Be it proteins, new superconductors, polymers or ferro-magnetic compounds, fundamental research is laying foundations for possible technical applications as well as industrial exploitation.

At the ESRF, industrial companies can be involved in two different ways, depending on the type of research they are interested in. The first concerns beam time allocation for non-proprietary and fully-published research. The procedure is the same as for public laboratories: the applications are peer reviewed by review committees and, if accepted, beam time is free of charge. The second concerns the proprietary research, for which a fee is charged and where results are kept confidential and therefore cannot be presented here. 

The three main fields of applied and industrial research are:
- pharmaceuticals, with a dozen pharmaceutical companies, mostly for diffraction data collection on crystallised protein-ligand complexes;
- materials, with more than fifteen companies working in such different areas as chemistry, cements, cosmetics, glass, polymers, petroleum, metallurgy, etc. These companies are strongly interested in the possibilities of microanalysis, real-time monitoring, analyses of thick samples using hard X-rays as well as new X-ray imaging techniques (microtomography, phase contrast);
- microelectronics, with about ten companies. This field will soon benefit from a dedicated experimental station for trace detection on silicon wafers using the Total X-ray Reflection Fluorescence (TXRF) technique.

In this section, very interesting applications in the field of environmental research are also shown.

The use of synchrotron radiation for the characterisation of materials is certain to play a crucial role in the development of new materials in the close future. It is a policy of the ESRF that European industry should share the benefit of its facilities.



High-resolution strain scanning


Residual stresses can enhance or degrade the performance and service lifetimes of mechanical components subjected to fatigue stresses. The measurement of steep gradient residual stress fields in such components, particularly near surfaces and sharp features, is therefore an important, but technically difficult, engineering requirement.

Traditional X-ray and neutron diffraction techniques are both widely used for such engineering and research applications, but both have important limitations. Laboratory X-ray diffraction can give reasonable spatial resolution, but it only accesses the surface regions; neutron experiments can probe subsurface regions, but only at the cost of poor spatial resolution. Over the past year, experiments carried out at BM16 with P.J. Webster's group from the University of Salford have shown that the high energy beam produced by the third generation synchrotron source, coupled with accurate diffractometry and parallel beam optics, overcomes these limitations. The synchrotron beam size and tunable energy permit very high spatial resolution with controllable depth penetration. Furthermore, the parallel beam optics allow the simultaneous optimisation of both the spatial and angular resolution, leading to high precision in the measurement of the strain gradients and magnitudes respectively.

Figure 118 shows the distribution of normal strain against depth in an aluminium-alloy aeroengine turbine blade which has undergone surface treatment to provide a protective compressive stress at the surface. Small scratches that occur during service at the surface are not able to propagate and become cracks because the compressive in-plane stress tends to close up the damage. The small beam size allows the resolution of features on the scale of tens of microns, with a strain sensitivity of better than 10-5. The profile of a diffraction peak at different depths in the sample is shown in Figure 119; the excellent angular resolution makes detailed peak shape analysis possible, allowing the elucidation of the relationship between macroscopic strains and microstructure. The broad peaks near the surface indicate, for example, that marked deformation of the surface layers has resulted from the surface treatment.

High resolution is crucial in cases such as the example above, where the phenomena of interest take place over a length scale of the order of 100 microns. Such information allows the development of computer stress models, and can help in determining treatment conditions which produce the optimal residual stress distribution for component performance. These promising initial results indicate that synchrotron strain scanning should become the method of choice for strain mapping in light components with rapidly varying strain fields.


P. J. Webster (a), G. B. M. Vaughan (b), G. Mills (a), and W. P. Kang (a), to be published in Materials Science Forum.

(a) The Telford Research Institute for Structures and Materials Engineering, University of Salford (UK)
(b) ESRF



Micro-analysis of wood by diffraction and fluorescence techniques

The analysis of the inorganic contents in wood is of great interest for several different reasons. As the tree represents a large part of all the biomaterial synthesised on this planet, the study of its mineral content and metabolism is important for the understanding of the ecology and circulation of many mineral nutrients and for the environmental impact of anthropogenic generated inorganic components. The ability to analyse the distribution and pathways for mineral supply in the tree will facilitate the understanding of the metabolism and the biochemical processes in which metal atoms play an important role.

