Hi, One way to get rid of the temperature uncertainty is to introduce a second thermocouple inside the furnace and to fix it as close as possible from the sample. If the sample is a bulk sample:Drill a hole inside the sample and put the thermocouple inside the hole and if the the sample is made of powder put it inside the powder. The are some miniature thermocouple (0.25-0.5mm diameter), mineral insulated which are unlikely to react with the samples.
Patrick ----- Message d'origine ---- De : "Payzant, E. Andrew" <payza...@ornl.gov> À : Mikko Heikkilä <mikko.j.heikk...@helsinki.fi> Cc : Rietveld list <rietveld_l@ill.fr> Envoyé le : Mer 14 Octobre 2009, 16 h 39 min 58 s Objet : Re: HT XRD calibration All, The two calibration standard method was described by Andy Drews (Ford Research Lab) in Advances in X-ray Analysis (the proceedings of the Denver X-ray Conference), vol. 44, page 44 (2001). Scott Misture (Alfred Univ), Stephen Skinner (Imperial College), Christian Resch (Anton Paar), and I ran a workshop on non-ambient diffraction at the Denver X-ray Conference in 2008, where we discussed many of these issues. There was a lively discussion with the attendees regarding various experiences with HTXRD, and the difficulty in abtaining the "right" answer. Too often we are so focused on the precision and accuracy of our refinements that we neglect the temperature issue, to our peril. One issue with the HT16 hot stage is that you will likely need to recalibrate with every new heater strip - don't rely on them behaving the same. Even for a particular strip, use of an internal standard is best if you can use one that does not react with either the strip or the sample. If you push this furnace to temperatures above 1200degC or so, the platinum will creep and the tensioning will change, so that the height/temperature calibration may subsequently be off. I agree fully with the suggestion to use parallel beam optics to eliminate the sample height error. You do lose some angular resolution of course, at least on a laboratory diffractometer, but you take one variable (sample height) out of the refinement. The other BIG problem with the HT16 (and many other non-ambient stages) is knowing the temperature of the SAMPLE (as opposed to the strip). Powders have hopelessly poor thermal conductivity, and the gradients in these furnaces can be substantial. We rarely measure the temperature of the sample directly, and this is another reason why internal standards are very useful - you can use either a known CTE to estimate the delta T over the full temperature range, or a known phase transformation to determine delta T at a single point. I personally prefer the former approach myself. Note that a solid-solid transformation may give a different result than a melting point standard, since the liquid will change the thermal conductivity between the strip and the sample. You can find more discussion of these issues in Mark Rodriguez's chapter in the book "Industrial Applications of X-ray Diffraction", or mine in "Principles and Applications of Powder Diffraction". I chair the "non-ambient diffraction" subcommittee of the ICDD, and in that capacity I am looking for ways we can develop and standardize these methods. If any of you would like to participate in this exercise please let me know. Perhaps a special issue of Powder Diffraction could be devoted to in-situ high-temperature diffraction papers? Always looking for good ideas... Andrew On 10/8/09 1:40 AM, "Mikko Heikkilä" <mikko.j.heikk...@helsinki.fi> wrote: Hello Rory, Ian, and the rest. I've been planning to do a similar calibration for a long time for our Anton Paar HTK1200N but I've had to postpone it over and over again. The approach by John Evans is very intriguing to me, but I'm lacking a recent Topas version and I'm not sure if very old 2.1 will do it. However, I think that a way to circumvent the simultanious refinement of temperature shift and wrong height is to use parallel beam if you have the possibility. On one hand you could first determine the temperature behaviour as the peaks shouldn't really shift because of the wrong height when using parallel beam. When the temperature behaviour is known, change back to focusing geometry and calibrate the height. On the other hand it might be possible to get the correct height in different temperatures just by changing the sample holder position until it halves the direct beam intensity. This is possible at least for our ceramic sample holder, I'm not sure how this would work with Pt-strip in other ovens. I'm very interested about suitable test materials. I was planning to go with corundum and parallel beam first, but now that this thread is open, I'll wait a bit longer (again). Regarding the calibration with phase change temperature, Robert Dinnebier has some materials listed on his webpage: http://www.fkf.mpg.de/xray/html/temperature_calibration.html Please correct me if I've made much mistakes above, I'm still at rather beginning of my learning curve :-) Cheers, Mikko ian.mad...