Hydrothermal Synthesis of Advanced Ceramic Powders

Wojciech L. Suchanek1, a and Richard E. Riman2, b

1Sawyer Technical Materials, LLC, 35400 Lakeland Boulevard, Eastlake, OH 44095, USA2Rutgers University, Department of Materials Science and Engineering, 607 Taylor Road,

Piscataway, NJ 08855, USAawls@SawyerLLC.com, briman@email.rci.rutgers.edu

Keywords: Hydrothermal synthesis, review powder synthesis, morphology control,hydroxyapatite, PZT, alumina, zinc oxide, carbon nanotubes.

Abstract. This paper briefly reviews hydrothermal synthesis of ceramic powders and shows howunderstanding the underlying physico-chemical processes occurring in the aqueous solution can beused for engineering hydrothermal crystallization processes. Our overview covers the current statusof hydrothermal technology for inorganic powders with respect to types of materials prepared, abilityto control the process, and use in commercial manufacturing. General discussion is supported withspecific examples derived from our own research (hydroxyapatite, PZT, -Al2O3, ZnO, carbonnanotubes). Hydrothermal crystallization processes afford excellent control of morphology (e.g.,spherical, cubic, fibrous, and plate-like) size (from a couple of nanometers to tens of microns), anddegree of agglomeration. These characteristics can be controlled in wide ranges usingthermodynamic variables, such as reaction temperature, types and concentrations of the reactants, inaddition to non-thermodynamic (kinetic) variables, such as stirring speed. Moreover, the chemicalcomposition of the powders can be easily controlled from the perspective of stoichiometry andformation of solid solutions. Finally, hydrothermal technology affords the ability to achieve costeffective scale-up and commercial production.

Introduction

Hydrothermal research was initiated in the middle of the 19th century by geologists and was aimed atlaboratory simulations of natural hydrothermal phenomena. In the 20th century, hydrothermalsynthesis was clearly identified as an important technology for materials synthesis, predominantly inthe fields of hydrometallurgy and single crystal growth [1]. However, the severe (supercritical)reaction conditions required particularly for growing single crystals have discouraged extensiveresearch and commercialization for many materials. In recent years, commercial interest inhydrothermal synthesis has been revived in part because a steadily increasing large family ofmaterials, primarily ceramic powders, has emerged that can be prepared under mild conditions(T<350ºC, P<100 MPa). The growing number of scientific papers on hydrothermal synthesis ofceramic powders, which almost quadrupled between 2000 and 2004 (Fig. 1a), illustrates the risinginterest in this area, with China, Japan, and USA publishing most extensively (Fig. 1b).

Hydrothermal Synthesis as a Materials Synthesis Technology

Process Definition. Hydrothermal synthesis is a process that utilizes single or heterogeneous phasereactions in aqueous media at elevated temperature (T>25ºC) and pressure (P>100 kPa) tocrystallize ceramic materials directly from solution. However, researchers also use this term todescribe processes conducted at ambient conditions. Syntheses are usually conducted at autogeneouspressure, which corresponds to the saturated vapor pressure of the solution at the specifiedtemperature and composition of the hydrothermal solution. Upper limits of hydrothermal synthesisextend to over 1000ºC and 500 MPa pressure [2]. However, mild conditions are preferred forcommercial processes where temperatures are less than 350ºC and pressures less than

Advances in Science and Technology Vol. 45 (2006) pp. 184-193online at http://www.scientific.net© (2006) Trans Tech Publications, Switzerland

All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without thewritten permission of the publisher: Trans Tech Publications Ltd, Switzerland, www.ttp.net. (ID: 12.45.174.100-24/10/06,17:51:02)

Fig. 1. (a) Number of papers on hydrothermal processing of ceramic powders published between1989-2005 vs. all materials science-related papers on hydrothermal technology [4]; (b) geographicaldistribution of the research on hydrothermal processing of ceramic powders (1989-2005). Based oninformation available in Science Citation Index, Chemical Abstracts, and INSPEC databases.

approximately 50 MPa. The transition from mild to severe conditions is determined mostly bycorrosion and strength limits of the materials of construction that comprise the hydrothermal reactionvessels. Intensive research has led to a better understanding of hydrothermal chemistry, which hassignificantly reduced the reaction time, temperature, and pressure for hydrothermal crystallization ofmaterials (T<200ºC, P<1.5 MPa) [1-5]. This breakthrough has made hydrothermal synthesis moreeconomical since processes can be engineered using cost-effective and proven pressure reactortechnology and methodologies already established by the chemical process industry.

