key: cord-0695046-wl1epz3r
authors: Moliterni, Anna; Altamura, Davide; Lassandro, Rocco; Olieric, Vincent; Ferri, Gianmarco; Cardarelli, Francesco; Camposeo, Andrea; Pisignano, Dario; Anthony, John E.; Cristallografia, Cinzia Giannini Istituto di; CNR,; Institute, Paul Scherrer; NEST,; Superiore, Scuola Normale; Nanoscienze, Istituto; Fisica, Dipartimento di; Pisa, University of; Research, Center for Applied Energy; Kentucky, University of
title: Synthesis, crystal structure, polymorphism and microscopic luminescence properties of anthracene derivative compounds
date: 2020-12-05
journal: Acta crystallographica Section B, Structural science, crystal engineering and materials
DOI: 10.1107/s2052520620004424
sha: 899518a300e2e5e2d8a44baea01a6374bc33eda9
doc_id: 695046
cord_uid: wl1epz3r

Anthracene derivative compounds are currently investigated because of their unique physical properties (e.g., bright luminescence and emission tunability), which make them ideal candidates for advanced optoelectronic devices. Intermolecular interactions are the basis of the tunability of the optical and electronic properties of these compounds, whose prediction and exploitation benefit from the knowledge of the crystal structure and the packing architecture. Polymorphism can occur due to the weak intermolecular interactions, asking for detailed structural analysis clarifying the origin of observed material property modifications. Here, two silylethyne-substituted anthracene compounds are characterized by single-crystal synchrotron X-ray diffraction, identifying a new polymorph. Additionally, laser confocal microscopy and fluorescence lifetime imaging microscopy confirm the results obtained by the X-ray diffraction characterization, i.e., shifting the substituents towards the external benzene rings of the anthracene unit favours {pi}-{pi} interactions, impacting on both the morphology and the microscopic optical properties of the crystals. The compounds with more isolated anthracene units feature shorter lifetime and emission spectra more similar to those ones of isolated molecules. The crystallographic study, supported by the optical investigation, sheds light on the influence of non-covalent interactions on the crystal packing and luminescence properties of anthracene derivatives, providing a further step towards their efficient use as building blocks in active components of light sources and photonic networks.

Organic semiconductors have known a growing interest during the last few decades due to their exploitation as active layers of a new generation of optoelectronic devices, such as organic lightemitting diodes (OLEDs; Yersin, 2008; Liu et al., 2013) , organic solar cellar (OSCs; Palilis, et al., 2008; Dou et al., 2013; Ostroverkhova, 2016) , organic field-effect transistors (OFETs; Ito et al., 2003; Allard et al., 2008; Wang et al, 2012; Mei et al., 2013) . Understanding their optoelectronic properties and the correlations with mechanical and morphological properties opened the door to the development of mechanical flexible, easily produced and cheap components for photonics, electronics, and energy conversion (Griffith et al., 2010) .

Among the organic semiconductors, acenes are aromatic hydrocarbons consisting of linearly fused benzene rings; the smallest compound of the acene family is anthracene that can be easily obtained from anthracene oil, i.e., the coal-tar fraction that distils at a temperature above 270 °C. These compounds are especially interesting in their crystalline forms, which allow intriguing effects to be observed, such as polariton lasing (Kena-Cohen et al., 2010) . Intermolecular interactions play a major role in organic crystalline materials, not only determining molecular packing and in turn the optical and electrical properties of the solid-state material, but also its macroscopic properties, such as crystal habit, directly affecting light polarization, confinement and guiding (Camposeo et al., 2019) . For example, needleshaped crystals can exhibit optical waveguide properties with low propagation losses (Camposeo et al., 2019) . - interactions favour an efficient channel for charge mobility (Anthony et al., 2002; Chen et al., 2006; Yao et al., 2018) , and are therefore exploited in material design (da Silva Filho et al., 2005) .

