European Commission

Ion beam study of artand archaeological objects

A contribution by members of the COST G 1 Action

Edited byG. Demortier and A. Adriaens

e,-v-

Directorate-General for Research

2000 EUR 19218

sn

O£Z-11 °dd '9661 '~n!A~S 'Rn!A~S ~p PRP!Sl~A!UO S~UO!:>R:>!Iq"d

~p OpR!lR:j~l:>~S iosp~ 'oq:>RUlR;)-Z~W9D or pOR RzWIRds~"M °V °W) ,(J.JaULoal1l{.J.Ll1 PUl1 SJ.Ll1 OJ sanb!ul{.JaJ

S!S,(Zl1Ul1 ULl1aq UO!fO suo!Jl1.J!lddyu! ,SUO!:j:>Rl~:jU! ~IdUlRs UlR~g °SlO:jRl~I~:>:>R :>!:jR:jsO.I:j~I3, 'O;) or 'ORla (Jo LL£-£££ °dd (L861) (yO:jS!lg 'l~~I!H UlRPV 'op~ '~W!lD °A\ oD pOR :j:JRA\ od) sULl1aqO.L.J!UL UO! '(8.Laua-l{8!l{

fo suopl1.J!lddl1 PUl1 sazd!.Ju!.Ld U! ',A~oIo~Rq:>lR U! s~qoldol:>!W lR~I:>nu 10 sUO!:jR:>!IddV, 'oD 'l~!:Il°W~a (9)°8C;Z-1 °dd (Z861) L61 spol{Jaw PUl1 :LJSUI 'l.JnN ':jU~wplRqwog ~I:>9lRd

p~~lRq;) Aq S!SAIRUV IR:>!W~q;) uo ~U!:j~~W IRUO!:jRW~:jUI pUO:>~S ~q:ll° S~UW~~:>old '(op~) oD 'l~!:Il°W~a (.)° L91-1 °dd (9861) vI g l{.J.Ll1asaH s.J!s,(l{d U! spol{Jaw PUl1 :LJsuI 'l.JnN ',A~oIo~Rq:>lR pOR

S:)lR ~q:I U! S!SAIROR UlR~q uoI, :(op~) oH °W 'U~Sl~PUV 'oW 'nu~w 'oq;) 'ZI!~H 'oD 'I~sWV 'oq;) 'l~!ORqR'l (~)

°09-1v °dd '9661 '~n!A~S 'Rn!A~S ~p PRP!Sl~A!UO S~UO!:>R:>!Iq"d

~p OpR!lR:j~l:>~S iosp~ 'oq:>RUlR;)-Z~W9D or pOR RzWIRds~"M °V °W) ,AJ:I~WO~Rq:>lR pOR S:)lR O:j s~nb!U

-q:>~:j S!SAIROR UlR~q UO! 10 suo!:jR:>!IddV, U! ,suo!:j:>R~llR~I:>nu uo p~SRq S!SAIROR IR:ju~w~I3, 'ov 'SSf)! (E)°ZIv-L9£ °dd (C;L61) '(Z)£1 Md!ltaH '(8.LaUH .J!ULOJY 'od ':)lRpog 'oD 'l~!:Il°W~a 'oD '){:>u!ÙUo:>~a (.)

°mUlRN OOOC;-g 'xIRd RI ~p ~UlRa-~.I:j°N S~l1t!:j!Sl~A!UO s~:jyn:>Rd 'NMV'l (J

'CJ URla Aq p~qp:)s~p Sl01Ul~I~:):)U ~I:)~d U!p~:)npold ~lU A~q.r 'A1!A!1:)U IUnp!S~l AUR P!OAU 01 CA~W £: 01 I) q~nou~ MoI S! S~I:)!1IP,d

1U~P!:)U! ~q:) 10 A~l~U~ ~q.r 'CJ SW01U 1q~!I APll1S 01 VMN pUR 3DId 'SW01U AAU~q pUR

wn!p~W ~U!U!U1UO:) SIUP~1UW APll1S 01 ~Iqu1!ns u~q:) ~lU SH"M pUR 3XId .s1u~w~I~ AAU~q

