RICERCA ITALIANA     

Growth and properties of semiconducting-oxide based quasi one-dimensional nanocrystals

Università degli Studi di Lecce

Abstract

The present proposal concerns a base-research project whose main activities and objectives aim at: (i) developing new synthesis/nano-fabrication techniques for the self-assembly growth of quasi one-dimensional (1D) nanocrystals of high-gap metal oxide (MOX) semiconductors; (ii) understanding their fundamental growth mechanisms; and (iii) studying the effects of low dimensionality and gas-surface interactions on their electronic, vibrational and optical properties. In particular the project intend to investigate and develop self-assembly methods of crystallographic-oriented (epitaxial) and random (non epitaxial) free-standing quasi 1D nanocrystals based on chemical vapour deposition (CVD) processes or on the innovative combination of CVD methods and advanced nano-fabrication tools, like focused ion beam nano-fabrication. Both metal catalyst assisted growth, the so-called vapour-liquid-solid mechanism, and catalyst-free vapour crystallisation will be employed. Besides the growth/nanofabrication activity, the research will address for the first time thermo/fluid-dynamic and atomistic modelling of quasi 1D nanocrystal self-assembly. The potentialities of the proposed self-assembly growth/nano-fabrication technologies will be specifically focused on the realization of both epitaxial 1D nanocrystals (nanorods), 1D hetero-structures and ordered nanorod arrays, as well as of large ensembles of long non-epitaxial 1D systems (nanowires and nanobelts) deposited over wide substrate areas. MOX semiconductors to be investigated include ZnO, MgO, SnO2, and TiO2 compounds and ZnO/MgO, ZnO/ZnMgO quasi 1D heterostructures. Long-term applications of these systems comprise UV opto-electronics, photonics and innovative optical gas sensing nano-devices. Within the project, we intend to undertake a systematic investigation of the physical properties of as-grown nanorod and nanowire/nanobelt structures, with special attention to the effect of reduced dimensionality and quantum confinement on their electronic/phononic states and radiative excitation-recombination processes. This study will be undertaken using a combination of optical spectroscopic tools and theoretical calculations. Besides providing experimental data on fundamental physical properties of MOX-based quasi 1D nanocrystals, the characterization of morphological, structural and optical properties of as-grown quasi 1D nanocrystal structures will provide the ultimate feedback to assess the reliability of the proposed growth/fabrication technology. Special emphasis will be put on the study of the optical properties and radiative emissions of MOX-based 1D nanocrystals and structures under controlled gaseous environments. <<<

Principal Investigator

Nicola LOVERGINE Università degli Studi di LECCE

Research Objectives

The project aims at: (i) developing new synthesis/nano-fabrication techniques for the self-assembly growth of quasi one-dimensional (1D) nanocrystals of high-gap metal oxide (MOX) semiconductors; (ii) understanding their growth mechanisms; and (iii) studying the effects of low dimensionality on their electronic, vibrational and optical properties. The project wants to investigate and develop self-assembly of both crystallographic-oriented (epitaxial) and random (non epitaxial) free-standing quasi 1D nanocrystals based on chemical vapour deposition (CVD) processes or on the combination of CVD and nano-fabrication tools. The potentialities of the proposed technologies will be focused on the realization of both epitaxial 1D nanocrystals (nanorods), 1D hetero-structures and ordered arrays, as well as of large ensembles of long non-epitaxial 1D systems (nanowires and nanobelts) uniformly deposited over wide substrate areas. Long-term applications comprise UV opto-electronics, photonics and novel optical gas sensors. We intend to investigate the nanorod and nanowire/nanobelt properties, with special attention to the effect of reduced dimensionality and quantum confinement on electronic and phononic states. Besides providing experimental data on the physical properties of MOX-based quasi 1D nanocrystals, the characterization of morphological, structural and optical properties of as-grown quasi 1D nanocrystal structures will provide the ultimate feedback to assess the reliability of the proposed growth/fabrication technology. The project objectives can be detailed as follows:

Objective 1: development of self-assembly growth/nano-fabrication methods of quasi 1D nanocrystals, heterostructures and arrays

