Search

Since the first anodic porosification of semiconductors in the 1950s, numerous studies have reported a wide range of porous structures. The conditions used during the growth of pores can explained the shapes obtained but the formation mechanism and especially the influence of surface chemistry during porosification still raise many questions. It is in such context that the contribution of liquid ammonia (-55°C) is crucial. Indeed, this remarkable non-aqueous solvent allows, on the one hand, to obtain novel porous structures on n-InP and, on the other hand, it allows for the first time a specific and original interfacial chemistry, governed by the chemistry of nitrogen. In fact, in liquid ammonia, the dissolution of III-V semiconductors (GaAs, InP) is systematically done after a surface passivation step. This passivation step is particularly stable on InP and revealed the formation of a phosphazene type film (monolayer) resulting from the concomitant oxidation of the solvent and the semiconductor. The "multi-material" approach is a key factor, as it underlines the determinant role of the surface chemistry. In fact, at higher current densities, the porous structures of n-InP and n-GaAs show contrasting structures, which give an original behaviour compared to the aqueous medium. In order to determine the efficiency of the dissolution mechanism (at lower and higher current density), the quantity of indium (or gallium) salt will be quantified by Inductively coupled plasma optical emission spectroscopy analysis after ammonia evaporation. Moreover, the remarkable reproducibility of the anodic behaviour allowed physico-chemical characterisations of the porous surface, supported by scanning electron microscopy and the variation of the interfacial potential, photoluminescence and X-ray photoelectron spectroscopy. More specifically on n-InP, where an original structure was revealed, these analyses, during the early stages of porosification, led to the identification of a porosification mechanism in this solvent. Furthermore, at higher current densities, an amorphous film becomes thicker on the surface of the porous structure. Further porosification requires the permeability of this amorphous structure which would most probably result from the polymerisation of phosphazene since the presence of nitrogen could be clearly identified on the pore walls by Auger spectroscopy.

Due to its high purity (99.9995%) liquid ammonia (NH3 liq.) has found a distinguished place in electrochemical studies of III-V semiconductors (III-Vsc) at low temperature (-55°C and atmospheric preasure). NH3 liq. provides interfacial electrochemistry in an original environment strongly different from the aqueous medium. However, in both solvents, the concepts of interfacial electrochemistry are the same onto III-Vsc. The contribution of NH3 liq. is significant in the understanding of fundamental electrochemical reactions such as hydrogen evolution and oxygen photo-reduction mechanism on III-Vsc (InP and GaAs). The aim of this article is to describe why NH3 liq. is a powerful solvent to understand charge transfer mechanisms at the interface III-Vsc/electrolyte.

For the last several generations of CMOS technology, copper has been used widely as an interconnect material in the back end of the line (BEOL). [1] [2] [3] As the current nodes technology are decreasing continuously, several copper scaling challenges appeared such as the Cu barrier thinning, the copper electromigration... [4] [5] Hence, cobalt and ruthenium are proposed as replacement materials. Recently, Co has made inroads in microelectronics, specifically on-chip metallization, serving as a capping layer for Cu to diminish electromigration but also provide a better conductivity for line width inferior to 10 nm [6] [7] [8]. In this very challenging context, the control of Co surface oxidation process is strategic. Conformal Co seed characteristics (i.e electronic conductivity), preceding the following filling chemistry, are partially related to its surface oxide. Its understanding, then its control, will be determinant to access to a perfect and reliable filling of the nodes. To collect quantitative information, we carefully investigate the behavior of a pure Co metallic surface which is exposed to air oxidation. For this purpose, we have used a 400nm thick Co layer electrochemically plated using Kari© aveni acidic chemistry. The Co layer is electrochemically deposited on Co seed (3nm) /Ta/TaN barrier/Tetraethyl orthosilicate (TEOS). The latter, with its native oxide, is introduced in ultra-high vacuum (UHV) of a XPS Nexsa ThermoFisher spectrometer. Then, a totally deoxidized Co surface is obtained with a argon beam sputtering. The initial free oxide bulk metallic Co is then checked by XPS. Starting from this surface, we explore again by XPS, consequences of different air exposure times using the entrance prepation chamber of the spectrometer. According to the time air-exposure, the kinetic of the oxide growth mechanism can be evaluated. Our work shows clearly that the metallic Co surface evolves very quickly to an ultra-thin ( Moreover, a simulation of behaviors of this oxide surface coverage immersed in acidic solution will be proposed to understand how an ultra-thin Co seed layer interacts in Co filling solution. For both cases (air-exposure and cobalt filling solution) the results comparison and a discussion will be proposed . References [1] D. Edelstein et all"IEEE IEDM Tech. Digest," p. 773, 1997. [2] S. Venkatesan et all "In Electron Devices Meeting," p. 769, 1997. [3] S. Tyagi, et all, "In Electron Devices Meeting,," p. 567, 2000. [4] H.-J. Cho, et all "VLSI Technology, IEEE Symposium," p. 1, 2016. [5] S. Y. Wu et all "Electron Devices Meeting (IEDM)," p. 2, 2016. [6] V. P. Graciano et all Journal of The Electrochemical Society, vol. 166, p. 3246, 2019. [7] J. H.-C. Chen et all 2016 IEEE International Interconnect Technology Conference / Advanced Metallization Conference (IITC/AMC);," p. 12, 2016. [8] C. Adelmann et all "Interconnect Technology Conference/Advanced Metallization Conference (IITC/AMC);," p. 173, 2014. Figure 1

