Symposium : I
Advanced silicon materials research for electronic and photovoltaic applications II
| start at | Subject | Num. |
|---|---|---|
| Opening Session : S. Pizzini, E. Weber | ||
| 09:00 | Opening and Introduction by the Symposium Chairs Authors : G. Kissinger, S. Pizzini, H. Tu, H. Yamada-Kaneta Resume : . | 1 1 |
| 09:10 | Ulrich Goesele –35 years of research: From defects and diffusion to nanomaterials Authors : Margit Zacharias, Nanotechnology, IMTEK, Fac. of Engineering, Albert Ludwigs University Freiburg, Georges -Koehler-Allee 103, 79110 Freiburg, Germany Resume : . | 1 2 |
| Nanocrystalline Silicon : D. Macdonald, T. Peaker | ||
| 09:40 | Conduction Mechanisms in Hydrogenated Nanocrystalline Silicon Authors : Anna Cavallini*, Daniela Cavalcoli*, Gabriel Micard+, Barbara Thereiden+, Giso Hahn+, *Department of Physics, University of Bologna, Viale C. Berti-Pichat 6/II, 40127 Bologna, Italy; +University of Konstanz, Faculty of Sciences, Department of Physics, Jakob-Burckhardt-Str. 29, 78464 Konstanz, Germany Resume : Hydrogenated nanocrystalline silicon (nc-Si:H) is an attractive material for photovoltaic applications, nevertheless some of its physical properties have been investigated only in recent times. In particular, the investigation of the transport mechanisms has up to now led to controversial results. This is mainly due to the complexity of nc-Si:H, as several phases and many defects and impurities coexist. The doping process further increases the complexity of the system as dopant atoms can segregate at nanocrystals (ncs) or at the boundaries between different phases.
An extended study of the conduction mechanisms at microscopic level of nc-Si:H thin films is here reported. The films have been deposited by Low Energy Plasma Enhanced Chemical Vapor Deposition, at deposition temperatures from 200 to 400°C and SiH4 dilution ratios from 1% to 50%, which resulted in crystal fractions ranging from 25 to 75%. p-type and n-type doped layers were obtained by using B2H6 and PH3 gases, respectively.
Sub-micron resolution current maps have been obtained by conductive atomic force microscopy. In the undoped samples all the maps presented a clear evidence of enhanced conduction in the ncs, while the disordered tissue surrounding them was mostly non-conductive. The conduction, furthermore, occurs mainly at the ncs independently of the crystalline fraction of the films. Doped films show a quite different behavior: nanocrystals are still more conductive than the surrounding tissue, but their localization in the map is different from that of intrinsic films. These results have been compared with macroscopic conductivity measurements. A unified model to interpret both microscopic and macroscopic results is advanced. | 2 1 |
| 10:10 | Coffee Break | |
| 10:30 | Carrier dynamics and recombination kinetics in Si nanocrystals Authors : W.D.A.M. de Boer, T. Gregorkiewicz Van der Waals—Zeeman Institute, University of Amsterdam Resume : Si Nanocrystals (SiNCs) embedded in a SiO2 matrix are intensively studied as a possible basis for photovoltaic devices. In spite of that, carrier relaxation and recombination processes in SiNCs are still not fully understood. In order to investigate these processes, time-resolved optical experiments are performed in the configuration of transient induced absorption (TIA), using a pump-probe setup. The fs pump pulse creates e-h pairs, whose relaxation is probed by a secondary “white” pulse. The broad spectral range of the probe pulse can give information on excited state absorption and possibly on Auger and multi-exciton generation processes (when high excitation flux is used). Selection of different excitation conditions for different samples with average sizes ranging from 2.5-5.5 nm, and comparing obtained spectral and temporal dependences, gives information on these processes.
Because of the high level of complexity of the afore-mentioned experiment, photoluminescence (PL) up-conversion technique with fs resolution is applied on the same samples, with similar excitation conditions as the TIA experiment. Cross-correlation of the results obtained with these two different experimental techniques, will gives insights on relaxation and recombination processes. In that context, the coexistence and competition between carrier cooling and direct and indirect bandgap recombination will be discussed [1].
[1] W.D.A.M. de Boer et al., Nature Phot., under submission | 2 2 |
| 10:45 | Simulations of silicon nanocrystals embedded in oxide for nanoelectronic applications Authors : Dumitru ARMEANU, Yann LEROY and Anne-Sophie CORDAN InESS (UMR 7163, CNRS UdS) - ENSPS, Bd Sébastien Brant, BP 10413, F-67412 ILLKIRCH Cedex Resume : As the size of transistors and non volatile memories is continuously scaled down, devices with silicon nanocrystals offer interesting properties directly based on quantum effects.
Our model is based on the geometrical and physical parameters of spherical nanocrystals embedded in oxide (nanocrystal radius, distance between nanocrystals, oxide thickness, bulk band structure).
As an example, we consider a flash memory whose floating gate consists of a layer of silicon nanocrystals. The channel and gate electrodes are described through energy subbands, and the silicon nanocrystal through discrete energy levels.