We will study and optimise CVD methods for self-assembly of quasi 1D nanocrystals and structures. Two different types of free-standing 1D structures will be initially grown: (i) epitaxial nanorods of ZnO, MgO, ZnMgO, and 1D heterostructures, such as ZnO/MgO or ZnO/ZnMgO; (ii) randomly oriented nanowires and nanobelts of ZnO, SnO2, and similar high gap metal oxides. Doping of quasi 1D nanocrystals will be also performed by n-type dopant species. There are two ways to achieve the self-assembly of quasi 1D nanocrystals: metal catalyst assisted growth, through the so-called vapour-liquid-solid (VLS) mechanism, or through catalyst-free vapour crystallisation, called vapour-solid (VS) mechanism. Both methods will be employed and implemented into the CVD growth processes of high gap metal oxides. To compare with literature, growth of quasi 1D nanocrystals by chemical vapour transport (CVT) methods will be briefly investigated; it should be however, pointed out that despite CVT methods are simpler and avoid the chemistry complications of CVD they are not ideal in terms of flexibility and process reliability. A set of suitable precursors and alternative CVD reaction pathways will be thus identified and tested. CVD conditions for optimal physical properties of as-grown 1D structures will be determined. A qualifying objective to be achieved will be the ability to growth epitaxial nanorods in a precise and periodic fashion (nanorod arrays) for fabrication of two-dimensional photonic crystal waveguides: this requires a control over the position of the metal catalyst on the substrate surface before growth. The innovative aspect of the project is that we intend to achieve such catalyst position control by employing a focused ion beam (FIB) facility which will be installed at Unit n. 2. To the best of our knowledge this combination of nano-fabrication and growth methods has not yet been reported in the literature and promises to be a critically enabling technology for the most demanding applications of future nanorod-based devices.

Objective 2: development of thermo/fluid-dynamic and atomistic modelling of quasi 1D nanocrystals self-assembly

Besides the growth/nanofabrication activity, the research will address thermo/fluid-dynamic and atomistic modelling of the self-assembly growth. No such attempt has been reported so far. We will establish phenomenological growth models based on thermodynamics and fluid-dynamics laws, predicting vapour composition and supersaturation, as well as mass transport rates in CVD reactors. The development of thermo/fluid-dynamic models will be based on data obtained via in situ mass spectrometry study of the vapour phase. The possibility of investigating growth stability conditions, in terms of constitutional supercooling, will be also explored.
Atomistic models will be considered to investigate quasi 1D nucleation and growth mechanisms. These models will be applied to both randomly oriented nanowires/nanobelts and epitaxial nanorods. In the latter case, the presence of metal catalysts during the VLS growth will be considered. Of high scientific relevance will be the possibility of correlating the nucleation/growth conditions foreseen by an atomistic model with experimental results through morphological analyses. The role of surface energy in controlling the shape of nanocrystals will be investigated, assuming that the morphology should approximate the equilibrium shapes as predicted by Wulff's theorem. In this respect, a correlation will be sought after between what experimentally observed and what predicted by the Hartmann-Perdok theory and, more in general, by the theories of crystal stability.

Objective 3: determination of the electronic, vibrational and optical properties of quasi 1D nanocrystals, heterostructures and arrays.

An important activity of the program will be the study of the electronic, vibrational and optical properties of quasi 1D nanocrystals and the effect of reduced lateral dimensions and quantum confinement. Polarisation sensitive optical techniques will be utilized in order to extract the anisotropic dielectric response of the nanorods and evaluated with respect to electronic transitions. The electronic states will be also studied by conventional spectroscopic tools like PL, PLE and ABS, whilst single or few nanorod/nanobelts emission will be studied by micro-CL. This will allow us to determine the electronic energy levels in as-grown nanorods/nanobelts and compare them with theoretical calculations.
Similar to electronic states, the vibrational properties are changed by the nanocrystal reduced dimensions. This effect will be studied by micro-Raman and resonant Raman scattering. A second check-up on reduced dimensionality effects will thereby be obtained on the interatomic force constants.
Non-linear optical properties and relaxation dynamics of photoexcited carriers in as-grown 1D nanocrystals and heterostructures will be also studied. To this purpose, we will perform ultra-fast spectroscopy experiments. Besides determining the relaxation times of photo-excited carriers, this will allow us to measure the nanorod optical gain and spectral band, of great interest for UV lasing applications. Finally, measurements of the angular dispersion of light through period arrays of nanorods will allow to study their photon confinement properties.

