ABSTRACT

Sol-gel chemistry provides a novel production route forceramics and composites that have a variety of applica-tions in medicine, biology, and biochemistry. Advantagesof sol-gel-derived materials in these applications includesimple, low temperature production routes that are capa-

ble of achieving temperature, chemical, and radiation-inert porousmaterials with a wide range of structural and microstructural prop-erties. Furthermore, sol-gel-derived materials display remarkablecompatibility with biomacromolecules and are conveniently func-tionalized with a variety of coupling agents. The properties of thesematerials and their present and potential applications in medicineand biology are reviewed.

BACKGROUND

The term sol-gel indicates a low temperature production routewhereby a ceramic or ceramic composite material is formed.Originally developed for the nuclear industry to aid in fuel process-ing, sol-gel-derived materials have found a wide range of applica-tions including microelectronics, solar cells, intelligent coatings,1

batteries,2 and catalysis, as well as in clinical and analytical chem-istry. The term sol refers to a colloidal suspension of solid particlesthat, under the proper conditions, condense with one another orwith unreacted precursor to form an irregular, non-crystalline,three-dimensional network that spans the entire volume of a vessel.The resulting network can be described as a gel when a path can betraced from one boundary point of the vessel to another withoutleaving the solid phase, although further condensation may still beoccurring.

The gel network is most commonly very porous (sometimes inexcess of 99%), and is often denoted by the interstitial fluid. Thefluid phase in hydrogels is thus water or aqueous solution, in alco-gels it is alcoholic solutions, and in aerogels it is air or other gases.The gelation reaction can occur in any number of solvents and iscommonly catalyzed by changes in pH or stoichiometry of thereactants. Unlike many organic polymer hydrogels, solvent

Applications for Sol-Gel-DerivedMaterials in Medicine and Biology

B y J o h n F . C o n r o y , M a r y E . P o w e r , P a m e l a M . N o r r i s

U n i v e r s i t y o f V i r g i n i a

exchange in inorganic gels produces minimal changes in thedimension of the gels. Swelling is typically only on the order of afew percent, and if it does occur, is likely to lead to breakage ofthe gel.

The solid backbone of the network can be made from any of anumber of materials, but sol-gel processing is most commonwith metal oxides, and most specifically with silica, alumina,titania, and zirconia. The solid component is commonly derivedfrom the hydrolysis/condensation of the respective alkoxy metal-late (Figure 1).3, 4 The hydrolysis reaction refers to the substitu-tion of alkoxy groups with hydroxyl groups or water on individ-ual metal centers, while condensation is the chain lengtheningreaction that releases water derived from hydroxyl groups onneighboring molecules. In general, hydrolysis is quickest inacidic solutions, while condensation occurs quickest under basicconditions. In low pH solutions or in solutions that contain sub-stoichiometric amounts of water, it is possible to produce stable“prehydrolyzed” sols in which condensation progresses slowlyenough that the sol can be handled as a liquid and stored forseveral months. The surface of sol-gel-derived metal oxides isusually more reactive than that of the corresponding melt-derived bulk materials. For example, sol-gel-derived silica has amuch higher concentration of uncondensed surface silanolgroups than melt silicas. Condensation (and hence passivation),of the surface silanol groups can be driven by annealing the gelsat elevated temperatures after drying.5, 6

The physical microstructure of sol-gel-derived metal oxides ishighly dependent upon the production conditions.3 Reactant stoi-chiometry, temperature, catalyst, polarity of the reaction solution,and/or surfactant concentration all significantly influence thephysical properties of the final gel. Although these effects are oftenspecific to the particular reaction conditions, some general conclu-sions can be drawn. In general, gels formed under acidic condi-tions tend to be “polymeric” in nature, where condensation ofunreacted alkoxy silicate occurs preferentially at the end of longchains that grow until they entangle and condense with one

another. This type of gel tends to display smaller pore sizes andhigher elastic moduli for a given density. On the other hand, sol-gels formed under basic conditions tend to form larger clustersthat condense with one another as particles. Sol-gels formed underbasic conditions tend to display larger pore sizes and lower elasticmoduli. Surfactant concentration and solution polarizability canalso have a significant influence upon gel microstructure, andphase separation may occur during the gelation reaction and thisleads to unique microstructural features. SEM micrographs ofthree different large pore, phase-separated sample gels from our labare shown in Figure 2 to illustrate the range of accessiblemicrostructures.