10 UO!1nq!.I:Js!p ~q:) uo UO!1UUll01U! AJB1U~W~Idwo:) ~A!~ S~W!1~WOS AUW SH"M 'CL-S) Cwrl £:

01 UMOp l~1~Wt!!p Wt!~q) AIqw~ssu ~qoldol:)!W U U! u~){Bp~pun ~q AUW sUO!1U~!1S~AU! ~q:)

'AJBss~:)~u 11 .UO!1U!pull! ~q1 ~U!lnP ~lnss~ld :)p~qdsow1u 1U 1d~){ 'SUO!SU~W!P ~~lUI 10

s1:)ul~1IP, 01 p~nddu ~q AUW 3DId pUR 3XId .Ci') S~!1!Iunb ~s~q:) 10 1S0W ss~ssod CAIsno~u

-U1Inw!s p~sn S~W!1~WOS spoq1~w ~~lq:) CVMN) ~I:)!1IP,d p~~lUq:) U 10 UO!1:)~1~P ~q1 ~U!A

-IOAU! C£) SUO!1:)u~llU~I:)nu puu C£) CUO!ss!W~ AUl-UWWt!~ p~:)npu! U010ld) 3DId 'CJ CUO!S

-s!W~ AUl-X p~:)npu! U010ld) 3XId 'I~A~I :)!dO:)SOl:)!W 1U U~A~ '~U!IdWt!s ~UR 1noq:)!M 1:)ul

-~1IP, ~q:) 10 s1IP,d 1U~l~11!P 10 UO!1!sodwo:) IU:)!W~q:) ~q:) ~A!~ 01 ~Iqu pUR ~A!1:)nl1s~p-uou

AIU!UW ~q 01 ~l!nb~l spoq:)~w IU:)!1AIUUU ~s~q:) SW~1! IU:)!~OIo~uq:)lU ~nb!lln 01 P~!Iddu

u~qft\ 'U~WS11ul:) 1U~!:)UR 10 d~SURW){lOM ~q:) pUR1sl~pun pUR s1:)ul~1IP, 10 ~:)UUU~AOld~q:) U!UIdx~ 01 p~sn MOU ~lU SIUP~1UW IU:)!~OIo~uq:)lU 10 S~SAIUUR IU1U~w~I~ ~A!1u1!1URn()

uon;>npO.I:)uI

"Ç XIPu~ddyuI P~1sII u~~q ~Aeq S1U~1UO:) S1I "){ooq sguW~~:)old ~IIIA~S ~q11° :Jjed SI uoI1nqIIJuo:)1U~S~ld ~q,L "S1U~W~I~ AAe~q guIuIe1uo:) sl~AeI IeI:)IJl~dns APll1S 01 u~q1 pUR q1d~p uI A1

-I~u~gowoq ~q1 I°j),uo:) 01 Inl~sn AIuIew ~nbIuq:)~1 AJB1U~W~Idwo:) e sI (S81I) Ado:)soj),-:)~ds gup~11e:)S){:)eq ploJl~q1n"M "P~SAfRUR ~q Aew D'l ueq1l~qgIq 1qgI~M :)!W°1e ue q1IM

1u~w~I~ AUY l~A~I :)Ido:)SOl:)!W 1e U~A~ 'guIIdillBs AUR 1noq1IM 's1:)el~:Jje 10 s:Jjed MOllBU10 uoI1Isodwo:) fRll1:)e ~q1 's~1nu!W M~l e uI 'uIU1qo 01 ~IqIssod sI 1I 'A ~W £ URq1l~MoIsI Agl~U~ ~SOqM SU010ld guIsfl "s1:)el~:Jje 1u~I:)UR 10 SISAfRUR Ie1u~w~I~ ~q1l01 poq1~w

~1em:):)e pue PIdel AJ~A e ~q 01 p~AOld u~~q seq (3XId) uoIss!W~ Ael-X p~:)npuI u010ld

:J;}U.I:JsqV

(J l~!:J.I°W~a "O

sasod.lnd IB~!~OloaBq~.lB .l°J Smi PUB ~XId Jo IB!JUaSS~

9 x!pu~ddV

Basic principles of PIXE

The irradiation of alI materials with charged particles leads to an ionisation of electronicshells and consequently to emission ofX-rays whose energy is characteristic ofthe atomoThe technique using this phenomenon induced by protons is called PIXE. In principle,the analysis of alI the elements may be performed but, practically, only elements with anatomic weight greater than 20 may be quantitatively determined. Elements with an ato-mic weight lower than 20 afe more accurately analysed by detection of y-rays producedby excitation of the internaI structure of the atomic nuclei (PIGE) (3) or detection of char-ged particles arising from the interaction between the incident particle and the nucleus.The characteristic K X -rays of light elements afe indeed so highly absorbed in thesamplè that corrections cannot be made to obtain sufficient accuracy with PIXE.