Objective n. 4: characterisation of optical and radiative properties of quasi 1D nanocrystals and heterostructures under controlled gaseous environments.

Radiative and non-radiative recombination channels will be studied to determine activation energies of surface defects of importance for optical gas sensing applications. The occurrence of photo-absorption bands will be studied, along with nanorod/nanobelt luminescence changes produced by gas-surface interactions under controlled environmental conditions. This optical effects will be studied as function of nanorods/nanobelts size, shape, crystalline faces exposed to the surrounding gaseous species, surface oxygen vacancies and material pre-treatment under oxidizing or reducing atmosphere. Beside we will identify the active poisoning species and the optimal working temperature for the different MOX quasi 1D nanocrystals. <<<

First Results

The list of intermediate results expected at completion of FASE I is the following:

- Functional validation of a well defined set of precursors for the CVD growth of high quality nanorods and nanobelts of high gap MOX semiconductors
- Definition of substrate treatment procedures and metal catalyst preparation specifications before growth
- Definition of CVD growth parameter values (temperatures, gas flows, pressures, etc), growth procedures for the VLSE growth of highly oriented ZnO, MgO and ZnMgO epitaxial nanorods
- Definition of CVD growth parameter values (temperatures, gas flows, pressures, etc), growth procedures for the growth of high quality ZnO, and SnO2 nanowires/nanorods on large substrate areas
- Complete quantitative description of CVD reactors thermo/fluid-dynamic conditions and cross-checking of predictions with experimental results
- Quantitative morphological and structural definitions of nanorod/nanobelt as function of growth conditions and comparison with 1D nucleation description in terms of atomistic models
- Completion of the experimental set up for resonant Raman scattering measurements
- Completion of the experimental set up for visible PL characterisation of high gap MOX
- Demonstration of quantum confined effects on the electronic and vibrational properties of nanorods/nanowires structures and comparison between experimental energy levels and theoretical calculations
- Comprehension of how morphological and structural properties of nanorods and nanobelts influence their visible PL emission
- Experimental database of PL response to each target species NO2, CO and NH3 of ZnO, MgO nanorods and SnO2, ZnO nanobelts as a function of working temperature and concentrationsThe list of intermediate results expected at completion of FASE I is the following:

- Definition of substrate treatment procedures and metal catalyst nanocluster by FIB nano-fabrication.
- Definition of CVD growth parameter (temperatures, gas flows, pressures, etc) and procedures for the VLSE growth of ZnO/MgO and ZnO/ZnMgO 1D epitaxial heterostructures
- Definition of CVD growth parameter (temperatures, gas flows, pressures, etc) and procedures for the growth of high quality TiO2, and In2O3 nanowires/nanobelts on large substrate areas
- Growth of high quality 1D heterostructures and ordered arrays of ZnO, MgO nanorods with different type of periodic arrangements of interest of 2D-PCWs
- Quantitative morphological and structural characterisation of nanorod arrays and 1D heterostructures, as well as TiO2 and In2O3 nanowires/nanobelts as a function of growth conditions and comparison with 1D nucleation description in terms of refined atomistic models
- Demonstration of quantum confined effects on the electronic and vibrational properties of 1D heterostructures (0D systems) and comparison between experimental energy levels and theoretical calculations
- Measurements of optical gain and carrier dynamics in ZnO, MgO nanorods and 1D heterostructures
- Measurements of light confinement and dispersion laws in ordered arrays of epitaxial nanorods
- Comprehension of how morphological and structural properties of 1D heterostructures nanorods and TiO2 and In2O3 nanobelts influence their visible PL emission
- Completion of an experimental database of PL responses to each target species NO2, CO and NH3 of SnO2, ZnO nanobelts as a function of working temperature and concentrations <<<

