|
|---|
| IG Home | Newsletters | Group Meetings | Administration | Industrial Applications | Hints & Tips | Group Information | Site Map |
|
|---|
Reports - Spring Meeting, Canterbury 17th to 19th April 2007XRF sessions are reported elsewhere...
Co-crystals of Pharmaceutical Materials
The first talk in this well attended session, chaired by Anne Kavanagh and Roy Copley, was given by Bill Jones (Cambridge). His informative presentation on "Strategies for Designing and Making Cocrystals" started by describing a cocrystal as being made up of two or more non-ionized molecular species that are connected by non-covalent interactions. One potential use of cocrystals was illustrated by considering the moisture stability of cocrystals, with a cocrystal of caffeine being far more resistant to hydration than anhydrous caffeine itself. The talk then proceeded to look at the formation of cocrystals and the differences between solution and mechanical methods. Again using examples of caffeine with different formers, it was shown that the two techniques could yield either the same cocrystal product, different polymorphs of the same product or even different cocrystal stoichiometries. Liquid-assisted grinding was the next topic covered and this showed the benefits of adding a couple of drops of liquid to the grinding process. As well as offering a different route to known products, in the case of caffeine and succinic acid, it showed that products previously unobtainable could be prepared. In summary, grinding and liquid-assisted grinding were shown to increase the potential ease and ability to screen new forms. The value of crystal structure prediction was illustrated using the example of 10,11-dihydrocarbamazepine, where a lower energy structure than the one already found experimentally was calculated and subsequently obtained via grinding. Predicting cocrystal structures represents an even bigger challenge due to the added complexity but interestingly both carbamazepine and 10,11-dihydrocarbamazepine were found to form very similar cocrystals with acetic acid, a situation far removed from the solid-state differences found for the individual tricyclic compounds. The final topic was terahertz spectroscopy and its use to monitor the formation of cocrystals during grinding of phnazine and dicaroxylic acids. The presence of a liquid phase was shown to reduce the grinding time and to increases the crystallinity of the product. The next speaker was Keith Chadwick (Manchester) whose comprehensive and well illustrated presentation was entitled, "Nucleation and Phase Relationships for a Cocrystal". The talk described a detailed study on benzophenone (BZP) and diphenylamine (DPA) that form a cocrystal with a 1:1 stoichiometry. In 1933 the binary phase diagram of BZP and DPA was constructed and it showed it was possible to form two polymorphs of the cocrystal from the melt. Form I has a known crystal structure and form II is metastable. The two forms can be prepared from an equimolar melt, although the only way to obtain form II is to crash cool to 253K. Conversion from form II to form I can be monitored using powder diffraction. Raman microscopy suggests the presence of hydrogen bonded dimers in form II (as in form I) and a conformation of the individual BZP and DPA molecules similar to those in the crystal structures of the individual components. Grinding equimolar quantities of BZP and DPA produced form I and a more detailed understanding of the solid-state reaction was achieved when single crystals were placed in contact under a microscope and found to melt to give a metastable liquid phase that only recrystallised as the cocrystal when either agitated or seeded. It was postulated that this liquid phase was due to the existence of a metastable binary eutectic. The reaction can be observed macroscopically when powder samples of BZP and DPA are put in contact and a yellow liquid forms at the interface. The suggestion is that binary systems with a submerged eutectic below room temperature have the potential to form cocrystals by dry grinding. The remainder of the presentation looked at the cocrystallisation from solution. The ternary phase diagram for BZP, DPA and methanol was carefully constructed from gravimetric analysis and differential scanning calorimetry and was used to understand various crystallization methods. Nucleation studies were carried out not only in methanol but also toluene. The difference between the ideal and experimental solubilities is consistent with association of the two molecules in solution and the formation of hydrogen bonded dimers at high concentrations. This theory was supported by both IR and NMR experiments. Chris Frampton (Pharmorphix) gave the final talk, "Pharmaceutical Cocrystals: Panic to Panacea" which gave a valuable overview of the benefits and possible pitfalls of the pharmaceutical development of cocrystals. After a range of definitions of a cocrystal, it was suggested that the primary industrial driver was improving solid-state properties and that cocrystallisation was just one of the options open to achieve this. After citing a number of recent examples, it was explained that cocrystallisation can lead to improvements in numerous properties, such as morphology, apparent solubility, dissolution