Electrons 20 keV Protons 2 Me V

~m ~m

\)

Figure 2. Comparison of X-ray spectra obtainedby irradiation of apathite with 20 ke V electrons

and 2 MeV protons.

PIXE is very similar in its principles toX-ray analyses with an electron micro-probe. The main differences lie in: thetrajectories of the incident particles (pro-tons penetrate in any material by graduaIdecrease of their energy along a straighttrajectory while electrons zigzag along acomplicated path), the depths of the ana-lysed materiallayers (5 to lO Jlm underthe surface for PIXE but only 1 Jlm withconventional electron probes) (Figure 1),the sensitivity (2 orders of magnitudebetter for PIXE because proton brems-strahlung is nearly absent) (Figure 2).

126

al

'£6£ '(9861) v£ 'S;}Iq1!,L 1!:}1!0 ;m;}I:>nN pUl! 1!:}1!0

:>!UJmy pUl! 'çÇZ '(Ç861) ££ 'S;}Iq1!,L 1!:}1!0 ;m;}I:>nN pUl! 1!:}1!0 :>!UJo:}y "W 'Ul!~l.I.mH pUl! '0'0 'U;}qO;) (6)'ZVç-££ç 'dd (9L61) SdN6-ÇLll H;)9L '('P;})Ul!~JOW "11 pUl! Ul!~~no "1'C 'SJO:}1!J;}I;}:>:>Y IJ1!wS

Jo Suon1!:>nddy J1!~snpuI pUl! :>ynU;}!:>s ;}q:J uo ;}:>U;}J;}JUO;) q:JJn°d ;}q:J Jo S~U!P;};}:>OJd "0 ':'{:>U!UUO:>;}O (8)

(A8'f) ~5J8U3

9 ç .~"(wn!~I~H-~A~JH)

pJRqs ){:>ulq U Jo w~:>~ds '.iIXId 'uv ~Jn~!d

-<m:c.

no

~'S~1}1SU~}U1 JO U01}~!lUA ~3'lP.I ~lp

~}~1:J~ldd~ O} l~plO U1 ~fR:JS :J1U!lp!lU3'°I ~

U1 U~A13' S1 S1~ fR:J1:JJ~A ~lp }~lp ~}ON '~l}

-:J~ds A~l-X fR:J1dA} ~A13' ~M 'v ~ln3'1d uI

lq~!~/'t\ q~~ JoS}U~W~I~ ~;)UIJ puu 10U!W 10J A}!A!}!SU~S}U~!;)YJns U ~A~~;)U O} }U~!U~AUO;) ~lOW

~m SAU1-X 'l .Oç = Z O} dn S}U~W~I~ Jo

S~SAIUUU ~;)!nb 10J ~U!}S~l~}U! AIUO u~q}~m S~!}!SU~}U! UO!ss!W~ AU1-X )I .08 O}

ÇI W01J S~!l1?A l~qwnu ;)!W°}U }~~m} ~q}

u~q/'t\ ~pm!U~UW JO Sl~plO ç AIm~u AqAmA SUO!};)~S-SSOl;) UO!SS!W~ AU1-X ~SOq}

}Uq} ~Al~sqo UU;) ~1'i\ .£ ~lM!d U! U~A!~

~m S~U!I AU1-X D'l 10 D)I ~;)npold O}

SPI~!A }U~;)S~lOnlJ ~q} 10J P~};)~llO;) puuA ~W £ O} dn suo}old }U~P!;)U! 10J SUO!}

-;)~S-SSOl;) UO!}US!UO! Jo s~nYUA P~};)~I~S

'S}U;)W;)I;) snoIlB/I. uo A;)W I:: O} dn suo}oJd f.q p;):JnpoJd

.SgIq~:J gA!:JSn~qxg AIIP,gU U! (6) gIq~I!t1A~MOU gIP, m~p g:J~ln:J:JV 19pOm "MSSd;)'3