Two types of tunnel transport are considered here: indirect tunneling from the channel to the nanocrystal and then from the nanocrystal to the gate, and direct tunneling from the channel to the gate. This last has to be taken into account since its influence on the total current in the device has been shown.
Simulations of the electrical characteristics for various geometrical parameters provide useful results to understand physical phenomena in nanocrystal devices and to help the experimentalists to find the optimal parameters. | 2 3 |
| 11:00 | STRUCTURE AND ELECTRONIC PROPERTIES OF SILICON NANOCRYSTALS IN A SILICON CARBIDE MATRIX Authors : R. Gradmann, P. Löper, M. Künle, S. Janz, M. Hermle, S.W. Glunz Fraunhofer Institute for Solar Energy Systems, Heidenhofstr. 2, 79110 Freiburg, Germany Resume : Silicon nanocrystals (Si NC) in a dielectric matrix have promising properties for a top cell in a Si-based tandem solar cell. A control of the NCs diameter allows to adjust the bandgap of the material. We deposit layers of a-SiC:H and Si rich a-SixC1-x:H by PECVD following a multilayer approach. Si NCs form during thermal annealing and are size controlled by the thickness of the Si rich layer. The stoichiometry, the annealing temperatures and the heating rate were varied in order to control size and density of Si NCs.
In this work, we correlate the structural features of the SiC films with their electrical and optoelectronic properties. Without doping samples crystallize at 1000°C. Doping up to 100sccm (B2H6 in H2) decreases the crystallization temperature down to 900°C. In XRD measurements, peaks of Si (111) and SiC (002) become sharper with increasing dopant flux. In addition, FTIR data shows a significantly narrower Si-C stretching peak of higher doped samples. The low doped, not annealed sample has a resistivity of 4.4E10 cm. Annealing at 1000°C reduces the resistivity to 3.5E6 cm. Additional doping leads to 5.1E1 cm. Obviously, high doping clearly decreases resistivity, promoted by temperature treatment. The effect of doping on the shape of the optical gap, the urbach energy, midgap state density and activation energy are investigated with UV-Vis spectroscopy, the Constant Photocurrent Method and temperature-activated (80K – 600K) conductivity measurements. | 2 4 |
| 11:15 | Electrical properties of Si nanocrystals for advanced solar cell concepts Authors : B. Rezgui, A. Sibai, T. Nychyporuk, A. Poncet, M. Lemiti, and G. Bremond Université de Lyon, Institut des Nanotechnologies de Lyon INL-UMR 5270, CNRS, INSA de Lyon, Villeurbanne, F-69621, France Resume : One possibility to improve the conversion efficiency of Si based solar cells is the use of multiple absorbers consisting of silicon quantum dots (Si QDs) embedded in dielectric matrix. In the Si QDs absorber material charge carrier confinement results in effective band gap energies above the bulk band gap of 1.12 eV, tunable by the Si QD size. Recently, Si-based tandem solar cell made from engineered silicon nanostructures has been proposed as one of the promising technologies towards the photovoltaic efficiency enhancement [1]. Much effort was spent on material fabrication in the past in order to control the band gap of these nanostructures [2]. However, work is currently focussing on optimisation of charge carriers transport, a central issue to efficient solar energy conversion. In solar cells application involving charge carrier extraction through insulated silicon quantum dots, tunneling transport plays a dominant role. In this case, the distance among the quantum dots must be close enough to permit electron tunneling transport. To further explore the potential of the Si QDs thin films for photovoltaic application, we performed photocurrent measurements on silicon nitride films containing Si QDs prepared by plasma enhanced chemical vapour deposition under optimized growth conditions. The photocurrent spectra have been acquired using a Jobin-Yvon (HRS-2) monochromator and photogenerated current was collected using a current amplifier (Keithley 428) with a lock-in amplifier technique to improve the signal/noise ratio. The results will be discussed and compared to those reported in the literature in order to obtain a whole physical picture of electronic charge transport in this nanomaterial. Complementary characterization techniques such as photoluminescence spectroscopy and transmission electron microscopy are used to investigate the structural characteristics and optical properties of our composite structures in order to support the carrier transport results.
This work points out the limits imposed on charge transport in silicon quantum dot structures. This study shows that photocourant spectroscopy is a promising tool to investigate the electrical properties of Si QDs embedded in SiNx.
