International
Tables for
Crystallography
Volume I
X-ray absorption spectroscopy and related techniques
Edited by C. T. Chantler, F. Boscherini and B. Bunker

International Tables for Crystallography (2022). Vol. I. Early view chapter
https://doi.org/10.1107/S1574870720003523

Semiconductors

Krystyna Lawniczak-Jablonskaa*

aInstitute of Physics, Polish Academy of Sciences, Lotnikow 32/46, 02-668 Warsaw, Poland
Correspondence e-mail: jablo@ifpan.edu.pl

In this chapter, examples of the most important properties of semiconductors (crystal structures, defects, dopants, energy gaps and their changes with alloying, nanostructures and magnetic properties) which can be obtained by applying X-ray absorption spectroscopy are presented with reference to the original papers. Owing to the limited space, the author's choice is not to consider this as a comprehensive overview, but as an example of the unique information that this technique offers.

Keywords: semiconductors.

1. Physical properties of semiconductors

Semiconductors are defined as crystalline or amorphous solids with variable conductivity. Their electrical properties depend on impurities, defects or intentionally introduced dopants, and are well described by band theory. The existence of an energy gap between valence and conduction bands is characteristic of semiconductors. In this respect, they are divided into wide-gap and narrow-gap semiconductors. Depending on the elemental composition, one can distinguish elemental (for example, silicon and germanium, the most popular) and compound (for example, GaAs, InSb or ZnO) semiconductors. By alloying elements or compounds, one can adjust many important material properties, including lattice parameters and band-gap energy. Therefore, binary and ternary alloys (for example, SiGe, InGaAs and GaAsP) have attracted much attention. Physical properties can be also changed by the dimensions of the material (quantum confinement effects), and nowadays many studies are devoted to semiconductor nanostructures (multilayers, quantum dots and nanowires). Scientific interest in a variety of semiconductor materials is driven by their increasing exploitation in electronic, optoelectronic and photonic devices. A comprehensive overview of the utilization of X-ray absorption techniques in studies of the structural properties of semiconductors is presented in Schnohr & Ridgway (2015[link]). Here, the most important results of these studies will briefly be described with reference to the original work.

2. Semiconductor alloys

X-ray diffraction (XRD) is a well established technique to study the crystal structures of materials and provides information on the average long-range crystallographic order in matter. In the case of well ordered materials, XRD is the best tool to obtain crystallographic data. X-ray absorption spectroscopy (XAS) offers information about the short-range atomic order (SRO) around a chosen element; therefore, it is the unique technique in the case when the short-range order differs from the long-range order. Such a case frequently occurs when alloying elemental and binary semiconductors. Many groups of materials are completely soluble when mixed with another group. The resulting alloy usually has the same crystal structure as the two parent compounds. The lattice parameter varies linearly with the composition in accordance with Vegard's law. Assuming that many material properties, in particular the band gap, directly depend on the SRO, experimental confirmation of how Vegard's law is realized on the atomic scale is very important. Here, the extended X-ray absorption fine-structure (EXAFS) technique finds unique application. Vegard's law can be satisfied according to two models. Firstly, all pairs of neighbouring atoms have the same bond length, proportional to the concentration x, and the bond angles in the crystal remain unchanged or, secondly, the bond length between given pairs of atoms does not change with concentration, but the lattice mismatch is accommodated by a change in the bond angles. Owing to its chemical sensitivity, EXAFS provides information about the specific distance of an element from its first and second neighbour atoms and its change with concentration in the case of binary and ternary alloys.

A good example of a binary alloy with 4% lattice mismatch is SiGe. This alloy crystallizes in the diamond structure and almost perfectly obeys Vegard's law. EXAFS studies of the SRO unambiguously proved that different pairs of atoms had different bond lengths (Aubry et al., 1999[link]). Nevertheless, on a change in composition, the Si—Si bond changes much less than the Ge—Ge bond. This indicates that the Si—Si bond is much stiffer and harder to stretch than the Ge—Ge bond. The lattice mismatch is accommodated mostly by bond bending. The distance between Si and Ge remains unchanged in all ranges of concentrations. Vegard's law is satisfied predominantly according to the second model.

