|

Exploring polymorphism in molecular compounds using high pressure

High-Pressure Crystallography
Francesca P. A. Fabbiani Emmy-Noether Jr. Research Group

Introduction to high-pressure Dr. Michael Ruf
Experimental setup Product Manager, SC-XRD Madison, WI, USA Dr. Francesca P. A. Fabbiani

 High-pressure crystallography with Collecting high-pressure data, emphasis on single-crystal X-ray structure determination solving and refining structures  High-pressure crystallization  Solid-state polymorphism  Molecular compounds with emphasis on pharmaceuticals and Molecular organic
materials, single crystals
Introduction to
High-Pressure Research

High-Pressure Research
Range of pressures and temperatures now accessible with static compression techniques in the laboratory. Planetary science and physics (e.g. minerals, perovskites, clathrates, ices – N.B. 15 polymorphs of H O Synthesis of novel materials (e.g. superhard materials, nanoporous materials, MOFs) Mao H , Hemley R J PNAS 2007;104:9114-9115
2007 by National Academy of Sciences Chemistry/molecular compounds Phase diagram of water (e.g. amino acids, molecular magnets, pharmaceuticals) Proteins, pressure-induced denaturation, folding Source: Martin Chaplin, http://www.lsbu.ac.uk/water/phase.html

High-Pressure Research
Industrial applications, e.g. pressure treatment of food Origin of life/ sustained life at the bottom of the oceans Source: Wikimedia Commons Access new materials, probe intermolecular interactions Understand polymorphic Source: Wikimedia Commons Interplay with theory Source: Wikimedia Commons High-Pressure Scale
(1000 atmospheres ≈ 1 kbar = 100 MPa = 0.1 GPa) intergalactic inside a light 10-16 10-5 10-4 10-2 10-1 Molecular materials
Orders of magnitude available for pressure variation >> temperature variation Experimental Setup
Pressure Generation for in situ XRD:
Diamond-Anvil Cell
First diamond anvil cell at NIST Gaithersburg Museum Aluminium DAC mounted on a DAC manufactured by Dr. H. Ahsbahs F. P. A. Fabbiani et al. CrystEngComm (2010), 12, 2354-2360 Source: G. J. Piermarini, Wikimedia Commons Pressure Generation for in situ XRD:
Diamond-Anvil Cell
P= F/A (kgms-2/m2 = N/m-2 = Pa)
Tightening screws Choose your DAC carefully: backing plates material, diamond cut, purity and culet size, gasket hole size, DAC opening angle DAC Loading
Prepare gasket, i.e. sample chamber Equipment for DAC preparation
• EDM machine (spark eroder) or Load sample together with pressure laser to drill gasket hole calibrant, e.g. ruby chip • Stereomicroscope with good (fluorescence) or quartz single crystal magnification, large working distance and polariser/analyser • Small spectrometer with large Sample can be in the solid (direct working distance to monitor ruby compression in an hydrostatic fluorescence if using ruby as medium), liquid or gas phase (in situ pressure calibrant crystal growth). A solution can also be loaded for in situ crystal growth DAC Loading
Three main sample loading methods available
Sample loading methods depend on: Nature of the sample under investigation Aim of experimental investigation The focus here is on molecular organic materials for single-crystal X-ray diffraction "Fathers" of high-pressure research: G. H. J. A. Tammann (1861-1938), P. W. Bridgman (1882-1961) DAC Loading
Approach 1: Direct compression in a hydrostatic medium Good for studying polymorphism in small molecules Less effective for studying polymorphism in larger and/or rigid molecules: kinetic barrier associated with molecular rearrangement is usually large Good for studying evolution of structure as a function of pressure, for obtaining p-T phase diagrams and isothermal equation of states Choice of hydrostatic medium: solubility/ freezing pressure considerations DAC Loading
Approach 2: In situ crystallisation and growing from the melt Excellent method for crystallising new polymorphs of compounds with melting points < 40°C and for comparing with low-temperature structures Not effective for higher melting organic compounds, which can decompose before the onset of melting DAC Loading
