Spectrochimica Acta Part B 59 (2004) 1611 – 1618 www.elsevier.com/locate/sab
A portable semi-micro-X-ray fluorescence spectrometer for archaeometrical studies $ Ch. Zarkadas*, A.G. Karydas Laboratory for Material Analysis, Institute of Nuclear Physics, N.C.S.R. bDemokritosQ, Athens 153 10, Greece Received 1 November 2003; accepted 1 May 2004 Available online 11 September 2004
Abstract A portable semi-micro-X-ray fluorescence (A-XRF) spectrometer was developed in the Laboratory for Material Analysis of the N.C.S.R bDemokritosQ. It utilizes a novel end-window, battery-operated, low-power X-ray tube (40 kV, 40 AA) with Au as anode material, a peltier cooled Si-PIN X-ray detector and associated electronics. The unique design of the probe-like X-ray tube anode allows very close coupling of any optical component to the tube anode, as well as to the sample position. A 240 Am pin-hole collimator was used to form the semimicrobeam. Monte Carlo calculations, as well as several sets of measurements, were performed, in order to determine the optimum geometrical and operational parameters. Preliminary results about the performance of our spectrometer are presented and compared to those reported in the literature for other micro-XRF instruments utilizing various optical elements (pin-holes, poly-capillary lenses) for focusing Xrays. The potential of this semi-micro-XRF spectrometer in the archaeometrical research is also discussed. D 2004 Elsevier B.V. All rights reserved. Keywords: A-XRF; Transmission anode; Portable; Non-destructive; Jewel; Soldering
1. Introduction The ability to produce X-ray microbeams to be used as analytical microprobes dates back to the early 1960s [1]. First attempts were based on the modification of commercially available X-ray spectrometers having as major components high power X-ray tubes, flat or curved analyzing crystals and various types of detectors [2–7]. Microanalysis was usually performed by insertion of differently shaped apertures, in various stages of the X-ray beam path, such as common pin-hole collimators and slits, lettering pen tips [2,4] and hypodermic needles [3]. In the following years advances in microfocus X-ray tubes and rotating anode systems accompanied by a rapid development of X-ray optical elements [1,8–14] have enabled the construction of $
This paper was presented at the International Congress on X-Ray Optics and Microanalysis (ICXOM XVII), held in Chamonix, Mont Blanc, France, 22–26 September 2003, and is published in the special issue of Spectrochimica Acta Part B, dedicated to that conference. *Corresponding author. Tel.: +30 210 6518770, fax: +30 210 6511215. E-mail address:
[email protected] (Ch. Zarkadas). 0584-8547/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.sab.2004.05.029
micro-XRF spectrometers for laboratory use [15–38] and boosted the analytical capabilities of synchrotron radiation facilities [39 40]. Nowadays, the application of micro-X-ray fluorescence (A-XRF) has substantially increased the potential of the conventional XRF technique providing also the possibility of two-dimensional mapping and three-dimensional imaging. In the field of archaeometry, the A-XRF technique has proven to be a valuable tool, since the reduction of beam sizes to the Am scale has increased significantly the analytical spatial resolution and allows the study of small details and/or remains on the artifacts. Several publications dealing with the analysis of objects of artistic or historical value by means of A-XRF have been released over the past few years, however, only few describe instrumentation appropriate for in-situ A-XRF analysis of archaeological materials [36,41]. The in-situ A-XRF analysis has eventually become one of the basic requirements of archaeologists and conservators, since the majority of objects of interest are usually immovable or difficult to be transported to a laboratory. Therefore, the demand for versatile, efficient, low weight, portable A-XRF equipment is continuously
1612
Ch. Zarkadas, A.G. Karydas / Spectrochimica Acta Part B 59 (2004) 1611–1618
growing. The construction of such devices requires careful selection over a wide range of available components [38] (micro-focus low-power X-ray tubes, small dimension detectors and associated electronics) and their integration into a compact and easily transportable instrument. Towards this direction this paper describes the development and study of a semi-micro-portable XRF spectrometer based on a low power X-ray tube, a pin-hole collimator and a Peltier cooled Si-PIN detector. The individual characteristics of the excitation source and their influence in the output tube spectrum are thoroughly examined. Optimization of geometrical and instrumental parameters is achieved through a series of measurements and Monte Carlo simulations. The performance of the spectrometer is presented in terms of elemental excitation yields (counts/s/(Ag/cm2)) and absolute detectable masses (pg) obtained from thin target measurements. The results obtained are compared to those of other AXRF instruments reported in the literature. The possibilities and limitations of this portable spectrometer in the analysis of metal alloys are also discussed. As a first archaeometrical application, we have examined and identified the type of the soldering used to join two parts of a pair of earrings dating to the Hellenistic period (2nd B.C.).
