Meeting Report - May 2008, BGS, Keyworth.
Sponsored by:

The meeting programme is on another page.

Report on XRD meeting following on the next day.

Delegate Photograph

Morning Session

Welcome
Dave Taylor, BCA Industrial Group

Dave provided the welcome for the day from the Industrial Group of the BCA for this second joint XRF meeting with the RSC Atomic Spectroscopy Group. He thanked all the sponsors for their kind contributions, the BGS for use of the excellent facilities and Mark Ingham and his collegues for their help with set up and local arrangements.

Chair, Morning Session
Andy Scothern, RSC.

Left to Right: Nick Marsh, John Boyle, Margaret West, John Mansell and Andy Scothern (Chair).

The first talk of the morning was by Nick Marsh (Leicester University), on the preparation and presentation of mineral samples for XRF analysis. The usual requirement is a fine powder (but if you have to crush the sample, is the technique nondestructive?) and there are several stages to this. Drying should be undertaken with care: volatile elements such as mercury can be lost, as can essential water from hydrated minerals. Crushing and milling is very dependent on mineral properties such as hardness, brittleness, cleavage and density, all but the last of which can be anisotropic. The final stage is to convert the powder into a form suitable for analysis. This may be a loose powder, a pressed pellet, a fused bead, even a dust on a filter. These all have their particular characteristics. It is always worth thinking about critical thickness (or thinness, in the case of dust on a filter). Mineralogical effects - for example, when one grain type is very different in its absorption of X-rays to others - can be very significant in pressed pellet samples. A fused bead will get round this difficulty. All in all, it was an interesting and thought-provoking talk.

The second speaker of the session was John Mansell (Omya UK Ltd), who spoke about the analysis of limestones by EDXRF. This study included elements from major to trace levels, and included samples from a large number of quarries, so that the provenance of unknown samples could be established. The XRF method was very much quicker than previous methods (ICP or DCP-based) which required preliminary dissolution of the sample. Not quite all elements could be analysed - in particular, the sensitivity needed for Cd is really beyond the instrument, and Ba is difficult because of line overlaps.

Margaret West (West Analytical Solutions Ltd) spoke next, describing the background to a new BSI standard for glass-making sands. She described the parameters of the specification for a glass-making sand, and how the new version of BS2975 had been put together by the Society for Glass Technology. This is in two parts: sampling and testing; and chemical analysis. A wide range of physical properties is specified, which need to be tested. There is a detailed description of a sampling protocol, because sampling (surprise, surprise!) is one of the trickiest problems in the analysis. Grinding media are discussed, and the preparation of samples for presentation to the XRF. For some purposes, wet chemical analysis (inevitably involving HF digestions) is unavoidable, but mostly XRF is a lot easier. Standards and reference materials are, of course, needed, and a range has been made available by the Society for Glass Technology in conjunction with BAS Ltd.

The final talk in the morning session was by John Boyle (University of Liverpool), who described the analysis of lake sediment cores by EDXRF. These samples can be very variable, and a general approach is needed. Many elements must be determined, often in a small sample (less than 1 g - the cores may have to be divided among many researchers). Particle size is rarely a problem because it is normally very small anyway. Matrix effects are definitely a problem, but one for which corrections can usually be made. Background measurements can be awkward, as the peaks are close together, but there are ways around this difficulty. Line overlaps are another problem, but again one which can be overcome with care.

This closed the morning session, after which we enjoyed an excellent buffet lunch while we looked at the manufacturers' exhibits.

David Beveridge
Harman Technology limited

Afternoon Sessions

Chair, Afternoon Session 1
Ros Schwarz.(Oxford Instruments Analytical)

Left to Right: Elke Adriaenssenns, Owen Butler, Ros Schwarz (Chair), Neil Eatherington and Graham Martin.

Neil Eatherington, (British Geological Survey, Nottingham)
When Sample preparation is not an option.