The purpose of this work is to study in detail the effects of monoatomic and gas cluster ion beam (GCIB) on a semiconducting electrode (InP) thanks to the correlation between X-Ray Photoelectron Spectroscopy (XPS) and electrochemistry. Such approach allows estimating the overall change in the electrical, chemical and structural properties of the modified semiconductor (SC). Recent developments in ion sputtering source enable a better control of the perturbations induced at the surface of the material. Furthermore, anodic dissolution will be presented as a tool to recover the initial characteristics of the modified electrode. The interaction of ions with matter and especially surfaces has been an active topic for both fundamental and applied research over the past decades and is still relevant today. Thanks to ion bombardments, various types of surfaces can be chemically, mechanically, optically or electrically modified. However, the understanding of the inherent phenomena and physical/chemical mechanisms involved remains incomplete. When ions, whose energy range between 1 and 10 000 eV, interact with a surface, many processes can occur and the energy distribution (dissipation) could differ markedly from one material to another. Nevertheless, whatever the source of modifications considered, the common fundamental question is the nature of the effects induced by the perturbation and its evaluation (area, depth). The consequences of ion bombardment on the surface of a SC may have repercussions: formation of ordered nanostructures via low-energy ion-beam irradiation as a bottom-up nano-manufacturing technique; ion implantation for SC device technology; optical lithography using ion beam or even functionalization of 2D materials. This topic has gained specific interest in the field of surface analysis where depth-profiling is required (e.g. XPS, Auger Spectroscopy or Secondary Ion Mass Spectroscopy). In contrast with monoatomic ions (e.g. Ar+), cluster ions (e.g. Arn +) do not penetrate deeply into target materials. Therefore, the energy of their impact is dissipated within the first few nanometers of the surface, resulting in low sputtering yields and minimal surface damage. However, the effects of these new generations of GCIB are not yet fully understood. The strong originality of our approach is to investigate the surface modifications induced not only by UHV characterization technique (XPS) but also with the use of solid/liquid interface electrochemical response for which SC have specific behavior. These efficient and quantitative tools provide relevant information on the modification performed at the SC surface: metallization, amorphization, preferential sputtering. The high number of parameters that govern the interfacial electrochemistry of the SC/electrolyte system (band bending, SC doping level, double layer capacitance, space charge layer, SC electrical impedance, photopotential) make this approach pertinent and sensitive to evaluate the perturbation amplitude. An example is given on the figure where the metallization of an InP electrode can be observed in vacuum after successive Arn + etching sequences by the shift of the Fermi level (XPS valence band region) and in solution by the narrowing of the depletion region and the flattening of the Mott-Schottky plot. In addition, a very important characteristic of SC interfacial electrochemistry is its availability to generate in situ anodic dissolution at nanoscale. Thereby, it represents an accurate and quantitative tool to estimate the thickness of the modified region. Moreover, using well-controlled dissolution, the initial properties can be retrieved. These results are compared to those obtained after annealing under UHV. Thus, the electrochemical responses will be associated directly to surface chemical information such as: loss of stoichiometry, Ar implantation, shift of the Fermi Energy level position, photo-peak broadening. Figure 1

This paper describes some fundamental concepts of Fermi level equalization between InP (n- and p- types) and two contrasted surface configurations: the liquid interaction on the one hand and the ultra high vacuum on the other. At open circuit potential, the junction InP / liquid was gradually illuminated according to the luminous flux. A decrease of the band bending was observed for both types using n & p photopotentials in acidic aqueous electrolyte. A quasi saturation of the difference was observed for sufficient light intensities. Despite strongly different surface conditions, the evolution of the band alignment of n- and p- doped InP toward electrolyte gives rise to a difference between the open circuit potential under illumination which is very close to the peaks positions separation observed in ultra high vacuum for XPS measurements. These observations allow comparison of the different band bending configurations of n & p-InP surfaces in liquid and in vacuum.

