Uniform doping of metal oxide nanowires using solid state diffusion
Supporting Information for
Uniform doping of metal oxide nanowires using solid state diffusion Joaquin Resasco1, Neil P. Dasgupta2,3, Josep Roque Rosell4, Jinghua Guo4, and Peidong Yang2,5* 1
Department of Chemical Engineering and 2Department of Chemistry, University of California, Berkeley, California 94720, United States. 3Department of Mechanical Engineering, University of Michigan, Ann Arbor, Michigan 48109, United States. Advanced Light Source, Lawrence Berkeley National Laboratory, Berkeley, California 94720, United States. 5Materials Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720, United States Table S1. EXAFS Curve Fitting Parameters Path R(Å) N σ2(Å2) Mn-O 2.03 4 .00386 Mn-O 2.15 2 .00428 Mn-Ti 3.30 4 .0104 Mn-Ti 3.79 4 .0127
ΔE0(eV) 2.6576
S0 0.8607
Rf (%) 1.6
The coordination number N was taken to be fixed for the crystalline material. S0 and ΔE0 were fit to the same value for all paths as they are properties of the central atom. X-Ray Photoelectron Spectroscopy Sample Source MnO This work Ref S1 Ref S2 Mn2O3 This work Ref S1 Ref S2 MnO2 This work Ref S1 Ref S2 As deposited Annealed
Δ2p1/2 (eV) 6.0 6.0 5.4 10.1 10.0 10.5 12.0 11.8 11.9 10.75 6.2
S1
Δ3s (eV) 6.1 6.0 6.1 5.2 5.1 5.4 4.7 4.5 4.5 5.3 6.0
Figure S1: XPS surv vey spectra of MnOx AL LD films deeposited usinng the Mangganese bis(N N,N’diisoprop pylacetamidiinate) follow wing a 60 sec Ar sputtter. The lackk of carbon in the specctrum reflects th he thermal stability s of th he amidinatee precursor.
Figure S2 2: (A) Thick kness of MnO Ox ALD film ms measuredd by TEM annd ellipsomeetry as a funnction of numbeer of cycless. A linear growth g rate of o ~1.0 Å/c ycle was obbserved. (B) Growth ratte per cycle of MnO M x ALD films underr increasing pulse p time oof the Mn am midinate preccursor. Saturration was obseerved after 1.0 s. S2
Figure S3: (A) TEM M image of a core-shelll TiO2|MnO x nanowire after 100 cyycles, show wing a conformaal coating. Scale S bar is 20 2 nm. (B,C)) STEM mo de EDS mappping and coorrespondingg line scan of core-shell c naanowire afterr 200 cycless showing coonfinement of the Mn to the shell oof the wire and the Ti in thee core. Scalee bar is 100 nm. n
Figure S4 4: XPS specctra of the Ti T 2p region,, showing thhe binding eenergy and ppeak structurre are unchangeed before an nd after the conversion prrocess.
S3
Figure S5: EDS linee scans of Mn:TiO M owires show wing the Mnn:Ti intensityy ratio acrosss the 2 nano ngth (B) of the wire, in ndicating thee homogeneoous incorporation of thee Mn diameter (A) and len dopant. Correspondin C ng elementall maps are sh hown in Figu gure 1f and 1g.
Figure S6 6: (A,B) STE EM mode ellemental map ps of Ti and Zr in conveerted wire.
S4
Calculattion of Expeected Mn Co oncentration
Figure S7 7: Measured d and calculaated atomic concentratio c on of Mn in converted wires as a funnction of ALD film f thickness.
S5
Ternary y MnTiO3 Ph hase Forma ation
Figure S8: S X-Ray Diffraction patterns fo or convertedd wire sam mples at diffferent anneealing temperatu ures. The teernary phasee MnTiO3 is formed at 1050°C. * iindicates difffraction linees for MnTiO3.
9: (A,B) TEM M images off nanowires converted att 1050°C. Nanowires phhase segregatte to Figure S9 form axiaal segments of pure MnT TiO3 and TiO O2. Axial seggmentation iis likely due to high diffu fusion 2+ rates of Mn M along c-axis. c EDS confirms rellative stoichiiometry of M Mn and Ti inn segments.
S6
Figure S10: (A,B) HR RTEM imag ges of nanow wires convertted at 1050°C, showing the interfacee between MnTiO3 and d TiO2. (C-E E) FFTs of TiiO2 section, MnTiO3 secction, and intterface respectiv vely. (F) Moiiré image shows epitaxiaal interface bbetween MnnTiO3 110 annd TiO2 001 1 which caan form due to t the similaarity of the oxygen sublaattice in the ttwo structurees. Howeverr, edge disllocations aree observed att the interfacce.