References
[1] G. Conibeer et al., Thin Solid Films 511-512 (2006) 654-662
[2] B. Rezgui, A. Sibai, T. Nychyporuk, M. Lemiti, and G. Brémond, J. Lumin. 129, 1744 (2009) | 2 5 |
| 11:30 | From silicon rich oxide to oxygen rich silicon for future PV application Authors : S. Mirabella,(1) G. Di Martino, (2,3) I. Crupi,(1) M. Miritello, (1) R. Lo Savio, (2) S. Gibilisco, (2) M. A. Di Stefano,(4) S. Di Marco,(4) F. Simone,(2) F. Priolo (1,2,3). (1) MATIS CNR-INFM, Catania, ITALY; (2) Dipartimento di Fisica ed Astronomia, Università di Catania, ITALY; (3) Scuola Superiore di Catania, Catania, ITALY; (4) IMS R&D, STMicroelectronics, Catania, ITALY. Resume : Si nanostructures will be exploited for increasing the energy conversion efficiency of PV cells. Among them, Si quantum dots (QDs) are usually obtained by annealing a silicon rich oxide (SRO) layer to let the Si exceeding the oxide stoichiometry precipitate into nanoclusters. The optical bandgap can be modulated by changing the Si excess or the QD phase (amorphous or crystalline). [1] Still, for efficiently harvesting the sun-power, SRO shows a quite high optical bandgap (above 2 eV), and, moreover, a rather high resistivity, preventing any real application in PV cells.
Such limits of SRO can be overtaken by increasing the Si content above 50% towards a new system, more rightly called oxygen rich silicon (ORS). Here, we present a large comparison of SRO and ORS in terms of both light absorption and electrical transport properties. Samples, with Si content in the 40-100 % range, were obtained through Chemical Vapor Deposition or Magnetron Sputtering, and analyzed by optical spectrophotometry, Raman and Rutherford Backscattering Spectroscopy, four-point probe measurement. By adjusting the Si content and the annealing process, a full matching between the optical bandgap and the rainbow spectrum can be obtained. Still, only the Si richer films, after a proper doping with B, showed a resistivity tunable in the 0.1-10000 Ohm•cm range. The chance of a good compromise between absorption and transport will be discussed for advanced PV.
[1] Mirabella et al., JAP 106 (2009) 103505 | 2 6 |
| 11:45 | Quantum confinement phenomena in nanocrystalline silicon Authors : L. Bagolini(1,2), A. Mattoni (1), L. Colombo(1,3), E. Poliani(4), S. Sanguinetti(4) (1) Istituto Officina dei Materiali del CNR UOS SLACS, Monserrato (Ca), Italy (2) Dipartimento di Energetica, Università di Roma ``La Sapienza\'\', Italy (3) Dipartimento di Fisica, Università di Cagliari, Italy (4) L-NESS and Dipartimento di Scienza dei Materiali, Università di Milano Bicocca, Italy Resume : Nanocrystalline silicon, consisting of a mixture of nanocrystals (nc-Si) embedded into a hydrogenated amorphous phase (a-Si:H), is of great technological relevance. By controlling the size of the grains and the microstructure features, it is in principle possible through quantum confinement to control and tune the optical absorption properties of the material [1]. The fundamental physical problem consists in understanding whether multidimensional confinement of both carriers (electrons and holes) can take place in nc-Si/a-Si:H mixtures. This is indeed a difficult task because of the disorder associated to the a-Si phase and to the interface features.
In this work [2], we predict theoretically and show experimentally the occurrence of quantum confinement in hydrogenated nanocrystalline silicon. We prove that only valence states (positively charged carriers) are confined effectively within the nanograins. The emission associated to confined states is verified by photoluminescence experiments on nanocrystalline samples with controlled grain size. According to present study, we propose nanocrystalline silicon as a promising material for oxygen-free optoelectronics, silicon-based memories and photovoltaics.
[1] L. Bagolini, A. Mattoni, and L. Colombo, Appl. Phys. Lett., vol. 94 053115 (2009).
[2] L. Bagolini et al., submitted for publication. | 2 7 |
| 12:00 | Stoichiometry of silicon rich dielectrics for silicon nanocluster formation Authors : J. Barreto, A. Morales and C. Dominguez Centro Nacional de Microelectronica IMB-CNM (CSIC), Barcelona, Spain M. Peralvarez and B. Garrido EME, Departament d\'Electronica, Universitat de Barcelona, Marti i Franques 1, 08028 Barcelona, Spain Resume : Silicon photonics has been bred by several techniques including Chemical Vapour Deposition and sputtering amongst others in order to synthesize silicon nanoclusters with CMOS-compatible technologies. Most of these techniques end up relying on the formation of nanoclusters through the diffusion and segregation of silicon atoms in a silicon-rich dielectric matrix. In a homogeneous distribution the precise knowledge of the stoichiometry of the silicon rich dielectric layer is fundamental in order to model the cluster growth.
In this work we present a parallel analysis on silicon rich dielectric layers obtained by different methods. X-Ray Photoelectron and Fourier-Transform Spectroscopy, along with Energy-Filtered Transmission Electron Microscopy and ellipsometry are used to characterize LPCVD and PECVD samples in the same theoretical silicon excess range.