III-V materials have yet proved to be promising candidates for photovoltaic applications. Solar cell technologies based on III-V semiconductors are competitive in term of fabrication cost but also have the potential to reach the highest photovoltaic efficiencies. To overcome the 30% conversion limitation established for a single junction, the Shockley Queisser limit [1], due to different losses that cannot be avoided (transmission, thermalization, emission…), different solar cell architectures have been proposed. One of them, multi-junctions, are using a stack of absorbing materials enabling to cover a wider part of the solar spectrum with high efficiency due to their optimized bang-gaps. Among other recent concepts, intermediate band solar cells have gained interest with predicted efficiency greater than 60% (under maximum concentration) and almost 48.2% under one sun [2], similar to the efficiency potential of a 3 material stack, but with a single material. By implementing quantum dots at the surface of a III-V single layer, an intermediate band can be generated [3, 4] and tuned by changing the quantum dots nature, size and shape. More complex structures with multi-stacked quantum dots have also been studied [5]. The present work focuses on the preparation of III-V quantum dots surfaces using wet chemistry. Not only dots evolution but also the development of a specific soaking to eliminate the wetting layer, detrimental for optoelectronic properties, will be shown. Numerous studies and formulation have been reported for chemical etching of III-V semiconductors [6]. Nevertheless, the remaining challenge concerns the implementation of this chemical engineering at nanometric scale [7], in consistency with the miniaturization trend and requirements of the materials, structures and devices. The study is carried out on InAs and InGaAs quantum dots grown by MBE on GaAs substrates. They present different surface densities. The structure of the samples is presented figure 1 where the wetting layer, essential for the quantum dots epitaxial growth, is illustrated. A multi-technique methodology combining XPS, AFM and nano-Auger is employed to precisely determine the surface chemical and morphologic modifications from micrometric to nanometric scale. Thanks to XPS, the evolutions of the fine chemistries of quantum dots and their surface are accessible, giving crucial information about the dissolution processes and enabling to re-adjust the experimental conditions. In Figure 1 b, the In3d photopeak evolution with the chemical treatment employed on a sample presenting only the wetting layer is shown. Indium atoms are only present in the wetting layer in this case. Thus it represents the tracking element indicative of the maintenance, thinning or elimination of this layer. The complementary use of AFM brings the evidence of the dots conservation or disappearance as illustrated Figure 1c. Nano-Auger provides additional information on local chemistry. Various formulations, in acidic or base media, are developed to enable a differential dissolution between the wetting layer and the quantum dots. Both oxidation-deoxidation cycling and direct dipping procedures were experimented and optimized taking account the differences in oxides solubility. The various scenarios likely to occur during deoxidation are presented Figure 1a, the elimination of QD being also possible by the dissolution of the wetting layer below. Finally, a comparison of capabilities for nano-etching and possible selectivity of the different solutions considered here will be presented. References [1] W. Shockley and H.J. Queisser, J. Appl. Phys. 32 (1961), 510. [2]A. Luque, A. Marti, Phys. Rev .Lett. 78 (1997) 5014-5017. [3] T. Sogabe, Y. Shoji, M. Ohba, K. Yoshida, R.Tamaki, H.-F. Hong, C.-H. Wu, C.-T. Kuo, S. Tomić, and Y. Okada, Scientific Reports (Nature Publishing Group), 4, (2014)/4/25. doi:10.1038/srep04792. [4] E. Lopez, A. Datas, I. Ramiro, P.G. Linares, E. Antolin, I. Artacho, A. Marti, A. Luque, Y. Shoji, T. Sogabe, A. Ogura, Y. Okada, Sol. Energy Mater Sol. Cells 149 (2016) 15-18. [5] T. Kada, S. Asahi, T. Kaizu, Y. Harada and T. Kita, Phys. Rev. B 91 (2015) 201303(R). [6] A.R. Clawson, Reports: a review journal, Mater. Sci. Eng. 31 (2001) 1-438. [7] D. H. van Dorp, S. Arnauts, M. Laitinen, T. Sajavaara, J. Meersschaut, T. Conard and John J. Kelly, Appl. Surf. Sci. 465, (2018), Pages 596-606. Figure 1

The today increase demand for high-performance microelectronic products requires new reliable interconnection metallization for microelectronic packaging. Copper electroplating is a cost-effective technology, mainly adopted to manufacture interconnects in integrated semiconductor devices [1-3]. In copper 3D interconnections, Damascene or Through Silicon Via (TSV) technologies; numerous factors have an influence on the filling and microstructural results, as the copper bath composition, the applied current density and waveform, the plating cell hydrodynamics…, conditioning the final resistivity of the copper layer. These parameters must be more and more optimised to achieve the filling of a large challenging range via dimensions and very aggressive aspect–ratio. Nevertheless whatever the design of the vias, the substrate composition, the plating cell and other parameters, the free-defect filling of interconnections is obtained only by bottom-up mechanism, where the copper deposition is promoted by superfilling effect and limited on the top, thanks to the addition of two or three organic additives. Without organic additives, only conformal or sub-conformal platings can be grown, causing voids, defects, lack of adherence and poor compactness. The superfilling mechanisms, depending on the additive composition, their relative concentration and the competition between them, are well described and simulated by the curvature-enhanced acceleration accumulation (CEAC) mechanism [4]. Determined first from steady-state measurements, this model permit to predict the evolution of via filling by copper [5], with no information concerning additives incorporation or signature throughout this layer. Indeed, the use of additives introduces also impurities and additives breakdown products during the copper growing. These contaminating agents create defects during the crystallisation like vacancies or dilatation sites of dislocation and limit grain boundary motion and growth during the self-annealing of the copper layer. These phenomena are well deal in the works of Koh et al. [6], and Stangl et al. [7] and particularly well observed with the use of aged bathes. Strangl et al. [7] have demonstrated the incorporation differences between a regular plated copper film and a pulse plated. Favry et al. [5] have also investigated the organic adsorption and inclusion during the copper growing as a function of the additive concentrations or current density. By the combination of electrochemical measurements and SIMS analysis, they have proved an increase of organic inclusion during a regime modification in the plating process. In this present research work, the influence of the applied waveform on the organic additives incorporations is discussed. The work focuses on the organic trace analyses by complementary analysis methods: XPS and SIMS. Firstly, XPS for X-ray photoelectron spectroscopy is a surface (10 nm) sensitive quantitative technique. The resulting spectrum allows the determination of the elementary and chemical composition of the analysed material and the concentration of the different detected elements. The XPS sensitivity is around 0.1 at.%, allowing a chemical cartography of the sample surface composition just after the bath extraction, with the determination of additive adsorptions during an OCP time. Then thanks to an ionic abrasion during the acquisition, a depth analysis is possible, profiling the elements concentration in the bulk material and not only on the surface. Secondly, based on heavy ions bombardment, SIMS for secondary ion mass spectrometry allows a surface composition analysis by sputtering, collecting and mass analysing of ejected secondary ions. This destructive method enables the determination of the elemental, isotopic and molecular composition of a surface or a bulk. The SIMS sensitivity is around 10-4 and 1 at%, but the quantitative analyse is limited. After electrochemical considerations concerning the influence of applied current density waveforms, this research work presents a comparative study of organic incorporations by XPS and SIMS analyses for different copper layers, taking into account the applied waveforms and the bath composition. [1] M. Paunovic at al, Fundamental Considerations in Modern Electroplating 5th edition, edited by M. Schlesinger and M. Paunovic, Wiley edition (2010) [2] M. Datta, Electrochemical processing technologies and their impact in microelectronic packaging in Microelectronic Packaging, edited by M. Datta, T. Osaka and J.W. Schultze, CRC Press edition (2005) [3] J.W. Dini et al, Electrodeposition of copper in Modern Electroplating 5th edition, edited by M. Schlesinger and M. Paunovic, Wiley edition (2010) [4] P.M. Vereecken et al, IBM J. Res. & Dev., 49 1 (2005) [5] E. Favry et al, Electrochimica Acta 53 (2008) 7004-7011 [6] L.T. Koh et al, Microelectronics Journal 33 (2002) 229-234 [7] M. Stangl et al, Microelectronic Engineering 84 (2007) 54-59 Figure 1