2. Theoretical The simplest and probably most inexpensive means for producing a sub-mm or even a micro-X-ray beam are pinhole collimators and apertures. Pinhole collimators or apertures, although not the most efficient optics, have been widely used [2–8,15,17–19,28,33,36] to produce X-ray beams down to 2–10 Am [8]. For hard X-rays pin-hole collimators are typically laser drilled discs of suitable materials such as W or Pt, having usually a thickness of less than 1 mm. The size of the beam spot and the flux obtained depends crucially on the size of the electrons focal spot, the anode to pin-hole distance (s 1), the pin-hole to sample distance (s 2), the pin-hole diameter (d) and the thickness of the collimating disc (h). It is a common knowledge and general rule that the formation of a beam size close to the diameter of the pin-hole requires small angular divergences and consequently the placing of the pin-hole close to the sample position. However, a more detailed study of the effect of the aforementioned geometrical parameters on the beam spot size can be performed only by application of computerbased simulations. Simulated beam profiles by means of a Monte Carlo algorithm developed in this work are presented in Fig. 1a,b for variable anode to pin-hole distances and pinhole diameters, respectively. For these simulations a fixed anode to sample distance of 14 mm and a Gaussian shaped electrons focal spot having a Full Width at Half Maximum (FWHM) of 150 Am was assumed. As can be seen from Fig. 1a, for a pin-hole collimator of thickness h=600 Am and pin-hole diameter of d=240 Am (larger than the electron focal spot), an increase in the ratio s 1/s 2 from 0.5 up to 2.5, reduces
Fig. 1. Monte Carlo simulations of beam profiles in the case of: (a) pin-hole collimator of thickness h=600 Am, pin-hole diameter d=240 Am and different anode to pin-hole (s 1)–pin-hole to sample distances (s 2); (b) pinhole collimator of thickness h=600 Am, anode to pin-hole distance (s 1)=10 mm, and variable pin-hole diameters d. Simulations refer to a Gaussian electrons focal shape of 150 Am at FWHM and for a total anode to sample distance of 14 mm.
the degree of the anode spot magnification through the pinhole from almost 500% to about 200%. Additionally, from Fig. 1b it can be observed that towards smaller pin-hole’s diameters there is a progressive change of the beam spot profile from a step-like distribution to a distribution exhibiting appreciable tailing due to umbra and penumbra effects [19]. According also to our simulations, the FWHM of the exciting beam is not influenced significantly by changes of the collimator’s thickness within the range of manufacturing capabilities (0.6–1 mm).
3. Experimental 3.1. The A-XRF spectrometer The A-XRF spectrometer designed and constructed in the Laboratory for Material Analysis N.C.S.R bDemokritosQ is shown in Fig. 2. The exciting source of this equipment is
Ch. Zarkadas, A.G. Karydas / Spectrochimica Acta Part B 59 (2004) 1611–1618
1613
Fig. 2. Photograph of the semi-micro-XRF spectrometer: (a) PRS-400 X-ray tube (see also Fig. 3); (b) XR-100 CR Amptek detector; (c) Optical microscope; (d) X–Y positioning stage; (e) X–Y–Z positioning stage; (f) Al housing; (g) Bronze shielding of the Be probe; (h) W/Cu alloy probe shielding; (i) Pt disc collimator with 240 Am pin-hole; (j) W/Cu detector collimator.