The analysis show that independently on the obtaining method the initial concentration of silicon excess can be used to estimate the evolution of the nanoclusters. However secondary parameters such as the obtaining temperature and the nitrogen concentration in the layer have to be taken into account. Therefore, experimental parameters such as the flow ratio between reactant gases or the refractive index prove to be insufficient if samples obtained by different methods are compared. | 2 8 |
| 12:15 | Lunch Break | |
| Silicon Quantum Wires : K. Kakimoto, T. Mchedlidze | ||
| 14:00 | Direct electrolytic reduction of porous nanometre SiO2 in molten CaCl2 for the production of silicon nanowires Authors : Juanyu Yang, Shigang Lu*, Hailing Tu Resume : Silicon is a key semiconductor material and plays essential roles in integrated circuit, silicon chips and solar cells. Silicon nanowires as an one-dimension nanometer material have attracted much attention because of their novel physical and chemical properties. Several methods have been developed for the fabrication of silicon nanowires, such as a laser-ablation metal-catalytic method, an oxide-assisted method, and a solution technique. In this study, the authors have proposed a novel of crystalline Si nanowire growth by electrolysis in molten CaCl2 at 900C using the nanometer SiO2 as a starting material. It was confirmed that the nanowire growth proceeds through the formation of the nanostructured wires consisting of silicon core and amorphous SiO2 shell. The as-synthesized nanowires show the microstructure very similar to that prepared by the other method such as metal-catalytic method and the oxide-assisted method. On the other hand, this method takes a number of merits such as high productivity, low cost, convenient operation, and reduction of contaminants. It can provide a great possibility for the mass production of the Si nanowires. Furthermore, by changing experimental temperature, this electrochemical approach is revealed to be applicable to produce a variety of nanostructured silicon such as nanotubes, nanowires and nanoparticles consisting of Si and SiO2. Or by mixing powders of different metal, containing metal other structured silicon can be produced. | 3 1 |
| 14:15 | Photoluminescence Properties of Silicon Nanowires Produced by Wet Chemical Etching Using Varying Etching Conditions (Times, Temperatures) Authors : Felix Voigt 1,2, Gottfried H. Bauer2, Vladimir Sivakov1, Andreas Berger1,3, Silke Christiansen1,4 1. Institute of Photonic Technology, Jena, Germany. 2. Institute of Physics, Carl-von-Ossietzky University, Oldenburg, Germany. 3. Max Planck Institute of Microstructure Physics, Halle, Germany 4. Max Planck Institute for the Science of Light, Erlangen, Germany Resume : Silicon nanowire (Si-NW) samples were prepared by Wet Chemical Etching of crystalline silicon wafers using various etching times. The diameters of these SiNWs ranged from 30 to 200 nm. Photoluminescence (PL) measurements were performed with excitation at 488 nm and a photon flux density of 8.3e17 s^-1 cm^-2. According to the diameter sizes > 10 nm, from quantum confinement theory no shift in PL peak energy compared to crystalline silicon is expected. However, PL measurements show peak emission energies in the range 1.4 to 1.6 eV. Moreover, a monotonic increase of the PL emission peak energy with etching time was found. After further treatment of the samples with HF, substantial PL emission was still detectable and the measured PL peak was pinned at 1.4 eV, irrespective of etching time. This forms the key observation of this contribution. We explain the observations by the hypothesis that the remaining PL emission is generated by nanosize structures located at the rough sidewalls of the Si-NWs. This hypothesis is supported by Transmission Electron Microscopy. Furthermore, we explain the shift of the PL emission peak with etching time for the untreated samples by superposition of two competing contributions to PL, originating on the one hand from nanostructured sidewalls of the Si-NWs and on the other hand from surface states of SiOxHy at the Si-NW surfaces. | 3 2 |
| 14:30 | Controlling the density of silicon nanowires grown on glass substrates for photovoltaic applications Authors : Benedict O’DONNELL, Jinyoun CHO, Linwei YU, Pere ROCA i CABARROCAS LPICM, Ecole Polytechnique, CNRS, 91128 Palaiseau, France Resume : Silicon nanowires are a promising material for future solar cells as they offer good carrier mobility, trap light efficiently and can be grown cheaply in commercial reactors (1). Although understanding of the role the substrate and growth conditions play on the properties of individual wires has advanced dramatically in recent years, limited progress has been made towards controlling their density - a crucial property in optimising light trapping effects and achieving conformal deposition of the subsequent layers in the cell. Most approaches rely on colloids or lithography, which are unsuitable for an economical, one-step process. We present evidence that the density of nanowires can be modified in a single pump-down process during which the cell is also made. Metal was sputtered onto ZnO-covered glass substrates with different surface textures and stoichiometries, and thermally diffused over their surface. SEM imaging revealed conditions under which the density of metal drops on the ZnO surface had decreased. By exposing these samples to a silane and TMB plasma, low-density batches of p-type Si nanowires were grown and covered with n-type amorphous Si to produce solar cells. Absorption measurements of the wires and I(V) characteristics of the cells will be presented and the implications of the wire density on light trapping and energy conversion efficiency will be discussed.
