A new method for direct preparation of tin dioxide ...mswool/publications/jmr_official_reprint.pdf · A new method for direct preparation of tin dioxide nanocomposite materials T.A

A new method for direct preparation of tin dioxidenanocomposite materials

T.A. Miller, S.D. Bakrania, C. Perez, and M.S. Wooldridgea)

Department of Mechanical Engineering, University of Michigan, Ann Arbor, Michigan 48109-2125

(Received 31 March 2005; accepted 29 June 2005)

In the current work, a novel combustion method is demonstrated for the directsynthesis of nanocomposite materials. Specifically doped tin dioxide (SnO2) powderswere selected for the demonstration studies due to the key role SnO2 plays insemiconductor gas sensors and the strong sensitivity of doped SnO2 to nanocompositeproperties. The synthesis approach combines solid and gas-phase precursors to stagethe decomposition and particle nucleation processes. A range of synthesis conditionsand four material systems were examined in the study: gold–tin dioxide, palladium–tindioxide, copper–tin dioxide, and aluminum–tin dioxide. Several additive precursorswere considered including four metal acetates and two pure metals. The nanocompositematerials produced were examined for morphology, phase, composition, and latticespacing using transmission and scanning electron microscopy, x-ray diffractometry, andenergy-dispersive spectroscopy. The results using the combustion synthesis approachindicate good control of the nanocomposite properties, such as the average SnO2

crystallite size, which ranged from 5.8 to 17 nm.

I. INTRODUCTION

Nanocomposite materials have the potential to dra-matically improve many engineering systems. Forexample, Uematsu et al.1 have created gold–titania(Au-TiO2) nanocomposite powders that exhibit remark-able increases in catalytic activity for carbon monoxideoxidation as a function of the nanocomposite mor-phology. Afonso et al.2 have created bismuth–alumina(Bi-Al2O3) and copper–alumina (Cu-Al2O3) nanocom-posites that exhibit superior structural and nonlinear op-tical properties, making the materials attractive for all-optical switching devices. Nanocomposite materials mayalso enable a hydrogen (H2) economy, as Cui and Zhang3

have demonstrated, large H2 storage capacities usingcerium–nickel (Ce-Ni) nanocomposites.

Nanocomposites can also greatly impact the perfor-mance characteristics of semiconductor gas sensors.4 Tindioxide (SnO2) is the most important material used insolid-state gas detectors, and improving the performanceof tin dioxide sensors has been directly linked to additionof dopants (typically noble metals or metal oxides) to theSnO2 to create nanocomposite materials.5–11 SnO2

nanocomposites can be created using a variety of tech-niques including sol-gel processing,6,10,11 chemical va-por deposition,12 wet chemical deposition,13,14 sputtering

methods,15,16 gas-phase condensation,17 pulsed laser ab-lation,18 mechanochemical processing,19 and combustionsynthesis.20–22

Combustion methods can be quite powerful synthesistechniques, with demonstrated ability to control particlesize, size distribution, phase, and composition.23,24 Com-bustion processes can be scaled to high production rates(on the order of g/h to kg/h),25,26 and combustion meth-ods rank among the few techniques that have been dem-onstrated to directly produce both thick and thin SnO2

films.27 The objective of this study is to demonstrate anovel combustion synthesis approach which can be usedto produce a broad range of SnO2 nanocomposite mate-rials with good control of the nanocomposite properties(e.g., morphology, average SnO2 particle size, dopantloading, etc.). Doped tin dioxide material systems wereselected for study due to the considerable promise tindioxide has in advanced gas-sensing applications and theextraordinary sensitivity of doped-SnO2 sensors to thetype, location, state, dispersion, and loading of the addi-tives (see references 4, 5, 11, 28, and 29 and referencestherein).

The synthesis approach is based on staged decompo-sition of particle precursor reactants and formation ofnanoparticles of multiple condensed-phase materials.Combustion synthesis generally uses gas- or liquid-phasereactants,23,24 excluding self-propagating high-temperature combustion synthesis (SHS) approaches.In this work, we demonstrate for the first time the com-bined use of solid- and gas-phase reactants in flame syn-thesis of nanocomposite materials. The use of multiple

a)Address all correspondence to this author.e-mail: [email protected]

DOI: 10.1557/JMR.2005.0375

J. Mater. Res., Vol. 20, No. 11, Nov 2005 © 2005 Materials Research Society 2977

precursors and multiple phases allows for greater flex-ibility and control of the final product properties. Forexample, by using precursors with different decomposi-tion and particle nucleation rates, the distribution of theadditives in the host or support material (SnO2 in thiscase) can be affected. The use of solid-phase precursorsfor the additives significantly expands the range of pre-cursor materials that can be considered for combustionsynthesis processing, including the use of less toxic ma-terials and a large number of precursors that do not con-tain chlorine. Chlorine contamination is a particular con-cern for doped-SnO2 used in gas-sensing applications.30,31

In the following sections, the synthetic approach isdescribed. The properties of the nanocomposite productpowders, including morphology, average particle size,and composition, are examined using a variety of tech-niques including transmission electron microscopy(TEM) imaging, TEM energy-dispersive spectroscopy(EDS), scanning electron microscopy x-ray energy-dispersive spectroscopy (SEM EDS), and x-ray diffrac-tometry (XRD). As reducing the average SnO2 crystallitesize to below 10 nm is a key goal to improve gas-sensorperformance,32,33 the effects of the additives on SnO2

crystallite size are presented and discussed in detail.

II. EXPERIMENTAL

The nanostructured powders were generated using thecombustion synthesis facility shown in Fig. 1. The facil-ity consists of three major components: the burner used

to create the high-temperature synthesis environment, thebubbler system used to deliver the tin dioxide precursorto the burner, and the particle feed system (PFS) used todeliver the additive precursor to the burner. A detaileddescription of the burner and the results of characteriza-tion studies can be found in Wooldridge et al.34 In thisstudy, all tin dioxide materials were produced usingtetramethyl tin (TMT) as the precursor for SnO2. A de-tailed description of the TMT bubbler system and resultsof characterization experiments for synthesis of undopedSn, SnO, and SnO2 nanoparticles can be found in Hallet al.35,36 Each of the facility components are describedbriefly below.

