CC BY
95
Ca1-x-yYbxPryF2+x+y solid solution powders as a promising materials for crystalline silicon solar energetics
S.V. Kuznetsov1, O. A. Morozov2, V.G. Gorieva2, M.N. Mayakova1, M.A. Marisov2, V.V. Voronov1, A.D. Yapryntsev3, V.K. Ivanov3, E.I. Madirov2, A. S. Nizamutdinov2, V. V. Semashko2, P.P. Fedorov1
1Prokhorov General Physics Institute, RAS, 38 Vavilova str., Moscow, 119991 Russia 2 Kazan Federal University, 18 Kremljovskaya, Kazan, 420008 Russia 3Kurnakov Institute of General and Inorganic Chemistry, RAS, 31 Leninsky pr., Moscow,119991 Russia
PACS 42.70.-a, 81.20.Fw DOI 10.17586/2220-8054-2018-9-2-259-265
We have synthesized single-phase powders of Cai—x—yYbxPryF2+x+y solid solutions with an average particle size of about 35 nm by co-precipitation from aqueous nitrate solutions. After annealing at 600 °C during 1 hour, the particle size was increased up to 150-200 nm. Individual luminescence bands of praseodymium are distinguishable in the luminescence spectrum. The intensity of the luminescence of ytterbium increased by a factor of 1000 in comparison with the unannealed samples. The highest luminescence intensity of ytterbium was detected for the Ca0.9495Yb0.0500Pr0.0005F2.0505-
Keywords: calcium fluoride, rare earths fluorides, chemical synthesis, nanoparticles, down-conversion luminophores, solar cells.
Received: 12 March 2018 Revised: 21 March 2018
1. Introduction
Solar energy is one of the most abundant energy sources [1,2] and crystalline silicon solar panels became a quite traditional tool for its utilization. The use of silicon has its major advantage in its chemical availability, as well as well-developed manufacturing and recycling technology, including recycling of the rare-earth-containing components. Other solar panels, such as GaAs, CdTe, Cu(In,Ga)Se2 multiple layer devices, can also be utilized for the solar energy conversion [1], but their manufacturing and recycling is more complicated and expensive due to their toxicity. In other words, the use of GaAs, CdTe, and/or Cu(In,Ga)Se2 doesn't obey "green chemistry" principles. Unfortunately, silicon solar panels possess relatively low coefficients for converting solar power to electric energy (under 25 % even for the best samples [1,2]). This is why silicon solar cells are frequently modified by luminescent coatings. These phosphor films can transform solar energy from the spectral region of photon nonsusceptibility (ultra-violet and visible range) of crystalline silicon solar cells to the region of their photoresponse (infra-red range).
Down-conversion luminescence, especially quantum cutting mechanism is one of the promising pathways of increasing crystalline silicon solar cell efficiency [1,3]. Quantum cutting allows for splitting each energy UV photon to two infra-red photons [4]. When Pr is excited at a wavelength of 185 nm, a quantum conversion efficiency of more than 100 % was obtained.
Later publications concerning down-conversion phenomenon also demonstrated quantum yields of more than 100 % for several materials [1].
Fluorite-type solid solutions M1-xRxF2+x (M = Ca, Sr, Ba, R - rare earth) exists in all MF2-RF3 systems [5]. Single-phase solid solutions in the systems Ca1-xRxF2+x are formed up to x = 0.2 [6], Sr1-xRxF2+x x = 0.6 [7]. For systems based on BaF2-RF3, single-phase solid solutions with low dopant content cannot be synthesized [8-10]. Effective up- and down-conversion luminescence is carried out through the interaction of the cations introduced in the matrix. Calcium fluoride matrix possesses certain technological advantages, for it has crystal lattice parameters close to that of silicon, so CaF2-based films can easier be attached to silicon substrates by epitaxy [11-13].
Thus it is of great interest to study CaF2 matrix with Pr3+-Yb3+ ions pair doped due to the existence of resonant transitions 3Po- 1G4 of Pr3+ and 2F7/2 - 2F5/2-transitions of Yb3+ ions [14] which is promising for quantum cutting by means of energy transfer from Pr3+ ions to Yb3+ ions. Therefore, the goal of the present study was the investigation of chemical and physical properties of Ca1-x-yYbxPryF2+x+y solid solutions for determining the chemical compositions suitable for increasing the efficiency of crystalline silicon solar panels.