EXAFS studies performed for many ternary alloys of compounds of elements from groups III and V and from groups II and VI confirmed the bimodal character of the bond length postulated in Mikkelsen & Boyce (1983[link]) as characteristic of the semiconductor alloy. This was proved for alloys crystallized in the zincblende and wurtzite structures. Nevertheless, clustering and phase segregation, strain and atomic ordering owing to the specific preparation conditions may influence the structural parameters of the alloy (see, for example, Woicik et al., 2012[link]). In general, the semiconductor alloys follow the Pauling and Huggins postulate that the atomic radii remain unchanged and the bond lengths for different atomic pairs are independent of composition x and identical to those of the parent materials (Pauling & Huggins, 1934[link]). In semiconductor alloys the change in the bond angles accommodates the lattice mismatch. However, detailed studies of many semiconductor alloys, mixed-cation as well as mixed-anion, showed that the Pauling limit is not strictly obeyed, but that a small linear change of the element-specific bond lengths with the composition x was usually observed in EXAFS studies (Boyce & Mikkelsen, 1987[link]; Letardi et al., 1987[link]). The original discovery of EXAFS that semiconductor bond lengths are nearly preserved, even in highly lattice-mismatched alloys, was proved by other techniques and has been well described by several theoretical models. The unique capabilities of the EXAFS technique allow the direct experimental determination of the element-specific first nearest-neighbour distance for many semiconductor alloys. The changes in bond angles can be estimated from the first and second nearest-neighbour distances (see, for example, Schnohr et al., 2008[link]). It was shown for InGaP that the anion–cation–anion (P—Ga—P and P—In—P) bond angles do not change much, but that all cation–anion–cation (In—P—In, In—P—Ga and Ga—P—Ga) bond angles deviate significantly from the crystallographic values and change linearly with the composition x.

Knowledge of the element-specific bond length enables specific atomic radii for a given element in a given type of matrix to be determined. The major factor determining the effective value of the atomic (or ionic) radius is the nature of the binding interaction (covalent, ionic, metallic or an intermediate type) and the local coordination geometry (for example tetrahedral or octahedral). An interesting example of estimation of the tetrahedral covalent (TC) radii of manganese, iron, cobalt and nickel in their alloys with ZnS was presented in Iwanowski et al. (1998[link]). The analysis of EXAFS data provided the bond lengths of both cation species (see, for zinc and iron, Fig. 1[link]a). The tetrahedral radius of zinc taken from other measurements was 1.225 Å and the advantage of a common anion enabled the TC radii of the transition metals to be derived, which differ significantly from the metallic radii (Fig. 1[link]b). TC radii exhibit a significant decrease with increased 3d shell filling. This indicated a diminishing contribution of the 3d electrons of the transition metal from manganese to nickel to the formation of the ionic covalent bond in the compounds studied.

[Figure 1]

Figure 1

(a) The bimodal interatomic distances in ZnFeS alloy. (b) Covalent tetrahedral radii of 3d metals in ZnS reported in Iwanowski et al. (1998[link]) (full triangles) and metallic radii (open circles).

Analysis of X-ray absorption near-edge structure (XANES) also provides unique information about the physical properties of semiconductor alloys, particularly with regard to the distribution of local density of states (DOS). Systematic studies of XANES have been presented in Lawniczak-Jablonska et al. (1996[link]) for alloys based on ZnS and ZnSe with a transition metal (manganese, iron, cobalt or zinc) as the admixed element. It was shown that the structure of the cation 4p DOS, reproduced by K-edge XANES, is mainly determined by the type of anion and does not depend greatly on the type of 3d metal. Moreover, XANES does not change with the alloy concentration x. Therefore, the changes of the cation distribution within the second coordination shell do not lead to pronounced differences in the spectra. The effective charge transfer for the 3d cation was estimated from measured chemical shifts of the respective K edges, and was within 2–2.5 e. Additionally, direct evidence for hybridization between the 3d metal states and the p states of sulfur was found, and an empirical correlation between the shape of the 3d metal K edges and the solubility limit of the metal in the investigated solid matrices was proposed.

