Advertisement

Chemistry Africa

, Volume 1, Issue 1–2, pp 67–77 | Cite as

Completely Unexpected Coordination Selectivity of Copper Iodide for Thioether Over Ethynyl

  • Antoine Bonnot
  • Frank Juvenal
  • Adrien Schlachter
  • Daniel Fortin
  • Pierre D. Harvey
Original Article
  • 68 Downloads

Abstract

The reactivity of the tetradentate ligand bis(p-thiomethylphenylacetylene) (MeSC6H4C≡C–C≡CC6H4SMe; L2) towards the CuI salt is compared to that for the known organometallic analogue trans-bis(p-thiomethylethynylbenzene)bis(trimethyl-phosphine)platinum(II) (trans-Pt(PMe3)2(C≡CC6H4SMe)2; L1). While L1 with CuI form a highly luminescent porous 2D coordination polymer (CP) of general formula ([Cu4I4]L1 · EtCN)n (CP1; Juvenal et al. in Inorg Chem 55:11096–11109, 2016) exhibiting both Cu(η2–C≡C) and Cu–S bonds, L2 reacts with CuI to produce a luminescent non-porous 2D CP exhibiting the general formula ([Cu4I4]{L2}3)n, CP2, which does not use the highly expected Cu(η2–C≡C) linkage, relying strictly upon Cu–S coordination. An examination of the X-ray structures for both L2 and CP2 indicates that CP2 network is built upon an expansion of the L2 lattice (plane sliding and slight L2L2 distance separation) resembling to a sort of template effect. CP2 has been characterized by TGA, UV–Vis, emission spectroscopy, and photophysics, which are accompanied by DFT and TDDFT computations.

Graphical abstract

Keywords

Coordination polymer Luminescence DFT compuations Template Copper Thioether 

1 Introduction

There is a wealth of examples dated from the 1990’s where copper halide salts selectively coordinate the alkynyl function (η2–C≡C) over the thioether one in bidentate (R–C ≡C···S–R′) ligands to form organometallic species with a dandling uncoordinated thioether [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]. However, this trend does not exclude the several examples where both coordination centers are used to generate polymetallic (cyclic and acyclic) species and polymeric solids [14, 15, 16, 17, 18]. The bis(phenylacetylene) motif (PhC≡C─C≡CPh) is also known to efficiently bind copper halides (CuX; X = Cl) in a (η2–C≡C) fashion [19, 20]. Upon the incorporation two methyl thioethers onto this motif, an emissive multidentate ligand is obtained (L2; Fig. 1) [21], which is also known to be prone to generate luminescent (rhodium) organometallic complexes [22], but the SMe groups do not coordinate the metal at all.
Fig. 1

Top: structures of L1 and L2. Bottom: stick representation of the 2D CP1 [23, 24]. The staircase Cu4I4 cluster is indicated by the green oval. The solvent molecules (MeCN or EtCN) are not represented but are present in the voids between the L1’s and clusters

Structurally related to L2, a recent investigation on an emissive and rigid organometallic ligand L1 (trans-Pt(PMe3)2(C≡CC6H4SMe)2; Fig. 1) was conducted in order to construct heterobimetallic coordination polymers (CPs) exhibiting both porous frameworks and exhibiting a strong luminescence [23, 24]. Figure 1 exhibits such an example ([Cu4I4]L1 · solvent)n (CP1), which was reported for potential sensor applications for organic vapour. In addition to this case, the presence of a Pt atom is also attractive for possible catalytic applications. The outcome of this study is that the formation of CPs when reacting L1 with CuX salts (X = Cl, Br, I) in various solvents (MeCN, EtCN and PhCN) is that despite the apparent steric constraint around the Pt-center, Cu(η2–C≡C) binding is always observed. When Cu–S coordinations also occur, porous or non-porous 2D frameworks are systematically generated [23, 24]. This preference for Cu(η2–C≡C) linkage over Cu–S is again fully consistent with earlier literature observation mentioned above [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18]. However, very rare examples of 2D and 3D CPs produced with alkynyl-containing dithioethers (here RC6H4SCH2C≡CCH2C6H4R; R = H, CH3) and CuX salts (X = I in this case) through Cu–S only, also exist (2 examples) [25, 26].

In this bref investigation, we now report a complete opposite selectivity when reacting L2 with CuI where no Cu(η2–C≡C) bond is observed. In this case, a 2D luminescent CP of general formula ([Cu4I4]{L2}3)n (CP2) is isolated. The explanation for this unexpected observation stems from the shape of L2, which in the solid state stacks side-by-side, in a similar manner to that observed in CP2. The emissive excited state has been characterized with the help of density functional theory (DFT) and time-dependent DFT (TDDFT) computations.

2 Experimental Section

2.1 Materials

CuI, p-thiomethylphenylacetylene, and 2-methyltetrahydrofuran (2MeTHF) were purchased from Aldrich. L2 was synthesized from a modified procedure (placed below) from the literature [27].