:S:J:Jgllg :J!:JS!A!:J~Igl ~U!pnI:JU! SgI!:J

-:JgÇOld :JUgp!:JU! gSgq:J Aq Sg:J~:JS AIP.UO!:J~:JS

:J!Ul°:J~ gq:J 10 UO!:J~qJnpgd gq:J 'sngI:Jnu

gq:J 10 PIgy qmoIno:J gq:J U! gI!:J:JgÇold

gq:J 10 g~~q:J A:J!:J°IgA gq:J P~ UO!:J:Jgugpq:Jdgp gq:J 'ssoI A~lgUg gq:J :Juno:J:J~ o:JU!

Sg){e:J uo!:J~In:JfB:J g:JgIdmo:J gqJ. 'sIgpom

fBlgAgS Aq AIfB:J!:JglOgq:J pg:J:J!pgld ngMAlgA OSfB gIP, AgqJ. 'pgqS!Iqnd Uggq gA~q

s:JUgmgIg fB:J!Ulgq:J SnOlgmnU UO suo:Jold

q:J!M Sg!:J!SUg:JU! UO!:J:Jnpold A~l-X lOl ~:J~P

UO!:J:JgS-SSOl:J pglnS~gm 10 :Juno~ :JS~A V

.(8) mnn;)~A U! p~;)npOl}U! ~U!~q moli (.;)1~ 'Sm~1! fJ~n~M.~ç~1~;)P1U! 'l~d~d ~){!Ù S1;)~Çqo ~I!~~lJ lO (.;)1~ 'S~U!1U~d 'S){ooq 's1u~wq;)md ~){!Ù ~~mIfJ~A ~U11u~A~ld '~lnss~ld ;)p~qdsoW1~ 1~ p~WJoJl~d ~q f..~m s~sf..fBU'B 3XId '~loWJ~qJJnd

SAUJ-X Dì puu D)I 10 S~9!su~:}U! ~A9uI~~ "£ ~Jn~!d

[1\8V'11] J.6.J9ua uo~o.Jd :1U8PIOUIr Z L O

Red Shard (Braive, Belgium)

103 ~.~ ,l'-

fl'f f'iiì 102 I ~"E ~. j::J ,O ~

Iu u

; I

<Il~

101

Figure 4b. PIXE spectrumof a red shard (Braive-

10° Belgium):o 1 2 3 4 5 6 7 8 9 10

Energy (keV)

Clay (Braive. Belgium)

103 ~ in ~ ~.J. ~ o L.;: ~~~ ~

I I I : 'i:~ ~ ILth u

,l'-"iJ: 102 Il ~c::J ~'

.§. ~

~ I

10'

Figure 4c. PIXE spectrum10° of clay of Braive.

o 1 2 3 4 5 6 7 8 9 10

Energy (keV)

Excavated Bronze (Mali)

"" "".f 'f .3 'l'".. .s:l'

103 "".3 li ~ .3.3 I

I . n ,..-J -J -'li .c .c'iiì I:!: '!;Il. Il. Il.

§ 102 ""I I I I )t;" ~

O I:!: o ,U C C

~ (I) (I)

~ 101 I

Figure 4d. PIXE spectrumof an African bronze

10° (Mali).o 5 10 15 20 25 30

Energy (keV)

128

Silver Coi n (XIX Century, France)

~ ~ ," " ~103 U U -

. = ,..-' -' -'.ò .ò .òDo Do Do

~ ~ ~$l 102 -J -'c: ' ,~ 01 01O :I: :I:

~uW~

10'

Figure 4e. PIXE spectrum

of a silver coin (France,

10° 1805).o 5 10 15 20 25 30

Energy (keV)

Zoomorphic Pendant

10'2 MeV Praton Beam Ka

I Cu

KpI La Au

10' Lp

LyAg

Ka KpI

:2 103Q)

~

102

Figure 4f. PIXE spectrumof a gold jewellery item

10' (Mesoamerican).o 5 10 15 20 25 30

Energy (keV)

Meteorite

103~"

:Il" di.!. LL.~. U.!. :Il"U eLL.

