Evolution of structure and spectroscopic properties of a new 1,3-diacetylpyrene polymorph with temperature and pressure

A new polymorph of 1,3-diacetylpyrene that is luminescent in the solid state and a prominent negative thermal expansion material has been obtained from its melt. A thorough structural characterization of this new crystal form was performed in a wide temperature and pressure range using single-crystal X-ray diffraction. Structural studies have been combined with steady-state UV–Vis spectroscopy and periodic density functional theory calculations. A previously published methodology of crystal placement in a diamond anvil cell has been successfully applied in predicting optimal 2°AP-β sample orientation, ensuring >80% data coverage and enabling unrestrained Hirshfeld atom refinements for high-pressure structures as well as analysis of anharmonic oscillations.


Chapter S1
Supporting materials

S1 Crystallization procedures
Recrystallizations from several solvents and their mixtures have been attempted in order to obtain single crystals.In each case, a mixture of both polymorphs was obtained.Most crystallizations yielded light yellow, thin, hexagonal plates of 2 • AP-α and orange adhesions of thin needles or very small blocks of 2 • AP-β.The majority of crystals were growing on each other and were of small size, which made them impossible to separate and could only be distinguished by their luminescence.
The best way to obtain single crystals of 2 • AP-β is to melt crystals of any 2 • AP polymorph.For this experiment, LinkamScientific TMS94 heating table was used while placed under a microscope.Crystals were placed in a sealed glass vial and quickly (20 • C/min) heated to 170 • C. Then temperature was slowly (5 • C/min) increased to 190 • C. At this temperature, some crystals were starting to melt, but after waiting 5 minutes, there were still many that didn't.The temperature was slowly (5 • C/min) increased to 195 • C, and this temperature was held for another 10 minutes.At this temperature, small yellow particles were found at the top of the vial, which suggests the sublimation of this compound.The temperature was once again slowly (5 • C/min) increased to 202 • C and held for 10 minutes to make sure that all of the crystals had melted.After that, everything was quickly (20 • C/min) cooled to room temperature.This procedure yielded large, orange, solidified chunks on the bottom of the vial and small, yellow particles on the walls in the upper part of the vial.After cutting the chunks from the bottom of the vial, smaller, irregular blocks were obtained and identified as single crystals of a new polymorphic form by X-ray diffraction.Table S1.1:List of solvents and their mixtures used for recrystallizations of 2

S2 Melting point determination
Due to the method of obtaining the 2 • AP-β polymorph, determination of the melting point temperature was performed for a couple of single crystals at atmospheric pressure.The purity and quality of the crystal were confirmed by a short X-ray diffraction experiment (unit cell determination).Crystals were placed on a siliconized glass wafer on a LinkamScientific TMS94 hot stage and then slowly heated at a rate of 2 • C/min in air under a microscope.Such a procedure was repeated twice on fresh single crystals.The melting point was determined at 174 • C. Melting points for polymorphs of 2 • AP compound were also redetermined using the MP70 Melting Point System capillary apparatus (Mettler Toledo) with a heating rate of 5 • C/min, yielding results within 1 • C of the observations for single-crystals.The melting point temperatures observed for the formerly described α polymorphs of 1,3-diacetylpyrene isomers were in excellent agreement with the former observations by [Rajagopal et al., 2014] within 2 • C.

S3.1 X-ray diffraction experiments at atmospheric pressure
For the variable temperature experiments, 5 crystals of 2 • AP-β were used: • crystal 1 -an orange single crystal of dimensions 0.130×0.185×0.216mm, mounted on a glass capillary with a trace of epoxy resin.
• crystal 3 -a yellow single crystal of dimensions 0.093×0.135×0.175mm, mounted on a nylon loop.
• crystal 4 -a yellow single crystal of dimensions 0.058×0.160×0.185mm, mounted on a nylon loop with a trace of epoxy resin.
• crystal 5 -a yellow single crystal of dimensions 0.044×0.130×0.169mm, mounted on a nylon loop with a trace of epoxy resin.
X-ray diffraction experiments at variable temperatures were performed using a Rigaku Oxford Diffraction Supernova diffractometer together with an Oxford Cryosystem 700 series temperature adapter filled with liquid nitrogen.The data were collected in CrysAlisPro [Rigaku-Oxford-Diffraction, 2019].The following 10 different measurement series were performed using different radiations: • using Cu Kα (λ = 1.54184Å): in 275 K (crystal 2) every 10 K from 110 to 300 K (crystal 3) every 10 K from 290 to 180 K (crystal 4) every 10 K from 190 to 270 K (crystal 4) every 10 K from 130 to 90 K (crystal 4) in 293 K (crystal 5) • using Mo Kα (λ = 0.71073 Å): in 100K (crystal 1) every 10 K from 300 to 390 K (crystal 4) every 20 K from 370 to 290 K (crystal 4), (only unit-cell parameters scans) every 10 K from 110 to 90 K (crystal 4) After analyzing results from 110-300 K (Cu Kα) series, it was noted that unit cell constant c started decreasing with increasing temperature.To check the consistency of such anomalous behavior further experimental series involving cooling from 290 to 180 K and warming from 190 to 270 K (Cu Kα) were performed.After obtaining consistent results and for further analysis, the 300-390 K and 370-290 K (Mo Kα) series were performed.A setup with Mo Kα radiation source was chosen in order to allow high temperature measurements.
Another observation after analyzing results from 110-300 K (Cu Kα) series was a discrepancy between unit cell parameters obtained at 110 K and at 100K (Mo Kα) from formerly conducted experiment.In order to check the consistency of unit cell constants at low temperatures, the 110-90 K (Mo Kα) and 130-90 K (Cu Kα) series were performed.This time crystal was cooled down slowly (placed on a diffractometer at room temperature and cooled down in the same rate as cryosystem), as opposed to previous experiments, when crystal was placed directly under 100 K liquid nitrogen stream.
Table S3.1:Unit cell parameters of structures studied in multi-temperature diffraction measurements.The letter in the last column means respectively: S -structure on complete data; T -tentative structure on incomplete data; P -pre-experiment.Systematic extinctions consistent with P 2 1 /c space group.

