In-situ XRF analysis as a diagnostic analytical tool in...

A.G. Karydas, V. Kantarelou

• Nuclear Science and Instrumentation Laboratory, IAEA Laboratories, A-2444 Seibersdorf

• Institute of Nuclear and Particle Physics, NCSR “Demokritos”, Aghia Paraskevi, Athens, Greece

In-situ XRF analysis as a diagnostic analytical tool in the conservation field

Andreas Karydas, ICTP, 14th of July 2015

Outline and the PROMET analytical campaigns across the Mediterranean

Outline

1. The PROMET project

2. The micro-XRF mobile instrumentation

2. Accuracy and pitfalls of micro-XRF analysis

3. PROMET campaigns:

- Ancient Messene (2006)

- Malta, Armoury Palace (2006)

- Damascus National Museum (2007)

- Numismatic Museum, Yarmouk University, Irbid, Jordan


PROMET FP6:2005-2008 Aim:

To develop Prototype innovative and advanced analytical methods to survey large collections of metal objects in-situ, making it possible to pinpoint conservation needs without any risk of damaging the artefacts

Efficient, versatile and mobile analytical methodologies: Micro-XRF and Laser Induced Breakdown Spectroscopy

LIBS related tasks were carried out by Prof. D. Anglos FORTH-IESL, Crete

Coordinator: Prof. V. Argyropoulos (TEI, Athens)

24 partners, including Turkey, Syria, Jordan, Morocco, Italy, France, Spain, Czech Republic


To develop, optimize and calibrate the analytical performance of aninnovative portable micro-XRF spectrometer

To develop and improve analysis procedures, protocols and thestandardization of the method

To apply the micro-XRF spectrometer for systematic technologicaland conservation related studies of museum metal collections at theMediterranean region:

• The study of the manufacture technology of metal alloys• Non – invasive characterization of corrosion products• Contribution to the assessment of innovative protective coatings

Demokritos objectives within PROMET:


μ-XRF spectrometer: Principle of operation

X-ray tube

Focusing X-ray device: Polycapillary X-Ray lensy

Sample

X-Ray detector

Development and application of

portable micro-XRF unit

Customized design of ARTAX by Bruker

Nano AXS


Versatility: In-situ Micro-XRF analyses

Laboratory test of TEI coupon

Numismatic Museum of Yarmouk UniversityIrbid, Nov. 2008

X-ray Detector

Laser pointer

X-ray lens

Headed Eagle lapis lazuli and gold 3000 B.C. Early Bronze Age

Damascus National Museum, Syria, October 2007


Pitfalls: Interference of XRF signal with diffraction peaks, QC/QA of micro-XRF data

6 8 10 12 14

101

102

103

104

Bragg peaks

Fe

50kV, 600A, 50s

Cu

Au

Au

#1#2Au

Cou

nts

Energy (keV)

Diffraction peaks Heterogeneity at the micro-scale Definition of the scanning area

that represents the alloy bulk composition

Alloys Filters/ Thickness (μm)Ti (23.6 ± 0.2) Co (17.7 ± 1.3) Pd (11.3 ± 0.3)

Gold x xSilver x x xCopper x x

5 10 15 20 25100

101

102

103

104

105

Ag

Au

Au

Au

Ag

Cu

Rh

Au

SBL

Cou

nts

Energy (keV)

SBLnorm_unfiltered_100 sec/600 A SBL_filtered_(Ti+Co)

Analytical range using He atmoshpere

1 2 3 4

200

400

600

800

1000

1200

1400

Ceramic sample

Ca

K

Helium No Helium

Si

Al

Cou

nts

Energy (keV)

50kV, 600mA, 100s

The improvement in the intensity of Al-K and Si-K characteristic X-ray lines issignificant, 22 and 7.3 times, respectively.