It is also known that non-covalent self-assembly of acenes can be controlled by the insertion of suitable molecular substituents, by which crystal properties can be effectively tailored (Anthony, 2005 (Anthony, , 2008 . Therefore, engineering of the active material relies on the in-depth knowledge of crystal structure and non-covalent interactions. Here, we provide such a detailed characterization through single crystal synchrotron X-ray diffraction for two silylethyne-substituted anthracene compounds, i.e., 1,2,3,4-Tetrafluoro-5,8-bis(trimethylsilylethynyl)anthracene and 9,10-bis(triisopropylsilylethynyl)anthracene (F4 TMS ANT and TIPS ANT, respectively), previously studied by Camposeo et al. (2019) , and Anthony and Parkin (2016) , respectively. The structural characterization of F4 TMS ANT carried out by Camposeo et al. (2019) is confirmed and improved in terms of structure model accuracy, and a detailed description of the intermolecular interactions is provided as well. In addition, a new polymorph of TIPS ANT is identified (named TIPS ANTp in this work), revealing that different molecular packing can result from the same synthesis procedure (Landis et al., 2005) . The occurrence of polymorphism in case of TIPS Anthracene compounds has been recently observed also by Bhattacharyya and Datta (2017) . The position of the molecular substituents towards the external benzene rings is shown to influence the crystal morphology (i.e., needle and plate shape in the case of F4 TMS ANT and TIPS Laser confocal microscopy and fluorescence lifetime imaging microscopy show for the two compounds significant differences in the luminescence properties, along with the uniform emission throughout both the compounds. The lifetime measurements here performed evidence decay times of the photoluminescence (PL) an order of magnitude longer for F4 TMS ANT with respect to TIPS ANTp.

The unlike PL features of the two compounds are in agreement with their dissimilar molecular arrangement suggested by the crystallographic study.

The accurate structural description here provided is a further step in view of tailoring crystal morphology and optical properties to achieve the sought compromise between molecular stabilization and optimal performance of the organic semiconductor (Gu et al., 2012) and/or its coupling to an optical network (Camposeo et al., 2019) .

Synthesis of F4 TMS ANT has been described in (Camposeo et al., 2019) ; the synthesis method for obtaining TIPS ANTp has been reported by Landis and co-workers (Landis et al., 2005) .

Single-crystal X-ray diffraction data were collected at the beamline PXIII (X06DA-PXIII, http://www.psi.ch/sls/pxiii/) at the Swiss Light Source (SLS), Villigen, Switzerland, using a Parallel Robotics Inspired (PRIGo) multi-axis goniometer (Waltersperger et al., 2015) and a PILATUS 2M-F detector. Data collections were performed at room temperature (T = 296 K) on selected crystals of F4 TMS ANT and TIPS ANTp, mounted on litholoops (Molecular Dimensions). For each crystal, complete data were obtained by merging two 360º  scans at =0° and  =30° of PRIGo. In shutterless mode, a 360º data set was collected in 3 min (beam energy of 17 keV, = 0.72932 Å, focus size 90  50 m 2 , 0.25 sec of exposure time per frame, 0.5° scan angle).

Main data collection details are given in Table 1 . Partial data sets were individually processed by XDS (Kabsch, 2010) , a software organized in eight subroutines able to carry out the main data reduction steps; the corresponding XDS_ASCII.HKL reflection files were scaled and merged by the XSCALE subroutine (Kabsch, 2010) . Structure solution was carried out by Direct Methods using SIR2019 and refined by SHELXL2014/7 (Sheldrick, 2015) . All non-hydrogen atoms were refined anisotropically. The carbon-bound H atoms were placed on geometrically calculated positions and refined using a riding-model approximation.