For years, the effectiveness of the Damascene process has allowed the semiconductor industry to successfully keep the pace with the aggressive scaling of Moore’s law. However, ever increasing costs have raised the need for alternative technologies bringing advantages in terms of performance gain, time to market and cost competitiveness. Thus, 3D chip stacking technologies have now become a field of extensive research [1-2]. The today improvements are mainly devoted on the current process optimization or the development of new methods to elaborate barriers, seed-layers and fillings compatible with technological advances and the more and more aggressive via aspect-ratio [3]. Generally made in copper by CVD, PVD, ALD or electrografted technics, the seed-layer plays an important role during the via filling. Its principal criteria are to be uniform in term of thickness and continuous to ensure a good electrical contact during the electrochemical process [4]. During the filling process, the wafers and thus the seed-layer are first pre-wetted using DI water or etching solutions and then immerged in the copper bath. Traditionally an idle time is maintain before applying the current density, to favour the species exchange. However, few studies have an interest for the reaction during this idle time and in more general consideration for the copper seed-layer stability on the copper electrolytic bath. Hence, the copper seed-layer is dissolved in the plating bath due to electrochemical reaction with the oxygen dissolved in the bath [5-7]. In this research work, the stability of the seed-layer and its dissolution rate are studied combining in-situ and ex-situ technics (figure 1). Open circuit potential versus time is an interesting approach for this study, allowing direct information on the surface modification during the idle time. Quartz microbalance measurements complete the analyses, giving direct information of weigh lost resulting from dissolution. As ex-situ characterization technics, SEM-EDS, XPS and RBS measurement were performed to characterize the surface states and the chemical composition after the idle time. Several parameters are particularly highlighted as the bath composition and its concentration and the rotation rate of the substrate. [1] T.P. Moffat, D. Wheeler et al., IBM J. Res. Dev., 49, 19 (2005) [2] P.M. Vereecken, R.A. Binstead, H. Deligianni, P.C. Andricacos, IBM J. Res. Dev, 49, 3 (2005) [3] M. Datta, Electrochemical processing technologies and their impact in microelectronic packaging in Microelectronic Packaging, edited by M. Datta, T. Osaka and J.W. Schultze, CRC Press edition (2005) [4] Part II Chip metallization in Microelectronic Packaging, edited by M. Datta, T. Osaka and J.W. Schultze, CRC Press edition (2005) [5] R. Schumacher et al, J. Electroanal. Chem. 219 (1987) 311-3174 [6] K.Y. Electrochimica Acta, 38, 14 (1993) 2121-2127 [7] A.G. Zelinsky et al, Corrosion Science 46 (2004) 1083–1093 Figure 1

Indium phosphide semiconductor (InP) had a direct band gap of 1.34 eV which was suitable for solar spectrum [1]. Moreover, the low surface recombination velocity [2] made InP a relevant candidate for solar-driven water decomposition. However InP suffered from (photo)corrosion due to dissolved oxygen which was still present [2],[3],[4]. At open circuit O2 behaved as a strong oxidized agent [1]. For n- type, the migration of photogenerated holes to the surface induced a competition between water oxidation reaction and self-oxidation of InP which led to passivated layer. In this work, passivated layers on InP were monitored by using anodic treatments as well in aqueous electrolyte as in liquid ammonia (atmospheric pressure, -55°C). In aqueous media, at pH 9, the growth and formation of thin and well covering InPO4 -like anodic films were evidenced [5] , [6]. In liquid ammonia the formation of a phosphazene film was shown on the surface [7],[8]. The perfect coverage of the surface by the phosphazene-like film was revealed by XPS analyses and photoluminescence [9]. The aim of this paper was to quantify the robustness of such passivated layers under photoanodic currents. By using atomic absorption spectroscopy (AAS) techniques, preliminary results showed that the dissolution level of InP was kept to very low or non-detectable level when the surface was recovered by the phosphazene film. The results are very encouraging since it proved that the phosphazene layer avoids InP dissolution that would authorize InP to behave as a photoanode. In this work, we proceed in the same way with InPO4 -like anodic films but also with hybrids layers formed sequentially in both electrolytes [10]. For each case, stability of InP layers are quantified by AAS after anodic photo-electrochemical treatment. [1] T. Wang, J. Gong, Angew. Int. Ed. (2015), 54, 10718 [2] L. GaO, Y. Cui, J. Wang, A. Cavalli, A. Standing, T. T. T. Vu, M. A. Verheijen, J. E. M. Haverkort, E. P. A. M. Bakkers, P. H. L. Notten, Nano Lett. 14 (2014) 3715 [3] A. Etcheberry, J. Gautron, J.L. Sculfort, Appl. Phys. Lett. 46 (8) (1985) 744 [4] Photoelectrochemical Solar Fuel Production, from basic principles to advanced devices, S. Giménez, J. Bisquert (Editors). [5] N. Simon, I. Gerard, C. Mathieu, A. Etcheberry, Electrochim.Acta 47 (16) (2002) 2621. [6] N. Simon, N.C. Quach, A.M. Gonc¸alves, A. Etcheberry, J. Electrochem. Soc. 154 (5) (2007). [7] A-M. Gonçalves, N. Mézailles, C. Mathieu, P. Le Floch, A. Etcheberry, Chem. Mat. 22 (2010) 3114 [8] H. R. Allcock, in Chemistry and Application of polyphosphazenes, (Eds: A. J. Willey and Sons), 2003. [9] F. Proise, F. Pardo, A. -L. Joudrier, C. Njel, J. Alvarez, A. Delamarre, A. -M. Gonçalves, A. Lemaître, N. Bardou, C. Dupuis, A. Etcheberry, J. -F. Guillemoles, J. -L. Pelouard, InP-based nano solar cells, Proc. SPIE 8981, (2014); doi:10.1117/12.2039295. [10] A-M Goncalves, N. Simon, C. Mathieu, A. Etcheberry, C. R. Chimie 11 (2008) 1037