The burner is a multielement diffusion flame burner(or Hencken burner) that is used to produce steady, lami-nar, high-temperature conditions by combusting hydro-gen (H2) and oxygen (O2) reactants dilute in argon (Ar)at atmospheric pressure. A reducing or oxidizing envi-ronment can be created for the synthesis conditions byvarying the reactant ratios.36 The 2.54 cm × 2.54 cmsquare burner consists of a hastalloy honeycomb supportthrough which stainless steel hypodermic needles are in-serted at systematic intervals. Hydrogen flows throughthe needles, as well as dilute O2 in Ar flow through theremainder of the channels of the honeycomb. The H2 andO2 mix rapidly outside the surface of the burner, leadingto a primary flame that has a slightly dimpled flamesurface. Approximately 3–5 mm above the surface of theburner (i.e., about 1 mm above the flame sheet), theconditions are uniform in temperature, pressure, and

FIG. 1. Schematic of the combustion synthesis facility used to generate the nanocomposite materials.

T.A. Miller et al.: A new method for direct preparation of tin dioxide nanocomposite materials

J. Mater. Res., Vol. 20, No. 11, Nov 20052978

composition. To minimize entrainment of room air, theactive area of the burner is surrounded by a co-flow ofnitrogen. A square optical chimney (3.8 × 3.8 × 34 cm)can also be used to extend the high-temperature regionabove the burner. In this study, no particle precursorreactants are introduced to the burner via the H2 or O2

manifold. All particle precursor reactants are directed tothe burner using the secondary fuel tube (see Fig. 1).

The secondary fuel tube (0.85 mm i.d.), is located atthe center of the burner, and the particle precursors forthe SnO2 and the additives are supplied via two deliverysystems. The vapor-phase precursor for tin dioxide iscreated by bubbling argon through a liquid reservoir ofTMT. The argon flow rate is monitored using a calibratedrotameter, and the Ar leaves the reservoir saturated inTMT. When the reservoir is at room temperature, theflow yields a mixture of 21–23% TMT (mole basis) inAr.36 In this study, all the experiments were conductedwith the TMT reservoir at room temperature with oneexception, where the reservoir was cooled to 0 °C result-ing in a mixture of approximately 5% TMT (mole basis)in Ar.

The solid-phase precursor reactants for the additivematerials are introduced to the secondary fuel tube viathe particle feed system. The PFS consists of an entrain-ment column, a syringe, and a syringe injection pump.Similar feed systems have been developed by othergroups for coal-particle combustion studies.37,38 The PFSused in this work is based upon those designs and modi-fied for synthesis studies. The glass column has a gasinlet diameter of 3.70 mm and an outlet diameter of1.07 mm, with a maximum diameter of 98.25 mm. Argonis used as the carrier gas through the column and isregulated by a calibrated rotameter. Additive precursorparticles are injected into the centerline of the column viaan open-ended syringe (4.5 mm i.d., BD 1 ml U-100,Franklin Lakes, NJ). The injection feed rate is controlledby a syringe pump (Medfusion 2001, Diluth, IA). A har-monic actuator is used to improve the steadiness of theparticle delivery from the syringe. For each experiment,the argon flow rate to the PFS is set at a constant value(400 ml min−1) and the syringe pump is set at a constantrate of plunger displacement (1 ml/h). These conditionscorrespond to a particle feed rate of approximately 2 g/h.Once the additive precursor particles are entrained in Ar,the particle flow is mixed with the TMT/Ar flow beforeentering the secondary fuel tube via an L-junction (seeFig. 1). The precursor flow then exits the secondary fueltube above the surface of the burner as a jet (with aReynolds number of Re ≅ 580), where the reactants forma secondary flame, which is a diffusion flame. Thenanocomposite powders are formed as products of thesecondary flame. Additional description of the PFS canbe found in Miller39 and in Miller et al.40,41

Samples of the final product powders were collected

for ex situ analysis in the exhaust region above theburner. Bulk samples were collected at a height of 27 cmabove the burner surface by thermophoretic depositiononto a water-cooled cold plate for sampling times ofapproximately 10 min. Discrete samples were collectedusing thermophoretic deposition directly onto TEM grids(Electron Microscopy Sciences, carbon film, 300-meshcopper or 300-mesh nickel [Hatfield, PA]) placed at27 cm above the burner surface for sampling times of lessthan one second. Sample preparations for each analyticaltechnique used in the study are described below.

Powder samples were analyzed for composition,phase, and average crystallite size using a powder XRD(an automated Scintag Theta-Theta XRD [Scintag, Inc.,Cupertino, CA], or a Rigaku double-crystal XRD [To-kyo, Japan]). Powder samples of approximately 40 mgwere obtained from the cold plate and were dispersedwith methanol (∼0.02 ml) into a paste form. Approxi-mately 0.5 ml of the paste was spread onto glass slidesand dried at room temperature for a minimum of 10 min.Spectral scans for phase identification and for averageadditive particle size were obtained over a 2� range of15–85° at a scan rate of 5° 2�/min using increments of0.02° 2� and Cu K� radiation (� � 1.5405 Å). Spectralscans for average crystallite size for SnO2 were measuredover a 2� range of 22–31° at a scan rate of 0.5° 2�/minusing increments of 0.02° 2� and Cu K� radiation (� �1.5405 Å). Peak positions and relative intensities of thepowder patterns were identified by comparison with ref-erence spectra from the International Center for Diffrac-tion Data (ICDD, Newton Square, PA).42

Bulk samples were analyzed using SEM-EDS (PhilipsXL30 field emission gun scanning electron microscope)for elemental composition and loadings. The bulksamples (0.2 mg) for SEM were dispersed with water(0.05 ml) onto a conductive, silver-painted aluminumsample stand and dried for a minimum of 12 h.