2.2 Synthesis of Bis(1,4-bis(4-(methylthio)phenyl)buta-1,3-diyne (L2)

To a 50 mL double neck round bottom flask, 5 mL of MeCN was introduced and the solution was stirred at room temperature under N2 atmosphere. Simultaneously, 170 mg of CuCl2 (1.0 mmol, 1 eq.) and 150 μL of tetramethylethylenediamine, TMEDA (1.0 mmol, 1 eq.) were introduced. The solution is stirred and then deep blue color is observed. After 20 min, 296 mg of p-thiomethylphenylacetylene (2.0 mmol, 2 eq.) is added with 150 μL of TMEDA once again. The solution was heated at 60 °C for 4 h. The solution was slowly cooled down to room temperature and filtered the precipitate. The precipitated was washed with 20 mL of water and 5 mL of ether. A yellow powder was obtained (282 mg, yield = 95%). IR: 2982 (νC–H), 2244 cm−1C≡C). 1H NMR (300 MHz, CDCl3): δ 7.42 (d, J = 8.6 Hz, 4H, Ar), 7.18 (d, J = 8.6 Hz, 4H, Ar), 2.49 (s, 6H, S–CH3) ppm. 13C NMR (300 MHz, CDCl3): δ 141.00 (Ar), 132.83 (Ar), 125.82 (Ar), 118.01 (Ar), 81.77 (C≡C), 74.26 (C≡C), 15.34 ppm. λabs(2MeTHF) = 361 (ε = 730), 336 (ε = 885), 317 nm (ε = 755 L.mol−1.cm−1).

2.3 Preparation of ([Cu4I4]{L2}3)n (CP2)

To 10 mL vial, 50 mg of L2 (0.17 mmol, 1 eq.) and 65 mg of CuI (0.34 mmol, 2 eq.) are introduced in 5 mL of degassed MeCN. The vial is sealed and covered with parafilm. The solution was stirred for 30 min at room temperature, and heated to 60 °C. The solution was slowly cooled down to the room temperature. Small crystals are formed after 12 h. The solution was slowly evaporated at room temperature leading to light yellow crystals. The crystals are washed with 10 mL of methanol, and dried. (82 mg, yield = 71%). Chem. Anal. Theory; C: 39.38, H: 2.55, S: 11.67%, found; C: 40.42, H: 2.61, S: 11.47%. IR: 2982, v(C–H); 2244, v(C≡C).

2.4 Instruments

Solid state UV–Vis spectra were measure using a Varian Cary 50 spectrophotometer at 298 and 77 K using raised-angle transmittance apparatus and a homemade 77 K sample-holder. Steady state emission and excitation spectra as well as chromaticity were measured on Edinburgh Instruments FLS980 Phosphorimeter equipped with single monochromators or a PhosphorimeterQuantaMaster 400 from Photon Technology International (PTI). The lifetime measurements were made with an Edinburgh Instruments FLS980 Phosphorimeter equipped with “flash” pulsed Lamp. The Frequency of the pulse can be adjusted from 1 to 100 Hz. All lifetime values were obtained from deconvolution and distribution lifetime analysis. The Quantum Yield measurements were performed with the Horiba Fluorolog III. This instrument is equipped of an integration sphere which allows the direct measurements of emission quantum yields. All samples were crushed to powder prior to use, and the homogeneity of the sample was verified by comparing the X-ray powder diffraction patterns with that calculated ones from the single crystal X-ray data. These spectra were corrected for instrument response. The TGA traces were acquired on a Perkin Elmer TGA 7 apparatus in the temperature range between 20 and 900 °C at 10 °C/min under nitrogen atmosphere. The figures have been treated by Origin software.

2.5 Computations

The density functional theory (DFT) and time dependent density functional theory (TD-DFT) calculations were performed with Gaussian 09 [28] at the Université de Sherbrooke with the Mammouth supercomputer supported by Le Réseau Québécois De Calculs Hautes Performances. All. cif files from X-ray crystal structures have been used like optimized structure for calculations. The DFT (singlet and triplet energy states) as well as TD-DFT calculations [29, 30, 31, 32, 33, 34, 35, 36, 37, 38] were carried out using the B3LYP functional. VDZ (valence double ζ) with SBKJC effective core potentials were used for all Cu and I [39, 40, 41, 42, 43, 44]. The calculated absorption spectra were obtained from GaussSum 3.0 [45].