Tree fibres and wood also constitute the basic material for the paper and pulp industry which is one of the most important industrial sectors in the world today, and the metal content in the fibres has a great impact on the chemical processing of fibres. Particularly for the process of pulp bleaching, the metal ion content plays a vital role. In recent years there has been a rapid development of new processes for pulp bleaching, so called TCF (Totally Chlorine Free) processes where hydrogen peroxide has been used as a bleaching agent. The effects of different metal ions present in pulp during hydrogen peroxide bleaching are not fully understood, but metals such as Fe, Mn and Cu are considered to have a negative effect while the presence of Mg and Ca has a positive effect. A lot of investigations have been carried out in order to understand the impact of different chemical and mechanical processes on the metal content and distribution in the pulp.

However, there is still a lack of knowledge on the primary distribution of metal ions in the fibre itself. One reason for this is the inability to apply conventional micro-analytical methods, like Scanning Electron Microscopy, directly on wood tissues in a quantitative way. In recent years, micro-beam X-ray Fluorescence has been applied for the study of trace element distribution in wood, on a micro-scale. This technique has proven to be very suitable for this type of material, and accurate and reproducible quantitative results have been achieved. However, the only instrument able to provide us with detailed information of the trace elemental distribution at various parts and components in individual fibres is the Microfocus beamline (ID13) at the ESRF. This instrument has a flux density sufficient for trace element analysis on a microscale level down to PPB (Parts Per Billion) levels, but also the ability to simultaneously generate structural information on this level through so called micro-diffraction technique.

A series of samples of wood tissues has been studied by means of microdiffraction and microfluorescence at the Microfocus beamline. The aim of the study was to investigate the degree of crystallinity, and preferential orientation of the cellulose molecules in the tissue at a microscopic level and in correlation to the distribution of trace metals within the tissue. The samples, which consisted of sections of wood tissue, transversal, radial and longitudinal cuts, had been chemically treated in different ways (H2SO4, EDTA, etc.) in order to affect the metal distributions within the tissue.

Preliminary evaluation shows strong variation of the trace element distribution and significant differences between different traces. The distribution and average concentration level of the traces were strongly affected by the chemical treatment. Figure 120 shows the distribution of manganese (left) over a few cells at the border between early wood and late wood from spruce. The figure at the right shows the density variation, as recorded from the scattered radiation, over the same area.

The distribution of trace elements mainly goes along with the cell walls; however, some of the elements seem to appear in localised areas rather then being evenly distributed in the organic material. For the transversal cuts (such as Figure 120) the fibre-axis is perpendicular to the image plane. In these cases, the diffracted radiation operates as an indicator for the fibre-axis orientation. While scanning over an individual cell wall, there is a small but significant indication of a shift in the fibre orientation.

In order to understand the metabolism and translocation of the trace elements, it is necessary to determine exactly where in the tissue these elements appear and what the micro-distribution looks like. Figure 121 shows an example of the distribution of calcium and titanium around a "transportation" channel in the tissue. While calcium seems to be related to the organic part, titanium appears in clusters (small inorganic particles) which seem to have accumulated around the channel. In same cases the diffracted radiation has indicated the presence of small inorganic crystals (titanium-oxide, quartz, etc).

Another example of the importance of trace element analysis in plant tissue, beside the TCF-problem, is the recent discovery of manganese accumulations in fungus-infected (Heterobasidium annosum) areas in spruce trees. Spruce and pine trees are affected worldwide by this fungus that accumulate manganese in and around infection zones.

The economical and environmental impact only from these two examples are enormous and the ability to study the intricate relation between polymers and heavy metals are of great importance, not only for paper and wood industry, but for a large part of the polymeric industry in general and also for the development of new material based on polymers.


A. Rindby (a) and P. Engström (a, b), to be published.

(a) Chalmers University of Technology, Göteborg (Sweden)
(b) ESRF




Trace element analysis on Si wafers

Worldwide, Integrated Circuit (IC) manufacturers are rapidly moving toward technologies aiming at 0.18 micron integration with even faster growth of device complexity.