Alkali elements are essential for batteries and solar cells. In batteries, alkali ions are typically intercalated in layered 2D structures, such as transition metal sulfides and selenides, in a reversible fashion 1–4. In solar cells based on Cu(In,Ga)Se2 and Cu2ZnSn(S,Se)4 semiconductors, Na2Sex and K2Sex compounds are sources of extrinsic doping and act as surfactants during or after the growth, ending up as segregations at the grain boundaries, where they are also thought to passivate electronic defects 5–10. Alkali (poly)chalcogenide solutions are also traditionally employed for photoelectrochemical characterization of semiconductors because they offer a wide range of redox states 11,12 and stabilize against photoanodic dissolution 13,14. More recently, they have also been adopted as electrolytes in quantum dot sensitized solar cells 15–17. However, despite their technological relevance, little is known about the single alkali selenide compounds beyond the pioneering works of Zintl and Klemm 18,19 and subsequent structural characterizations 20–24. In this presentation, our recent work on (i) synthesis, (ii) characterization (iii) reactivity to oxygen and (iv) utilization of alkali (poly)selenides is reviewed. The synthesis is performed in liquid ammonia at atmospheric pressure and -55 °C. The progress of the reaction is monitored electrochemically, in order to unveil the kinetics. The decrease of solvated electrons concentration after selenium addition is measured directly by means of cyclic voltammetry at the surface of a Pt microelectrode. The compounds are characterised by X-ray photoelectron, Raman and electronic spectroscopies. Broadening of the Se 3d XPS peaks with the increase of the number of Se atoms in the chain suggests a limited extent of negative charge delocalization among the Se atoms in the solid (poly)selenides. A direct band-gap of 1.83 eV and pseudo-direct band-gap of 2.05 eV are identified for Na2Se and K2Se, respectively. The full width at half maximum of the fluorescence emissions is on the order of 300 meV, and the emission peaks occur at more than 0.5 eV below the respected bandgaps, revealing a pronounced defect-assisted radiative recombination. The extreme sensitivity of the compounds to oxidation is studied and discussed with respect to chalcogenide grain boundary passivation. Additionally, extrinsic doping of epitaxial and polycrystalline CuInSe2 films is investigated by means of a novel gas-phase route 25 that utilizes the synthesised compounds. The resulting films are characterised by photoluminescence spectroscopy, revealing the emergence of a distinct peak at ca. 100 meV higher energy than typical 1 eV peak of Cu-poor CuInSe2, irrespective of the alkali source. Lastly, some of the species likely involved in the alkali gas-phase transport and incorporation in CuInSe2are identified for the first time by Knudsen effusion mass spectrometry. References: 1 W. B. Johnson and W. L. Worrell, Synthetic Metals, 1982, 4, 225–248. 2 S. Kalluri, K. H. Seng, Z. Guo, A. Du, K. Konstantinov, H. K. Liu and S. X. Dou, Scientific Reports, 2015, 5, 11989. 3 X. Yang et al., Nanoscale, 2015, 7, 10198–10203. 4 W.-H. Ryu et al., ACS Nano, 2016, 10, 3257–3266. 5 I. Repins et al., Progress in Photovoltaics: Research and Applications, 2008, 16, 235–239. 6 A. Chirilă et al., Nature Materials, 2013, 12, 1107–1111. 7 P. Jackson et al., phys. stat. sol. (RRL), 2015, 9, 28–31. 8 C. M. Sutter-Fella et al., Chemistry of Materials, 2014, 26, 1420–1425. 9 F. Werner et al., Journal of Applied Physics, 2016, 119, 173103. 10 D. Braunger, D. Hariskos, G. Bilger, U. Rau and H. W. Schock, Thin Solid Films, 2000, 361–362, 161–166. 11 S. Licht, Solar Energy Materials and Solar Cells, 1995, 38, 305–319. 12 S. Licht, Journal of The Electrochemical Society, 1995, 142, 1546. 13 A. B. Ellis et al., J. Am. Chem. Soc., 1976, 98, 6855–6866. 14 F. Forouzan and S. Licht, J. Electrochem. Soc., 1995, 142, 1539–1545. 15 G. Hodes, J. Phys. Chem. C, 2008, 112, 17778–17787. 16 V. Chakrapani et al., J. Am. Chem. Soc., 2011, 133, 9607–9615. 17 L. A. King et al., J. Phys. Chem. C, 2014, 118, 14555–14561. 18 E. Zintl et al., Zeitschrift für Elektrochemie und angewandte physikalische Chemie, 1934, 40, 588–593. 19 W. Klemm et al., Z. Anorg. Allg. Chem., 1939, 241, 281–304. 20 H. Föppl et al., Z. anorg. allg. Chem., 1962, 314, 12–20. 21 P. Böttcher, Z. anorg. allg. Chem., 1977, 432, 167–172. 22 I. Schewe-Miller and P. Böttcher, Zeitschrift für Kristallographie, 1991, 196, 137–151. 23 V. Müller et al., Zeitschrift für Naturforschung B, 1992, 47b, 205–210. 24 J. Getzschmann et al., Zeitschrift für Kristallographie - New Crystal Structures, 1997, 212, 87–87. 25 D. Colombara et al., Manuscript under review, 2016.