shown in Fig. 3. It is a 1.6-W battery operated end-window Xray tube with a grounded Au anode (model PRS-400, Photoelectron, USA), which was initially manufactured for medical purposes. The X-ray tube consists of a metal body ending to a 200 Am thick Be—needle like—probe. A protective layer of Ni and an additional layer of Chrome Nitride on top of the Ni layer cover the interior of the probe. The gold anode layer, of 0.9 Am nominal thickness, is deposited onto the Ni protective layer and covers the internal radius of the half sphere at the end point of the probe. The electron focal spot onto the anode has a FWHM approximately equal to 150 Am. Focusing of the electron beam can be performed manually by use of two X–Y potentiometers. Operational voltage values of 0, 30, 35, 40 kV and tube current values of 0, 1.25, 2.5, 5, 10, 20, 40 AA can be selected. The Be probe’s end cover is made of a tungsten–copper alloy rod (W 72%, Cu 28%, Goodfellow) lying on a X–Y positioning stage (MellesGriot). A 240 Am pin-hole Pt disc is tightly fixed in a groove at the front face of the tungsten–
Fig. 3. Drawing of the PRS-400 end-window, gold-anode, X-ray tube.
copper alloy component. In this configuration, the Pt disc can move relatively to the anode, therefore exact alignment of the pin-hole with respect to the electron focal spot can be achieved. Fluorescence radiation emitted by the sample is recorded by a Si-PIN Peltier cooled X-ray detector (XR-100CR Amptek, Bedford, MA, USA) of 25 mm2 crystal area and 300 Am crystal thickness, having a resolution of 190 eV at the Mn-Ka X-ray peak. An external tungsten–copper collimator restricts the detector’s effective crystal area to about 19 mm2. Data acquisition is performed by means of the pocket MCA (model MCA8000A, Amptek) connected to the serial port of a laptop computer. Visualization of the analyzed spot on the object’s surface is performed by using an optical microscope (Edmund optics, 15 magnification) with a calibrated crosshair ruler. The microscope is aligned in such a way, so that its focal point coincides with the intersection point of the detector’s axis and the tube’s Be probe axis. For laboratory or in-situ measurements of small objects, a X–Y–Z positioning stage (MellesGriot) is being used. In the case of large and immovable objects, the spectrometer can be placed onto a transportable X–Y–Z base. In order to integrate all the A-XRF components in their final geometry, the requirement that the spectrometer must be able to analyze objects with various surface shapes was of prior consideration. Consequently, a distance of at least a few mm should be available between the pin-hole and the sample position. According to our simulations, presented in Fig. 1a, a satisfactory beam spot size can be achieved by placing the pin-hole collimator at a distance from the anode, which is at least 2.5 times the distance of the pin-hole to the sample position. Under these considerations, an anode to aperture distance of 9.7 mm, an aperture to sample distance
1614
Ch. Zarkadas, A.G. Karydas / Spectrochimica Acta Part B 59 (2004) 1611–1618
of 4 mm and a sample to detector distance of 12 mm were adopted. 3.2. The X-ray tube spectrum The description of the X-ray tube emission spectrum can be considered as a prerequisite for performing an accurate quantitative analysis. In some cases, the description of the tube spectrum is not required to be performed in absolute terms, but at least to fit qualitatively the various spectrum components. In order to obtain an analytical description of the PRS400 tube spectrum, direct measurements were performed at 30 kV operational voltage. The obtained tube spectrum was simulated by applying a least squares fitting procedure based on the semi-empirical model proposed by Ebel et al. [42]. This model describes transmission X-ray tube spectra having an anode thickness, which is much greater than the maximum electron penetration depth. In our case, the mean range of electron penetration n ave(E), which corresponds to an emitted photon of energy E, is almost 0.25 Am for 10 keV electrons in gold and thus comparable to the anode’s thickness n anode. Therefore, we have modified the selfabsorption factor of Ref. [42] by assuming that the forward emitted photons are exponentially attenuated in a distance equal to (n anode n ave(E)). The simulated spectrum was next corrected for the absorption of the emitted X-rays in the air path between the anode and the detector, the absorption in the probe’s cover materials and for the detector’s intrinsic efficiency. Excellent agreement was found between the simulated and the measured spectrum presented in Fig. 4. 3.3. Experimental determination of optimum operational parameters One of the main characteristics of the X-ray tube used in this work is the possibility of performing manual focusing of the electron beam onto the gold anode material by means Fig. 5. Scan of the output tube flux, as a function of the X–Y grid settings: (a) Coarse scan; (b) Fine scan.