The microscale optical properties of F4 TMS ANT and TIPS ANTp were investigated by laser confocal microscopy (LCM). To this aim, spatially-resolved emission spectra were measured by an inverted microscope equipped with a laser confocal scanning head (FV1000, Olympus), by exciting samples with a continuous wave laser emitting at 405 nm through either a 10× objective (Olympus UPLSAPO) with 0.4 numerical aperture (NA) or a 60× objective (Olympus PLAPON) with NA=1.42. The excitation power was in the range 10-50 µW. The laser was focussed to a diffraction-limited spot onto the sample. The photoluminescence was collected by the same objective and measured by a photomultiplier (Olympus). The sampling speeds during the measurements of the fluorescence micrographs were in the interval 15-80 µs/µm. Typically, higher excitation power and sampling speed were used for F4 TMS ANT samples. In order to measure the spatially-resolved PL spectra, confocal micrographs were collected at various emission wavelengths with a spectral bandwidth of 2 nm. A polarization analyser positioned on the optical path of the emission was used for measuring polarized PL spectra. To this aim, samples were excited by a linearly polarized laser, with polarization direction parallel to the long axis of the TIPS ANTp platelets. The PL lifetime for the crystalline samples was investigated by confocal fluorescence lifetime imaging microscopy (FLIM). This analysis was carried out by an inverted microscope with confocal head (TCS SP5, Leica Microsystem) and either a 40× (NA = 1.25) or a 100× objective (Fluotar, NA = 1.3). Samples were optically excited by a 405 nm pulsed diode laser (Picoquant, Sepia Multichannel picosecond diode laser, maximum average power 30 µW) operating at 40 MHz, whereas the fluorescence intensity was measured by a photomultiplier tube interfaced with a Time Correlated Single Photon Counting setup (PicoHarp 300, PicoQuant, Berlin).

The detection rate was kept in the interval 10 5 -10 6 counts/s, while lifetime time signals were timeintegrated until reaching an average value of the order of 10 2 counts in each area of the scanned micrographs. Spectral filters with variable bandwidths were exploited in order to measure the PL lifetime in different spectral intervals. The measured temporal decay profiles of the PL were fitted to exponential functions convoluted with the instrumental response function, that is assumed to be Gaussian with full width at half maximum of 280 ps.

For confocal analysis the crystalline samples were placed on top of a glass coverslip (thickness 150 µm), while for macroscopic optical characterization the crystalline samples were placed on the surface of a 1×1 cm 2 quartz substrates (thickness 1 mm). Crystalline samples with low inhomogeneities as visible by bright and dark field optical microscopy were selected for the measurements. Absorption spectra were measured by using a UV-visible spectrophotometer (Lambda 950, Perkin Elmer). The samples were mounted on a sample holder for solid state specimens that is made by a metallic plate with a central clear part and two clamps which block two edges of the quartz substrate, while leaving the central part of the substrate free for optical access. The incident optical beam was properly masked in order to have a spot size matching the area of the crystalline sample.

Both F4 TMS ANT and TIPS ANTp were solved and refined by single-crystal synchrotron X-ray diffraction data. According to the literature, TIPS ANT crystallized in the centrosymmetric space group Pbca and the related crystallographic data were deposited at the Cambridge Crystallographic Data Centre (CCDC), with deposition number CCDC 962668 (Anthony and Parkin, 2016) ; in case of TIPS ANT, data collection was carried out at a safe temperature of T= 250 K because the authors observed that a destructive phase transition occurred for crystals cooled to 240 K. In the present work, in order to investigate the occurrence of phase transitions caused by cooling, for TIPS ANTp two diffraction experiments were carried out, at room temperature and at 250 K, respectively. The analysis of the corresponding sets of diffraction data revealed no changes in the crystal structure. Consequently, the results here presented concern only room temperature measurements.

The structure characterization presented in this work in case of F4 TMS ANT confirmed the crystal structure results described by Camposeo et al. (2019) , [whose corresponding CIF file was deposited at the Cambridge Crystallographic Data Centre (CCDC) with deposition number CCDC 1838578 and for the sake of completeness provided also as supplementary material (i.e., file 1838578.cif)] while for TIPS ANTp enabled to identify a new polymorph of TIPS ANT, that crystallized in the centrosymmetric space group P-1.