Timescale

24 months

National and international background

Quasi one-dimensional (1D) semiconductor nanocrystals represent the forefront of today's solid state physics and technology. These systems, having two of their dimensions comparable to the wavelengths of the electronic or phononic wavefunctions, are expected to show a variety of quantum confinement effects, such as density of state singularities, molecular-like states extending over large distances, high luminescence efficiencies and lower lasing threshold. Also, their high surface-to-volume ratio allows to exploit the role of surface states (and their ambient-driven changes) in determining carrier optical excitation/recombination phenomena.
Despite these attractive properties, until very recently, not too many studies were performed on 1D semiconductor systems, whilst two-dimensional structures (quantum wells) have been under study already for more than two decades and quasi 0-dimensional (0D) objects (quantum dots) have been in the focus of researchers for nearly a decade. The main reason for such discrepancy resides in the difficulty of fabricating quasi 1D nanocrystals. In the recent past these structures have been generally grown on and stabilised by the surface of a 3Dimensional (3D) substrate, with their major dimension running in the surface plane [1], with the consequence, however, that interactions with the 3D substrate can be quite strong and may dominate the 1D effects. Free standing quasi 1D nanocrystals having negligible interaction with the substrate can be now fabricated by a variety of self-assembly methods, and thus there is the unique chance to observe quantum effects in these structures and to use them for applications to novel types of opto-electronic or photonic nano-devices and nano-sensors.
Free standing quasi 1D nanocrystals consist of cylindrical- or prismatic-shaped crystals, with lateral size ranging within 1-100 nm and length from a few hundreds nanometer up to several micron (so called nanorods) or even millimeter (nanowires and nanobelts/nanoribbons). In general, two types of free standing quasi 1D nanocrystals are considered: (a) crystallographic-oriented (i.e. epitaxial) quasi 1D-nanocrystals laying with their major dimension normal to the surface of single crystalline substrates, so-called epitaxial nanorod (or nanopillar) structures, and (b) non-epitaxial (i.e., randomly oriented) ensembles of long nanowires or nanobelts.
The synthesis/fabrication of semiconductor epitaxial nanorods in highly oriented and ordered arrangements represents a major requirement of future opto-electronic and photonic applications. A distinctive advantage of an epitaxial nanorod structure is the possibility to change the material composition and/or intentional doping along the nanorod length (i.e. growth direction): quasi 1D heterostructures (such as superlattices and quantum dots) and p-n junctions could be thus fabricated, leading to new nanoscale optoelectronic devices. The ability to insert an optically active quantum dot into a 1D nanorod semiconductor structure is nowadays considered the ideal way to fabricate single-photon quantum devices [2]. A second emerging field of potential applications for epitaxial nanorods is the fabrication of two-dimensional photonic crystal waveguides (2D-PCWs). Photonic crystals have recently attracted a great deal of interest due to their ability to strongly influence the dispersion of light. Under favourable circumstances, such structures possess a photonic band gap, that is a frequency range where electromagnetic-wave propagation is forbidden in the crystal, independently of the light propagation direction [3,4]. Recent theoretical work [5] has shown that regular rows or periodic (i.e., rectangular or honeycomb-like) arrangements of parallel dielectric nanorods in air behave as a 2D-PCW, due to the light total internal reflection, whilst the guided mode dispersion is affected by the nanorod periodicity. Any ‘defect' breaking the photonic crystal periodicity produces photonic states localised at the defects [6], allowing the application of the nanorod-based 2D-PCWs as resonators for nano-lasers [5]. Also, a sharp transition between waveguided modes and the localised photonic states could be induced by small changes of the material dielectric constant at the defect site, leading to the realisation of fast photon flux switches [6].
Randomly oriented nanowires/nanobelts become instead important when nanocrystal high surface-to-volume ratios are required over large areas or high volume fill factors, as in gas sensing applications. Also, long nanowires/nanobelts could be used in assembling quasi 1D circuits by atomic force microscope manipulation or microfluidic methods [7].
Owing to these attractive assets a large scientific and technological interest has recently arose on 1D nanocrystals made of high gap semiconducting metal oxides (MOX). Zinc oxide (ZnO) is the most representative among these materials. It presents fundamental advantages over other semiconductor compounds: (i) a RT band gap of 3.37 eV; (ii) a unique combination of piezoelectric, conducting and optical properties; (iii) the largest exciton binding energy of all II-VI and III-V semiconductors, allowing excitonic stimulated emission up to 550 K; and (iv) a large bond energy of 1.89 eV, leading to good thermal and chemical stability. Furthermore, the ternary system CdO-ZnO-MgO covers a large band gap range (MgO has an energy gap of 4 eV), with a smaller variation of lattice parameters than in III-group nitrides. Other high gap MOX of technological interest include tin oxide (SnO2), titanium oxide (TiO2), gallium oxide (GaO) and indium oxide (In2O3) for chemical sensing applications.
