maxillofacial prosthetics

Polyurethane Elastomers as Maxillofacial Prosthetic Materials

A. JON GOLDBERG, ROBERT G. CRAIG,* and FRANK E. FILISKOtDepartment of Restorative Dentistry, The University of Connecticut Health Center,Farmington, Connecticut 06032, USA, *Department of Dental Materials, School of DentalMaterials, School of Dentistry, The University of AMIichigan, Ann Arbor, Michigan 48109, USA.and tDepartment of Materials and Metallurgical Engineering, College of Engineering,The University of Michigan, Ann Arbor, Michigan 48109, USA

A series of polyurethane elastomers based onan aliphatic diisocyanate and a polyether mac-roglycol was polymerized with various cross-link densities and OH/NCO ratios. Stoichio-metries yielding between 8,600 and 12,900 gm/mole/crosslink and an OH/NCO ratio of 1.1resulted in polymers with the low modulus, yethigh strength and elongation necessary formaxillofacial applications.

J Dent Res 57 (4) :563-569, April 1978.

There are no completely satisfactory maxillo-facial reconstructive materials. The best maxil-lofacial prosthesis available today, althoughexcellent when originally delivered, deterior-ates in 6 to 12 months to the point where it re-quires replacement.'-4 This deterioration isassociated with either degradation of mechani-cal properties or changes in appearance. Thedeficient properties which most commonlycause mechanical failure are tear resistanceand general stiffening of the elastic material asa result of migration and leaching of the plasti-cizer. Degradation of esthetic qualities can bea result of color changes in the base polymercaused by oxidation and ultraviolet light, colorchanges in pigments and dyes used to char-acterize the prosthesis, or adsorption of dirt,grease, or cosmetics onto the surface and sub-sequent diffusion into the polymer.

Currently, the most widely used types ofmaxillofacial materials include rigid poly-(methyl methacrylate), plasticized poly (vinylchloride) or plasticized vinyl chloride-vinylacetate copolymers, and silicone rubbers. Sev-

Received for publication September 1, 1977.Accepted for publication October 25, 1977.This investigation was supported by USPHS Train-

ing Grant DE-00181 from the National Institute ofDental Research, National Institutes of Health, Bethes-da, Md 20014.

Portions of this article were presented at the AADRMeeting, New York, 1975.

Vol. 57 No. 4

eral other materials have recently been sug-gested including a latex-dispersed elastomer,5a silphenylene polymer,6 and a polycarbonate/silicone rubber block copolymer.7

Polyurethane elastomers have great po-tential as maxillofacial materials. This hy-pothesis is based on their inherent environ-mental stability, high tear resistance, lowmodulus without the use of plasticizers, andgood ultimate strength and elongation. Theycan accept intrinsic coloring and are amenableto maxillofacial processing techniques. Mostimportantly, the structure of the polyurethanepolymer can be varied to optimize the desiredproperties and as a result polyurethanes arecurrently used in a variety of biomaterial appli-cations.&21

Polyurethanes are based on componentdiisocyanates, macroglycols, chain extenders,and crosslinking agents. In order to develop asatisfactory maxillofacial material the opti-mum combination of chemistry and morphol-ogy must be utilized. Comprehensive litera-ture reviews of polyurethane structure-proper-ty relations and chemistry, including discus-sions of molecular weight, intermolecularforces, chain stiffness, tendency toward crystal-lization, and intermolecular bonding, can befound elsewhere.22-25 One unexpected aspectof crosslinking, however, does warrant briefdescription. Thermoset urethane elastomers areproduced through the introduction of primarychemical crosslinks. In most practical poly-urethane systems, the crosslinking is a com-plex combination of bonds resulting from tri-functional chain extenders, allophanate andbiuret links, and physical crosslinks associatedwith paracrystalline domains.26'27 However, ifthe reactants are combined in stoichiometricratios, and the reactions are preferentially cat-alyzed, a known, controlled morphology can be