Fig. 4. Measured and simulated output X-ray spectrum of the PRS-400 X-ray tube at 30 kV.
of two adjustable X–Y potentiometers. Since we refer to a transmission anode X-ray tube, a change in the electron focal spot may affect the tube’s output X-ray intensity depending on the uniformity of the anode layer. Therefore, the optimum X–Y potentiometer settings, which provide the highest available X-ray flux while at the same time minimize parasitic X-ray lines originating from the Be probe’s internal cover materials (mainly of Ni), have to be experimentally determined. Towards this aim, the X-ray detector, properly collimated, was placed at a distance of 25 cm from the edge of the uncollimated Be probe. A Mo foil of 50 mg/cm2 was inserted in front of the detector. In this geometry the detector’s input count rate (ICR) was recorded for 30-s
Ch. Zarkadas, A.G. Karydas / Spectrochimica Acta Part B 59 (2004) 1611–1618
intervals for various X–Y grid values at 30 kV and 1.25 AA tube settings. The data obtained after a coarse and fine X–Y scan are displayed in Fig. 5a,b. From these data it was concluded that the tube’s maximum output flux is almost identical for five pairs of X–Y values. For each one of these pairs, scattered beam spectra were accumulated at 30 kV operational voltage, using a polyethylene scatterer. A maximum Au-La/Ni-Ka ratio equal to about seven was found for one of them, which was adopted as the optimum pair of X–Y potentiometer settings. In order to ensure now the maximum output photon rate also through the pin-hole of the Pt disc collimator, an alignment procedure was next applied. The tube’s tungsten– copper alloy end cover was put in position and the collimated detector was placed directly in front of the pinhole at a distance of 4 cm. With the X-ray tube operating at 30 kV and using the X–Y positioning stage, a X–Y scan was performed in 40 Am steps until the maximum count rate was observed. The spectral quality of the exciting beam after the alignment was examined in our final geometry by measuring again a polyethylene scatterer at 30 kV (Fig. 6). In order to reduce further the Cr-Ka and Ni-Ka signals originating from the Be probe’s interior layers, an Au foil of 8.8 mg/cm2 thickness was inserted in the incident X-ray beam path. This resulted to a complete elimination of the Cr-Ka signal (Fig. 6) and to an improvement of the Au-La/Ni-Ka ratio by a factor of 4.1 compared to the unfiltered exciting beam. Subsequently, the insertion of the Au filter causes a reduction in the absolute Au-La yield by a factor of 2.5, which is, however, inevitable for the minimization of parasitic X-ray lines. Finally, the dimensions of the exciting beam on the sample position were determined by fluorescence measurements of a 50 Am Tungsten wire (stretched by use of two screws and placed onto the X–Y–Z positioning stage (Melles Griot)), scanning the X-ray beam in 75 Am steps. The results of the X and Y beam scans are presented in Fig. 7a and b.
1615
Fig. 7. Beam profiles obtained by scanning the incident beam using a 50Am W wire: (a) X-axis beam profile; (b) Y-axis beam profile.
Apparently, the beam at the sample position has a rather elliptical shape with the two axes having a FWHM of about 450 and 530 Am, respectively. Moreover, the volcano-like shape indicates probably a non-uniform anode layer. The maximum irradiated area at a distance of 4 mm from the pin-hole is estimated to be 0.5 mm2 with the core of the beam being focused in an area of about 0.2 mm2.
