Laser cleaning the Abergavenny Hoard: Silver Coins from the time of William the Conqueror (45-51)

Laser Cleaning the Abergavenny Hoard: Silver Coins from the Time of William the Conqueror

Mary Davis

Department of Archaeology & Numismatics, National Museum ofWales, Cathays Park, Cardiff CF10 3NP, Wales, UK.

Abstract. The Abergavenny Hoard is a recently discoveredcollection of late Saxon and Norman silver coins, the detailsof which were badly obscured by corrosion products and ironconcretions. A Q-switched Nd:YAG laser, using near infraredradiation at 1.064 μm, was employed to clean the hoard Detail retained on the surface of the coins after lasercleaning included rough out marks and polishing marks fromthe original die, as well as the legend. From this evidencesurface damage appears to have been minimal, and thetreatment very successful.

1. Introduction

In April 2002 a scattered hoard of 199 silver pennieswas found by metal-detectorists in a field nearAbergavenny, Monmouthshire in Wales. It is not clearwhether they had been deliberately hidden or lost. Whenit was discovered, the hoard was heavily encrusted withiron deposits; within these were preserved traces offabric suggesting that the coins had originally beenheld in a cloth bag. The total value of the hoard cameto sixteen shillings and seven pence (16s 7d, or£0.83p) which would have represented several months’wages for most of the population. The hoard wasdeclared as ‘Treasure’ under current legislation forEngland and Wales, and has been acquired by theNational Museum of Wales (NMW).

Fig. 1. Some of the coin hoard before treatment. (Averagediameter of coins = 19mm)

The hoard includes coins of the Anglo-Saxon king Edwardthe Confessor (1042-66) and the Norman king William theConqueror (1066-87). Anglo-Saxon and Norman coins forman unique historical source: each names its place ofminting and the moneyer responsible. People had accessto a network of mints across England (there were nonein Wales) and every few years existing money was calledin to be re-minted with a new design. The King took acut on each occasion.

Fig. 2. Microphotograph (x13) of the woven fabric preservedwithin the iron incrustations, possibly the remains of a bagor purse and SEM micrograph of the constituent bast fibres(scale bar = 100µm)

2. Condition and Preliminary Cleaning

When the coins were discovered they were covered iniron concretions; these were obscuring most of thedetail, and were much harder than the underlyingsilver. It was important to remove the concretions toobtain valuable numismatic information, but mechanicalcleaning methods such as using a scalpel would havedamaged the underlying silver. Chemical cleaning wasalso tried; tests were carried out using 10% citricacid and 10% oxalic acid. The oxalic acid was much moreeffective and succeeded in removing some of thecorrosion and in separating many of the coins but stillfailed to shift the majority of iron incrustations fromthe coins’ surface and much of the detail remainedimpossible to decipher. Several of the coins were alsocracked or broken; physical pressure or chemicalcleaning would have caused further damage to these. Although laser cleaning has been successfully usedon metal sculptures made of aluminium [1] and lead [2],cleaning metals, and especially those fromarchaeological contexts has been approached with somecaution; the complexity and variations of surface dirtand corrosion over altered metal cores means that theeffects of laser treatment need to be evaluated foreach situation. However, tests carried out by Pini etal [3] did give some indications that laser cleaningcould be successful on these coins. In addition,some experimental tests had been carried out at NMW in2000 on a variety of materials using a Q-switchedNd:YAG laser. One successful treatment had been the

removal of green corrosion products from the surface ofa very delicate piece of gilded bronze [4]. Laser cleaning in this case looked like a possiblesuccessful treatment for several reasons: the coinsshowed a significant contrast between the dull red,black and green corrosion products on their surface andthe underlying white metal. This suggested that theremoval of the corrosion by the laser could be selfdetermined when the overlying incrustations and oxideswere ablated and the white metal exposed, as had beenseen for the cleaning of both archaeological tin andsilver by Pini et al [3]. Another feature which couldcontribute to the success of laser cleaning in thisinstance was the fact that the incrustations on themajority of the coins were not a result of thecorrosion of the metal itself but of firmly adheredcontamination.

3. Observations on Cleaning

Two coins were taken to the Conservation Centre at theNational Museums Liverpool, where some preliminarytests using a Q-switched Nd:YAG laser with a greenlaser at a wavelength of 532 nm and infrared (IR) laserat 1064 nm were undertaken. The short pulse duration(5-10ns) was important to avoid heat damage, as the‘laser interaction with the metals has to be shortenough to produce a fast removal of encrustation,avoiding heat conduction into the substrate’ [1]. Thegreen laser appeared to have a very damaging effect onthe silver, removing the concretions but altering theappearance of the silver surface which looked moreporous, less shiny and much less compact; it alsodestroyed some of the detail. The IR laser was muchless damaging, but the cleaning still appeared a bittoo harsh.

NMW hired a Q-switched Nd:YAG laser to experimentfurther with cleaning the coins. In order to reduce theenergy levels emitted, an aperture was made which keptthis at 33mJ – one third of the normal lowest rateemitted by the machine. The working distance was mostlybetween 20-25 cm. This was varied: for softer reddishbrown iron corrosion products the laser was movedfurther away; too close and it tended to reduce thebrown to a harder black substance. For very hard, thickblack incrustations the laser was moved closer: thiswould make the beam smaller and so it could be moreaccurately directed, and it did help to quicken thedislodging of larger lumps. Water was applied to thesurface and this appeared to help considerably in theremoval of both types of corrosion, small lumpssometimes detaching themselves from the surface. Movingthe object round and cleaning it from a slightlydifferent angle also helped dislodge corrosion fromdetailed areas and where more stubborn incrustationswere present.