563

564 GOLDBERG, CRAIG & FILISKO

N

zE 4.0

J

o 3.0

z 2.0

LU

3 6 9 12 15

Mc X 10-3

FIG 1.-100% tensile modulwlecular weight per crosslink forurethane) elastomer. (Adapted fron

developed. This approach was usec

to evaluate the effect of crosslink d(tensile properties of poly(ester ure

interesting minimum occurred in100% modulus versus molecularcrosslink, reproduced in Figure 1.is contrary to the increasing m

crosslink density seen with hydrocamers, and can be explained on thesecondary bonding present in polthanes). As crosslinks are addedmer, the initial effect is to inhibiof the polyurethane chains, decamount of secondary bonding, an

modulus and strength. With cc

creased crosslink concentrationreached (molecular weight per croproximately 5,000) where the cros

selves start to support a significanthe load. Further increases in cros

now raise the 100% modulus andOn the basis of the knowle

urethane elastomers it was hypoth.polyether macroglycol in combinaaliphatic diisocyanate could provicsary environmental stability for a r

material. This polyether-aliphaticinherently flexible. If, in the bulksecondary bonding, and moduluspressed, as described above, yet usile properties maintained, a signproved prosthetic material could b

* LD2699, now designated AdipreneDupont, Wilmington, De. Lot numbers 1

t Aldrich Chemical Co., Milwaukee,ber 051217, 99 +%.

: Celanese Chemical Company, New§ Catalyst T-9, M&T Chemicals, Inc.* * Drierite, W. A. Hammond Drie

Xenia, Oh.tt Linde 5A molecular sieve, 1/6-inc

The aim of this study was to formulate,prepare, and optimize the tensile properties ofsuch a polymer, through control of the cross-

- link density and NCO/OH ratio.

Materials and Methods

Poly(tetramethylene oxide) glycol, PTM-EG, and hexamethylene diisocyanate, HMDI,were selected as the macroglycol and diisocya-18 21 24nate respectively. These components were re-ceived in the prepolymer form* having a mo-

s versus molecular weight of approximately 1,000 gm/a poly(ester mole, and contain 4.75 ± 0.15% active NCOn Pigott.) end groups. A prepolymer system was selected

because it can be cast into open molds, a pre-I by Pigott28 ferred technique in producing maxillofacialensity on the prostheses. The PTMEG/HMDI pre-polymer.thanes). An was light amber in color and was a viscousthe plot of liquid at room temperature. The pre-polymerweight per was cured with the difunctional polyol 1,4-This artifact butanediol,t and trifunctional trimethylol pro-odulus with pane,+ TMP. Because of the low reaction rateirbon elasto- between the aliphatic diisocyanate and polyolbasis of the curing agent, a stannous octoate catalyst,§ was

Iy(ester ure- incorporated at the time of mixing.to the poly- The prepolymer was received sealed underit alignment nitrogen. After each use, the metal containercreasing the was purged with dry nitrogen and resealed.Id hence the This procedure was also used with the curingntinued in- agents and catalyst, which were stored in aa point is desiccator between uses. The nitrogen was

)sslink of ap- dried by bubbling it through sulfuric acid, pass-sslinks them- ing it through a column of sodium hydroxideIt amount of pellets, and finally through a column of anhy-;slink density drous calcium sulfate.** It was critical thatI strength. moisture contamination be avoided because ofIge of poly- the hydrophilic nature of the reactants. Con-esized that a taminated materials could upset the stoichio-tion with an metric balance of the polyurethane reaction.le the neces- The prepolymer casting technique was de-maxillofacial veloped from the supplier's recommendationsmolecule is as well as suggestions from other researchers,29polymer, the and consisted of the following steps:could be de- 1. The desired ratios of 1,4-butanediol andiltimate ten- trimethylolpropane were mixed. TMP is aLificantly im- solid at room temperature, but if both polyolse realized. are heated and mixed, the resulting solution

is liquid at room temperature, facilitating fur-LW520, E. I. ther handling.

W07, 122, 131. 2. Twenty four hours before sample prep-

York, NY. aration, the catalyst was accurately weighed., Rahway, NJ. into the curing agent mixture. Approximatelyrite Company, one tablespoon of molecular sievett was addedch pellets. per liter of catalyst-cure to remove any remain-

IJ Dent Res April 1978

POLYURETHANES AS MAXILLOFACIAL MATERIALS 565

ing moisture. The entire mixture was slowlyrotated for 24 hours before sample preparation.

