STUDY ON NANOMETER ZINC PARTICLES FABRICATED BY GAS EVAPORATION METHOD

ACTA METALLURGICA SINICA (ENGLISH LETTERSVol 16 No. 3 pp 161-168 June 2003STUDY ON NANOMETER ZINC PARTICLES FABRICATED BYGAS EVAPORATION METHODManuscript, received 19 September 2002; in revised form 6 January 200, eiaA.A.A. Salch. X.. Zhai, Y.C. Zhai.Y. Fu and A.L. ZhatSchool of Materials and Metallurgy, Northeastern University, ShenyangNanometer in particles urith mean diameters 12-100nm made by evaporating its pow-ders in argon gas were studied mainly by X-ray diffraction and electron microscopyThey are collected at various distances and those factors infiuencing the mean particlesize were studied. The optimal synthetic conditiong were obtained, i. e, evaporationtemperature is 1200C; argon Row rate is 0.4m/h; amount of pounder charged is 3gdistance from evaporation source is 10cm. It was found that the size of particles wasgoverned by argon flow rate, evaporation temperature, amount of metal charged anddistance from the source. The size increases remarkably with distance in the spacewhere no metal vapor exists. This implies that the crystallites grow by coalescence.Electron micrographs and diffraction patterns are reproduced to show the size, shapeand state of oxidations. Nanoparticles with definite crystal habits were sometime.observed among those with irregular onesKEY WORDs gas evaporation method, nanometer Zn, XRD,2'EM1. IntroductionNanomaterials are those with a crystal sizc in the region of 1-100nm, which can beproduced by gas evaporation method(1-41. This method has had a rather long historystretching back to the formation and use of"smokes"such as carbon or bismuth"blacksfor a variety of applications(5-71. PfundlBI evaporated zinc in air at low pressure andobtained a soot like substances zinc black, which was used as an absorber of infraredradiation. Uyeda and Kimotol9) was able to vary the average particle size by changingthe gas pressure, Wadaobserved and qualitatively explained variations of particle sizesoccurring when different noble gases were usedOf many methods to prepare nanometer Zn particles, the gas evaporation method seemssurface, but production cost is fairly high. The chemical methods take precedence overthe former in a large amount produced at low cost. The gas evaporation method can buThe principle is very simple and dircct, but there exist many difficulties in practice]It can be used most frequently for the production of dry (i.e. not in liquid suspensionmetal nanoparticles with very low impurity content is based on the condensation of ace ofling gaelium that acts both as a cooling agent and a carril山中国煤化工mevapour nucleates homogeneously as a result of collCNMH Gncrt gas atomsGrowth of particles within the convective gas flow occurs via coalcscenceli5-221162Previous investigators[5-6, 23-25, have prepared ultrafine Zn particles by evaporatingfrorn tungsten basket, spiral, conical, helix or boat tungsten heaters were adopted, while inthe present investigation, nanometer Zn particles was produced by the direct evaporationfrom crucible in which the heat source was an electric furnace; experimental conditions forpreparation of nanometer Zn have been studied first; then observation by X-ray diffraction and transmission electron microscopy(TEM) have been made about structure andInorphology of Zn particles condensed onto inner wall and onto collector2. ExperimentalThe powder of zinc metal used in the present study was of purity 99.994% and that ofargon gas, of 99.99%. An alumina crucible containing about 5 to 15g of metal zinc powderof 370um in mean grain size was placed in an alumina tube This outer alumina tube wassct in a horizontal electric furnace. The work chamber, in which the particles were made,was a metal cylinder whose inner diameter and height were 360cm and 500cm, respecc-olletThis chamber was evacuated and back filled with argon gas through a sensor iow controllerup to the desired pressure during the condensation. Water was used to cool inner wall ofwork chamber to make deposition easy. The temperature was monitored by a Pt-ptrhthermocouple attached to the bottom, and was controlled within +10C. a new techniqueof collecting samples at various heights in a vapor was developed, and confirmed that theparticle size was different in different heights. A simple device 50cm in height consists offive sheets of stainless steel which were placed at a distance of 10cm from each other andwere supported and welded to a double tube, thus allowing the water Hows through oncedownward and back through other upward. These sheets were made in a rectangular case6cm in length and 1.5cm in width and they cooled by water to promote the sticking ofparticles on them. This collector was placed vertically in the work chamber. The verticaldistance from the crucible to the collection positions was fixed. Vapor particles, which hitthe rectangle condensed there as well as they condensed on the wall of the work chamber.The structures of the samples collected from the inner wall of the work chamber and fromthe collector wore identified by