The information is prepared on the basis of data from the information-analytical system RSF, informative part is represented in the author's edition. All rights belong to the authors, the use or reprinting of materials is permitted only with the prior consent of the authors.

Project titlePostsynthetic doping of nanosilicon: a route for obtaining non-cubic modifications of silicon, p-n junctions and creation of optoelectronic devices.

AffiliationFederal State Budgetary Educational Institution of Higher Education Lomonosov Moscow State University,

Research area 03 - CHEMISTRY AND MATERIAL SCIENCES, 03-601 - Chemistry of new inorganic functional and nano-dimensional materials

Keywordsnanosilicon, colloidal semiconductors nanocrystals, doping, nanoparticles films, conductivity, p-n junction

Annotation
The project aims at the control of silicon nanoparticles’ (nc-Si) properties by doping them using the process of gas-phase diffusion of dopant precursor. We have previously obtained boron- and phosphorus-doped nc-Si by using the aforementioned method. This year (2021), we discovered that using sulfur as a dopant in certain modes of this method leads to recrystallization of nc-Si and formation of faceted nano- and micro-particles with a predominance of hexagonal form of symmetry. All of these nano- and micro-particles consist of silicon with addition of sulfur (<1%). At the same time, IR-spectra change drastically with appearance of direct band gap that is smaller than the indirect band gap of bulk silicon. Development of a method for obtaining the direct-gap modification of silicon is an important task, because it opens the way to obtaining of efficient interband luminescence of silicon, which until now has only been observed under the quantum confinement conditions (e.g., in quantum dots or silicon nanowires, where the difference between direct and indirect bandgaps gets erased, but the band gap energy also increases and luminescence shifts from IR to visible part of the spectrum). Based on the entirety of the obtained data, we presume that we have obtained a mixture of sulfur-doped silicon’s polymorph modifications intermediate between cubic (normal, Si-I) and hexagonal (Si-IV with the hexagonal diamond structure, lonsdaleite) forms of silicon. In the Project, it is planned to determine the reason of such changes and find the right conditions for synthesis or isolation of pure non-cubic phase silicon. To accomplish these goals, a set of research methods will be used: HRTEM, electron diffraction, X-ray diffraction, X-ray fluorescence, absorption spectroscopy, Raman spectroscopy and luminescence spectroscopy. Furthermore, a study will be conducted on the effect of using other chalcogens as dopants. To study doped nc-Si and obtain films and structures from them a sol will be prepared from synthesized particles, which gives a significant advantage for working with nanoparticles and allows creating films with unlimited surface area by using cheap printing methods. This is promising, for example, for the creation of solar panels. By using layer etching of nanoparticles and composition analysis after each stage of etching the radial distribution of impurities in nc-Si will be defined. Films and structures will be printed from nc-Si sols; structures will be made by combining nc-Si films with donor (B) and acceptor (P, S, Se, Te) impurities to create p-n junctions. Films made of cubic and non-cubic nc-Si will be used to create heterostructures. For the first time the study of electrical properties of heterostructures of cubic and non-cubic nc-Si will be conducted, along with the temperature dependence of these properties. Moreover, photoelectrical properties of such structures and p-n junctions made from nc-Si will be studied and the possibility of using them as photodetectors and basis for solar panels. Recrystallization of nc-Si because of sulfur doping not only changes nanoparticles’ properties, it also leads to the formation of faceted long (>100 µm) and thin rods with hexagonal section that can be easily observed with optical microscope. These rods have lower sulfur content than nanoparticles. In the Project, it is planned to create a micromanipulator to retrieve these rods from the synthesized sampled and study them afterwards. Raman spectroscopy for different parts of the rod and measurement of electrical properties will be performed. By diffusion of acceptor impurity from the melt, single p-n junctions will be formed on the rod; their electrical and photoelectrical properties will be studied.

Expected results
A new method for obtaining the particles of hexagonal silicon Si-IV and/or intermediate polymorph modifications between Si-IV and Si-I will be developed. Unlike the already known methods, a new method is simple and easily scalable, which makes it easy for other researches to obtain, use and study these new phases of silicon. Conditions for synthesis and/or isolation of particles of silicon’s direct band gap modification in pure form will be determined, which will provide a material for a new route to silicon light-emitting devices. Optical, electrical and photoelectrical properties of aforementioned phases and electrical and photoelectrical properties of heterostructures made of these phases in combination with cubic silicon will be determined. Conditions for reproducible printing of p-n junctions from nanosilicon with the use of gas dynamic spraying will be found. Electrical and photoelectrical properties of printed samples will be studied. This will contribute to the development of printed silicon photovoltaic devices. Conditions for obtaining microdiodes based on the silicon microrods will be developed and their electrical and photoelectrical properties will be determined, which can be useful for electronics.