2. Experimental
We used 99.99 wt% pure ytterbium and praseodymium nitrate hexahydrates, calcium nitrate tetrahydrate (all reagents were manufactured by LANHIT, Russia), 99.99 wt% pure 46-49 wt% aqueous hydrogen fluoride (SigmaTec, Russia) and double distilled water as the starting materials, without further purification. Fluoritetype CaF2:Yb:Pr solid solution precipitates were synthesized according to our previously published protocol [15]. 0.5 M calcium, praseodymium, ytterbium nitrate aqueous solutions were added dropwise to the two-fold excess of 5.0 M aqueous HF under vigorous magnetic stirring. The process of co-precipitation is based on the fact that the solubility of alkaline-earth and rare-earth fluorides in water is low (about 10-5-10-4 M) [16,17]. After CaF2:Yb:Pr precipitate was formed, the mother solution was decanted. Precipitates formed were washed with double distilled water until negative nitrate anion test with diphenylamine. This qualitative reaction is based on the oxidation of diphenylamine, which has a sensitivity to nitrate ions. Colorless diphenyl benzidine is first formed, which upon further oxidation is converted into a diphenyl benzidine quinoid derivative having a blue color. Obtained powders were dried in air at 45 °C for 5 hours and annealed in platinum crucibles (also in air) at 600 °C for 1 hour.
All samples were analyzed using a Bruker D8 Advanced (Cu Ka radiation) diffractometer. Particle sizes and morphology were determined by a Carl Zeiss NVision 40 scanning electron microscope. Differential thermal analysis (DTA) and differential thermal gravimetric analysis (DTG) was performed on a MOM Q-1500 D derivato-graph with platinum crucible on air (rate of heating/cooling - 10 degrees/min). Fluorescence of the samples was excited by a 443 nm cw laser diode with a pumping power density of 152.8 W/cm2. The luminescence spectra were recorded with Stellarnet EPP 2000 spectrometer using a 5 mm filter (ZhS-16) and corrected for the spectral sensitivity of spectrometer. The luminescence kinetics were registered with the use of scanning monochromator with 600 l/mm diffraction grating and under pulsed excitation with 444 nm laser light from OPO system Lotis TII LT2211. The luminescence light was converted by photomultiplier tube FEU62 and digitalized by BORDO oscilloscope.
3. Results
Typical X-ray diffraction pattern of CaF2:Yb:Pr unannealing sample is presented in Fig. 1a. All observed reflections are in a good agreement with JCPDS Card #77-2245 data, i.e., XRD results confirmed that synthesized samples contain only one fluorite-type phase. According to our previous experimental data for similar samples, the luminescence intensity of unannealed samples was low. One of the main reasons for this is the presence of hydroxyl anions in the crystal lattice, which result in quenching of luminescence. We chose the annealing temperature based on DTA/DTG study (Fig. 2). It's clear, that the dehydration proceeds in several stages and finishes at about 500-550 °C. Weight loss upon was 3.7 wt%. Based on this we chose the 600 °C and 1 hour as annealing conditions. Fig. 1b represents XRD data for annealed sample. Lines in the X-ray pattern of the sample dried at 45 °C (Fig. 1a), were much broader than the lines for the sample annealed at 600 °C (Fig. 1b). Annealing leads to an increased crystallite size. We calculated the unit cell parameters for the samples before and after annealing (Table 1). Annealing at 600 °C resulted in an anticipated small decrease in the unit cell parameter value in comparison with the data for the same composition samples dried at 45 °C: see also [18] for the similar dehydration effect studies. The weight loss upon for dehydration in Table 1 and in TG data in DTA/DTG study were similar.
Typical scanning electron microscopy (SEM) images of the samples dried at 45 °C are presented in Fig. 3a,b,c. SEM results confirm that particle sizes vary from 20 nm to 80 nm (average particle size is about 35 nm). Fig. 3b presents SEM image taken in EBSD regime, which allow for atomic contrast regime. The uniform color of the image on Fig. 3b confirms the single phase composition of the powders. Annealing at 600 ° C caused the increase in the particle sizes up to ^ 150-200 nm (Fig. 3d,e). Evaluation of chemical composition based on EDX analysis is summarized in Table 2. The content of praseodymium in the solid solutions is close to it's nominal content in initial aqueous solution. The content of ytterbium fluoride in solid solutions is somewhat higher than in aqueous solution. The EDX composition determination error is 0.5 mol.%. This means that the concentration of praseodymium in the samples may also be greater than in the case of ytterbium. The mean value of ytterbium distribution coefficient between solid and aqueous phases is about 1.45. This means that the crystallization proceeds in an incongruent manner.