2.6 X-ray Crystallography

Clear light yellow Needle-like specimen of L2 were measured on a Bruker Apex DUO system equipped with a Cu Kα ImuS micro-focus source with MX optics (λ = 1.54186 Å). A total of 1727 frames were collected for L2. The total exposure time was 4.80 h. The frames were integrated with the Bruker SAINT [46] software package using a wide-frame algorithm. The integration of the data for L2 using a triclinic unit cell yielded a total of 4113 reflections to a maximum θ angle of 70.84° (0.82 Å resolution), of which 2727 were independent (average redundancy 1.508, completeness = 95.7%, Rint = 1.65%, Rsig = 3.57%) and 2526 (92.63%) were greater than 2σ(F2). Data were corrected for absorption effects using the multi-scan method (SADABS) [46]. The structure was solved and refined using the Bruker SHELXTL Software Package, using the space group of P − 1, with Z = 2 for L2. Clear light yellow prism-like specimen of CP2 were measured on a APEX DUO system equipped with a TRIUMPH curved-crystal monochromator and a Mo Kα fine-focus tube (λ = 0.71073 Å). A total of 797 frames were collected. The frames were integrated with the Bruker SAINT [46] software package using a wide-frame algorithm. The integration of the data using a triclinic unit cell for CP2 yielded a total of 7436 reflections to a maximum θ angle of 26.41° (0.80 Å resolution), of which 5535 were independent (average redundancy 1.343, completeness = 99.8%, Rint = 1.74%, Rsig = 4.00%) and 4652 (84.05%) were greater than 2σ(F2). Data were corrected for absorption effects using the multi-scan method (SADABS) [46]. The structure of the CP was solved and refined using the Bruker SHELXTL Software Package, using the space group of P -1, with Z = 2 for CP2. The crystallographic data for L2 and CP2 are placed in the Supporting Information.

2.7 X-ray Powder Diffraction Pattern Measurements

Prior to the photophysical investigation, L2 and CP2 were checked for mixed phase. They mixed with a small amount of paratone oil and cut to approximately 0.3 × 0.3 × 0.3 mm3. The samples were placed on a sample holder mounted at 173.2 K on a Bruker APEX DUO X-ray diffractometer. 6 correlated runs per sample with Phi Scan of 360 degrees and exposure times of 270 s were collected with the Cu micro-focus anode (1.54184 Å) and the CCD APEX II detector at 150 mm distance. These runs, from −12 to −72°2θ and 6 to 36°ω, were then treated and integrated with the XRW2 Eval Bruker software to produce WAXD diffraction patterns from 2.5 to 82°2θ. The patterns were treated with Diffrac.Eva version 2.0 from Bruker.

3 Results and Discussion

3.1 Synthesis and X-ray Structures

The reaction scheme for the synthesis of CP2 is shown in Scheme 1. Crystals suitable for X-ray analyses were obtained from a slow evaporation of a MeCN solution (Figs. 2, 3).
Scheme 1

Synthesis of L1, and CP2; (i) CuCl2, TMEDA, MeCN, N2, 4 h; (ii) CuI, MeCN, 12 h

Fig. 2

ORTEP representation of L2 in the crystal at 173 K stressing on the S···S separations for adjacent ligands. The thermal ellipsoids are set at 50% probability. Yellow = S, blue = C, white = H (color figure online)

Fig. 3

Structure of a fragment of CP2 at 173 K. Top: side view of the 2D layer. Bottom: top view the 2D framework stressing on the S···S separations for adjacent ligands (note that in this image the Cu–S bonds have been removed for a better visualization of the cluster. Yellow = S, brown = Cu, purple = I, grey = C. Selected bond lengths (in Å): S3–Cu2: 2.3982(8); Cu2–S2, 2.3647(9); Cu2–Cu1, 2.7721(6); Cu1–I1, 2.5769(5); I1-Cu2, 2.5852(5); Cu2–I2, 2.6085(5); I2–Cu1, 2.6943(5); Cu1-Cu1, 2.7351(6); Cu1–I2, 2.6479 (5); Cu1–S1, 2.3338(9) (color figure online)

During the course of this study, the X-ray structure of L2 was revisited, but at a different temperature (Fig. 2), because of the template effect discovered (below). L2 stacks side-by-side with S···S separations of 4.130 and 5.073 Å, and a dihedral angle for the C6H4··· C6H4 planes, γ, of 51.90° at 173 K. These values are essentially identical to that of the literature (d(S···S) = 4.109 and 5.057 Å; γ = 51.78°) at 120 K [22]. This weak compressibility indicates that the stacking is rather compact.

Concurrently, CP2 forms a centro-symmetric 2D framework of a general formula ([Cu4I4]{L2}3)n (Fig. 3). Numerous features are striking. First, despite unambiguous literature evidence for the favoured selectivity for Cu(η2–C≡C) coordination bond first, this structure exhibits only Cu–S bonding. Second, the secondary building unit (SBU) is a tetranuclear Cu4I4 cluster adopting a staircase structure, which is identical to that observed for the related CP1 built upon L1 (see structure of L1 in Fig. 1) with a general formula ([Cu4I4]L1 · solvent)n (solvent = MeCN, EtCN) [23, 24]. This trait is somewhat unusual since the structure of the SBU is highly variable going from the rhomboid Cu2I2, to the closed and open cubane Cu4I4, to the staircase Cu4I4, to the hexagonal Cu6I6, to the closed side-fused dicubane Cu8I8, to the polymeric staircase (Cu2I2)n geometries, just to state a few, and that a slight structural modification of the bridging dithioether ligand can lead to a complete change in SBU and even CP dimensionality [47, 48, 49]. Third, the L2 ligand crystalizes side-by-side with an astonishing resemblance with the crystal structure of L2 itself. Indeed, the d(S···S) values are 3.849, 4.704 and 5.630 Å (Fig. 4), for which the average value (4.73 Å), while the d(S···S) separation in a repetitive unit is 13.534 Å (13.534/3 = 4.51 Å).
Fig. 4

