of 8

SANDRA Portable XRF Ruvalcaba

Published on 2 weeks ago | Categories: Documents | Downloads: 0 | Comments: 0
206 views

Comments

Content

Resear Res earch ch Art Articl icle e Rece Re ceii ve ved: 25 Oc Octo tobe berr 20 2009 09

Rev evii se sed: 28 Fe Febr brua uary ry 20 2010 10

Acce Ac cept pte ed: 1 Mar arch ch 20 2010 10

Publ Pu blii she shed d on onll inei inei n Wi Will ey ey In Intter ersc scii enc ence e: 7 Ma Mayy 20 2010 10

(www.interscience.com) (www.inter science.com) DOI 10.1002/xrs.1257

SANDRA: a portable XRF system for the study of Me Mexi xica can n cu cult ltur ural al he heri rita tage ge†  J. L. Ruvalcaba Sil, D. Ra Ramí míre rezz Mi Mira rand nda, a, V. Ag Agui uila larr Me Melo lo an and d F. Pi Pica cazo zo ∗

This work presents the portable X-ray system system SANDRA (Sistema de Anali Analisis sis No Destr Destructiv uctivo o por RAyos X or System for Non Destructive Analysis using X-rays) developed at the Physics Institute of the UNAM, Mexico, for the study of Mexican cultural heritage collections. The X-ray fluorescence (XRF) SANDRA device can use 75 W Mo, Rh and W X-ray tubes and Amptek Si-PIN andCd-Te det detect ectorsthat orsthat areselec areselectedand tedand com combin bined ed dep depend endingon ingon the ele elemen mentalrange talrange det detect ectionrequir ionrequireme ementsand ntsand thespeci thespecific fic proble pro blem m to be stu studie died. d. In thi thiss pap paper, er, a ful fulll des descri cripti ption on and cha charac racter teriza izatio tion n of thi thiss sys system tem,, sen sensit sitivi ivitie tiess for the X-r X-ray ay tub tubes es and detectors as well as the detection limits are discussed. Examples of applications to technological studies on pre-Columbian c 201 metalli meta llicc art artifa ifactsand ctsand ana analys lysis is of col color or mat materi erialsof alsof anc ancien ientt Mex Mexica ican n cod codex ex areshown areshown.. Copy Copyrig right ht  2010 0 Joh John n Wil Wiley ey & Son Sons, s, Ltd Ltd..

Introduction

 3    3    8  

X-ray fluo X-ray fluore resce scence(XRF)is nce(XRF)is a fas fastt andhigh sen sensit sitivi ivity ty mul multie tielem lemenental technique; it can be applied directly for non-destructive and non-invasive studies.[1,2]  This technique is well established as a basic analytical tool for cultural heritage applications for more than 30 years.[3–13] Although there are many museums and conservation labor laborator atories ies equ equippe ipped d with stat static ic or port portable able XRF equi equipmen pments ts in developed countries, nowadays, the use of portable devices for the characterization of objects and collections of museums, librar lib raries ies,, or in ge gener neral, al, mat mater erial ialss fro from m the arc archae haeolo ologic gical al re recor cord d or art history is fortunately increasing in countries of Latin America. Since Sin ce the ear early ly wor works ks usi using ng por portab table le XRF sys system temss[1,3,5,7,11] , many impr improvem ovements ents have been achie achieved. ved.[1,2,14–16] Wit With h the developm deve lopment ent of Si-PI Si-PIN N and silic silicon on drift detectors, detectors,[17–19] the systems syste ms beca became me light lighter er and the port portabili ability ty was sign significan ificantly tly improved for  in sit situ u   measurements. Thus, several systems have been developed in many countries, mainly in Europe. Currently, different types of X-ray sources – including radioactive sources and X-ray tubes – and detectors are used. [20–25] XRF apparatus are ext extens ensive ively ly app applie lied d for the cha charac racte teriz rizati ation on of all kin kind d [26–33] of cult cultural ural heri heritage tage’s ’s mate materials rials.. Often, Ofte n, XRF analyses are compleme comp lemented nted by othe otherr port portable able spec spectrom tromete eters, rs, such as the [13–34] Raman ones, and also using other X-ray based techniques such as particle induced X-ray emission spectroscopy (PIXE). [35,36]  The last improvem improvements ents on XRF systems consider the developme development nt of devi devices ces com combinin bining g X-ra X-rayy diff diffract raction ion (XRD (XRD)) and XRF in sit situ u.[37–39] Despite the rich cultural heritage of the Latin American area and the improvement improvementss in the experimenta experimentall devi devices, ces, very few portable XRF equipments have been developed in the last years in thi thiss re regio gion. n.[40–42] This paper presents the portable X-ray system SANDRA SAND RA (Sist (Sistemade emade Anali AnalisisNo sisNo Dest Destruct ructivopor ivopor RAyo RAyoss X orSystem forNon Dest Destruct ructiveAnalysisusing iveAnalysisusing X-rays X-rays)) deve develope loped d at thePhysics Institute Insti tute of the UNAM, Mexico, for the study of outstanding outstanding Mexican cultural heritage collections. The main applications of  the SANDRA system are related to pre-Columbian and colonial manuscript analyses, paintings for palettes studies and painter’s techniqu tech niques es from colo colonial nial to mode modern rn per periods, iods, 16th– 17th cent centurie uriess polychrome colonial sculpture, pre-Columbian metallic artifacts (gold, bronzes), pre-Hispanic turquoises and green stone objects,

 X-Ray Spectrom. 2010 2010,, 39  39,, 338 – 345

as well as ancient photographs. Some examples of applications are also discussed in this paper.