3. Defects in semiconductors

A particular feature of semiconductors is the dependence of the conductivity on defects in the crystal structure. The existence of defects influences the DOS distribution and consequently many physical properties. The perfect tool to monitor these changes is XAS. Nice examples of such studies are presented for group III nitrides, SiC, diamond and other wide-gap semiconductors in Schnohr & Ridgway (2015[link]) (p. 49 and references therein). The proper choice of the X-ray absorption edge allows the DOS of the conduction band around atoms of interest to be selectively monitored. By comparing a so-called perfect crystal, with a very low content of defects, with a crystal with intentionally or unintentionally introduced defects, the source and kind of defects can be identified. Moreover, taking advantage of the linear polarization of synchrotron radiation, angle-dependence studies can examine the atomic order separately along the c axis and in the c plane. This is important for binary and ternary compounds with wurtzite or hexagonal crystal structures, where these bonds differ (Lawniczak-Jablonska et al., 1997[link]; Katsikini et al., 1998[link]). In the case of crystalline layers one can examine the influence of crystal-growth techniques and substrate types on the bond lengths and the bond structure, as well as estimate the kind of defects formed. Standard EXAFS studies (Lawniczak-Jablonska et al., 2001[link]) for GaN bulk crystals and several epilayers resulted in a direct estimate of the bond lengths in the c plane and along the c axis. The latter were found to be weaker and easier to break in all investigated samples. This may be the reason why most of the observed defects in GaN are located along the c axis. In addition, from XANES and EXAFS analysis it was shown that the N atoms in GaN layers are surrounded by Ga atoms, which are mostly located in the proper lattice sites. In contrast, the atomic order around Ga atoms, particularly along the c axis, suggests that N vacancies and other defects are present.

An example of this influence, that of the intentionally generated point defects in highly oriented pyrolytic boron nitride (HOPBN) on the DOS distribution of boron (B K edge), is presented in Caretti & Jiménez (2011[link]). The investigated HOPBN sample was subject to Ar+ bombardment (in an ultrahigh vacuum environment) and subsequent air exposure. The results indicate a significant amorphization owing to the impinging energetic Ar+ ions, which preferentially sputter N atoms to create vacant N sites and interstitials.

Owing to the fact that the energy positions of the absorption edge and the shape of the XANES spectrum are very characteristic of a given phase, it can be used as a `fingerprint' of the given phase. Moreover, in the case of a mixed-phase sample, the weighted average of the set of spectra of pure phases can be used to estimate the content of given phases (Katsikini et al., 1998[link]).

4. Dopants

The elemental selectivity of XAS is its main advantage in studies of dopants. The impact of dopants on the properties of semiconductors are responsible for their wide application, from electronics (transistors and diodes) to photonics and optoelectronics. The dopant concentrations that can be detected using XAS depend on many factors. The most important, which cannot be improved, are the atomic numbers of the dopant and the matrix. Heavy elements in a light matrix can be detected even at a concentration of 1015–1014 atoms cm−3. The other experimental factors (the brilliance of the photon source, the detector) are continuously under improvement and push this limit down.

In most of the reported studies the fluorescence detection mode was used, offering a good signal-to-noise ratio. The main disadvantage of XAS is that the signal from all dopant atoms of a given element is measured, but the dopants can create different clusters or occupy different sites in the crystal structures. To tackle this problem, structural simulations nowadays offer great help. A historical overview of the application of XAS to the location of dopants is presented in Schnohr & Ridgway (2015[link]) (p. 77), starting from the first report of the location of arsenic in an amorphous silicon matrix (Knights et al., 1977[link]). A nice example of solving the problem of the location of manganese implanted into crystalline silicon with the help of simulations is presented in Wolska et al. (2007[link]). The analysis of XANES and EXAFS spectra and the consideration of models of several possible manganese locations proved that manganese ions implanted into bulk silicon with a dose of 1016 atoms cm−3 and with an energy of 160 keV do not form metallic or oxide inclusions. Moreover, models assuming the location of manganese in a substitutional or interstitial position in the silicon lattice resulted in theoretical spectra that differed from the experimental spectra (Fig. 2[link]a). Both XANES and EXAFS spectra are in reasonable agreement with the model, which assumes the formation of clusters with an SRO close to strained and defect Mn–Si compounds with five to eight nearest-neighbour Si atoms (Fig. 2[link]b).