1 Yu L, O’Donnell B, Alet P, Conesa-Boj S, Peir F,Arbiol J, Roca P,. Nanotechnology 20 (2009) 2 | 3 3 |
| 14:45 | Ultrasharp conical Si nanowires Authors : J. Cervenka1, M. Ledinsky1, H. Stuchlikova1, J. Stuchlik1, S. Bakardjieva2, K. Hruska1, J. Holovsky1, Z. Vyborny1, A. Fejfar1, and J. Kocka1 1 Institute of Physics, Academy of Sciences of the Czech Republic, v. v. i., Cukrovarnicka 10, 162 53 Prague 6, Czech Republic 2 Institute of Inorganic Chemistry, Academy of Sciences of the Czech Republic, v. v. i., 25068 Rez, Czech Republic Resume : Silicon nanowires have attracted tremendous attention in the last decade not only due to their special physical properties but mainly for their potential applications in electronic, photonic, thermoelectric and sensing devices. Despite of many possible applications, the current growth of Si nanowires using a metal catalysis and chemical vapor deposition (CVD) or molecular beam epitaxy (MBE) is slow and expensive for a large scale production.
Here we present a new way of preparing Si nanowires by plasma-enhanced chemical vapor deposition (PECVD) using gold nanoparticle catalysis that we have recently developed [1]. This method produces Si nanowires with very fast growth rates, two orders of magnitude faster than by CVD and MBE techniques, while using significantly lower substrate temperatures. In addition, Si nanowires can be grown in the form of ultrasharp cones with a typical tip diameter below 10 nm. Si nanowires are composed of crystalline and amorphous Si as confirmed by Raman spectroscopy and transmission electron microscope. A systematic study of Si nanowires morphologies by scanning and transmission electron microscopies under different growth conditions will be presented. Finally, we will address optical properties of Si nanowire layers, which have shown very good antireflection effects comparable to black silicon.
References:
[1] J. Cervenka, M. Ledinsky, H. Stuchlikova, J. Stuchlik, Z. Vyborny, J. Holovsky, K. Hruska, A. Fejfar, and J. Kocka, Phys. Status Solidi RRL 4, 37–39 (2010). | 3 4 |
| 15:00 | Confined growth of silicon nanowires for the realization of low cost solar cell Authors : D. Buttard1,2, L. Dupré1, T. Bernardin1,3, M. Zelsmann3 1 CEA-Grenoble/INAC/SP2M/SiNaPS, 17 rue des martyrs, 38000 Grenoble, France 2 Université Joseph Fourier, Grenoble-1, 38000 Grenoble France. 3 LTM-CNRS c/o CEA-LETI-MINATEC, 17 rue des martyrs, 38000 Grenoble, France Resume : The future use of nanostructures (carbon nanotubes, quantum dots, nanowires…) in devices will only occur if the growth of these nanostructures is self-organized, with a high density and controlled in terms of dimensions and localization in space. Currently, many nanoobject are made and studied for their physical properties but these studies are in general done on one or few nanoobjects and uncontrolled in terms of direction and localization.
We have developed generic high-density (1010 cm2) and self-organized matrix of nanoporous alumina on to large area (100 mm diameter wafer). This matrix will locate and guide wires during their growth. Furthermore, the good mechanical, thermal and chemical stability of the matrix should also facilitate their integration in devices. For example we fabricate porous alumina membrane using a double anodization procedure (hole period around 92 nm, we domains of few microns).
After the realization of nanoporous alumina on to silicon substrate, catalyst was deposed in to the pores and then the silicon nanowires have been growth by catalytic mode in a CVD reactor (10 to 50 nm diameter wires) [1]. The use of a porous alumina membrane enables to grow wires on non-aligned substrates (effective growth direction different from the preferential growth direction), with a good regularity, which can be used for low cost solar cell.
In this paper we will present results from elaboration and characterization of high density silicon nanowires embedded in nanochannels of nanoporous alumina template. We will also present and discussed results from structural characterisation by electron microscopy and by x-ray diffraction measurements.
[1] D. Buttard, T. David P. Gentile, F. Dhalluin, T. Baron, Phys. Stat. Sol. RRL 1, 19 (2009). | 3 5 |
| 15:15 | I-V characteristics of silicon during electrochemical anodization Authors : A. GHARBI A. SOUIFI B. REMAKI Resume : The property of selective formation of porous silicon PS is presented which is a basis step for the local electrical isolation process. In fact, when a silicon wafer of different regions of doping concentration is anodized at a fixed potential, the dissolution current will be more important in regions of higher doping level and PS formation can only take place there. The analysis of anodic current-potential I-V characteristics of p-type silicon in the condi-tions corresponding to porous silicon formation shows a potential shift in the I-V characteristics with the silicon doping concentration which can explain the selectivity of PS formation towards the doping level of silicon. In ad-dition, it is seen that the mechanism at the origin of PS formation on p-type silicon in concentrated hydrofluoric acid solution is determined by the charge exchange at the silicon surface over a Schottky barrier which is formed at the interface through a thermionic process. | 3 6 |
| 15:30 | Coffee Break | |
| Point Defects I : B. G. Svensson, Y. Kamiura | ||
| 16:00 | The impact of dopant compensation on the boron-oxygen defect in crystalline silicon Authors : D. Macdonald, A. Liu, F. Rougieux, A. Cuevas School of Engineering, The Australian National University, Canberra ACT 0200 Australia B. Lim and J. Schmidt Institut fur Solarenergieforschung Hamlen (ISFH), Am Ohrberg 1, D-31860 Emmerthal, Germany Resume : Significant levels of dopant compensation are likely to occur in new forms of solar-grade silicon feedstocks. The existence of additional boron in compensated p-type silicon raises the prospect of increased recombination due to the presence of the well-known boron-oxygen defect [1], making such material less attractive for solar cells. However, a recent study [2] has hinted that the extent of this defect is in fact not determined by the total concentration of boron acceptors in a wafer, but by the net doping, leading to the postulation of boron-phosphorus pairs (B-P) in compensated wafers.