X-ray diffraction with Cuka radiation in the 20 rangeof 25 to 60. Samples for TEM were prepared by dispersing particles in HC3HC2OH(2.5mol ethanol)solution using ultrasonic bath( DXM-CSPR III) for approximately 1h tobreak down the agglomerates. A microscope grid was dipped in the suspension and slowlywithdrawn and, after drying under lamp it was ready for viewing. This technique is calledrandom sampling. The observation of morphology of the particles and electron diffractionpatterns was carried out on a TEM(Philips; Model EM400T, Holland) operating with aighest accelerating voltage of 100kv. Micrographs were taken at different magnificationsaccording to the particle size. The average particle size was evaluated by XRD and TEMThe particle sizes range from 12 to 100nm, depending on the production conditions. Therange of particle sizes was remarkable3. Result and Discussionblck\s soon as the zinc powder was inserted into tl中国煤化工 pour of the zincck was observed to start from the vent, in theCNMH Gy a single zoneThe lack of the outer zone may be naturally understood because no vapor is allowed to godownward in the case of a crucible as there is a strong upwards convection current, alongthe outside of crucible. The inner front possibly corresponds to the outer region of thesingle-zone. The colour of the particles did not vary by varying the evaporation condition;this means that the colour of the deposited particles remained the same, the normal colourof nanometer Zn particlesll4). Typical value of the density of the Zn nanopowders was3.523g/cm. The presence of oxygen before diffraction study might lead to the formationof zinc oxidc which in turn would, to a smaller density or the extremely low value mightbe resulted from the irregular surface morphologyIn consideration of the influence of the various factors, dependence of the crystal sizeand habit on the position of collection is remarkable. As a result, the size increases withdistance from the source until saturation sets at a, certain distance as seen in Fig. 1, whichindicates that the particles grow where no vapor exists. This means that the particles growy ordinary mechanism of vapor growth only during the initial stage of gas evaporationFurther growth takes place by the coalescence whcn thcy collide with one another in theupper part of a smoke where is no vapor. Coalescence takes place provided that thetemperature is high enough and the surface is clean. Since the surface oxide hinderscoalescence, irregular shapes might be found in an impure atmosphere/ l. When particlesare collected at affixed point, their size varies with the evaporation ternperature, the gaspressure and the amount of metal charged, as shown in Figs. 2, 3 and 4, respectively. Thesize seemed to become larger when evaporation was carried out at high temperature orwhen the evaporated amount of metal was larger, however, at higher source temperaturethe density of particles with the growth layer is larger, so that coalescence occurs moreto form large crystals by means of an increase of the amount charged due lo was neededfrequently. As the amount of the charged zinc powder increased, longer timeevaporation time, and extended the growi h time, thus large particles were forned. Bvarying the gas pressure, the size of particles is changed in a controlled way. The particlesize was small at low pressures and bccomc larger at higher pressures/19-21, 251. At the argonAow rate of 0.2 to 0.4m /h, the mean diameter was decreased as a result of diminishingrowth times by rapid cooling of argon gas, which freezes the structure of primary particles.With increasing argon flow rate from 0.4 to 1.0m /h, the mean diameter becomes largerhese results, we believe, can be explained simply by the higher gas pressurc yieldingmore cfficicnt confinement of the particles to the growth region. Specifically, the size wasabout 62nrn at 0.2m/h and increased to 94nm at 1.0m/h. On the other hand, the lowAow rate of argon gas is due to significant, reduction in collision of hot "particles, andhence inhibiting agglomerationThe powder X-ray diffraction pattern of the nanometer Zn(Fig5a)showed four strongreflections with plane spacings of 2. 4767, 2. 3110, 2.0934 and 1.6886, each of which corresponds to(002),(100),(101)and(102)planes of closed-backed hexagonal (hcp)Znstructure and amorphous Zno. The different peaks were narrow. It showed that it wasZn. The strong intensity of diffraction peaks indicates that the as-prepared zn nanometeris well crystallized, while the low intensity and low background around 31.5 and 34.5,of 20, particularly, for the peaks due to 1 and 2 planes. This suggests that the expan-sion of the presence of a ccrtain amount of zinc oximetallic nanormaterials(7, 24. The characteristic pe!