The discrete samples were studied using TEM (JEOL2010F field emission gun analytical electron microscopeor Philips CM12 analytical electron microscope) to iden-tify particle morphology (including particle size) andhigh-resolution TEM (JEOL 3011 high-resolution elec-tron microscope) for detailed examination of latticestructure. Samples were also examined using TEM-EDSfor elemental analysis and the determination of dopantparticle location. All materials were sampled onto copperTEM grids, with the exception of powders which couldcontain copper as an additive. The latter materials weresampled onto nickel TEM grids to eliminate interferenceduring EDS determination of elemental compositions.

III. RESULTS

Four material systems were examined in the study:gold–tin dioxide, palladium–tin dioxide, copper–tin


J. Mater. Res., Vol. 20, No. 11, Nov 2005 2979

dioxide, and aluminum–tin dioxide. Tables I and II sum-marize the operating conditions and particle precursorproperties for each synthesis system. All compressedgases (H2, O2, Ar, N2) were obtained from CryogenicGases, with purities >99.99%. Throughout the study, theH2 and O2 flow rates were set at fixed values of 2.78 and1.46 ml min−1, respectively. Nitrogen was used as ashroud gas for all experiments at a fixed flow rate of28.3 ml min−1. The tetramethyl tin was obtained in liquidform (98% assay, Alfa Aesar, Ward Hill, MA), and theargon flow rate through the TMT bubbler was set at afixed rate of 63.5 ml min−1. All solid-phase precursorreactants were sieved to less than 45 �m (except bis-(dibenzylideneacetone)palladium, which was sieved toless than 125 �m) to facilitate the particle flow throughthe system. The results for each material system are pre-sented below.

A. Gold-doped tin dioxide powders

As seen in Table II, several studies were conducted onthe Au-SnO2 material system. A typical XRD pattern forthe baseline gold acetate/TMT system (Case 2) is pre-sented in Fig. 2. For comparison, spectra for undopedSnO2 obtained at the same baseline synthesis conditionsare provided in the lower half of the figure. Throughoutthe material systems studied in this work, the peaks of theundoped and doped SnO2 indexed to the cassiterite phaseof tin dioxide. Peaks attributable to tin monoxide or me-tallic tin were never observed. No phase changes to theSnO2 were observed by altering the synthesis conditions(e.g., longer residence times, etc.). The additional peaksin the gold acetate/TMT system consistently indexed topure metallic gold.

Figure 3 shows representative TEM images of the Au/SnO2 materials for the baseline gold acetate/TMT system(Case 2). As seen in the images, there are two generalmorphologies present in the Au-doped samples: largerhigh-contrast spherical particles and aggregates consist-ing of small crystalline primary particles. The smallerparticles were identified as tin dioxide and the high-contrast spherical particles were identified as gold using

TEM–EDS analysis. The aggregated structure indicates arelatively slow sintering rate between the SnO2 primaryparticles at the synthesis conditions studied. The mor-phology of the SnO2 particles is consistent with that ob-served previously for synthesis of undoped SnO2 at simi-lar synthesis conditions.35,36,43

As seen in Fig. 3, the larger Au particles were well-integrated into the SnO2 aggregates. The gold particles inthe nanocomposite material system were sparsely located(∼3–7 particles per 25 �m2 area on the TEM grid,Case 2 conditions) and never appeared as gold aggre-gates, indicating that when Au interparticle collisions oc-curred, the sintering between the gold particles wasrapid. The geometric mean diameter of the gold particlesproduced from the baseline gold acetate/TMT system was83 nm, as determined from the TEM images. This averageAu particle size is based on particles that were clearly ob-servable in the TEM images. Smaller Au particles are moredifficult to distinguish from the SnO2 particles without EDSanalysis (as seen in the Case 4 results presented below),consequently the average gold particle size determined inthis manner is considered a high estimate.

The interaction of the gold additives with the SnO2

particles was examined using high-resolution TEM (seeFig. 4). The presence of gold in the SnO2 lattice structurecan cause a dislocation in the tin dioxide lattice, alteringthe d spacings. Consequently, the HRTEM images wereused to measure the spacing between the lattice fringes inthree categories of SnO2 particles: undoped SnO2 par-ticles, Au-doped SnO2 particles far from a gold particle(Case 2 conditions), and Au-doped SnO2 particles adja-cent to additive gold particles (Case 2 conditions). Thespacings were consistent for the first two cases (4 Å), butthe spacings became smaller when the tin dioxideparticles were next to the gold particles (3.5 Å). The Auadditive affects the local crystalline structure of theSnO2, although whether the changes are due to Au mi-gration into the tin dioxide structure or the manner inwhich the SnO2 crystallites form near the Au additives(for example, if the SnO2 particles form by heteroge-neous condensation onto existing Au particles) cannot bedetermined from the HRTEM imaging.

TABLE I. Summary of parametric studies conducted.

Synthesis conditionsCaseno(s).

Chimney(yes/no)

Ar flow rate(l min−1)

Molefraction ofTMT (%)

Baseline undoped SnO2 condition 1 N 17.1 22Baseline doped SnO2 condition 2, 6, 7, 8, 9, 11 N 17.1 22Long residence time conditions 3, 10, 12 Y 17.1 22Lower Ar flow rate to primary flame conditions 4 N 11.4 22Lower TMT conditions 5 N 17.1 5

Note: Case numbers correspond to those listed in Table II. The argon flow rate listed is the flow rate to the primary flame, and the mole fraction of TMTis the amount of TMT in the argon prior to merging with the particle feed flow.



Beyond the baseline gold acetate/TMT conditions,the synthesis environment was examined to determinethe effects of the burner operating conditions on theproperties of the nanocomposites produced. Specifically,

lowering the argon flow rate to the primary flame, the useof the optical chimney and lowering the concentration ofTMT to the secondary flame were each investigatedwhile holding all other operating conditions constant.Lowering the argon flow rate to the primary flame in-creases the temperature of the primary flame and affectsthe residence time of the particles. The use of the quartzchimney extends the high-temperature region above thesurface of the burner. The effects of the changes in op-erating conditions on the average SnO2 particle size arediscussed below in Sec. III E. The general morphology ofthe nanocomposites, Au particles integrated in SnO2 ag-gregates, was unchanged for any of the synthesis condi-tions. However, the use of the optical chimney toextend particle residence times at high temperatures(Case 3) led to larger aggregate structures (see Fig. 5).

