding molecule have been found and are within the range for expected values [6]. Among the strong intermolecular interactions of compounds 4 and 5 are edge-to-face and C–H/π interactions (Fig. 4). The edge-to-face interactions refer to the ortho-hydrogen atom of the Ph–C≡C ligand with the electron density of a phenyl group in the meso-position of an adjacent molecule (4: 2.8 Å, 5: 2.8 Å). The C–H/π interaction refers to the para-hydrogen atom of the Ph–C≡C ligand with the electron density of an adjacent porphyrin ring at the β-position (4: 2.8 Å, 5: 2.7 Å). Intermolecular interactions of compound 6 include C–H/π and π–π interactions, whereas the two most striking interactions are of CH/π type (Fig. 5). The first refers to a para-hydrogen atom of a phenyl group in the meso-position of an adjacent molecule with the electron density of the C≡C bond of the Ph–C≡C ligand (2.8 Å) and the second CH/π-type interaction refers to a hydrogen atom in the β-position of an adjacent porphyrin ring system with the electron density of the phenyl ring of Ph–C≡C (2.8 Å). Similar interactions are found for the series (Ph–C≡C)₄E(IV) [37], [38], [39]. However, in (Ph–C≡C)₄E(IV), the C≡C bond lengths are comparable and the E(IV)–C≡C angles are much wider (range: 174–177°) than that of (Ph–C≡C)₂E(IV)TPP (range: 163–167°).

3.3 DFT calculations

In order to understand the difference of the angles E(IV)–C≡C between (Ph-C≡C)₂E(IV)TPP and (Ph–C≡C)₄E(IV), DFT calculations were carried out. A simplified molecular system was used by replacing the four phenyl groups on the porphyrin moiety by four hydrogen atoms. The computations revealed that in vacuo the coordination sphere of the metal center E(IV) is an octahedron with an angle E(IV)–C≡C of 180°. Corresponding calculated bond lengths and their respective bond orders are listed in Table 3. These show that C≡C and especially E(IV)–C bonds are weakened and in good agreement with the observed values for 4, 5 and 6 (Table 2).

Furthermore, IR vibration analysis resulted in values for the bending vibration along E(IV)–C≡C between 80 and 93 cm⁻¹. Finally, although the angle E(IV)–C≡C is 180° for all calculated porphyrins, a weakened E(IV)–C bond and a low frequency of E(IV)–C≡C bending vibration seem to be a good precondition for easy bending along E(IV)–C≡C. And indeed, further DFT calculations showed that whenever one molecule of benzene is added to the system, the resulting weak interactions already lead to an E(IV)–C≡C angle of 174° (Fig. 6).

Fig. 6:

Calculated influence of benzene on the geometry of E(IV)–C≡C in vacuo.

In order to estimate the required energy for bending, a rigid PES scan along the Sn(IV)–C≡C angle (180 to 165°) was performed. They revealed that energies up to 10 kJ mol⁻¹ are sufficient to deflect the respective Sn–C≡C angle. This low value is in agreement with the above assumption that only weak interactions are necessary for bending along E(IV)–C≡C. We therefore assume that the slightly extended σ bond lengths of E(IV)–C for compounds 4, 5 and 6 (Table 2), as compared to reported values of (Ph–C≡C)₄E(IV) [37], [38], [39], represent weakened E(IV)–C bonds. These weakened E(IV)–C bonds enable the angles E(IV)–C≡C to be bent to about 165° by the above-mentioned packing interactions. It also explains why the angles along E(IV)–C≡C of (Ph–C≡C)₄E(IV) are around 175°, as there appears to be less weakening of E(IV)–C bonds.

3.4 UV/Vis spectra

The steady-state absorption spectra of compounds 4, 5 and 6 were recorded in benzene under inert atmosphere and are shown in Fig. 7. Compound 4 is known in the literature; however, molar absorption coefficients have not been reported so far to the best of our knowledge [5]. Maxima of absorption bands and molar absorption coefficients are listed in Table 4.

Fig. 7:

Steady-state absorption spectra of compounds 4, 5 and 6 in benzene.

A comparison of the absorption maxima of compounds 4, 5 and 6 with those of the corresponding dichlorides [1], [36], [40] shows bathochromic shifts for all bands of at least 10 to maximally 30 nm. This is in agreement with the assumption that more electropositive ligands decrease the required absorption energies [1], [41]. This might be caused by the increased donation of electron density into the metal center, which further increases the energies of the transitions involving highest occupied molecular orbital (HOMO) levels of the porphyrin π system. A comparison of calculated orbital energies of 4, 5 and 6 with those of 1, 2 and 3 revealed that the transitions involving orbital energies of HOMO–1, lowest unoccupied molecular orbital (LUMO), LUMO+1 of 4, 5 and 6 are raised by approx. 0.3 (eV) and the energies of the HOMO orbitals of 4, 5 and 6 are raised by approx.0.4 (eV). This results in lower energy gaps and therefore leads to bathochromic shifts for 4, 5 and 6.

