Nanoporous thin films with controllable nanopores ...

Home

Search

Collections

Journals

About

Contact us

My IOPscience

Nanoporous thin films with controllable nanopores processed from vertically aligned nanocomposites

This content has been downloaded from IOPscience. Please scroll down to see the full text. 2010 Nanotechnology 21 285606 (http://iopscience.iop.org/0957-4484/21/28/285606) View the table of contents for this issue, or go to the journal homepage for more

Download details: IP Address: 221.10.40.236 This content was downloaded on 28/09/2013 at 23:28

Please note that terms and conditions apply.

IOP PUBLISHING

NANOTECHNOLOGY

Nanotechnology 21 (2010) 285606 (7pp)

doi:10.1088/0957-4484/21/28/285606

Nanoporous thin films with controllable nanopores processed from vertically aligned nanocomposites Zhenxing Bi1 , Osman Anderoglu2 , Xinghang Zhang2 , Judith L MacManus-Driscoll3, Hao Yang4 , Quanxi Jia4 and Haiyan Wang1 1

Department of Electrical and Computer Engineering, Texas A&M University, College Station, TX 77843, USA 2 Department of Mechanical Engineering, Texas A&M University, College Station, TX 77843, USA 3 Department of Materials Science and Metallurgy, University of Cambridge, Pembroke Street, Cambridge CB2 3QZ, UK 4 Center for Integrated Nanotechnology, Los Alamos National Laboratory, Los Alamos, NM 87545, USA E-mail: [email protected]

Received 11 February 2010, in final form 26 May 2010 Published 28 June 2010 Online at stacks.iop.org/Nano/21/285606 Abstract Porous thin films with ordered nanopores have been processed by thermal treatment on vertically aligned nanocomposites (VAN), e.g., (BiFeO3 )0.5 :(Sm2 O3 )0.5 VAN thin films. Uniformly distributed nanopores with an average diameter of 60 nm and 150 nm were formed at the bottom and top of the nanoporous films, respectively. Controllable porosity can be achieved by adjusting the microstructure of VAN (BiFeO3 ):(Sm2 O3 ) thin films and the annealing parameters. In situ heating experiments within a transmission electron microscope (TEM) column at temperatures from 25 to 850 ◦ C, provides significant insights into the phase transformation, evaporation and structure reconstruction during the annealing. The in situ experiments also demonstrate the possibility of processing vertically aligned nanopores (VANP) with one phase stable in a columnar structure. These nanoporous thin films with controllable pore size and density could be promising candidates for thin film membranes and catalysis for fuel cell and gas sensor applications. (Some figures in this article are in colour only in the electronic version)

applications. It was reported that the same porous material system processed by different techniques results in a large variation in the electrochemical performance of the cathode membrane for solid oxide fuel cells (SOFCs) [12–15]. Most of the porous structures were either directly deposited by solutionbased techniques or other deposition techniques [16–18], or achieved by post-deposition processes such as chemical etching [19, 20], electrochemical processes [21, 22], high resolution photolithography patterning [23], and e-beam etching [24]. Most of these techniques have difficulties in achieving an ordered nanoscale porosity with controlled pore density and

1. Introduction Nanoporous membranes and thin films have received much research interest because of their potential applications as porous catalysis membranes in fuel cells [1–5], gas sensors [6–8], and porous anodic alumina oxide (AAO) membrane templates for processing nanotubes and nanowires [9–11]. Processing porous thin films with geometrically controllable nanopores, i.e., controlled pore density and pore size distribution, is important in achieving consistent chemical catalytic properties of thin film membranes for fuel cells and gas sensor 0957-4484/10/285606+07$30.00

1

© 2010 IOP Publishing Ltd Printed in the UK & the USA

Nanotechnology 21 (2010) 285606

Z Bi et al

Figure 1. TEM analysis on (BFO)0.5 :(SmO)0.5 VAN thin films. (a) Cross-sectional TEM images of (BFO)0.5 :(SmO)0.5 VAN thin film with (b) zoom-in high resolution TEM images and the corresponding selected-area diffraction pattern as an inset. (c) Plan-view TEM image of (BFO)0.5 :(SmO)0.5 VAN thin film with (d) zoom-in high resolution TEM image and the corresponding selected-area diffraction pattern as an inset. The epitaxial orientation relationships are determined to be BFO(002) SmO(004) STO(002) and BFO(200) SmO(220) STO(200).