The control of surface quality or interface chemistry is one of the key steps for a lot of new advanced industrial products with high added value. This is particularly true in the case of III-V semiconductors (III-Vs) technology for which surface passivation plays a key role in device performances, low temperature and low damage processes having been developed over time to reduce surface effects. With the introduction of III-V’s on large scale Silicon platforms and device scaling to nanometer range, passivation question will be on the front page again [1] . Phosphorus nitride insulating films as passivation layers on InP seem to be very attractive to realize MISFETs with high channel electron mobilities and low interface state densities. This paper presents an original electroless approach of polyphosphazen film formation on InP in liquid ammonia (-55 °C). The chemical surface modifications are closed to those observed by electrochemistry[2],[3]. These findings are protected by a CNRS patent filed by us in 2013 [4] . Moreover, the chemical stability of the surface is also demonstrated by XPS analyses. Indeed no air ageing of the surface is observed after three months of air exposure. The electroless mode is attractive to the industry since it avoids any electrical (potential and current) control at the interface. This paper proposes a novel electroless passivation route in liquid ammonia through a “Polyphosphazen like” film formation which could be adapted to a large scale production since no electrical contact is involved at the interface to provide a full and efficient chemical passivation. [1]“Interface engineering and chemistry of Hf-based high-k dielectrics on III–V substrates”, G. He, X. Chen, Z. Sun, Surface Science Reports 68 (2013) 107. [2] A-M. Gonçalves, N. Mézailles, C. Mathieu, P. Le Floch, A. Etcheberry, « «Fully Protective yet functionalizable monolayer onto InP” Chemistry of Materials.22 (2010) 3114-3120. [3] A-M. Gonçalves, C.Njel, C. Mathieu, D.Aureau, A. Etcheberry. « Phosphazene like film formation on InP in liquid ammonia (223K). Thin Solid Films 538 (2013) 21–24. [4] “Polyphosphazène electroless sur semiconducteurs III-P”, réf CNRS 03024-01 BFF100427CG A. Etcheberry, A-M. Gonçalves, C. Mathieu, J. Vigneron, N. Mézailles, P. Le Floch

ZnO is a wide band gap (3.32 eV) semiconductor with n-type conductivity, which is used in many devices for different applications, including gas sensors, transparent electrodes, and acousto- and electro-optical devices. Recently, the combination of ZnO with boron-doped-diamond (BDD) was used in order to make heterojunctions [1], surface acoustic-wave (SAW) devices [2] and biosensors electrodes [3]. On the majority of presents works, ZnO was deposited on BDD by chemical vapor deposition (CVD) [4], magnetron sputtering [5] or atomic layer deposition (ALD) [6]. The electrodeposition represents an interesting new way for the deposition of ZnO on BDD. The electrodeposition of ZnO has been successfully performed on many conductive substrates [7] but only one time on BDD [8]. This chemical route presents a few advantages such as low cost and possibility to obtain a large range of structures from the dense layer up to well-defined nanostructures. The electrodeposition of a metal oxide such as ZnO consists to a redox process at the electrode surface, to raise the local pH and then to precipitate an oxide or hydroxide phase. Three different precursors can be used to produce hydroxide at electrode surface: hydrogen peroxide, nitrite salt or dissolved oxygen. This last one is used in our work. Dissolved oxygen is reduced, raising the local pH on the conductive diamond electrode and causing the precipitation of zinc hydroxide [9]. Then the dehydration of hydroxide leads to zinc oxide according to the following scheme. O2 + 2 H2O + 4 e- -> 4 OH- (1) Zn2+ + 2 OH- -> Zn(OH)2 (2) Zn(OH)2 -> ZnO + H2O (3) In this work, zinc chloride (ZnCl2) as precursor is used for the first time on BDD and potassium chloride (KCl) as support electrolyte. Our aim was to study the influence of deposition parameters such as zinc precursor concentration, substrate’s surface chemistry or bath temperature, on ZnO morphologies. The resulting ZnO deposits were characterized using X-ray Diffraction (XRD) and scanning electron microscope (SEM). The effect of [Zn2+] is illustrated on figures 1 and 2. As evidencing by XRD (Fig. 1), deposits obtained at 60°C, on as-grown BDD substrate, with 1 mM and 5 mM of Zn2+ correspond to well-crystallized zinc oxide. In both cases, the intensity of diffraction peak (002) indicates a preferential growth to the c-axis. The diffraction peaks from the substrate are labeled with an asterisk. SEM images (fig. 2) confirm the preferential orientation along c-axis. However, ZnO morphologies are completely different by changing Zn2+ concentration. Indeed, with a concentration of 5 mM, micrometric hexagonal grains are obtained while with a concentration ≤1 mM, nanometric pyramids are grown. Finally, our work shows that electrodeposition of ZnO is fully applicable on diamond substrate. And, parameters such as concentration of Zn2+, BDD’s surface chemistry or bath temperature, allow modifying the morphology of deposits. References: [1] Wang, C. X.; Yang, G. W.; Zhang, T. C.; Liu, H. W.; Han, Y. H.; Luo, J. F.; Gao, C. X.; Zou, G. T. Diam. Relat. Mater. 2003, 12, 1548–1552. [2] Luo, J. T.; Zeng, F.; Pan, F.; Li, H. F.; Niu, J. B.; Jia, R.; Liu, M. Appl. Surf. Sci. 2010, 256, 3081–3085. [3] Zhao, Z.; Lei, W.; Zhang, X.; Wang, B.; Jiang, H. Sensors (Basel). 2010, 10, 1216–1231. [4] Sang, D.; Li, H.; Cheng, S. Appl. Surf. Sci. 2011, 258, 333–336. [5] Yu, Q.; Li, H. D.; Zou, G. T. Diam. Relat. Mater. 2011, 20, 351–354. [6] Hikavyy, a.; Clauws, P.; Vanbesien, K.; De Visschere, P.; Williams, O. a.; Daenen, M.; Haenen, K.; Butler, J. E.; Feygelson, T. Diam. Relat. Mater. 2007, 16, 983–986. [7] Lincot, D. Thin Solid Films 2005, 487, 40–48. [8] Chatterjee, A.; Foord, J. Diam. Relat. Mater. 2006, 15, 664–667. [9] Peulon, S.; Lincot, D. 1996, 05, 166–170. Figure 1