[Figure 2]

Figure 2

(a) The experimental spectrum (open squares) and simulation of Mn K-edge spectra for manganese location in silicon. MnSi is the substitutional position, Mnint are interstitial positions with different distances from the Si atom (R) and MnSi is the compound. (b) Simulation of Mn K-edge spectra for MnSi and a compound without the second shell of Mn atoms together with the experimental spectrum.

5. Energy gap, valence and conduction bands

A very characteristic feature of semiconductors is the existence of an energy gap between the valence and conduction bands. XAS and X-ray emission spectroscopy (XES) directly probe the partial DOS of the conduction band (CB) and valence band (VB), respectively. Overlapping the XES and XAS spectra with reference to the core level provides a direct measurement of the energy positions of the VB and CB states in semiconductor materials. It is particularly important for materials with a high level of structural disorder. In Yu et al. (2009[link]), GaN1−xAsx films were studied over the complete composition range. It was shown that films with a composition of 0.17 < x < 0.75 were amorphous, while those outside this range were crystalline (either single-crystalline or polycrystalline). The optical absorption results showed a continuously monotonic decrease in the band gap as the arsenic content increased. However, the absolute movement of the CB and VB of the GaN1−xAsx alloys cannot be derived from optical measurements. The effective use of soft XANES and XES showed that the reduction in the band gap can be primarily attributed to the downwards movement of the CB minimum for alloys with x > 0.2 and to the upwards movement of the VB maximum for alloys with x < 0.2. The unusual electronic structure and the capability to control the locations of the CB and VB edges offers the opportunity to use these alloys in novel solar-power conversion devices.

Combining resonant inelastic X-ray scattering (RIXS) and XANES, it is possible to estimate not only the semiconductor band-gap value but also its type. The pioneering work demonstrating this for CdO is presented in Demchenko et al. (2010[link]). The presented data set shows a progressively varying partial k mixing of initial and final states near the threshold and thus a varying incoherent line shape (Fig. 3[link]a). Overlapping of the XANES spectrum with RIXS makes it possible to estimate both direct (2.4 eV) and indirect (0.9 eV) band-gap values (Fig. 3[link]b) in CdO.

[Figure 3]

Figure 3

(a) Normalized to maxima and vertically offset RIXS spectra of a CdO film measured for excitation energies around the O K edge (533 eV) as indicated on the right. (b) Enlarged threshold region of oxygen 2p→1s XES and 1s→2p XAS spectra for a CdO film with the quantitative identification of direct and indirect band gaps. The excitation energies were 528.5 eV (thick black line), 534 eV (thick red line) and 553.5 eV (thick green line). The thin black and red lines correspond to the partially coherent fraction (standard): 528.5 and 534 eV, respectively (for details, see Demchenko et al., 2010[link]). The XES energy scale was aligned to the XAS energy scale using the elastic peak of the emission spectra (right).

Moreover, the features in the experimental spectra are well reproduced by calculations within the real-space multiple-scattering formalism (FEFF code) for a cluster of about six and ten coordination shells around the absorber for the L3 edge of Cd and the K edge of O.

6. Nanostructures

Low-dimensional semiconductor technology plays an important role in microelectronics and optoelectronics. The variation in dimensions opens a new degree of freedom in the manufacture of semiconductor devices, for example semiconductor lasers and light-emitting diodes. Knowledge of strain, chemical composition, interface quality and atomic ordering are of great importance to understand the growth mechanism, as well as the electronic and optical properties, of nanostructures. To be suitable for devices, the nano­structures are embedded in a superlattice and capping plays a crucial role in strain modification and atomic diffusion. Strain is closely related to the composition, shape and aspect ratio of the nanostructures. The nanostructures, the substrate and the matrix apply shear stress to each other. Methods exploiting the element sensitivity of XAS offer the unique possibility of disentangling the origin of the observed strain and estimating the elemental composition of each of the layers. The properties of nanostructures grown using different technologies are frequently studied by the methods of atomic force microscopy (AFM), high-resolution transmission electron microscopy, photo-luminescence and Raman scattering. However, the understanding and control of the morphology of the interface, particularly in buried layers where AFM cannot be applied, is still insufficient.