S7
Transient Diffusion of Mn in TiO2 Nanowires The transient diffusion of Mn in the TiO2 nanowire can be described by Fick’s Second Law:
Where is the concentration of Mn and x is the distance from the edge of the nanowire. The solution for diffusion into a finite medium solution can be obtained numerically and is available in for simple geometric shapes in Gurney-Lurie charts. The numerical solution would predict a more rapid conversion than the solution for a semi-infinite medium but otherwise similar diffusion profiles. Therefore as a simple approximation, the nanowire can be considered a semiinfinite medium to yield an analytical solution: erf
2
can be taken as unity as the nanowire shell is manganese The surface concentration of Mn ( oxide. The initial concentration of Mn in the nanowire ( ) is zero. From single crystal studies the diffusion coefficients of various transition metals in rutile TiO2 at different partial pressures of oxygen are known.3 The diffusion coefficients of interest here are those orthogonal to the c-axis, which are lower than those parallel to the c-axis. The diffusion coefficients were also measured to be substantially higher at low partial pressures of oxygen. The diffusion coefficient of Mn in TiO2 is quite high: 1.1*10-10 cm2/s at 900°C in air. As the diffusion distance is quite small as the nanowire diameter is ~100 nm, the nanowires readily become fully converted at high temperature. To observe the transient diffusion behavior the conversion process was investigated at lower temperatures. The expected distribution of Mn for the semi-infinite medium solution at T=600°C is shown in Figure S11, assuming a diffusion coefficient of 2.7*10-12 cm2/s for Mn diffusion orthogonal to the c-axis in an Ar environment. STEM elemental maps for incomplete conversion for different conversion times are shown in Figure S12. We see that at elevated temperatures the conversion is very rapid, and that the expected diffusion profiles are reproduced by the experimental data.
S8
Figure S11: Mn distrributions in the t TiO2 nan nowire as a fu function of tiime for convversion at 6000°C in an Ar environment. The nanow wire is assum med to be a ssemi-infinitee medium. T The nanowiree becomes fully converrted in a relaatively short conversion time.
Figure S12: (A-D) ST TEM mode EDS E elemen ntal maps forr transient coonversion off nanowire aafter n of conversiion at 600°C C in Ar. 0, 1, 10, and 100 min
S9
Mechaniism for Difffusion Proceess From sin ngle crystal studies s on meetal ion diffu usion in rutille TiO2, the mechanism of diffusionn is suggested d to be dependent on thee charge state of the ion. Divalent im mpurity ions diffuse rapiddly along opeen channels parallel to th he c-axis and d are pushedd into substittutional Ti siites by intersstitial Ti ions. Divalent D ions therefore have h a large anisotropy iin diffusion ccoefficient ffor different crystallog graphic direcctions. Trivaalent and tetrravalent ionss dissolve suubstitutionallly and diffusse interstitiaally. Both mechanisms are a importantt for mixed-vvalent impurrity ions succh as Mn. Howeverr, in this systtem, the con ncentration grradient existts in the radiial direction,, orthogonal to the c-axis. Therefore it is likely that t the diffu usion of Mn from the sheell into the w wire core is driven by y Mn atoms dissolving substitutionallly and diffuusing throughh an interstittial mechaniism.3 The largeer diffusion coefficients c observed for conversionn in an Ar ennvironment aare a result oof changes in i point defeect concentraations, nameely oxygen vvacancies andd interstitial Ti atoms, w with partial prressure of ox xygen. In air, the diffusio on coefficiennts are reducced due to thhe smaller pooint defect co oncentration and Mn is not n reduced to t the Mn2+ ooxidation staate as shownn in XPS in Figure S13, howeverr, the mechan nism for difffusion is likeely unchangeed.
Figure S13: (A,B) XP PS spectra of o the Mn 2p and 3s regioons, indicatiing the satelllite peak disttance in the 2p region and the t 3s multip plet peak spllitting. The vvalues are coonsistent witth a Mn3+ n state. oxidation
S10
References: S1. Gorlin, Y.; Jaramillo, T. J. Am. Chem. Soc. 2010, 132, 13612. S2. Dicastro, V.; Polzonetti, G. J. Electron Spectrosc. Relat. Phenom. 1989, 48, 117. S3. Sasaki, J.; Peterson, N.; Hoshino, K. J. Phys. Chem. Solids 1985, 46, 1267.