3. The prepolymer was carefully weighedinto a disposable polypropylene beaker andheated to 100 C under a 5 mm Hg vacuum ina 2,000-ml reaction kettle. Degassing was con-

sidered complete when excessive bubbling ofthe prepolymer stopped.

4. Simultaneously with step 3, a stoichio-metric excess of the cure-catalyst mixture was

poured into a second disposable polypropylenebeaker, heated and degassed in a reactionkettle.

5. When thoroughly degassed and heated to100 C, the prepolymer was removed from thereaction kettle and reweighed. Using an eye-

dropper, the cure-catalyst mixture was care-

fully weighed into the prepolymer. The cure-

catalyst mixture was added to within 1% ofthe desired amount. This step was accomp-

lished in 60 seconds.6. The prepolymer-cure-catalyst mixture was

mixed for 60 seconds with a variable speedstainless steel propeller. Mixing speed was con-

trolled to allow maximum mixing withoutwhipping bubbles into the liquid polymer.

7. The liquid polymer was replaced intothe reaction kettle and degassed for an addi-tional 120 seconds at 100 C and 5 mm Hg.

8. The liquid polymer was then cast intopreheated, fluorocarbon coated,* stainless steelmolds. Tensile specimens for the polyether-aliphatic prepolymer were compression-mold-ed and cured in the press at 100 C for 8 hours.

The optimum stoichiometry of the ali-phatic-polyether elastomer for maxillofacialapplication was determined by simultaneouslyvarying the crosslink density and prepolymerto curing agent ratio. The crosslink density wascontrolled by reacting the prepolymer withvarious 1,4-butanediol/TMP ratios. Increasedamounts of TMP resulted in higher cross-

linked densities. The various ratios of 1,4-butanediol/TMP along with the resulting mo-

lecular weights per crosslink, Mc, are listedin Table 1. The total equivalent weight of cur-

iilg agent to prepolymer, identified as the OH/NCO ratio, was varied from a 10% deficiency,to stoichiometric amount, to 10% excess.

Five dumbbell-shaped tensile specimenswere prepared with each different 1,4-butane-diol/TMP and OH/NCO combination, for a

total of 105 specimens. Each group of five

$ Fluoro Glide, Chempast, Inc., Wayne, NJ.Instron Corporation, Canton, Ma.

TABLE 1

1,4-BUTANEDIOL/TRIMETHYLOLPROPANE RATIOSAND RESULTING MOLECULAR WEIGHTS

PER CROS SLINK

1,4-Butanediol/TMP Molecular WeightRatio Per Crosslink, Mc

0/100 2,60025/75 3,40050/50 5,10060/40 6,40070/30 8,60080/20 12,900100/0 No primary crosslinking

specimens was poured from one batch of ma-terial, and compression-formed in the samemold. Catalyst concentration was adjusted foreach batch so that all specimens were molded10 to 15 minutes after mixing of the prepoly-mer and curing agent. The fluorocarbon-coatedstainless steel mold was recoated after everyfour to six uses. All samples were inspected at5 X magnification for nicks, tears, and bubbles.Defective samples were discarded. Molded ten-sile specimens had gauge dimensions of 3.2 X3.2 X 25.4 mm (I8 X I/8 X1 inch). All tensilesamples were tested on a constant strain-ratetesting machine,t at room temperature, usinga crosshead speed of 5 cm/min. Strain wasmeasured with an incremental extensometer,tusing a one-inch gauge length.

Results

Overall, there was a greater variation inproperties with the 1,4-butanediol/TMP ratiothan with the OH/NCO ratio. This effect isunderstandable since the former ratio has amore drastic effect on crosslink density. Withincreasing relative amounts of the difunctionalcuring agent 1,4-butanediol, and consequentlya decrease in crosslink density, there was a gen-eral increase in strength and a fourfold increasein elongation. The final elongation is plotted asa function of molecular weight per crosslink,Mc, for the three different OH/NCO ratios inFigure 2.