Fig. 3. A coin prior to laser cleaning (William I, Londonmint)

The sound made by the interaction of the laser with thesurface became quite a good indicator of how drasticthe treatment was, and helped define the best workingdistance on particular types of surface. The surface of the silver itself was superficiallyaltered by the use of the laser, whether or not waterwas used. It acquired a whitish coloured bloom, andthis inturn also obscured the detail.

Fig. 4. The same coin after laser cleaning showing the creamcoloured ‘bloom’ on the surface and the cleaned coin afterthe bloom had been removed

This bloom was quite loosely attached to the surfaceand could be rubbed away. In practice, this layer wasremoved from the coin in the lab with a soft glassbristle brush under a low-powered optical microscope.Cotton wool swabs and other mild abrasives worked – butthe glass bristle brush required minimal use and wasable to clean around the detailed legend of the coin invery little time while leaving no discernible scratchesto the surface under low-powered magnification. At thispoint any tiny remaining incrustations which the lasercleaning had failed to remove could often be flickedoff with a scalpel.

When this layer had been removed the silver was verycompact and shiny. The detail retained on the surfaceof the coins was excellent and included rough out marksand polishing marks on the original die as well as thelegend. From these results surface damage appeared tohave been minimal, and considerably less than wouldhave occurred with mechanical cleaning or chemicaltreatments. It was also relatively time-efficient.

Fig. 5. Two cleaned coins; the first (Edward, Taunton mint)showing rough out marks from when the die was made which arestill visible on the struck coin, the second (Edward, Londonmint) with remnant silver sulphide corrosion on the surface.

The laser was most successful at removing ironconcretions adhering directly to the silver surface. Italso removed green copper corrosion products from thecoins where this had occurred. It did not remove silversulphide so well. This was only present on a few of thespecimens, but as it had occurred as a direct result ofcorrosion of the silver itself, the metal in theseareas had a slightly rougher, pitted surface. The laserwas not able to remove all the darker corrosion fromindents and pores, and a greyish sheen was left on someof the coins.

The laser was a particularly good cleaning methodwhen it came to dealing with cracked and broken coins.Pressure applied by mechanical cleaning or the effectsof chemical cleaning could easily have resulted in lossor damage to these much weakened specimens. Becausethere was no physical pressure, broken edges andcracked areas could be cleaned confidently and maximumdetail retained.

Fig. 6. A cleaned coin with a crack running down the centre(Edward, London mint) and a corroded and broken coin duringlaser cleaning (William irregular)

The surface of the coins was examined under a scanningelectron microscope with an energy dispersive X-rayspectrometer (SEM-EDS) to look in more detail at thebloom and the cleaned surfaces. SEM-EDS showed nosignificant chemical alteration to the silver and thebloom is likely to have been an oxide; under very highmagnification it was composed of a layer of looselyformed crystals and abrasive marks were visible wherethese had been brushed away. However, under lowermagnifications there was very little difference betweenthe areas with the bloom and those which were fullycleaned – except that the detail was clearer.

Fig 9. SEM micrographs of a coin under lower (scale bar =1mm) and higher magnification (scale bar = 20µm), after lasercleaning and with half with the oxide bloom still present

4. Conclusions

Hoards from western Britain are rare, and theAbergavenny Hoard has produced many previouslyunrecorded combinations of mint, moneyer, and issue. Itincludes products of 36 identifiable mints, as well assome irregular issues which cannot at present belocated. The numismatist felt that this was a very successfultreatment for this particular hoard, where a largeamount of information was gained from cleaning thecoins. Ethically, it did less damage than many chemicalmethods would have done, and was deemed an appropriatemethod of conservation for these types of artefact.

Acknowledgements. Martin Cooper and Sam Sportun (Theconservation centre, National Museums Liverpool)offered help, advice and undertook initial cleaningexperiments. Andy Charlton (Lynton Lasers) gave muchsupport and advice for the project and supplied the

‘home-made’ aperture. Edward Besly (Numismatist, NMW)encouraged the work and supplied numismatic informationand Bob Child (Head of conservation, NMW) bothencouraged the work and obtained funding for the lasertreatment.

References

1. M Cooper, in Laser Cleaning in Conservation, Edited byM. Cooper 74, 1998. Butterworth-Heinemann, Oxford 2. A Naylor, in Journal of Cultural Heritage, Proceedingsof the International Conference Lacona III, Lasersin the Conservation of Artworks III, Edited by R Salimbeniand G Bonsanti,, S145, 2000.3. R Pini, S Siano, R Salimbeni, M Pasquinucci, M miccioin Journal of Cultural Heritage, Proceedings of theInternational Conference Lacona III, Lasers in theConservation of Artworks III, Edited by R Salimbeni and GBonsanti,, S129, 2000.4. Y Ever-Hadani, Unpublished MA Project, University ofDurham, 2000.