Characteristic luminescence spectra of the samples before and after annealing are presented in Fig. 4a. The luminescence intensity of the unannealed sample appeared to be very low as it has been predicted and intensity of Yb3+ luminescence at the wavelengths around 1 micrometer was almost at the level of noise. In the visible spectral range, we have registered broad luminescence band located in the region of praseodymium luminescence. After annealing/dehydration stage, the luminescence spectra have changed. In the visible spectral range, characteristic
Fig. 1. XRD patterns of CaF2:Yb (5.0mol%):Pr (0.1 mol%) powder: dried at 45 0C (a), and annealed at 600 0 C (b)
Table 1. XRD results for CaF2:Yb:Pr solid solutions
No* Nominal doping Pr/Yb, mol. % After drying at 45 0 C After annealing at 600 0 C Weight loss upon, wt.%
a, A a, A
1 0.01/7.0 5.4757(7) 5.4700(6) 4.5
2 0.01/10.0 5.463(2) 5.460(3) 3.8
3 0.01/12.0 5.4799(4) 5.4774(4) 4.2
4 0.05/5.0 5.461(2) 5.464(1) 4.3
5 0.05/7.0 5.476(1) 5.4694(7) 4.1
6 0.05/10 5.4741(8) 5.47643(8) 4.7
7 0.05/12.0 5.4757(7) 5.472(1) 4.1
8 0.05/20.0 5.4826(7) 5.4850(8) 4.8
9 0.1/1 5.460(2) 5.4627(2) 5.7
10 0,1/5 5.4747(3) 5.4737(6) 5.6
11 0.1/10 5.490(2) 5.4759(4) 5.5
12 0.1/20 5.4802(9) 5.481(1) 4.9
13 1/5 5.475(1) 5.469(1) 5.6
14 1/10 5.4731(5) 5.4804(3) 2.7
* - The sample labeling used in Table 1 is maintained across the whole paper for the reader's convenience.
Fig. 2. DTA/DTG analysis of CaF2:Yb(7.0 mol.%):Pr(0.01 mol.%) precipitate. T - temperature, G - weight, DTA - differential thermal analysis, DTG - differential thermal gravimetry. The weights at annealing was 3.7 wt.%
Table 2. EDX results for CaF2:Yb:Pr samples
Sample number -temperature of drying Pr/Yb, mol.% The effective praseodymium/ytterbium distribution coefficient
Nominal doping Doping based on EDX
10-45 0.1/5.0 */6.8 -/1.36
10-600 0.1/10.0 */7.1 -/1.42
11-600 0.1/10.0 */15.5 -/1.55
13-45 1.0/5.0 1.1/7.9 1.1/1.58
14-45 1.0/10.0 1.1/14.9 1.1/1.49
* The content of praseodymium is below the limit of reliable determination.
spectral lines of praseodymium ions have been resolved which agree with literature data [19] and consistent with data for other fluoride systems containing Pr3+ [20-22]. These lines correspond to numerous transitions from 3Pj manifold to 3HJ and 3FJ manifolds. The intensity of ytterbium luminescence at around 1-micrometer spectral range has increased one thousand fold. It means that annealing stage is an efficient tool for dehydration and increasing of ytterbium luminescence.
Luminescence spectra for series of annealed samples with different Pr3+ and Yb3+ ions content have been investigated. Characteristic spectra are presented in Fig. 4b. The inset shows luminescence lines in visible spectral range in enlarged intensity scale. The known Pr3+ ion luminescence lines are registered in the visible spectral range. The intense luminescence within 900-1100 nm range corresponding to 2F5/2—2F7/2 transitions of Yb3+ ions was detected. This indicates a successful transfer of excitation energy from praseodymium to ytterbium. Intense luminescence of ytterbium at A = 1 ^m overlaps the photosensitivity region of crystalline silicon solar cells [1,3]. We observe a decrease of Pr3+ luminescence intensity with higher Yb3+ content for samples with low levels of Pr3+ doping which proves down-conversion approach for the selected ions pair. This also indicates the applicability of this approach to increase the efficiency of solar cells. At the same time, the decrease of shorter wavelength peak of Yb3+ luminescence is observed with the increase of Yb3+ content, which speaks about reabsorption process.