中国煤化 served in otherEn the Fig. 5b isobserved at20=36°,388°,43°and54.2°. The paYHCNMHGaks 1,2,6l648 is not consistent with Zn and may, due to ZnO. The general features of the diffractionpattern are not the same as seen in Fig 5a. There are marked decreases and increases inthe peak intensities as well as grain sizes, specifically, 98nm and 66nm respectively. Thedisagreement between them is attributed to the different preparation conditions employedevaporation temperature, distance and amount of metal charged. However, all the sampleshaving similar oxidation properties are detected as ZnOIt is of great interest to study crystallites having clear-cut habits formed in an ex-tremely clean atmosphere, but the formation of clear-cut habits is disturbed by slightoxidation or surface adsorption, which reduces the anisotropy of the surface energy. Eachzinc crystal must be covered with a zinc oxide layer as thin as a few nanometer when samples were transferred through air before diffraction study. This slight amount of oxygenmay coat the nucleated particles epitaxially and thereby prevent further growth except byagglomeration (23 However, the oxygen content in nanometer zinc prepared by gas evapo-ration is so small, which is due to the fact that the whole preparation process was carriedout in the argon gas. Despite the fact that the oxidation is the most remarkable duringfabrication, it protects the nanoparticles from further oxidation. The oxygen content does9097050010203040506090010001100120013001400Evaporation temperature.ocFig 1 Mean diameters of particles collected atFig 2 Relation between the thean diametersvarious distances(evaporation conditionsand the evaporation temperature(amountamount of metal charged 10g)04m3/h)00800049800000000002040.608120246810121416Argon gas fow rate, m/hAmount of metal charged, gFig 3 Mean diametersFig 4 Relation between the mean diameters and(T=1200.C and amount of metal charged中国煤化工 harged(T=1200COMm /b)not increase even if the nanomaterials arc kept in air for one year because the oxygenoxide layer on the particle surface prevents the interior from oxidation, but is fairly resis-tant to electronic conduction. Consequently, metallic nanomaterials usually have a staticelectrification. If the depth of the oxide layer is not enough, cracks on the surface oxidelayer are possibly created by discharge between particles, which expose the particle partsto air again and cause a quick oxidation (271. In general, oxygen slows down the growth ofnanoparticles, is deleterious to their mechanical properties and diminishes their density.36Fig 5 X-ray diffraction patterns of nanometer Zn particles at evaporation temperature of1300oC (a)and at 1200oC with the distance of 20cm(b)(amount of metal charged of10g and argon flow rate of 0.4m/h)Fig 6a-c show TEM micrographs of nanometer Zn particles and the corresponding electron diffraction pattern Fig 6d from particle Fig 6a, with the reflection indexed with thehcp structure, the electron diffraction pattern, which also reveals weak reflection. Diffrac-tions lines due to oxides were found. The oxide formed on the zinc particles even whenthese were produced under the cleanest possible conditions. These photographs suggestthat the zinc particles have the shapes of a polyhedral, spherical and particles having irregular shapes, which exhibited striking aggregation habits. For all gas evaporation particlessmaller than about 20nm, the particles look almost spherical. For larger sizes, however,the picture is quite differentllLl. These particles may grow by collision(coagulation)andcoalescence(sintering) in the work chamber. If sintering is fastcr, single particles are pro.duced while if collisions are faster, agglomerate particles are formed[27].However, theseparticles collected are nearly aggregated and have various sizes, but the aggregates appearto be made up of individual particles having diameter of about 10nm. Most applicationsrequire that long-chain aggregates be avoided. However, aggregates can, at times, be beneficial, for example to improve powder flow, but these aggregates must have a large fractaldimension, i. e,, the aggregate must appear overall spherical in shape[ 28. It can be deducedfrom the diffraction patterns and the corresponding micrographs that a given irregularparticles having its c-axis parallel to the particle. One can see good strong spotty ringslead to zinc and weak diffuse rils lead to zinc oxNanometer with definite crystalhabits was sometimes observed among those with中国煤化工 vious that zincatoms leave the evaporation source in all directionCNMH Gthe wall of thework chamber like soot. In this course, they collide with gas atoms and lose energy. Macro-16scopically speaking, the metal vapour is cooled in the gas as it diffuses from the source 25When collision occurs between two of the atoms on the surface than single atoms. Suchpairs then act as nuclei of condensation. The condensation takes place in the region wherethe vapor is cooled below the critical temperature of saturation. Once the nucleation hastaken place the growth stage begins. The single crystals of Zn observed must arise from anatom-by-atom accretion process, the growth process terminating when the population ofsingle atoms has been depleted. If the transit time of the material is made sufficiently longby reducing the How rate of argon to zero, then clumping can occur and probably dose sobefore the sample