Approach 2: In situ crystallisation and growing from the melt DAC Loading
Approach 3: In situ crystallisation from solution Excellent method for crystallising new polymorphs and solvates. No limitation to low melting point or small molecules Can vary solvent, pressure, temperature and concentration Prerequisite: relatively high solubility of the solute; solubility of solute increases with increasing temperature and decreases with increasing pressure Introduction to high-pressure research Experimental setup Collecting high-pressure data, solving and refining structures Molecular organic materials, single crystals
Collecting high-pressure
data, solving and refining
structures
Data Collection
Centre on the diffractometer, 2-step Set up @ GZG Göttingen
procedure: optical centring and direct beam centring (Dawson et al. 2004) or diffractometric measurements (King & Finger 1979, Dera & Short collimator and long beamstop
Choose suitable data collection Diffracted beam D strategy (and wavelength, if applicable) and exposure time; Data collection in
transmission mode
maximise no. frames per run (this helps during data integration if integrating with Bruker software) A. Dawson et al. J. Appl. Cryst. (2004), 37, 410-416 H. E. King & L. W. Finger J. Appl. Cryst. (1979), 12, 374-378 P. Dera & A. Katrusiak J. Appl. Cryst. (1999), 32, 510-515 Data Collection
For one orientation of the DAC, the accessible region of reciprocal space is determined by the detector distance and the DAC opening angle By a combination of ω-scans using different orientations of the DAC in φ, and different orientations of the Steel support of the DAC detector in 2θ, about ⅓ of all starting to obscure the reflections can be collected (Angel et al. 1992). This can be increased by collecting more data with a different orientation of the DAC with respect to χ (see later) R. J. Angel et al. Phase Transitions (1992), 39, 13-32 Data Collection
Example of data collection strategy for a 3-circle diffractometer, detector distance = 7 cm and DAC perpendicular to the beam @ phi = 0°, ½ DAC opening angle = 45° 2 Theta Omega Phi Sweep Scan direction Important: check hardware limits; check diffraction limits and adjust 2 Theta and Omega accordingly Data Collection
Set up @ GZG Göttingen
X-ray Crystallography
Diffracted beam D Limited sampling of reciprocal space  data completeness and resolution structure solution Sample scattering power (and size) structure refinement Data Completeness
Coverage of reciprocal space Increasing Data Completeness
Rotate by 120°
Rotate by 120° On a 3-circle
and collect
and collect
Rotation of the DAC DAC with large opening angle + non-diffracting backing plates Data collection with short-wavelength radiation (see later) Careful orientation of the crystal in the DAC (ambient p) Multiple or twinned crystals F. P. A. Fabbiani et al. CrystEngComm (2010), 12, 2354-2360 Data Processing
Data indexing identify reflections arising from sample Harvesting reflections If large background variations, reduce the number of images and runs Choose an appropriate value It is useful to exclude certain regions, e.g. Be rings, gasket rings; if sample reflections are scarce and indexing is difficult, try omitting high-resolution regions, where diamond reflections are more abundant. This can also be achieved through a reciprocal lattice viewer (recommended) Data Processing
Data indexing identify reflections arising from sample Reciprocal lattice viewer The reciprocal lattice viewer is an invaluable tool for indexing, for assessing data quality and for twin-spotting. Reflections can be conveniently assigned to different groups and exported for indexing with external programs, e.g. CELL_NOW Data Processing