Analytical performance: Elemental sensitivity

600 μA50 kV

Thin Targets ~ 50 μg/cm²

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28

0.01

0.1

1

10

S

K-sensitivities, He atmosphere K-sensitivities, No Helium L-sensitivities, No Helium

SiAl

W-Au

Pt

Ag-Sn

Pb

CoFe

CrV

CaK Se Rb

BrSrY

Nb

GaGeNi

Cl

Cu

Sens

itivi

ty: c

ps/(

g/cm

2 )

Energy (keV)

AgSn

Analytical performance: Spatial resolution

0 5 10 15 20 2530

40

50

60

70

80

90

100

FWH

M (

m)

Energy (keV)

Filtered_Ni (25m) Unfiltered

Calibration methodology

1sin1)()(),(),()()()(

kdkairikikiiki EEfdEFEEAEEETEIwGEI


0 2 4 6 8 10 12 14 1610-2

10-1

100Thin targets

Inte

nsity

(cps

/A

)

Energy (KeV)

Theory (FPA) (K lines) Experimental (K lines) Theory (FPA) (L lines) Experimental (L lines)

(a)

Experimental/simulated pure element thick/thin elemental intensities


Kantarelou et al., XRS, 2015

0 5 10 15 20 25-50

-40

-30

-20

-10

0

10

20

30

40

50 (c)

(%) D

evia

tion

Energy (KeV)

Thick targets (K lines) Thick targets (L lines) Thin targets (K lines) Thin targets (L lines)

2 4 6 8 10 12 14 16 18 20 22 24 26 28 300.0

0.2

0.4

0.6

0.8

1.0 (d)

Tran

smis

sion

(a.u

.)

Energy (keV)

Results of the fitting procedure

Estimated Lens transmission efficiency


Kantarelou et al., XRS, 2015

4 6 8 10 12 14 16 18 20 22 24 26 280.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3 (b)

I J K L CNR91 CNR92 CNR152 ABQAQ ABSBL ABLLI ABKMF

Fact

or (K

i)

Energy (keV)0 2 4 6 8 10 12 14 16 18 20

0.5

1.0

1.5

2.0

2.5

3.0

620 1412 89 BAM 1831 a4 c3 d3 f3 e3

Fact

or (K

i)

Energy (keV)

(a)

0 2 4 6 8 10 12 14 16 180.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

K lines L lines

mea

n va

lue

of K

i

Energy (keV)

Accuracy/Quantification of CH related materials

Glasses

Glasses

Gold/Silver alloys


Assessment of micro-XRF analysis accuracy

A. Heginbotham et al., An Evaluation of Inter-Laboratory Reproducibility for Quantitative XRF of Historic Copper Alloys, Proceedings of the International Conference on Metal Conservation, METAL 2010, pp 178-188, Edited by Paul Mardikian, Claudia Chemello, Cristopher Watters and Peter Hull, 11-15 October 2010, Charleston, South Carolina, USA

Micro-XRF (~50μm) Milli –XRF (3 mm)Validation with respect to Cu based RMs


Semi-QA and Diagnostic Micro-XRF AnalysisMethodology:

Line and area scans to obtain in reasonable measuring time (1x1 mm2, 50 μm step, 10s/step,~1.5h) intensity maps of the detected characteristic X-ray lines

Filtered excitation

Analysis of corroded area vs corrosion free area

Variation of the K/L or L/M elemental intensity ratios in single spot, line or area scan measurements


Semi-QA and Diagnostic Micro-XRF Analysis

Identification of the spatial coexistence of different elements,fingerprints of certain corrosion products or of manufacturetechniques.

Spatial distribution of individual elements

Results obtained:

Identification of the presence of certain minor to traceelements that may support provenance and manufacturestudies of the metal

Estimation on a semi-quantitative basis of the elements enriched or depleted from the surface

Rough estimation of the depths that a certain element is located, namely, on the surface, near surface (~2-10 μm) or below ~10 μm.


0 1 2 3 4 5

0.0

0.2

0.4

0.6

0.8

1.0

1.2

FI

#47#9

Inte

nsity

Position (mm)

Cl-K Cu-K

4 8 12 16 20 24100

101

102

103

104

105

RhRh

Pb-LPb-L

Pile-upsSn-L

+Ca-

K

Cu-K

Sn-K

CuE

P

Cl

Cu-K

Cou

nts

Energy (keV)

# 9 # 47

Analysis of Copper coupon corrosion products

Artificially and naturally aged bronze coupon: (Cu: 91.3%, Sn: 7.5%, Pb: 1.0%)

50kV, 600μA, 30s/step,0.1mm/step,50 measurements

#9 : green area#47: pale green area

Artificially and naturally aged silver coupon:Ag: 92%Cu: 6.5% Pb: 1.5%

50kV, 600μA, 30s/step, 0.1mm/step,50 measurements

Silver coupon (prepared-characterized by Prof. G. M. Ingo, Polytechnico of Milano) A - Paratacamite