TEM-EDS analysis of the materials produced usingthe lower argon flow rate (Case 4) yielded an interestingresult. In addition to the morphology of Au particlesintegrated into the SnO2 aggregates, Au signals weredetected from aggregates where no discrete Au particlecould be identified. Figure 6 shows an example of suchan aggregate. This indication of smaller Au particles orthe possibility of Au within the SnO2 crystallites is notlikely to be exclusive to Case 4 synthesis conditions, asis discussed further in Sec. IV.

Gold particles (aerodynamic particle sizes of 0.5–0.8 �m,

FIG. 2. XRD spectra of Au-doped SnO2 powder (Case 2 conditions,produced using gold acetate as the additive precursor). XRD spectrafor undoped SnO2 (Case 1) are provided for comparison in the lowerportion of the figure. The diffraction patterns and lattice parameters forthe cassiterite phase of SnO2 (21-1250)42 and metallic gold (4-784)42

are also indicated for reference.

TABLE II. Synthesis conditions studied and results for average SnO2 crystallite size determined by analysis of XRD spectra.

Caseno. Synthesis conditions

Materialsystem Additive precursor

AverageSnO2

crystallitesize (nm)

1 Baseline undoped SnO2 SnO2 None 12 ± 1.02 Baseline doped SnO2 SnO2/Au Gold acetate (Au(O2CCH3)3)

(Alfa Aesar, 99.9%, sieved to <45 �m)9.6 ± 1.1

3 Long residence time SnO2/Au Gold acetate (Au(O2CCH3)3)(Alfa Aesar, 99.9%, sieved to <45 �m)

9.2 ± 0.2

4 Lower Ar flow rate toprimary flame

SnO2/Au Gold acetate (Au(O2CCH3)3)(Alfa Aesar, 99.9%, sieved to <45 �m)

13 ± 0.5

5 Lower TMT SnO2/Au Gold acetate (Au(O2CCH3)3)(Alfa Aesar, 99.9%, sieved to <45 �m)

5.8 ± 1.1

6 Baseline doped SnO2 SnO2/Au Gold (Au)(Alfa Aesar, 99.96%, APS 0.5–0.8 �m*)

...

7 Baseline doped SnO2 SnO2/Pd Palladium acetate (Pd(O2CCH3)2)(Sigma-Aldrich, 98%, sieved to <45 �m)

8.6 ± 0.3

8 Baseline doped SnO2 SnO2/Pd Palladium (Pd)(Alfa Aesar, 99.95%, APS 1.0–1.5 �m*)

17.0 ± 2.2

9 Baseline doped SnO2 SnO2/Pd Bis(dibenzylideneacetone) palladium, (Pd(C17H14O)2)(Alfa Aesar, sieved to <125 �m)

7.6 ± 0.4

10 Long residence time SnO2/Pd Bis(dibenzylideneacetone) palladium (Pd(C17H14O)2)(Alfa Aesar, sieved to <125 �m)

11 ± 0.3

11 Baseline doped SnO2 SnO2/CuxO Copper acetate (Cu(O2CCH3)2)(Alfa Aesar, 99.999%, sieved to <45 �m)

16 ± 0.3

12 Long residence time SnO2/AlxOy Aluminum acetate ((C2H3O2)2AlOH)(Sigma-Aldrich, sieved to <45 �m)

6.9 ± 0.1

*As per manufacturer specifications.



as per manufacturer specifications) were also examinedas a potential direct source for Au additives in the tindioxide powders. For these experiments, the majority ofthe precursor gold particles appeared to flow through thesystem with little interaction with the TMT/SnO2 system.TEM images of the product powder confirm few Auparticles integrated with SnO2 aggregates, and the Auparticles were typically larger than 250 nm in diameter.

FIG. 5. TEM image of Au-doped SnO2 nanocomposite (Case 3 con-ditions, produced using gold acetate as the additive precursor).

FIG. 3. TEM images of Au-doped SnO2 powders (Case 2 conditions,produced using gold acetate as the additive precursor). The high-contrast spherical particles are Au, as identified using TEM-EDS.

FIG. 4. HRTEM image of Au-doped SnO2 nanocomposite (Case 2conditions, produced using gold acetate as the additive precursor). Thedark region is an Au particle. The lattice fringe spacings of the tindioxide particles (identified in the image) were used to examine theeffects of the Au additives on the SnO2 crystallite structure.



Gold particles are often used to color glass, and thesize of the gold particles controls the color, where 10 nmgold particles result in a pink glass, 10–20 nm gold par-ticles result in a violet-red glass, and 20–50 nm goldparticles result in a deep purple glass.44 The undopedSnO2 powders produced in this work were generallygray/white. The Au-doped SnO2 powders produced usinggold acetate were pink. The Au-SnO2 materials producedusing pure gold as the additive precursor were more goldthan pink, consistent with less integration of the goldwith the SnO2 and larger gold particle sizes.

B. Palladium-doped tin dioxide powders

Three palladium precursors were examined as SnO2

additives in this work: palladium acetate, pure palladiumand bis(dibenzylideneacetone)palladium. A typical XRDpattern for the baseline palladium acetate/TMT system(Case 7 conditions) is shown in Fig. 7. The peaks of theXRD spectra index to the cassiterite phase of tin dioxideand metallic palladium. No palladium oxides were ob-served. The morphology of the Pd-SnO2 nanocompositesis similar to that of the gold acetate/TMT system, as seenin Fig. 8 where palladium particles are integrated intoSnO2 aggregates. The composition of the palladium par-ticles was confirmed via TEM-EDS, and the TEM im-ages indicated palladium particles on the order of 50 nm.

