Article Cite This: J. Phys. Chem. A 2018, 122, 4628−4634
pubs.acs.org/JPCA
Structural and Spectroscopic Investigations of Two [Cu4X6]2− (X = Cl−, Br−) Clusters: A Joint Theoretical and Experimental Work Camille Latouche,* Romain Gautier,* Romain Génois, and Florian Massuyeau Institut des Matériaux Jean Rouxel (IMN), Université de Nantes, CNRS, 2 rue de la Houssinière, BP 32229, 44322 Nantes Cedex 03, France S Supporting Information *
ABSTRACT: Herein we report a joint experimental and theoretical investigation on two tetranuclear Cu(I) clusters stabilized by halide ligands. These clusters are of high interest due to their spectroscopic and optical properties, more precisely both clusters exhibit thermochromism. The compounds synthesized by the hydrothermal method have been characterized by singlecrystal X-ray diffraction, UV−visible spectroscopy and quantum calculations. Modeled structures have been investigated by means of DFT and TD-DFT methods. Anharmonic computations have been performed to better achieve the vibrational investigation. Computations of the triplet excited states permit us to get more insights into the structure and electronic structure of the excited states responsible for the luminescence properties. Calculations are in agreement with the observed phosphorescence wavelengths.
■
INTRODUCTION Nowadays materials exhibiting photoluminescence properties as a response to stimuli such as temperature, electricity or grinding have demonstrated their promises for several types of applications as sensors, memories or display devices.1−11 As a matter of fact, the past years have been effective to target new compounds devoted for these applications and chemists have provided much effort to synthesize and design new functional materials. On these grounds, computationally and experimentally oriented chemists joined forces to obtain new compounds and especially new clusters exhibiting strong luminescence properties, based on Cu+/Ag+ cluster types (formally d10).12−25 Among all these clusters, it should be mentioned that some of them are able to trap anion species (ion, molecule) allowing us to shape new architectures together with promising optical properties.26−34 Analyzing these assemblies becomes mandatory and it allows a rationalization of the phenomena occurring when spectroscopic properties (optical, vibrational) are observed. Recently, some of us have reported the synthesis, the characterization, and a DFT rationalization of the observed properties of the [Cu4Br6]2− cluster.35 In the latter investigation, two phosphorescence wavelengths were observed and the computations retrieved with an acceptable accuracy both of them allowing a good description of the triplet excited states. In this paper, we © 2018 American Chemical Society
present a combined theoretical and experimental investigation on two copper halide clusters, i.e., [Cu4Cl6]2− (1) in in [C6H16N2]3[Cu4Cl6][Cu2Cl6]·H2O (1a) and [Cu4Br6]2− (2) in [C6H16N2]3[Cu4Br6][Cu2Br6] (2a). For compound 1, the electronic structure is described, a band assignment of the observed absorption band using the TD-DFT method is proposed, IR and Raman spectra are simulated using anharmonic computations and the optical properties issued from the triplet excited states are rationalized. As stated before, the optical properties of 2 have already been reported; therefore, we mainly focus our attention on the vibrational properties for this compound. Finally, the possibility to entrap an hydride into 1 has been investigated in order to evaluate the capability of the cluster to be a potential candidate for hydrogen storage. All the results from DFT computations are compared with the available experimental ones.
■
EXPERIMENTAL SECTION Synthesis. The compound [C6H16N2]3[Cu4Cl6][Cu2Cl6], H2O (1a) was synthesized by a hydrothermal method from a mixture of 7.87 mmol of Cu metal, 4.43 mmol of N,N′Received: March 19, 2018 Revised: April 26, 2018 Published: April 26, 2018 4628
DOI: 10.1021/acs.jpca.8b02663 J. Phys. Chem. A 2018, 122, 4628−4634
The Journal of Physical Chemistry A
■
dimethylpiperazine in 3 mL of HCl 37%. The mixture was heated at 180 °C for 24 h and slowly cooled to room temperature at the rate of 10 °C/h using a 23 mL Teflon-lined stainless steel autoclave. Single crystals suitable for single crystal X-ray diffraction were recovered by filtration. Compound 2a was synthesized as previously described in the literature.35 Structure Determination. The structure determination of compound 1a was carried out from single-crystal X-ray diffraction with a Bruker-Nonius Kappa CCD diffractometer (monochromated Mo Kα radiation). Absorption corrections were carried out using SADABS.36 The crystal structure was determined by direct methods and was completed by Fourier difference syntheses with SIR2004.37 SHELXL-2013 was used to refine the crystal structures, and anisotropic displacement parameters were considered.38Additional symmetry elements were checked with the program PLATON.39 The crystallographic data for compound 1a are summarized in Table 1.