102 ; .~ x ~'" ,.L..Lr- Zc: \:Il" ~ ..L ~U 1-.-

~ Z

~ 101 I I

Figure 4g. PIXE spectrum10° of a meteorite sample.

o 5 10 15

Energy (keV)

129

Each element ionised with the incident protons gives several X-ray signals. When Klines afe selected far the analysis we may choose two different signals Ka and Kj3 withknown intensity ratios. If L lines afe selected (for heavy elements) ODe may select La,Lj3 or L y, depending on the eventuality of coinciding energies from various elements;see below far Zn and Au interferences.

The analysed depth of the irradiated artefact is less than l O ~m, even for particles cros-sing 25 to 30 ~m in the material due to the decrease ofX-ray production with decreasingenergy of the projectiles. Archaeological applications afe then restricted to non-corrodedmaterials or to drillings involving partial destruction of the sample. PIXE is very fastand accurate for the non-destructive characterisation of the main components of narrowregions on gold jewellery artefacts in a non-vacuum 'milliprobe' arrangement: goldjewellery items afe indeed well preserved and not affected by a surface corrosion in adepth comparable with the thickness of the analysed layer.

Experimental procedure for PIXE

Figure 5. 'l'ypical arrangement installed at LARN (Namur) for non-vacuum PIXE. The incident protonbeam crosses a thin foil of Al before reaching the target sample situated at l cm in the air. The X-rays ofthe sample afe collected in the main detector. A lateral monitor collects X-rays of argon (in air) which gives

relative intensities of the incident proton beam.

Figure 5 illustrates the experimental set up for non-vacuum PIXE at LARN (Namur,Belgium). The proton current from the accelerator is maintained at a low level (0.1 tolO DA) in order to keep the total counting rate of X-rays compatible with the best ener-gy resolution of the solid state Si(Li) detector (500 to 700 counts per second). The pro-blem of different counting rates in characteristic X-far peaks of main and minor ele-ments can be solved bere by using appropriate filters. Foils of selected materials afe usedto reduce selectively the X-rays of the main component, and therefore enhance the rela-tive intensities of alI other useful X-far lines.

130

1£1

p~:)uln:)Iu:) :)SlIj ~IP. SU01:)Ul:)U~:)UO:) (I) :u01:)uln:)Iu:) ~q:) u1 U01:)U:)1:)S1qdos 10 sd~:)s JUl~A~S

~u1sn P~u1UIl~:)~p u~q:) ~IP. s~Idwus ~q:) 10 S1SAJUUU ~q:) ~ullnp P~:):)~II°:) s~r:a1su~:)u1 AUl- X ~q:) UV .~:)u~l~l~l U su u~){U:) s1 punodwo:) JU1:)l~WWO:) UMOq V .P~Mouol ~q O:) u~q:)

~AUq U01:)1sodwo:) JUn:):)u ~q:) u1 U01S1:)~ld ~q:) ~U1A~1q:)U u1 sd~:)s JUl~A~S .s:)u~w~I~ l~q:)OlJU 10 s~1:)1su~:)u1 AUl-X ~q:) s~Y1Pow s:)u~w~I~ OM:) ~s~q:) uuq:) ~lOW ~u1ll1U:)uo:) AOUU uu

u1 s:)u~uodwo:) OM:) AIUO 10 SU01:)Ul:)U~:)UO:) ~A1:)uI~l ~q:) u1 U01:)U1IP.A AUV .SPI~1A ~:)U~:)S~l

-onu AUl-X A.rnpuo:)~s 'U01:)wosqu AUl-X JU1:)U~l~JJ1P 'u01:):)npold AUl-X lol u01:):)~S-SSOl:)'~~UUl ~I1:):)~rold ::)uno:):)u 0:)u1 sl~:)~wuIP.d JUw~:)u1 :)~~IP.:) ~q:) UU ~U1){U:) s01:)UJ U~IP. ){U~d

~q:) woll p~:)uln:)Iu:) ~q O:) ~AUq su01:)Ul:)U~:)uo:) Iun:):)u ~q,L .SU~IP. ){U~d ~Ar:auI~l 10 u01:)uu1W

-l~:)~p ~q:) O:) p~:)np~l s1 :)u~w~rosu~w q:)u~ 'Apu~nb~suoJ .s:):)ul~:)IP. P~1Pn:)s ~q:) lol p~snsu01:)1Puo:) JU:)1IJ~WO~~ ~q:) u1 p~:)U1pUll1 ~q AUW puu '~IquI1UAu AlJU1:)l~WWO:) ~IP. s:):)ul~:)IP.