pore size distribution. Some of the techniques require multiple steps (e.g., high resolution lithography patterning) and can be very expensive (e.g., e-beam etching). Our recent work on vertically aligned nanocomposites (VAN) opens a brand new avenue in achieving ordered nanopore structures. Several VAN structures including BiFeO3 :Sm2 O3 (BFO:SmO) [25–27], La0.7 Sr0.3 MnO3 :ZnO (LSMO:ZnO) [28], and other systems have been demonstrated [29–31]. Both phases in VAN structures have grown epitaxially with each other on substrates and formed a unique nano-checkerboard structure in-plane. It shows vertical strain control along the vertical interface which allows highly strained films to be achieved in thick films. The two phases in the VAN structure are selected based on their similar growth kinetics, thermodynamic stability and epitaxial growth on given substrates [25]. Using this two-phase VAN ordered structure as a template, it is highly possible to process ordered nanopore structures by eliminating one phase in the two-phase VAN structure. Nanopore geometry control including the pore density and size can be achieved by: (1) adjusting the phase compositions or deposition parameters to obtain different column widths in the VAN structures [27]; (2) optimizing the thermal treatment parameters such as annealing temperature and duration. In this paper, we present our initial success in processing nanoporous films by both furnace annealing and in situ heating experiments in a transmission electron microscope (TEM). The VAN system selected for this study is (BFO)0.5 :(SmO)0.5 film for a demonstration. BFO has a melting point of 817–825 ◦ C, which is lower than that of SmO (2300 ◦ C) [32–34]. BFO can be decomposed and vaporized, and ordered nanopores will be formed during annealing. The effects of annealing temperature on the new nanoporous structures were also investigated to explore the formation mechanisms of the nanopores.

2. Experimental methods (BFO)0.5 :(SmO)0.5 VAN thin films were deposited on single crystal SrTiO3 (STO)(001) substrates in a pulsed laser deposition (PLD) system with a KrF excimer laser (Lambda Physik Compex Pro 205, λ = 248 nm). The laser beam was focused to obtain an energy density (4 J cm−2 ) at an incidence angle of 45◦ . The (BFO)0.5 :(SmO)0.5 targets were prepared by a conventional powder mixing method followed by target sintering with flowing oxygen (0.150 sccm). An optimized substrate temperature (650 ± 5 ◦ C) and an oxygen pressure of 0.2 Torr were maintained during deposition. Other details can be found in a previous paper [27]. The film thickness is controlled at 150 nm for this study. The microstructure of the as-deposited and annealed films was characterized by x-ray diffraction (XRD) (BRUKER D8 powder x-ray diffractometer), scanning electron microscopy (SEM) (FEI SEM with field emission electron gun), atomic force microscopy (AFM) and high resolution TEM. A standard cross-sectional view and plan-view TEM sample preparation procedure was used here, including manual grinding and thinning of the sample by a final ion milling step (PIPS 691 precision ion polishing system, 3.7 keV). Both the conventional TEM characterization and in situ heating experiments were carried out using JEOL-2010 (LaB6 electron gun) with a Gatan ORIUS CCD camera and a point ˚ A Gatan-628 side entry, furnace type resolution of 2.34 A. single tilt heating TEM specimen holder was used for the in situ heating experiment. The temperature was measured using a thermocouple attached to the furnace and the maximum operating temperature is 1300 ◦ C. The heating current and the ramping rate were carefully calibrated by considering the temperature gap between the specimen and the thermal sensor. 2