The crystal structure solution step was carried out by SIR2019 that exploits the information on cell parameters, diffraction intensities and expected chemical formula to determine the space group and solve the structure by Direct Methods (Giacovazzo, 2014) . Crystal structures were refined using full-matrix least-squares techniques by SHELXL2014/7 (Sheldrick, 2015) . Non-hydrogen atoms were refined anisotropically while hydrogen atoms were geometrically positioned and constrained to ride on their parent C atoms with the following bond lengths constraints: C−H=0.96 Å and C−H = 0.93 Å for methyl and aromatic H atoms, respectively. The isotropic U value constraint Uiso(H)=k Ueq(C) was set, with k=1.5 and 1.2 for methyl and aromatic H atoms, respectively; a rotating group model was applied for methyl groups.

Main crystal data and details on data collection and structure refinement are summarized in Table 1 that, in case of F4 TMS ANT, provides also the results obtained by Camposeo et al. (2019) (see the corresponding CIF file) for the sake of comparison.

F4 TMS ANT crystallized in the centrosymmetric space group P21/c, with one molecule in the asymmetric unit (see Figure 1 ), and TIPS ANTp crystallized in the centrosymmetric space group P-1, with half a molecule in the asymmetric unit (see Figure 2 ). In case of F4 TMS ANT the space group determination was automatically carried out by SIR2019 by considering the Laue group compatible with the geometry of the unit cell and assigning a probability value to each related extinction symbol, taking into account a statistical analysis carried out on the experimental intensities; at the end of this step the most plausible space group was graphically selected. For both compounds, the structure solution process was automatically performed by SIR2019. Crystal structure refinement was carried out by SHELXL2014/7 by applying full-matrix least-squares techniques. Non-hydrogen atoms were refined anisotropically while a riding-model approximation was applied in case of hydrogen atoms: H atoms were geometrically positioned at the bond distances C−H=0.96 Å and C−H=0.93 Å for methyl and aromatic H atoms, respectively and allowed to ride on their respective parent C atoms. In case of methyl group, a rotating group model was assumed and the torsion angle defining its orientation about the Si−C bond [in case of F4 TMS ANT] or the C−C bond [in case of TIPS ANTp], was refined. The isotropic U value satisfied the following constraints: Uiso(H)=k Ueq(C), with k=1.5 and 1.2 for methyl and aromatic H atoms, respectively.

Main crystallographic data are given in Table 1 ; additional tables, concerning refined fractional atomic coordinates and displacement parameters, bond distances and angles, and torsion angles, were provided in the Supplementary Information.

For both compounds no presence of classical hydrogen bonds was detected; the crystal packing suggested that the main intermolecular interactions were weak and, in case of F4 TMS ANT, consisted of - interactions (Janiak, 2000; Meyer et al., 2003; Martinez and Iverson, 2012) and Caryl−F···H−C interactions (Meyer et al., 2003) ; for TIPS ANTp, they were aromatic interactions, weaker than those involved in F4 TMS ANT, and C−H··· interactions (Meyer et al., 2003; Nishio, 2004; Nishio et al., 2012) .

In case of F4 TMS ANT, the refined crystal structure here described is similar to that obtained by Camposeo et al. (2019) , as shown in Figure 3 , providing the overlay of the two structure solutions, having an r.m.s. deviation of 0.04 Å (SIR2019; Burla et al., 2015) . Our results, compared to the structure model by Camposeo et al. (2019) , are characterized by a not negligible improvement in terms of C−C bond precision and agreement factors; the bond precision, in case of our results, is 0.0032 Å instead of 0.0061 Å, while the R[F 2 > 2σ(F 2 )], wR(F 2 ) were 0.048 and 0.148 instead of 0.062 and 0.175, respectively (see Table 1 ).