Merits of Hydrothermal Synthesis. Hydrothermal synthesis offers many advantages overconventional and non-conventional ceramic synthetic methods. All forms of ceramics can beprepared with hydrothermal synthesis, namely powders, fibers, and single crystals, monolithicceramic bodies, and coatings on metals, polymers, and ceramics. From the standpoint of ceramicpowder production, there are far fewer time- and energy-consuming processing steps since high-temperature calcination, mixing, and milling steps are either not necessary or minimized. Moreover,the ability to precipitate already crystallized powders directly from solution regulates the rate anduniformity of nucleation, growth and aging, which results in improved control of size andmorphology of crystallites and significantly reduced aggregation levels, that is not possible withmany other synthesis processes [6]. The elimination/reduction of aggregates combined with narrowparticle size distributions in the starting powders leads to optimized and reproducible properties ofceramics because of better microstructure control. Precise control of powder morphology can alsobe significant. For instance, powders with crystallites having well-developed shapes corresponding toparticular crystallographic directions, such as whiskers, plates, or cubes, can be oriented to formmaterials with single crystal-like properties, such as polymer-ceramic composites or texturedceramic-ceramic composites with anisotropic properties. Another important advantage of thehydrothermal synthesis is that the purity of hydrothermally synthesized powders significantly exceedsthe purity of the starting materials. It is because the hydrothermal crystallization is a self-purifyingprocess, during which the growing crystals/crystallites tend to reject impurities present in the growthenvironment. The impurities are subsequently removed from the system together with thecrystallizing solution, which does not take place during other synthesis routes, such as high-temperature calcination.

Hydrothermal processing can take place in a wide variety of combinations of aqueous and solventmixture-based systems. In general, processing with liquids allows for automation of a wide range ofunit operations such as charging, transportation, mixing and product separation. Moreover, relativeto solid state processes, liquids give a possibility for acceleration of diffusion, adsorption, reaction

Others11%

Italy 2%

Taiwan 2%Germany 2%

India 3%France 3%

UK 3%

USSR/Russia4%

USA11%

China44%

Japan15%

1990 1992 1994 1996 1998 2000 2002 2004 20060

100

200

300

400

500

600

Num

ber o

f Sci

entif

ic P

aper

s

Publication Year

All papers on hydrothermal technology Papers on hydrothermal synthesis of powders only

(a)

(b)

11th International Ceramics Congress185

rate and crystallization, especially under hydrothermal conditions [4]. However, unlike manyadvanced methods that can prepare a large variety of forms and chemical compounds, such aschemical vapor-based methods, the respective costs for instrumentation, energy and precursors arefar less for hydrothermal methods. Hydrothermal methods are more environmentally benign thanmany other synthesis methods, which can be attributed in part to energy conserving low processingtemperatures, absence of milling, ability to recycle waste, and safe and convenient disposal of wastethat cannot be recycled [4]. The low reaction temperatures also avoid other problems encounteredwith high temperature processes, for example poor stoichiometry control due to volatilization ofcomponents (e.g., Pb volatilization in Pb-based ceramics).

Materials synthesized under hydrothermal conditions often exhibit differences in point defectswhen compared to materials prepared by high temperature synthesis methods. For instance, inbarium titanate, hydroxyapatite, or -quartz, water-related lattice defects are among the mostcommon impurities and their concentration determines essential properties of these materials. Theproblem of water incorporation can be overcome by either properly adjusting the synthesisconditions or by use of non-aqueous solvents (solvothermal processing). Another importanttechnological advantage of the hydrothermal technique is its capability for continuous materialsproduction, which can be particularly useful in continuous fabrication of ceramic powders [7].Moreover, hydrothermal crystallization can be monitored in-situ using a range of techniques [8-9],which allow determination of the crystal growth mechanism and better control of the hydrothermalsynthesis process.