B - ChloroargyriteC - Silver (oxide)

A – GreenB - WhiteC - Black

A.G. Karydas et al, PROMET Book, 2008

Analysis of metal corrosion products

2 3 4 5 6 7 8 9 10100

101

102

103

104

105

Fe + CuEPCa

Ag-LCu-K

Cl

Cu-K

Cou

nts

Energy (keV)

A - Green B - White C - Black

4 8 12 16 20 24100

101

102

103

104

105

Ca-

K

Pb-LPb

-L

Pb-L

Pile

-ups

Rh-

K

Ag-

L

Ag-K

Cu-

K

Ag-K

esc-

Cu

Cl-K

Cu-

K

Cou

nts

Energy (keV)

B A C

0.00 0.25 0.50 0.75 1.00

0.00

0.25

0.50

0.75

1.00 0.00

0.25

0.50

0.75

1.00

DC B

A

Cu-KAg

-L

Cl-K

1 2 3 4 5100

101

102

103

104

105

ClS

Ag

Black Metal

Cou

nts

Energy (keV)

Silver Bowl 1400 -1300 BC Late Bronze Age

Damascus Archaeological museum:Analysis of silver tarnishing

5 10 15 20 25100

101

102

103

104

105

CaCl Au RhS

Ag

AgAg

AuAu

Cu

Cu

Black Metal

Coun

ts

Energy (keV)

Tarnish: corrosion mainly caused by the sulfur in the air

Thickness of the layer: ~ 0.5 μm

Baal Gods

Manufacture technology (compositional analysis, raw materials)

Identification of corrosion products

Late Bronze Age, 1400 B.C. Ugarit site: Issues addressed

Gilding technique

PROMET, Damascus, Syria: Gilded Bronze figurines

Kantarelou et al., JAAS, 2015

El God

Compositional analysis of bronze metalConcentrations (wt. %)

Fe: 0.26 ± 0.03Cu: 92.5 ± 1.0Zn: 0.60 ± 0.06As: 0.10 ± 0.01Sn: 6.32 ± 0.30Pb: 0.040 ± 0.004

Cu rich corrosion

Sn rich corrosion

Bronze metal


1 100.00.20.40.60.81.01.21.41.61.82.02.22.42.6

Ag-K/Ag-L

Theory (norm to pure Ag) Reference Alloys - Fischer El god (obj 1) Baal God (obj 4) Baal God (obj 5)

Inte

nsity

ratio

Thickness / um

Compositional analysis of gold foil

Thickness of the gold foil > 10 micrometers Two (2) main compositional groups: Ag: 14-15% or 4-5%Cu: generally <2-3%


Imaging of the elements distribution on the gold foil

5 10 15 20 25101

102

103

104

105

106

Cu SnFeCaPbPU's

KPbPb

Cu

Sn

Fe

Rh

Ca

Cu

Cou

nts

Energy (keV)

Eail God, Late Bronze Age 1400-1300 B.C.PROMET, Damascus, Syria: Corrosion products

Malachite

5 10 15 20 25100

101

102

103

104

105

Cu SnFeCaAsZnSe

AsFe

Zn

SnRh

RhSe

Ca

Cu

Cu

Cou

nts

Energy (keV)

Sn rich layer

Copper oxide

Trace elements:Zn, As, Se

5 10 15 20 25100

101

102

103

104

105

106

CuSnFeAsFe

Sn

Rh

RhPU'sAs

AsSnCu

Cu

Cou

nts

Energy (keV)

Copper-base bracelet, #216, (Ottoman Period),General + Pitting + Crevice Corrosion, Activecorrosion have caused serious crack

A. Arafat et al., Journal of Cultural Heritage,http://dx.doi.org/10.1016/j.culher.2012.07.003,14 (3), (2013) 261-269

5 10 15 20 25100

102

104

RhFe+E

PCu

Zn

Pb+As

Ni

Cu

Zn

Pb

CuZnPbFeNiAs

Cu

Cou

nts

Energy (keV)