JU:)1~OIo~uq:)IP. u1 :)s~l~:)u1 10 s:)u~w~I~ ~q:) 10 :)sow ~U1U1U:)UO:) S~Idwus ~:)u~l~l~l w~poW

'PI\?,MJ01:)q~fBJ:)S ~lom sI s~p~:)Rm

~:)u~l~l~l UMOU:){ q:)IM uospRdmo:)

'(lO:):)~:)~p ~q:) pUR :)~~I\?,:) ~q:) U~~M:)~q

sIRp~:)Rm ~q:) Ul SAR1-X 10 UOl:)wosqR

AIufBm pUR S~I~l~U~ AR1-X :)u~1~111P

101 A:)U~I:)Yl~ s:)1 'lO:):)~:)~p ~q:) 10 ~I~UR

PIIoS 'A:)ISU~:)Ul mR~q uo:)old) suoI:)-:)~1l0:) IR:)P:J~mo~~ ~q:) I~ ~:)RIn:)~:)

O:) ~uIARq PloAR O:) l~plO UI 'uol:)RIn:)~:)

Aq SISA~UR ~Al:)R:)I:)URnb MoI~ O:) ~IqRI

-fBAR ~I\?, R:)RP ~:)lsAqd InJ~sn ~q:) IIV

'l~nY ;)U!Z ~ JO ~;)U~S~ld ~~

U! p~~;)~nO;) ~;)~J~~ PIog ~ Jo w~;)~ds 3XId '9 ~m2!d

Jaqwnu lauue4:>v20L. B9L9Ç;-ao 21-~

's){Ugd gdu:)sg pue wns 'uondJosqu

AUl-X 'SUO!~:)gS-SSOl:) AUl-X :SU q:)ns

~uno:):)u O~U! Slg~gillBJBd ItIlgAgS Sg){U~

Sg!~!SUg~U! AUl-X ItIn~:)u 10 uo!~u~nd-ma:) gq.L 'SUgJB ){Ugd g~uln:)ItI:) o~ pue'SUO!~gl ){Ugd lItI Al!~UgP! o~ 'A g~ 9 o~ dn

wnnu!~uo:) u o~ gs!l ~U!A!~ (U!~!lO 8un]

-l/v.JjSSWiJ.Jq UOIJ:)gIg Ampuo:)gS wollAIU!UW) puno1'ii~:)uq ItIn~:)u gq1 ~:)uIJqnso~ 19p1O U! pgu!illBXg S! wru~:)gds AUl-Xq:)Ug 'lg~ndwo:) IItlws u 10 dIgq gq1 q1!A\

'l~:JIY ~A!:J~~I~s ~q:J U! p~~npoldS~U!1 uz 10 ~~u~s~ld ~q:J U! '~V 'uV 'nJ

~u!U!u:Juo~ :J~ul~:jJU uu 10 (A ~W 8"Z=dg:)

UO!:JU!PUll! uo:Jold ~U!lllP p~u!U:Jqo Uj:J

-~~ds [t'~!dA:J s~:Juj:Jsnn! 9 ~ln~!d '~lq!S

-sodw! u~q:J S! :J~~lU:J ~q:J U! (s):Ju~w~l~

q~ns 10 u°!:Juu!Wl~:J~P ~q:J 'wnJ"!~~ds ~q:J

U! lU~ddu (s)l~:JIY ~q:J 10 (s):Ju~w~l~ ~q:J

10 SAUl-X ~!:JS!l~:J~UlUq~ ~q:J SV .~~u~~

-s~lonlJ AUl-X [t'u°!:J!PPU UU O:J ~s!l ~A!~

AUW lO:J~~:J~P ~q:J puu :J~~lU:J P~:JU!PUll!