For F4 TMS ANT ( Figure 1 ) the anthracene core consists of three planar-like six-membered rings. The mean distance from the least-squares plane calculated for 18 atoms, including 14 C-atoms of anthracene group and the four F-atoms, was 0.0074 Å and the largest deviation was at C14 (0.023 Å) and at F4 (0.019 Å). A slight bending was observed for the lateral chains involving the alkyne group: the distances from the least-squares plane in case of C15, C16 and Si1 were 0.024, 0.047 and 0.117 Å, respectively, while in case of C20, C21 and Si2 were 0.074, 0.135 and 0.280 Å, respectively. As usual for anthracenyl units, the peripheral aromatic rings were distorted from the hexagonal geometry with shortening of some bond distances (Kovalski et al., 2018) : the bond distances C5−C6, C7−C8, C1−C14, C12−C13 were 1.34, 1.346, 1.378 and 1.364 Å, respectively, while the rest of bond distances were between 1.389 and 1.445 Å. The mean value of the C−C bonds was 1.4098 Å (ring C1 C2 C11 C12 C13 C14), 1.4042 Å (ring C2 C3 C4 C9 C10 C11) and 1.3938 Å (ring C4 C5 C6 C7 C8 C9). The crystal packing evidenced the presence of weak Caryl−F···H−C interactions (Meyer et al., 2003) and parallel offset -interactions (Janiak, 2000; Meyer et al., 2003; Martinez and Iverson, 2012) , as shown in Figures 4 and 5. These kinds of -arrangements are energetically more stable and favoured than the parallel face-centred stacked ones (Martinez and Iverson, 2012) . The minimum distance between centroids of the aromatic rings was 3.698 Å (see Figure 5 ) and the minimum interplanar distance was 3.422 Å, in agreement with the typical values of the interplanar distances for -interactions, belonging to the range 3.3-3.8 Å (Janiak, 2000) . The parallel offset -interactions were responsible for stacking arrangements along a (see Figure 4 ). Caryl−F···H−C interactions contributed to stabilize the crystal structure (Meyer et al., 2003) and the values of the distance F·· C belonged to the range observed for this kind of interactions (i.e., 3.3-3.6 Å, see Meyer et al., 2003) .

In case of TIPS ANTp, the mean distance from the least-squares plane, calculated for the C-atoms of anthracene group, was 0.0041 Å and the largest deviation was at C3 (0.006 Å) and at C4 (0.006 Å). The bending observed for the lateral chain involving the alkyne group was larger than in case of F4 TMS ANT, probably due to the larger number of methyl groups: the distances from the least-squares plane in case of C6, C1 and Si1 were 0.048, 0.136 and 0.409 Å, respectively. Also for TIPS ANTp the peripheral aromatic rings were distorted from the hexagonal geometry: the bond distances C10−C12 and C9−C11 were 1.358 and 1.356 Å, respectively, while the rest of bond distances were between 1.411 and 1.430 Å. The mean value of the C-C bonds in the three aromatic rings shown in Figure 2 is 1.4005 Å (ring C4 C10 C12 C11i C9i C3i), 1.4173 Å (ring C3 C2 C4 C3i C2i C4i) and 1.4005 Å (ring C11 C9 C3 C4i C10i C12i). A view along a of the crystal packing is given in Figure 6 . Differently from TIPS ANT [Anthony and Parkin (2016) ], the crystal packing in TIPS ANTp (present work) did not reveal any edge-to-face interaction; the main intermolecular contacts were weak interactions between parallel aromatic rings and C−H··· interactions (Meyer et al., 2003; Nishio, 2004; Nishio et al., 2012) . The last ones, as observed by Nishio (2004) , are entropically favoured and contribute to stabilize the crystal structure. The parallel aromatic rings are characterized by a large offset (the shortest distance between centroids of parallel aromatic rings is 4.726 Å, see Figure 7 , and the minimum interplanar distance is 2.471 Å); consequently, these interactions, in spite of the short interplanar distance, were weaker than the parallel-offset -interactions detected in F4 TMS ANT, and due to the large distance between centroids are not the typical - interactions (Janiak, 2000) . If compared with F4 TMS ANT, the weaker -interactions in case of TIPS ANTp could lead to a reduction of the charge mobility, which should be confirmed by proper electrical characterization (this study is beyond the goal of this paper). The values of C-H··· distances shown in Figure 7 belonged to the distance range observed for this kind of interactions (i.e., 3.3−4.1 Å, see Meyer et al., 2003 , Hattab et al., 2010 .