Data indexing identify reflections arising from sample Data integration mask out regions of detector obscured by the DAC; choose appropriate resolution; background correction The following are recommendations based on personal experience. There is no "one-fits-all" strategy that will work for every sample: try different options to optimise your integration. Once the integration parameters have been optimised, I would strongly recommend performing successive integration cycles ("UB matrix update") for best intensities and unit cell parameters Data Processing
Data indexing identify reflections arising from sample Data integration mask out regions of detector obscured by the DAC; choose appropriate resolution; background correction Try to keep the box size small. If problems with convergence: uncheck this option If background is jumpy  choose high frequency; otherwise reduce Twins can be easily handled Data Processing
Data indexing identify reflections arising from sample Data integration mask out regions of detector obscured by the DAC; choose appropriate resolution; background correction This option might be useful for synchrotron data Threshold for strong reflections: lower this to, e.g. 8 for weak data For using dynamic masks generated with an external program Data Processing
Typical frames from CCD Area Detector Powder ring from almost invisible at 2θ = 0 (gasket and Powder ring from Shading from DAC opening Dynamic masks: A. Dawson et al. J. Appl. Cryst. (2004), 37, 410–416; N. Casati et al. J. Appl. Cryst. (2004), 40, 620-630 Data Processing
Data indexing identify reflections arising from sample Data integration mask out regions of detector obscured by the DAC; choose appropriate resolution; background correction Generate dynamic masks "on the fly", e.g. with the Bruker SAINT integration software, V8.07A run from the command line ("Advanced options") Input DAC geometry Data Processing
Data indexing identify reflections arising from sample Data integration mask out regions of detector obscured by the DAC; choose appropriate resolution; background correction Scaling and absorption correction 2-stage procedure: analytical correction for DAC components and gasket shadowing, see programs by S. Parsons, A. Katrusiak and R. J. Angel; multiscan correction to correct for other systematic errors and for scaling, e.g. SADABS. Beware of outliers, e.g. diamond reflections! Space group determination difficulty related to completeness, redundancy, resolution and crystal orientation  systematic absences are not always present Data Processing
Structure solution direct methods, global optimisation methods (borrowed from powder diffraction), molecular replacement, etc.: numerous programs available! Data merging crucial step; robust-resistant and experimental (1/σ2) weighting scheme with SORTAV SORTAV: R. H. Blessing J. Appl. Cryst. (1995), 30, 421–426 Refinement
Very high-quality high-pressure data can be collected nowadays. It is nevertheless important to be realistic during refinement. Refinement of ADPs for all non-H atoms might not be possible Most commonly encountered problem: low data to parameter ratio; restraints are your friends: treat them well and be generous Always investigate outliers before omitting reflections: go back to the original frames The following are examples taken from my own research CRYSTALS: P. W. Betteridge et al. J. Appl. Cryst. (2003), 36, 1487 SHELXL: G. M. Sheldrick Acta Cryst. (2008), A64, 112-122 Refinement
Example of problematic data, lab source In situ crystallisation study Cell setting, space group Monoclinic, P2 /n Constraints: 2 rigid bodies a, b, c (Å) 7.630(2) 17.209(3) 7.3708(11) 1 isotropic parameter Multi-scan abs. correction T No. of measured, independent and observed [F > 4σ(F)] reflections No. of parameters R [F > 4σ(F)], wR (F2, all reflections) (Å-1) and completeness (%) Structure solution: DASH Refinement program: CRYSTALS Refinement
Example of good data, lab source In situ crystallisation study Cell setting, space group Monoclinic, P2 /c a, b, c (Å) 8.9537(11) 5.4541(6) 13.610(4) SIMU, DELU, DFIX [for (N)H positions] restraints Multi-scan abs. correction T No. of measured, independent and observed [F > 4σ(F)] reflections No. of parameters and restraints R [F > 4σ(F)], wR (F2, all reflections) (Å-1) and completeness (%) Structure solution: Sir92 Refinement program: CRYSTALS Refinement