4. Results and discussion 4.1. Thin target measurements and MDLs in metal alloys
Fig. 6. Spectra of the scattered tube exciting X-ray beam from a thick polyethylene material at 30 kV, before and after the insertion of an 8.8 mg/ cm2 Au filter in the exciting beam’s path.
The performance of the spectrometer was initially evaluated in terms of elemental sensitivities and minimum detection limits (MDLs) derived from thin target measurements. The thin targets were prepared by evaporation of pure elements or compounds onto an 8 Am stretched Kapton foil and measured at 30 kV, 40 AA for 1000 s live measuring time. The elemental sensitivities expressed in counts/s/(Ag/ cm2) and the MDLs expressed in absolute masses (pg), respectively, are presented in Table 1. Elemental sensitivities
1616
Ch. Zarkadas, A.G. Karydas / Spectrochimica Acta Part B 59 (2004) 1611–1618
Table 1 Elemental sensitivities (counts/s/(Ag/cm2)) and MDLs (pg) obtained from thin target measurements at 40 kV, 30 AA for 1000 s live measuring time Target
Thickness (Ag/cm2)
Sensitivity (counts/s/(Ag/cm2))
MDL (pg)
Ca Ti Mn Cu
51 109 218 289
0.096 0.19 0.43 0.79
548 419 219 248
in the range 0.1–0.8 counts/s/(Ag/cm2) were obtained for the elements Ca–Cu, with the corresponding MDLs ranging from 0.25 to 0.55 ng. Similar thin target yields and MDLs situated at the 1–10 ng level have been also reported in the literature for a compact XRF instrument with a Mo tube and 300 Am pin-hole, providing an irradiated sample area of 0.5 mm2 and operating at 45 kV, 0.2–0.5 mA [36]. Our elemental sensitivities are also comparable to those reported for a laboratory A-XRF spectrometer utilizing a rotating Cu, Mo anode X-ray tube and capillary optics [33]. In this case, the reported MDLs range down to 0.1–1 pg since the irradiated sample area is smaller by a factor of 1100 due to the use of a 15 Am capillary. However, it should be noted that the operating conditions of the laboratory scale instrument described in Ref. [33] refer to a tube that exceeds the maximum power capability of our tube by a factor of 15 000. In order to demonstrate further the analytical capabilities and limitations of our spectrometer in the case of metal alloy matrices, measurements of a high stainless steel and of a high purity gold standard sample were performed at 30 and 35 kV, respectively. Both samples were measured for 1000 s
Fig. 9. Comparison between the X-ray spectra obtained from the line scan of the soldering part on the earring’s surface.
live time and for a total counting rate of about 1200 cps at a tube current of 20 AA. For the high stainless steel, the MDLs obtained for some elements (Cr, Ni) range between 100 and 400 ppm. However, there are certain limitations in the detection of others, associated either with the polychromatic nature of the exciting tube radiation or with the inherent properties of the type of the detector used [36,43]. In the energy region 9–13 keV the detection ability of the spectrometer is restricted only to major constituents due to the presence of the scattered exciting Au-L lines. In the energy region below 8 keV, the presence of the intense Fe matrix lines prevents the sensitivity in the detection of lower Z elements to a certain extent, due to incomplete charge collection, escape peaks and shoulder peaks that have to be
Fig. 8. (a) Pair of gold earrings from the Benaki museum jewellery collection dating to the Hellenistic period (2nd B.C.); (b) Detail of the earring’s surface, depicting the soldering area and the positions analyzed (a–e).