Ceramic Powder Compositions and Morphologies. A variety of ceramic powders have beensynthesized by hydrothermal methods. Most common are oxide materials, both simple oxides, suchas ZrO2, TiO2, SiO2, ZnO, Fe2O3, Al2O3, CeO2, SnO2, Sb2O5, Co3O4, HfO2, etc., and complexoxides, such as BaTiO3, SrTiO3, PZT, PbTiO3, KNbO3, KTaO3, LiNbO3, ferrites, apatites,tungstates, vanadates, molybdates, zeolites, etc., some of which are metastable compounds, whichcannot be obtained using classical synthesis methods at high temperatures. Hydrothermal synthesis ofa variety of oxide solid solutions and doped compositions is common. The hydrothermal technique isalso well suited for non-oxides, such as pure elements (for example Si, Ge, Te, Ni, diamond, carbonnanotubes), selenides (CdSe, HgSe, CoSe2, NiSe2, CsCuSe4), tellurides (CdTe, Bi2Te3, CuxTey,AgxTey), sulfides (CuS, ZnS, CdS, PbS, PbSnS3), fluorides, nitrides (cubic BN, hexagonal BN),aresenides (InAs, GaAs), etc.

Crystalline products can be in a variety of forms, such as equiaxed (for example cubes, spherical),elongated (fibers, whiskers, nanorods, nanotubes), plates, nanoribbons, nanobelts, etc. Core-shellparticles (CdS/ZnO, CoAl2O4/ZnAl2O4) and composite powders consisting of a mix of at least twodifferent powders (TiO2/HAp, TiO2/SiO2) can be also prepared. Quite often, the size andmorphology can be engineered over a wide range and even adopt nonequilibrium morphologies.

Design of Hydrothermal Experiments

A majority of the hydrothermal synthesis work that has been done in the past has used Edisonian trialand error methods for process development. This type of experimental approach suffers from itstime-consuming nature and the inability to clearly discern between processes that are controlled byeither thermodynamics or kinetics. Thermodynamic modeling can be used to design a process to bethermodynamically favored using fundamental principles instead of the Edisonian methods [10-11].For a given precursor system, the effects of concentration, temperature and pressure can be modelledto define the processing variable space over which the phase(s) of interest are stable. Moreover,many different types of precursor systems can be compared and experiments can be designed tomake materials that have never been previously prepared in hydrothermal solution. Such a designapproach requires far fewer experiments than Edisonian methods. It can be accomplished usingcommercially available computer software, such as OLI software, which also contains a database of

Advances in Science and Technology Vol. 45 186

thermodynamic properties for many common systems. The methodology for these calculations havebeen reviewed elsewhere [10-11].

With the thermodynamic variables processing space well defined for the material of interest, arange of conditions can be explored for control of reaction and crystallization kinetics for thepurpose of developing a process suitable to produce the desired form of the ceramic powders (e.g,particular size, morphology, aggregation level). Thermodynamic processing variables, such astemperature, pH, concentrations of reactants and additives, determine not only the processing spacefor a given material but also influence both reaction and crystallization kinetics. The phenomena thatunderlie the size and morphology control using the thermodynamic variables are the overallnucleation and growth rates, which control crystal size, and the competitive growth rates alongprincipal crystallographic directions that control morphology. Size of the crystals can be thuscontrolled by varying temperature and concentration. Crystal morphology and size can beadditionally affected by surfactants, which can adsorb on specific crystallographic faces and solvents,which adsorb similarly as well as regulate solubility. Unfortunately, changing the thermodynamicsynthesis variables is constrained by the phase boundaries in a specific phase diagram. Therefore, itmay or may not be possible to exploit all sizes and morphologies for a certain material by modifyingonly the thermodynamic variables. However, non-thermodynamic variables are also extremelyimportant when operating in thermodynamically limited processing variable space. For instance,changing the stirring speed during powder synthesis can change the particle size by orders ofmagnitude [12].