The SAN SANDRA DRA Sys System tem Our actual SANDRA system is the result of a previous prototype with a Rover system with Si-PIN and CZT X-ray detectors from Amptek. [41,42] In th the e pa past st ye year ars, s, it ha hass be been en re reno nova vate ted d an and d new detectors detectors may be used used.. To impr improve ove our syste system, m, seve several ral comparisons with PIXE and RBS measurements measurements have been carried out on book’s decorations from 17th to 19th centuries, [43] Maya pottery and pre-Hispanic copper artifacts. [44] SANDRA SAND RA syste system m has a typic typical al geo geomet metric ric confi configura guration; tion; the detector is set at 45 ◦ from the direction of the excitation X-rays (Fig. 1). The available X-ray tubes have Mo, Rh and W anodes with 125 µm bery berylliumwindow lliumwindow.. Themaximumpowerof theX-ray tube tubess is 75 W (50 kV, 1.5 mA). mA). They corr correspo espond nd to the XTF5 XTF5011 011 model from Oxford Instruments. The X-ray tube is powered by a high voltage power supply XLG50P100 from Spellman. The X-ray tube is mounted on an X–Y–Z support, so it can be manually moved 3 cm in each direction in front of the studied object to smoothly reach the selected region of analysis. Three small fans and a main fan at the bottom of the X-ray tube provide the necessary cooling for the X-ray tube. A thermocouple is provided in the X-ray tube for temperature monitoring. The maximum X-ray tube operation temperature temperatu re is 50 ◦ C.  The SANDRA system is mounted on an articulating arm boom stand that ensures high mobility and flexibility in the movement (Fig. 2). With this setup, it is possible to reach corners and difficult locationson locat ionson thestudiedobjec thestudiedobject. t. Thesystemcan be usedfor analy analysis sis onthe ve verti rticalandhoriz calandhorizont ontalplane alplaness orit canbe til tiltedas tedas nec necess essary ary..



Correspond Corre spondenceto: enceto: J. L. Ruval RuvalcabaSil, cabaSil, Instit Institutode utode Física Física,, Univ Universida ersidad d Nacion Nacional  al   Autonoma de Mexico (UNAM), Apdo. Postal 20-364, Mexico DF 01000, Mexico. E-mail: [email protected]fisica.unam.mx  Instituto de Física, Universidad Nacional Nacional Autonoma de Mexico (UNAM), (UNAM), Apdo. Postall 20-36 Posta 20-364, 4, MexicoDF 01000 01000,, Mexico

† Articlepublishedas partof the specia speciall issueon Porta Portable ble X-ra X-rayy Spect Spectromet rometers. ers.

c  2010 John Wiley & Sons, Ltd. Copyright 

SANDRA portable XRF system

X-ray tube

X-Y-Z support

fans

beam shutter focusable lasers

region of analysis

webcam

Si-PIN detector

Figure 1. The SANDRA XRF system during the study of a collection of  Mexican historic photography. Main components are pointedout.

Figure 2. The SANDRA XRF system supported on its articulating arm boom stand. Analysis of the Colombino codice.

 The X-ray beam diameter is determined by a lead collimator.  The beam diameters may be 0.5, 1.0, 1.5 and 2.0 mm. Two X-ray detectors can be used: a Si-PIN detector (XR-100-CR model) and a Cd-Te detector (XR-100T-Cd-Te model) from Amptek. The Si-PIN detectorhas a 6 mm2 activearea,a 500 µm thickness and a 0.5 µm Be window while theCd-Tedetector dimensions are5 × 5 × 1 mm (25 mm2 active area) and 4 µm Be window. The energy resolution at 5.9 keV (Mn-K α  line) of the Si-PIN and Cd-Te detectors are 180 and 360 eV, respectively. A conic piece in aluminum covers the detector head for its protection and it could be used to set a He jet for light elements detection improvement. The cone has a cylindrical apertureof 4 mm of diameter. Under this configuration, this cone is working as a collimator only for the Cd-Te detector, as the exposed area is then 12.5 mm 2 , a half of the total detector active area.  The X-ray detector signal is processed by a PX4 digital pulse processor from Amptek before reaching the laptop computer. Recently, the complete X123 Si-PIN spectrometer system from Amptek has been tested as well. This system provides together the X-ray detector, the amplifier and the power supply in a small packing and it is directly connected to the computer by a USB

 X-Ray Spectrom. 2010, 39, 338–345

cable. The performance of this Si-PIN detector is similar to the one described above. For the SANDRA operation a LED security light turns on when X-ray tube is working. The X-rays fromthe tube can bestopped by a shutter. To determine the region of analysis, two focusable laser pointers intercept at 8 mm from the X-ray tube exit collimator. An X-ray screen may be used to set properly the working distance and the laser pointers. The total distance from the X-ray tube exit window to the sample surface is 9 cm while the X-rays emitted from the sample have to travel 1.5 cm in air before reaching the detector. The incident X-ray beam spot at the sample surface was measured for 1.0, 1.5 and 2 mm diameter using 250 µm diameter Au wire and mounted on a XY translation stage with 10-µm resolution. We determined that there is no significant scattering and the increase in the beam diameter is lower to 15% of the collimator diameter. Considering the previous parameters, Rh and Mo K peaks from the X-ray tube loss only 1% of their intensities in the air path while L lines from Mo and Rh have a significant absorption and only 3.6 and 12.6% may reach the sample, respectively. For this reason, the corresponding X-ray scattering is very weak and it is not often detected. The Mo-L peak may overlap S-K and Pb-M peaks, whereas the Cl-K peak may be overlapped by Rh-L peak. On the contrary, W Lα  and Lβ  lines show a lower absorption in air (about 10%), then the scattering peaks are frequently observed in the spectra using the W X-ray tube. Nevertheless, the Mo, Rh and W L peaks may be easily removedusingfilters in theincident X-ray path, such as a foil of 12 µm of Cr or Zn. On the other hand, only the X-rays emitted from sample by the lighter elements are absorbed significantly in the 1.5 cm air path. For Al, Si, S, Cl, K, Ca, Ti, Fe and Cu K α the transmitted X-rays are 5, 24, 36, 51, 61, 76, 85, 92, 97 and 99%, respectively. This fact means that if lowenergy X-rays detection is required forthe analysis with our equipment, it will be necessary to use a He jet at the X-ray detector and at the path of the X-ray incident beam as well. Finally, to improve the data record during the analysis and to observe the irradiated region at the object surface simultaneously to the X-ray spectra acquisitions, a webcam with medium resolution is used. In this way, both an image and the corresponding spectra are stored in the computer.

