Worldwide distribution of Total Reflection X-ray Fluorescence instrumentation and its different fields of application: A survey

Highlights

• According to the survey world maps show the distribution of TXRF equipment.

• Nearly 700 individual units are running actually in 57 countries of 6 continents.

• Users work at 200 universities, 60 synchrotron-beamlines, and 150 semiconductor fabs.

• 13 fields of applications (e.g. environmental, chemical) are evaluated statistically.

• Manufacturers, conference members and authors lead to 3 different pie-charts.

Abstract

A survey was carried out with users and manufacturers of Total Reflection X-ray Fluorescence instrumentation in order to demonstrate the worldwide distribution of TXRF equipment and the different fields of applications. In general, TXRF users come from universities and scientific institutes, from working places at synchrotron beam-lines, or laboratories in semiconductor fabs. TXRF instrumentation is distributed in more than 50 countries on six continents and is applied at about 200 institutes and laboratories. The number of running desktop instruments amounts to nearly 300 units. About 60 beamlines run working places dedicated to TXRF. About 300 floor-mounted instruments are estimated to be used in about 150 fabs of the semiconductor industry. In total, 13 different fields of applications could be registered statistically from three different aspects.

Introduction

Total Reflection X-ray Fluorescence (TXRF) was proposed as a special method for X-ray spectral analysis by Yoneda and Horiuchi in 1971. For a further development of the method within 20 years after, four scientists got the respected Bunsen–Kirchhoff prize of the “Gesellschaft Deutscher Chemiker” in 1991: H. Aiginger, P. Wobrauschek, J. Knoth, and H. Schwenke. Since 1986 several international workshops and conferences were held biannually. The 15th conference was held in Osaka, Japan, with nearly 200 attendees. Five different manufacturers built TXRF instruments during these years; only two have been left, but another three manufacturers started recently. Complete instruments as well as individual components are commercially available. A special module has been distributed as a low-cost attachment for existing X-ray equipment. Manufactured by the Atominstitut in Vienna, Austria, it has been installed in developing countries around the whole world under the auspices of the International Atomic Energy Agency (IAEA). It may be stressed that these devices are non-profit systems for academic use only.

In 2012, a working group was created as an open discussion forum for international communication in the field of TXRF and related methods. Facebook was chosen as the social network due to its widespread availability and its ease of use. The main facebook page describes the group as “Open forum for specialists in X-ray spectroscopic techniques where you can share knowledge, experiences, questions, and lines of the future of this technology”. The number of group members was 50 at the beginning of 2014; the administrator is Ramon Fernandez-Ruiz.

The only monograph on TXRF was published by John Wiley & Sons in 1997. The second edition will appear in 2014. Different review articles on TXRF or book contributions have also been published within the last decades summarizing new developments and results [1], [2], [3]. Specific articles deal with further developments such as excitation with synchrotron radiation [4], with semiconductor and wafer analysis [5] and with biological applications [6]. Some series appear periodically reporting on the progress in X-ray spectrometry, including TXRF; e.g. [7], [8].

TXRF is a universal method of multielement analysis which is quite economical with respect to costs of instrumentation and maintenance, as well as low in time consumption. TXRF is a non-consumptive method like all X-ray techniques. On the other hand, an entirely non-destructive analysis is mostly impossible because of the need for minute specimens.

The method is applicable to a great variety of sample materials, and a wide scope of applications can be outlined. Usually a minute specimen, i.e., a single particle, some grains of a fine powder, a thin layer or some small droplets, has to be deposited in the center of a flat and clean carrier. Mostly, a separate quartz-glass or Plexiglas carrier is used for each sample so that no memory effects occur. Quantification is performed by internal standardization, either by the addition of the analyte element itself or of another element (i.e., previously not present in the sample). No labor-intensive calibration is required. Because of the small sample volume needed for analysis, no matrix effects occur. As a result of these measures, quantification becomes easy and reliable and can be performed in a large dynamic range of 5 orders of magnitude.