impinges on the substrate(281. However, there is insuficient informationcurrently available to enable us to suggest a model for this accretion processThe mean particle size, estimated by X-ray diffraction ranges from 12 to 100nm, beingquite close to those obtaincd from TEM observation. It was reported(2, 13, 21] that thepresence of ZnO could affect their crystal shape by destroying their habits and becamevery rough and irregular, for example, when about 1% of air is ad- mixed in the inert gas,166nsnmGFig 6 TEM micrographs of nanometer Zn particles(a, b, c) and an electron diffractionpattern of photo (a)(d). Polyhedral, sphernother formsas canbe seen in photographs were found bt中国煤化工 all particlesre likely to be considerably oxidized andCNMHGystalhabit was observed. Broad rings in(d)probaur1uviuing that theparticle is not a zinc single crystalthe particle size becomes smaller in addition to the loss of definite morphology. They maybe interpreted as follows: in the former the reduction in size is caused by the oxide filmsor adsorbed molecules over each particles, which are most likely to hinder the coalescencegrowth. In the latter the oxide is formed on t he surfaces of particles. Since it has a veryhigh melting point, the coalescence is disturbed when particles collide with one another.For this reason, the surfaces become rough and irregular. It is generally accepted thatthe impurity content of even less than 1% exerts a considerable influcnce on the particlesize, although no quantitative measurements have as yet been made. It should be bornein mind that the effect of impurities is almost always present and the layers are zinc oxideformed when particles were exposed to air, before diffraction study. This consideration isreasonable, and agreement with the results shown above4. ConclusionsThe nanometer Zn particles(12-100nm in size) with different parameters were syn-thesized by gas evaporation and the effect of operating variables was investigated. It isindicated that the gas flow-rate and evaporation temperature havc a dominant effect on theparticle size. The average particle size increases with increasing Ar gas and evaporationtemperature, tends to increase slightly with amount of metal charged increasing and alsoposition of collection. We found that the particles collected closer to the source tended tobe somewhat smaller than those condensed farther away. The linear relation exists in thecoalescence mechanis growth in which two collided particles unite with a definite latticeorientation as to minimize their interface. The most difficult problem encountered duringthe experirnents is that the particles of zinc could not be protected from oxidation, even ina considerably pure atmosphere, as indicated by X-ray diffraction and electron microscopyThe subtle effect of oxygen on the crystalline shape of the particles is that it coated theNucleated particles with epitaxial ZnO layer and thereby prevented further growth expectby agglomeration. In this case, the particles became remarkably irregular and rough inshape. It may be reasonably concluded thal the oxidation of particles, while they werebeen condensed in the argon was negligible under the present experimental conditions andthat the particles were slightly oxidized when they were exposed to air before diffractionstudy in all size ranges, the particles showed simple geometric shapesREFERENCES1 R Uyeda, J. Crystal Growth 24 /25(1974)692 N. Wada, J. Appl. Phgs. 7(1968)1283 Y. Fukano, Appl. Pl4 M. Kato, J. Appl. Phys, 15(1975)7575S. Gao, D. Cui, B Huang and M. Jiang, J. Crystal Growth 192(1998)896 Y Saito, K. M. and R Uyeda, J. Appl. Phys. 19(1980)16037 A. Tasaki, S. Tomiyama, S Iida, N. Wada and R,, J. Appl. Phys. 4(1965)7078A. H Pfund, J. Opt. Soc. Amer. 23(1933)3759 R. Uyeda and O. Kimoto, J. Appl. Phys. 18(1944)7610 B.A. Nagasawa and S. Ogawa,J. Phys. Soc. 15(1960)111 C.G. Granqvist and R.A. Buhraman, J. Appl. Phys. 47rV凵中国煤化工13 S Kasukabe, S. Yatsuya and R Uyeda, J. Appl. Phys. 13(1974)1714. MHG12 R Uyeda, Progress in Materials Science 35(1991)116811 Y Saito, S. Yatsuya, K. Mihama and R. Uyeda, J. Appl. Playw. 17(1978)29115 S. Yatsuya, S. Kasukabe and R Uyeda, J. Crystal Growth 24/25(1974)31916 G. Skandan, Y.J. Chen, N. Glumac and B H. Kear, Nanostructured Materials 11(1999)14917 S.Y. Chun, N. Wakiya, K. Shinozaki and N Mizutani,J Mater. Res. 13(1998)211018 R w. Siegel, Mate. Sci. Eng. A168(1993)189.19 H Eifert, D Kupp and B Ginther, Powder Metallurgy 43(2000)31020 B. H. Gunther, Int. of Powder Metallurgy 35 (1999). 5321 J D. Eversole and H P. Broida, J. Appl. Phys. 45(1974)59622 K. Kimoto and I Nishida, J. Appl. Phys. 6(1967)104723 I Kubo, J. Phys. Soc. Japan 16( 1961)235824 J.M. Ntep, S.S. Hassani, A. Lusson, A.T. Carli, D. Ballutaud and G. Didier, R T. Riboult, J. CrystalGrouth207(1999)30.25 V.G. Gryaznov and L I. Trusov, Progress in Material Science 37(1993)29326T. Ohno, S. Yatsuya and R. Uyeda, J. Appl. Phys. 15 (1976)121327 M. Rajalakshmi, K.A. Arora, B.S. Bender and S Mahamuni, J. Appl. Phys. 87(2000)2445:8 L R. Axelbaum, Powder Metallurgy 43(2000)323中国煤化工CNMHG