Characterization of the XRF System  The full characterization of the SANDRA system was required to determine their advantages and limitations, as well as the most suitable configurations. Measurements on a full set of standard reference materials from MicromatterCo. were carried outfor theMo, Rh andW X-ray tubes and the Si-PIN and Cd-Te detectors. The elemental range covered fromSito SnforK and L lines and PbandAuL lines.From the X-ray fluorescence spectra,the sensitivitywas calculated[1] fromtheratio of the peak area (counts) and the product of the concentration (µg/cm2 ) and the time (s). For calculation of sensitivities, peak  areas were measured using AXIL program. Figures 3 and 4 show the sensitivity curves as a function of the X-ray energy for the Si-PIN and the Cd-Te for the Mo, Rh and W under the samevoltageand currentconditions (45 kV and 0.3 mA). Filters were not used. From thesensitivitycurves, it is clear that theW X-ray tube gives rise to higher sensitivities for both detectors from 3 to 8 keV (K to Cu) mainly due to the W–L lines excitation while for the higher

c  2010 John Wiley & Sons, Ltd. Copyright 

 

www.interscience.com/journal/xrs

 3    3    9  

J. L. Ruvalcaba Sil etal .

Mo Rh W

0

   ) 10   s    *    2

  m   c    /   g      µ    /   s 10-1    t   n   u   o   c    (   y    t    i   v    i    t    i   s   n 10-2   e    S

Ag Fe

Zn As

Br

V

Sn Sr

Ca Cl Si

Si-Pin K lines Si-Pin L lines Cd-Te K lines Cd-Te L lines

  y    t    i 0   v    i    t    i 10   s   n   e   s   e    F   y    b    d 10-1   e   z    i    l   a   m   r   o   n   y    t    i   v 10-2    i    t    i   s   n   e    S

Zn

Br

Fe

Sr

Au Pb

V Ca

Ag

Cl

Sn Si

Mo X-ray tube

Si-Pin detector 10-3

-3

10 0

2

4

6

8

10 12 14 16 18 20 22 24 26 28 30

0

2

4

6

8

X-ray energy (keV)

10 12 14 16 18 20 22 24 26 28 X-ray energy (keV)

Figure 3. Sensitivity curves as a function of the X-ray energy using the Si-PIN detector andthe Mo,Rh and W X-ray tubes under similarconditions (0.3 mA,45 kV). A setof Micromatter Co.thin reference materials wasused.

Figure 5. Comparisonbetween the sensitivitiesnormalizedby Fe sensitivity forthe Si-PINand Cd-Te detectors forthe Mo X-ray tube under thesame power conditions (0.3 mA, 45 kV).

101 Si

Mo Rh W

   )   s    *

Ag

   2

  m   c 100    /   g      µ    /   s    t   n   u   o   c    (   y    t    i -1   v    i    t 10    i   s   n   e    S

Fe

10

   )   m   c    /   g      µ 0    (    t 10    i   m    i    l   n   o    i    t   c   e    t   e    D -1

   2

Sn

Zn

1

As Br Sr

V Ca

Sn Ag Cl

10

Cl

Ca Ti Fe

Cd-Te detector 2

4

6

8

10 12 14 16 18 20 22 24 26 28 30

10

15

20

25

X-ray energy (keV)

 3   4   0  

Br

30

35

Si-Pin Cd-Te

Sr

Mo X-ray tube

10-2 0

Zn

40

45

50

55

Atomic number Z

Figure 4. Sensitivity curves as a function of the X-ray energy using the Cd-Te detector and theMo, Rh andW X-ray tubes under similar conditions (0.3 mA,45 kV). A setof Micromatter Co.thin reference materials wasused.

Figure 6. Detection limits comparison for the Si-PIN and Cd-Te detectors forthe Mo X-raytube under thesame excitationconditions (0.3 mA,45 kV). A set of Micromatter Co. thin reference materials was used.

energy range (10– 26 keV)the sensitivies aresimilarby comparison to the Mo and Rh X-ray tubes. On the other hand, despite a rapid decrease in the sensitivity for the Si-PIN detector and the Rh tube for X-ray energies higher than 18 keV, thesensitiviesare very similar forthe Mo andRh X-ray tubes from Si to Sr. For energies lower than 14 keV, the Mo-K lines areenergetically closerto theK-edge of theelements in this range than the Rh-K lines (and therefore the respective photoelectric cross sections are higher). However, the Rh anode source offers more intense bremsstrahlung radiation in the samerange which is actually the most significant contribution in the ionization process of the corresponding elements. For these reasons the sensitivities are comparable. A comparison between the sensitivities normalized by Fe sensitivity for the Si-PIN and the Cd-Te detectors using the Mo X-ray tube is shown in Fig. 5. It is observed that the normalized sensitivities forthe Si-PINdetector arehigher in thelightelements range (the factor decreases from 3 for Cl when the X-ray energy

increases). In the medium X-ray energy range (from V to As), the normalized sensitivities are very similar and they are slightly higher for the Cd-Te detector from 11 keV. Nevertheless, for X-ray energies higher than 18 keV the normalized sensitivities for Cd-Te are higher than the Si-PIN normalized sensitivities from a factor of 3.5 up to 7 as X-ray energy increases. The Cd-Te detection starts from Cl because of the low energy X-ray absorption in its thick window (4 µm), but the Cd-Te efficiency increases rapidly for higher energies improving its sensitivity.  The detection limits[2] calculated as the ratio between the product of the concentration by three times the square root of  the background and the peak area, for the Mo X-ray tube and for both detectors, show that the lower detection limits are reached in all the cases for the Si-PIN detector (Fig. 6). Peak to background ratios are in general higher for the Si-PIN detector than for the Cd-Te detector. For Fe, these ratios are 0.97 and 0.7 for the Si-PIN and the Cd-Te detector, respectively, while for Ag the same ratios are 0.92 and 0.57. Despite the higher sensitivity for the Cd-Te in

www.interscience.com/journal/xrs

 

c  2010 John Wiley & Sons, Ltd. Copyright 

 