Palladium particles (aerodynamic particle sizes of 1.0–1.5 �m as per manufacturer specifications) behaved insimilar fashion to the pure gold precursor materials. Thepalladium appeared to have little interaction with theSnO2 based on TEM imaging which showed only fewlarger (on the same order as the precursor Pd material)

palladium particles mixed with the SnO2. However,analysis of the XRD spectra showed that the averageSnO2 crystallite size was affected by the Pd precursor,which is discussed further below.

A larger metal organic precursor for palladium,bis(dibenzylideneacetone)palladium, was examined tofurther investigate the effects of precursor decompositionchemistry on the Pd-SnO2 nanocomposite properties.When the baseline SnO2-doped synthesis conditionswere used (Case 9), no palladium signals were observedusing either XRD or TEM-EDS. When the high-temperature regions were extended using the opticalchimney (Case 10 conditions), the XRD pattern did in-dicate the presence of palladium in the product materials.However, TEM imaging with EDS analysis failed toidentify any discrete Pd particles.

C. Copper-doped tin dioxide powders

Representative XRD spectra of copper acetate-TMTmaterials are shown in Figure 9. The peaks index to thecassiterite phase of tin dioxide and a mix of CuO andCu2O (cupric oxide and cuprous oxide, respectively).TEM imaging, as shown in Fig. 10, indicates similarintegrated aggregate and additive morphology to that ob-served with the previous metal acetate precursors. On thebasis of the TEM images and EDS analysis, the particlescontaining copper were on the order of 50 nm and some-what faceted in shape.

D. Aluminum-doped tin dioxide powders

Last in the series of metal acetate additive precursorsexamined was aluminum acetate. XRD analysis of the

FIG. 6. TEM image of Au-doped SnO2 nanocomposite (Case 4 con-ditions, produced using gold acetate as the additive precursor). EDSanalysis of the aggregate indicated the presence of Au in the structure.

FIG. 7. XRD spectra of Pd-doped SnO2 powder (Case 7 conditions,produced using palladium acetate as the additive precursor). The dif-fraction patterns and lattice parameters for the cassiterite phase ofSnO2 (21-1250)42 and metallic palladium (Pd, 5-681)42 are providedfor reference.



materials produced indexed to the cassiterite phase oftin dioxide; however, no aluminum or aluminum oxidepeaks could be identified in the XRD spectra. Similarly,the TEM imaging showed aggregated SnO2 particleswith no EDS peaks attributable to aluminum.

E. Average SnO2 crystallite size

The average crystallite size of the tin dioxide particleswas determined for the material systems and synthesisconditions studied using the Scherrer equation:

dp =0.9�

�1�2 cos�,

where dp is the average crystallite size, � is the wave-length of the source emission (1.542 Å for this study),�1/2 is the full width at half-maximum of the peak usedfor the analysis, and � is the XRD scattering angle of thepeak. Here, the 110 feature of the SnO2 spectra was used,

FIG. 8. TEM images of Pd-doped SnO2 nanocomposites (Case 7 con-ditions, produced using palladium acetate as the additive precursor).EDS analysis of the particles identified in (a) by the arrows indicatesthe presence of palladium. EDS analysis of the aggregate shown in (b)also indicates the presence of Pd.

FIG. 9. XRD spectra of CuxO-doped SnO2 powder (Case 11 condi-tions, produced using copper acetate as the additive precursor). Thediffraction patterns and lattice parameters for the cassiterite phase ofSnO2 (21-1250),42 copper(II) oxide (cupric oxide, Cu, 5-661),42 andcopper(I) oxide (cuprous oxide, Cu2O, 5-667)42 are provided for ref-erence.

FIG. 10. TEM image of CuxO-doped SnO2 nanocomposites (Case 11conditions, produced using copper acetate particles as the additiveprecursor). EDS analysis of the particles identified in the image by thearrows indicates the presence of copper.



where � � 13.3°,45 although additional peaks were ex-amined for some spectra to verify the dp determinations.A summary of the XRD results for dp is provided inTable II. Note that the signal for the (110) peak for SnO2

was too low for the pure gold/TMT system to obtain areasonable estimate for dp. Repeatability studies wereperformed for the undoped and gold acetate-doped tindioxide systems. The samples showed little experiment-to-experiment variability (<10%). The total uncertaintyin the crystallite size measurements for each system isprovided in Table II, where the total uncertainty is pri-marily due to experimental variability and the curve-fitting process used to determine the peak height.

As seen in Table II, the average SnO2 crystallite sizespans 5.8–17 nm, with the gold acetate/lower TMT sys-tem (Case 5) yielding the smallest SnO2 particles and thepure palladium/TMT system (Case 8) yielding the largestparticles. In general, the metal acetates led to smalleraverage SnO2 particle sizes compared to the undopedSnO2 (which had an average crystallite size of 12 nm),with the exception of the copper acetate precursor and thegold acetate precursor/lower argon flow to the primaryflame conditions (Case 4). As stated earlier the lowerargon flow leads to higher primary flame temperaturesand potentially longer particle residence times. Thelarger average SnO2 crystallite size is consistent withadditional sintering that would occur with longer times athigh temperatures. Note also that the average crystallitesize for the undoped SnO2 is consistent with previousdirect measurements made using TEM imaging.36

Table II includes interesting results for the averageSnO2 crystallite size when the TEM, EDS, and XRD dataare also considered. Specifically, the pure palladium ap-peared to have little integration with the SnO2 particlesbased on TEM imaging. However, the pure palladiumprecursor had a marked affect of increasing the averageSnO2 crystallite size compared to both undoped anddoped baseline synthesis conditions. Additionally, al-though aluminum was not detected via TEM-EDS orXRD spectra in the materials produced using aluminumacetate, the average SnO2 crystallite size for these mate-rials was dramatically reduced compared to undopedSnO2. The bis(dibenzylideneacetone)palladium precur-sor had a similar effect, where palladium did not appearin the XRD or TEM-EDS spectra for Case 9, but theaverage SnO2 crystallite size was reduced compared topure SnO2.