rate, chemical and mechanical stability, bulk density and hygroscopicity. The design of cocrystals can start with an analysis of the contacts or synthons of the active pharmaceutical ingredient (API) and interrogation of the Cambridge Structural Database to assess what interactions might be expected. Identifying suitable cocrystal formers needs to take into account numerous factors, including safety and toxicity, molecular weight, reactivity and pKa. The presentation then addressed the issue of sample preparation and the advantages of grinding and solvent-assisted grinding were again commented upon. In addition, the benefits of slurrying to produce cocrystals of caffeine with barbital, tartaric acid and succinic acid were outlined. In terms of high-throughput analysis of solid-state forms, the danger of just using XRPD to identify what can be isomorphous forms was explained, requiring an additional characterization technique. Raman spectroscopy has been developed into a combinatorial screening system which is used in conjunction with clustering algorithms to analyse the data produced. A description of work with carbamazepine and 70 solvents was used to illustrate the ability to identify new forms through clustering but also the danger of salt formation. Grinding and slurrying experiments of the relatively insoluble anti-malarial, artemisinin with cocrystal formers was used to warn of API reactivity, where a unique powder pattern was shown to be a reaction product. Early results of grinding and slurrying experiments of dipyridamole with aspirin were followed by a final comment on the topic of intellectual property, where it was noted that novel cocrystal forms can be patented in the same way as other solid forms if advantageous properties arise. The session was closed by thanking the speakers for providing both interesting and enlightening presentations on what is increasingly a topic of interest to the pharmaceutical industry. Roy Copley 
XRF/XRD joint session on thin films.
Thin Films and Coatings by XRF and XRD: an overview Tom began by saying XRD, XRF and XRR are universally acknowledged to be more direct and less model-dependant than optical methods. However, practical difficulties have limited their application in a production environment. XRF can analyse all film types and can determine elemental composition and obtain quantitative analysis. XRD can analyse crystalline and polycrystalline material whilst XRR can analyse all thin film types. Thin film analysis with X-rays became firmly recognised as a result of the 1954 Paper by I G Parratt dealing with surface studies using X-ray Reflectivity. Reflectivity measurements from oxide layers on copper had been obtained and simulated reflectivity curves produced. [It was shown that the peak to peak separation of the measured 'oscillations' was proportional to the wavelength of the X-rays and inversely proportional to the thickness of the layer]. In the semi-conductor industry, X-ray metrology for thin films is driven by the use of ultra-thin layers for new gate dielectrics, the use of localized stress to enhance channel mobility and the use of novel low-k dielectric materials. All these present metrology problems that are poorly served by classical optical methods. Tom told everyone about the Gartner Hype Cycle, devised by 'marketeers' to describe the path to technological maturity for a product! To begin with there is a technology trigger, followed by a peak of inflated expectation, followed by a trough of disillusionment, a slope of enlightenment and a plateau of productivity! Commercial GAXRD made an appearance in the 1980's - it gets rid of the 'high' backgrounds observed when thin films are analysed by traditional XRPD. HRXRD is an enabling technology in band-gap and strain engineering for III-V optoelectronics and is an industry standard metrology technique for III-V epitaxial structures. XRF from light elements is strongly absorbed so the detected intensity is independent of film thickness. Conversely, the intensity from heavy elements is dependant upon thickness. The substrate signal is attenuated by a thin film and is thus dependant upon film thickness. XRR metrology offers several advantages. It is independent of the optical parameters n and k, it can measure opaque metal films and is sensitive to interfacial layers and interfacial roughnesses. The technique is very good for low k dielectrics. On the negative side it generally uses a large spot size and needs a flat surface for analysis. Layer thickness information can be determined very accurately by XRR and these can be used as calibration standards for XRF. It leads the way for micro-focussed XRF calibrated by XRR. Tom concluded by saying that the semi-conductor industry provides the technological 'push' to analyse smaller and smaller features. In 2003, 90 nm features were of interest. It is predicted that in 2011+ features as small as 10 nm will need to be analysed. In line with these increasing demands there have been exciting developments over the last 10 years. Improvements in the focussing of X-ray beams combined with advances in detector technologies has reduced XRD, XRF and XRR probe sizes to a few hundreds of, and in some cases less than 100, microns. Advanced Solid-state X-ray