Z Bi et al

the BFO and SmO phases in the corresponding selectedarea diffraction (SAD) pattern and Fast Fourier Transfer patterns (as inset). The orientation relationships are determined to be BFO(002) SmO(004) STO(002) and BFO(200) SmO(220) STO(200). The (BFO)0.5 :(SmO)0.5 VAN thin films were then ex situ annealed at various temperatures and durations to explore the annealing effects on the pore formation. The first annealing was at 1000 ◦ C for 1 h. X-ray diffraction (XRD) was first conducted to explore the microstructure variation introduced by high temperature annealing. Obvious differences can be found from the XRD data in figure 2 for both the as-deposited VAN thin films and the sample after annealing. For the as-deposited sample, both BFO and SmO phase grown are highly textured, mainly along the STO(00l ) direction. After annealing, the main peaks are indexed as SmO(00l ) peaks and no obvious BFO(00l ) peaks are observed. This indicates that a large fraction of the BFO phase has decomposed and evaporated during the high temperature annealing, leaving SmO as the major phase in the annealed sample. The intensity of the SmO(800) peak was increased after the thermal treatment, which suggests that high temperature annealing also improves the SmO phase crystallinity at the time of pore formation. Microstructures of the annealed (BFO)0.5 :(SmO)0.5 thin films were examined by using cross-sectional TEM. One such image in figure 3(a) shows that highly ordered nanopores have formed in the film with an average pore diameter of 60 nm, for those close to the film–substrate interface, and 150 nm, for those on the top of the film. High temperature annealing on (BFO)0.5 :(SmO)0.5 VAN thin films forms a unique bi-layer nanopore structure. The spacing between the neighboring pores is around 50 nm for both layers of pores. Based on the pore size and distribution, the estimated total film porosity is around 40%. The corresponding selected-area diffraction (SAD) pattern of both the film and substrate shown

Figure 2. XRD θ –2θ scans of (BFO)0.5 :(SmO)0.5 nanocomposite thin films before and after annealing at 1000 ◦ C for 1 h.

While ramping up the temperature the electron beam was switched off to avoid a thermal radiation effect.

3. Results and discussion A set of (BFO)0.5 :(SmO)0.5 VAN films with film thickness of 150 nm were processed with identical processing conditions for the multiple annealing experiments. Figure 1 shows high resolution TEM images in both cross-sectional view (figures 1(a) and (b)) and plan-view (figures 1(c) and (d)) for an as-deposited (BFO)0.5 :(SmO)0.5 thin film. A clear vertically aligned columnar structure was observed. BFO and SmO phases/columns alternate with each other. The average width of the BFO/SmO columns is around 10 nm. A remarkable feature of the (BFO)0.5 :(SmO)0.5 films is the spontaneous phase ordering (figures 1(b) and (d)). Both phases grew epitaxially on SrTiO3 (STO)(001) substrates, evidenced by the distinct diffraction dots from

Figure 3. TEM image of (BFO)0.5 :(SmO)0.5 nanocomposite thin film after annealing at 1000 ◦ C for 1 h showing a bi-layer porous microstructure. Below are (b) selected-area diffraction pattern from both the film and the substrate; and fast Fourier transform patterns from the (c) STO substrate; (d) BFO region; and (e) SmO region.

3


Z Bi et al

Figure 4. Schematic diagrams illustrate a possible pore formation mechanism in (BFO)0.5 :(SmO)0.5 nanocomposite during high temperature annealing. (a)–(c) are cross-sectional view; (d)–(f) are plan-view. Areas in red correspond to BFO phase and areas in yellow correspond to SmO phase. (a) and (d) illustrates the initial VAN structure with a diameter of d . (b) and (e) represent the initial decomposition and vaporization process with pore size d1 = d . (c) and (f) describe the nanopore growth and the coalescence of the SmO phase (d3 > d2 > d1 ). d is the width of BFO columns before phase coalescence; d1 is the diameter of the top open and bottom closed pores during annealing; d2 is the diameter of the bottom closed pores after annealing; d3 is the diameter of the top open pores after annealing; d is the width of BFO column left after phase coalescence.

in figure 3(b) demonstrate the highly epitaxial SmO phase in the annealed film without clear diffraction from BFO. This observation is consistent with the above XRD result. The Fast Fourier Transform patterns processed from selected areas in the high resolution TEM image also suggest that the majority of the phase is SmO with only a very small portion of areas identified as BFO. This unique porous structure is mainly related to the original spontaneously ordered vertical columnar structure of the VAN template. Through annealing, the well aligned vertical nanocolumnar structure has been converted into a nanoporous film with uniformly distributed ordered nanopores. The diameter of the pores is approximately six times the width of the column in the as-deposited VAN films. Due to the different melting temperatures of BFO (Tm = 817– 825 ◦ C) and SmO (Tm = 2300 ◦ C), BFO will decompose and evaporate before SmO starts to decompose. The schematics in figure 4 show the possible scenario for the pore formation during annealing from both cross-sectional view ((a)–(c)) and plan-view ((d)–(f)). During the high temperature annealing (1000 ◦ C), the BFO phase went through a decomposition– vaporization–diffusion process while the SmO phase went through reconstruction. The top open pores came from the BFO phase vaporization on the top portion of the BFO columns, and the bottom closed pores were formed through the BFO phase decomposition and diffusion to free space (figures 4(b) and (e), d1 ≈ d ). The closed pores at the film– substrate interface are typically smaller than the open pores on the top of the film since the bottom pore formation is also limited by the diffusion process (figures 4(c) and (f), d3 > d2 ).