Fig. 2. (a) and (b) SEM photographs of submicron ZnO powders synthesized hydrothermally withoutstirring (magnification is the same in both cases). (c) and (d) SEM photographs of PZT crystalsprepared hydrothermally at 150ºC for 24 h at moderate stirring. Total concentration of (Zr+Ti) was0.1 m and the Pb:Zr:Ti ratio was 1.1:0.52:0.48. Concentration of the TMAH mineralizer was (c)0.5 m, (d) 1.0 m.

(a)

(b)

(c)

5 m

(d)

5 m

11th International Ceramics Congress187

Exploiting Synthesis Variables: Representative Systems

Representative examples of hydrothermally synthesized oxide powders, such as hydroxyapatite,PZT, and ZnO, all derived from our own research, are discussed in this section. These examplesdemonstrate the ability of the hydrothermal technology to control chemical composition, aggregationlevel, size and morphology in wide range, resulting in powder characteristics that are difficult toachieve by classical powder synthesis methods.

Thermodynamic Variables (ZnO, PZT, HAp). The technique of using thermodynamicvariables, such as temperature, concentrations of reactant, various additives and/or solvents, is themost widely used method to control powder size and morphology under hydrothermal conditions.

In our first example, ZnO powders were synthesized from aqueous solutions containing dissolvedZn salts in the presence of basic mineralizers at temperatures ranging between 150-300ºC undersaturated vapor pressures without stirring. Concentrations of the zinc salts varied between 0.3 m and6.0 m, concentration of the mineralizers were set between 0 and 12.0 m. The synthesized powderswere in most cases pure-phase ZnO (zincite) with equiaxed morphology and low aggregation level.Size of the crystallites ranged between tens of nanometers and tens of microns. Fig. 2a,b shows twoexamples of powders with different crystallite sizes. Within the range of ZnO phase stability, withincreasing temperature, and zinc concentration, size of the crystals generally increased.

Changing reactants concentration can be used not only to control size of the crystallites but alsoto control their shape. Fibers or cubes of lead zirconate titanate (PZT, chemical formula Pb(ZrxTi1-

x)O3, where x=0-1) can form during the hydrothermal synthesis at 150ºC from precursors derivedfrom Zr and Ti alkoxides in the presence of TMAH mineralizer (Fig. 2c,d). TMAH concentration of0.5 m results in the formation of PZT fibers while 1 m-TMAH yields PZT cubes [13].

The experiments with hydroxyapatite powders [HAp, chemical formula Ca10(PO4)6(OH)2] were

Fig. 3. HAp nanosized crystals preparedhydrothermally at 200ºC for 24 h undermoderate stirring in (a) 50 vol. % 2-propanol (aq); (b) in pure water with 1 wt%KCl; (c) HAp whiskers synthesizedhydrothermally at 200ºC for 5 h undermoderate stirring.

200 nm200 nm200 nm200 nm

200 nm200 nm200 nm200 nm 10 m

(a)

(b)

(c)

Advances in Science and Technology Vol. 45 188

focused on size and morphology control by adjusting chemical additives. Very fine, nearly nanosizedHAp crystals have been synthesized at 200ºC for 24 h under autogenous pressure of about 2 MPaunder moderate stirring, using the method described elsewhere [14]. HAp crystals synthesized in50 vol.% 2-propanol(aq) had low aspect ratios ranging between 2-3 and diameters between 20-40 nm (Fig. 3a). Conversely, uniform, nanosized needles (dimensions of about 20 nm by 100-160 nm, aspect ratio of 5-8) were formed when 1 wt% KCl additive was used (Fig. 3b). HApcrystals prepared under similar conditions but without additives, were 20 nm by 50-100 nm in size,yielding aspect ratios between 3-5.