Unfortunately, the effort to increase the down-conversion yield by an increase of Pr3+ ion content does not yield positive results. In Fig. 5a one can see that luminescence lines intensity of Pr3+ ions increase while the
(e)
Fig. 3. (a) SEM image of sample 10 after drying at 45 °C. (b) SEM image (EBSD regime) of sample 13 after drying at 45 °C. (c) SEM image of sample 14 after drying at 45 °C. (d) SEM image of sample 10 after annealing at 600 °C. (e) SEM image of sample 11 after annealing at 600 °C.
Wavelength.
(a)
Wavelength, (b)
Fig. 4. Luminescence spectra for sample of CaF2 with Pr3+ (0.1 %) and Yb (5 %) before and after annealing (a), luminescence of a series of CaF2 :Pr:Yb samples annealed at 600 °C (b)
intensity of Yb3+ ions drops for samples with 5 % of Yb3+ when increasing the Pr3+ content from 0.05 % to 1 %. Apparently, we have the opposing process of radiationless decay of Yb3+ ions excited state.
In order to clarify this, we have studied the Yb3+ luminescence decay kinetics in the series of samples with constant Yb3+ content and varying Pr3+ content. In the Fig. 5b luminescence decay curves of Yb3+ detected at 978 nm are presented. It is clearly seen that Yb3+ luminescence lifetime drops with increased Pr3+ ion content.
Fig. 5. Luminescence spectra (a) and Yb3+ luminescence decay (b) for samples of CaF2 with constant Yb3+ content (10 %) and various Pr + content
Apparently, this quenching occurs due to close location of 2F5/2 and 1G4 manifolds for Yb3+ and Pr3+ ions respectively in the band gap of CaF2 matrix which provides energy transfer from Yb3+ to Pr3+ ions. This process brings more restrictions to search for optimal Yb and Pr ions composition in CaF2 for means of efficient down-conversion. The highest level of luminescence intensity among all the studied samples was observed for the composition CaF2:Pr (0.05 mol.%):Yb (5.0 mol.%).
4. Conclusions
Single-phase precipitates with an average particle size of about 35 nm were formed upon precipitation of Ca1-x-y YbxPryF2+x+y solid solutions from the corresponding aqueous nitrate solutions. Crystallization of fluoride phases proceeds according to the incongruent mechanism. The mean value of the ytterbium distribution coefficient
was about 1.45. These samples dried at 45 °C had weak luminescence intensities. The annealing step for dehydration was found to be necessary for intensity luminescence increasing. According to DTA/DTG study, the temperature of annealing must be higher than 550 °C. The annealing conditions chosen for further study were 600 °C, 1 hour. After annealing, the particle size increased up to 150-200 nm. Individual luminescence bands of praseodymium are distinguishable in the luminescence spectrum. The intensity of the luminescence of ytterbium increased by 1000-fold. At the same time, we have observed the quenching of Yb3+ ions luminescence at high levels of Pr3+ content apparently due to energy transfer from 2F5/2 manifold of Yb3+ to *G4 manifold of Pr3+ ion. The highest luminescence intensity of ytterbium was determined for a Cao.9495Ybo.o5ooPro.ooo5F2.o5o5 solid solution. This composition is substantial interest for further experiments aimed at the creation of luminescent coatings on crystalline silicon designed to increase the efficiency of solar panels.
Acknowledgements
The study was funded by a grant from the Russian Science Foundation (project # 17-73-20352). Authors thank E. V. Chernova for their help in preparation of this manuscript.
References
[1] Huang X., Han S., Huang W., Liu X. Enhancing solar cell efficiency: the search for luminescent materials as spectral converters. Chem. Soc. Rev., 2013, 42, P. 173-201.
[2] Green M.A., Bremner S.P. Energy conversion approaches and materials for high-efficiency photovoltaics. Nature Mater., 2017, 16, P. 23-34.
[3] Richardson B.S. Enhancing the performance of silicon solar cells via the application of passive luminescence conversion layers. Solar Energy Materials & Solar Cells, 2006, 90, P. 2329-2337.
[4] Piper W.W., DeLuca J.A., Ham F.S. Cascade fluorescent decay in Pr3+-doped fluorides: Achievement of a quantum yield greater than unity for emission of visible light. J. Lumin., 1974, 8, P. 344-348.