During the decomposition and vaporization of BFO, SmO columns (as well as a small amount of residual BFO phase) may have combined together to form more continuous frames with much larger nanopores (figures 4(c) and (f), d > d ). The high resolution SEM image in figure 5(a) presents the surface pore density of the annealed film. These surface pores are the ones shown on the top layer in the annealed film (shown in figure 3(a)). Based on the contrast difference (area with dark contrast corresponding to pores), the surface porosity is estimated to be 35%, which is close to the value estimated by the cross-sectional TEM study. The distribution of the surface pore sizes is plotted in figure 5(b). Most of the surface pores are in the range 100–200 nm, which is larger than the ones at the substrate–film interface. AFM scanning on the nanoporous thin film sample (not shown here) also confirmed that the average diameter of the top pores is around 150 nm with a depth of 30 nm. To fully understand this bi-layer nanopore structure formation mechanism during the annealing process, we conducted in situ heating experiments in TEM. It is a powerful tool to monitor the microstructure evolution at an atomic scale during the annealing process. It is noted that the in situ annealing experiment deals with a semi-2D sample foil (with foil thickness around 50–100 nm depending on the regions in the foil) which might not completely represent the formation mechanisms of the 3D sample. However this in situ experiment can at least reveal the decomposition and evaporation process of the BFO phases in real time. The movie snapshots from the two-stage annealing, from 25 ◦ C to 600 ◦ C and from 600 ◦ C to 850 ◦ C, are shown in figures 6(a) 4


Z Bi et al

Figure 5. (a) High resolution scanning electron microscopy analysis on (BFO)0.5 :(SmO)0.5 nanocomposite thin film annealed at 1000 ◦ C for 1 h show ordered nanopores on the film surface. (b) Surface pore density distribution reveals an average pore size of about 150 nm.

and (b), respectively. During the experiment, the ramping rate was gently controlled to avoid a sharp temperature difference between the TEM specimen and the sample stage. From 25 to 600 ◦ C (figure 6(a)), there is no obvious evaporation observed until the temperature reaches 600 ◦ C with some indication of decomposition and vaporization of the BFO phases. From 600 to 850 ◦ C, as shown in figure 6(b), the snapshots show obvious material evaporation and structure variation during the annealing process. It is clear that the vaporization of the BFO phases starts uniformly throughout the BFO columns. Starting from 750 ◦ C, the majority of the BFO phase has vaporized and left nanopores along the sides of the SmO columns. The SmO columns are largely intact. Different from the ex situ annealed sample in figure 3, the vaporization of BFO nanopores did not alter the SmO nanocolumns during the in situ experiments. It shows that the well-separated individual nanopores did not coalesce at lower temperature from 650 to 750 ◦ C. Starting from 750 up to 850 ◦ C, the nanopores start to coalesce with their neighboring ones and the SmO nanocolumns begin to lose their vertical alignment and form larger SmO grains around the nanopores. It is worth pointing out that the experimental temperature is lower than the melting temperature of SmO. This suggests that there is a significant amount of material diffusion promoted by the extra thermal energy, the energy stored at the nanocolumn boundaries and the surface energy of the nanopores, which results in the coalescence of the SmO grains even around 800 ◦ C. This in situ heating experiment definitely provides useful information for understanding the annealing process and a clear insight that it is possible to process ordered vertically aligned nanopores by precisely controlling the annealing process on the VAN structure. Regarding the materials system selection, a large number of VAN structures have been demonstrated to date [25–31]. This gives a vast choice in the material selection of the nanoporous structures. Besides annealing, the secondary phase can be removed by chemical etching or other approaches which are currently under investigation.

Figure 6. In situ heating cross-sectional TEM analysis on (BFO)0.5 :(SmO)0.5 VAN thin films. (a) Snapshots taken during temperature ramping from room temperature to 600 ◦ C. No obvious decomposition and evaporation is observed. (b) Snapshots taken from 600 to 850 ◦ C show the sequence of decomposition, evaporation and pore recombination.