Comparison of the packing of the ligand L2 within its own crystal and CP2. Note that the H-atom and the Cu4I4 SBU are not shown for clarity

These dimensions compare favourably to the average d(S···S) value observed in the L2 crystal (4.60 Å). Similarly, two of the three coordinated L2 ligands exhibit γ = 63.25° (somewhat larger than that for L2; 51.90°), whereas the third one is 0°. The comparison of the L2 arrangements in the L2 crystal and within CP2 indicates that a seemingly “L2 layer” sliding and a small L2···L2 separation enlargement take place in order to adapt for the incorporation of the staircase Cu4I4 SBU. Similarly, the approximate size of the Cu4I4 cluster (largest Cu···Cu separation = 6.782 Å) and cluster···cluster distance (Cu···Cu gap = 6.757 Å) account for 13.54 Å (i.e. average space available for 3 L2 ligands = 4.51 Å).

In addition, this easy adaptability of the L2 lattice to incorporate the Cu4I4 cluster is also translated by a small change in porosity (the percentage of voids in the lattices, according to Mercury, are 25.9 and 28.3% for L2 and CP2, respectively). The similarity of the atom–atom separations cited above and the void spaces in the lattice explain why no Cu(η2–C≡C) coordination bond is formed in this case, despite the wealth of literature data indicating a clear preference for the Cu(η2–C≡C) over the Cu–S linkage for these hetero-polydentate ligands. This outcome results from a serendipity template effect of the L2 crystal and the rather large “brochette” of possible SBU’s that can offer the (Cu2I2) synthon, and seems to the best of our knowledge, totally unprecedented. Some physical and optical properties of this new CP are now described.

3.2 Thermal Properties

The thermal stability of CP2 was tested by TGA (Fig. 5). The first weight loss is observed in the vicinity of 190 °C. Between 190 and 220 °C, a loss of ~ 1% is observed corresponding to a methyl group (theory = 0.91%). Then between 220 and 305 °C, another ~ 2% mass is lost which can be attributed to the loss of two other methyl groups (theory = 1.83%). Between 305 and 470 °C, a mass loss of ~ 19% is depicted and correspond approximately to the loss of one ligand L2 (theory = 17.9%). Finally after a quasi-gradual weight loss, a plateau starts being defined with a residual mass of 48% at 900 °C (i.e. the limit of the instrument). This residual is most likely CuI (theory = 46.3%), which is known to exhibit a melting and boiling point respectively at 606 and 1290 °C.
Fig. 5

TGA trace (black) and its 1st derivative (grey) for CP2. Scan rate = 10 °C/min

3.3 Photophysical Properties of L2

For comparison purposes with CP2, a brief description of the optical (Fig. 6) and photophysical properties (Table 1) is provided. These features are unambiguously associated with emissive ππ* excited states but are verified below.
Fig. 6

Excitation (blue), emission (red) and absorbance (black) spectra of L2 in 2MeTHF at 298 and 77 K (color figure online)

Table 1

0-0 peak positions and emission lifetimes

L2

λabs (0–0)

λem (0–0)

Lifetimes

298 K

360 nm

364 nm

τF = 0.32 ns

77 K

368 nm

369 nm

τF = 0.62 ns

500 nm

τP = 680 μs

The emission lifetimes occurring in the nanosecond (ns) time scale along with the small Stoke shift (1–4 nm) support the assignment for fluorescence, τF, whereas the μs time scale the 132 nm spectral gap between the absorption at 368 nm and the 0-0 peak at 500 nm indicate the presence of phosphorescence, τP. The shorter values for τF (normally ranging from 1 to 10 ns) and τP (usually found in the ms to s time scale) are consistent with a small heavy atom effect arising from the S-atoms. The nature of these singlet and triplet excited states has been confirmed by DFT and TDDFT computations. After the geometry optimisation of L2, the shapes and atomic contributions of the frontier MOs have been calculated in both the ground (S0: HOMO and LUMO = highest occupied and lowest unoccupied molecular orbital, respectively) and lowest energy triplet (T1: LSOMO and HSOMO = lowest semi-occupied and highest semi-occupied molecular orbitals, respectively) excited states (Fig. 7, see SI for more MOs and the atomic contributions). These MO are best described at filled and empty π orbitals as anticipated. TDDFT compute the position of the 0–0 peak at 373 nm, agreeing with the values of 368 and 369 nm observed at 77 K (Table 1), a temperature where the contributions of “hot bands” is minimal. The calculated oscillator strength (f) is 1.50 meaning that the transition is strongly allowed by symmetry, and is composed of almost entirely on HOMO → LUMO (95%). The first 100 computed electronic transitions are placed in the SI for convenience.
Fig. 7

Representations of the frontier MOs and semi-occupied MOs for L2 in its ground state (S0) and triplet state (T1), respectively based on their optimized geometries