X-Ray Spectrom. 2010, 39, 338–345

SANDRA portable XRF system

104 Cl

   )   g    /   g      µ    ( 103    t    i   m    i    l   n   o    i    t   c   e    t   e 102    D

complemented by Raman spectroscopy, Mid-Fourier-transform infrared (FTIR) in situ and VIS-NIR light spectrometry. Our SANDRA system has been extensively used for the study of  Mexican cultural heritagecollections since the construction of our first prototype several years ago. Due to thesuccess of this device, other apparatus have been constructed. Actually, three devices are operational and they are used in collaboration with the most relevant Mexican museums. The main applications are related to studies of pre-Columbian and Colonial manuscripts, paintings for paletteandpainter’stechniquesfromColonialtomodernpainting, polychrome colonial sculptures, pre-Columbian metallic artifacts (gold, bronzes), pre-Hispanicturquoises and green stone pieces, as well as historic photographs. Recent applications include metallic threads and textiles. In this work, two examples of our researches using the SANDRA system are presented.

Sediment (SRM 2711, 2704) Tomato leaves (SRM 1573a) Gold alloy (588/340)

Si

Au Cu Ag

Pb

Sr Br

Ca

Mo X-ray tube Si-pin detector

Fe

101 0

2

4

6

8

10

12

14

16

18

20

22

24

26

X-ray energy (keV)

Figure 7. Detection limits for the Si-PIN and Cd-Te detectors for the Mo X-ray tube (0.3 mA, 45 kV) from solid reference materials NIST SRM 2704, 2711, 1573a and a Degussa gold alloy (Au 585/Ag 340).

the high energy range, the background in the spectrum is much higher and the corresponding detection limits are larger. On the other hand, several reference standard materials from NIST were used to evaluate the detection limits for solid samples analysis: SRM 2704 Buffalo river sediment, SRM 2711 Montana sediment unpurified, SRM 1573a Tomato leaves and a homogeneous gold alloy reference 585/340 from Degussa were irradiated by 5 min using the Mo X-ray tube and the SiPIN detector. The X-ray tube power parameters are 45 kV and 0.3 mA. Considering the same detection limit criteria as above, the graph shown in Fig. 7 was obtained. For sediments and the plant reference, thedetection limits may reach 40 µg/g for Cu and 25 µg/g for other heavier elements, like Sr. Detection limit for Fe increases to 100 µg/g while for Ca and Cl, the detection limits are 400 and 1000 µg/g, respectively. We may expect similar results for other materials with a matrix similar to sediments (stone, glass). For the gold matrix, Cu has a detection limit of 450 µg/g and Ag may reach 2400 µg/g while the Au may attain 1340 µg/g. From the previous measurements, we can conclude that a Mo or Rh X-ray tube with a Si-PIN detector represent the best combination for a first general analysis including light elements. Cd-Te detector may be useful for the detection of higher X-ray energies or heavier elements. The combined use of these X-ray detectors may also be considered.

SomeApplicationstoCulturalHeritageStudies For research in cultural heritage and a first characterization of a collection or specific objects, XRF  in situ is a powerful analytical tool. XRF analysis provides enough information in short time to establish the nature of the materials (inorganic or organic) and the use of complementary techniques. After a collection study by XRF, it is possible to select representative objects for further studies in the laboratory or, if necessary, to determine a strategy of sampling with minimum damage to the objects. This analytical approach has been adopted for the study of Mexican collections by our interdisciplinary group. After the application of imaging techniques and microscopical examination, XRF is applied and

 X-Ray Spectrom. 2010, 39, 338–345

Pre-Columbian gold and silver artifacts

In Mexico, few pre-Columbian artifacts and collections of precious metals are preserved. Most important items are kept in national museumsandduetotheirimportanceitisverydifficulttotransport them to laboratories for analytical studies, and sampling is not allowed or it is very limited. In some cases, non-destructive and non-invasivestudieswerecarriedoutbyPIXEonareducedamount of gold items[45] and previously only few artifacts were irradiated in situ usingportable XRF.[46] Forthis reason, scarce information on Mesoamerican gold work exists and there were insufficient data concerning the silver artifacts before our research. Certainly, the richest and most important collection with an archaeological context from Mesoamerica is the Tomb 7 from MonteAlban,centralOaxacaintheSouthofMexico,corresponding to the Mixtec Culture of the late postclassic period (1200–1521 A.D.). In the case of this collection, few electron microscope analyses (SEM-EDS)werecarriedout on a small numberof samples from selected items. [47] Recently, a set of portable equipments were transported to the Museo de las Culturas de Oaxaca in Oaxaca City to perform a non-destructive and non-invasive characterization. The state of conservation of thecollection is very good. Then, for the gold artifacts and most of the silver items there were unexpected important patina effects on the analytical results. Considering the sensitivity data, the Cd-Te detector and the Mo X-ray tube combination was chosen for this study as the metallic alloys are composed by medium and heavy elements. In particular, the detection of Ag K lines is important for this study as they provide more representative information of the bulk composition than Ag-L lines. In this way, more than 600 XRF analyses were performed on 49 artifacts within 5 days in the museum during the usual opening hours. X-ray tube conditions were 45 kV and 0.15 mA using a 1.5 mm beam spot. A spectrum required 60 s per region of analysis. Thenumber of measurements on each object varieddepending on the complexity of the artifact and its heterogeneity. Most complex artifacts such as necklaces required many analyses, and then the X-ray tube filament current wasincreased to 0.3 mA to reduce theacquisition time to 30 s and to obtain suitable spectra. A Degussa gold alloys set (including the Au/Ag contents 585/340, 750/120, 750/40, 585/140 and 900/40) as well as silver alloys (0.925, 0.720 sterling Ag) were irradiated under the same conditions to get a suitable system calibration.[48] AXIL program was used for X-ray peak area calculations. The elemental concentrations were obtainedfollowing a procedure described by Karydas.[26]

c  2010 John Wiley & Sons, Ltd. Copyright 

 

www.interscience.com/journal/xrs

 3   4  1 

J. L. Ruvalcaba Sil etal .

211

213

212

358 357 356 359

210

211

215

355

209

360

354

353

207

350 350

208

352

348

Figure 8. Goldpendant of Xochipilligod, piece A. Other four similarpieces completethe original set.Mixtec culture, Oaxaca,Mexico. ca 15th century.