represented via Hirshfeld surfaces, using the CrystalExplorer17 software . The Hirshfeld surface offers a useful tool for measuring the space occupied by a molecule in a crystal and summarizing information on all intermolecular interactions. In Figures S1 and S2 the Hirshfeld surface mapped over dnorm is shown in case of F4 TMS ANT and TIPS ANTp, respectively. The conventions for the surface colours are the following ones: blue, white and red colours identify the interatomic contacts as longer, at van der Waals separations and short, respectively. In both the figures the blue colour is predominant. No red region is observed for TIPS ANTp while in case of F4 TMS ANT a very small and faint red region, indicated by red arrows in Figure S1a and zoomed in Figure S1b Figure 8c) . Indeed, the variation (Δλp) of the PL peak wavelength (λp) along the length of the needle of F4 TMS ANT is <3 nm (Δλp/λp=0.5%), and a similar spatial stability is found for the PL full width at half maximum (measured values are in the interval 102-104 nm). The PL spectrum of a single platelet crystal of TIPS ANTp is, instead, much more structured, with peaks at 454 nm, 474 nm, 500 nm, 508 nm and 538 nm (Figure 8b ). Some of these PL peaks (the ones at 474 nm and 508 nm) are close to those of the PL spectrum of molecules of TIPS ANTp in solution (see Figure S3 in the Supporting Information, where PL peaks at 446 nm, 475 nm and 507 nm can be identified for the molecule in solution) (del Valle et al., 2002) . In addition, polarized PL spectra shown in Figure S4 in the Supporting Information highlight a variation of the shape of the emission spectrum of TIPS ANTp upon changing the polarization of collected light. This analysis unveils the presence of a peak, in the high energy tail of the PL spectrum, that is polarized in a direction parallel to the short axis of the crystal face, and a dependence of the intensity ratio of the peaks at 500 nm and 508 nm on polarization. The structured shape of the spectrum and the polarization dependence are indicative of the presence of different emissive species. Here we point out that the confocal microscopy measurements allow some spectral features of the spectrum below 475 nm to be unveiled for crystalline samples of TIPS ANTp, which were previously masked by self-absorption (Camposeo et al. 2019) . This is highlighted in Figure S5a where the emission spectra of TIPS ANTp, measured by vertically shifting the high-numerical objective (NA=1.42) along the crystal thickness (z axis in Figure S5 ), are compared. This analysis highlights the increasing contribution of self-absorption as the excitation focal spot is shifted into the crystal, resulting in a decrease of the intensity of the high energy transitions with respects to the low energy one ( Figure   S5b ). The larger self-absorption effect in case of TIPS ANTp is also favoured by the strong overlap between the absorption and PL spectra, more prominent compared to F4 TMS ANT (see Figure S6 ).

Similarly to F4 TMS ANT, the PL spectrum does not feature significant variations within individual platelet crystal, as obtained by spatially-resolved photoluminescence (Figure 8d ). The different shape of the crystalline samples (needle vs platelet) was demonstrated to determine a different light transport behaviour (Camposeo et al. 2019 ), namely a more efficient self-waveguiding of the emitted light in needles compared to platelets.

The PL lifetime of F4 TMS ANT and TIPS ANTp are shown in Figure 9a and 9b, respectively. While a long lifetime in the range 50-80 ns is estimated for F4 TMS ANT, the PL lifetime of TIPS ANTp is about an order of magnitude shorter, being in the interval 1-4 ns (compare insets of Figure 9a ,b). The longer PL lifetime in case of F4 TMS ANT could be due to π-π interactions and the columnar stacking of chromophores. Interestingly, also the spectral dependence of the lifetime is different for the two compounds: the temporal decay of the PL of F4 TMS ANT is constant throughout the spectral range of the emission, whereas for TIPS ANTp we observed an increase of the lifetime by red-shifting the PL wavelengths (Figures 9a,b) . More specifically, the lifetime for the high energy components of TIPS ANTp (in the interval 450-475 nm) is about 0.8 ns, and it increases to about 4 ns for the low energy ones (525-600 nm). For the sake of comparison, we recall that solutions of TIPS ANT molecules at low concentrations have been reported to show a wavelength-independent lifetime of about 6.6 ns (Pun et al. 2018) . Measurements performed at various z-positions of the objective do not evidence significant differences of the PL decay curves (the estimated lifetime being around 2 ns for all z values), ruling out potential effects related to self-absorption and re-emission ( Figure S7 of the Supporting Information).