Detector for the analysis of Thin-layered Structures Joachim began by saying structural analysis of thin layered structures involves different X-ray scattering techniques like rocking curve, reciprocal space mapping or reflectivity measurements. These methods require high angular resolution and a high dynamical intensity range as compared with standard measurements on polycrystalline materials. X-ray reflectivity measurements usually require parallel beam geometry and high resolution with crystal mono-chromators. Reflected intensities can be very high, up to 100x106 cps and so there is a requirement to use beam attenuators or reduce the X-ray source intensity. Historically, these measurements are time consuming to obtain and so what is now needed is an ultra-fast solution - 'one detector for all applications'. Joachim informed everyone about a recently introduced solid state detector for thin film analysis - based upon the very latest pixel detection technology. Development was through the Medipix 2 Project which was run by a large international consortium. The detector has 256 x 256 pixel resolution (65536 pixels with each pixel covering an area of 55 x 55 microns). It has 97% count linearity up to 100,000 cps per pixel or 25 million cps per strip. There is no need for a beam attenuator. The detector can operate in 0 D or 1 D and resolution can be further improved by reducing the active area from 6 to 1 mm2. XRR scans can be carried out in 0 D mode with an analysis time of typically 10 mins. Rocking curve scans typically take 10 to 15 minutes with traditional detectors. With this ultra-fast solution scans can be obtained in 10 or so seconds for the same measurement quality. The detector can be combined with all diffracted beam optics and due to its superior resolution and unmatched dynamic range it can be used for all types of X-ray scattering techniques. More information about this detector and its applications can be found in a Paper by Paul Fewster - J.App. Cryst 38(2005) 62. Up-To-Date XRD-Techniques for investigating ultra-thin films and ultra-small features The rapid progress in nanotechnology and nanomaterials has resulted in an increasing demand to characterize ultra thin films, nanostructure and organic thin films with laboratory X-ray instrumentation. Hugues went on to say that this data can be used as preliminary sample screening before synchrotron / neutron measurements or to monitor and improve sample growth. Typically, X-ray probe sizes are 50 to 100 microns. A omega-2-theta goniometer is used with a Eulerian cradle, a ceramic X-ray tube and 50 micron mono-capillary. Hugues explained how stress can be measured for copper wire structures using the Cu(331) cystallographic reflection. For the semi-conductor industry SiGe thin films have been studied and the strain in silicon films determined. XRF with a 100 micron spot has been used to investigate Cr(5nm)/Al(200nm)/MoNx(10nm)/CdTe(3 micron) structures. For very thin layers, the VANTEC-1 with a sealed gas filled tube and linear detector has been used. X-ray Probes of the layer and interface structure of nano-scale films for Opto-Electronics and Spintronics Brian began by saying that very high resolution XRD developed in the 1980's due to the need to measure composition and lattice strain in thin semi-conductor films. He showed examples of diffraction profiles taken from the early days and from recent times. A recent interest is the determination of the composition of AlxGa(1-x)As thin films on GaAs. Aluminium has been detected at the 1% level. Graded SiGe structures have also been investigated and Brian showed by use of reciprocal space maps, evidence for the asymmetric relaxation of test structures. He illustrated the application of laboratory-based grazing incidence in-plane diffraction to obtain independent twist and tilt data from GaN films grown on thin AlxGa1-xN buffers on sapphire. Brian went on to discuss grazing incidence X-ray reflectivity, This has now become a standard tool for the measurement of layer thickness and interface width in nanoscale thickness metal films for spintronic devices and data storage. Measurement of the off-specular scatter enables topological roughness to be distinguished from chemical intermixing across interfaces. He illustrated this by showing the measurements obtained for ultra-thin epitaxial layers of Co, Cu and Ni on silicon. Grazing incidence scattering becomes particularly powerful when combined with grazing incidence fluorescence measurements. This was illustrated in a study of the effect of Bi as a surfactant in the sputtering of smooth Fe50Mn50 exchange bias films. By a simple consideration of the Bi fluorescence signal or by more complex modelling, it has been shown that Bi does act as a surfactant under conditions of one or two monolayer coverage. Mark Farnworth
IG Keynote