A major concept of the technology leading to the formation of self-organizing quantum dots (QDs) consists of the growth of several monolayers (MLs) of one semiconductor in a matrix or on top of another semiconductor with highly mismatched lattice parameters. The lattice mismatch between the substrate and the overgrown layer allows the formation of self-assembled QDs through the Stranski–Krastanov mechanism. Owing to the well elaborated growth technology of silicon and germanium crystals and their lattice mismatch (about 4%), particular attention was paid to the Ge/Si system and a few examples of unique information obtained by XAS for this system will be presented. Studies of III–V and I–VI nano­structures are reviewed in Schnohr & Ridgway (2015[link]) (p. 247, p. 269 and references therein). Recognizing the advantages of XAS as a local atomic order probe, XAS has been applied by many scientists to study the local arrangement of Ge atoms that are uncapped and buried in silicon, while also exploring the linear polarization of synchrotron radiation. The local structure inside the grown QDs is not easily accessible. Boscherini and coworkers (see, for example, Boscherini et al., 2000[link]) were the first to investigate the structure of epitaxial uncapped germanium QDs using XAS. In a series of papers, they reported the local structure of germanium QDs grown in Si(001) and Si(111) oriented substrates at temperatures above 400°C under different conditions. They found a pronounced Ge–Si intermixing. The intermixing was higher in the sample with only a wetting layer (up to 50%). The authors proposed that the silicon content in an island is limited both by the diffusion factor and by the fact that the lattice in the island is more relaxed than in the wetting layer, reducing the driving force for intermixing. Measurements were performed for both a parallel and a perpendicular orientation of the sample to linearly polarized synchrotron radiation, but no significant differences were found. Modifications of radial distributions of atoms in plane and out of plane for samples with different numbers of germanium monolayers have been reported in papers by Demchenko et al. (2004[link], 2007[link]) for the case of capped QDs. The QDs were formed in germanium monolayers (ML) grown with different thicknesses at very low temperature (210°C). The presence of a silicon cap on the germanium ML induces additional stresses and modifies the shape and composition of the formed structures. The formation of a monocrystalline germanium core inside the QDs was suggested, and some intermixing of Ge and Si atoms was found only at the surface of the germanium QDs. It was concluded that the reduction of the germanium layer growth temperature to 210°C limited silicon interdiffusion inside the QDs. Moreover, by increasing the thickness of the germanium layer, a partial relaxation of the lattice strain inside the QDs occurred. A difference in the relaxation process for Ge—Ge bonds was observed in plane and out of plane. Knowledge of the atomic concentration inside the QDs allows a prediction of the wavelengths corresponding to the emission from islands in the experimental photoluminescence spectra. XAS provides good input data for the calculation of the band-edge alignment in Si/GeSi/Si(001) structures with self-assembled islands. In addition to EXAFS, XANES spectra were also analyzed (Kolobov et al., 2002[link]) for uncapped and capped QDs and were fitted by a linear combination of reference bulk samples. The obtained information confirmed the EXAFS results.

In the case of nanowires, unique results have been achieved by exploiting XANES together with de-excitation spectroscopy (XES and X-ray excited optical luminescence; XEOL). Sham et al. (2004[link]) reported XEOL and XES studies of silicon nanowires (SiNWs) with excitations at the Si K and L3,2 edges, respectively. Both XEOL and XES show that the surface oxide plays a significant role in the electronic structure and optical properties of SiNWs. The observed chemical and morphology-dependent luminescence was attributable to emission from the encapsulating silicon oxide, quantum-confined silicon crystallites of various sizes embedded in the oxide layer and the silicon oxide interface. XES clearly shows the presence of a relatively thick oxide layer covering the SiNW and the DOS tailing across the Fermi level. This has significant implications for the electronic and optical properties of SiNWs.

