To journal index
1 August 2011

Using XRF Analysis on Two Staffordshire Hoard Objects

X-ray Fluorescence (XRF) is a technique that is widely used in the heritage sector to identify the elements present on the surface of an object as it is a non-destructive technique. Once the objects from the Staffordshire Hoard have been cleaned they are analysed in order to discover the composition of the metals used. Birmingham Museum and Art Gallery possesses a Bruker Tracer III-V handheld XRF analyser. This has the advantage of being portable so that it can be taken to the objects if they are large or located elsewhere. However when working on smaller objects the slight disadvantage is the 5mm beam size, so it cannot be used to pinpoint small areas for analysis. Therefore we have a partnership with the British Museum conservation and scientific departments where more detailed and accurate analysis will take place in the future. This technique in combination with other analysis may be able to provide links between objects.

XRF machines work by producing an X-ray that excites atoms on the surface of the object, in order to stabilise themselves these atoms then release energy which can be detected by the machine. Each element produces energy with a different wavelength, thus allowing the machine to tell them apart. The machine also counts each instance of energy being released for each element, thereby producing a spectrum with peaks whose size are dependant on the quantity of each element present. This can then be converted into percentage values for the composition of the object.

I recently analysed two hoard objects and presented here are the findings. I also captured this on a video blog, so you will be able to see me using the analyser.

K353 Gold Pommel


Pommel of a sword or knife

K353 Overview after edited

The object shown in fig. 1 above is the pommel of a sword or knife, a pommel is the part found at the very end of the handle. This pommel has been decorated with twisted wires in a technique known as filigree. The bright yellow colour of the metal suggested that it was very pure gold with little silver or copper present that can make it slightly pink or white. Analysis showed this to be the case with a large gold (Au) peak present on the spectrum shown below in fig. 2 and only very small peaks for copper (Cu) and silver (Ag). The result given was 94% gold which is around 22 carats. This is far higher than is usually used in most modern jewellery as gold is very soft unless it is alloyed with other metals. This may explain why the surface of this object is very worn in places.

Graph - XRF spectrum for the surface of K353

Fig. 2 XRF spectrum for the surface of K353

K353 bottom of pommel

Fig. 3 K353 Overview bottom after edited

On the interior of this pommel was a green powdery layer, this is indicative of copper corrosion and shows that the gold was placed around a copper alloy core to strengthen it. This is what allowed the Anglo-Saxon craftsman to use a thin layer of high carat gold without it being easily bent out of shape. Analysis of this green material did indeed produce copper but also tin (Sn) with a fairly large peak. An alloy of copper and tin produces bronze, a very strong metal that was widely used for many different objects although it had been superceeded by iron and steel for blades by this point. The high peak for iron (Fe) may well be from the soil which has not yet been excavated from the pommel interior at this point.

Graph - XRF spectrum for the interior layer of K353

Fig. 4 XRF spectrum for the interior layer of K353

K64 Silver Edging

The object shown below in fig. 5 is a decorative silver edging that would have been riveted to a larger object. It is inlaid with a substance known as niello; a compound containing silver and/or copper and sulphur. It has also been gilded along the left side.

Decorative silver edging

Fig. 5 K64 exterior after captioned edited

The spectrum produced on the silver surface inlaid with niello shown in fig. 6 indicates that the surface is mostly silver as would be expected. Although sulphur will be present as part of the niello it is unlikely to appear on this spectrum as it is a light element which produces low amounts of energy which may be absorbed before it reaches the XRF’s detector. This is where the British Museum science department will come into play, using their analytical equipment to further test this object’s composition.

Graph - XRF spectrum for the silver surface of K64

Fig. 6 XRF spectrum for the silver surface of K64

The gilded surface shown on the left side of fig. 7 produced the spectrum shown in fig. 8 with large peaks in both gold and silver. This is because it is not only detecting the very thin layer of gold but also the silver underneath. This may be because the gilding is worn and scratched in some areas or simply because the gold is so thin that the X-rays can penetrate it to the silver beneath. With gilded surfaces one element we particularly look for is mercury, as this was commonly used to aid the gilding process at this time. However it would be present in such small quantities that our XRF may not be able to detect it even if it were present. This is another area that can be looked at in the future at the British Museum.


K64 Gilded surface detail

fig. 7 K64 Gilded surface detail after x7.5 edited

Graph - XRF Spectrum for the gilded edge of K64

Fig. 8 XRF Spectrum for the gilded edge of K64

To conclude, although our analysis is relatively quick and simple it already gives us a good idea of the composition of these spectacular objects. As detailed briefly in this blog, this data can tell us more about the methods and materials involved in their manufacture. However further work is required and will be ongoing for many years, both here and in other institutions such as the British Museum, to glean as much information as possible from this highly significant find.

Graeme McArthur, Intern for the Hoard Conservation Team at BMAG and student at the University College of London.








Great Blog! By far one of the more complete survey results that I have come across. I was wondering if you had several pieces of, lets gold or copper artifacts, and they were all cut from the same source,lets say a copper pot, would all the artifacts have identical XRF results or would they vary slightly?

I’ve recently read that the transparent red stones remaining in the Westness brooch are glass, rather than garnets as is often assumed. I’m wondering if any of the many “garnets” in the Staffordshire hoard have been analyzed, and if so, what the results are.