Annotation of the results obtained in 2023
During the next stage of the project, the apparatus for producing amorphous NS by decomposition of SiH4 in glow discharge was constructed. The optimal conditions for producing amorphous NS were determined. 21 syntheses on NS was conducted on this apparatus. In 19 of them, the product is fully amorphous. Flow rates of SiH4, Ar, H2 were varied; five syntheses with addition of oxygen-containing gases, such as air, O2 and SO2, were conducted. Addition of Ar lowers the power of discharge, but does not affect the throughput or conversion rate of SiH4. Increase in flow rate of H2 and SiH4 increases the discharge power. Conversion rate of SiH4 reaches 94%. For NS samples that were synthesized without the oxidant gas, the surface passivation is mostly hydridic, as shows by IR-spectroscopy data. Based on SEM results, particles are of round shape, grouped in agglomerates and have the mean size of 20-35 nm. According to IR spectroscopy data, nanoparticles obtained from the plasma-chemical decomposition of SiH4 contain dozens of at.% of H, which greatly influences the recrystallization. Spectroscopic analysis is not suitable for samples synthesized with SO2, since it is impossible to account for amount of H bonded to S. Vacuum annealing leads to breaking of up to 95% of Si-H bonds at 570C°. Series of doping annealings of new NS in S, Se and Te vapors. Annealing in S vapors results in growth of rods at varying temperatures. Addition of about 15 at.% hydrogen to NS synthesis increases the yield of rods, and removal of hydrogen reduces it. In case of oxidized NS, the situation is opposite – addition of H2 reduces the yield of rods by two orders of magnitude. Approximation of H and O atoms content in synthesis, made with use of IR spectroscopy, shows that optimal ratio of H/O is in 1.5-3 range. Optimal annealing temperature is 850C°, as increasing or lowering the temperature reduces the yield and the size of the rods. Annealing of samples with different crystallinity shows that it is not a dominating factor for the microrod growth. It was found out that continuous annealing with S and Se (675C°, 15 hr) leads to initial crystallization. Annealings at 725C° for 3 days in vapors of S demonstrate that in case of low temperature and long synthesis time hydrogen increases not the yield of rods, but their size. Position of new absorption feature maximum shifts toward the red part of the spectrum as the annealing temperature grows. Maxima for all the studied samples, except for the one obtained at 675C°, belong to 2.8-3 eV range and do not correlate with the annealing temperature. The upper limit of the band is 3.3-3.4 eV. Micromanipulator for isolation of singular rods from samples was constructed. By using this device, some of the rods were positioned on the ends of fibers with FC°-connector to obtain the absorption spectra of single rods. Picosecond supercontinuum was used as a light source. Aside from the absorption edge, the peak at 1.08-1.09 eV was observed on spectra. The method for doping half of the microrods with Al for manufacturing of p-n junction was developed, using vacuum-thermal sputtering on the half of the rod. High quality p-n junctions with photoresponse to microscope illumination were produced. I-V curves were measured in the voltage range from -14 to +14 V and approximated with the model of two opposed Schottky diodes. I-V curves of rods held in microtweezers and dipped into a Ga drop are asymmetrical and possess a large region with low conductivity, low ideality factors and high-energy barriers. I-V curves of rods with Ti electrodes are symmetrical and have a bend in range of +- 500 mV after which they become mostly linear; the height of silicon-titan border barrier is 0.3 eV and the ideality factor close to 1. Heating samples reduces the aforementioned bend but the resistance of the rod increases since all the dopant is already electrically activated. Barrier energy increases with temperature; this is attributed to the existence of interface layer through which the charge carriers are tunneling. After annealing the rod in the hydrogen atmosphere at 300C° for 30 minutes the bend on the I-V curve completely disappears, which means that it was caused by the traps on the rod’s surface. Meanwhile, barrier energies does not change and the ideality factor gets closer to 1. Photoconductivity spectra of Si microrod with Ti electrodes was obtained at 140K temperature under irradiation with a xenon lamp and an applied voltage of 0.1-1 V. Photoconductivity slowly decreases as the photon energy lowers from 2.8 to 1.5 eV, after which the decrease is faster. The test annealing with phosphorus was carried out after moving to hydrogenated nanoparticles. The formation of nanoparticles of a new Si3P4 compound with the defective zinc blende structure and the cell parameter of 5.04 was observed at the annealing range of 400-900C°. The preparation of solid solutions based on Si3P4 will make it possible to control the concentration of Si vacancies and consequently create the narrow-gap semiconductor.