The combination of the unique properties of high gap MOX with the low dimensional properties of free standing quasi 1D nanocrystals could result into a new class of devices, such as UV emitting nanolasers, field emission nanodevices and optical gas sensors.
Indeed, RT lasing action at 385 nm has been reported for ZnO epitaxial nanorods [8]. 1D heterostructures based on epitaxial ZnO/MgO and ZnO/ZnMgO nanorods are now considered potential candidates to realise 0D systems (quantum dots in nanorods)) for opto-electronic nanodevices emitting in the UV spectral range. The fabrication of 2D-PCWs based on ZnO or MgO epitaxial nanorods for realization of fast photon switches operating in the UV range has not been attempted yet, but efforts in that direction are already reported [14]. The use of ZnO nanorods as field emission tips have been also reported [9]. The sensitivity of ZnO optical recombination to surface states could benefit of the nanocrystal high surface-to-volume ratio. Surface states have been identified as a possible cause of the visible emission in ZnO nanowires [10]. A very interesting result is the progressive increase of the green light emission intensity as the ZnO nanorod/nanowire diameter decreases [8]. This suggests that the reduced dimensionality of the system could play a role in the onset of visible luminescence spectra of high gap metal oxides, both due to quantum size effects or by increasing the nanowire surface-to-volume ratio. It has been also shown that the luminescence of SnO2 nanobelts changes in a reproducible and reversible way when exposed to traces of NO2. Besides, the response seems highly selective towards humidity and other polluting species like CO and NH3. Ionosorbed gaseous species that create surface states can quench the material luminescence by destroying radiative recombination paths. Finally, 2D-PCWs made of epitaxial ZnO or other high gap MOX nanorods could give rise to an new class of optical gas sensors based on the gas-driven changes of the material dielectric constant.
Epitaxial nanorods can in principle be obtained by a variety of nano-technology methods. A typical top-down approach to the fabrication of nano-scale structures would utilize nano-lithographic techniques, such as electron beam lithography and proximal probe patterning in combination with dry-etching methods to remove excess material from planar structures. However, top-down methods do not warrant a sufficiently low dimensionality, the resolution limit of nano-lithographic techniques being around 100 nm, i.e. well above the size below which quantum-confinement effects appear. Moreover, the resulting nano-crystals may suffer from process damages and surface roughness that could seriously deteriorate their optical properties. In the case of long free standing nanowires or nanobelts, top-down methods are even more difficult to use.
Bottom-up techniques, achieving quasi 1D nanocrystals by self-assembly growth mechanisms have greater potentials in terms of crystalline perfection, cost and high productivity. One of the most promising self-assembly method is the so-called metal catalyst assisted vapour-liquid-solid (VLS) mechanism. Self-assembly of Si and Ge nanowires has been demonstrated by Lieber in 1998 [11,12] using Fe and Au nanoclusters as catalysts. VLS epitaxy (VLSE) has been reported since then for III-V compounds free-standing nanorod grown by metalorganic vapour phase epitaxy (MOVPE) and molecular beam epitaxy (MBE) [2]. The feasibility of VLSE for the growth of ZnO nanorods has been demonstrated in 2001 by Yang and co-workers [8] using a chemical vapour transport (CVT) process and Au as catalyst. Alternative catalysts, like Cu [13], have been proposed, but no clear data about their advantages/disadvantages exist. The possibility of selecting specific areas of the substrate where the nanorod VLSE nucleation and growth occur has been demonstrated using Au-patterned substrates [8,14]; no control of the single catalyst nanocluster position and size has been reported instead and much remains to be done on this topic. Reports on the VLSE growth of other semiconducting oxides, like MgO, CdO or the ternary alloy ZnMgO, are still lacking in the literature, as for 1D heterostructures like ZnO/MgO and ZnO/ZnMgO. In this respect a great amount of work both on the growth/fabrication and the characterisation of ZnO-based nanorod structures is still needed.
Quasi 1D nanocrystals can be obtained also by catalyst-free self-assembly growth processes. Noteworthy, the growth of well-aligned ZnO nanorods has been also achieved by pyrolytic MOCVD [15] without the use of a metal catalyst and attributed to a vapour-solid (VS) growth mechanism, although its exact details are still not clear. Similarly, the growth of long (up to mm) non-epitaxial nanobelts of ZnO, CdO, SnO2, and In2O3 was demonstrated by thermal evaporation under controlled conditions without the presence of a metal catalyst [16]. However, whilst catalyst-free methods are interesting for studying the basic physical properties of quasi 1D nanocrystals, they do not allow control of the nanocrystal nucleation position and density; most device applications (i.e., device integration and PCWs) require instead site-specific nucleation, as well as a process that remains site-specific with time.