Ch. Zarkadas, A.G. Karydas / Spectrochimica Acta Part B 59 (2004) 1611–1618
1617
5. Summary–conclusions
Fig. 10. Cu/Au and Ag/Au intensity ratios obtained by the measurements performed at the soldering part of the earring’s surface.
taken into account if accurate analysis of adjacent elements is to be performed (e.g, for Mn). For elements with characteristic X-ray lines above 13 keV the efficiency of the Si-PIN detector decreases compared to the efficiency of a conventional Si(Li) detector and excitation is performed only by the continuous bremsstrahlung radiation. However, due to the high Z anode material of the tube, bremsstrahlung radiation is sufficient to excite elements such as Mo and Nb down to 100 ppm concentrations. In the case of the high purity gold sample, a MDL of about 330 ppm was obtained for Ag. 4.2. An archaeometrical application: identification of hard soldering A pair of gold earrings from the Benaki museum jewellery collection, dating to the Hellenistic period (Fig. 8a,b), was measured in our A-XRF setup at 30 kV operating voltage and 20 AA tube current. A line scan was performed at the soldering part on the object’s surface with the corresponding spectra shown in Fig. 9. The increased Cu/ Au and Ag/Au ratios observed in the analysis of spot (c) (see Fig. 10) prove that a hard soldering procedure was applied. In this case, a lower melting point brazing ternary Au–Ag–Cu alloy with increased Ag and Cu concentrations compared to the corresponding concentrations of the earring’s alloy was used to join the two parts of both earrings. This brazing procedure can be clearly discriminated from the procedure of using a copper powder compound (chrysocolla or malachite) under reducing conditions to join the parts of the earrings through the solid state diffusion-bonding of copper [44]. If this was the case, then only an increased Cu/Au intensity ratio should be observed in the joining area. In order to check the absence of Cd, which has also been reported to be used in a brazing alloy, the analysis of the soldering area was repeated at 35 kV. However, no Cd was detected at the level of our minimum detection limits (about 200 ppm).
The development and study of a portable semi-micro-Xray fluorescence spectrometer has been described. The spectrometer is based on a novel end-window low-power transmission gold-anode X-ray tube and a Si-PIN peltier cooled X-ray detector. The unique design of the tube’s probe allows very close coupling of any optical component and makes the use of pin-hole collimators feasible and efficient. Sensitivities obtained from thin target measurements are similar to those reported in the literature for other compact or laboratory scale micro-XRF instruments. The size of the X-ray beam provides adequate spatial resolution required in many archaeometrical applications. As an example, the spectrometer was used successfully for the identification of the soldering technique that was applied in a pair of Hellenistic gold earrings. The use of a pin-hole aperture with diameter 100 Am is really feasible, without a significant worsening of the presented performance. As a next step, the replacement of the pin-hole collimator with a pollycapillary lens is expected to increase significantly the spatial resolution and the absolute MDLs of our spectrometer. Acknowledgments This work was granted by the bThera Foundation, Peter M. NomikosQ. The authors would like to thank P.M. Nomikos for his continuous support and encouragement, as well as for the supply of the PRS-400 X-ray source and associated electronics, G. Konstantato for his continuous administrative support, M. Dinsmore for fruitful discussions, N. Divis for his valuable contribution during the development of the spectrometer and N. Kotzamani from the Benaki Museum (Athens, Greece) for allowing us to analyze the pair of gold earrings. References [1] P.A. Pella, M. Lankosz, Highlights of X-ray spectrometry for microanalysis, X-ray Spectrom. 26 (1997) 327 – 332. [2] I. Adler, J. Axelrod, J. Branco, Further application of the intermediate X-ray probe, Adv. X-ray Anal. 2 (1960) 167 – 173. [3] R.D. Sloan, The X-ray spectrographic analysis of thin films by the milliprobe technique, Adv. X-ray Anal. 5 (1962) 512 – 515. [4] K.F.J. Heinrich, X-ray probe with collimation of the secondary beam, Adv. X-ray Anal. 8 (1962) 516 – 526. [5] W.J. Wittig, The use of an X-ray fluorescence semi-microprobe attachment in metallurgy, Adv. X-ray Anal. 8 (1965) (64) 248 – 258. [6] E.P. Bertin, Evaluation and Application of an improved slit probe for the X-ray secondary emission spectrometer, Adv. X-ray Anal. 8 (1965) 231 – 247. [7] T. Shiraiwa, N. Fujino, Micro fluorescent X-ray analyzer, Adv. X-ray Anal. 11 (1967) 95 – 104. [8] G.E. Ice, Microbeam-forming methods for synchrotron radiation, X-ray Spectrom. 26 (1997) 315 – 326. [9] A. Bjeoumikhov, N. Langhoff, R. Wedell, V. Beloglazov, N. Lebed’ev, N. Skibina, New generation of polycapillary lenses: manufacture and applications, X-ray Spectrom. 32 (2003) 172 – 178.
