24April2013

XRF Meeting Report 24^th April 2013

A joint BCA/ RSC Atomic Spectroscopy Group Meeting at University of Leicester.

A PDF version of this report (frep13.pdf) with extra photographs can be downloaded from the foot of this page. Clicking the file name will load the file into your browser, clicking the small blue down arrow to the far right of the file name will download a copy to your computer.

Some of the presentations at this meeting (as PDF files) are available to view.

Delegate Photograph.

Click on photograph for larger image!

Morning Session.

The first session of the meeting was a practical demonstration by Oliver Meyer and Rainer Schramm, of Fluxana, of techniques for the preparation of solid samples for XRF without using a fusion. In many cases, fusion is the best option, but it is not the simplest. Loose powder samples can give good results, but all to often do not, because of variable packing and/or particle-size effects. Pellets can sometimes be pressed without a binder (like the KBr discs often

Morning Session demonstrators and speaker, from left to right:

Oliver Meyer, Ros Schwarz(Chair), Allan Finley, Frederic Davidts, Anne-Catherine Breton, Rainer Schramm and Margaret West (Chair)

After the coffee break, Anne-Catherine Breton from Claisse spoke about the development of a procedure for the XRF analysis of iron ore samples. The present method, ISO9516-1:2003, covers a fair range of elements (not including iron), but over calibration ranges which are often inadequate. The first calibration used standards made with pure oxides and a comparison was made with a second calibration using CRM’s as standards. The new method includes iron and widens the calibrated range for almost all elements. The 29 standards used are 20 iron ores and 9 other materials. The XRF analysis uses the LiF (200) crystal wherever possible; count times are designed to give the required accuracy and precision. Sample preparation is a fusion using a borate flux - a different mixture of lithium metaborate and tetraborate to the ISO procedure - at 1050 °C in a 95Pt/5Au crucible. Ammonium nitrate (rather than sodium nitrate, which precludes Na determination) is added as oxidant, and ammonium iodide as non-wetting additive. Copies of a leaflet on this process were available from the Claisse stand.

The final talk of the morning session was Ros Schwarz's introduction to the participant sample. Portions of this were distributed to any participant who wanted it. This is intended to act as an aid to methods development and training, and is not a proficiency test or Round Robin. Participants are asked to determine as many elements as they wish by any suitable XRF method or other technique, and to return the results by 30th September 2013. The origin of the sample was not disclosed in detail, but it is a composite from more than one brownfield site. Accessory minerals include sand and clays.

David Beveridge

Afternoon Session.

Afternoon Session speakers, from left to right:

Garry Smith, David Beveridge (Chair), Chris Calam, Colin Slater, Steve Davies and Dave Taylor (Chair)

Fortified by a good lunch we enjoyed two papers offering good advice for delegates who had collected their “participant sample” from Ros.

Being in the audience for Steve Davies’s original talk on Specimen Support Films back in 1995 it was interesting to note that not much had changed. Using 3 international standards for oil applications (ASTM D6443, ISO 20884 & IP503) Steve showed that hydrocarbon films were “cleaner” than those bearing O2 and benefited from better transmission per unit area. The chemical resistance data showed that the films hadn’t changed but in the petrochemical industry the samples are different. Fuels now have increasing O2 bearing additives and samples containing hydrocarbon mixed products should not be placed on hydrocarbon films. For oil samples 3.5 μm Mylar was recommended. For those outside the petrochemical industry interested in standardless analysis, hydrocarbon films are cleaner and have better transmission. Whichever film is selected, the analyst was advised to test before use as leakage may have an adverse effect on the service support costs.

Colin Slater’s talk on Standardless Analysis gave us useful advice when charged with answering the time old question “What is in my sample?” When no standards are available, we now have fundamental parameters at our disposal. Colin steered through challenges such as matrix effects and showed that any XRF FP data is as good as the information that is known about the sample. Using an Aldrich molecular sieve, analytical data was shown to improve as more facts concerning the sample were introduced into the software package. For water/ethanol/oil based samples, the Rh Kα Compton ratio technique was offered as a valuable way to deal with very low density materials. Finally, Colin gave examples of how to analyse layer thickness in samples.

Margaret West.

After our break for tea (and yummy cake!), Garry Smith, XRF Application Specialist with SciMed, discussed the difficulties inherent in carrying out Liquid Waste Analysis. He highlighted the general problem caused by the presence of two phases in these samples; suspended solids and immiscible liquids, which will start to settle out and separate as soon as the sample is poured into an XRF cup. He emphasized how difficult this makes it to fulfil the basic requirement that the analysis should be representative of the bulk material, because the radiation from each element in the sample will have a different penetration depth. Penetration depth increases with increasing photon energy, but decreases with average atomic number of the matrix and sample density. With light matrix liquid samples the penetration depths can be dramatically different on the scale of a sample cup. For example Cr Kα radiation will come from up to 4.33mm in oil or 1.44mm in water, but Cl Kα will come from the bottom 0.36mm in oil or only 0.17mm in water. He showed graphs of an oil and water mix with the Cl predominately in water and Cr predominately in oil where over ten minutes the Cl signal increased by about 70% and the Cr signal decreased by 33% as the water layer separated to the bottom of the cup. He then mentioned two options to tackle the problem; separating the phases by centrifuging or filtration before analysis; or using a solid binder to immobilise the phases. He suggested binders such as graphite powder or activated alumina for this purpose.

The final talk of the day, given by Chris Calam from Thermo Scientific Portable Analytical Instrumentation, took us out of the laboratory into the African bush, away from any sources of power. His talk on Field Portable Sample Preparation Tools for Mineral Analysis in Remote Location discussed the challenges of surveying for ore bodies. Field Portable (FP)-XRF is used to detect elements such as Cu, traces of which can reach the surface above an exploitable deposit. In the first instance, FP-XRF is used on a transect to look for a “Cu anomaly”; levels significantly above background. Chris showed us plots demonstrating how well FP-XRF can detect Cu anomalies with no sample preparation at all (compared to laboratory results for samples from the same transect analysed by XRF and ICP-MS). However, when the exploration moves on to resource estimation and reliable quantitative analysis is required, sample preparation is needed as figures from FP-XRF can vary markedly over one piece of a heterogeneous rock. Comparing FP-XRF field and laboratory XRF and ICP-MS measurements graphically illustrated the improvement in correlation obtained when some sample preparation was carried out in the field. For example, the relative standard deviation of results for nickel improved from 0.797 to 0.961 and for Mg improved from 0.417 to 0.950; also emphasizing how particularly important this is for light elements. The battery-operated sample preparation tools he suggested included a direct rock sampler which grinds material from the surface of an outcrop into a collection tube and a portable mill. Other preparation equipment was human-powered, such as a crusher with a large hammer and a sample press, also operated by sharp blows with a hammer (up to about 10 tons). These items are marketed in a kit.

We ended the meeting by expressing our thanks to the local organisers from the University of Leicester, Nick Marsh and Cheryl Haidon and to the members of the XRF committee.

Ros Schwarz.