3.4 Photophysical Properties of CP2

CP2 is found emissive at 298 and 77 K (Fig. 8) with an emission quantum yield of 17% at 298 K (using an integration sphere). The μs time scale for the lifetimes (τe; Table 2) and the large spectral gap between the absorption and emission (~ 180 nm) indicate that this luminescence arises from a triplet excited states. The emission decays appears bi-exponential but do not constitute a surprise for this type of ligand. Indeed in recent detailed investigation [50], it was recently demonstrated that the emission decays in ligands exhibiting the general structure analogous to L1 were all bi- or poly-exponential depending on the dihedral angles formed by the C6H4C≡C planes. Indeed in CP2, L2 adopts two conformations (two L2 with γ = 63.25° and one with 0°). The nature of the excited states was also addressed using DFT and TDDFT computations using the X-ray data. For simplicity and the less calculation time, the representative fragment used was the cyclic unit (Me2S)4[Cu4I4](μ-L2)2[Cu4I4](Me2S)4 where the Me2S end groups were optimized and the selected L2 were those with γ = 63.25° (Fig. 9).
Fig. 8

Solid state absorption (black; reflectance), excitation (blue), and emission (red) spectra of CP2 at 298 and 77 K. The chromaticity diagrams and data are placed in the SI. Note that the abrupt change in absorbance at 350 nm is due to the change in lamp) (color figure online)

Table 2

τe data (± 5%) for CP2 at 298 and 77 K (Bi = pre-exponential factor)

λem(nm) 298 K

Bi

τe (μs) 298 K

χ 2

Φe (%) 298 K

λem(nm) 77 K

Bi

τe (μs) 77 K

χ 2

481

0.0034

0.43

1.023

17

499

0.0066

1.88

1.084

0.0038

1.52

0.0004

12.3

Fig. 9

Representations of the frontier MOs and semi-occupied MOs for CP2 in its ground state (S0) and triplet state (T1), respectively, using a fragment ((Me2S)4[Cu4I4](μ-L2)2[Cu4I4](Me2S)4) based on its X-ray data. See SI for more MOs

In the ground state (S0) the filled MOs are located within the Cu4I4 core, whereas the empty ones exhibit atomic contributions of the π-systems in both bridging L2 ligands (Table 3). Any combination of the low-energy electronic transitions (i.e. HOMO → LUMO, HOMO-1 → LUMO, HOMO → LUMO + 1) creates metal-halide-to-ligand charge transfer excited states M/XLCT, an outcome considered common in CuX-dithoether assemblies (X = halide) [47, 48, 49]. Similarly, the LSOMO and HSOMO exhibit atomic contributions respectively localized mainly in one of the Cu4I4 clusters and one of the L2 ligands of the (Me2S)4[Cu4I4](m-L2)2[Cu4I4](Me2S)4 also indicating that the triplet excited state T1 is an 3M/XLCT*. The TDDFT computations (Table 4) place the lowest energy singlet–singlet transition at 396 nm with a weak oscillator strength value (f). This value is consistent with the weak feature at ~ 415 nm depicted in the excitation spectrum at 298 K, but also the ill-defined shoulder in this region in the absorption spectra (Fig. 8).
Table 3

Atomic contributions (%) of the various fragments to the frontier MOs in CP2.a

Ground state (S0)

Fragments

H-1

HOMO

LUMO

L + 1

L2

12.7

12.6

86.9

90.6

Cu atoms

40.6

41.2

12.6

7.0

I atoms

46.7

46.2

0.5

2.4

Triplet excited state (T1)

Fragments

LS-1

LSOMO

HSOMO

HS + 1

L2

13.9

12.8

92.1

87.9

Cu atoms

39.7

40.4

6.7

10.4

I atoms

46.4

46.8

1.3

1.6

aSee SI for more MOs

Table 4

Calculated positions, oscillator strengths (f) and major contributions of the first 10 singlet–singlet electronic transitions for CP2 (see SI for the first 100 transitions)

λ (nm)

F

Major contributions (%; H = HOMO, L = LUMO)

396.3

0.006

H → L (89)

391.9

0.057

H → L+1 (89)

389.4

0.015

H-1 → L (92)

384.8

0.043

H-1 → L+1 (92)

360.8

0.034

H-3 → L (55)

356.1

0.001

H-3 → L (13), H-3 → L+1 (61)

354.6

0.469

H-5 → L (29), H-4 → L (17), H-3 → L (10), H-3 → L+1 (10), H-2 → L (24)

353.3

0.104

H-5 → L (10), H-2 → L (53), H-2 → L+1 (12)

352.7

0.320

H-5 → L+1 (39), H-4 → L+1 (14), H-2 → L+1 (16)

349.6

0.024

H-5 → L+1 (13), H-2 → L+1 (53)