1618
Ch. Zarkadas, A.G. Karydas / Spectrochimica Acta Part B 59 (2004) 1611–1618
[10] N. Gao, I.Y. Ponomarev, Pollycapilary X-ray optics: manufacturing status, characterization and the future of the technology, X-ray Spectrom. 32 (2003) 186 – 194. [11] D.H. Bilderback, Review of capillary X-ray optics from the 2nd International Capillary Optics Meeting, X-ray Spectrom. 32 (2003) 195 – 207. [12] G. Hirsch, Metal capillary optics: novel fabrication methods and characterization, X-ray Spectrom. 32 (2003) 229 – 238. [13] M. Haschke, M. Haller, Examination of poly-capillary lenses for their use in micro-XRF spectrometers, X-ray Spectrom. 32 (2003) 239 – 247. [14] C. Macdonald, W.M. Gibson, Applications and advances in polycapillary optics, X-ray Spectrom. 32 (2003) 258 – 268. [15] M.C. Nichols, R.W. Ryon, An X-ray microfluorescence analysis system with diffraction capabilities, Adv. X-ray Anal. 29 (1986) 423 – 426. [16] A. Rinby, Applications of fiber technique in the X-ray region, NIM A 249 (1986) 536 – 540. [17] D.R. Boehme, X-ray microfluorescence of geological materials, Adv. X-ray Anal. 30 (1987) 39 – 44. [18] M.C. Nichols, D. Boehme, R. Ryon, D. Wherry, B. Cross, G. Aden, Parameters affecting X-ray microfluorescence analysis, Adv. X-ray Anal. 30 (1987) 45 – 51. [19] D.C Wherry, B.J. Cross, T.H. Briggs, An automated X-ray microfluorescence materials analysis system, Adv. X-ray Anal. 21 (1988) 93 – 98. [20] N. Yamamoto, Y. Hosokawa, Development of an innovative 5 Am phi focused X-ray-beam energy-despersive spectrometer and its applications, Jpn. J. Appl. Phys. Part II-Lett. 27 (1988) L2203 – L2206. [21] P. Engstrfm, S. Larson, A. Rindby, B. Stocklassa, A 200 Am X-ray microbeam spectrometer, NIM B 36 (1989) 222 – 226. [22] A. Rindby, P. Engstrfm, S. Larrson, B. Stocklassa, Microbeam technique for energy despersive X-ray fluorescence, X-ray Spectrom. 18 (1989) 109 – 112. [23] D.A. Carpenter, M.A. Taylor, C.E. Holcombe, Application of a laboratory X-ray microprobe to materials analysis, Adv. X-ray Anal. 32 (1989) 115 – 120. [24] S. Larsson, P. Engstrom, A. Rinby, B. Stocklassa, X-ray capillary microbeam spectrometer, Adv. X-ray Anal. 33 (1990) 623 – 628. [25] N. Shakir, S. Larsson, P. Engstrfm, A. Rindby, X-ray Microbeam spectroscopy: a biological application, NIM B 52 (1990) 194 – 198. [26] D.A. Carpenter, M.A. Taylor, Fast, high resolution X-ray microfluorescence imaging, Adv. X-ray Anal. 34 (1991) 217 – 221. [27] K. Furuta, Y. Nakayama, M. Shoji, H. Nakano, Y. Hosokawa, Intensity of X-ray microbeam formed by a hollow glass pipe, Rev. Sci. Instrum. 62 (1991) 828 – 829. [28] P.A. Pella, L. Feng, Fabrication and selected applications of a NIST X-ray microfluorescence spectrometer, Adv. X-ray Anal. 35 (1992) 1063 – 1067. [29] A. Rindby, P. Voglis, G. Nillson, B. Stocklassa, Software development for X-ray microbeam spectroscopy, Adv. X-ray Anal. 35 (1992) 1247 – 1251.