These future technologies rely on key parameters such as film homogeneity and absence of surface contamination. The detection of contaminants at a level of 1010 at./cm2 (equivalent to 10-5 monolayers) is no longer adequate, whilst short cycle time and mapping of 300 mm wafer is increasingly needed.

Traditionally, contamination detection is performed with Total Reflection X-Ray Fluorescence excited by rotating anode X-ray tubes. These sources are at the limit of their performance and cannot be improved further. For this reason, SR-TXRF (Synchrotron Radiation excited Total Reflection X-Ray Fluorescence analysis) is indicated as a viable solution for the needs of IC manufacturing industries of the next decade.

A pilot test on the possibilities offered in this field by the ESRF has been conducted at ID32 where a basic optical set-up consisting of an Si(111) double-crystal monochromator and collimating slits has been used. In opposition to conventional SR-TXRF geometry, where samples are mounted vertically to increase the angular acceptance of the detector that has to be aligned along the horizontal polarisation direction, samples were mounted horizontally. A high-purity Ge detector with reduced dead layer was placed at about 50 mm from the impinging point of the beam on the wafer with a Teflon collimator of 5 mm diameter. Measurements were performed in air under a laminar flow hood providing an environment better than class 100.

Ca, Ti, Cr, Fe, Ni, Cu and Zn contaminants on a pre-calibrated wafer were analysed and Lower Limits of Detectability (LLD) were determined. For the Ni atoms a LLD of 2 x 108 at./cm2 has been obtained, a result comparable with the best ones obtained on a second generation synchrotron radiation source using multilayers, focusing optics and vertical sample geometry. A remarkable feature of this pilot test is that the elastic peak contribution is of the same order of magnitude as fluorescence, due to about 1011 at./cm2 as shown in Figure 122. Considering a penetration depth of the radiation of 30 Å into the wafer, this means that from one impurity atom we obtained the same signal as the one given by 105 atoms of silicon substrate; the measured detectability limit therefore translates in the possibility of seeing an impurity atom out of ten million substrate atoms.

The obtained results are very promising for synchrotron radiation trace element analysis: using appropriate optics and detectors we estimate that it would be possible to lower the LLD to about 106 at./cm2. Since however such low levels of detectability are not of immediate interest to IC manufacturing industries, these improved performances will be translated instead into lower cycle time mapping with an affordable measuring time.


F. Comin (a), L. Ortega (a), V. Formoso (a), A. Stierle (a), to be published.

(a) ESRF




In situ X-ray diffraction studies of electrochemical lithium intercalation in the Li-Mn-O and Li-Co-O systems


These systems have been widely studied as materials for the positive electrodes of rechargeable lithium batteries. In situ X-ray diffraction has been shown to be extremely useful in characterising structural modifications of insertion compounds over a wide range of intercalate compositions. This technique allows a real-time characterisation of the insertion compounds without the drawbacks of the ex situ method, which includes a need for a large number of samples, the risk of atmospheric contamination and the adverse effects of self-discharge.

X-ray diffraction experiments have been performed during lithium intercalation and deintercalation on the Li-Mn-O and Li-Co-O systems, using the high-resolution powder diffraction beamline at the ESRF (BM16). Diffraction measurements were made at an X-ray energy of 30 keV in transmission on complete, working, 1 mm-thick Li-ion batteries using plastic lithium electrolyte and built with the Bellcore technology (Figure 123). The batteries were cycled using either galvanostatic or potentiodynamic modes, controlled by the MacPile system. Data were collected using the nine channel crystal analyser stage developed for the beamline to provide excellent resolution with good statistics.

In the Li-Mn-O system, lithium extraction from the cubic spinel LiMn2O4 was investigated, which occurs in the 3­5 V range. The results confirm the two-phase reaction occurring for extraction of ca. 0.5 Li. In addition, a transition appearing at the beginning of the deintercalation, which had been little studied previously, was investigated in detail. The main results are as follows:

This redox step is observed at 3.95 V on charge, and 3.3 V on discharge.

It has very slow kinetics, as shown by the evolution of diffraction line widths with time.