5


Z Bi et al

4. Conclusion In summary, we have presented a new approach in processing nanoporous thin films with ordered nanopores from (BFO)0.5 :(SmO)0.5 VAN thin films. High resolution TEM reveals that, after annealing at 1000 ◦ C for 1 h, a bi-layer nanopore structure with open pores on the top and closed pores on the bottom has formed. The in situ heating experiment in TEM demonstrates that the BFO phases went through a decomposition–vaporization–diffusion process starting from 600 ◦ C during which the SmO phases remain mostly intact. Above 750 ◦ C the nanopores start to coalesce into large pores and the SmO columns start to recombine into large grains around the nanopores. This demonstrates that it is feasible to process vertically aligned nanopores or nanoporous films with controlled nanopore density by controlling the annealing process of VAN structures.

[12]

[13]

[14]

[15]

[16]

Acknowledgments

[17]

This work is supported by the US National Science Foundation (Ceramic Program, NSF-0709831 and 1007969). [18]

References [19]

[1] Corbin S F, Clemmer R M C and Yang Q 2009 Development and characterization of porous composites for solid oxide fuel cell anode conduction layers using ceramic-filled highly porous Ni foam J. Am. Ceram. Soc. 92 331–7 [2] Teng X W, Liang X Y, Rahman S and Yang H 2005 Porous nanoparticle membranes: synthesis and application as fuel-cell catalysts Adv. Mater. 17 2237–41 [3] Jou S and Wu T H 2008 Thin porous Ni-YSZ films as anodes for a solid oxide fuel cell J. Phys. Chem. Solids 69 2804–12 [4] Kjeang E, Michel R, Harrington D A, Djilali N and Sinton D 2008 A microfluidic fuel cell with flow-through porous electrodes J. Am. Chem. Soc. 130 4000–6 [5] Yoshikawa M, Boden A, Sparr M and Lindbergh G 2006 Experimental determination of effective surface area and conductivities in the porous anode of molten carbonate fuel cell J. Power Sources 158 94–102 [6] Zhang Q, Saraf L V, Smitha J R, Jha P and Hua F 2009 An invisible bend sensor based on porous crosslinked polyelectrolyte film Sensors Actuators A 151 154–8 [7] Aroutiounian V, Arakelyan V, Galstyan V, Martirosyan K and Soukiassian P 2009 Hydrogen sensor made of porous silicon and covered by TiO2−x or ZnOAl thin film IEEE Sensors J. 9 9–12 [8] Martinez H M, Rincon N E, Torres J and Alfonso J E 2008 Porous silicon thin film as CO sensor Microelectron. J. 39 1354–5 [9] Houser J E and Hebert K R 2009 The role of viscous flow of oxide in the growth of self-ordered porous anodic alumina films Nature Mater. 8 415–20 [10] Yang F, Xie B Y, Sun J Z, Jin H K and Wang M 2008 Preparation and wettability of Zn–Al layered double hydroxide film directly grown on highly porous anodic alumina oxide template Mater. Lett. 62 1302–4 [11] Tian M L, Xu S Y, Wang J G, Kumar N, Wertz E, Li Q, Campbell P M, Chan M H W and Mallouk T E 2005