Functionalization of sol-gel-derived metal oxides is quite easydue to both the reactive surface and liquid precursor of the solid.The increased reactivity of sol-gel-derived materials removes theneed for traditional activation steps,7, 8 and the liquid precursorguarantees even perfusion of a wide variety of silane couplingagents that can be introduced at various times during the sol-gelproduction process. Furthermore, by alternating condensation andfunctionalization steps, it is possible to pattern three-dimensionalstructures with functionalization domains.9 The most commonapproach to surface functionalization involves silane couplingagents that, despite difficulty with reproducibility, are widely avail-able and the subject of much research due to their commercialimportance.10

Liquid-filled gels have traditionally been dried either by simpleevaporation or by supercritical drying of the liquid component.Evaporation is the most convenient and inexpensive drying route,but it is difficult to perform with anything other than thin films,microparticles, and small monoliths. The dry gel is called a xerogelto denote a change in gel structure. During evaporation, a capil-lary pressure develops within the pores of the gel and draws thepore walls together. On the microscale, this collapses the pores andreduces the mean pore diameter and bulk porosity. On themacroscale, gradients in capillary pressure are often strong enoughto fracture the gel and large macroscopic monolithic pieces are dif-

Figure 1. An illustration describing a standard sol-gel process.

ficult to obtain. Even though the dry volume of xerogels is com-monly between 1/2 to 1/30th the original volume of the gel, xero-gels often retain porosities in excess of 30% and they retain muchof the surface reactivity of the undried gels.

Supercritical drying, on the other hand, essentially preserves thestructure of the initial liquid-filled gel and is capable of producinglarge monolithic pieces. By taking the interstitial liquid supercriti-cal, the capillary pressure is brought to zero and the gel micro- andmacro-structure are preserved during removal of the supercriticalfluid. The resulting material is called an aerogel and is one of themost extraordinary room temperature solids known to humanity.Aerogels have the lowest density, lowest thermal conductivity, low-est dielectric constant, lowest acoustic velocity, and lowest refrac-tive index of any known solid. Although the primary commercialapplication of aerogels is thermal insulation in double pane win-dows, they have also been proposed in various microelectronic,acoustic, and personal care product11 applications.

CURRENT APPLICATIONS

COLUMN CHROMATOGRAPHY STATIONARY PHASES:

In the past, the most widespread application of silicon chem-istry to medicine and biology was in chromatography column sta-tionary phases for clinical chemistry and bioanalytical chemistry.Despite stability over a limited pH range, modified silica particlesand gels have been used as stationary phases in a wide range ofchromatographic applications.12 Recently, the sol-gel productionroute has begun to find novel applications in analytical chro-matography.13-15 Furthermore, the importance of the sol-gel pro-duction route may dramatically increase with the further develop-ment of microanalytical systems.

One of the earlier and more successful examples of a sol-gel-derived stationary phase is the Zorbax line originally developed atDuPont for size exclusion chromatography.16 Zorbax is the tradename for a series of sol-gel-derived microspheres that are formed bycausing the gelation reaction to occur in a “mold” consisting ofreverse phase (water in oil) micelles. The geometry and size of thegel particles are determined by the microemulsion, and gelmicrostructure and surface functionality by the stoichiometry of the

reactants. It should be noted that the commercially available Zorbaxparticles are commonly sintered prior to sale and thus do not dis-play the high reactivity of the native gel without reactivation.8

A more recent example of a sol-gel derived column stationaryphase has been described by Guo and Colon for open tubular liq-uid chromatography.13, 14 Many capillary columns are functional-ized (fixed) when they are solids, often with discouraging results.These chemically bonded stationary phases provide low phaseratios, poor pH stability, and the results are often unreliable.However, functionalization directly in the sol during gelationyields stationary phases with higher capacities and good pH stabil-ity. It has been postulated that the direct incorporation of themodification into the gel-derived silica is responsible for theincreased performance and control of the reaction conditions.

In addition to such particulate chromatography stationary phas-es, a number of researchers have examined monolithic stationaryphases for various HPLC and gas chromatography applications.Current research includes examining the effects of various surfacemodifications, developing larger pore diameter materials,17 anddeveloping hydrolysis-resistant materials other than silica for chro-matographic applications.18, 19

CHEMICAL SENSORS:

Sol-gel-derived materials have recently spurred much interest inboth medical and environmental chemical sensing. Sol-gel-derivedsilica is commonly used to entrap and/or covalently bind macro-molecules and small molecule dyes within a highly porous solidnetwork. An extensive body of literature has recently beenreviewed.20, 21 The sol-gel matrix serves the dual purpose of confin-ing large quantities of these molecules within a three-dimensionalmatrix yet allowing easy communication with the external environ-ment by diffusion through the pores. Furthermore, the reactivity ofbiological macromolecules such as enzymes within sol-gel matriceshas been shown to be largely preserved, and a wide range of suchmolecules have been incorporated into sol-gel-derived glasses.