It seems to be a two-phase reaction, showing the existence of a previously uncharacterised intermediate LixMn2O4 phase.

LiCoO2 is one of the most promising high-energy density positive electrode materials for commercial applications. Whilst showing many favourable attributes including high energy density, good power rates, low self-discharge and excellent cycle life, there still exist many unanswered questions as to the electrochemical performance of this layered material. One of these elusive questions is the existence and structure of the end member CoO2 upon complete lithium deintercalation. X-ray diffraction data collected during intercalation and deintercalation have allowed us to follow precisely all of the structural changes (Figures 124 and 125) and to characterise the CoO2 phase.


M. Anne (a, d), Y. Chabre (b), F. Le Cras (a), R. Palacin (c), L. Seguin (c), P. Strobel (a), J. M. Tarascon (c), G. Vaughan (d), EPDIC-5, Parma, 25-28 May 1997.

(a) Laboratoire de Cristallographie, CNRS, Grenoble (France)
(b) Lab. de Spectrométrie Physique, UJF and CNRS, Grenoble (France)
(c) LRCS, Université de Picardie and CNRS, Amiens (France)
(d) ESRF




In situ EXAFS investigation of the catalyst role of metallic nanoparticles

Gaseous atmosphere sensing has become an industrial concern to control and design specific manufacturing processes, to produce safe domestic hardware on behalf of environmental needs. In case of devices using solid-gas interactions, the reaction mechanism frequently implies an electrical signal alteration of the sensing element. A sensor can then be made of a combination of a small sensing element, which can save electronic equipment for electrical signal processing. As heating is generally necessary to reach the sensor working conditions, extensive research efforts have been made in the last past years to obtain sensing elements saving heating electrical power. Processing sensing material as thin film can provide an answer for this need and many research groups are now involved in this field.

SnO2 is frequently used as an active element for gas detection. By dispersing a low concentration of fine metallic particles (Pd, Pt, Cu...) on the surface of the SnO2 grains, the sensitivity and selectivity can be improved (Figure 126). For carbon monoxide, the interaction phenomena with the composite can be described using a simple model where the desorption of reducing species induces a modification in the local environment of the metallic element and its oxidation state, leading to an electronic charge transfer between the clusters and the SnO2 support. Another model describes the metallic particles as catalysts which activate the interaction between gas and SnO2 surface. These models nevertheless need to be verified.

We have used X-ray absorption to perform an in situ study of the evolution of the local environment and of the oxidation state of metallic elements added in thin films of Pt doped tin oxide deposited on an SiO2 substrate by a Pyrosol process. A homogeneous dispersion of metallic aggregates (3-5 nm) within the film is obtained by this process. Figure 127 shows the TEM picture for a Pt 12% doped SnO2 film. The in situ EXAFS experiments have been made in a specific reaction cell, heated up to 400 °C and put in contact with air or with a polluting gaseous mixture containing small amounts of reducing gas (Air + 300ppm CO, N2 + 900ppm CO). Such a diluted cluster system inside a matrix can only be investigated by a high flux and high brilliance synchrotron source like the ESRF. The experiment has been done at the ESRF on the beamline BM32 in fluorescence mode using the new detector at the Pt LIII edge.

Figure 128 shows the evolution of the white line under various working conditions. We can see that the platinum exhibits two states, either oxidised or reduced, depending on the gas. In the case of pure air (respectively N2+CO) the XANES and EXAFS spectra show that the platinum is either in an oxidised state (Pt-O bonding) or in a reduced state (Pt-X bonding, X = CO or Pt). The change in the white line is particularly pronounced for the smallest particle concentration, i.e. the finest particles.

A study of the kinetics of the oxidation/reduction process was also performed using the evolution of the white line as a function of gas, temperature and time. The 6% Pt sample was thermally cycled between 400 °C and 25 °C, alternating pure air or CO in nitrogen at each temperature. This analysis has shown the ability of the finest Pt particles (3% and 6%) to change the oxidation state in contact with air + 300 ppm CO and to correlate this behaviour with the good sensitivity to CO (see Figure 126).