[30] D.A. Carpenter, M.A. Taylor, R.L. Lawson, High resolution microfluorescence imaging with a laboratory-based instrument, J. Trace Microprobe Tech. 13 (1995) 141 – 161. [31] B. Holynska, B. Ostachowitz, J. Ostachowitz, A. Ostrowski, J. Ptasinski, D. Wergynek, Multifunctional system for energy-dispersive X-ray fluorescence analysis, J. Trace Microprobe Tech. 13 (1995) 163 – 176. [32] A. Attaelman, P. Voglis, A. Rindby, S. Larsson, P. Engstrfm, Improved capillary optics applied to microbeam X-ray fluorescence: resolution and sensitivity, Rev. Sci. Instrum. 66 (1995) 24 – 27. [33] K. Janssens, B. Vekemans, L. Vincze, F. Adams, A. Rindby, A microXRF spectrometer based on a rotating anode generator and capillary optics, Spectrochim. Acta Part B 51 (1996) 1661 – 1678. [34] X. Ding, Y. He, Y. Yan, X-ray source for X-ray microfluorescence using a monolithic X-ray focusing lens combined with aperture optics, X-ray Spectrom. 26 (1997) 374 – 379. [35] Y. Hosokawa, S. Ozawa, H. Nakazawa, Y. Nakayama, An X-ray guide tube and a desk-top scanning X-ray analytical microscope, X-ray Spectrom. 26 (1997) 380 – 387. [36] G. Vittiglio, K. Janssens, B. Vekemans, F. Adams, A. Oost, A compact small-beam XRF instrument for in situ analysis of objects of historical and/or artistic value, Spectrochim. Acta Part B 54 (1999) 1697 – 1710. [37] S. Bichlmeier, K. Janssens, J. Heckel, P. Hoffmann, H.M. Ortner, Comparative material characterization of historical and industrial samples by using a compact micro-XRF spectrometer, X-ray Spectrom. 31 (2002) 87 – 91. [38] S. Bichlmeier, K. Janssens, J. Heckel, D. Gibson, P. Hoffmann, H.M. Ortner, Component selection for a compact micro-XRF spectrometer, X-ray Spectrom. 30 (2001) 8 – 14. [39] P. Suortti, W. Thomlinson, Medical applications of synchrotron radiation, Phys. Med. Biol. 48 (2003) 1 – 35. [40] G.E. Brown, N.C. Sturchio, An overview of synchrotron radiation applications to low temperature geochemistry and environmental science, Rev. Mineral. Geochem. 49 (2002) 1 – 115. [41] K. Janssens, G. Vittiglio, I. Deraedt, A. Aerts, B. Vekemans, L. Vincze, F. Wei, I. Deryck, O. Schalm, F. Adams, A. Rindby, A. Knochel, A. Simionovici, A. Snigirev, Use of microscopic XRF for non-destructive analysis in art and archaeometry, X-ray Spectrom. 29 (2000) 73 – 91. [42] H. Ebel, M.F. Ebel, C. Pfhn, Bernd Schogmann, Spectra of X-ray tubes with transmission anodes for fundamental parameter analysis, Adv. X-ray Anal. 36 (1993) 81 – 88. [43] A.G. Karydas, C. Zarkadas, A. Kyriakis, J. Pantazis, A. Huber, R. Redus, C. Potiriadis, T. Paradellis, Improvement of peak-to-background ratio in PIXE and XRF methods using thin Si-PIN detectors, X-ray Spectrom. 32 (2003) 93 – 105. [44] G. Demortier, The soldering of gold in antiquity, in Materials Res. Soc., Materials issues in art and archaeology, Mater. Res. Soc. Symp. Proc. 123 (1988) 193 – 198.