Neil acknowledged that for the majority of applications XRFS requires sample preparation to get the best from the technique, but what happens when preparation is not an option? In-situ samples together with historic or valuable samples require sensitive handling and cannot be destroyed. The problems with in-situ analysis stem from the moisture content of samples together with particle size and mineralogical effects.

Drying of samples was discussed with attention drawn to the problems of using a microwave oven (elevated temperatures of up to 200°C) and the loss of volatile analytes from environmental and soil samples. Conventional drying at 40°C was ruled out because of the time taken to achieve a dry sample (<5% moisture) measured in days rather than minutes.

Several case studies were presented featuring marine sediment cores, panned stream concentrates and in-situ analysis. Data was compared between handheld XRFS costing ~ £35k and specialist equipment such as an ITRAX core scanner costing ~£200k with the conclusion that the handheld was useful as a screening tool and has some success at quantification. Overall the correlations between fully prepared samples and in-situ samples were generally good.

Prof Graham Martin, (Victoria and Albert Museum, London)
Heritage and XRFS -unlikely companions?

This was a different presentation where Graham avoided posting the usual spectra, graphs and data but instead concentrated on the application of ED-XRF within the heritage sector. Graham began with an overview of the Victoria and Albert Museum and its associated institutes and explained that a large challenge was communicating the applications to his colleagues with arts based backgrounds.

The sheer scale and diversity of the exhibits and potential sampling sites were discussed together with the fact that invasive and chemical techniques were offered as a last resort but usually not at all. The analytical equipment used was presented from the Tracor DuBois pigment analyser from 1983 to the present equipment of the Bruker ArtTax, a portable (in 3 cases), open-sourced analyser mounted on a highly flexible mechanical arm that requires high operator skill levels.

The types of sampling were explored detailing pigments from an unused artist palette from 1880 that effectively provided standards for pigments through to illuminated manuscripts, ceramics and more recently the exploration of polymers in more contempory exhibits.

Health and safety aspects were also considered by demonstrating the elevated arsenic contamination from the process of taxidermy to the contamination of mercury in felt hats. Indeed it was suggested that if one hat was thrown into a reservoir it had the potential to poison one million litres of water at present WHO levels of mercury contamination.

Owen Butler, (HSL, Buxton) & Elke Adriaenssens, (VMM, Antwerp, Belgium)
Analysis of heavy metals in suspended airborne particles collected on filter papers by XRF techniques.

Owen began by summarising the work of the HSL at the Buxton site which has favourable scientific and engineering facilities. As an agency of the Health and Safety Executive, part of the governments Department of Work and Pensions the analytical services unit undertakes wide ranging tests and is involved with regulatory and investigative measurements, method development, proficiency testing and some commercial work.

The workplace risk to the inhalation of aerosol and particulates down to sub-micron size and subsequent sampling problems were discussed. In particular was noted the problem caused by ergonomic work practices of sampling to mimic inhalation. Technical innovations such as a unique design of filter cups to accept varying diameter papers from 13-48 mm, together with the problems of moving from WD-XRFS to ED-XRFS analysis and the subsequent bias introduced with the smaller beam size were presented.

A new model for sampling particles, the Minisampler employing a 13 mm filter located adjacent to the mouth was demonstrated and validated using a model and automated welding rig.

Future interest lies in the analysis of nano-particles with spot sizes down to that of a pinhead.

Elke introduced her work with VMM, the Belgium equivalent of the EA, reporting on air and water quality and the general state of the environment. Air quality was measured in a network of automatic and semiautomatic stations for a variety of parameters including PM10 and PM2.5 particulate fractions. The automatic stations provide continuous real-time data available on the website www.vmm.be .

Two populations of heavy metal data for 2007 were described, urban/rural and industrial. The rural/urban mean and maximum values were low whilst the industrial values were high requiring a wide ranging application to process more than 7500 samples annually. Technical problems were overcome by the attention to measuring conditions and the lack of commercially available standards overcome by a method of in-house standard preparation. The calibration was described as secondary and validated by ICP-MS or AAS techniques.

