What Has Changed in XRF Technology in the Last 20 Years?

• Look at my personal XRF time line
• See how the technology has changed over the last 20 or so years
• The new hobby of home XRF

My XRF Time Line

• 1966 Sheffield University Metallurgy Department Philips PW 1310 XRF with added electro-mechanical print out. Couldn’t find an image
• This unit was about a 1.4m cube, weighed over a tonne and had two 2m electronic racks, one for the HV and the other the scaler displays and controls
• 4 place sample changer with spinner and about 10 adjustable pre-set positions for the crystals

​

​

My XRF Time Line

• 2 X-Ray tubes, Rh for general work and Cr for light element
• Choice of air, vacuum or He path
• Argon-methane flow gas
• Scanning a sample gave a chart recorder output 3 metres long which took about 1 hour
• Used a set of look up books to convert 2Θ to element

My XRF Time Line

• 1994 Sheffield Assay Office we bought a Spectro X-Test jewellery analyser with a proportional counter with ~950eV resolution on Mn. Bought another about a year later
• Had a 386 desktop PC and could be used for metals analysis and plating thickness
• Excellent performance but we suddenly found that some solders used were under carat and now failed assay!

​

​

Technology Changes�Different energy dispersive detector

Technology Changes�Proportional Counter v SDD

SDD detector ( Si(Li)) vs Pin diode

Technology Changes

• The Silicon Drift Detector really was the next big change for EDX systems around 2004
• Better resolution, higher count rates, a real game changer
• Coupled with the new technology of X-ray capillary optics now we have fast scanning at count rates nearing 500Kcps

Silicon Drift Detector

• Thermoelectrically cooled (Peltier)
• Be window or Carbon based material
• Light element capability (i.e. P, Si, Al)
• Count rates: 10,000~1,000,000 cps
• Resolution: 120~170eV at Mn K a
• 10+ years life

​

Introduction to µ-XRF�Type of µ-XRF

• Conventional XRF Micro XRF with collimator Micro XRF with optics

Typical spot size and analytical task

Poly-capillary focusing optics

• Channel size in microns
• Large open-area: up to 80%
• Compact structure
• Light weight
• Robust

Poly-capillary focusing optics

• Sub-million to multi-million capillary channels
• Each individual channel is aligned to the same point – the focal point
• Focal point exists on both the input side and the output side
• High flux density gain – up to five orders of magnitude higher than a pinhole collimator and at least two orders of magnitude higher than a monocapillary optic
• The focal spot size is independent on the source size

f_output

f_input

Typical Scanning Applications �PCB of a PDA with integrated circuits – look inside

Mapping area: 116x68 mm², Step size: 90 µm Measuring time: 15 ms/pixel, 5.5 h total

Technology Changes

• Miniaturisation has now given us hand held XRF with excellent performance for a very wide range of sample types
• The latest version Silicon Drift Detector with a graphene window now extends the capabilities of HHXRF to see Na in samples even without the use of He

​

Technology Changes

• Polarised XRF is now widely used in current instruments giving significantly lower detection limits
• Particularly useful in the petro-chemical industry to analyse for light elements in fuels

​

Technology Changes

• In WDXRF there hasn’t really been any radical changes in the last few years
• Instruments are smaller with lower power and running costs in a bench top model
• Minor tweaks in crystals and geometries give improved performance
• Newer crystals for light elements now allow us to test even down to Be

Technology Changes

• Count rates have increased, 1Mcps is the now possible
• We can of course now carry out a surface scan of a sample in some WDXRF systems
• Whilst this can be quite slow it does have the benefits of lower detection limits and ability to see lighter elements than EDXRF

Technology Changes

• The often forgotten XRF system is of course TXRF, not possible without all the technology changes

​

Technology Changes

• Whilst it is a niche market it has benefitted greatly from the new generation SDDs giving detection limits for some elements in some matrices of 1pg
• For trace analysis for some elements this is an excellent technique with negligible running cost compared to ICP-OES/ MS

​

The New Hobby, Home XRF

• For me the best part of the new technology is you can have your own Hand Held XRF
• It means old analysts like me can still keep the grey cells turning over on the odd project like medieval Fe making slags and ores
• Its only when you have to start paying for your own consumables you think laterally to save money
• So this Yorkshireman tries to do XRF on the cheap

Typical Early Fe Making Slag Find

Sample Crushing

• Hand crushing in an agate pestle and mortar several dozen samples is a possibility
• Is it easier to drill the samples with a masonry drill? Yes it is.
• Or adapt an angle grinder to put a “catching tube” on it? SiO2 / Al2O3 contamination?
• Just work with solid samples and test several areas?

Fe Slag Sample Analysis

• Reference materials?
• Modern iron making slag standards of course have very little Fe in them whereas historical slags can have 30% or more
• In my stash of bits I have some geological standards and I can buy some pure Fe₂O₃
• Make my own reference standards by mixing powders

My Geological Reference Materials

Home Made Reference Materials

Comparison of Fe calibration Graphs Using the Same Spectra, Loose Powders

As Fe Regression = 0.992892

As Fe2O3 Regression = 0.952421

Fe Slag Sample Analysis, Sample Cups

Eventually I ran out of all the spare cups I had acquired.

Got these off an auction site, plastic cups for beauty products

Use an “O” ring to hold on the Mylar type film

Pressed Pellets

• Can buy an XRF pellet press and die for £5K-£10K secondhand
• https://shop.unigreenscheme.co.uk/
• Garage type hydraulic press about £100-£150 but still need a die
• Back to auction site, you can buy a “herb press” for about £10
• I made a press with a piece of steel bar and 2 bolts

Pressed Pellets

• Pressing Pellets

​

Pressed Pellets

Fused Beads

• As part of my research into early iron making slags ideally I need to make fused beads
• So Li salts? Pt crucibles and moulds? Furnace?
• Far too expensive
• Can you use graphite crucibles for fusing XRF beads? Several papers available on this
• Paper by Malvern Panalytical using Borax and NaNo₃ to fuse Iron ore

Fused Beads

• Need a furnace
• Borax has a much lower melting point than Li salt of course
• Microwave furnace might be hot enough
• Glass kiln furnace from an auction site £27
• Ceramic lined with SiC / SiN easily gets to 1000C

Fused Beads

Fused Beads

Cu 22mm Plumbing Stop End

• Casting mould? Pt? Too expensive
• I am only interested in elements Mg-Fe
• Would a Cu pipe stop end make a mould?
• In fact can I melt in these at 900C?

​

0.4g Fe2O3 Fused in 4g Borax, Cast into Cu “Mould”

Blue spectrum top of the bead, Red spectrum bottom

​

Unfiltered X-Ray tube so will see Cu and Ni from the tube

0.4g Fe2O3 Fused in 4g Borax, Cast into Cu “Mould”, Expanded View

0.4g Fe2O3 Fused in 4g Borax, Cast into Al and Cu “Mould”, Expanded View

Blue spectrum top of the bead cast into Al mould

Red spectrum top of the bead cast into Cu mould

Unfiltered X-Ray tube so will see Cu and Ni from the tube

Conclusion

• It may possible to carry out XRF analysis on geological samples using much cheaper equipment
• In some countries this can be important where simple, inexpensive equipment can help with their analysis
• How much will Li salts be in few years time when they are all being used for car batteries?

20 Years of XRF

• Thank you for Your Time
• Any Questions?