For any given crosslink density, there wasa relatively small change in tensile propertieswith OH/NCO ratio. This observation wasparticularly true under 250% elongation. Fig-ure 3 contains the stress-elongation data for thesamples prepared with a 70/30 1,4-butanediol/TMP ratio, and is representative of the effect

1. ol. 57- No. 4


00!otD

J

6O0

500

400

300

200

100

OH/NCO= 1.I

V0.9

- D- - - o

KI -EI

2 4 6 8 10 12 ' XMOLECULAR WEIGHT PER CROSSLINK,Mc X10-3

FIG 2.-Ultimate elongation versus molecu-lar weight per crosslink for the aliphatic/poly-ether system; OH/NCO = 0.9, 1.0, 1.1.

of varying the OH/NCO ratio with a givenMc. There was a slight depression of the 100%modulus (stress at 100% elongation), but anincrease in ultimate tensile strength and finalelongation, with increasing OH/NCO ratio.

The stress-elongation curves representingthe various 1 ,4-butanediol/TMP ratios, butwith a common OH/NCO ratio of 1.0 aregrouped in Figure 4. This figure, in essence,shows the change in tensile properties withcrosslink density. The calculated molecularweight per crosslink is shown for each curve.

The stress-elongation curves measured atconstant crosshead speed and temperaturewere typically nonlinear, and represented thetime-dependent viscoelastic properties of theseelastomers. For maxillofacial prosthetic appli-cation the modulus of the material is of partic-ular importance, because flexibility affects bio-compatibility as well as esthetics. In order to

14

12

10

F::z

V1

8 OH/ NCO............

.0.9

6 _ -1.0

6[1.

4

200 300

% ELONGATION

400

FIG 3.-Stress-elongation curves for thealiphatic/polyether system; 1,4-butanediol/TMP = 70/30, OH/NCO - 0.9, 1.0, 1.1.

make a better comparison of the time-depend-cnt stress-elongation data, the equilibriumvalues of the modulus were calculated follow-ing the scheme of Smith and Magnusson.30

The first step in calculating an equilib-rium modulus is to establish a criteria whichdemonstrates that the values do in fact repre-sent equilibrium behavior. The Smith andMagnusson treatment centers on the stress-strain equation predicted by the kinetic theoryof elasticity rubber.

S=VRT (a- 1/a2)

where S is the engineering stress, V is the molesof effective chains per units volume, R is thegas constant, T is the absolute temperature,and a is the extension ratio equal to [A 1/1o] +1. These authors showed that a plot of aSversus 8, where 8 equals Al /lo, yields a straightline for most elastomers, and the slope yields a

14

12

10

8Ez8

*d

UrI,, 6

A4

2

0

Mc =

2,900

n100

200 300

% ELONGATION

400 500

FIG 4. Stress-elongation curves for thealiphatic/polyether system; OH/NCO = 1.0.

precise value for the modulus, E. The value ofE thus calculated represents an equilibriumbehavior if the slope of the log (aS) versuslog (8) plot is unity. This stems from the factthat the exponent, n, in the equation

caS = E8n

is unity for an equilibrium elastic response.The calculated value of E should be close to3G, where G is the bulk modulus obtainedfrom the slope of a S versus a- 1/a2 plot.

Using the stress-elongation data from Fig-ure 4, plots of aS versus 8 (shown in Fig 5),S versus a-1/a2, and log (aS) versus log (8)were made. The slopes of the log (aS) versuslog (8) plots were linear, indicating that an

I Dent Res April 1978

X


IE

14

Lf

uL

12

10

8

0

&

FIG 5.-Sa versus 8 for the aliphatic/poly-ether system calculated from the data inFigure 4.

equilibrium modulus was indeed determined.The modulus, E, calculated from the aS versus8 plots, the bulk modulus, G, calculated fromthe S versus a-l/a2 plots, and the slopes oflog (aS) versus log (8), n, are listed inTable 2.

Discussion

Although described in detail in the Meth-ods section, several aspects of the technique

used to polymerize the polyurethane elastomerswarrant discussion. First, the importance ofavoiding contamination by moisture cannot beunderstated. The presence of moisture duringthe reaction can completely eliminate the re-activity of the isocyanate groups on the ure-thane prepolymers. In the severe case, thepolymerization can be minimal, resulting in aviscous liquid as a final product. Minor con-tamination allows polymerization, but the re-sultant molecular weight and crosslink densityare below the theoretical values. Moisture con-tamination was minimized by maintaining allreactants and the catalyst in an atmosphere ofdry nitrogen, and storing the containers in adesiccator. The use of the 5-Angstrom molec-ular sieve was an additional aid in removingmoisture from the premixed curing agents. Allhandling and mixing was performed rapidly,and containers were purged and resealed asquickly as possible. Preparation of the ure-thane materials was suspended when condi-tions of high humidity existed in the labor-atory.