Publications

1. Bubenov S.S., Vinokurov A.A., Yudin I.V., Popelensky V.M., Kononov N.N., Kuznetsova T.A., Dorofeev S.G. High-Throughput Low Frequency Reactor for Non-Thermal Plasma Synthesis of Amorphous Silicon Nanoparticles Silicon, 2023 г. (year - 2023) https://doi.org/10.1007/s12633-023-02797-6

2. Nikiforova P.K., Bubenos S.S., Platonov V.B., Kumskov A.S., Kononov N.N., Kuznetsova T.A., Dorofeev S.G. Isolation of cubic Si3P4 in the form of nanocrystals Beilstein Journal of Nanotechnology, т. 14, 971-979 с. (year - 2023) https://doi.org/10.3762/bjnano.14.80

3. Дорофеев С.Г., Винокуров А.А., Бубенов С.С., Попеленский В.М., Сапков И.В., Кононов Н.Н., Кузнецова Т.А. ПОСТСИНТЕТИЧЕСКОЕ ЛЕГИРОВАНИЕ НАНОКРЕМНИЯ: СТРУКТУРА, ОПТИЧЕСКИЕ И ЭЛЕКТРИЧЕСКИЕ СВОЙСТВА VIII Всероссийская конференция по наноматериалам "НАНО2023", с. 25-26 (year - 2023)