One example is the entrapment of the proteins cytochrome c,copper-zinc superoxide dismutase, and myoglobin by Ellerby et al.and Yamanaka et al.22, 23 Simple admixing of the proteins in a pre-hydrolyzed sol followed by gelation resulted in a solid, three-

Figure 2. Sample gel macrostructures obtained using the biocompatible production route described in the text.Although the densities of these samples are identical, the variable macrostructure will provide different physicaland transport properties. The images show a clear transition between a polymeric (left) and a particulate (right)

gel induced solely by surfactant (polyethylene glycol) concentration during condensation.

dimensional, protein-containing matrix in which the reactivity andspectral characteristics of the proteins are only slightly changed.Similar approaches have been used for a variety of biopolymers,including redox proteins24, 25 and immunoglobulins.26 Both mono-lithic and optical waveguide27 devices have been produced usingthese approaches.

PERMSELECTIVE AND ION-EXCHANGE MEMBRANES:

Both permselective and ion-exchange membranes have beenformed using sol-gel derived materials. The applications of thesematerials in medicine and biology include ion exchange duringpharmaceutical production, ion selective electrodes for clinical diag-nostics, and water treatment and desalinization. Functionalizationof the metal oxide backbone with charged or reactive organic modi-fiers is required to achieve the requisite properties.28, 29

FUTURE APPLICATIONS

MICROANALYTICAL SYSTEMS:

The relative importance of sol-gel-derived materials in analyticalsystems appears to be increasing with the development of micro-analysis, as sol-gel processing provides a convenient route for theproduction of porous media in a variety of applications.Microanalytical devices exploit microfabrication techniques fromthe semiconductor industry in order to create miniature analyticaland bioanalytical devices for point-of-care medical diagnosis andenvironmental chemical analysis. The small sizes of these devicesyield equipment that is inexpensive, fast, disposable, portable, andmassively parallel, but at the same time raise novel production andhandling issues for their components. For example, microscaleanalytical systems exhibit the same reliance upon porous media astheir macroscale counterparts, and yet traditional, well-character-ized porous materials and methods are incompatible with manymicrofabricated devices. This is due to handling difficulties, pH ortemperature instability, the need for successive loading and/orwashing steps, contamination by further microfabrication process-ing, or the need to access these materials with UV light. Currentmicroscale analytical systems where porous media are of criticalimportance are outlined in Table 1.

Sol-gel-derived silica possesses several material properties thatmake it a very attractive candidate for microanalytical systems,including covalent bonding with common microelectronic materi-als such as silica, temperature stability up to ~700°C, an easy tohandle liquid precursor state, a well-developed surface modifica-tion chemistry, and a microstructure that is largely tunable with

production conditions. The gelation reaction is also particularlywell-suited for incorporation into microanalytical devices.Gelation occurs at room temperature, which prevents damage toother thermally sensitive structures. Furthermore, gelation can becatalyzed by relatively mild changes in pH or water content and,under selected conditions, the catalysts need not be added in situ.The reaction conditions can be chosen such that gelation occurs intimes ranging from seconds to weeks. Thus, a single liquid phasecould be loaded into the device and used following a spontaneousgelation, without further rinsing, washing, addition of catalyst, orsteps to promote adhesion to the microanalytical substrate.

A direct comparison of an organic polymer such as polyacry-lamide and silica compatibility with microanalytical systems canbe made based upon micro-gel electrophoresis systems.30, 31 Inorder to obtain sufficient adhesion of polyacrylamide to the capil-lary walls, a primer step was required. A low molecular weightpolyacrylamide was bound to an alkoxy silicate and injected intothe device for covalent immobilization. Only thereafter could thepolyacrylamide precursor solution be injected. Furthermore, thecatalyst for the polymerization (gelation), reaction had to bedrawn into the device and diffusion mixed into the volume ofinterest. Diffusion is a relatively slow process, and can often resultin large heterogeneity in the final gel. These multiple reactions andhandling issues can be avoided by working with sol-gel-derived sil-ica materials that provide covalent bonding of a ready-to-useporous medium in a single step.