The weak spectral intensity and the small calculated f value are consistent with the quasi inexistent MO overlap between the Cu4I4 core and the π-system of L2 within the CP meaning that the low-energy electronic transitions are forbidden by symmetry (here lack of overlap). Some ill-defined weak vibronic components in the vicinity of 400-450 nm are observed but the lifetime of these features (presumably fluorescence). However, the τF value could not be extracted because they fell below the detection limit of the instrument (IRF = 90 ps). This outcome is expected as CP2 includes the Cu4I4 cluster inducing a stronger heavy atom effect on the excited states. This effect is well demonstrating by comparing the triplet lifetimes (τP) of L2 and τe of CP2 at 77 K decreasing from 680 μs down to 1.9 and 13.3 μs, respectively. By applying a ratio 13.3 (CP2)/680 (L2) of 0.0193 to τF for L2 at 77 K (i.e. 620 ps), an anticipated τF value of ~ 12 ps is indeed predicted for CP2 (outside the detection limit). By plotting the calculated f values as a function of the positions of the electronic transitions (i.e. 0–0 peaks), one obtains a simulated spectrum excluding the vibronic progressions (Fig. 10), which is reminiscent of that obtained experimentally (Fig. 8).
Fig. 10

Bar graph reporting the calculated oscillator strength, f, and calculated position of the 100st electronic transitions calculated by TDDFT for CP2 (bar graph; f = computed oscillator strength). The black line is generated by assigning an arbitrary FWHM = 1000 cm−1 to the transition. (FWHM = full width at half maximum)

In addition, Fig. 8 also suggests that the singlet and triplet emission strongly overlap ought to the rather very large bandwidth (~ 170 nm at 298 K, and ~ 200 nm at 77 K). The position of the triplet state can be estimated by either calculating the energy difference between S0 and T1 (3.08 eV) or between the LSOMO and HSOMO (2.47 eV, Fig. 9). Based on these calculated data, the position of the triplet emission is 404 or 502 nm, depending which data set one uses. In both cases, these values are larger than that of 396 nm calculated for CP2 (Table 2). The conclusion of these computations and comparisons is that this novel emissive CP2 material exhibits singlet and triplet M/XLCT excited states.

3.5 Supplementary Discussion

The coordination of the staircase Cu4I4 cluster into the crystalline L2 matrix proceeds in a convenient serendipity, and a template manner. The similarity in percentage of voids in the lattices (i.e. 25.9 and 28.3% for L2 and CP2, respectively) is also translated by a similarity in C···C separations between adjacent C≡C–C≡C parallel chains (ranging from 4.0 to 4.8 Å). This structural feature is most likely associated with the non-nil dihedral angles γ. Knowing that L2 attached onto gold nanoparticles can successfully be photo-polymerized upon UV illumination (i.e. λexc = 250 nm) to form a coloured conjugated polymer (λabs ~ 600 nm) of general formula (–C≡C–CR=CR–)n (R = C6H4SMe), when the tilt angle (40-60°) permits a C···C separation smaller than 4 Å [51]. For the L2 crystal and CP2, this distance condition is never met and unsurprisingly both remain white upon UV irradiation and are luminescent (i.e. no polymerization nor emission quenching is observed). It is not true that all diacetynyl compounds exhibit C···C separation longer than 4 Å, and consequently these species are vulnerable to photo-additions (1 + 4 or 1 + 2). The presence of the staircase Cu4I4 cluster can be perceived as a protective device during to coordination assembling. First, the geometry can separate the C-atoms at distances longer than what could lead to C–C bond formation. Second, the external heavy atom effect (just the fact that it is in the vicinity) can decrease the excited state lifetimes and promote intersystem crossing in a way that the 1ππ* manifold are efficiently deactivated. Third, in the situation where the coordination led to the formation of heavily mixed excited states (where the atomic contribution are located significantly on both the ligand and the SBU), then the nature (and its intrinsic reactivity) of its 1ππ* excited state will be deeply altered, not to say precluded. Again, this situation was not observed in L2 (see Fig. 9 and Table 3) but was noted for other CP exhibiting the SC6H4R moiety (R = alkyl group) on the thioether ligands [47, 48, 49].

Moreover, ideally one could comments on the comparison of the Cu–S and Cu(η2–C≡C) bond energies. Unfortunately, literature data on these properties are rather scarce and non-homogeneous. A qualitative comparison is still attempted here. The computed bond energy (DFT, B3LYP) for CH3Cu(η2–HC≡CH) is 72.8 kJ/mol [52], but to the best of our knowledge no similar computations were performed between a cupper(I) and a thioether. Experimental Cu–S values for Cu(II) and various dithiocarbamates are known [53], and a typical value turns around 17.7 kJ/mol [54]. Concurrently, the bond energy between copper surface (110) with thiacyclohexane and thiacyclopentane are respectively 57.9 and 96.5 kJ/mol [55]. Intuitively, one may expect that these scarce data suggest that the Cu(I)-thiosulfur bond energy be placed somewhere between these values (17.7–96.5 kJ/mol), which seems, on average (“57 kJ/mol”), to be just under the calculated Cu(η2–HC≡CH) bond strenght (72.8 kJ/mol) [52]. This comparison is consistent with the very high literature tendency to observe a clear selectivity of Cu(I) toward ethynyl over the thioether center, as indicated in the Introduction. However, the closeness of these values is also consistent that exceptions may occur, as demonstrated in this work.