Solutions are normally evaporated prior to analysis with the result of improved detection limits. Non-volatile liquids can be investigated directly but with lower detection power. For improved detection limits, their matrix has to be digested. In the case of complex samples, a decomposition and separation of the matrix become necessary. A lot of preparation techniques which are well-tried and tested for approved methods like ET-AAS or ICP-OES can also be applied for TXRF.

Simultaneous multielement detection is possible with detection limits in the low pg-range. Up to 80 elements can be determined, and low detection limits below 10 pg can be reached for many elements, with the exception of light or low-Z elements (Z < 11). For these elements, fine vacuum and ultrathin windows are needed. The detection of light and also of transition elements needs further and special efforts.

TXRF's unique capability for micro- or even ultra-microanalyses of small samples should be underlined. When only small sample amounts are available such a method is absolutely essential. No additional instrumental device is needed for TXRF as is the case for ET-AAS and ICP-MS. A minute specimen of about 10 μL or 10 μg has only to be deposited on a flat glass carrier.

In addition, TXRF is effectively applied to trace analyses of the elements when larger sample amounts are available. Aqueous solutions, high-purity acids, and body fluids can be analyzed down to the pg/mL region. Only a few droplets – representative for the sample – need to be pipetted on flat carriers and dried by simple evaporation. The total spectrum is recorded simultaneously, so that even an element that may be in the sample unexpectedly will be detected and no element will be overlooked.

Depth profiles of biogenous materials can be recorded after sectioning of the material with a freezing microtome. Sections of micrometer thickness can be investigated. Stratified near-surface layers of material science can be characterized by a stepwise planar sputter-etching and subsequent analysis of the remaining surfaces by TXRF. This technique of depth profiling is suitable for thin layers of nanometer thickness deposited on wafers, but it is destructive and time-consuming.

In contrast to this technique, another method related to TXRF is also suitable for deposited or implanted layers in the lower nm-range, however, it is non-destructive and moreover is fast. This variant is called Grazing Incidence-XRF (GI-XRF). For micro- and trace analyses, only one fixed angle position is used in order to realize excitation under total reflection. But for surface and thin-layer analyses by GI-XRF, the sample must be tilted in several steps of an angle scan. The technique is restricted to optically flat samples, e.g., wafers and glass disks which are simply or repeatedly coated or implanted. The evaluation carried out by an appropriate software-program requires a model of the system with several parameters for the individual layers. This model is varied by an iterative process until calculations and measurements are in an acceptable correspondence. Samples of a certain roughness (above 0.1 μm) blur the effect of total reflection but can be treated by special algorithms. New applications are concerned with semiconducting layers and even thin polymeric films.

The competitiveness of TXRF for micro- and trace analyses as well as for surface and thin-layer analyses with other efficient and well-established methods of atomic spectroscopy is undisputed. With respect to its capabilities, TXRF has surpassed conventional XRF by far and has attained a leading position in atomic spectroscopy. The outstanding features compete very well with those of inductively coupled plasma-mass spectrometry (ICP-MS), with electrothermal-atomic absorption spectrometry (ET-AAS), and with instrumental neutron activation analysis (INAA). For several applications, TXRF is superior because of its simplicity and rapidity. The detection power may be inferior but for many applications it is sufficient. In the field of surface analyses, TXRF is highly effective in the contamination control of wafers and therefore is a widespread analytical tool in the semiconductor industry. For analyses of thin stratified layers, GI-XRF is able to compete with the reference methods Rutherford backscattering (RBS) and secondary ion mass spectrometry (SIMS).

TXRF's simplicity of operation should be emphasized, supported by automated sample changing, measurement, and evaluation. The running costs are quite low and the maintenance of the instrumentation is rather easy. Several combinations of total-reflection with other spectral analytical methods such as XANES, XRR, XRD, and XPS show a present trend and future prospects of TXRF. Standardization by ISO, ASTM, or DIN is just being developed.