CELL IMMOBILIZATION MATRICES:

The immobilization of cells within solid, three-dimensionalmatrices has been proposed for a number of cell-based applicationsincluding sensor systems, bioreactor supports, and tissue engineer-ing. Sol-gel derived materials are likely candidates for these sys-tems due to their high biocompatibility, large surface area, andopen pore structure. Cells most likely interact with the matrix sur-face through hydrogen bonding and charge-charge interactionswith ionized surface silanol groups.

In the implant community, the systemic and cyto-compatibil-ity of melt alumina and silica are renown.32, 33 Indeed, the bio-compatibility of many silica-based glasses is so high that theyare described as “bioactive.”34, 35 The host microenvironmentinside bioactive glasses is so favorable that cell and fiber pene-tration occurs spontaneously after implantation, from both softand hard tissue.36 As mentioned earlier, sol-gel-derived silica andalumina are not identical to their melt cousins. However, meltsilica and alumina are often “activated” in vivo through exposure

Table 1. Sample microanalytical systems and their reliance upon porous media.

Stationary PhaseCapillary Electrophoresis

Capillary Immunoassays

Size Exclusion Chromatography

Electrochemical Cells

Perm-Selective Membranes

Capillary Electrochromatography

Supported Liquid Membranes

Nucleic Acid " Biochips"

Surface Modification,Pore Size

Surface Modification,Pore Size

Support Liquid Junction

Stationary Phase

Solid Phase Support

Select Permeability of Ions

Stationary Phase

Support Liquid Membrane

Solid Phase Support

Mechanical Stability, Ion Permeability

Ion Permeability

Surface Modification,Pore Size

Mechanical Stability

Mechanical Stability, Surface Modification

Surface Modification,Pore Size 11-15

22

37-38

39

23

16

24

17-20

to high pH conditions and yet retain their biocompati-bility. It is thus strongly suspected that sol-gel-derivedsilica and alumina will provide similar cytocompatibility.

Traditional sol-gel methods involve extremes of pH andhigh concentrations of alcohol, both of which can destroybiological molecules. However, preparation of silica solscan be accomplished without the addition of alcohol at thestart of the reaction. In addition, reaction conditions canbe adjusted with a buffer to a biologically compatible pH,after the acid catalyzed hydrolysis of alkoxy metallate pre-cursor and before the biological molecule is added.Although this work is still in its infancy, several groupshave investigated cell and cell extract immobilization insol-gel matrices.

Extracts of Thiobacillus thioxidans were doped into asol-gel and the reduction of H2S was tested for andobserved, indicating that the biological activity of theenzymes remained intact.37 This group also trapped extracts ofPseudomonas stutzeri and Acinetobacter johnsonii, importantorganisms in anaerobic denitrification, in a sol-gel-derivedmonolith.21 Enzymatic activity was monitored by following theconversion of nitrate to nitrite, albeit at somewhat lower levelsthan controls. In another study, Inama et al.,38 investigated thetrapping of Saccharomyces cerevisiae in thin layers of silica sol-gelon glass surfaces. The authors showed that the trapped yeast cellsurvived and remained enzymatically active for at least 15months. Braun et al.,39 entrapped whole E. coli cells in a sol-geland observed that the cells are trapped as aggregates, the size ofwhich increased with cell concentration. In this study however,cell viability was not reported.

In our laboratory, we added genetically engineered E.coli to amacroporous sol-gel-derived matrix40 and subjected them toaerosols of bacteriophage virus containing the gene required toactivate Green Fluorescent Protein in the entrapped bacteria(Figure 3).41 Fluorescence was detected microscopically and indi-cated that the bacteria remained viable and that the virus was ableto penetrate the porous matrix.

ACOUSTIC APPLICATIONS

Supercritically-dried aerogels have been proposed for severalapplications such as impedance matching layers for air-coupled

medical ultrasound imaging.42-48 Since the acoustic impedance ofaerogel is essentially tunable over the range between the acousticimpedance of air and many solid materials, it may provide a tailor-made material for maximizing acoustic transmission between differ-ent media. Such tuning has been done primarily through manipu-lation of the production reaction conditions, and a wide range ofsol-gel densities and acoustic velocities have been achieved.