The aim of this work is to systematically study the generation of boron-oxygen defects in deliberately compensated Czochralski-grown silicon that is otherwise extremely pure. The concentration of boron-oxygen defects is measured directly via carrier lifetime measurements on compensated and non-compensated wafers of similar net doping, giving a direct comparison. The rate at which the defects are formed under illumination is also reported. Combining these measurements confirms that the defect concentration is indeed determined only by the net doping. However, we find that the proposed presence of B-P pairs is not consistent with our measurements of the majority carrier mobilities, or the position of the characteristic lifetime ‘crossover point’ caused by the dissociation of iron-boron pairs. We also consider the implications of the results for compensated n-type silicon wafers.
[1] J. Schmidt and K. Bothe, Physical Review B 69 (2004) 024107.
[2] R. Kopecek, J. Arumughan, K. Peter, E. A. Good, J. Libal, M. Acciarri and S. Binetti, Proceedings 23rd PVSEC, Valencia, Spain (2008) 1855. | 4 1 |
| 16:30 | Spectroscopic and electrical characterization of compensated silicon Authors : S. Binetti1, A. Le Donne1, M. Acciarri1, D. Macdonald2, T. Iwai3 and M. Tajima3 1 CNISM and Dept. of Material Science, University of Milano-Bicocca, via Cozzi 53, 20125 Milano, Italy 2 Dept. of Engineering, College of Engineering and Computer Science, Australian National University, Canberra, Australian Capital Territory 0200, Australia 3 Institute of Space and Astronautical Science / JAXA, Sagamihara 229-8510, JAPAN Resume : It is widely known that PV industry is expected to remain mostly based on crystalline silicon at least for the next decade, therefore a substantial increase in the Si production is needed both in order to sustain the high PV growth rate and to decrease the cost per Watt peak. For this aim, efforts were made in order to develop processes for the purification of metallurgical grade Si, which results in low cost but less pure Si, the so-called solar grade silicon (SoG-Si). Besides a variety of metallic impurities, SoG-Si often contains a large amount of the doping elements. Thereby, the effect of the compensation, has to be taken into account, since this can impact solar cell performance. In this work, the spectroscopic and electrical features of compensated p-type Si samples were monitored by different techniques. A photoluminescence analysis in the temperature range between 10 and 30 K allows the identification of the donor-acceptor pair recombination related peak. A detailed analysis of this peak has been carried out in view of its using as fingerprint of the presence of compensation in SoG-Si. Furthermore, a spectroscopic analysis of photoluminescence at liquid He temperature enables the determination of donor and acceptor impurity concentrations. The impurity concentrations obtained by optical method have been also compared with those coming from two different electrical techniques, the former based on carrier lifetime and the latter based on Hall mobility measurements. | 4 2 |
| 16:45 | Study of the Light-Induced-Degradation in compensated n-type silicon Authors : Thomas Schutz-Kuchly (CEA-INES), Jordi Veirman (CEA-INES), Yannick Veschetti (CEA-INES), Sébastien Dubois (CEA-INES), Dick Heslinga (CEA-INES) ,Olivier Palais (IM2NP) Resume : The need of the Photovoltaic industry in low cost silicon (Si) leads to the use of Si purified via the metallurgical route (SoGM-Si). This material will be compensated where both doping species are present, i.e. Boron (B) and Phosphorus (P) at high concentrations.
p-type SoGM-Si is very sensitive to Light-Induced-Degradation (LID) effects. LID is due to the formation under illumination of a recombination-active center associating a B atom and an Oxygen dimmer (O2). Studies conducted on p-type compensated Si have shown that increasing the compensation level reduces the LID effect leading to the hypothesis of Boron-Phosphorus (B-P) pairs. Then, compensated n-Si containing high B and O should not be sensitive to LID effect since all the B is compensated.