S5.1 Deposition numbers
Structures determined for the purpose of this paper were deposited as individual entries within Cambridge Structural Database [Groom et al., 2016].Exact deposition number for each structure have been presented in Table S5.2.

S6 Luminescence measurement details
The steady-state UV-vis fluorescence spectra were recorded with a Labram HR800 (Horiba JobinYvon) spectrometer coupled with an Olympus BX61 confocal microscope and equipped with a Peltier-cooled CCD detector (1024 × 256 pixel).A diode pumped, frequency doubled Nd:YAG laser 405 nm (output power 100 mW) was used as the excitation source.Calibration of the instrument was performed using a 520 cm −1 Raman signal of a silicon wafer.The spectra were collected in the 1atm -4.3 GPa pressure range for the same single crystal specimen in a DAC which was used in in-house x-ray diffraction experiments.

S7 Periodic DFT details
Periodic calculations were made in Crystal17 [Dovesi et al., 2018], using the DFT method with B3LYP functional with the Grimme dispersion correction (D3) and the 6-311G basis set.In all cases, strict truncation criteria for the Coulomb and Exchange sums (10 −7 , 10 −7 , 10 −7 , 10 −7 , 10 −29 ) were applied to ensure proper convergence of the calculations.The experimental geometry of the 275 K structure was optimized without temperature restraints.Starting from the same geometry, EOS calculations were also performed, producing 10 theoretical structures in pressure range from 0.06 to 4.08 GPa.For each structure single-point calculations were performed in order to obtain band gaps spans and lattice energies.Interaction energy calculations, taking into account the BSSE correction, were performed for molecules forming π-stacking interactions.
Table S7.1:Unit cell parameters of structures predicted by periodic DFT calculations under applied pressure.

( a )
Vial after melting experiments.(b) Crystals of 2 • AP-β on the bottom of the vial.

Figure S1. 2 :
Figure S1.2:The appearance of the vial and the crystals inside it after the melting experiment, observed under UV light (λ = 395 nm).

Figure S3. 1 :
Figure S3.1:Map of all possible crystal orientations with histogram of completenesses that can be theoretically achieved for our monoclinic sample with DAC opening angle of 52 • and radiation wavelength λ = 0.41618 Å [Tchoń and Makal, 2021].Theoretically predicted completeness of a sample placed in DACOne20 in the actual experimental orientation is 93% (marked by black square).

Figure S3. 2 :
FigureS3.2:The strategy employed to place the crystal in a favorable orientation.According to the potency map (FigureS3.1)for the monoclinic sample at the given experimental conditions, maximal data coverage can be obtained when (100), (001) and (010) directions are not aligned with the DAC axis.(a) A small droplet of the epoxy resin (light yellow) is made on one culet of the DAC (blue) and left to dry.The crystal (orange), after preliminary face-indexing, is propped up to avoid alignment of the crystallographic directions with the DAC main axis, i.e. to avoid (100), (001) and (010) crystal faces being parallel to the diamond face.A trace amount of fresh glue might be used to prevent the crystal from sliding.(b) The DAC is then furnished with the gasket (gray), ruby chip (red), filled with pressuretransmitting medium (light-gray), and closed in preparation for an XRD experiment.

Figure S3. 3 :
Figure S3.3:Layers 1kl for high-pressure experiments in 0.85 and 1.7 GPa.Decrease of intensity and elongation of reflections indicate sample detoriation.
Figure S4.1:Fractal dimension plots [Meindl and Henn, 2008] for structures with and without anharmonic atomic displacement parameters refined.Ideal parabolic shape suggests a random distribution of differential density, i.e. a well-fitting structure model.

Figure S8. 1 :
Figure S8.1:Evolution of the unit cell parameters with external stimuli.Purple points, signed as PRE, were determined from the short pre-experiments (P), based on less than 30 reflections at each temperature point; this resulted in substantial uncertainties.

Table S4 .
1: A listing of refined Gramm-Charlier parameters describing correction for anharmonicity in thermal motions of O1 and O2 atoms at 1 atm and 0.9 GPa.Statistically significant coefficients are represented in bold.

Table S5 .
2: CCDC deposition numbers of crystal structures determined for the purpose of these studies, sorted by crystal phase and exerted pressure.
[Macrae et al., 2008]RMSD and maximal difference, calculated in Mercury[Macrae et al., 2008]in "molecule overlay" mode, of molecules, which were obtained experimentally and theoretically predicted.Energies of frontier crystalline orbitals, resulting band gap spans and related absorption wavelengths, based on periodic DFT calculations.

Table S10
TableS10.3:Individualcomponents of the energy of CH...O interactions B and the distances between the molecules creating them.Molecules forming this interaction are connected through the -x, y+1/2, -z+1/2 symmetry operation.TableS10.5:Individualcomponents of the energy of CH...O interactions D and the distances between the molecules creating them.Molecules forming this interaction are connected through the x, y, z symmetry operation.TableS10.6:Individual components of the energy of CH...O interactions E and the distances between the molecules creating them.Molecules forming this interaction are connected through the -x, -y, -z symmetry operation.
.2: Individual components of the energy of CH...O interactions A and the distances between the molecules creating them.Molecules forming this interaction are connected through the -x, y+1/2, -z+1/2 symmetry operation.