In addition to fine HAp powders, the hydrothermal technique allows synthesis of larger HApcrystals, such as micron-sized HAp whiskers. The HAp whiskers can be prepared at 200ºC for 5hunder the pressure of about 2 MPa, under moderate stirring, from aqueous solutions containingvarious concentrations of Ca(OH)2, H3PO4, and lactic acid [15]. In this experimental set-up, Ca ionsare being chelated by the lactic acid, preventing precipitation of amorphous calcium phosphate atambient temperatures. Precipitation at elevated temperatures yields large HAp crystals. Despiteacidic pH in the starting solutions, the concentrations of the reactants were still within the stabilityfield of HAp at 200ºC [14]. The diameters, lengths, and aspect ratios of the synthesized HApwhiskers were in the range of 1-10 m, 30-60 m, and 5-35, respectively (Fig. 3c). With increasingconcentration of the lactic acid in the starting solution, whisker diameters increased and aspect ratiosdecreased while lengths remained almost unchanged. The Ca/P molar ratios of the HAp whiskerswere in the range of 1.59-1.62, as compared with 1.67 for stoichiometric HAp. They increasedslightly with increasing concentration of the lactic acid and were almost independent of pH and theCa/P ratio in the starting solution.

Non-thermodynamic Variables (PZT) Non-thermodynamic synthesis variables, such as stirringrate and method of precursor were found to be essential in morphology control of the PZT crystals.Hydrothermal synthesis of PZT cube-shaped crystallites was performed at 150ºC for 2-24 h in stirredautoclaves at a stirring speed ranging from 200-1700 rpm [12]. Concentration of the

Fig. 4. (a) Phase equilibria in the Pb-Zr-Ti-K-H2O system at 150ºC as functions ofprecursor and mineralizer concentrations.The symbol, , denotes experimentalreaction conditions; (b) SEM photograph ofPZT particulates, synthesized at 150ºC for24 h at 700 rpm stirring; (c) Average particlesize of PZT particulates as a function ofstirring speed during hydrothermal synthesisat 150ºC for 24 h. Error bars denotestandard deviation values.

(c)

5 µm5 µm5 µm5 µm

0 200 400 600 800 1000 1200 1400 1600 18000

2

4

6

8

10

12

14 TiO

2-based precursor

Alkoxide-derived precursor

Ave

rage

Par

ticle

Siz

e of

PZT

(µm

)

Stirring Speed (rpm)

0

-0.8

-1.6

-2.4

-3.20 2 4 6 8 10 12 14

[KOH] (mol/kg H2O)

Log

[Ti]

No PZT

PbO+PZT

0% < Yield < 99%

PZT 70/30, yield > 99%

• •

PbO

All aqueous species0

-0.8

-1.6

-2.4

-3.20 2 4 6 8 10 12 14

[KOH] (mol/kg H2O)

Log

[Ti]

No PZT

PbO+PZT

0% < Yield < 99%

PZT 70/30, yield > 99%PZT 70/30, yield > 99%

• •

PbO

All aqueous species

(a) (b)

11th International Ceramics Congress189

chemicals used was well within the calculated stability field of the PZT phase, which is shown inFig. 4a. With increasing stirring speed, size of the PZT crystals became smaller and they exhibitedbetter-defined cube habit (Fig. 4b). This effect was more strongly pronounced in the case of usingthe TiO2-based precursor: increasing the stirring speed from 200 to 1700 rpm has reduced theparticle size of PZT crystals almost 2 orders of magnitude, from over 10 m to a submicron range(Fig. 4c). Generally speaking, stirring during crystal growth leads to an increase in the probability ofspontaneous nucleation, a decrease in supersaturation inhomogenities, and an increase of the growthrate [10-11]. The final size of the crystals is then determined by a balance between the nucleation andgrowth rates. In this particular case, the nucleation rate seems to dominate over the growth rate,therefore the crystal size decreased with increasing stirring speed.

Hydrothermal Hybrid Techniques

In order to additionally enhance the reaction kinetics or the ability to make new materials, a greatamount of work has been done to hybridize the hydrothermal technique with microwaves(microwave-hydrothermal processing), electrochemistry (hydrothermal-electrochemical synthesis),ultrasound (hydrothermal-sonochemical synthesis), mechanochemistry (mechanochemical-hydrothermal synthesis), optical radiation (hydrothermal-photochemical synthesis), and hot-pressing(hydrothermal hot pressing), as reviewed elsewhere [1-4]. Below we show an example of a novelhydrothermal hybrid technique, which was developed by the authors [16-17].