@csiro.au wrote: > Hi Rory, > > At the moment we are in the same position i.e. attempting to calibrate the > temperature and systematic offsets in an HTK10 furnace with Pt strip heater. > One approach is to take materials which have a well defined solid-solid phase > transition or melting point and observe this as a function of set > temperature. Unfortunately, most materials only provide one or two > calibration points and hence you will need to heat several samples in order > to cover a wide temperature range. > > At the moment we (and by "we" I mean my PhD student Mark Styles is doing all > the work and I stick my head in occasionally and say "how's it going?") have > taken an 'external standard' approach (we cannot add a standard to our sample > and risk reaction). In this we heat a material with known thermal coefficient > of expansion (TCE), measure the unit cell edge a function of temperature and > then determine a calibration curve. However, the flaw in this approach will > be the impact of other systematic errors (e.g. sample displacement) on the > observed peak positions and hence cell edge. > > John Evans (Durham Uni) uses a modified TCE approach in which he adds two > standards with different TCE's (I haven't got the references to hand, but > this is described in one of his papers) and normalises observed cell edge (or > cell volume) to that at the starting temperature (To). Instead of using the > absolute cell edge (volume) for each phase, he then uses the difference > between the two compounds as a way of eliminating the systematic errors which > should affect both equally. > > If you pardon the pun, this is a hot topic for us at the moment - we are > coming up to some synchrotron time where T calibration will be important - we > have not completely sorted out the fine details but intend to:- > > (1) Use a mixture of NIST silicon and corundum > (2) Heat to various temperatures over the range of interest > (3) Use the 'surface analysis' approach (Stinton and Evans) which relies on > use of the Topas Rietveld software. In this, the cell edge is not refined, > but defined as a function of applied temperature, the temperature offset (a > refinable parameter) and the published TCE equations for each phase. > > What we hope to end up with is an analytical method which is quick and easy > to apply to produce a T calibration for every new heater - i.e. it should not > be an onerous task so that it can be carried out frequently. > > I hope this helps - we still have much fine detail to sort out. > > Cheers > > o----------------------oo0oo---------------------------o > Ian Madsen > Team Leader - Diffraction Science > CSIRO Minerals > Box 312 > Clayton South 3169 > Victoria > AUSTRALIA > Phone +61 3 9545 8785 direct > +61 3 9545 8500 switch > +61 (0) 417 554 935 mobile > FAX +61 3 9562 8919 > Email ian.mad...@csiro.au > o----------------------oo0oo---------------------------o > > -----Original Message----- > From: r.m.wil...@qmul.ac.uk [mailto:r.m.wil...@qmul.ac.uk] > Sent: Thursday, 8 October 2009 3:14 AM > To: rietveld_l@ill.fr > Subject: HT XRD calibration > > > Dear All > We have an Anton Paar HTK16 furnace with a Pt strip. I am > hoping to calibrate it both thermally and mechanically (height shift) > using either alpha alumina or ceria so as to better identify HT > phases. Have any of you experience of this, or better suggestions for > calibrants? My only serious worry is whether there is any potential > for ceria to react with, the Pt strip? > (Perhaps under vacuum?) I'd welcome any advice or suggestions. > Yours > Rory > > ************************************ > Dr R.M. Wilson, > Experimental Officer & RPS > School of Eng. and Materials, > G27 & G30 X-ray Laboratories, > Francis Bancroft Building > Queen Mary College > University of London > Mile End Road > LONDON E1 4NS > UK > Tel: +44 (0)20 7882 7938 > Fax: +44 (0)20 7882 7931 > ************************************ > > > -- Mikko Heikkilä, M.Sc. Laboratory of Inorganic Chemistry Department of Chemistry P.O.Box 55 (A.I. Virtasen aukio 1) FI-00014 University of Helsinki phone: +358 9 191 50216 fax: +358 9 191 50198 mobile: +358 440 572 372 email: mikko.j.heikk...@helsinki.fi -- E. Andrew Payzant Senior R&D Staff Member High Temperature Materials Laboratory Oak Ridge National Laboratory 1 Bethel Valley Road PO Box 2008, MS 6064 Building 4515, Room 113 Oak Ridge, TN, 37831-6064 ph: (865) 574-6538 FAX: (865) 574-3940 web: <http://www.ms.ornl.gov/DTP/payzant.shtml> __________________________________________________ Do You Yahoo!? En finir avec le spam? Yahoo! Mail vous offre la meilleure protection possible contre les messages non sollicités http://mail.yahoo.fr Yahoo! Mail