A temperature-dependent reaction mechanism could be proposed. At an elevated temperature, platinum oxide reacts with CO and changes to metallic Pt, which itself reacts with chemisorbed oxygen on SnO2-x:

PtO2 + 2CO Pt + 2CO2 (1)
and Pt + SnO2-x PtO2 + SnO2-x' (2)

Oxygen in gaseous then reacts with the surface of SnO2 to restore the stoichiometry:

Ogas + SnO2-x' SnO2-x (3)

In such a case, no significant change in electrical properties is observed. However at low temperature the reaction (2) is slow. Consequently platinum is not re-oxidised. At about 100 °C a great proportion of Pt is reduced and reacts as a donor of electrons, leading to a conductance peak at low temperature (see Figure 126). This mechanism can be involved at all temperatures under N2/CO, due to the absence of gaseous oxygen.


M. Gaidi (a), M. Labeau (a), B. Chenevier (a) and J.-L. Hazemann (b), to be published.

(a) Laboratoire Matériaux et Génie Physique, INPG, CNRS, Grenoble (France)
(b) LGIT, CNRS and Laboratoire de Cristallographie CNRS Grenoble (France)




Crystallinity of plastic containers using microfocus X-ray diffraction

Physical properties, such as strength and gas-transport properties, of artefacts made from poly(ethylene terephthalate) (PET) depend on the degree of polymer orientation and crystallinity, and its variation in the artefact. The manufacture of PET containers involves first the injection moulding of an unoriented amorphous preform followed by blowing the preform into a cold shaped mould to form the container. This process is carried out via various thermal routes, but for all routes there is a significant variation in the temperature profile through the wall of the preform at the moment of extension in the blowing zone. In addition, the contour shape of the container and the complex blowing process itself produces considerable variation in wall thickness and direction of stretching at different positions in the container. The polymer therefore experiences a wide variation in draw temperature, draw ratio and draw direction at different positions of the container and at different depths in the wall, resulting in a wide variation of orientation and crystallinity in the local polymer structure. In order to understand the mechanical and diffusion properties of the container, it is necessary to be able to characterise how this structure can vary.

In an effort to improve efficiency and quality control in the processing of bulk polymers in the manufacture of products such as containers from PET (which have a high performance specification), a preliminary study of the dependence of polymer orientation and crystallinity on manufacturing conditions has been undertaken by a team led by ICI scientists. In order to probe variations in structure through the container wall with sufficient spatial resolution, and thus allow the establishment of the connections between local properties and the manufacturing process, investigations were carried out at the ESRF on the Microfocus beamline ID13. A beam of 2.3 µm diameter (fwhm) with a divergence of ~ 3 mrad was obtained at the exit of tapered glass capillary focusing optics. Also crucial for the experiment was the availability of a computer-controlled stage which allowed the specimen to be tracked in two directions perpendicular to the beam with a resolution of 0.1 µm. Diffraction data were recorded every 30 µm across the full width of the sample (about 1mm thick) on a Photonic Science CCD X-ray detector which enables data for d-spacings in the range 15 to 1,5 Å to be recorded with an exposure time of 20 seconds.

Selected frames from the longitudinally sliced samples are displayed in Figure 129. The patterns show a marked variation in crystallinity through the thickness of the wall. Particularly at the inside edge there are strong oriented crystalline reflections superimposed on a fairly isotropic amorphous halo. The intensification of the crystal reflections around the equator is consistent with preferential chain alignment along the longitudinal axis of the container. The intensity of the crystal reflections diminishes towards the centre of the wall thickness leaving an almost isotropic amorphous diffraction halo. Nearer the outer edge of the wall there is an intensification of the halo on the equator indicating an increasing degree of orientation in the disordered chains. The oriented crystalline pattern is not that normally seen from highly oriented PET fibres. The strongest features in the full diffraction pattern are the (100) and (1-10) reflections and there is an absence of any (010) reflections near the equator. This would be consistent with chain alignment in the longitudinal axis but with a preference for the (100) crystal planes to lie parallel to the plane of the bottle wall.