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

6

Penetrating the oxide barrier in situ and separating freestanding porous anodic alumina films in one step Nano Lett. 5 697–703 Joo J H and Choi G M 2007 Open-circuit voltage of ceria-based thin film SOFC supported on nano-porous alumina Solid State Ion. 178 1602–7 Ma X Q, Dai J X, Zhang H, Roth J, Xiao T D and Reisner D E 2005 Solid oxide fuel cell development by using novel plasma spray techniques J. Fuel Cell Sci. Technol. 2 190–6 Lin B, Sun W P, Xie K, Dong Y C, Dong D H, Liu X Q, Gao J F and Meng G Y 2008 A cathode-supported SOFC with thin Ce0.8 Sm0.2 O1.9 electrolyte prepared by a suspension spray J. Alloys Compounds 465 285–90 Ding D, Li L, Feng K, Liu Z B and Xia C R 2009 High performance Ni–Sm2 O3 cermet anodes for intermediate-temperature solid oxide fuel cells J. Power Sources 187 400–2 Liu J P, Li Y Y, Huang X T, Li G Y and Li Z K 2008 Layered double hydroxide nano- and microstructures grown directly on metal substrates and their calcined products for application as Li-ion battery electrodes Adv. Funct. Mater. 18 1448–58 Gao Y F, Nagai A, Masuda Y, Sato F, Seo W S and Koumoto K 2006 Surface precipitation of highly porous hydrotalcite-like film on Al from a zinc aqueous solution Langmuir 22 3521–7 Hu G S, Huang J Q, Orkoulas G and Christofides P D 2009 Investigation of film surface roughness and porosity dependence on lattice size in a porous thin film deposition process Phys. Rev. E 80 041122 Hayashi Y et al 2008 Comprehensive chemistry designs in porous SiOCH film stacks and plasma etching gases for damageless Cu interconnects in advanced ULSI devices IEEE Trans. Semicond. Manuf. 21 469–80 Solanki C S, Bilyalov R R, Poortmans J, Celis J P and Nijs J 2003 Effect of the composition of electrolyte on separation of porous silicon film by electrochemical etching Phys. Status Solidi a 197 507–11 Chou S L, Wang J Z, Liu H K and Dou S X 2008 Electrochemical deposition of porous Co3 O4 nanostructured thin film for lithium-ion battery J. Power Sources 182 359–64 Djurfors B, Broughton J N, Brett M J and Ivey D G 2006 Electrochemical oxidation of Mn/MnO films: mechanism of porous film growth J. Electrochem. Soc. 153 A64–8 Ohmukai M, Okada K and Tsutsumi Y 2005 Patterned porous silicon formed with photolithography J. Mater. Sci., Mater. Electron. 16 119–21 Nakao M, Oku S, Tamamura T, Yasui K and Masuda H 1999 GaAs and InP nanohole arrays fabricated by reactive beam etching using highly ordered alumina membranes Japan. J. Appl. Phys. 1 38 1052–5 MacManus-Driscoll J L, Zerrer P, Wang H Y, Yang H, Yoon J, Fouchet A, Yu R, Blamire M G and Jia Q X 2008 Strain control and spontaneous phase ordering in vertical nanocomposite heteroepitaxial thin films Nature Mater. 7 314–20 Yang H, Wang H Y, Yoon J, Wang Y Q, Jain M, Feldmann D M, Dowden P C, MacManus-Driscoll J L and Jia Q X 2009 Vertical interface effect on the physical properties of self-assembled nanocomposite epitaxial films Adv. Mater. 21 3794–8 Bi Z X, Lee J H, Yang H, Jia Q X, MacManus-Driscoll J L and Wang H Y 2009 Tunable lattice strain in vertically aligned


Z Bi et al

nanocomposite (BiFeO3 )x :(Sm2 O3 )1−x thin films J. Appl. Phys. 106 094309 [28] Kang B S, Wang H, MacManus-Driscoll J L, Li Y, Jia Q X, Mihut I and Betts J B 2006 Low field magnetotransport properties of (La0.7 Sr0.3 MnO3 )0.5 :(ZnO)0.5 nanocomposite films Appl. Phys. Lett. 88 192514 [29] Moshnyaga V et al 2003 Structural phase transition at the percolation threshold in epitaxial (La0.7 Ca0.3 MnO3 )1−X :(MgO) X nanocomposite films Nature Mater. 2 247–52 [30] Zheng H et al 2004 Multiferroic BaTiO3 –CoFe2 O4 nanostructures Science 303 661–3

[31] Zheng H M, Straub F, Zhan Q, Yang P L, Hsieh W K, Zavaliche F, Chu Y H, Dahmen U and Ramesh R 2006 Self-assembled growth of BiFeO3 –CoFe2 O4 nanostructures Adv. Mater. 18 2747–52 [32] Tyholdt F, Jorgensen S, Fjellvag H and Gunnaes A E 2005 Synthesis of oriented BiFeO3 thin films by chemical solution deposition: phase, texture, and microstructural development J. Mater. Res. 20 2127–39 [33] Mukherje J I and Wang F F Y 1971 Kinetics of solid-state reaction of Bi2 O3 and Fe2 O3 J. Am. Ceram. Soc. 54 31 [34] Andreev O V, Vysokikh A S and Vaulin V G 2008 Sm2 S3 –Sm2 O3 phase diagram Russ. J. Inorg. Chem. 53 1320–4

7