The combination of XAS and XRD methods known as the diffraction anomalous fine-structure (DAFS) method offers additional unique possibilities for studying nanostructures. Namely, DAFS enables the determination of the valence state and local structure of a selected element at a specific crystalline site and/or phase in a nanostructure. A nice example is the determination of the SRO of a single atomic type in a sample of mixed amorphous and nanocrystalline phases of germanium (Frenkel et al., 2002[link]). EXAFS yields information about the SRO of all Ge atoms in the sample, while DAFS determines the SRO of only the ordered fraction. Therefore, DAFS can be used to disentangle the contributions of ordered and disordered states of the same element. It was shown that the first-shell distance distribution is bimodal, being shorter in the crystalline phase and longer in the rest of the germanium-rich phase. Moreover, DAFS spectroscopy allowed the separation of strain and composition and the detection of atomic ordering inside SiGe nano-islands, as was demonstrated by Richard et al. (2009[link]). The authors studied a series of Ge/Si(001) nano-island samples: pyramids and domes on nominal and pre-patterned surfaces. For free-standing domes, it was shown that the germanium content strongly depends on the growth conditions, with a tendency to increase from the bottom to the top of the nano-islands. In pyramidal islands, a mixing of ordered and random phases was observed. For small capped pyramids, the atomic order was closest to the ordered GeSi lattice with 50% germanium concentration. It was shown that DAFS spectroscopy is the only nondestructive method that allows the actual germanium content and the in-plane and out-of-plane strain to be recovered and the detection of atomic ordering. The great disadvantages of the DAFS method are complicated spectrometer adjustment, time-consuming data collection and a difficult data-analysis procedure. Recent progress in synchrotron science has encouraged the development of a more efficient way to collect good-quality DAFS spectra in a reasonable time (Kawaguchi et al., 2014[link]).

7. Diluted magnetic semiconductors

XAS, particularly when used with circularly polarized radiation, offers unique information in studies of diluted magnetic semiconductors (DMSs), as presented, for example, in Schnohr & Ridgway (2015[link]) (p. 313). DMSs are a class of semiconductors with an immersed magnetic element to introduce magnetic properties into the semiconductor. The interest in DMS is stimulated by the expectation that manipulation of the electron spin, as an alternative to manipulation of the electron charge, can be used for the storage of information in semiconducting devices. A very important issue is the location of the magnetic ions in the semiconductor matrix, as it is correlated with all important physical properties. It has already been proved that EXAFS is the perfect tool to solve this issue. The failure to produce DMSs which are ferro­magnetic at room temperature (RT) resulted in an increasing interest in the study of granular materials with RT ferromagnetic properties. XAS, which is element-selective and sensitive to local atomic order, is also a very useful tool for detecting nanoclusters with different structures and compositions. Moreover, in the case of an element with a magnetic moment and due to the existence of dichroism, applying circularly polarized radiation in XAS studies offers the possibility of estimating the orbital and spin moment of the magnetic element.

The observed dichroism is the consequence of spin conservation in dipole transition, which means that a spin-up electron can only be promoted to a spin-up empty state and vice versa. Since left- and right-polarized radiation only interacts with electrons with spin up or down, respectively, in the case of an element with a difference in the occupancy of the spin-up and spin-down empty states differences in the spectra will be observed.

X-ray magnetic circular dichroism (XMCD) is defined as the difference in the absorption cross section between left and right circularly polarized X-rays (Schütz et al., 1987[link]). XMCD has developed into a widely used technique for the element-specific characterization of magnetic materials. Quantitative evaluation of magnetic moments, separated into spin and orbital contributions, is possible from integrals over XMCD spectra by applying the so-called sum rules (Thole et al., 1992[link]; Carra et al., 1993[link]). Owing to the requirement for a high magnetic field to introduce magnetic order in DMS, it is not frequently used for studies of magnetic ions in semiconductors. Nevertheless, several studies have been reported for DMS, particularly for GaMnAs. The spectroscopic features of Mn-XMCD appeared to be very sensitive to changes in the environment around manganese (Wu, 2005[link]; Schnohr & Ridgway, 2015[link], p. 332). This promises the potential to use XMCD for the unambiguous determination of impurity distribution in DMS through the strong interplay between experiment and theory.