Annotation of the results obtained in 2022
In the course of work, a series of doping annealings of nanosilicon were carried out in sulfur vapor in evacuated quartz ampoules. The amount of sulfur put in ampoules varied in range of 5-30 at.%, annealing temperature was in range of 500-900C and the annealing time varied from 1 to 5 h. SEM was used to determine the optimal conditions for growing rods with hexagonal sections: 20 at.%S, 850C, 3 h. Diameter of the rods varies in range of 5-20 μm, length from dozens to hunderds of μm. Lowering sulfur concentration or annealing time leads to lower amount of recrystallized materials, e.g. rods. Increase in annealing time does not change the result but increase in sulfur amount increases the concentration of rods in druses, while rods’ surface becomes more smooth and wavy. The lowest temperature at which the recrystallization was observed is 800C. It is also the lowest temperature at which the rods form, although their diameter decreases to 1-2 μm and length to 3-60 μm, rods become sporadic. Moreover, test annealing with selenium and tellurium (20 at.% both) were conducted at 850C for 4.5h. Going from sulfur to selenium results in absence of rods, but faceted crystallites with size of 0.1-1 μm are observed, their shape resembles truncated octahedral. In case of annealing with tellurium, same faceted crystallites are observed, but unlike the selenium samples, there are also crystallites in shape of parallelepipeds and tetrahedral microrods 30-50 μm long and 2-5 μm wide and also druses. After test annealings, another series of them were conducted, using selenium, tellurium and also mixtures of two or all three chalcogens. Mixtures with sulfur and tellurium result in formation of hexahedral and tetrahedral rods correspondingly, but in some cases, structures that are more convoluted were observed. Data was obtained that indirectly indicates that micron rods grow from amorphous silicon, and not nanocrystals. It is possible that in future one will be able to decrease the recrystallization temperature by using pure amorphous nanosilicon. Crystal structure of samples was studied by electronic diffraction and compared to data from ICSD and ICDD databases and computational data. In case of SiTe1 sample (20at. %, 850С, 4.5h), all interplanar distances within the accuracy of determination correspond to those in the cubic phase. Moreover, electronic diffraction pictures are perfectly indexed in accordance with cubic structure. SiS8 (20 at. %, 850С, 5h) sample is characterized by a greater variety of interplanar distances, among which there are specific for hexagonal phase "satellites" of the reflection of the cubic phase at 3.14 Å and other reflections corresponding to the hexagonal phases Si-IV, 4H-Si, 6H-Si. In SiSe1 (16.7 at.%, 850С, 4.5h) and SiS6 (20 at.%, 825С, 4.5h) samples most of the studied crystallites belong to the cubic phase, but in some cases the diffraction pattern could only be attributed to the hexagonal structure of 4H-Si. The found minimal temperature of nanosilicon recrystallization in sulfur vapors (800C) is still higher that the decomposition temperature of hexagonal Si-IV (750C), therefore, silicon phase mixtures form. Therefore, the separation of particles in sols/suspensions by various methods was carried out. The effective way of separation was found – filtration through the stainless steel nets with gap size of 30 μm. To isolate the hexahedral rods in sulfur samples, two nets are sufficient for efficient removal, to isolate smaller tetrahedral rods using four nets together. Removing rods from the net can be done with the ultrasonic bath. The content of chalcogens was determined using X-ray fluorescence analysis with total external reflection on sapphire substrates (TXRF). One cycle of etching reduces the average radius of silicon particles by 2 nm. Content analysis was done after each cycle. Concentration of sulfur before etching is in range from 1 to 5 at.% and does not correlate with synthesis conditions or sulfur content after etching. After etching out the surface layer sulfur content decreases by an order of magnitude and amount to 0.1 to 0.6 at.%, which still exceeds the equilibrium solubility of sulfur in silicon (~2*10^-4 at.%). After this, concentration of sulfur remains the same after etching out at least two more layers, which indicates the uniform distribution of sulfur along the particle radius below the surface oxide layer. Higher annealing time leads to the increased amount of sulfur in etched samples. Concentration of heavier chalcogens until etching is in range of 0.7-9 (Se) and 0.3-7 (Te) at.% and also do not correlate with synthesis conditions. In mixed systems, concertations of sulfur and selenium always end up much higher than the one of tellurium, and if equal amount of sulfur and selenium are put in ampoule, the resulting concentration of sulfur is higher. After the first etching in HF, the concentration of chalcogens, with the exception of two samples (Se 20 at.% and Te 20 at.%, 850C, 5h) increases significantly, moreover, formless holes on the surface of crystallites, which resemble the holes on cheese, were found by means of SEM. In case of tellurium, particles with most of the volume etched out were observed. Such “cheese holes” seem to be the result of the inclusion dissolution. In sulfur samples, the only difference between SEM images is the gradual rounding of the edges. For samples with sulfur and selenium, the absorption spectra have a pronounced absorption band with a maximum at 1.8–2.8 eV. These bands are well described by the Drude equation, which determines absorption on free charge carriers; however, the calculated free-electron concentrations were too high—they were several times higher than the concentrations of chalcogens in the samples. It is possible that this band is related to the formation of miniband and transitions of electrons from it to Г-point of the conduction band after the light absorption. One could expect an energy of 2.28–2.8 eV for such transition, which was in good agreement with the observed positions of the maxima. The absorbance maximum of ~2 eV was also predicted for a supersaturated chalcogen solution in silicon, but its maxima is at lower energies than the one of silicon bandgap, i.e. it belong to IR region. For electrical measurements, particles were deposited on chips by dropping sols in methanol. Moreover, one heterostructure (nc-Si/SiS6f) was deposited by the gas-dynamic method from acetonitrile sols, consisting of a layer of initial nanosilicon in contact with one electrode and an overlapping layer of nanosilicon, recrystallized in sulfur vapor (20 at.%, 825C, 1 h), from which the rods were previously removed by filtration, in contact with another electrode. The measurements were carried out in an ultrahigh vacuum in the chamber of an Auger spectrometer. Before measurements, the films were heated from 1 to 6 minutes at 700–900C until the maximum conductivity was obtained. The temperature dependences of the conductivity of all the studied samples have an activation character. The conductivity activation energies of recrystallized samples are an order of magnitude lower than the ones in samples without recrystallization. The I–V curves of the sample Se 20 at.%, 850C, 5 h are linear, the I–V curves of the other films are approximated by a power function, which indicates the limitation of the current by the space charge. Exponential components are added to the I–V curves of films with rods, which result in the asymmetry of the I–V curves, are associated with the formation of barriers, and charge tunneling. Asymmetry can be caused by an asymmetric accumulation of rods near the electrodes, it is advisable in the future to separate the rods before depositing the films and apply them separately. Films made of recrystallized in sulfur and selenium vapor particles are characterized by long-term retention of high conductivity when exposed to air. The I–V curves of the aforementioned heterostructure at various temperatures are close to linear functions, but above 300C spontaneous jumps are observed between several different I–V curves, which become more frequent with increase in temperature.

Publications

1. Popelensky V.M., Chernysheva G.S., Kononov N.N., Bubenov S.S., Vinokurov A.A., Dorofeev S.G. Doping of nanocrystalline silicon with sulfur by gas-phase diffusion Inorganic Chemistry Communications, 141, 109602, 1-7 (year - 2022) https://doi.org/10.1016/j.inoche.2022.109602

2. Vinokurov A.A., Popelensky V.M., Bubenov S.S., Kononov N.N., Cherednichenko K.A., Kuznetsova T.A., Dorofeev S.G. Recrystallization of Si Nanoparticles in Presence of Chalcogens: Improved Electrical and Optical Properties Materials, 15, 24, 8842 (year - 2022) https://doi.org/10.3390/ma15248842

3. Popelensky V.M., Dorofeev S.G., Vinokurov A.A., Bubenov S.S., Kononov N.N., Cherednichenko K.A. Рекристаллизация кремния в присутствии халькогенов Красновидово, 1, 111-112 (year - 2022)