8th RC Competition

The 2016 Reynolds Cup was supported by The Clay Minerals Society, the Deutsche Ton- und Tonmineralgruppe e.V. (DTTG), and the Technical University Bergakademie Freiberg.

Once more the Reynolds Cup has seen a big response in the community of mineralogists. Whereas in 2002 from 40 registrants 15 submitted their results (rate 38%) the number increased again to 83 registrants from 25 countries with 69 sent in results (rate 83.1%).

Eighth (2016) Reynolds Cup

As usual the three samples of the 8th RC2016 were comprised of mixtures of purified, natural and synthetic minerals commonly found in clay bearing rocks and soils that represent realistic mineral assemblages. The specific aim was to evaluate proficiency of analysts working on clay-rich materials representing a shale, a clay-rich tailing, and a weathered/altered ultrabasic rock. This information was relayed to the contestants with their samples. The non-clay minerals and clay minerals/phyllosilicates were judged in groups as in previous competitions.

The three top contestants were announced at the 53rd Annual CMS meeting in Atlanta, Georgia, in June 2016 and presented with plaques. Here they are:

First Place Winners

The team of QMINERAL & KU Leuven / ONDRAF-NIRAS, Heverlee, Belgium

Presenting the Cup: Rieko Adriaens (left) and Gilles Mertens (right)

First place, bias 50.7 %

Rieko and Gilles have formed a creative team and worked permanently on the improvement of their techniques. In previous contest they repeatedly reached ranks close to the top 3, and this year they occupied the throne. Congratulations!

Second Place Winners

Peter Self and Mark Raven, CSIRO Adelaide, Australia

Second Place, bias 54.2 %

Mark and Peter are working in the field of applied mineralogy and are very experienced in phase analysis of various materials, not only clays. The team already reached the third place in 2008 and won 5th contest in 2010. As active members of the CMS they also contributed to the improvement of methods in our community by supplying nice special clay minerals from Australia to the Source Clay Repository.

Third Place Winners

Stephen Hillier, Ian Phillips, Helen Pendlowski, Nia Gray

The Hutton Institute, Aberdeen, Scotland

Third Place, bias 57.1%

Since 2002 no Reynolds Cup contest without this team in the top 3 ranks – Steve’s group has demonstrated that superior quality mineral analysis can be provided persistently. This year only the rigid application of the rules for discriminating illite from illite-smectite interstratification, a orgiation of a few percent swelling layers, accounted a significant bias to the evaluation of the results. Without this diminutive uncertainty the team would have won once more with significant margin. Especially impressive are the great results presented by the group for the most complex and complicated sample 8/3.

Traditionally, the RC organizer asks the participants who came close to the top 3 ranks if they agree to be identified to the public, to acknowledge their superior quality work, too. This year we have a shared 4th place:

Fourth Place Winners

Carolin Podlech, Markus Peltz, Georg Grathoff, University of Greifswald, Germany

Shared Forth Place, bias 80.0%

This newly formed team, consistig of two students and one experienced researcher, participated for the first time in the RC Contest. Well Done!

Fourth Place Winners

Kristian Ufer, Reiner Dohrmann. Jan Dietel and Stephan Kaufhold (absent),BGR Hanover, Germany

Shared Forth Place, bias 81.4%

The BGR team participated in the RC continuously since 2002, was placed second in the first contest, third in 2014, and in all other contests always ranked very close to the top three. The group works actively in orgelopment of methods, and other teams who used the BGMN software will have profited by their efforts. Thanks!

How the Top 3 teams did the job

Rieko and Gilles, the QMinerals and KU Leuven team:

As in most laboratories, our mineral quantification procedure relies on ‘calculating’ an X-ray diffraction pattern and matching it to the measured X-ray diffraction pattern. When reviewing our submissions of previous RC editions, we noticed that, even when using the right compositions and abundances, we could not perfectly reconstruct our experimental X-ray diffraction patterns. The precision of our X-ray diffractogram was therefore not representative enough for the provided mineral mixture. Hence, we decided not only to use the traditional McCrone milling to improve the quality of the diffraction data but also to apply spray-drying. The initial sample was split in four representative parts. Two parts were subjected to size separation in three different size fractions and measured “as such” without any further treatment as we felt this is crucial for correct identification. The large advantage of this procedure is that when treated carefully, the material can be re-used. These two splits (ca. 2g) were then used to prepare powder diffraction patterns used for quantification. The sample was McCrone milled twice and spray-dried. Al2O3 was measured following the external standard method. One split was used for chemical major element analysis. All samples were measured using a traditional Phillips PW1050/37 goniometer connected to a PW1830 generator equipped with Cu-Ka radiation and proportional detector type PW3011/11. For quantification, we traditionally use both Rietveld refinement (TOPAS) and full pattern fitting software (QUANTA) depending on the sample composition. Recently we started using a PONKCS-assisted Rietveld procedure in TOPAS (Bruker) whereby all crystalline minerals are refined as usual. Clays however are modeled as “fixed”, non-refineable structures.