A similar behaviour of the lifetime was reported also for other acenes, such as anthracene and tetracene (Ahn et al. 2008 and Camposeo et al. 2010) , and was attributed to excitonic and defect or trapped states.

Overall, the results of the microscopic PL measurements are consistent with the crystallographic analysis and support the different molecular packing of the investigated compounds in crystalline samples. In fact, F4 TMS ANT crystallizes as elongated needle-like samples, with parallel offset π-π interaction and columnar stacking of chromophores. These configurations might lead to excimers (Liu et al. 2016) with resulting red-shifted, broad, long-lasting emission. On the contrary, the increased separation of the molecules in crystals of TIPS ANTp and the weaker molecular interactions, allow the electronic properties of the individual molecules to be partially preserved in the crystalline samples, which show structured PL spectra ascribable to different emitting species. 

The crystal structure of two silylethyne-substituted anthracene compounds [i.e., 1,2,3,4-Tetrafluoro-5,8-bis(trimethylsilylethynyl)anthracene and a new polymorph of 9,10bis(triisopropylsilylethynyl)anthracene] was determined by single-crystal synchrotron diffraction to identify main factors influencing the optical properties of these organic semiconductors. The crystallographic study revealed that the two compounds were characterized by different intermolecular interactions, responsible for dissimilar luminescence effects. The crystal morphology also affects the optical properties of the two compounds: needle-shape crystals are characterized by  interactions and show broad and long-lasting photoluminescence. This was not found in case of the second crystal that was platelet shaped, for which a brighter PL with shorter lifetime was also observed, that could be exploited for the development of efficient light-emitting components and optical sensors. Cell refinement: XDS (Kabsch, 2010) ; data reduction: XDS (Kabsch, 2010) ; program used to solve structure: SIR2019 ; program used to refine structure: SHELXL2014/7 (Sheldrick, 2014) ; molecular graphics: SIR2019 , Mercury ; software used to prepare material for publication: WinGX (Farrugia, 2012) and publCIF (Westrip, 2010) . Special details Geometry. All esds (except the esd in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell esds are taken into account individually in the estimation of esds in distances, angles and torsion angles; correlations between esds in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell esds is used for estimating esds involving l.s. planes. Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > 2σ(F 2 ) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger. Positive values of z correspond to positioning the excitation laser focal spot into the crystal thickness as shown in Figure S5 . The continuous black lines are fits to data by exponential functions convoluted with the instrumental response function. These measurements were performed by using the 100× objective with NA=1.3.

SHELXL2014/7 (Sheldrick, 2015), WinGX (Farrugia, 2012), publCIF (Westrip, 2010)

Phasing in Crystallography: A Modern Perspective

The CH/ Hydrogen Bond: Implication in Crystal Engineering In The Importance of Pi-Interactions in Crystal Engineering: Frontiers in Crystal Engineering, First Edition

Electronic Processes of Organic Crystals

C4-C5-C6-C7 0

C5-C6-C7-C8 −0.2 (4) C2-C11-C12-C13

C5-C6-C7-F3 −

F2-C6-C7-F3 0.3 (3) C20-C12-C13-C14

F3-C7-C8-F4 −0.5 (3) C11-C12-C13-C14 −0

C6-C7-C8-F4

C6-C1-Si1

Si1-C5-H5 107

Si1-C7-H7 106

Si1-C8-H8 106

C7-Si1-C5-C13 −60

C1-Si1-C5-C15 −66

C8-Si1-C5-C15 48

C7-Si1-C5-C15

C1-Si1-C7-C16 −162

C4-C10-C12-C11 i −0.1 (2) Symmetry code: (i) −x, −y+1