Elke went on to describe the equipment used for heavy metal determinations commenting on the inflexibility of using two X-ray tubes for the WD-XRFS technique. A change of equipment to a PANalytical Epsilon 5 ED-(P)-XRFS with a Gd tube allowed for the analysis of As, Cd, Cr, Cu, Mn, Ni, Pb, Sb and Zn at 30 minutes per sample. The advantage of the ED analysis was a lower detection level, meeting BS EN 14902 requirements except for Cd. Validation between ED-(P)-XRFS and ICP-MS showed good correlation for As, Cd, Ni and Pb at higher concentrations. For Ni and Pb the correlation is still acceptable at lower concentrations, for Cd and As the correlation is not good in the low concentration range.

Simon Carter
BGS

Chair, Afternoon Session 2
Dave Taylor.

Left to Right: Andy Scothern, Alison Burke, Dave Taylor (Chair) and Richard Meeres.

Richard Meeres, (Bureau of Analysed Samples Ltd).
A History of the Development of Certified Reference Materials.

This presentation discussed the history of the company, from its inception in 1912 as Ridsdale & Co to the Bureau of Analysed Samples (BAS) in 1935. In 1950, the Honorary Advisory Committee (HAC) was created to provide an unbiased wealth of experience, make recommendations, evaluate results and approve Certified Reference Materials (CRMs). The HAC has met twice a year since its creation, and BAS also organises Triennial Meetings of UK Co-operating Analysts to gather comments and ideas for the future.

The types of CRMs produced consist of iron & steels, non-ferrous metals & alloys, iron & non-ferrous ores, ferro-alloys, slags & refractories and ceramic materials & minerals.

Alison Burke, (Huntsman Pigments).
Routine Analysis of TiO₂ by WD XRF.

Huntsman Pigments has seven factories around the world which together produce approximately half a million tonnes of titanium dioxide each year. Around 60% is used in coatings (paint, etc) with the remainder being used in products such as plastics, inks, toothpaste and cosmetics. Sample types requiring analysis are the final packed pigment, calciner discharge, in-process samples and ores. Several techniques are used for elemental determination; however XRF provides a simple, rapid and easy to use method.

The Harmonisation Project of Analysis was initiated in 1998 in order to ensure inter-site reproducibility precision targets for each XRF element. Initially blind testing was carried out every six months to establish levels of agreement using z-score statistics. Now established harmonisation is maintained by the project sharing monthly validation for all seven laboratories, carrying out annual blind tests, hosting telecons every two months, providing operator training and international compliance audits.

Andy Scothern, (Saint-Gobain Gypsum).
XRF Analysis in the Gypsum Industry.

Gypsum markets consist largely of construction (37%), renovation (29%), household (19%) and industry (15%). Saint-Gobain Gypsum (SGG) currently employs in excess of 200,000 people, with two out of three working outside France. Research and Development centres are based mainly in France, but also in Germany, the United States and South America. The Technical Centre is based in East Leake and concentrates on process research, product research and product development.

Gypsum consists of several phases - gypsum (CaSO₄.2H₂O), hemi-hydrate ( CaSO₄.1/2 H₂O), and both soluble and insoluble forms of anhydrite (CaSO₄). By applying temperature to this material, it phases through gypsum to hemi-hydrate to soluble anhydrite finally to insoluble anhydrite and back again by the addition of water. The reduction of water, and therefore the reduction in heat required for the final products are crucial and SGG is trying to find ways to implement this.

Samples are dried overnight at 40°C in order to remove any free water and to convert any soluble anhydrite to hemi-hydrate. Proximate analysis was used in the past to record the weight loss on heating; however, thermogravimetric (TGA) monitoring is now used. Problems with this technique include the inability to measure insoluble anhydrite, as there is no water left to lose, and difficulties when clays or additives are present. WD-XRF is now used for total SO₃ content in anhydrite as it is more accurate than TGA, assuming all SO₃ is present as sulphate.

Other work in progress included analysis of cement boards, semi-quantitative analysis on powders and liquids and trace metals in dusts.

Leian Grimsley
BGS