 Two of the artifacts, including the regions of analysis, are shown in Figs 8 and 9: a gold pedant and a silver finger ornament. The corresponding results are plotted in the graphs of Figs 10 and 11, respectively. The high homogeneity of the alloys composition is not surprising.[45,49]  These artifacts were casted by false filigree lost wax procedure. For the gold pendant, the X-ray incident angle was modified until a grazing geometry of irradiation in flat areas was achieved; nevertheless, gold enrichment at the surface of the gold pendant was not observed. [50]  The mean composition of this artifact is Au 46.9%, Ag 32.8%, Cu 20.3%, similar to other three identical pendants of the collection. They may be produced in the same workshop with the same gold alloy. Their composition is quite differentif wecompareit with other pieces of thecollection(mean Au 77%, Ag 15%, Cu 8%). Concerning the silver ring elemental composition, this is similar to the rest of the silver items of the collection, with Cu average contents of 2% and very low amounts of Au. The same alloy was used for all the parts of the finger ornament, except for one of  the bells (region 356), attached by a plastic wire. We can assert that after the discovery of the tomb, this bell was added to the ornament butthis modification to complete theartifactis notright, as it does not match the expected composition of the artifact.  These are just few results[49] ; a detailed paper with the full set of artifacts and data of the collection is in progress. Studies of early colonial codices

 3   4  2  

Inthepre-Hispanicandearlycolonialcodicesanimportantamount of knowledge is preserved including religion and traditions of  the ancient Mexican people. Most of these manuscripts were destroyed during the catholic evangelization in the 16th and 17th centuries. Someof these documents survived, and nowadays they are kept in European collections. In the Mexican collections, only one pre-Hispanic codice, the Colombino, is preserved in the

www.interscience.com/journal/xrs

 

351

Figure 9. Silver finger ornament of eagle. Other three similar pieces completethe original set. Mixtec culture,Oaxaca, Mexico.ca 15th century.

65  Cu  Ag  Au

60 55 50    )    %    ( 45   n   o 40    i    t   a   r    t   n 35   e   c   n   o 30    C

25 20 15 10   6    7   8   9   0   1   2   3   4    5   6    7   8   9   0   1   2   2  0   2  0   2  0   2  0   2  1   2  1   2  1   2  1   2  1   2  1   2  1   2  1   2  1   2  1   2  2   2  2   2  2

Figure 10. Elemental contents of Cu, Ag and Au determined by XRF for the Xochipilli god pendant, piece A.

Biblioteca Nacional de Antropologia e Historia (INAH) with about other 40 colonial codices written after the Spanish conquest. A research project on this kind of manuscripts has been developed to determine the materials used in these documents, to understand how they were written and to propose preventive conservation strategies. In fact,there are scarce analyses on codice materials.[42,51]  The SANDRA system and other portable equipments were transported to the security areas of the library to carry out the analysis of the mostimportant codicesof theMexicancollection.In

c  2010 John Wiley & Sons, Ltd. Copyright 

 

X-Ray Spectrom. 2010, 39, 338–345

SANDRA portable XRF system

100 95 90    ) 85    %    (   n 80   o    i    t   a   r    t   n   e   c 10   n   o    C

Cu Ag Au

1

0.1   8   9    5  0    5  1    5  2    5  3    5  4    5    5    5  6    5    7    5  8    5  9   6  0   3  4   3  4   3   3   3   3   3   3   3   3   3   3   3

Figure 11. Elemental contents of Cu, Ag and Au determined by XRF for the silver finger ornament of eagle.

the first stage of research, the Colombino codice and two colonial codices (de la Cruz Badiano, Azoyu) were studied. [51] In this case, we discuss the analysis of one colonial codice. Among the most important early colonial manuscripts, the dela CruzBadiano codicewas written in 1552 in theFranciscan’s school for noble Indians (Santa Cruz de Tlatelolco) in Mexico City just 31 years after the conquest of the Aztec capital, Tenochtitlan. This European book style codice was discovered in the Vatican Library and it was given as a present to Mexico by the Pope in 1990.  The main subject of this document is the pre-Hispanic traditions of use of plants and minerals for medical purposes.[36]  The manuscript, written with iron-gallic inks within minium margins

(confirmed by our XRF analysis [52] ), has colored drawings of local plants, with their names written in red ink in ancient Mexican language – Nahuatl – and the description of the recipes written in Latin (Fig. 12).Specialattention wasfocusedon thecolor materials.  The first XRF analyses were carried out using the Cd-Te detector and the Mo X-ray tube (45 kV and 0.3 mA using 1-mm beam spot) as mineral pigmentscould be found.After a first general analysis of  several pages at the beginning, middle and end of the document, only arsenic in onekind of yellowcolor, probably dueto orpiment, and high iron contents (earth pigments) in brown color were detected. Then, green, blue, red and other kinds of yellow color may be organic as their spectra are similar to the paper one. [52] A second round of analysis and newmeasurements were carried out using the Si-PIN detector and similar X-ray tube conditions. As an example, Fig. 12 shows the pages 9 and 12 recto of the de al Cruz Badiano codex with the analyzed regions. The main results are plotted in the graph of Fig. 13.  The identification of two yellow was confirmed, one organic and the other one (B191) fitting the orpiment composition (As, S). The white color (B137, B195) is related to Ca and S signals, i.e. gypsum. Brown and ochre are clearly correlated to high amounts of Fe, and in some cases Mn, typical of earth pigments (B139, B194). Sometimes they can be mixed with orpiment. Despite the variability observed for the green, blue and red colors, mainly due to the overlapping of the color layers and color combinations with gypsum and orpiment, the intensities of the main detected elements for these colors are very similar to the paper signals. Considering the previous data collected with the Cd-Te detector, we conclude that these colors must be organic (dyes) and the presence of Si, K and Ca indicate that clays could be used to fix the dye, such as in the Maya blue manufacture with indigo and palisgorskite.[53]  This preparation of colors is related to pre-Hispanic origin and it does not correspond to European