F. Additive loadings

The additive loadings within the nanocomposite ma-terials were determined using SEM-EDS for selected ma-terials and conditions. The results are summarized inTable III. Molar Sn/O ratios of 1:2 confirmed the tindioxide stoichiometry for the palladium acetate and goldacetate materials. The oxygen concentrations were higher

for the copper acetate materials, which is consistent withthe forms of copper oxide observed in the XRD spectra.The loadings of the additives are comparable with thoserequired to achieve significantly improved SnO2 sensorperformance.4

IV. DISCUSSION

The location of the metal additives in the nanocom-posite system can be as critical as the composition andphase of the additives, particularly for tin dioxide sen-sors as demonstrated by Bittencourt et al.46 and Cabotet al.11,29 For example, Bittencourt et al.46 studied thedifferences in sensor performance between uniform dis-tributions of Pt additives in the SnO2 versus Pt additivessputtered onto the surface of thick-film SnO2 sensors.When the Pt additives were sputtered, the Pt particleswere within 3 �m of the surface of the film, which were20–25 �m thick. The sensors made with uniformly dis-tributed Pt particles considerably outperformed (by ap-proximately a factor of 4) the films where the Pt additiveswere localized near the surface of the sensor.

Cabot et al.29 examined the effects of the location ofadditives in SnO2 nanocomposites at a higher resolutionlength scale. Cabot and co-workers determined thatdoped SnO2 synthesized using sol-gel methods led tonanocomposites where the Au additives were present asAu nanoparticles integrated with the SnO2 aggregates (aswere observed in this work) and as smaller concentra-tions of Au distributed on the SnO2 crystallites. Theyfound TEM imaging did not capture the presence of thesesmaller units of gold using either bright-field imaging orGatan Image Filtering techniques. Cabot et al.29 attrib-uted the modifications to the SnO2 sensor performance tothe smaller more localized Au additives.

Several factors indicate the metals additives are likelypresent within the SnO2 for the nanocomposites pro-duced in the current work. First, the size of the SnO2

primary particles can be an indication of the presence ofadditives in the structure of the SnO2 nanoparticles.Metal additives are well known to modify the SnO2 graingrowth kinetics. Specifically, when the crystallite size isreduced compared to undoped SnO2, the inhibited-growth effect has been attributed to the concentration ofadditives localized on the oxide surface.29 Second, usingthe aluminum acetate precursor (Case 12), yielded ma-terials where metal particles could not be identified in theSnO2 aggregates using TEM-EDS or XRD. Studies ofnickel-doped SnO2 by Epifani et al.47 using sol-gel proc-essing yielded similar results, where the grain size of theSnO2 was reduced, but no dopant particles or dopantXRD peaks were detected. Given the high boiling pointsof the metal and metal oxides for this system (e.g., Tbp,Al

� 2519 °C, Tbp,Al2O3 ≅ 3000 °C49), condensed phasealuminum or aluminum oxide particles should be present



in the materials sampled on the cold plate and the TEMgrids (which were located in the exhaust region of theburner). Additionally, when only aluminum acetate wasburned (without TMT), the nanoparticles producedshowed the presence of aluminum using TEM-EDSanalysis. The aluminum acetate clearly affected the av-erage crystallite size for SnO2 (decreasing the averagecrystallite size by nearly a factor of 2 compared to theundoped SnO2 case). In all, the data lead to the conclu-sion that aluminum is present in the SnO2 at levels thatare below the detectible limit for TEM-EDS and XRDanalyses. Similar arguments can be applied to the resultsfor Cases 9 and 10, where bis(dibenzylideneacetone)pal-ladium was used as the additive precursor.

The third line of reasoning supporting the presence ofthe additive materials in the SnO2 nanoparticles is due tothe high-temperature combustion environment used tocreate the materials. Metal acetates generally decomposeat low temperatures (e.g., the decomposition temperatureof palladium acetate is 220 °C48), and the vapor pressureof the pure metals can be high at the temperatures foundin the combustion synthesis environment (e.g., aluminumhas a vapor pressure of 100 Pa at 1544 °C).49 Conse-quently, metal and metal organic species are likely pres-ent in the gas phase. The high number density of SnO2

particles provides sites for condensation of the metalsand conversely, condensed metal particles provide sitesfor SnO2 deposition and SnO2 particle growth.

V. CONCLUSIONS

A broad range of nanocomposite materials can be pro-duced using combined gas- and solid-phase precursors ina combustion synthesis system. The combustion synthe-sis method presented here exhibited significant flexibilitywith respect to control of the nanocomposite properties,such as the average SnO2 crystallite size and the type ofadditive. For example, changing the operating conditionsof the reactor (by decreasing the initial TMT reactantconcentrations for the TMT/gold acetate system, Case 5)allowed over a factor of two decrease in the averageSnO2 crystallite size (compared to the higher-temperature TMT/gold acetate system, Case 4), from 13to 5.8 nm. Similarly, changing the solid-phase precursorcan be used to control the average crystallite size of the

SnO2. For example, the TMT/bis(dibenzilideneacetone)palladium precursor yielded an average SnO2 crystallitesize of 7.6 nm. The TMT/pure metallic palladium yieldedan average SnO2 crystallite size of 17 nm. The results ofthis study and the level of control of the nanocompositeproperties indicates that combined-phase combustionsynthesis methods have potential to create a broad rangeof nanocomposite materials, considerably beyond thescope of the SnO2 sensing materials presented in thiswork.

ACKNOWLEDGMENT

The authors acknowledge the generous support of theNational Science Foundation (DMI 0329631).

REFERENCES

1. T. Uematsu, L. Fan, T. Maruyama, N. Ichikuni, and S. Shimazu:New application of spray reaction technique to the preparation ofsupported gold catalysts for environmental catalysis. J. Mol.Catal. A: Chem. 182, 209 (2002).