The aim of this work is to study if solar cells realized on n-type compensated material can be subject to LID, because recent studies have shown a doubt on the B-P pairs. 2 mV drop in Voc being related to a volumic degradation has been observed on n-type SoGM-Si cells. This unexpected degradation led us to study the effect of LID on Boron-Phosphorus compensated n-type Czochralski Si ingot which only contains B, P and O. Because of the low segregation coefficient of P, this ingot presents a resistivity varying from 0.45 ohm.cm to 0.10 ohm.cm. The LID effect will be studied on the cells fabricated in order to check if the assumption of B-P pairs is valid or if compensated n-type Si can be subject to LID involving BOi2 pairs. | 4 3 |
| 17:00 | Hall mobility drops in disordered Boron-doped Czochralski Silicon compensated by thermal donors activation Authors : J. Veirman, S. Dubois, N. Enjalbert, J-P. Garandet, D. Heslinga CEA, LITEN, INES, 50 av du Lac Léman, F-73377, Le Bourget du Lac, France M. Lemiti Laboratoire de Physique de la Matière, UMR-CNRS 5511, INSA de Lyon, Bât. B. Pascal, 7 avenue Jean Capelle, 69621, Villeurbanne, Cedex, France Resume : The use of Solar-Grade Silicon (SoG-Si) purified by metallurgical routes is identified as a major cost-reduction opportunity for the photovoltaic industry. Because dopants are not easily removed during its purification, SoG-Si is compensated i.e. it contains both donor and acceptor doping impurities at high concentrations.
Compensation, by decreasing the recombination strength of harmful species, can lead to carrier lifetime improvements in SoG-Si. Conversely, the majority carrier mobility (µ) is likely to be lower than in uncompensated Si since SoG-Si contains more ionized doping impurities for a given carrier density. The aim of this work is to study the variations of µ with the Compensation Level (Cl) and to determine whether current µ models correctly describe µ in compensated Si. For this purpose, µ was obtained by Hall measurements in Boron-doped Electronic Czochralski Si where high Cl were achieved through the controlled activation of the Thermal Donors (TD). We found that µ models cannot reproduce our data beyond a threshold Cl at which µ starts to decrease steeply.
We attribute these drops to a highly heterogeneous TD distribution resulting in spatial variations of resistivity/type of conductivity that can alter the Hall voltage or directly µ (space charge scattering). We suggest that such drops may occur beyond some Cl whatever the doping impurities under consideration, but that their onset should be dependent on the crystallisation process. | 4 4 |
| 17:15 | Atomistic simulations of point defect diffusion in Si and SiGe Authors : Pascal Pochet , Damien Caliste, Konstantin Rushchanskii, Frédéric Lançon and Thierry Deutsch Laboratoire de simulation atomistique (L_Sim), SP2M, INAC, CEA, 38054 Grenoble cedex 9, France Resume : Point defect engineering (PDE) is a method used to control species diffusion in material during various microelectronic processes [1]. Point defect diffusion mechanisms as well as their associated energetics are thus key parameters to design proper PDE strategies.
In this paper, we will report kinetic lattice Monte Carlo simulations of vacancy-assisted diffusion in silicon and silicon germanium alloys. The formation and migration energies for vacancies in SiGe alloys have been calculated by means of ab initio calculations. This system has been carefully treated, taking into account the Jahn-Teller effect and the influence of composition and stress.
In a first part, we will show that the observed temperature dependence for vacancy migration energy in silicon [2] is explained by the existence of three diffusion regimes for divacancies. This characteristic has been rationalized with an analytical model [3]. The exact position of this temperature regime strongly depends on vacancy concentration. Moreover, the model accounts well for the regimes observed after He implantation in silicon [4]. Thus this model can be helpful for processes were such implantations are used.
In a second part, we will present results on the stress enhanced vacancy-mediated diffusion in biaxially deformed (100) SiGe-films. This phenomenon is usually quantified through the Q\' strain derivative introduced by Cowern et al. [5]. Based on our calculations and on a new analysis of the published experimental data, we will show that Q\' is unexpectedly temperature [6] and strain dependent. These informations are again mandatory to design an appropriate PDE strategy in strained SiGe nano-devices.
[1] Shao L, Liu J, Chen Q.Y.and Chu W.-K, Mater. Sci. Eng. Rep. 42,. 65-114 (2003).
[2] H. Bracht, J. F. Pedersen, N. Zangenberg, A. N. Larsen, E. E. Haller, G. Lulli, and M. Posselt, Phys. Rev. Lett. 91, 245502 (2003).
[3] D. Caliste and P. Pochet Phys. Rev. Lett., 97 135901 (2006)
[4] M. L. David, M. F. Beaufort, and J. F. Barbot, J. Appl. Phys. 93, 1438 (2003).
[5] N. E. B. Cowern, P. C. Zalm, P. van der Sluis, D. J. Gravesteijn, and W. B. de Boer, Phys. Rev. Lett. 72, 2585 (1994).