188 141 189

143

190

194

191

142

192 193

136 195

139 140

137 138

131 133 134

198 196

197

132

135 199

 3   4   3  

Figure 12. Nine and 12 recto pages of the dela Cruz Badiano codice (1552). XRF analyzed regions are shown.

 X-Ray Spectrom. 2010, 39, 338–345

c  2010 John Wiley & Sons, Ltd. Copyright 

 

www.interscience.com/journal/xrs

J. L. Ruvalcaba Sil etal .

folio 9r

white

folio 12r

brown

brown white

yellow    )   s    t   n   u   o   c    (   y    t    i   s   n   e    t   n    i

  y   a   r      X

green

104

blue red

green

Si S K Ca Fe As Pb

paper 103

102

101   1   6   8   0    7   2   9   1   9   2   3   4    5   9  0   1  3   1  3   1  3   1  4   1  3   1  4   1  3   1  9   1  8   1  9   1  9   1  9   1  9   1    D    D    D    D    D    D    D    D    D    D    D    D    D    D    B    B    B    B    B    B    B    B    B    B    B    B    B    B

Figure 13. X-ray intensities from the XRF spectra of the analyzed regions of Fig. 12, dela Cruz Badiano codice.

lacquers.[54] It is clear that the use of dyes and pigments for coloring followed a specific pattern, like in the case of the red color; minium (lead tetraoxide) never appears in the figures and it was exclusively used in the red margins and mixed in the red inks probably with a red dye.  The de la Cruz Badiano codice shows the syncretism of the preColumbian and European traditions of writing at the beginning of  the colonial period in Mexico. The Europeandocument format and book style, written with iron-gallic inks, integrates representations of plants with pre-Hispanic iconographical elements and a selectiveuse of dyesand pigmentsfollowing the pre-Columbian codice traditions. Only in the late colonial codices, dyes are replaced by European pigments.[42,51] Further analyses on the codices collection using XRF and other complementary techniques, such as Raman and Mid-FTIR in situ are in progress.

Conclusions

 3   4  4 

 This paper showed the main analytical features of the portable XRF spectrometer SANDRA, developed in the Instituto the Fisica, UNAM, Mexico. It has been used forvarious applications including technical studies of metallic artifacts (gold, bronzes), analyses of  pre-Columbian and colonial manuscripts, studies of painting for palette and painter’s techniques, conservation studies of historic photographs as wellas characterizationof pre-Hispanic turquoises and green stone objects.  The SANDRA system provides outstanding information for materials identification and use of materials, technological studies and for conservation and restoration purposes. Under the actual configuration of the SANDRA system, the Mo and/or Rh X-ray tubes combined with the Si-PIN detector is the most appropriate setup for a first general study. The combination of detectors and X-ray tubes providescomplementary information for cultural heritage with a convenient set of detected elements and X-ray energy ranges.  This device represents a suitable choice using low cost semiconductor X-ray detectors and standard X-ray tubes. This fact has to be considered for limited budgets like in the case of  Latin American countries.

www.interscience.com/journal/xrs

 

Acknowledgements

 This researchis part of the interdisciplinary Mexican project MOVIL: Non-destructive methodologies for the In situ Study of the Cultural  Heritage   supported by CONACYT Mexico grant U49839-R. The ´ authors thank K. Lopez, J. Beristain, J. G. Morales and J. C. Pineda for their technical support as well as E. Hernandez Vazquez for the de la Cruz Badiano  codex and the gold and silver artifacts photographs.  The studyof goldand silverpre-Columbian itemswas carriedout in the Museo Regional de Oaxaca, Oaxaca, Mexico with thecollaboration of the INAH Center Oaxaca and the Conservation Metals ´ y Workshop of Escuela Nacional de Restauracion, Conservaci on Museografia (INAH), as well as the Instituto de Investigaciones ´ Antropol ogicas and Instituto de Investigaciones Este´ ticas, UNAM.  The studies of Mexican codexes were carried out in collaboration with the Biblioteca Nacional de Antropologia e Historia, INAH ´ and Laboratorio de Diagn ostico de Obras de Arte of Instituto de Investigaciones Est´eticas, UNAM.

References [1] K. Janssens, X-ray based methods of analysis, in   Non Destructive Microanalysis of Cultural Heritage Materials, vol. 42,   Wilson and  Wilsons Comprehensive AnalyticalChemistry  (Eds: K. Janssens,R. Van Grieken), Elsevier Science: Amsterdam, 2004, pp 129. [2] A. G. Karydas, X. Brecoulaki,Th. Pantazis,E. Aloupi,V. Argyropoulos, D. Kotzamani,R. Bernard, Ch.Zarkadas, Th.Paradellis,Importance of  in-situ EDXRF measurements in the preservation and conservation of Material Culture, in   X-Rays for Archaeology  (Eds: M. Uda, G. Demortier, I. Nakai), Springer: Dordrecht, The Netherlands, 2005, 27. [3] R. Cesareo,G. E. Gigante,P. Canegallo,A. Castaellano,J. S. Iwanczyk, A. Dabrowski, Nucl.Instrum.Meth. A 1996, 380, 440. [4] A. Castellano, R. Cesareo, Nucl. Instrum. Meth. B 1997, 129, 281. [5] A. Longoni, C. Fiorini, P. Leutenegger, S. Scuti, G. Fronterotta, L. Struder, ¨ P. Lechner, Nucl. Instrum. Meth. A 1998, 409, 407. [6] R. Cesareo, G. E. Gigante, A. L. Hanson, Nucl. Instrum. Meth. B 1998, 145, 434. [7] J. L. Ferrero, C. Rold´an, N. Ardid, E. Navarro,  Nucl. Instrum. Meth. A 1999, 422, 868. [8] J. Kunicki-Goldfinger, J. Kierzek, A. Kasprzak, B. Małoz˙ ewska-Bu´cko,  X-Ray Spectrom. 2000, 29, 310.