The influence of potential was followed by chrono-amperometric experiements and then copper nuclei were physically and chemically analyzed. Thanks to SEM-FEG observations, the nuclei formation and growth has been finely depicted for a large range of applied potentials. Auger investigations have allowed the chemical determination of the nuclei composition and XPS measurements the determination of the global copper layer composition. Analyses by Atomic Absorption Spectroscopy have completed this study by an of a mean copper layer thickness. Thanks to the combination of these different characterization techniques, the influence of the potential on the copper nucleation has been determined. The final aim of this study is to optimize the best plating parameters to obtain a homogeneous, uniform, hiding and adherent copper layer on molybdenum substrates, with a controlled thickness, lower than 100nm.

In liquid ammonia (-55°C), a stable phosphazene like film has been already revealed on InP. This work reports, for the first time, the evidence of different stages of this film on InP in this relevant non aqueous solvent. These results were carried out by a fully coupled approach with cyclic voltammetry, capacitance measurements and XPS analyses.

Our purpose is to deposit a very well-controlled thickness of electroless nickel to form contact lines for silicon solar cells. For that, an accurate understanding of the nickel electroless mechanism is necessary. The main step is to control the beginning of nickel nucleation at very short times. In this study, a commercial bath was employed. First, we have demonstrated that a too oxidized silicon surface block the nickel deposition. Then, the first steps of nucleation were finely studied using both SEM-FEG and XPS analyses. And to finish, a nickel electroless deposition mechanism was proposed.

Molybdenum substrates are mainly used for photovoltaic applications, and specifically as support for absorber deposition, for thin film solar cells. For these applications, one of the technologic key points is the electrodeposition of a copper layer. Furthermore, in the microelectronic domain, OMG Ultra Pure Chemicals (UPC) has developed a specific chemistry (OMniCuTM chemistry), enhanced by organic additives, dedicated to the copper electrodeposition on copper substrates (Damascene and Through Silicon Vias applications) (1,2). The aim of this article is to determine the efficiency of this commercial chemistry to coat a homogeneous, uniform, hiding and adherent copper layer on molybdenum substrates, with a monitoring thickness, lower than 100nm.

This manuscript presents an original XPS study of mesoporous p-type silicon samples. The use of X-ray photoelectron spectroscopy helps in understanding the different chemical environments of silicon atoms generated during porous-silicon formation and their evolution in time. Thus, the oxidation state and the contamination can be precisely known and controlled. The loss of the flatness of the surface is followed by the modification of the plasmon peaks and the morphology obtained is monitored by SEM.

Soft chemical processes such as screen printing, sol-gel, electrochemical or chemical bath-deposition, enable low cost and easily scalable processes for thin layer solar cells. For the CIGSe based solar cells, with the standard Mo/Cu(In,Ga)Se2/CdS/i-ZnO/n+-ZnO structure, all layers of the p-n heterojunction and the transparent conductive oxide window can be grown by aqueous solution processes. Based on our previous work and the literature it was shown that cells with electrodeposited Cu(In,Ga)(S,Se)2-based absorbers can present conversion efficiencies up to 13.8 %. When ZnO:Cl window layer is deposited by electrodeposition with a coevaporated CIGS-based absorber, efficiency above 15% can be reached. In this work we will mainly focus on the application of this soft deposition technique for the preparation of the buffer layer. Electrochemical investigation of In-S layers is carried out. XPS analysis show the presence of both S and O. Conversion efficiencies of complete devices up to 9.5 % are obtained.

A protective monolayer "phosphazene" like film on InP has been already successfully evidenced by a controlled anodic process in liquid ammonia (NH3 Liq.). In order to understand the formation mechanism of this passivating film, in-situ comparisons of electrochemical interface responses are performed as well on InP as GaAs. Current-voltage and interfacial capacity measurements are explored on both semiconductors. Similarities of electrochemical behaviours between these two semiconductors are striking. But contrary to InP, the initial surface state is electrochemically recovered by an in-situ cathodic treatment on GaAs. This absolute reversible electrochemical behaviour is definitively significant for GaAs.

Filling of Through-Silicon Vias (TSV) is governed, similarly to the well-known Damascene process, by several plating parameters and filling mechanisms. However, due to their larger dimensions, the diffusion of species into this particular type of via is becoming the main obstacle for the superfilling phenomenon generation. It is thus of prime importance to adjust the electroplating process parameters in order to take into account the TSV specificity. In this paper, the impact on the filling mechanism of four plating parameters - plating time, current density, wafer rotation and accelerator concentration - was investigated using two distinct methodologies: a fundamental analysis of the plating parameters on a gold planar electrode and their direct impact on the via filling.