~q:J U~~M:J~q l~:JI!l ~q:J 10 ~~u~s~ld ~qJ.

using a linear mIe with the characteristic peak areas, (2) these results afe then used tocaIculate the stopping power, and the X-ray absorption in this new materiaI, (3) this pro-cedure is repeated until a new step gives a modification lower than the statistical accu-racy and (5) finaIly the secondary fluorescence is calculated. Even with a reference mate-riai containing elements whose concentrations may vary from a factor 3 compared withthose in the sample, the complete procedure of calculation on the X-ray spectra finallygives concentrations with an accuracy of 5 % or better. A control has been made usingPIGE. In this latter case, y-rays produced in the 20-30 first microns under the irradiatedsurface afe not absorbed in the material itself because of their high energy (100 ke V andmore). As cross-sections for y-ray emissions on medium and heavy elements afe around3 orders ofmagnitude lower than X-ray ones, this procedure takes a long time and is onlyused as an ascertaining procedure in limited cases.

Nuclear reactions leading to the emission of charged particles may also be used for thequantitative determination of light elements. Characteristic protons spectra induced bydeuteron irradiation allow depth profile analysis of Si, S, Al... in matrices containingheavy elements (3).

AlI the calculations in PIXE experiments afe based on the hypothesis that the material ishomogeneous in alI the depth of the analysed material (5-10 ~m). Nevertheless, hete-rogeneity in those thicknesses may be identified by comparing the relative intensities ofalI the characteristic X-ray lines of each elemento Elements concentrated at the surface(or under a surface layer) show Ka/Kf3 and La/Lf3 intensity ratios greater (or lower) thanthose calculated by the basic programme (lO).

Basic princlples of RBS

Rutherford backscattering spectroscopy is based on the elastic collision of an incidentcharged projectile with the nucleus of an atomo In this elastic process the incident andthe collided nucleus remain in their fundamental energy level after the interaction: thewhole kinetic energy is then conserved. Equations of conservation lead to a single rela-tion between the energy of the backscattered particle (emitted angle close to 180°) andthe incident energy:

2( M-m = Eo M + rr

Esc

where m and M afe the masses of the projectile and of the nucleus respectively (in thisprocedure m < M). One observes that Esc is proportional to Eo and is closer to Eo whenm is very light by comparison with M. If this parameter is to be chosen to distinguishscattered particles of mass m from various nuclei of mass M, m must be the highest aspossible but a1ways smaller than M.

(lO) Demortier, G., 'Differential PIXE analysis of Mesoamerican jewellery items', proceedings ofthe IBA-12 lnternational Conference, (Phoenix, Arizona, 1995), accepted far publication in Nucl. Inst7: andMethods.

132

Practically, far both acceleration and detection purposes, the incident particle is of a pro-ton (mass 1) or an alpha particle (mass 4). We then give in Figures 7 (a and b) the scat-tered energies of protons and a-particles on various nuclei. In this figures we bave alsotaken into account that the probability of elastic scattering is proportional to the squareof the atomic number of the target. This true proportionality to Z2 is expected far highZ values but large variations would arise far low Z nuclei (for example carbon and oxy-gen). Note the logarithmic scale in the ordinates.

Figure 7 a. Calculated energies Esc of backscattered protons (Ei = 2.5 Me V) on various masses.

100

IO"O""Q3

'>."O

Q)...

~"iOuIn

Q)>

.~

""Q3a:

0.1

0.01o QS 1 1.5 2.52

Esc in MeV

Figure 7b. Calculated energies Esc of backscattered a-particles (Ei = 2.5 Me V) on various masses.

133

The relative intensities of scattered particles on Au and O afe then in a ratio of 97.5. RBSis then 2 orders of magnitude more sensitive for gold than for oxygen. Comparing Figure6a with 6b, one clearly observes that the mass separation is better for a-particles than forprotons but this separation becomes often insufficient to distinguish neighbouring ele-ments like Cu and Zn, and even Ag and Au.

The study of materials using RBS is not straightforward for bulk material, due to thedecrease in energy of the incident projectiles when they penetrate in the material. Themeasured energy of a particle scattered in a layer below the surface is not given by thesimple formula due to the energy loss of the projectile when penetrating the material andthe energy loss of the scattered particle before going out of the target in the backscatteredstraight trajectory. Typical RBS spectra of a-particles on gold, silver and bronze homo-geneous alloys afe given in figures 8b, 9b and 10b; ODe observes that the safe identifi-cation of ODe element is only easy for the heaviest of them.