c  2010 John Wiley & Sons, Ltd. Copyright 

 

X-Ray Spectrom. 2010, 39, 338–345

SANDRA portable XRF system [9] K. Janssens, G. Vittiglio, I. Deraedt, A. Aerts, B. Vekemans, L. Vincze, F. Wei, I. Deryck, O. Schalm, F. Adams, A. Rindby, A. Kn¨ochel, A. Simionovici, A. Sinigirev, X-Ray Spectrom. 2000, 29, 73. [10] M. Mantler, M. Schreiner, X-Ray Spectrom. 2000, 29, 3. [11] P. Moioli, C. Seccaroni, X-Ray Spectrom. 2000, 29, 48. [12] J. L. Ferrero, C. Rold´a n, D. Juanes, E. Rollano, C. Morera,   X-Ray  Spectrom. 2002, 31, 441. [13] C. Ricci, I. Borgia, B. G. Brunetti, C. Miliani, A. Sgamellotti, C. Seccaroni, P. Passalacqua, J. Raman Spect. 2004, 35, 616. [14] Ch. Zarkadas, A. G. Karydas, X-Ray Spectrom. 2004, 33, 447. [15] V. Desnica, M. Schreiner, X-Ray Spectrom. 2006, 35, 280. [16] Ph.J Potts, M. West. (Eds.) Portable X-ray Fluorescence Spectrometry: Capabilities for In Situ Analysis, Royal Society of Chemistry: Cambridge, UK, 2008. [17] A. Sokolov, A. Loupilov, V. Gostilo, X-Ray Spectrom. 2004, 33, 462. [18] M. Ferretti, Nucl.Instrum.Meth. B 2004, 226, 453. [19] A. G. Karydas, Ch. Zarkadas, A. Kyriakis, J. Pantazis, A. Huber, R. Redus, C. Potiriadis, T. Paradellis, X-Ray Spectrom. 2003, 32, 93. [20] Z. Sz¨okefalvi-Nagy, I. Demeter, A. Kocsonya, I. Kov´acs, Nucl.Instrum. Meth. B 2004, 226, 56. [21] C. Rold´an, J. Coll, J. L. Ferrero, D. Juanes, X-Ray Spectrom. 2004, 33, 28. [22] A. S. Serebryakov,E. L.Demchenko,V. I. Koudryashov,A. D.Sokolov, Nucl. Instrum. Meth. B 2004, 213, 699. ˇ ˇ [23] K. Tantrakarn, N. Kato, A. Hokura, I. Nakai, Y. Fujii, S. GluˇsScevi S.  X-Ray Spectrom. 2009, 38, 121. [24] K. Uhlir, M. Griesser, G. Buzanich, P. Wobrauschek, C. Streli, D. Wegrzynek,A. Markowicz, E. Chinea-Cano, X-RayS pectrom. 2008, 37 , 450. [25] I. Nakai, S. Yamada, Y. Terada, Y. Shindo, T. Utaka, X-Ray Spectrom. 2005, 34, 46. [26] A. G. Karydas, D. Kotzamani, R. Bernard, J. N. Barrandon, Ch. Zarkadas. Nucl.Instrum. Meth. B 2004, 226, 15. [27] C. Vazquez-Calvo, B. G´omez Tubío, M. Alvarez de Buergo, I. Ortega Feliu, R. Fort, M. A. Respaldiza, X-Ray Spectrom. 2008, 37 , 399. [28] M. Uda, Nucl.Instrum. Meth. B 2004, 226, 75. [29] J. L. Ferrero, C. Roldan, D. Juanes, J. Carballo, J. Pereira, M. Ardid. J. L. Lluch, R. Vives, Nucl.Instrum.Meth. B 2004, 213, 729. [30] R. Cesareo, A. Brunetti, S. Ridolfi, X-Ray Spectrom. 2008, 37 , 309. [31] A. Castellano, G. Buccolieri, S. Quarta, M. Donativi, X-Ray Spectrom. 2006, 35, 276. ¨ [32] S. Rohrs, H. Stege, X-Ray Spectrom. 2004, 33, 396. [33] A. Guilherme,A. Cavaco, S. Pessanha, M. Costa,M. L. Carvalho, X-Ray  Spectrom. 2008, 37 , 444. [34] M. Aceto, A. Agostino, E. Boccaleri, A. Cerutti Garlanda,   X-Ray  Spectrom. 2008, 37 , 286. [35] L. Pappalardo, A. G. Karydas, N. Kotzamani, G. Pappalardo, F. P. Romano, Ch. Zarkadas. Nucl.Instrum. Meth. B 2005, 239, 114. [36] A. Gianoncelli, J. Castaing, A. Bouquillon, A. Polvorinos, P. Walter,  X-Ray Spectrom. 2006, 35, 365. [37] A. Gianoncelli, J. Castaing, L. Ortega, E. Dooryh´ee, J. Salomon, P. Walter, J.-L. Hodeau, P. Bordet, X-Ray Spectrom. 2008, 37 , 418. [38] M. Uda, A. Ishizaki, R. Satoh, K. Okada, Y. Nakajima, D. Yamashita, K. Ohashi, Y. Sakuraba, A. Shimono, D. Kojima,  Nucl. Instrum. Meth. B 2005, 239, 77. [39] G. Chiari, Ph. Sarrazin, Portable Non Invasive XRD/XRF instrument:a new way of looking at objects surface,  9th International Conference on NDT of Art, Art2008, Jerusalem Israel, 25–30 May   2008, http://www.ndt.net/search/docs.php3?MainSource =65 (accessed in 2010). [40] C. R. Appoloni, P. S. Parreira, F. Lopes, Thirteen years of activities on Art; Archaeometry and Cultural Heritage Conservation at the State University of Londrina Applied Nuclear Physics Laboratory, ´  Latino Americano sobre Proceedings of LASMAC2007. 1o Simp osio