2. C.N. Afonso, J. Solis, R. Serna, J. Gonzalo, J.M. Ballesterosa, andJ.C.G. de Sande: Pulsed laser deposition of nanocomposite thinfilms for photonic applications. Proc. SPIE Int. Soc. Opt. Eng.3618, 453 (1999).

3. Z. Cui and Z. Zhang: Ce-Ni nanoparticles with shell structure forhydrogen storage. Nanostruct. Mater. 7, 355 (1996).

4. T.A. Miller, S.D. Bakrania, C. Perez, and M.S. Wooldridge: Tindioxide/metal/metal oxide nanocomposites for gas sensor ap-plications. In Functional Nanomaterials, edited by E. Rosenbergand K.E. Geckeler (American Scientific Publishers, StevensonRanch, CA) (in press).

5. N. Yamazoe: New approaches for improving semiconductor gassensors. Sens. Actuators, B 5, 7 (1991).

6. F.E. Krius, H. Fissan, and A. Peled: Synthesis of nanoparticles inthe gas phase for electronic, optical and magnetic applications—Areview. J. Aerosol Sci. 29, 511 (1998).

7. W. Göpel and K.D. Schierbaum: SnO2 sensors: Current status andfuture prospects. Sens. Actuators, B 26, 1 (1995).

8. E. Kanazawa, M. Kugishima, K. Shimanoe, Y. Kanmura,Y. Teraoka, N. Miura, and N. Yamazoe: Mixed-potential typeN2O sensor using stabilized zirconia- and SnO2-based sensingelectrode. Sens. Actuators, B 75, 121 (2001).

9. I. Sayago, J. Gutiérrez, L. Arés, J.I. Robla, M.C. Horrillo,J. Getino, J. Rino, and J.A. Agapito: Effect of additives in tin ox-ide on the sensitivity and selectivity to NOx and CO. Sens. Ac-tuators, B 26, 19 (1995).

10. P. Siciliano: Preparation, characterisation and applications of thinfilms for gas sensors prepared by cheap chemical method. Sens.Actuators, B 70, 153 (2000).

TABLE III. Average additive loadings in the SnO2 nanocomposites as determined using SEM XEDS analysis.

Caseno.

Materialsystem

Additiveprecursor

Metal/Sn ratio

Average(mole basis)

Standarddeviation

(mole basis)Average

(mass basis)

Standarddeviation

(mass basis)

2 SnO2/Au Gold acetate 0.059 0.015 0.097 0.0247 SnO2/Pd Palladium acetate 0.268 0.036 0.24 0.032

11 SnO2/CuxO Copper acetate 0.217 0.025 0.116 0.013



11. A. Cabot, J. Arbiol, J.R. Morante, U. Weimar, N. Bârsan, andW. Göpel: Analysis of the noble metal catalytic additives intro-duced by impregnation of as obtained SnO2 sol-gel nanocrystalsfor gas sensors. Sens. Actuators, B 70, 87 (2000).

12. C.G. Borman and R.G. Gordon: Reactive pathways in the chemi-cal vapor deposition of tin oxide films by tetramethyltin oxidation.J. Electrochem. Soc. 136, 3820 (1989).

13. H. Unuma, H. Tahabatake, K. Watanabe, T. Ogata, andM. Sugawara: Preparation of SnO2 thin films by the oxidative-soak-coating method. J. Mater. Sci. Lett. 21, 1241 (2002).

14. A.K. Mukhopadhyay, P. Mitra, A.P. Chatterjee, and H.S. Maiti:Tin dioxide thin film gas sensor. Ceram. Int. 26, 123 (2000).

15. J. Calderer, P. Molinàs, J. Sueiras, E. Llobet, X. Vilanova,X. Correig, F. Masana, and A. Rodríguez: Synthesis and charac-terization of metal suboxides for gas sensors. Microelectron. Re-liab. 40, 807 (2000).

16. Y-S. Choe: New gas sensing mechanism for SnO2 thin-film gassensors fabricated by using dual ion beam sputtering. Sens. Ac-tuators, B 77, 200 (2001).

17. V.M. Jiménez, A.R. González-Elipe, J.P. Espinós, A. Justo, andA. Fernández: Synthesis of SnO and SnO2 nanocrystalline pow-ders by the gas phase condensation method. Sens. Actuators, B 31,29 (1996).

18. S. Nicoletti, L. Dori, G.C. Cardinali, and A. Parisini: Gas sensorsfor air quality monitoring: Realization and characterization of un-doped and noble metal-doped SnO2 thin sensing films depositedby the pulsed laser ablation. Sens. Actuators, B 60, 90 (1999).

19. L.M. Cukrov, T. Tsuzuki, and P.G. McCormick: SnO2 nanopar-ticles prepared by mechanochemical processing. Scripta Mater.44, 1787 (2001).

20. D. Lindackers, C. Janzen, B. Rellinghaus, E.F. Wassermann, andP. Roth: Synthesis of Al2O3 and SnO2 particles by oxidation ofmetalorganic precursors in premixed H2/O2/Ar low pressureflames. Nanostruct. Mater. 10, 1247 (1998).

21. W. Zhu and S.E. Pratsinis: Synthesis of SiO2 and SnO2 particlesin diffusion flame reactors. AIChE J. 43, 2657 (1997).

22. G. Skandan, N. Glumac, Y.-J. Chen, F. Cosandey, E. Heims, andB.H. Kear: Low-pressure flame deposition of nanostructured ox-ide films. J. Am. Ceram. Soc. 81, 2753 (1998).

23. M.S. Wooldridge: Gas-phase combustion synthesis of particles.Prog. Energy Combust. Sci. 24, 63 (1998).

24. S.E. Pratsinis: Flame aerosol synthesis of ceramic powders. Prog.Energy Combust. Sci. 24, 197 (1998).

25. R.M. Laine, R. Baranwal, T. Hinklin, D. Treadwell, A. Sutorik,C. Bickmore, K. Waldner, and S.S. Neo: Making nanosized oxidepowders by flame spray pyrolysis, in Novel Synthesis and Proc-essing of Ceramics, edited by H. Suzuki, K. Komeya, and K.Uematsu (Key Engineering Materials Trans. Tech. Publ. Ltd.,Zurich, Switzerland, 1998), p. 17.