In this paper we report, for the first time, a galvanostatic control in 1M HCl aqueous solution onto a InP surface that had been previously entirely nitrogenated. The surface nitrogenation required an anodic electrochemical treatment in acidic liquid ammonia (NH3 liq.). An homogeneous covering film with " P-N " terminations was obtained onto the InP surface properly deoxidized. This thin film is notable for its lack of air ageing and also for its chemical stability in HCl. A galvanostatic control at different current density (10 mA.cm-2, 100mA.cm-2 and 300 mA.cm-2) in 1M HCl was used to identify the effect of this " P-N " terminations on the porosification process. In comparison to a bare surface, a contrasted evolution of the interfacial potential was clearly observed for low current density. As expected, a crystal oriented pore morphology (CO) is obtained onto a bare InP surface whereas "oscillating" current line oriented pore morphology (CLO) is kept onto a surface entirely recovered by a " P-N " film. The presence of the nitrogenated film can therefore be significant on the pore growth evolution in 1M HCl aqueous solution.

A new route of InP passivation was clearly ruled by an anodic electrochemical process in liquid ammonia as a relevant nonaqueous solvent. The coupling approach using both galvanostatic control and XPS analyses evidenced that this first protective film upon InP was definitely associated to one coating monolayer of general formula [ (H2N)P=N-]n. The electrochemical stability of the complete covering thin film was tested during hydrogen evolution in HCl aqueous solution. In spite of the high cathodic charge, unexpectedly no cathodic decomposition of the coated InP semiconductor was detected by XPS analyses. The electrical conductivity of this protective film and its electrochemical stability under negative overvoltage was then reported. In contrast to a bare InP, the electroless immersion of the protected InP in K2PtCl4 solution did not lead to the formation of metallic Pt cluster onto the substrate. On the contrary, the presence of amino groups in the film offered a unique opportunity for Pt (+II) grafting. A hybrid inorganic structure was proposed from the N/Pt atomic surface ratio, slightly higher than 2.0. The stability of the covered InP substrate as well as the ultra thin passivated film upon functionalization were also demonstrated from XPS analyses.

In this paper, we report an unexpected dissolution phenomenon at n-InP electrodes during porous etchiong in 1M HCl under potentiostatic consitions (5V vs. Ag/AgCl). Although the scanning electron microscope examinations show the typical tubular and regular pores that exhibit a depth proportional to the anodic charge, further chemical analyses of the dissolved material performed after the pores formation by atomic absorption spectroscopy reveal a dual dissolution process. An electrochemical etching process exhibiting a valence of 8 is evidenced. Howeverit is surprisingly coupled to a chemical dissolution that correspnds to the half the dissolved InP.

Depending on the applied electrochemical parameters, various anodic films can be grown onto InP in aqueous media. In borate buffer at pH = 9, the formation of homogeneous and thin InPO4-like films occurred when anodic treatment is performed with a low current density, while with higher imposed current configuration porous and thicker layers are grown. In this work, AFM measurements coupled with XPS analyses and capacitance-voltage investigations have been used to study the relationship between morphology, composition and electrical properties of different anodic oxide films performed on InP surface, with current densities ranging from 1 mA.cm-2 to 100 mA.cm-2.

Porous InP layer are obtained in HCl. Kinetic aspect associated to the growth are considered. XPS shows that specific Cl chemistry seems to be linked to the porous layer growth. Dosage of dissolution product by atomic spectroscopy demonstrated that a very interesting delay phenomenon for the observation of the final In concentration in solution is observed. This phenomenon can be understood as the consequence of the presence of a confined medium at the time of the porous layer inside the pore. This situation seems to generate a dissolution process with strong electron injection.

The electrodeposition of multinary compounds is of key importance for future technologies. Cu-In-Se system represents a case example, with a great potential in the field of photovoltaics. At IRDEP we achieved efficiencies up to 11.5% based on electrodeposited Cu-In-Se material. In the present paper, nucleation and growth steps of CuInSe2 are studied by polarization, impedance spectroscopy and SEM. The bulk composition is measured by x-ray fluorescence. The chemical environment of Cu, Se and In and the surface composition are analysed by XPS. A quasi-instantaneous 3D nucleation is observed. Binary Cu-Se nuclei are first obtained, which induce the incorporation of indium leading to CuInSe2. With increasing growth, secondary phases and Se(0) are obtained. At the threshold of potential corresponding to indium incorporation, a blocking adsorbate is formed, which is reduced at higher cathodic overvoltage into Cu-In-Se compound. These results are discussed in terms of general behavior of multinary compound deposition.

The anodic electrochemical treatment in liquid ammonia of InP (n- and p-) leads to a simultaneous oxidation of both semiconductor and solvent. The oxidation of InP takes place through a passivating layer revealed as well by electrochemistry as by XPS. The chemical composition of InP surface is therefore gradually modified according the anodic charge. The composition of the film layer is supposed to be close to HN=P-NH2, a promising compound stable as well in liquid ammonia as in water.

Growth, formation and stability of anodic oxides obtained on n- InP were investigated by coupling electrochemical methods and XPS analyses. Using transient photocurrent and Mott-Schottky behavior, as in situ probes, we have pointed out several steps for InP oxidation process, related to a gradual oxide coverage evidenced by XPS characterizations. The current density chosen to perform the oxidation governs the resulting chemical composition, the texture and the electrical properties of the oxide. Low applied current densities lead to the formation of thin (around few nanometers) and homogeneous films, while high current densities induce thicker and less homogeneous oxides. The resulting electrical properties also strongly depend on the anodization mode. The thinner homogeneous layers, grown with low current densities, exhibit the best passivating features.