[41]

[42]

[43]

[44]

[45] [46] [47] [48] [49]

[50]

[51]

[52]

[53]

[54]

˜  de M´  etodos Físicos e Químicos em Arqueologia, Arte e Conservac¸ ao Patrimˆonio Cultural , CD. ISBN 978-85-98196-81-7, Sao Paulo, 2008. J. L. Ruvalcaba Sil, La pintura de caballete vista a trav´es de la fluorescencia de rayos X, in La Materiadel Arte: Jos´  e Maria Velazco y  Hermenegildo Bustos (Eds: T. Falcony, S. Zetina), CONACULTA-INBA Museo Nacional de Arte: M´exico, 2004, pp 81. J. L. Ruvalcaba, C. Gonz´alez Tirado, Ana´alisis in situ de documentos ´ historicos mediante un sistema port´atil de XRF, in  La Ciencia de Materiales y su Impacto en la Arqueología, vol. 2 (Eds: D. Mendoza, J. Arenasy, V. Rodriguezcoord), Lagares Ed., Academia Mexicana de ´ Ciencia de Materiales A.C.: Mexico, 2005, pp 55. ´ alezTirado.FRXport´atil L. TornerMorales,J. L. Ruvalcaba Sil,C. Gonz´ ´ y PIXE como Tecnicas Complementarias para el An´alisis de Libros Antiguos: Estudio de Guardas y Cantos Decorados, in La Ciencia de Materiales y su Impacto en la Arqueología, vol. 3 (Eds: D. Mendoza, J. Arenas, V. Rodriguez, J. L. Ruvalcaba), Lagares Ed., Academia Mexicana de Ciencia de Materiales A.C.: M´exico, 2006, pp 91. N. Schulze, ‘‘For Whom the Bells Tolls’’ Mexican copper bells from  Templo Mayor Offering; Analysis of the production process and it cultural context, in  Materials Issues in Art and Archaeology VIII , vol. 1047 (Eds:P. Vandiver, B. McCarthy, R. Tykot, J. L. Ruvalcaba-Sil, F. Casadio), Materials Research Society: Warrendale, Pennsylvania, 2008, pp 195. J. L. Ruvalcaba-Sil, G. Demortier, A. Oliver.  International Journal of  PIXE  1995, 5, 273. R. Cesareo, G. E. Gigante, J. S. Iwanczyk, M. A. Rosales, M. Aliphat, ´ P. Avila. Rev.Mex. Fís. 1994, 2, 301. G. A. Camacho-Bragado, M. Ortega-Aviles, M. A. Velasco, M. Jose´ Yacaman, J. Met. 2005, 57 (7), 19. A. G. Karydas, Ann. Chim. 2007, 97 (7), 419. G. Pe˜nuelas, J. L. Ruvalcaba, J. Contreras, E. Hern´andez, E. Ortiz,Non Destructive   In situ   Analysis of Gold and Silver Artifacts from the  Tomb 7 of Monte Alban, Oaxaca, Mexico. Proceedings of the 37th International Symposiumon Archaeometry , Siena, 2008, in print. G. Demortier, J. L. Ruvalcaba-Sil, Non-destructive ion beam techniques for the depth profiling of elements in Amerindian gold jewellery artifacts, in   Cultural Heritage Conservation and  Environmental Impact Assessment by Non-destructive Testing and  Microanalysis (Eds: R. Van. Grieken, K. Janssens), A.A. Balkema Publishers: London, 2005, pp 91. ´ ´ E. Hern´andez, S. Zetina, J. L. Ruvalcaba, M. Lopez C´eceres,T. Falcon, C. Gonz´alez, E. Arroyo, Non destructive   In situ   Study of Mexican Codexes: Methodologyand FirstResults of Materials Analysis forthe Colombino and Azoyucodexes,Proceedingsofthe37thInternational  Symposiumon Archaeometry   Siena, 2008, in print. ´ S. Zetina, J. L. Ruvalcaba, T. Falc´on, E. Hernandez, C. Gonz´alez, E. Arroyo. Painting syncretism: a non destructive analysis of the badiano codex,   9th International Conference on NDT of Art, Art2008, Jerusalem Israel, 25–30 May   2008, http://www.ndt.net/search/docs.php3?MainSource =65 (accessed in 2010). M. S´a nchez del Río, P. Martinetto, A. Somogyi, C. ReyesValerio, E. Dooryhee, ´ N. Peltier, L. Alianelli, B. Moignard, L. Pichon,  T. Calligaro, J.-C. Drane, Spectrochim. Acta Part B: Atom. Spect. 2004, 59, 1619. E. Arroyo, T. Falc´on,E. Herna´ ndez,S. Zetina,A. Nieto, J. L. Ruvalcaba, XVI century colonial panel paintings from New Spain: material reference standards and non destructive analysis for mexican retablos,   9th International Conference on NDT  of Art, ART2008, Jerusalem, Israel, 25–30 May   2008, http://www.ndt.net/search/docs.php3?MainSource =65 (accessed in 2010).

 3   4   5 

 X-Ray Spectrom. 2010, 39, 338–345

c  2010 John Wiley & Sons, Ltd. Copyright 

 

www.interscience.com/journal/xrs

Sponsor Documents


Recommended

No recommend documents

Or use your account on DocShare.tips

Hide

Forgot your password?

Or register your new account on DocShare.tips

Hide

Lost your password? Please enter your email address. You will receive a link to create a new password.

Back to log-in

Close