S11
Uniform doping of metal oxide nanowires using solid state diffusion Joaquin Resasco1, Neil P. Dasgupta2,3, Josep Roque Rosell4, Jinghua Guo4, and Peidong Yang2,5* 1
Department of Chemical Engineering and 2Department of Chemistry, University of California, Berkeley, California 94720, United States. 3Department of Mechanical Engineering, University of Michigan, Ann Arbor, Michigan 48109, United States. Advanced Light Source, Lawrence Berkeley National Laboratory, Berkeley, California 94720, United States. 5Materials Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720, United States Table S1. EXAFS Curve Fitting Parameters Path R(Å) N σ2(Å2) Mn-O 2.03 4 .00386 Mn-O 2.15 2 .00428 Mn-Ti 3.30 4 .0104 Mn-Ti 3.79 4 .0127
ΔE0(eV) 2.6576
S0 0.8607
Rf (%) 1.6
The coordination number N was taken to be fixed for the crystalline material. S0 and ΔE0 were fit to the same value for all paths as they are properties of the central atom. X-Ray Photoelectron Spectroscopy Sample Source MnO This work Ref S1 Ref S2 Mn2O3 This work Ref S1 Ref S2 MnO2 This work Ref S1 Ref S2 As deposited Annealed
Δ2p1/2 (eV) 6.0 6.0 5.4 10.1 10.0 10.5 12.0 11.8 11.9 10.75 6.2
S1
Δ3s (eV) 6.1 6.0 6.1 5.2 5.1 5.4 4.7 4.5 4.5 5.3 6.0
Figure S1: XPS surv vey spectra of MnOx AL LD films deeposited usinng the Mangganese bis(N N,N’diisoprop pylacetamidiinate) follow wing a 60 sec Ar sputtter. The lackk of carbon in the specctrum reflects th he thermal stability s of th he amidinatee precursor.
Figure S2 2: (A) Thick kness of MnO Ox ALD film ms measuredd by TEM annd ellipsomeetry as a funnction of numbeer of cycless. A linear growth g rate of o ~1.0 Å/c ycle was obbserved. (B) Growth ratte per cycle of MnO M x ALD films underr increasing pulse p time oof the Mn am midinate preccursor. Saturration was obseerved after 1.0 s. S2
Figure S3: (A) TEM M image of a core-shelll TiO2|MnO x nanowire after 100 cyycles, show wing a conformaal coating. Scale S bar is 20 2 nm. (B,C)) STEM mo de EDS mappping and coorrespondingg line scan of core-shell c naanowire afterr 200 cycless showing coonfinement of the Mn to the shell oof the wire and the Ti in thee core. Scalee bar is 100 nm. n
Figure S4 4: XPS specctra of the Ti T 2p region,, showing thhe binding eenergy and ppeak structurre are unchangeed before an nd after the conversion prrocess.
S3
Figure S5: EDS linee scans of Mn:TiO M owires show wing the Mnn:Ti intensityy ratio acrosss the 2 nano ngth (B) of the wire, in ndicating thee homogeneoous incorporation of thee Mn diameter (A) and len dopant. Correspondin C ng elementall maps are sh hown in Figu gure 1f and 1g.
Figure S6 6: (A,B) STE EM mode ellemental map ps of Ti and Zr in conveerted wire.
S4
Calculattion of Expeected Mn Co oncentration
Figure S7 7: Measured d and calculaated atomic concentratio c on of Mn in converted wires as a funnction of ALD film f thickness.
S5
Ternary y MnTiO3 Ph hase Forma ation
Figure S8: S X-Ray Diffraction patterns fo or convertedd wire sam mples at diffferent anneealing temperatu ures. The teernary phasee MnTiO3 is formed at 1050°C. * iindicates difffraction linees for MnTiO3.
9: (A,B) TEM M images off nanowires converted att 1050°C. Nanowires phhase segregatte to Figure S9 form axiaal segments of pure MnT TiO3 and TiO O2. Axial seggmentation iis likely due to high diffu fusion 2+ rates of Mn M along c-axis. c EDS confirms rellative stoichiiometry of M Mn and Ti inn segments.
S6
Figure S10: (A,B) HR RTEM imag ges of nanow wires convertted at 1050°C, showing the interfacee between MnTiO3 and d TiO2. (C-E E) FFTs of TiiO2 section, MnTiO3 secction, and intterface respectiv vely. (F) Moiiré image shows epitaxiaal interface bbetween MnnTiO3 110 annd TiO2 001 1 which caan form due to t the similaarity of the oxygen sublaattice in the ttwo structurees. Howeverr, edge disllocations aree observed att the interfacce.