Cu: 76.08%Zn: 21.7%Ni: 0.80%Sn: 0.43%Pb: 0.43%Fe: 0.39%As: 0.17%

5 10 15 20100

102

104

Cl

Cu

Cu

PbPb+A

s PU'sFe+E

PCu

Zn

CuClZnPbAsRh

Cou

nts

Energy (keV)

PROMET at Irbid, JordanUmm Qais artefacts

Atakamite, Para-atakamite

5 10 15 20 25100

101

102

103

104

105

Rh

Pb+AsZn

Zn

Cu

CaClPU'sFe

+EPC

u

CuZnAsPbCaCl

Rh

Cu

Cou

nts

Energy (keV)

Copper oxides

Compositional analysis of different typology copper based artifacts through time

Surface characterization of high tin bronze mirrors

Combined microXRF and LIBS analysis for enhancing in-depthelemental distribution

PROMET in Ancient Messene - Objectives:


Dynamic combination of micro-XRF and LIBS

LASER

Spectrograph

Fiber

Mirror

260 270 280 290 300 310 320 330 340

3x

PbCu

As

AsAs Cu

Cu Cu}

Cu

Cu

LIBS

X-ray Detector

Polycapillary

0 5 10 15 20 25 30101

102

103

104

105

106

107

Sn-K

Sn-K

Rh-K

Cu-K

Rh-K

Pb-L

Pb-L

Pb-L

Zn-K

Zn

-K

Cu-K

Cu-SUM peaks

Sn-L

Ni-K

Coun

ts

Energy

RC36/11 Brass50 kV, 600 A, 150secUnfiltered radiation

μXRF

V. Kantarelou et al., METAL-07, Vol. 2, Innovative investigation of metal artifacts, pp. 35-41, (2007)


Courtesy of D. Anglos

0.77

1.0

0.55

0.33

0.110

0.84 1.100.600.360.12

Cu-K

Position (mm)

Posi

tion

(mm

)

0300006000090000120000150000180000210000240000

1.0

0.77

0.55

0.33

0.110

0.84 1.100.600.360.120

Sn-L

Position (mm)

Posit

ion

(mm

)

50093813751813225026883125356340001.0

0.77

0.55

0.33

0.110

0.84 1.100.600.360.120

Sn-K

Position (mm)

Posi

tion

(mm

)200288375463550638725813900

1.0

0.77

0.55

0.33

0.110

0.84 1.100.600.360.12

Pb-L

Position (mm)

Posi

tion

(mm

)

0875017500262503500043750525006125070000

20 pulses

Micro-XRF elemental mapping of the LIBS ablated area

W2 (%): Cu: 82.1, Sn: 6.9, Pb: 10.7

PROMET at Ancient Messene

1.9 3.3

2.3

1.7

1.0

0.3

3.0

2.61.10.40

Cu-K

Relative x-position (mm)

Rel

ativ

e y-

posi

tion

(mm

)

0

1.1E5

2.2E5

3.4E5

4.5E5

5.6E5

6.8E5

7.9E5

9E5

0

3.0

2.3

1.7

1.0

0.3

2.6 3.31.91.10.4

Pb-L

Relative x position (mm)

Rel

ativ

e y

post

ion

(mm

)

0

4400

8800

1.3E4

1.8E4

2.2E4

2.6E4

3.1E4

3.5E4

-XRF analysis of High Tin Bronzes PROMET at Ancient Messene

0.3

1.0

1.7

3.0

2.3

0 1.9 3.32.60.4 1.1

Sn-K

Relative x position (mm)R

elat

ive

y po

sitio

n (m

m)

5E2

9.4E2

1.4E3

1.8E3

2.3E3

2.7E3

3.1E3

3.6E3

4E3

Ancient Messene, Greece:Manufacture of high tin bronze mirrors

Mirror 1 (M1)

(Cu: 70.6%, Sn:26.0%, Pb:3.4 % )

Mirror 2 (M2)(Cu: 70.2%, Sn:22.9 %, Pb: 6.8 % )

Examined areas : metal, black, silverish Examined areas : metal, black, silverish,green and light grey

2nd c. BC

Clean Silverish Black0.40.60.81.01.21.41.61.82.02.2

M1

Rel

ativ

e In

tens

ity (a

.u.) CuK

SnKSnLPbL

Clea

n

Silv

eris

h

Blac

k

Gre

en

Ligh

tGre

en

0.0

0.5

1.0

1.5

2.0

2.5

3.0

M2

Rel

ativ

e in

tens

ity (a

.u.)