Industrial Group Keynote Lecture Wed 18th April 2007 Robert began by saying that one of the most important needs of industry is to produce computer models for the behaviour of products and processes. The modelling of materials is an important element of this. For 10,000 years, material properties have been obtained from studies of chemical composition and of how the materials are affected by temperature and pressure. It was only in 1976 when a new 'tool' came on the scene - surface free energy. The origins of catalysts are borne out of the knowledge of surface free energy. For 100 nm structures the surface free energy is 1 % of the Total Free Energy - a relatively small contribution. Nanomaterials can be zero dimensional (quantum dots) or 1 or 2 or 3 dimensional. 'Surface shell' reconstruction is used for the study of nanomaterials. For nanomaterials the 'surface' can be a very significant part of the sample volume. Computer models are used. However, there needs to be experimental constraints otherwise the models produced are nonsense. X-ray Diffraction can be used to determine lattice parameters (peak positions), crystal structure (positions and intensities) and 'real' structure from the full XRD profile. 'Line' profiles can be used to separate size and strain broadening. They can also be used to determine defect density and dislocation density. Robert went on to say that there is a distortion of the lattice parameter as the sample reduces in size - towards 0.6x0.6x0.6 nm. Nanostructures can be analysed with Transmission Electron Microscopy (TEM), Field Emission Gun - Secondary Electron Microscopy (FEG-SEM) and synchrotron focussed X-ray beams. XRD analysis of the core structure and microstructure gives information about stacking fault density / types, size / strain and preferred orientation. Analysis of the 'surface shell' give information about the Pair Distribution Function (PDF) and the 'shell' can also be analysed with X-ray Photoelectron Spectroscopy (XPS) and described using computer models. Mark Farnworth
Diffraction from surfaces and Two Dimensional Crystallography
Structural Studies of Ordered Mesoporous Silica in Channelled Substrates Mesoporous silicates are produced when a surfactant is used as a template to produce a silica backbone. The surfactant is added to a sol-gel mix. Rod-like silica 'micelles' are produced. Mesoporous silica is used in areas such as electronics, catalysts, photonics and medical applications. Donna described research centred on establishing a relationship between etched silicon channels and their effect on mesoporous thin films. Using lithographically patterned (100) oriented substrates and a dip-coating method produced reproducible alignment of the pores across the substrate surface with the pores aligned parallel to the long side of channels. Etched silicon substrates or pillared silicon substrates are used and it was found that the latter produces poorer film qualities. Small angle X-ray diffraction (SAXS) and reflectivity (XRR) measurements showed strain effects when compared with simple films. A supercritical fluid inclusion technique is used to incorporate cobalt into the pores. High resolution Secondary Electron Microscopy was used for imaging. They discovered a high degree of pore order, large domain sizes, strong adherence to the channels, good thermal stability and that quality was dependant on the channel dimensions. The results indicate that the topographic structure can be used to engineer the placement and directionality of the pore structure. X-ray Characterisation of Nanomagnetic materials Tom began by saying that there is now a capability to produce patterned elements and nanostructures which allows the control of lateral dimensions. Patterning is achieved using Focussed Ion Beam (FIB) or lithography using photoresist. Arrays of nanoscale elements form the background for fundamental studies of magnetism which are important in realising both magnetic memory and high-density storage media. Several techniques are used to analyse small groups of nanostructures or small areas of nanostructured materials. The problem is that small volume / small area analysis may not give information that is representative of the sample as a whole. However, by imaging arrays of nanoscale material in reciprocal space this problem can be overcome. Averaged array parameters can be determined. Rocking curve analysis is a good technique because the presence of a repeat pattern in the plane results in satellites of coherently scattered intensity equi-spaced on either side of the 'specular ridge'. The spacing of these satellites gives the periodicity of the nanostructures, the width of satellites the coherence and the 'envelope' of the scattered intensity the shape of nanostructures. The use of resonant magnetic scattering to probe the element specific magnetisation processes in arrays was presented. With magnetism, there is a need to know the scaling behaviours, anisotropies and the structure of domains. One avenue of research is to link the magnetic peak envelope obtained from Rocking Curve analysis to the magnetic size and shape and to compare with the physical size and shape. Tom clearly showed how X-ray data is providing new insights into the magnetic properties of these important systems. SAXS and GI applications using a novel modular laboratory system Peter's presentation covered the use of Grazing Incidence Small Angle X-ray Scattering (GISAXS) and SAXS for the nanostructure analysis of bulk materials, nanoparticle solutions and thin solid films. A high brilliance 50 W microfocus beam (Hecus X-ray Systems) is used with 1D and 2D detectors. Key issues are the micro-source size (point focussing), the small beam footprint at the detector and the detector sensitivity. GISAXS is used in reflection mode and a gas or liquid environment can be used. Peter concluded by say that this 'low power' system offers a performance that compares with high power rotating anode systems and small synchrotron sources. Mark Farnworth