The general trends can be rationalised in terms of a variation in temperature across the wall thickness at the moment of extension. At the outer edge of the wall, where the temperature is lowest, the degree of extension has produced a degree of orientation in the disordered chains but not enough orientation to enable them to crystallise. Towards the centre the higher temperature has allowed a degree of relaxation in the orientation of the amorphous chains. On the inside edge there is a long enough residence at a higher temperature to enable crystallinity to develop from the partially oriented amorphous matrix.

These studies demonstrate that an intense micrometre-sized X-ray beam can be used to investigate local textural variations in technologically important artefacts fabricated from polymer materials. Confidence is also added to the future application of these procedures in the investigation of structures as small as a few micrometres, e.g. at the interfaces in laminated materials.


C. Martin (a), A. Mahendrasingam (a), W. Fuller (a), J.L. Harvie (b), D.J. Blundell (c), J. Whitehead (c), R.J. Oldman (c), C. Riekel (d) and P. Engstrom (d), J. Synchrotron Radiation (1997) 4, 223-227.

(a) Department of Physics, Keele University (UK)
(b) ICI Polyesters, Wilton (UK)
(c) ICI Wilton Research and Technology Centre (UK)
(d) ESRF




Protein crystallography for pharmaceutical research

Structural biology has developed rapidly during the last decade, not only for its fundamental academic interest in molecular and cell biology, but also for biotechnological applications in pharmaceutical drug design and in the development of enzymes used in industrial processes. Synchrotron radiation has become a major tool for protein structure determinations since today about one third of the diffraction data are collected with this source.

Protein crystallography is one of the main activities at the ESRF. The collaboration with the EMBL outstation in Grenoble has played an important role in the development of beamlines, the biological infrastructure and the user support. By the end of 1998 there will be five experimental stations fully devoted to protein crystallography and another three partially devoted. They are equipped with high-performance detectors, in particular with CCD array detectors with fast read-out systems. But the most important is that the beamlines benefit from the qualities of the ESRF source that enable breakthroughs in various directions. The high brilliance combined with micro-analysis techniques make the collection of data on crystals as small as 103 µm3 possible, the low divergence allows the diffraction spots to be separated even for unit-cell parameters exceeding 100 nm (which is often the case for virus crystals), the optics elements provide high-energy resolution which is necessary for phasing the reflections by the MAD technique (multiple anomalous diffraction) and the pulsed nature of the beam permits nanosecond time-resolved data collections to follow, for example, rapid structural modifications such as the release of carbon monoxide in myoglobin. A review of the various possibilities offered by the ESRF in structural biology is given in the section on "Life Sciences", in these Highlights.

As an example of typical work we will mention the structure determination of a N-myristoyl transferase (NMT) led by a team from Zeneca Pharmaceuticals. NMT is a protein which catalyses the transfer of the fatty acid myristate from myristoyl CoA to the N-terminal glycine of substrate proteins. The enzyme in this study has been produced by Candida albicans, a yeast responsible for the majority of systemic fungal infections in immuno-compromised humans. NMT is also an essential enzyme for vegetative growth and should therefore constitute an attractive target for potential antifungal pharmaceuticals.

Various data collections on the native compound, on heavy-atom derivatives and on the Se-Met protein have been necessary in order to solve the structure at 0.28 nm resolution. The MAD data collection was performed at the ESRF (BM14 beamline), on the same crystal, at three different wavelengths around the Se absorption edge.

The protein fold displays on one face of the protein a long, curved, relatively uncharged groove, at the centre of which is a deep pocket. From the other side of the protein, the C-terminus is directed towards the pocket with the C-terminal leucine forming the floor of the pocket (Figure 130). The electrostatic character of the surface of the protein in and near the groove and pocket suggests that the groove and the pocket are the sites of substrate binding and the floor is the catalytic centre.


S. Weston (a), R. Camble (a), J. Colls (a), G. Rosenbrock (a), I. Taylor (a), M. Egerton (a), A. Tucker (a), A. Tunnicliffe (a), A. Mistry (a), F. Mancia (b), E. de la Fortelle (b), J. Irwin (b), G. Bricogne (b), R. Pauptit (a), submitted to Structure.

(a) Zeneca Pharmaceuticals, Macclesfield (UK)
(b) MRC, Cambridge (UK)