Example of good data, lab source Here would expect diamond overlaps Overlap with gasket Shaded reflections Increasing data quality – Part I
Synchrotron radiation Useful properties of synchrotron radiation: Brilliance  gain in diffracted intensity compared to a lab source, i.e. increase in resolution and Tuneable wavelength: short- wavelength radiation is accessible  less absorption and significant gain in completeness n = 2d sinθ Small source size: microfocussing is possible  very small samples can be investigated; reduction/elimination of gasket diffraction Refinement
Example of good data, synchrotron radiation In situ crystallisation study SIMU, DELU restraints Cell setting, space group All H-atoms could be located on difference Fourier maps Triclinic, P-1 a, b, c (Å) 6.7906(12) 7.3159(4) 15.8428(14) α, β, γ (°) 86.297(6) 78.924(11) 72.713(6) Multi-scan abs. correction T No. of measured, independent and 6429, 1533, 1294 observed [F > 4σ(F)] reflections No. of parameters and restraints R [F > 4σ(F)], wR (F2, all reflections) C H NO . 7(H O)
(Å-1) and completeness (%) Structure solution: Sir92 Refinement program: CRYSTALS Refinement
Example of good data on a large molecule, synchrotron radiation Compression study SIMU, DELU, DFIX restraints High pressure (1.0 GPa)
High pressure (1.0 GPa)
Cell setting, space group Orthorhombic, P2 2 2 a, b, c (Å) 15.9455(4) 21.0511(5) 23.8739(8) Multi-scan abs. correction T No. of measured, independent and 53048, 10032, 7534 observed [F > 4σ(F)] reflections No. of parameters and restraints R [F > 4σ(F)], wR (F2, all reflections) (Å-1) and completeness (%) C H CoN O P . 22 H O
Structure solution: (SHELXS) Refinement program: SHELXL Refinement
Example of good data on a large molecule, synchrotron radiation Ambient-pressure and temperature study SIMU, DELU, DFIX restraints Ambient pressure
Ambient pressure
Cell setting, space group Orthorhombic, P2 2 2 a, b, c (Å) 15.8260(9) 22.4438(13) 25.4429(16) Multi-scan abs. correction T No. of measured, independent and 101570, 26484, 19937 observed [F > 4σ(F)] reflections No. of parameters and restraints R [F > 4σ(F)], wR (F2, all reflections) (Å-1) and completeness (%) C H CoN O P . 23.5 H O
Structure solution: (SHELXS) Refinement program: SHELXL Refinement
Example of good data on a large molecule, synchrotron radiation Water ordering in channels at high pressure High pressure (1.0 GPa)
Ambient pressure
Electron density maps generated with shelXle, Fo-Fc @ 0.31 e-/Å3, Fo @ 0.98 e-/Å3 ShelXle: C. B. Hübschle et al. J. Appl. Cryst. (2011), 44, 1281-1284 Increasing data quality – Part II
Synchrotron radiation Useful properties of synchrotron radiation: Shorter wavelength  less absorption, more diffraction data and smaller diffraction angles Ag radiation (0.56087 Å) is now available as an air-cooled 30 W microsource (supplier: Incoatec) n = 2d sinθ  increase in data completeness, "cleaner" background Ag radiation in the lab
Comparative study on gabapentin heptahydrate, Incoatec Ag microsource vs. Mo sealed tube on a Bruker AXS Apex II diffractometer Ag-IµS, 90 µm beam Mo-sealed tube, 500 µm beam Exposure time (s/0.3°) <I> <I/σ> <Redundancy> <Completeness>/% R (I<2σ(I)) Number in parenthesis refer to the highest resolution shell (1.00 – 0.90 Å) Fore more information: http://www.incoatec.de/?id=101 Further Reading
Further Reading
High-Pressure Crystallography
Prof. Simon Parsons (Edinburgh) Dr. Heidrun Sowa (Göttingen) Dr. Jürgen Graf (Incoatec) Dr. Michael Ruf (Bruker) Question & Answer
Questions?
Type any questions you may have
into the Q&A panel, then click Send.
Thank you for joining!
A PDF copy of the slides and a link to the recording of this webinar will be emailed to you. For more Bruker SC-XRD webinars, visit

Source: https://bruker.cn/fileadmin/user_upload/8-PDF-Docs/X-rayDiffraction_ElementalAnalysis/SC-XRD/Webinars/Bruker_AXS_High-Pressure_Crystallography_Webinar_20111206.pdf