4 Conclusion

This work reported a drastic change in coordination selectivity of the CuI salt, normally favouring first the η2–(C≡C) binding over the Cu–S. To the best of our knowledge, this observation is extremely rare and is well explained by an unexpected template effect of the ligand crystal structure and the adaptability of the CuI salt to form a large variety of SBUs with different geometries. The resulting CP is luminescent with a quantum yield of 17% at 298 K, and this photophysical activity witnesses the absence of photo-polymerization of the diacethynyl ligand, assigning to the Cu4I4 cluster protecting role of against this common photochemical process.

Notes

Acknowledgements

This work was supported by the Natural Sciences and Engineering Research Council of Canada, the Fonds de recherche du Québec-Nature et technologies, Compute Canada and Calcul Québec, and the Centre Quebecois sur les Matériaux Fonctionnels.

Supplementary material

42250_2018_4_MOESM1_ESM.docx (2.4 mb)
Supplementary material 1 (DOCX 2472 kb)

References

  1. 1.
    Olbrich FA, Kopf JO, Weiss ER, Krebs AD, Müller S (1990) Acta Crystallogr C Cryst Struct Commun 46:1650–1652CrossRefGoogle Scholar
  2. 2.
    Olbrich F, Gröger G, Behrens U (1999) Zeitschrift für Kristallographie-New Crystal Structures 214:195–200CrossRefGoogle Scholar
  3. 3.
    Olbrich F, Schmidt G, Weiss E, Behrens U (1993) J Organomet Chem 456:299–303CrossRefGoogle Scholar
  4. 4.
    Gröger G, Olbrich F, Weiss E, Behrens U (1996) J Organomet Chem 514:81–86CrossRefGoogle Scholar
  5. 5.
    Gröger G, Olbrich F, Schulte P, Behrens U (1998) J Organomet Chem 557:251–258CrossRefGoogle Scholar
  6. 6.
    Olbrich F, Behrens U, Schmidt G, Weiss E (1993) J Organomet Chem 463:249–254CrossRefGoogle Scholar
  7. 7.
    Olbrich F, Behrens U, Gröger G, Weiss E (1993) J Organomet Chem 448:C10–C12CrossRefGoogle Scholar
  8. 8.
    Olbrich F, Behrens U, Weiss E (1994) J Organomet Chem 472:365–370CrossRefGoogle Scholar
  9. 9.
    Schmidt G, Behrens U (1996) J Organomet Chem 509:49–55CrossRefGoogle Scholar
  10. 10.
    Schulte P, Gröger G, Behrens U (1999) Z Anorg Allg Chem 625:1447–1452CrossRefGoogle Scholar
  11. 11.
    Schulte P, Gröger G, Behrens U (2000) Z Anorg Allg Chem 626:679–686CrossRefGoogle Scholar
  12. 12.
    Schulte P, Schmidt G, Kramer CP, Krebs A, Behrens U (1997) J Organomet Chem 530:95–100CrossRefGoogle Scholar
  13. 13.
    Schulte P, Behrens U, Olbrich F (2000) Z Anorg Allg Chem 626:1692–1696CrossRefGoogle Scholar
  14. 14.
    Brussaard Y, Olbrich F, Behrens U (1996) J Organomet Chem 519:115–123CrossRefGoogle Scholar
  15. 15.
    Gröger G, Behrens U, Olbrich F (2000) Organometallics 19:3354–3360CrossRefGoogle Scholar
  16. 16.
    Olbrich F, Schmidt G, Behrens U, Weiss E (1991) J Organomet Chem 418:421–429CrossRefGoogle Scholar
  17. 17.
    Olbrich F, Kopf J, Weiss E (1993) J Organomet Chem 456:293–298CrossRefGoogle Scholar
  18. 18.
    Schulte P, Gröger G, Behrens U (1999) J Organomet Chem 584:1–10CrossRefGoogle Scholar
  19. 19.
    Kuznetsova OA, Khmara EF, Filyakova VI, Uimin MA, Ermakov AE, Rhee CK, Charushin VN (2007) Russ J Gen Chem 77:404–408CrossRefGoogle Scholar
  20. 20.
    Shirtcliff LD, Haley MM, Herges R (2007) J Org Chem 72:2411–2418CrossRefPubMedGoogle Scholar
  21. 21.
    Biswas M, Nguyen P, Marder TB, Khundkar LR (1997) J Phys Chem A 101:1689–1695CrossRefGoogle Scholar
  22. 22.
    Steffen A, Ward RM, Tay MG, Edkins RM, Seeler F, van Leeuwen M, Pålsson LO, Beeby A, Batsanov AS, Howard JA, Marder TB (2014) Chem Eur J 20:3652–3666CrossRefPubMedGoogle Scholar
  23. 23.
    Juvenal F, Bonnot A, Fortin D, Harvey PD (2017) ACS Omega 2:7433–7443CrossRefGoogle Scholar
  24. 24.