Gold Allov600

500

400

'iiìC:Jf:- 300

"O""ai:>= 200

RBS-Proton SpectrumAuAgCu

100

o0.0 0.5 1.0 15 2.0 2.5

Energy (MeV)

Figure 8a. Typical spectrum of backscattered protons (Ei = 2.5 Me V) on a homogeneous gold-silver-cop-per alloy (Au 75 %; Ag 12.5 %; Cu 12.5 %).

Gold Alloy600

RBS-u Spectrum

AuAgCu

500

400

'iiìC:JO 300!:?-

"O"ai~ 200

100

o~nn

.-v.v 0.5 1.0 1.5 2.0 2.5

Energy (MeV)

Figure 8b. 1'ypical spectrum of backscattered a-particles (Ei = 2.5 Me V) on a homogeneous gold-silver-copper alloy (Au 75 %; Ag 12.5 %; Cu 12.5 %). A indicates the Esc on Au at surface, B and C for Ag andCu at surface. In the region D, signals of alI the elements (present in the bulk) afe superposed.

1~d.

---

0.0 0.5 1.0 1.5 2.0 2.5

Energy (MeV)

Figure 9a. Typical spectrum of backscattered protons (Ej = 2.5 Me V) on a homogeneous silver-copperalloy (Ag 90 %; Cu lO %).

Silver Alloy

300

200'Ci)'

rE:3o

!::!-

1:)

"Qj

>

100

0-0.0 0.5 1.0 1.5 2.0 2.5

Energy (MeV)

Figure 9b. Typical spectrum of backscattered a-particles (Ei = 2.5 Me V) on a homogeneous silver-cop-per alloy (Ag 90 %; Cu lO %).

As the energy 10ss of a-particles in materials is higher than that of protons, the total ana-lysed depth with a is much 10wer than with protons but the profile at the surface is moreeasy to determine by a-particles than by protons as it can be seen by comparing parts aand b of Figures 7, 8 and 9. The choice of a or protons and the choice of the typical inci-dent energy is then very dependent on the layer we intend to study.

RBS is very simple in use for the study of thin layers of heavy elements on bulk matrixcontaining light or medium elements but RBS becomes very complicated for materialscontaining elements with neighbourlng atomic masses and is of no use for light elementsin heavy matrices (lI).

(lI) Ruvalcaba-Sil, J. L, Demortier, G., Misaledis, P. ed., Application of particle and laser bearns in mate-rials technology (1995), pp. 463-470, Kiuwer Academic Pubi., and J. L. Ruvalcaba-Sil, G. Demortier,Ecaart Conference, Zurich 1995 (to be published).

135

RBS is then of very limited application for archaeological purposes but is idea1ly usedas a complementary method of PIXE as it will be discussed in the study of gold platingor gold depletion gilding in the chapter on meta1lic samples (i2),

Bronze

250

200

~~~

'Uì"E 150:Jo

~

"C"Qj 100~

-'"

/'

50;.--~-

~

o0.0 0.5 1.0 1.5 2.0 2.5

Energy (MeV)

Figure 1 Da. Typical spectrum of backscattered protons (Ei = 2.5 Me V) on a bronze artefact (Cu 78 %; Sn

18 %; Pb 4 %).

Bronze\ . , . == ~s. ~ 'RBS-u Spectrum

CuSnPb

250

200 \

\

'"

~

'Uì

§ 150~

"C~ 100

50-'_., ~=:~. .

-~;O", . I . I . I .~;--h I.0.0 0.5 1.0 15 2.0 2.5

Energy (MeV)

Figure lObo 1'ypical spectrum ofbackscattered a-particles (Ei = 2.5 MeV) on a bronze artefact (Cu 78 %;Sn 18 %; Pb 4 %).

(12) Demortier, G., 'mA applications to ancient metallic items' in Applications 01 ion beam analysis tech-niques to arts and archaeometry (M. A. Respaldiza and J. G6mez-Camacho, eds.). Secretariado dePublicaciones Universidad de Sevilla, Seville, 1996, pp. 91-114.

136