Detailed studies of samples with cubic GaMnAs and hexagonal MnAs inclusions have been reported by Lawniczak-Jablonska and coworkers (Schnohr & Ridgway, 2015[link], p. 334). The orbital and spin magnetic moments located on Mn atoms were examined as a function of the magnetic field. Particular interest was devoted to changes in magnetic moments on the formation of nanoclusters. Samples with homogeneously distributed Mn atoms and cubic and hexagonal nanoclusters were investigated. A pronounced increase in the orbital moment and the saturation of the spin moment were observed on the formation of nanoclusters. In samples without nanoclusters the spin moment increases linearly with the magnetic field, reaching 0.24 μB at 5 T. The possibility of changing the direction of the easy and hard axis of MnSb inclusions by a proper choice of the crystallographic direction of the substrate has also been demonstrated by XMCD studies (Lawniczak-Jablonska et al., 2011[link]).

8. Summary

XAS is a unique technique to study the physical properties of semiconductors. The most important attributes of this technique are the element selectivity and the sensitivity to local atomic order. This allows information to be obtained about dopant positions, the kinds of defects in the crystal structure and the bond lengths, as well as site occupation and atomic radii in given types of bond. Exploiting the polarization of synchrotron radiation, the crystal anisotropy as well as orbital and spin moment of the element under study can be estimated; moreover, selection rules for X-ray transitions offer information about the partial density of state distributions in the conduction and valence bands, as well as the energy gap. The support of advanced theoretical methods to simulate the XAS spectra of different models of atomic order has increased the range of application and the reliability of the method. Recent progress in synchrotron source construction has resulted in a relatively high intensity of the beam focused down to 10 nm. This opens the way for XAS studies of nano-objects (nanoXAS) and XAS microscopy. The increase in the beam intensity also allowed effective time-resolved studies exploiting the time structure of the synchrotron beam and synchronization with the optical laser which starts the observed processes. The future development of the XAS technique towards spatially resolved and time-resolved studies provides access to nano-objects and the dynamics of chemical reactions.