In our lab, we have conducted fundamental studies of acousticpropagation in the lightest of sol-gel-derived materials, namelyaerogel.49, 50 With a very low acoustic impedance, aerogel is easierto probe using air-coupled (as opposed to contact mode) transduc-ers. Furthermore, given the extremely low density of commonaerogels, the interstitial gas exerts a significant influence upon thebulk acoustic properties of the medium (Figure 4). The effects ofthe interstitial gas are quite unusual, due, in part, to the extremelysmall pore size of aerogel. Since the mean free path of commongases at STP is larger than the mean pore diameter inside aerogel,the interstitial gas dynamics are akin to the gas dynamics inside aUHV chamber. Traditional models for describing acoustic propa-gation in porous media are thus of limited value, and aerogels dis-play unique acoustic behavior. Examples of this unique behavior atultrasonic frequencies include: 1) a nearly linear increase in attenu-ation with frequency (constant Q) in the solid skeleton in theultrasonic frequency range,44 2) a decrease in acoustic velocity with

Figure 3. (A) SEM micrograph of E. coli (pET-gfp) adherent to and growing within the matrixof EBO8 aerogel. (B) SCLM micrograph of E. coli expressing GFP after contact withaerosolized bacteriophage l.

Figure 4. (Below) Acoustic velocity through anaerogel as a function of interstitial gas pressure.The acoustic velocity increases several fold withincreased pressure, in contrast to ideal gasseswhere acoustic velocity is independent of pres-sure. The aerogel, in effect, has a heightened sen-sitivity to pressure changes in the molecular gas.

compressive strain,43 3) unexplained attenuation bands at distinctultrasonic frequencies,42 and, 4) a decrease in attenuation by thesolid skeleton with increasing temperature.44 At the current time,understanding aerogel acoustic behavior is limiting the exploita-tion of aerogel in many potential applications.

CONCLUSION

The sol-gel production route is a convenient method for theproduction of ceramics and ceramic composites for medicine andbiology. The utility of sol-gel-derived materials is based upon theirlow temperature production route, tunable porosity andmicrostructure, ease of surface functionalization, and liquid pre-cursor state that allows even perfusion of dopant molecules andthe formation of complex shapes in situ by gelation molding. Thehigh porosity of hydrogels is critical to wet applications likebiopolymer immobilization and column chromatography, where aliquid-like environment is desirable. Likewise, in dry applicationslike acoustics, the high porosity of aerogels leads to an unexpected-ly low acoustic velocity and several unique acoustic properties.Through a better understanding of the fundamental physicalproperties of sol-gel-derived materials, the engineering applicationsof these materials will increase and improved instrumentation andanalytical techniques can be developed.

REFERENCES

1. M. J. Marquissee and L. S. Sandell, US 4,090,887, 1978.2. S. T. Mayer, J. L. Kaschmitter, and R. W. Pekala, US

5,601,938, 1997.3. J. Brinker and G. Scherer, Sol-Gel Science, Academic Press,

New York, 1989.4. Y. A. Attia, ed., Sol-Gel Processing and Applications, Plenum

Press, New York, 1994.5. B. Riegel, I. Hartmann, W. Kiefer, J. Gross, and J. Fricke, J.

Non-Cryst. Solids, 211 (1997) 294.6. M. W. Connor and C. Colmenares, J. Non-Cryst. Solids, 201

(1996) 76.7. W. A. Aue and C. R. Hastings, J. Chromatog., 42 (1969) 319.8. J. J. Kirkland and K. Koehler, US 5,108,595, 1992.9. A. van Blaaderen and A. Vrij, Langmuir, 8 (1992) 2921.10.E. P. Plueddemann, Silane Coupling Agents, Plenum Press,

New York, 1991.11.S. F. Rouanet, W. E. McGovern, W. Cao, J. M. Moses, A. L.

Carrillo, and I. M. Klotz, US 5,725,836, 1998.12.U. Deschler, P. Kleinschmit, and P. Panster, Agnew. Chem. Int.

Ed. Engl., 35 (1986) 1420.13.Y. Guo and A. Colon, Anal. Chem., 67 (1995) 2511.14.Y. Guo, G. A. Imahori, and L. A. Colon, J. Chromatogr. A,

744 (1996) 17.15.J.-T. Wu, P. Huang, M. X. Li, M. G. Qian, and D. M.

Lubman, Anal. Chem., 69 (1997) 320.16.J. J. Kirkland, US 3,782,075, 1974.17.H. Minakuchi, K. Nakanishi, and N. Soga, Anal. Chem., 68

(1996) 3498.18.Y. Hu and P. W. Carr, Anal. Chem., 70 (1998) 1934.19.A. M. Clausen and P. W. Carr, Anal. Chem., 70 (1998) 378.20.B. C. Dave, B. Dunn, J. S. Valentine, and J. I. Zink, Anal.