Mechanochemical-Hydrothermal Synthesis of Nanosized Hydroxyapatite Powders withControlled Chemical Composition. The mechanochemical-hydrothermal synthesis is a hybrid ofthe mechanochemical and hydrothermal synthesis methods and should be clearly distinguished fromthe mechanochemical synthesis, which involves only solid-state reactions. The mechanochemical-hydrothermal synthesis (sometimes called “wet” mechanochemical) takes advantage of the presenceof an aqueous solution in the system, which can actively participate in the mechanochemical reactionby acceleration of dissolution, diffusion, adsorption, reaction rate, and crystallization [4]. Themechanochemical activation of slurries can generate local zones of high temperatures (up to 450-700ºC) and high pressures due to friction effects and adiabatic heating of gas bubbles (if present inthe slurry), while the overall temperature is close to the room temperature. The mechanochemical-hydrothermal route produces comparable amounts of powders as the hydrothermal processing but itrequires lower temperature, i.e. room temperature, thus, there is no need for a pressure vessel andexternal heating, because the reaction is accelerated via comminuting methods.

A variety of undoped, carbonate-, sodium-and-carbonate-, and Mg-substituted crystallinehydroxyapatite powders (denoted hereafter HAp, CO3HAp, NaCO3HAp, and Mg-HAp,respectively) were prepared at room temperature via heterogeneous reactions between resepectivelyCa(OH)2, CaCO3, Na2CO3, and/or Mg(OH)2 powders and (NH4)2HPO4 aqueous solution using themechanochemical-hydrothermal route described in detail elsewhere [16-17]. In the case of the HApwithout substituting ions, the synthesis product was phase-pure, highly-crystalline HAp, which wasthermally stable at 900ºC and had nearly stoichiometric composition. Room temperature products ofcarbonated HAp synthesis were phase-pure CO3HAp and NaCO3HAp containing 2.5-12 wt% ofcarbonate ions in the lattice [16]. The purified powders substituted with Mg were phase-pure Mg-HAp containing 0.24-28.4 wt% of Mg [17]. Surprisingly, concentration of Mg in the Mg-HAppowders was higher than in the starting slurries, indicating preferential incorporation of Mg over Ca,which has never been reported before and seems to be related to the unique mechanochemical-hydrothermal synthesis conditions only. All synthesized HAp, CO3HAp, NaCO3HAp, and Mg-HAppowders, independently of their chemical compositions, exhibited high surface area ranging between82 and 269 m2/g, corresponding to crystalite size of 7 nm to 23 nm. The crystallites were ratherheavily agglomerated, with median particle sizes between 100 nm and 1.6 µm. Additional control ofcrystallization pH and use of surfactants allowed reducing the agglomerate size to about 90 nm [14].Our results emphasize importance of the aqueous solution, which actively participates in the

Advances in Science and Technology Vol. 45 190

synthesis reaction by dissolving at least one of the reactants, in our case (NH4)2HPO4, which is notobserved with conventional mechanochemical synthesis of HAp [16-17]. The mechanochemical-hydrothermal synthesis allows preparation of HAp powders with a variety of chemical compositionsand with well defined stoichiometry levels but is not well suited to control aggregation level, size andmorphology of the HAp crystallites.

Hydrothermal Synthesis of Non-Oxide Materials

The hydrothermal technique has been mostly known for preparation of oxide materials, which is alsothe main focus of this article. However, non-oxide materials, including a wide range of metals,semiconductors, intermetallics, selenides, tellurides, sulfides, fluorides, nitrides, arsenides, etc., canalso be synthesized from aqueous solutions in a variety of sizes and morphologies. Here we presentan example of multiwalled carbon nanotubes, which in addition to other forms of carbon (amorphouscarbon, diamond) can be synthesized under hydrothermal conditions from a variety of carbonsources, as demonstrated by our research [18].