A comparison of the absorption bands among compounds 4, 5 and 6 shows that the spectrum of compound 4 is the most bathochromic shifted followed by compounds 5 and 6. This trend was not found in the literature for the series of Cl₂E(IV)TPP and Ph₂E(IV)TPP [1], [36], [42]. However, the spectra of Ph₂E(IV)TPP were measured in different solvents (Si: CH₂Cl₂, Ge: PhCN, Sn: C₆H₆) and are therefore difficult to compare.

3.5 Luminescence spectra

The normalized luminescence spectra of compounds 4, 5 and 6, which were excited into the Soret band, are shown in Fig. 8. They were recorded at 298 K and at 77 K in a mixture of THF/toluene (v/v 6:4), which forms a glass at 77 K. The recorded maximum wavelengths of emission and the relative quantum yields of fluorescence and phosphorescence are listed in Table 5. Lifetimes of all fluorescence and phosphorescence signals were measured with TCSPC. The corresponding decay signals can be seen in Figs. 9 and 10. The obtained decay times are listed in Table 6 and were calculated using a single exponential fit determined by minimizing the reduced χ². All compounds displayed typical emission spectra of porphyrins, representing a mirror image of the Q bands of the absorption spectra and having relatively small Stokes shifts of 100–150 cm⁻¹. Phosphorescence was detected at around 830 nm only for compounds 5 and 6. Although phosphorescence was reported for a Si(IV) porphyrin [43], we were not able to detect any phosphorescence for 4. Table 5 shows decreasing fluorescence quantum yields and increasing phosphorescence quantum yields going for compounds 4–6. Lifetimes for fluorescence and phosphorescence are both shortening from 4 to 6 (Table 6).

Fig. 8:

Luminescence spectra of compounds 4, 5 and 6 in THF–toluene (v/v: 6:4) under excitation at λ=464 nm, 454 nm and 436 nm, respectively. (a) At 298 K, (b) at 77 K, inlay: phosphorescence of 5 and 6.

Fig. 9:

¹S→⁰S fluorescence decay of compounds 4, 5 and 6 in THF–toluene (6:4); excitation: 435 nm, emission: 4: 695 nm 5: 690 nm 6: 685 nm; IR=instrumental response, T=77 K.

Fig. 10:

¹T→⁰S phosphorescence decay of compounds 5 and 6 with the corresponding fit according to a mono-exponential decay model and the corresponding residuals; excitation: 392 nm, emission: 830 nm, T=77 K.

In order to compare this observation with the data from the literature, compounds 4, 5 and 6 are compared to the available data of the series Cl₂E(IV)octaethylporphyrin (OEP), which were recorded in 2-methyltetrahydrofuran (MTHF) [43]. As mentioned therein, many differences in the emission spectra appeared (shifted bands, different relative intensities, different phosphorescence yields) when different structures of the macrocycles were used. Therefore, only trends are compared. The analysis of the fluorescence quantum yields of Cl₂E(IV)OEP shows highest values for Si derivatives followed by Ge and Sn derivatives. Inverted are observed effects for phosphorescence with highest yields for the Sn derivatives followed by Ge and Si. The phosphorescence lifetimes are longest for the Si derivative followed by Ge and Sn.

Overall, similar trends are observed for our series of compounds 4–6. The observations of both series can be explained through the heavy atom effect of the metal center. The heavier the metal, the more pronounced is spin–orbit coupling, which depopulates singlet excited states and populates triplet excited states. As a consequence, fluorescence quantum yields become lower and phosphorescence quantum yields become higher if going from compound 4–6. Lifetimes of fluorescence and phosphorescence are both shortening due to the heavy atom effect.

In order to obtain information about the influence of the ligand Ph–C≡C on the luminescence properties, comparison with the reported photophysical properties of 1, 2 and 3 was undertaken. However, since most luminescence data in the literature [40], [43], [44], [45] are reported for different solvents (pyridine, EtOH, MTHF), comparison was difficult to perform.

The quantum yield and lifetime of fluorescence for compound 4 in THF–toluene (0.085, 8.64 ns) are significantly higher than the reported values of compound 1 in pyridine (0.027, 1.8 ns) [44].

The quantum yield of fluorescence for compound 5 in THF–toluene (0.064) is slightly lower than the quantum yield of compound 2 in EtOH (0.081) [40]. Although this difference is very small, it can be explained by the assumption that quantum yields for π–π* transitions are slightly higher in protic solvents than in aprotic ones [45], [46]. A comparison of the ¹S→⁰S fluorescence lifetimes also shows small differences (3.98 ns for 5, 4.7 ns for 2).

Maiti et al. [45] reported photophysical data of compound 3 in different solvents at room temperature. They determined fluorescence quantum yields for 3 of 0.018 (THF), 0.024 (toluene) and 0.032 (EtOH). A comparison of these data to the herein determined value of 0.018 (THF/toluene) for compound 6 shows similarity, also for its ¹S→⁰S fluorescence lifetimes (range: 1–2 ns). A comparison of the phosphorescence quantum yield of 0.002 for 6 with the reported one of 0.024 (MTHF) for 3 indicates an at least 10 times higher value for 3 [43].