[6] K. Z. Rushchanskii; P. Pochet and F. Lançon Appl. Phys. Lett. 92 152110 (2008) | 4 5 |
| 17:45 | A physical model of vacancy engineering for ultrashallow junction control Authors : Chihak Ahn/Newcastle University Joochul Yoon/Newcastle University Nick Bennett/Newcastle University Nick Cowern/Newcastle University Nikolas Zographos/Synopsys Switzerland LLC Resume : Co-implantation of non-dopant ions such as Si can be used to create separated vacancy-rich and interstitial-rich crystalline silicon regions – the vacancy-rich region near the surface and the interstitial-rich region deeper in the silicon – in order to modify the activation and diffusion behaviour of a subsequently implanted dopant. This approach can be beneficial for enhancing boron activation in ultra shallow junctions, since the excess of vacancies close to the entry surface of the Si ion beam acts to control transient enhanced diffusion and clustering of boron ions implanted after the Si implant step. We have investigated the B diffusion/activation in Si co-implanted Si using a combination of density functional theory (DFT) and continuum modelling. The binding energies of small vacancy clusters were calculated with DFT, and then used as key parameters in continuum modelling. The model predicts small vacancy clusters to be quite stable up to 800oC and Boron transient enhanced diffusion (TED) to be suppressed. In addition, the initial recombination between vacancies and B interstitials results in high B activation at low temperatures, which is consistent with experimental observation. The model reproduces the annealed B profiles for various co-implant doses and temperatures. | 4 6 |
| 18:00 | Structure and electronic properties of trivacancy and trivacancy-oxygen complexes in silicon Authors : V.P. Markevich1, A.R. Peaker1, B. Hamilton1, S.B. Lastovskii2, L.I. Murin2, J. Coutinho3, V.J.B. Torres3, L. Dobacewski4, and B.G. Svensson5 1 Photon Science Institute, University of Manchester, M60 1QD, UK 2 NAS of Belarus, P. Brovka Str. 19, Minsk 220072, Belarus 3 I3N, University of Aveiro, 3810-193 Aveiro, Portugal 4 Institute of Physics, 02-668 Warsaw, Poland 5 Oslo University, N-0316 Oslo, Norway Resume : The trivacancy (V3) is one of the most abundant defects in Si irradiated with high energy particles (ions, neutrons or electrons with E>5 MeV). It is of crucial importance in the radiation hardness of detectors used in high energy physics and medical radiation therapy. We have shown {V.P. Markevich et al., PRB 80, 235207 (2009)} that in the neutral charge state the trivacancy in Si has a fourfold coordinated (FFC) configuration with a lower formation energy than the (110) planar structure. In the FFC configuration V3 has trigonal symmetry and an acceptor level at Ec-0.075 eV. The interaction of trivacancies (mobile at T>200 C) with interstitial oxygen forms V3O complexes with the first and second acceptor levels at Ec-0.46 eV and Ec-0.34 eV.
We have undertaken DLTS and Laplace DLTS on n+p silicon structures irradiated with 6 MeV electrons and identified donor levels of the V3 and V3O complexes. It is found that both defects possess two donor levels in the (110) planar configurations. The first donor level is at Ev+0.19 eV and Ev+0.24 eV, and the second donor level is at Ev+0.11 eV and Ev+0.15 eV for the V3 and V3O complexes, respectively. The overall picture, including structural details, relative stability, and electrical levels, is accompanied and supported by ab-initio modeling studies. | 4 7 |
| 18:15 | Defects formed by pulsed laser annealing: electrical properties and depth profiles in n-type silicon measured by deep level transient spectroscopy Authors : D. Schindele (a), P. Pichler (b, a), J. Lorenz (b), P. Oesterlin (c), H. Ryssel (a,b) (a): Chair of Electron Devices, University of Erlangen-Nuremberg, Cauerstrasse 6, 91058 Erlangen, Germany (b): Fraunhofer Institute of Integrated Systems and Device Technology (IISB), Schottkystrasse 10, 91058 Erlangen, Germany (c): Innovavent GmbH, Bertha-von-Suttner-Strasse 5, 37085 Goettingen, Germany Resume : The ongoing downscaling of microelectronic devices requires dopant activation with diffusion kept to a minimum. One method considered in device processing is microsecond annealing by pulsed lasers. This is a highly non-equilibrium process during which intrinsic point defects generated at the surface diffuse into the material and form a variety of defects during cooling-down. Since the defects lie close to the surface, they may well affect the effective concentration and the formation of leakage currents in pn-junctions of advanced technologies. These defects may be clusters of intrinsic point defects or contain residual impurities like oxygen or dopants. In this work, a bare n-type silicon substrate (phosphorus concentration 5e14 cm-3) was annealed with a pulsed laser with maximum temperatures from about 800 K up to 1650 K. The defects remaining after annealing were characterized by deep level transient spectroscopy (DLTS) to obtain their activation energies and capture cross-sections for majority carriers in n-type silicon. Additionally, the depth profiles of these defects were measured via a variation of the pulse voltage. The dominating defects found have levels at
Ec-0.45 eV, Ec-0.41 eV, Ec-0.34 eV, Ec-0.3 eV, Ec-0.21 eV, and Ec-0.17 eV. Part of these defects can be attributed to PV, V2 and VO. | 4 8 |
| Back | ||
| Une réalisation advisa.fr |  |