CuKSnKSnLPbL

Silverish: Increase of SnLα, decrease of CuKa and PbLa : Sn surface enrichmentBlack : Significant increase of SnLα (Cassiterite?)Green : Increase of CuKa, Cu corrosion products (malachite?)Light green : Increase of PbLα, SnLα, decrease of Cu-K, (Lead-white? Cassiterite?)

Ancient Messene, Greece: Surface examination of high tin bronze mirrors

0 2 4 6 8 10

0

1

2

3

4

5

Black Silverish

Pure SnO2 layer SnO2+ Cu 31Sn8

Rel

ativ

e In

tens

ity (a

.u.)

Layer Thickness (m)

Sn-Lα

Ancient Messene, Greece:Black patina on high tin bronze mirrors

Silverish < 1μmBlack layer with a thickness of very few microns

Gilding technology

Authenticity issues

Compositional analysis of armory components (rivets)

Identification of surface corrosion products

Conservation related issues addressed:

PROMET at Armoury Palace, Malta


0.9

0.6

0.3

0

1.4 1.81.00.60.2

Cu-K


Rel

ativ

e y

posi

tion

(mm

)

022504500675090001.125E41.35E41.575E41.8E4

0.9

0.6

0.3

0

1.4 1.81.00.60.2

Fe-K


Rel

ativ

e y

posi

tion

(mm

)

1E52E53E54E55E56E57E58E59E5

0.9

0.6

0.3

0

1.4 1.81.00.60.20

Hg-L


Rela

tive

y po

sitio

n (m

m)

0

1.5E3

3E3

4.5E3

6E3

7.5E3

9E3

1.05E4

1.2E40.9

0.6

0.3

0

1.4 1.81.00.60.2

Au-L

Relative x position (mm)R

elat

ive

y po

sitio

n (m

m)

087501.75E42.625E43.5E44.375E45.25E46.125E47E4

Gilded Iron Alloy Falling Buff -XRF analysis of gilding areas

PROMET at Armoury Palace, Malta


The mean Au/Hgvalues werededuced among asubgroup of thearea map spots,where the Auintensity variesbetween 0-10%, 10-20%, 30-40% etcwith respect to itsmaximum value

Fire-Gilding technique: Hg to Au ratio

C. Degrigny et al. , Metal-07, pp. 26-34, (2007)

0.0 0.2 0.4 0.6 0.8 1.00.0

0.1

0.2

0.3

Hg-

L/A

u-L

Au-L relative intensity

# 6 # 7 #14

Palace Armoury, Malta

Conclusions

Micro XRF analysis can offer fast elemental distributionmaps, contributing thus both towards the identificationof surface corrosion products and manufacture techniquesas well

The operating conditions of the micro-XRF spectrometer require careful optimization per type of samples analyzed, that it is in many cases not a trivial and straightforward procedure.


PROMET Impact• The results of the project offer a cost effective

approach in identifying the conservation problems and needs of a metals collection using portable diagnostic techniques

• The achieved results can easily be applied to anymuseum setting world-wide due to the portabilityof the instruments developed whereas analyticalmethodologies are easily transferable.


AcknowledgementsTEI, Athens, Prof. V. Argyropoulos (coordinator) , M. Giannoulaki

IESL-FORTH, D. Anglos, A. Giakoumaki

Ancient Messene, Society of Messenian Archaeological Studies

Prof. Dr. P. Themelis, Director of Excavations

Museum of Ancient Messene, S. Polymenea, Conservator

National Museum at Damascus, Syria

Luda Mahfoud, Abeer Kurdab, Mayada El Saadi and Kasem Yahya

Directorate General of Antiquities and Museums, Mr Maher Azar

Malta Heritage, Dr. C. Degrigny, Dr. S. Golfomitsou, D. Vella

Yarmouk University, Dr Manar Bani Hani, Lina Khrees, Prof. Mohammed S. Shunnaq, Prof. Ziad Al-Saad

Royal Scientific Society of Jordan, Dr Abeer Arafat and

Umm Qais Museum, Mohammad Bashabsheh

In-situ XRF analysis as a diagnostic analytical tool in...

Documents

Transcript of In-situ XRF analysis as a diagnostic analytical tool in...