Hydrothermal Synthesis of the Carbon Nanotubes. During the hydrothermal treatment of C60

at 700ºC (100 MPa) in the presence of 3 wt% of nickel, high-quality open-ended multi-walledcarbon nanotubes were formed in the vicinity of the nickel particles [18]. These nanotubes hadtypically an outer diameter of 30-40 nm and wall thickness of 5 nm with a graphitization level similarto carbon nanotubes prepared by CVD techniques, as shown in Fig. 5b. In the follow-up works,cheaper carbon sources, such as polyethylene or amorphous carbon were used in hydrothermalsynthesis of the carbon nanotubes primarily in the presence of the Ni catalysts. Currently, a variety ofsizes of the multiwalled carbon nanotubes with yields approaching 100% and diameters rangingbetween 50 and 500 nm can be prepared by properly adjusting synthesis conditions of thehydrothermal treatment of very simple and cheap carbon sources. Nanotubes with diameters 50-100 nm are shown in Fig. 5a. The hydrothermal technology gives promise for development ofinexpensive mass-production of carbon nanotubes in the near future.

Fig. 5. (a) High-yields of the multiwalled carbon nanotubes with diameters of 50-100 nm synthesizedusing polyethylene carbon source; (b) TEM micrographs of typical carbon nanotubes produced byhydrothermal treatment of fullerene C60 with 3% Ni, 30% water at 700ºC for 168 h under 100 MPa.Insert shows graphite fringes in the nanotube wall.

2 µm

(a)

2 µm2 µm

(a)(a)

5 nm

(b) 20 nm

5 nm

(b) 20 nm

11th International Ceramics Congress191

Industrial Production

Several hydrothermal technologies, primarily for the production of single crystals, such as α-quartzfor frequency control and optical applications, ZnO for UV- and blue light-emitting devices, andKTiOPO4 for nonlinear optical applications, have already been developed that demonstrate thecommercial potential of the hydrothermal method. However, the largest potential growth area forcommercialization is ceramic powder production. The widely used Bayer process uses hydrothermalmethods to dissolve Bauxite and subsequently precipitate aluminum hydroxide, which is later heat-treated at high temperature to crystallize as α-alumina. In 1989, the worldwide production rate wasabout 43 million tons/year. The production of perovskite-based dielectrics and zirconia-basedstructural ceramics is a promising growth area for hydrothermal methods. Corporations such asCabot Corporation, Sakai Chemical Company, Murata Industries, Ferro Corporation and others haveestablished commercial hydrothermal production processes for making dielectric ceramic powdercompositions for capacitors.

Hydrothermal Production of -Al2O3 Powders. Sawyer Technical Materials, LLC recentlydeveloped hydrothermal production of high-purity -Al2O3 powders. Sawyer’s products line of the -Al2O3 powders called DiamoCor, comprises both undoped and Mn or Cr-doped compositions withthe following grain sizes: 40 m, 20-25 m, 10 m, 6 m, 3 m, 1 m, and 100-250 nm. Phasepurity of the powders is 100% of the -Al2O3 phase. Chemical purity ranges between 99.9% and>99.97%. The powders are being produced at Sawyer in hundreds of pounds batches, with very highreproducibility and a total production capacity of 100,000 lbs/month.

Representative corundum powders with a 10 m grain size are shown in Fig. 6a. The picturereveals typical as-synthesized powder, with uniform particle size distribution and well-defined crystalfaces in each grain, which is a single crystal of the -Al2O3 phase. The morphology of the

Fig. 6. -Al2O3 DiamoCor powders producedby Sawyer Technical Materials, LLC: (a) as-synthesized powders with median particle sizeof 10 m; (b) as-synthesized powders withmedian particle size of about 100 nm (c)typical particlesize distributions of variouspowders after milling/dispersing, nominalmedian particle sizes are marked.