[5] Sobolev B.P. The Rare Earth Trifluorides. Part 1. The High Temperature Chemistry of the Rare Earth Trifluorides. Institut d'Estudis Catalans, Barcelona, Spain, 2000, 520 pp.
[6] Fedorov P.P., Mayakova M.N., Kuznetsov S.V., et. al. Synthesis of CaF2-YF3 nanopowders by co-precipitation from aqueos solutions. Nanosystems: Physics, Chemistry, Mathematics, 2017, 8(4), P. 462-470.
[7] Mayakova M.N., Luginina A.A., Kuznetsov S.V., et. al. Synthesis of SrF2-YF3 nanopowders by co-precipitation from aqueous solutions. Mendeleev Communications, 2014, 24(6), P. 360-362.
[8] Kuznetsov S.V., Fedorov P.P., Voronov V.V., et. al. Synthesis of Ba4R3Fi7 (R stands for Rare-Earth Elements) Powders and Transparent Compacts on Their Base. Rus. J. Inorg. Chem., 2010, 55(4), P. 484-493.
[9] Fedorov P.P., Mayakova M.N., Kuznetzov S.V., et. al. Co-precipitation of yttrium and barium fluorides from aqueous solutions. Materials Research Bulletin, 2012, 47, P. 1794-1799.
[10] Karbowiak M., Cichos J. Does BaYF5 nanocrystal exist? - The BF2-YF3 solid solution revisited using photoluminescence spectroscopy. J. Alloy Compd., 2016, 673, P. 258-264.
[11] ElFajri A., Moussetad M., Tardy P., Barriere A.S. Electrical study of Pr3+ ion environment in Cai—xPrxF2+x thin films. Phys. Status solidi (a), 2000, 179, P. 373-386.
[12] Sokolov N.S., Suturin S.M. MBE growth of calcium and cadmium fluoride nanostructures on silicon. Appl. Surf. Sci., 2002, 175-176, P. 619-628.
[13] Blunier S., Zogg H., Maissen C., et. al. Lattice and thermal misfit dislocation in epitaxial CaF2/Si(111) and BaF2-CaF2/Si(111) structures. Phys. Rev. Lett., 1992, 68(24), P. 3599-3602.
[14] Yagoub M.Y.A., Swart H.C., Coetsee E. Concentration quenching, surface and spectral analyses of SrF2:Pr3+ prepared by different synthesis techniques. Opt. Mater., 2015, 42, P. 204-209.
[15] Fedorov P.P., Mayakova M.N., Kuznetsov S.V., Voronov V.V. Low temperature phase formation in the CaF2-HoF3 system. Russ. J. Inorg. Chem., 2017, 62(9), P. 1173-1176.
[16] Batsanova L.R. Rare-earth fluorides. Rus. Chem. Rev., 1971, 40(6), P. 465-484.
[17] Gmelin Handbuch der anorganischen Chemie. Syst. Nummer 39: Seltenerdelemente. Teil C.3: Sc, Y La und Lanthanide. Fluoride, Oxifluoride und zugehogige Alkalidoppelverbindungen. Berlin, Springer Vlg., 1976, 439 pp.
[18] Rozhnova Yu.A., Kuznetsov S.V., Luginina A.A., et. al., New Sri—x—yRx(NH4)yF2+x—y (R = Yb, Er) solid solution as precursor for high efficiency up-conversion luminophor and optical ceramics on the base of strontium fluoride. Mater. Chem. Phys., 2016, 172, P. 150-157.
[19] Dieke G.H., Crosswhite H.M. The spectra of the doubly and triply ionized rare earths. Appl. Opt., 1963, 2, P. 675-686.
[20] Kuzmanoski A., Pankratov V., Feldmann C. Energy transfer of the quantum-cutter couple Pr3+-Mn2+ in CaF2:Pr3+,Mn2+ nanoparticles. J. Lumin., 2016, 179, P. 555-561.
[21] Meijerink A., Wegh R., Vergeer P., Vlugt T. Photon management with lanthanides. Opt. Mater., 2006, 28, P. 575-581.
[22] Pudovkin M.S., Morozov O.A., Pavlov V.V., Korableva S.L., Lukinova E.V., Osin Y.N., Evtugyn V.G., Safiullin R.A., Semashko V.V. Physical Background for Luminescence Thermometry Sensors Based on
Pr3+
:LaF3 Crystalline Particles. J. Nanomaterials, 2017, Article
No 3108586.