Chem., 66 (1994) 1120A.21.D. Avnir, S. Braun, O. Lev, and M. Ottolenghi, Chem. Mater.,

6 (1994) 1605.22.L. M. Ellerby, C. R. Nishida, F. Nishida, S. A. Yamanaka, B.

Dunn, J. S. Valentine, and J. I. Zink, Science, 255 (1992)1113.

23.S. A. Yamanaka, C. R. Nishida, L. M. Ellerby, C. R. Nishida,B. Dunn, J. S. Valentine, and J. I. Zink, Chem. Mater., 4(1992) 495.

24.V. Glezer and O. Lev, J. Am. Chem. Soc., 115 (1993) 2533.25.S. Shtelzer and S. Braun, Biotechnol. Appl. Biochem., 19

(1994) 293.26.R. Wang, U. Narang, P. N. Prasad, and F. V. Bright, Anal.

Chem., 65 (1993) 2671.27.L. Yang and S. S. Saavedra, Anal. Chem., 67 (1995) 1307.28.C. Hsueh and M. Collinson, J. Electroanal. Chem., 420

(1997) 243.29.P. Lacan, C. Guizard, P. Le Gall, D. Wettling, and L. Cot, J.

Membrane Sci., 100 (1995) 99.30.A. T. Woolley and R. A. Mathies, Proc. Natl. Acad. Sci.

U.S.A., 91 (1994) 11348.31.A. T. Woolley and R. A. Mathies, Anal. Chem., 67 (1995)

3676.32.S. Hulbert, in An Introduction to Bioceramics (L. L. Hench

and J. Wilson, eds.), World Scientific, Singapore, 1993, 25.33.J. A. Miller, J. D. Talton, and S. Bhatia, in Clinical

Performance of Skeletal Prosthesis (L. L. Hench and J. Wilson,eds.), Chapman and Hall, London, 1996, 41.

34.V. Gross, R. Kline, H. J. Schmitz, and V. Strunz, CRC Crit.Rev. Biocomp., 4 (1988) 2.

35.L. L. Hench, in Bioceramics: Materials Characteristics versusIn-Vivo Behavior, Vol. 523 (P. Ducheyne and J. Lemmons,eds.), Ann. New York Acad. Sci., 1988, 54.

36.J. Wilson, G. H. Pigott, F. J. Schoen, and L. L. Hench, J.Biomed. Mater., 15 (1981) 805.

37.R. Armon, C. Dosoretz, J. Starosvetsky, F. Orshansky, and I.Saake, J. Biotech., 51 (1996) 279.

38.L. Inama, S. Dire, G. Carturan, and A. Cavazza, J. Biotech.,30 (1993) 197.

39.S. Braun, S. Rappoport, S. Shtelzer, R. Zusman, S.Druckmann, D. Avnir, and M. Ottolenghi, in Biotechnology:Bridging Research and Applications (D. Kamely, ed.), KluwerAcademic Publishers, 1993, 205.

40.K. Tadanaga, K. Iwashita, T. Minami, and N. Tohge, J. Sol-GelSci. Tech., 6 (1996) 107.

41.M. E. Power, P. M. Norris, and C. E. Daitch, The BIOGELCellular Support Matrix, SEAS Report No.UVA/532567/MAE00/101, July, 1999.

42.V. Gibiat, O. Lefeuvre, T. Woignier, J. Pelous, and J.Phalippou, J. Non-Cryst. Sol., 186 (1995) 244.

43.H. Altmann, T. Schlief, J. Gross, and J. Fricke, in UltrasonicInternational 91 Conf. Proc., (1991) 261.

44.A. Zimmerman, J. Gross, and J. Fricke, J. Non-Cryst. Sol. 186(1995) 238.

45.P. P. Narang, Appl. Acoust., 34 (1991) 249.46.J. Gross, J. Fricke, R. W. Pekala, and L. W. Hrubesh, Phys.

Rev. B, 45 (1992) 12774.47.J. Gross and J. Fricke, J. Non-Cryst. Sol., 145 (1992) 217.48.J. Gross, J. Fricke, and L. W. Hrubesh, J. Acoust. Soc. Am., 91

(1992) 2004.49.J. F. T. Conroy, B. Hosticka, S. C. Davis, and P. M. Norris,

Proc. 1998 IMECE Conf. (1998) 25.50.J. F. T. Conroy, B. Hosticka, S. C. Davis, A. N. Smith, and P.

M. Norris, Microscale Thermophysical Engineering, In Press