largest diff peak/hole (e Å−3)
RESULTS AND DISCUSSION
Structure Description. The new compound 1a is isostructural to the previously reported d10 copper chloride compound 2a.35 The crystal structure of 1a is described in the noncentrosymmetric space-group P212121. The compound is built from two clusters ([Cu2Cl6]4− and [Cu4Cl6]2− units). The Cu−Cl bond lengths are in the range 2.223(3) Å < d < 2.691(3) Å. In [Cu2Cl6]4− units, Cu−Cu distances are 2.6724(17) and 2.6431(16) Å < d < 3.066(2) Å for [Cu4Cl6]2− units. The cations, namely 1,4-dimethylpiperazine1,4-diium, together with water molecules are located between the different anionic species. Ground State Investigation. As the geometrical parameters of 2 have already been reported and discussed in ref 35, this section is mainly focused on the crystal and electronic structures of 1. The data from the simulated optimized geometries are compared to the X-ray diffraction ones. Similarly to cluster 2, the optimized simulated structure of the cluster with X = Cl is in reasonable agreement with respect to the experimental one. In this geometry, the four metal atoms form a tetrahedron in which the six edges are capped by a μ2-halide ligand. Consequently, each metal atom is linked to three halides in a trigonal-planar MX3 configuration (Figure 1). This statement seems of high interest since such fragments characterized up to now generally have a fourth ligand attach to each Cu, sometimes with an enchlatrated oxo ion.52−56
Table 1. Crystallographic Data for Compound 1a chemical formula space group a (Å) b (Å) c (Å) α (deg) β (deg) γ (deg) radiation 2Θ range for data collection (deg) reflections collected no. of data/restraints/parameters goodness-of-fit on F2 final R indexes [I ≥ 2σ(I)]
Article
[C6H16N2]3[Cu4Cl6][Cu2Cl6], H2O P212121 13.3441(13) 14.6101(9) 20.1415(19) 90 90 90 Mo Kα 12.84−51.58 31711 7409/0/387 1.043 R1 = 0.0569 wR2 = 0.1213 +1.11/−0.87
Optical Characterization. Diffuse reflectance spectra were collected from 250 to 2500 nm using a Varian Cary 5G spectrophotometer with a 60 mm integrating sphere. Absorbance spectra were calculated from reflectance measurements using the Kubelka−Munk function (a/S = (1 − R)2/2R, where a is the absorption coefficient, S the scattering coefficient, and R the reflectance). The photoluminescence spectra at low and room temperature were collected using a Spex Fluorolog-3 spectrofluorometer from Horiba Jobin Yvon. The excitation source is a 450 W Xe lamp. Computational Details. Ground state (GS) and triplet excited state (ES) computations have been performed on the models using the PBE040,41 functional together with the Def2TZVPD42 basis set. The Gaussian package has been used for all the calculations.43 All computations have been performed in a vacuum, and the geometries have been checked to be at the minimum of the potential energy surface by diagonalizing their Hessian for both ground and excited states. Simulated spectra have been modeled using the VMS package.44 Optical properties were investigated using TD-DFT. Calculations of IR and Raman vibrations and spectra have been performed with the GVPT245−50 model as implemented in the used version of Gaussian. Structures and orbital plots have been drawn with the GaussView software.51
Figure 1. Investigated compounds: left corresponds to 1(Cl) and right to 2(Br).
When the Td symmetry is enforced, the computed Cu−Cu and Cu−X distances match nicely the averaged observed ones (for 1 and 2, respectively: Cu−Cu (Å), 2.717 [2.843] and 2.749 [2.869]; Cu−X (Å), 2.329 [2.300] and 2.459 [2.410]) (Table 2). According to the crystal structures of each compound, the Table 2. Relevant Computed and Averaged Experimental (between Square Brackets) Parametric Data of Clusters 1 and 2 1 2
Cu−Cu (Å)
Cu-X (Å)
2.717 [2.843] 2.749 [2.869]
2.329 [2.300] 2.459 [2.410]
tetranuclear copper species is connected to a binuclear copper one through a long Cu−X contact. For compound 1, the latter distance is ca. 2.70 Å, which is by far larger than the other Cu− Cl distances (ca. 2.30) and is also much larger than the sum of covalent and atomic radii of Cu and Cl.57 This distance is also higher than the sum of the ionic radii of Cl and Cu in tetravalent coordination mode (2.40 Å).58 Furthermore, one 4629
DOI: 10.1021/acs.jpca.8b02663 J. Phys. Chem. A 2018, 122, 4628−4634
Article
The Journal of Physical Chemistry A
this crystal make very complicated the identification of the origin of the experimental signal. At this stage, the following results should be taken as fully predictive. To be the most accurate in the vibrational characterization of compounds 1 and 2, IR (Figure 3a) and Raman (Figure 3b) spectra have been simulated using the anharmonic approximation by keeping the same functional and basis set (Figure 3). Furthermore, the used methodology is the same as reported by some of us in previous reports where the agreements between the available experimental data and the computations were good.59,60 It should be pointed out that the anharmonic corrections to the fundamental bands are weak in this region (