Mark and Peter, the CSIRO team:

Preliminary back pressed XRD patterns of each of the supplied powder samples were run ‘as received’ on our XRD. Sub-samples of ~ 1.5g of each sample were then micronized under ethanol with a McCrone micronizing mill then oven dried at 60°C. After drying, the micronized samples were thoroughly mixed in an agate mortar and pestle to ensure homogeneity. The fine powders were lightly back pressed to minimise preferred orientation and XRD patterns collected on a PANalytical X’Pert Pro MPD using iron filtered cobalt K alpha radiation. Patterns were collected from 3 to 80° 2 theta at 0.017° steps. Total data collection time was ~30 minutes. The process of micronizing the samples under ethanol followed by oven drying partially dehydrates the swelling clay minerals resulting in broad asymmetric 00l peaks. The micronized samples were therefore calcium saturated to restore the 001 peaks of the swelling clay minerals to ~15Å. This was achieved by washing the micronized samples twice with 1M CaCl2, washing with deionised water followed by ethanol (centrifuged at 6000rpm after each step) before oven drying at 60°C. The Ca saturated samples were again thoroughly mixed in an agate mortar and pestle to ensure homogeneity and lightly back pressed into sample holders for XRD measurement. XRD patterns were then re-collected. Comparison of the XRD patterns before and after Ca saturation confirmed there were no water soluble phases present in any of the samples. A further 2g sub-samples of the as received materials were dispersed with 1M NaCl and centrifuged at various speeds to separate <0.2µm, 0.2-2µm and >2µm fractions. The fractions were again Ca saturated and pressed powder and oriented, magnesium saturated and glycerolated specimens were prepared to help identify the clay mineral species. During the centrifugation process, thin, dark particles were found floating on the surface and collected for XRD and SEM with EDX analysis. These particles were found to be amorphous, composed only of carbon and later shown to be anthracite. Quantification was performed using SIROQUANT version 4 with ‘calibrated’ HKL files prepared from several in-house clay standard materials. This was a critical part of the analysis due to the presence of the Mg-rich clay minerals; saponite, talc, antigorite, vermiculite and Mg-rich chlorite identified in sample RC8-3. Elemental analysis of a sub-sample of the Ca saturated materials were determined by Li borate fusion and XRF analysis. This was used to confirm the upper level of Ti oxide phases present and also assisted with the identification of fluorite, apatite and cassiterite in sample RC8-1.

Steve Hillier and the Hutton Institute team:

How did we do it? Basically, like last time in 2014 and pretty much the same as before that in 2012. Our quantification method is based around full pattern fitting of experimental patterns. So the key is having appropriate patterns for all the minerals that are identified. We typically begin by running unprepared ‘as received’ samples on the XRD, but although this may give some indication of the minerals present, we really only make these runs in the excitement of receiving the samples! They are not really useful for anything. Undoubtedly, the best diffraction patterns for phase ID and the subsequent quantification are those prepared next by McCrone milling followed by spray drying of a carefully split portion of the bulk as received samples. Because the spray drying eliminates preferred orientation it becomes possible to rely much more on intensity information as well as d-spacings during mineral identification. Having identified what minerals we can the next step is to start fitting patterns and this process itself also assists with further mineral identification as well as confirming if the standard patterns we are attempting to fit are appropriate for the actual minerals present in the sample. For some minerals in the RC8 samples were needed to make new standards, for example we didn’t have a topaz standard, or a mica that matched the trioctahedral mica in sample RC8-1, so we obtained some new specimens and made up some new standards. One of the big advantages of an experimental pattern fitting approach is that provided your standards are good matches to the unknowns (and the fitting itself is what tells you that) and you are confident of the calibrations for your minerals, there is absolutely no reason to dilute your sample with an internal standard; it just gets in the way! Because background points are also included in the fitting, it also means that experimental full pattern approaches can quantify amorphous phases directly. This can be a big advantage over other X-ray diffraction methods especially when disordered minerals like clays accompany amorphous phases. While fitting patterns we also looked closely at oriented clay fractions separated from small sub samples to identify the clay minerals present. This helps ensure that the random powder patterns chosen for the quantification of the whole bulk sample are as appropriate as possible. Finally, we also made a chemical analysis by XRF and this is used as a cross check on the final mineralogical analysis. The guiding principle here is that any mineralogical analysis must be compatible with the measured chemical composition of the sample. In fact on this occasion, we correctly questioned the accuracy of an XRF analysis with respect to an erroneous silica content provided for us by an external lab (we don’t have our own XRF). We were suspicious of the XRF result because our calculations based on assumed compositions for the minerals identified and their weight fractions indicated considerably less silica than the XRF lab reported; it will make us think twice about using that lab again. The main point however, is that even based on assumed mineral compositions this kind of cross cheek is probably the best way that there is to verify that your quantitative mineralogical analysis is a reasonable one. The one mistake that cost us a big bias penalty was a misidentification of an illite (with about 5% expandability) as an illite/smectite with about 10-15% expandability (Figure 1). In hindsight this was perhaps due to our over reliance on the full pattern fitting of the bulk samples, we should have examined the clay fraction more closely than we did, and it emphasises that quantification has to be conducted hand in hand with precise identification.

Figure 1. An Illite (red trace) like as was in sample RC8-2 versus an illite/smectite (blue trace) as we modelled as in RC8-2.

Once again like previous Reynolds Cups we can only say that we learned a lot about our methods and importantly how we can work to improve them further.

Thanks to the winners for sharing their knowledge, and thanks to all participants for their efforts in raising the level of mineral phase analysis.

2016 principal organizer

Reinhard Kleeberg

Freiberg, Germany