    Juvenal F, Langlois A, Bonnot A, Fortin D, Harvey PD (2016) Inorg Chem 55:11096–11109CrossRefPubMedGoogle Scholar
  25. 25.
    Aly SM, Pam A, Khatyr A, Knorr M, Rousselin Y, Kubicki MM, Bauer JO, Strohmann C, Harvey PD (2014) J Inorg Organomet Poly Mat 24:190–200CrossRefGoogle Scholar
  26. 26.
    Knorr M, Guyon F, Khatyr A, Daeschlein C, Strohmann C, Aly SM, Abd-El-Aziz AS, Fortin D, Harvey PD (2009) Dalton Trans (6):948–955Google Scholar
  27. 27.
    Cardolaccia T, Li Y, Schanze KS (2008) J Am Chem Soc 130:2535–2545CrossRefPubMedGoogle Scholar
  28. 28.
    Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, Montgomery JA Jr, Vreven TK, Kudin KN, Burant JC, Millam JM (2004) Gaussian 03, revision C. 02. Gaussian Inc, Wallingford, p 26Google Scholar
  29. 29.
    Hohenberg P, Kohn W (1964) Phys Rev 136:B864–B871CrossRefGoogle Scholar
  30. 30.
    Kohn W, Sham LJ (1965) J Phys Rev 140:A1133–A1138CrossRefGoogle Scholar
  31. 31.
    Parr RG, Yang W (1989) Density-functional theory of atoms and molecules. Oxford University Press, OxfordGoogle Scholar
  32. 32.
    Salahub DR, Zerner MC (1989) The challenge of d and f electrons: theory and computation. American Chemical Society, Washington, DCCrossRefGoogle Scholar
  33. 33.
    Bauernschmitt R, Ahlrichs R (1996) Chem Phys Lett 256:454–464CrossRefGoogle Scholar
  34. 34.
    Casida ME, Jamorski C, Casida KC, Salahub DR (1998) J Chem Phys 108:4439–4449CrossRefGoogle Scholar
  35. 35.
    Stratmann RE, Scuseria GE, Frisch MJ (1998) J Chem Phys 109:8218–8224CrossRefGoogle Scholar
  36. 36.
    Lee C, Yang W, Parr RG (1988) Phys Rev B 37:785–789CrossRefGoogle Scholar
  37. 37.
    Miehlich B, Savin A, Stoll H, Preuss H (1989) Chem Phys Lett 157:200–206CrossRefGoogle Scholar
  38. 38.
    Becke AD (1993) J Chem Phys 98:5648–5652CrossRefGoogle Scholar
  39. 39.
    Binkley JS, Pople JA, Hehre WJ (1980) J Am Chem Soc 102:939–947CrossRefGoogle Scholar
  40. 40.
    Gordon MS, Binkley JS, Pople JA, Pietro WJ, Hehre WJ (1982) J Am Chem Soc 104:2797–2803CrossRefGoogle Scholar
  41. 41.
    Pietro WJ, Francl MM, Hehre WJ, Defrees DJ, Pople JA, Binkley JS (1982) J Am Chem Soc 104:5039–5048CrossRefGoogle Scholar
  42. 42.
    Dobbs KD, Hehre WJ (1986) J Comput Chem 7:359–378CrossRefGoogle Scholar
  43. 43.
    Dobbs KD, Hehre WJ (1987) J Comput Chem 8:861–879CrossRefGoogle Scholar
  44. 44.
    Dobbs KD, Hehre WJ (1987) J Comput Chem 8:880–893CrossRefGoogle Scholar
  45. 45.
    O’Boyle NM, Tenderholt AL, Langner KM (2008) J Comp Chem 29:839–845CrossRefGoogle Scholar
  46. 46.
    Bruker AXS (2008) APEX2, V2008. 6, SADABS V2008/1, SAINT V7. 60A, SHELXTL V6. 14. Bruker AXS Inc., MadisonGoogle Scholar
  47. 47.
    Harvey PD, Knorr M (2016) J Inorg Organomet Polym Mater 26:1174–1197CrossRefGoogle Scholar
  48. 48.
    Harvey PD, Knorr M (2015) J Clust Sci 26:411–459CrossRefGoogle Scholar
  49. 49.
    Harvey PD, Knorr M (2010) Macromol Rapid Commun 31:808–826CrossRefPubMedGoogle Scholar
  50. 50.
    Marineau-Plante G, Juvenal F, Langlois A, Fortin D, Soldera A, Harvey PD (2018) Chem Commun 54:976–979CrossRefGoogle Scholar
  51. 51.
    Chan YH, Lin JT, Chen IW, Chen CH (2005) J Phys Chem B 109:19161–19168CrossRefPubMedGoogle Scholar
  52. 52.
    Kovacs A, Frenking G (1999) Organometallics 18:887–894CrossRefGoogle Scholar
  53. 53.
    Hill JO, Murray JP (1993) Rev Inorg Chem 13:183–197Google Scholar
  54. 54.
    Cavel KJ, Hill JO, Magee RJ (1980) J Chem Soc Dalton Trans 1638:1640Google Scholar
  55. 55.
    Thomas TM, Grimm FA, Carlson TA (1982) Agron PA 25:159–169Google Scholar

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Département de ChimieUniversite de SherbrookeSherbrookeCanada

Personalised recommendations