References

First citationAubry, J. C., Tyliszczak, T., Hitchcock, A. P., Baribeau, J. M. & Jackman, T. E. (1999). Phys. Rev. B, 59, 12872–12883.Google Scholar
First citationBoscherini, F., Capellini, G., Di Gaspare, L., Rosei, F., Motta, N. & Mobilio, S. (2000). Appl. Phys. Lett. 76, 682–684.Google Scholar
First citationBoyce, J. B. & Mikkelsen, J. C. (1987). In Ternary and Multinary Compounds: Proceedings of the 7th International Conference, edited by S. K. Deb & A. Zunger. Pittsburgh: Materials Research Society.Google Scholar
First citationCaretti, I. & Jiménez, I. (2011). J. Appl. Phys. 110, 023511.Google Scholar
First citationCarra, P., Thole, B. T., Altarelli, M. & Wang, X. (1993). Phys. Rev. Lett. 70, 694–697.Google Scholar
First citationDemchenko, I. N., Denlinger, J. D., Chernyshova, M., Yu, K. M., Speaks, D. T., Olalde-Velasco, P., Hemmers, O., Walukiewicz, W., Derkachova, A. & Lawniczak-Jablonska, K. (2010). Phys. Rev. B, 82, 075107.Google Scholar
First citationDemchenko, I. N., Lawniczak-Jablonska, K., Kret, S., Novikov, A. V., Laval, J., Zak, M., Szczepanska, A., Yablonskiy, A. N. & Krasilnik, Z. F. (2007). Nanotechnology, 18, 115711.Google Scholar
First citationDemchenko, I. N., Lawniczak-Jablonska, K., Piskorska, E., Zhuravlev, K. S., Nikiforov, A. I. & Welter, E. (2004). J. Alloys Compd. 382, 206–210.Google Scholar
First citationFrenkel, A. I., Kolobov, A. V., Robinson, I. K., Cross, J. O., Maeda, Y. & Bouldin, C. E. (2002). Phys. Rev. Lett. 89, 285503.Google Scholar
First citationIwanowski, R. J., Lawniczak-Jabłońska, K., Golacki, Z. & Traverse, A. (1998). Chem. Phys. Lett. 283, 313–318.Google Scholar
First citationKatsikini, M., Paloura, E. C. & Moustakas, T. D. (1998). J. Appl. Phys. 83, 1437–1445.Google Scholar
First citationKawaguchi, T., Fukuda, K., Tokuda, K., Shimada, K., Ichitsubo, T., Oishi, M., Mizuki, J. & Matsubara, E. (2014). J. Synchrotron Rad. 21, 1247–1251.Google Scholar
First citationKnights, J. C., Hayes, T. M. & Mikkelsen, J. C. Jr (1977). Phys. Rev. Lett. 39, 712–715.Google Scholar
First citationKolobov, A. V., Oyanagi, H., Wei, S., Brunner, K., Abstreiter, T. & Tanaka, K. (2002). Phys. Rev. B, 66, 075319.Google Scholar
First citationLawniczak-Jablonska, K., Iwanowski, R. J., Demchenko, I. V., Boettcher, T., Einfeldt, S., Hommel, D., Cortes, R. & Perera, R. C. C. (2001). J. Alloys Compd. 328, 77–83.Google Scholar
First citationLawniczak-Jablonska, K., Iwanowski, R. J., Golacki, Z., Traverse, A., Pizzini, S., Fontaine, A., Winter, I. & Hormes, J. (1996). Phys. Rev. B, 53, 1119–1128.Google Scholar
First citationLawniczak-Jablonska, K., Suski, T., Liliental-Weber, Z., Gullikson, E. M., Underwood, J. H., Perera, R. C. C. & Drummond, T. J. (1997). Appl. Phys. Lett. 70, 2711–2713.Google Scholar
First citationLawniczak-Jablonska, K., Wolska, A., Klepka, M. T., Kret, S., Gosk, J., Twardowski, A., Wasik, D., Kwiatkowski, A., Kurowska, B., Kowalski, B. J. & Sadowski, J. (2011). J. Appl. Phys. 109, 074308.Google Scholar
First citationLetardi, P., Motta, N. & Balzarotti, A. (1987). J. Phys. C Solid State Phys. 20, 2853–2884.Google Scholar
First citationMikkelsen, J. C. & Boyce, J. B. (1983). Phys. Rev. B, 28, 7130–7140.Google Scholar
First citationPauling, L. & Huggins, M. L. (1934). Z. Kristallogr. 87, 205.Google Scholar
First citationRichard, M.-I., Katcho, N. A., Proietti, M. G., Renevier, H., Favre-Nicolin, V., Zhong, Z., Chen, G., Stoffel, M., Schmidt, O., Renaud, G., Schülli, T. U. & Bauer, G. (2009). Eur. Phys. J. Spec. Top. 167, 3–10.Google Scholar
First citationSchnohr, C. S., Araujo, L. L., Kluth, P., Sprouster, D. J., Foran, G. J. & Ridgway, M. C. (2008). Phys. Rev. B, 78, 115201.Google Scholar
First citationSchnohr, C. S. & Ridgway, M. C. (2015). Editors. X-ray Absorption Spectroscopy of Semiconductors. Heidelberg: Springer.Google Scholar
First citationSchütz, G., Wagner, W., Wilhelm, W., Kienle, P., Zeller, R., Frahm, R. & Materlik, G. (1987). Phys. Rev. Lett. 58, 737–740.Google Scholar
First citationSham, T. K., Naftel, S. J., Kim, P. G., Sammynaiken, R., Tang, Y. H., Coulthard, I., Moewes, A., Freeland, J. W., Hu, Y.-F. & Lee, S. T. (2004). Phys. Rev. B, 70, 045313.Google Scholar
First citationThole, B. T., Carra, P., Sette, F. & van der Laan, G. (1992). Phys. Rev. Lett. 68, 1943–1946.Google Scholar
First citationWoicik, J. C., Ludwig, J. K. F. & Moustakas, T. D. (2012). Appl. Phys. Lett. 100, 162105.Google Scholar
First citationWolska, A., Lawniczak-Jablonska, K., Klepka, M., Walczak, M. S. & Misiuk, A. (2007). Phys. Rev. B, 75, 113201.Google Scholar
First citationWu, R. (2005). Phys. Rev. Lett. 94, 207201.Google Scholar
First citationYu, K. M., Novikov, S. V., Broesler, R., Demchenko, I. N., Denlinger, J. D., Liliental-Weber, Z., Luckert, F., Martin, R. W., Walukiewicz, W. & Foxon, C. T. (2009). J. Appl. Phys. 106, 103709.Google Scholar








































to end of page
to top of page