Both VLS and VS growth mechanisms are being applied to growth epitaxial nanorods and/or randomly oriented nanowires/nanobelts of high gap MOX, using different deposition methods, such as MBE [17], pulsed laser deposition [18], electro-deposition [19], and both CVT [8,16] and chemical vapour deposition (CVD) methods [20-25]. Whilst CVT methods are straightforward to use, the CVD ones are attractive in reason of their high versatility, higher reproducibility, relatively low costs and ability to growth on large area substrates. It should be noticed that the relevant literature is still very lacking, particularly as to the details of nanorod/nanowire initial nucleation stages and 1D growth. Also, no systematic investigation has been yet reported, to identify a range of suited precursors for the growth of high gap MOX nanorods/nanobelts by CVD methods, nor clear operative ranges of the main parameters (temperature, pressure, gas flow rates etc.) and of their influence on the nanocrystal physical and chemical properties have been established.
A particularly critical aspect of quasi 1D nanocrystals is their shape and dimensions, as wavefunctions and energy levels of both 1D-confined carriers and phonons depend on the crystal geometry and size. ZnO epitaxial nanorods have evidenced the occurrence of both cylindrical- or hexagonal-shaped nanocrystals, whilst CdO and SnO2 nanobelts have shown rectangular-shaped sections and curled appearance; however, no clear correlation of the crystal shape, facet crystallographic planes and dimensions with exact growth conditions has been reported.
The study of electronic, vibrational and carrier optical excitation/recombination properties of self-assembled nanorods and nanowires/nanobelts of high gap MOX nanocrystals is of critical importance to evidence/understand the effect of low dimensionality.
Several research groups have investigated the photoluminescence of ZnO nanorods [8, 26]. Strong near band-edge and deeper green band emissions are usually observed at RT; whilst UV emission can be attributed to both free and bound excitons or free-to-bound transitions [26], green emission bands are instead ascribed to a transition involving near-surface defect states [27], like oxygen vacancies [28]. The observation that these bands become intense with decreasing the size of the nanorods [27] is a clear indication of the high surface-to-volume ratio of these structures. However, quantum confinement effects on luminescence is not observed for nanorods with diameters above 100 nm, posing an upper limit on the nanorod size. Recent time-resolved measurements have further clarified the extent of this conclusion showing luminescence decay rates that decreases with increasing nanorod length as a result of exciton-polariton excitation [29]. Raman scattering has also been applied on ZnO-nanorods [30,31]. There has been on the other hand not much focus in the Raman data evaluation on accurate frequency determinations and/or check on selection rules. This is important, however, since deviations of these data from the standard 3D values would be an indication of quantum confinement effects. Raman signals from published papers indicate that one can be optimistic of being capable to perform single nanorod/nanowire experiments. In addition, surprisingly enough, resonant Raman scattering, which has the potential to increase the signals by orders of magnitude [32] has not been attempted yet.
Besides vibrational properties, which provide information on the material structure and composition (defects), the linear optical properties (reflectance, transmission) based on the structure optical constants are even more important as they determine the optoelectronic behaviour in light emitting devices or photonic crystals which could be fabricated based on the nanorods. The most simple techniques for the determination of the optical constants (or the dielectric function) are the reflectance based methods like Surface Differential Reflectance (SDR), Reflectance Anistropy Spectroscopy (RAS) or spectroscopic ellipsometry (SE). None of these techniques has been applied to oxide nanocrystals. However, even for the 3Dimensional bulk crystals the optical constant are only poorly known and often only in a limited spectral range. One of the reasons for that was the difficulty to grow these oxide-semiconductors in single crystal bulk form and therefore to produce well defined materials. The other difficulty is that the most important spectral characteristics of these materials occur in a region where measurements are technically more difficult. There are recent ellipsometric measurements on ZnO and MgZnO which however are limited to the band-edge [33,34]. Data at higher energies nevertheless are important, because they allow a global test on calculations, which then provide also a higher confidence for the low energy (optical) range. <<<