10 µm

(a)

10 µm10 µm

(a)(a)

500 nm

(b)

500 nm500 nm

(b)(b)

(c)

0.1 1 10 100

0

20

40

60

80

1003 µm 10 µm

25 µm

40 µm

160 nm1 µm

Cum

mul

ativ

e W

eigh

t Per

cent

Particle size (µm)

Advances in Science and Technology Vol. 45 192

powders in addition to their phase and chemical composition make DiamoCor -Al2O3 a veryaggressive and long-lasting abrasive material. The powders have been successfully used in industriallapping and in thermal-spray applications, due to the combination of chemical and phase purity,morphology and tight particle size distribution, which result in good flowability. The morphology ofsubmicron -Al2O3 powder is shown in Fig. 6b. The particle size distributions of selected powdersare revealed in Fig. 6c as cumulative curves.

The example of -Al2O3 DiamoCor indicates a commercial potential of the hydrothermaltechnique, which allows reproducible synthesis of large volumes of powders with a wide range ofcrystallite sizes and chemical compositions.

References

[1] K. Byrappa and M. Yoshimura: Handbook of Hydrothermal Technology (NoyesPublications/William Andrew Publishing LLC, U.S.A. 2001).

[2] R. Roy: J. Solid State Chem. Vol. 111 (1994), p. 11.

[3] S. Sômiya: Hydrothermal Reactions for Materials Science and Engineering. An Overview ofResearch in Japan (Elsevier Science Publishers Ltd., U.K. 1989).

[4] M. Yoshimura, W. L. Suchanek, and K. Byrappa: MRS Bull. Vol. 25 (2000), p. 17.

[5] B. Gersten, M. Lencka and R. E. Riman: Chem. Mater. Vol. 14 (2002), p. 1950.

[6] R. E. Riman: in High Performance Ceramics: Surface Chemistry in Processing Technology,edited by R. Pugh and L. Bergström (Marcel-Dekker, U.S.A. 1993), p. 29.

[7] W. J. Dawson: Ceram. Bull. Vol. 67 (1988), p. 1673.

[8] G. C. Ulmer and H. L. Barnes: Hydrothermal Experimental Techniques (Wiley-Interscience,U.S.A. 1987).

[9] I. Sunagawa, K. Tsukamoto, K. Maiwa, and K. Onuma: Prog. Crystal Growth and Charact.Vol. 30 (1995), p. 153.

[10] R. E. Riman, W. L. Suchanek, and M. M. Lencka: Ann. Chim. Sci. Mat. Vol. 27 (2002) p. 15.

[11] W. L. Suchanek, M. M. Lencka, and R. E. Riman: in Aqueous Systems at ElevatedTemperatures and Pressures: Physical Chemistry in Water, Steam, and Hydrothermal Solutions,edited by D. A. Palmer, R. Fernández-Prini, and A. H. Harvey (Elsevier Ltd. 2004), p. 717.

[12] W. L. Suchanek, M.M. Lencka, L. E. McCandlish, R. L. Pfeffer, M. Oledzka, K. Mikulka-Bolen, G. A. Rossetti, Jr. and R. E. Riman: Crystal Growth & Design Vol. 5 (2005), p. 1715.

[13] S.-B. Cho, M. Oledzka, and R. Riman: J. Crystal Growth Vol. 226 (2001), p. 313.

[14] R. E. Riman, W. L. Suchanek, K. Byrappa, C. S. Oakes, C.-W. Chen, and P. Shuk: Solid StateIonics Vol. 151 (2002), p. 393.

[15] W. Suchanek, H. Suda, M. Yashima, M. Kakihana, and M. Yoshimura: J. Mater. Res.Vol. 10(1995), p. 521.

[16] W. L. Suchanek, P. Shuk, K. Byrappa, R. E. Riman, K. S. TenHuisen, and V. F. Janas:Biomaterials Vol. 23 (2002), p. 699.

[17] W. L. Suchanek, K. Byrappa, P. Shuk, R. E. Riman, K. S. TenHuisen, and V. F. Janas:Biomaterials Vol. 25 (2004), p. 4647.

[18] W. L. Suchanek, J. Libera, Y. Gogotsi, and M. Yoshimura: J. Solid State Chem. Vol. 160(2001), p. 184.

11th International Ceramics Congress193