One Hundred and One Ways to Prepare an XRPD Sample
TEXT ?? A Standardless future for Quantitative XRPD Speakers - Left to Right: Steve Norval (Chair), Rob Hill, Chris Gilmore, Paul O'Meara. Random Mounts and Reproducibility? The Key to Standardless QPA Gordon began by saying flat-surface mounts of materials such as clays can be prepared that produce close-to-random diffraction patterns when recorded by a curved position sensitive detector (PSD) with a large solid angle. This means that QPA of materials prone to preferred orientation (such as clays) can become an easy routine. The 'standard' whole pattern need only be run once and placed in the databank. Only a flux correction needs to be applied to the data for the day of measurement. Gordon showed the results of QPA of peat obtained from Keri, Greece. For materials with well established structures, calculated pattern intensities can be compared with whole pattern results from the PSD to verify the randomness. In the case of unknown or poorly-defined structures the XRD-PSD method can still be used by simply comparing with ('real' material) experimental whole patterns in the databank, unlike an approach that requires calculated intensities from known structures. Thus, accuracy is controlled by 'best-matching' to the standards available. In effect, standardless QPA by XRD-PSD is simply a pattern comparison procedure. Quantitative Analysis of Mixtures using High Throughput Instrumentation without the use of standards Christopher described the PolySNAP computer program which uses the full powder diffraction pattern to carry out the quantitative analysis of mixtures from powder diffraction data. It does not, in general, use calibration standards and is capable of achieving an accuracy of 1% under favourable conditions. Experiments involving 2- and 3- phase mixtures of organic samples were carried out on a Bruker D8 GADDS system and on a Bruker D4 and the TOPAS software as a reference. Following data collection, the next step is 'visualisation'. Every pattern is compared with every other pattern The correlation matrix is converted into a 'distance' matrix and similarities discerned by the use of dentrograms. The final step is data processing in which quantitative analysis is obtained for all the input data. It was found that for a typical 2-phase system e.g. paracetamol and lactose PolySNAP gave better results than TOPAS. However, TOPAS gave better results for 3-phase mixtures. Christopher concluded by saying that PolySNAP can give mixture compositions, in a high throughput environment, to within 10%, quickly and without user intervention. Rietveld Analysis and its applications to the cement industry XRD and Rietveld analysis is the most powerful method for quantitative phase analysis. The Rietveld methodology has been known for some 40 years and it was the lack of computing power that prevented its breakthrough until relatively recent times. The method uses the full XRD pattern for analysis - all data points are used. The TOPAS software was used since there is then no need for standards. In the cement industry quality control of raw materials is required. Chemical and physical properties of the raw materials are determined by the phase composition rather than the elemental analysis. Rob showed how Rietveld analysis has been used to analyse Alite - caused by clinker dust in the pre-heated. Reitveld analysis has been used to determine the true clinker composition. Applications of Rietveld in Aluminium production Paul explained that the production of aluminium starts with the mining of its principle ore Bauxite. To convert the Bauxite to aluminium, it is washed, ground and dissolved in caustic soda at high pressure and temperature - the Bayer process. The resulting liquor contains a solution of sodium aluminate and undissolved bauxite residues known as 'red mud'. This needs to be disposed of carefully, usually on the seabed or as landfill. Bauxite contains three principle aluminium bearing constituents:- Gibbsite (an aluminium hydroxide) and Bohmite and Diaspore which are aluminium oxide hydroxides. The temperature and pressure in the digester are determined by the relative abundance of each of these.The sodium aluminate solution is pumped into a 'precipitator' where the alumina particles are separated. These are then heated at 1100°C to produce the white, pure alumina powder. Single line methods of quantitative analysis involve the use of Relative Intensity Ratio's (RIR's), internal standard, straight line method and matrix flushing. Rietveld involves whole pattern fitting and has been found to give very accurate results. The method is good when preferred orientation is known to be present, when standards are unavailable, when 'occupancy' varies and when peaks overlap. Paul concluded by saying the Rietveld method is used to control the manufacturing process and reduce the environmental impact. Mark Farnworth
|
Industrial Group of the BCA
(a 1.4Mb PDF file)