S7
Transient Diffusion of Mn in TiO2 Nanowires The transient diffusion of Mn in the TiO2 nanowire can be described by Fick’s Second Law:
Where is the concentration of Mn and x is the distance from the edge of the nanowire. The solution for diffusion into a finite medium solution can be obtained numerically and is available in for simple geometric shapes in Gurney-Lurie charts. The numerical solution would predict a more rapid conversion than the solution for a semi-infinite medium but otherwise similar diffusion profiles. Therefore as a simple approximation, the nanowire can be considered a semiinfinite medium to yield an analytical solution: erf
2
can be taken as unity as the nanowire shell is manganese The surface concentration of Mn ( oxide. The initial concentration of Mn in the nanowire ( ) is zero. From single crystal studies the diffusion coefficients of various transition metals in rutile TiO2 at different partial pressures of oxygen are known.3 The diffusion coefficients of interest here are those orthogonal to the c-axis, which are lower than those parallel to the c-axis. The diffusion coefficients were also measured to be substantially higher at low partial pressures of oxygen. The diffusion coefficient of Mn in TiO2 is quite high: 1.1*10-10 cm2/s at 900°C in air. As the diffusion distance is quite small as the nanowire diameter is ~100 nm, the nanowires readily become fully converted at high temperature. To observe the transient diffusion behavior the conversion process was investigated at lower temperatures. The expected distribution of Mn for the semi-infinite medium solution at T=600°C is shown in Figure S11, assuming a diffusion coefficient of 2.7*10-12 cm2/s for Mn diffusion orthogonal to the c-axis in an Ar environment. STEM elemental maps for incomplete conversion for different conversion times are shown in Figure S12. We see that at elevated temperatures the conversion is very rapid, and that the expected diffusion profiles are reproduced by the experimental data.
S8
Figure S11: Mn distrributions in the t TiO2 nan nowire as a fu function of tiime for convversion at 6000°C in an Ar environment. The nanow wire is assum med to be a ssemi-infinitee medium. T The nanowiree becomes fully converrted in a relaatively short conversion time.
Figure S12: (A-D) ST TEM mode EDS E elemen ntal maps forr transient coonversion off nanowire aafter n of conversiion at 600°C C in Ar. 0, 1, 10, and 100 min
S9
Mechaniism for Difffusion Proceess From sin ngle crystal studies s on meetal ion diffu usion in rutille TiO2, the mechanism of diffusionn is suggested d to be dependent on thee charge state of the ion. Divalent im mpurity ions diffuse rapiddly along opeen channels parallel to th he c-axis and d are pushedd into substittutional Ti siites by intersstitial Ti ions. Divalent D ions therefore have h a large anisotropy iin diffusion ccoefficient ffor different crystallog graphic direcctions. Trivaalent and tetrravalent ionss dissolve suubstitutionallly and diffusse interstitiaally. Both mechanisms are a importantt for mixed-vvalent impurrity ions succh as Mn. Howeverr, in this systtem, the con ncentration grradient existts in the radiial direction,, orthogonal to the c-axis. Therefore it is likely that t the diffu usion of Mn from the sheell into the w wire core is driven by y Mn atoms dissolving substitutionallly and diffuusing throughh an interstittial mechaniism.3 The largeer diffusion coefficients c observed for conversionn in an Ar ennvironment aare a result oof changes in i point defeect concentraations, nameely oxygen vvacancies andd interstitial Ti atoms, w with partial prressure of ox xygen. In air, the diffusio on coefficiennts are reducced due to thhe smaller pooint defect co oncentration and Mn is not n reduced to t the Mn2+ ooxidation staate as shownn in XPS in Figure S13, howeverr, the mechan nism for difffusion is likeely unchangeed.
Figure S13: (A,B) XP PS spectra of o the Mn 2p and 3s regioons, indicatiing the satelllite peak disttance in the 2p region and the t 3s multip plet peak spllitting. The vvalues are coonsistent witth a Mn3+ n state. oxidation
S10
References: S1. Gorlin, Y.; Jaramillo, T. J. Am. Chem. Soc. 2010, 132, 13612. S2. Dicastro, V.; Polzonetti, G. J. Electron Spectrosc. Relat. Phenom. 1989, 48, 117. S3. Sasaki, J.; Peterson, N.; Hoshino, K. J. Phys. Chem. Solids 1985, 46, 1267.
S11