26. R.M. Laine, K. Waldner, C. Bickmore, and D.R. Treadwell:Ultrafine powders by flame spray pyrolysis. U.S. Patent No. 5 958361, September 28, 1999.

27. G. Skandan, N. Glumac, Y-J. Chen, F. Cosandey, E. Heims, andB.H. Kear: Low-pressure flame deposition of nanostructured ox-ide films. J. Am. Ceram. Soc. 81, 2753 (1998).

28. C. Bittencourt, E. Llobet, P. Ivanov, X. Correig, X. Vilanova,J. Brezmes, J. Hubalek, K. Malysz, J.J. Pireaux, and J. Calderer:Influence of the doping method on the sensitivity of Pt-dopedscreen-printed SnO2 sensors. Sens. Actuators, B 97, 67 (2004).

29. A. Cabot, A. Diéguez, A. Romano-Rodríguez, J.R. Morante, andN. Bârsan: Influence of the catalytic introduction procedure on thenano-SnO2 gas sensor performances: Where and how stay thecatalytic atoms. Sens. Actuators, B 79, 98 (2001).

30. N. Sergent, P. Gélin, L. Périer-Cambry, H. Praliaud, andG. Thomas: Preparation and characterisation of high surface area

stannic oxides: Structural, textural and semiconducting properties.Sens. Actuators, B 84, 176 (2002).

31. J. Zhang and L. Gao: Synthesis and characterization of nanocrys-talline tin oxide by sol-gel method. J. Solid State Chem. 177, 1425(2004).

32. C. Xu, J. Tamaki, N. Miura, and N. Yamazoe: Grain size effectson gas sensitivity of porous SnO2-based elements. Sens. Actua-tors, B 3, 147 (1991).

33. G. Zhang and M. Liu: Effect of particle size and dopant on propertiesof SnO2-based gas sensors. Sens. Actuators, B 69, 144 (2000).

34. M.S. Wooldridge, P.V. Torek, M.T. Donovan, D.L. Hall,T.A. Miller, T.R. Palmer, and C.R. Schrock: An experimental in-vestigation of gas-phase combustion synthesis of SiO2 nanopar-ticles using a multi-element diffusion flame burner. Combust.Flame 131, 98 (2002).

35. D.L. Hall, P.V. Torek, C.R. Schrock, T.R. Palmer, andM.S. Wooldridge: Gas-phase combustion synthesis of tin oxidenanoparticles. Proceedings of the 2001 International Symposiumon Metastable, Mechanically Alloyed, and Nanocrystalline Mate-rials; (Trans. Tech. Publ., Zurich, Switzerland, 2002), p. 347.

36. D.L. Hall, A.A. Wang, K.T. Joy, T.A. Miller, and M.S. Wooldridge:Combustion synthesis and characterization of nanocrystalline tinand tin oxide (SnOx, x � 0–2) particles. J. Am. Ceram. Soc. 87,2033 (2004).

37. J. Ma: Soot formation during coal pyrolysis. Ph.D. Dissertation,Chemical Engineering Department, Brigham Young University,Salt Lake City, UT, 1996.

38. S. Niksa, R.E. Mitchell, K.R. Hencken, and D.A. Tichenor: Opti-cally determined temperatures, sizes, and velocities of individualcarbon particles under typical combustion conditions. Combust.Flame 60, 183 (1985).

39. T.A. Miller: Combustion synthesis of metal/metal oxide nanocom-posite materials. Ph.D. Dissertation, University of Michigan, AnnArbor, MI.

40. T.A. Miller, C.H. Chu, and M.S. Wooldridge: Demonstration of aparticle feed system for combustion synthesis of metal and metaloxide materials, Presented at the 2003 Technical Meeting of theEastern States Section of the Combustion Institute, State College,PA, October, 2003.

41. T.A. Miller, S.D. Bakrania, C. Perez, and M.S. Wooldridge: Anexperimental investigation of the use of solid-phase precursors fordirect synthesis of doped tin dioxide nanocomposite powders.Combust. Flame 2005 (submitted).

42. JCPDS Powder Diffraction Files: 21-1250, 4-784, 5-661, 5-667,5-681. International Centre for Diffraction Data: Newton Square,PA, 1990.

43. D.L. Hall: Gas-phase combustion synthesis of nanocrystalline tinand tin oxide particles. M.S. Thesis, Mechanical Engineering De-partment, University of Michigan, Ann Arbor, MI, 2001.

44. N.R. Wilk, Jr. and H.D. Schreiber: Optical properties of gold inacetate glasses. J. Non-Cryst. Solids 239, 192 (1998).

45. K. Ikohura and J. Watson: The Stannic Oxide Gas Sensor (CRCPress, Boca Raton, FL, 1994).

46. C. Bittencourt, E. Llobet, P. Ivanov, X. Correig, X. Vilanova,J. Brezmes, J. Hubalek, K. Malysz, J.J. Pireaux, and J. Calderer:Influence of the doping method on the sensitivity of Pt-dopedscreen-printed SnO2 sensors. Sens. Actuators, B 97, 67 (2004).

47. M. Epifani, M. Alvisi, L. Mirenghi, G. Leo, P. Siciliano, andL. Vasanelli: Sol-gel processing and characterization of pure andmetal-doped SnO2 thin films. J. Am. Ceram. Soc. 84, 48 (2001).

48. E. Makino and T. Shibata: Micromachining compatible metal pat-terning technique using localized decomposition of an organome-tallic compound by laser irradiation. J. Micromech. Microeng. 8,177 (1998).

49. CRC Handbook of Chemistry and Physics, edited by D.R. Lide(CRC Press, Boca Raton, FL, 1997).



Documents

A new method for direct preparation of tin dioxide ...mswool/publications/jmr_official_reprint.pdf · A new method for direct preparation of tin dioxide nanocomposite materials T.A