Oxidation of the catalyst from the stannousto stannic octoate form is also a potential de-terrent to achieving theoretical crosslink den-sity and molecular weight. As indicated in theIntroduction, the aliphatic urethane prepoly-mer required a catalyst for the curing reac-tion with 1,4-butanediol and trimethylol pro-pane. It would have been desirable to cure thepolyurethane without a catalyst, since this ad-ditive could potentially have a deleteriouseffect on the biocompatibility, and at highertemperatures, the hydrolytic stability of thefinal polymer. However, attempts at polymeri-zation without a catalyst at higher tempera-

BLE 2

EQUILIBRIUM MODULUS, E, BULK MODULUS, G, LINEARITY EXPONENT, n,FOR THE POLY(ETHER URETHANE) ELASTOMERS, OH/NCO = 1

G,MN/m2 3G,MN/m2 E,MN/m2Mc n (psi) (psi) (psi)

2,600

3,400

5,100

12.900

oo

1.070

0.992

0.968

1.002

0.943

1.21(176)0.96(140)0.48(70)1.34(194)1.88(273)

3.64(528)2.89(420)1.46(212)4.02(582)5.65(819)

4.14(600)3.39(492)1.55(225)4.78(693)7.52

(1,090)

Vol. 57 No. 4


tures for even as long as 24 hours were unsuc-cessful. Stannous octoate was selected becauseit has been successfully used as a catalyst withsiloxane polymers designed for biological en-vironments. In the presence of moisture, thestannous octoate can be oxidized to a statewhich not onlv eliminates its ability to func-tion as a catalyst, but actually inhibits the iso-cyanate-polyol reaction. This condition occursif the relative amount of stannic octoate ex-ceeds 8%. Unfortunately, the change is notreadily identified until all components havebeen mixed. The result is either no polymeriza-tion or a low molecular weight material, de-pending on the degree of contamination.Usually, the reaction can be accelerated withthe incorporation of additional catalyst, al-though this approach is precarious at best andcan lead to extremely fast reaction rates if ex-cessive catalyst is used.

It is apparent then that incorporation ofthe curing agents and catalyst into the heatedurethane prepolymer must be rapid to avoidmoisture contamination, but also accurate.This procedure was accomplished by re-weigh-ing the prepolymer after heating at 100 C andjust prior to addition of the cure-catalyst mix-ture. The weights were recorded to within 0.01gram and the necessary amount of cure-catalystwas calculated on this basis. With batches ofapproximately 50 gm, calculations indicatedthat the molecular weight per crosslink wascontrolled to within a value of 50 gm/mole/crosslink. The success in controlling morphol-ogy is reflected in the consistent change inproperties with Mc.

Table 2 lists the equilibrium modulusvalues for the poly(ether urethanes) preparedwith Mc values of 2,600 to infinity. These dataagree wvith those of Pigott28 in that initiallythere is a depression of the modulus, followedby an increase in the modulus with crosslink-ing density. The minimum modulus occursnear an Mc value of 5,000. As described ear-lier, the decrease in modulus is caused by theinterference in intermolecular bonds by thetrimethylolpropane curing agent. The expon-ent "n" values in Table 2, which theoreticallyshould be unity, indicate that equilibriummodulus values were in fact obtained.

The polyurethane elastomer based on thealiphatic diisocyanate HMDI, and a polyethermacroglycol, PTMEG, was selected on thebasis of its potential physical as well as me-chanical properties. The inherent chemistry of

the reactants would provide good environ-mental stability, but the main challenge wasto optimize, in one polymer, the mechanicalproperties necessary for a successful maxillo-facial material. Tensile data were the primarydeterminants in this optimization process.

The elongation data in Figure 2 suggestedthat an OH/NCO ratio of 1.1 would be desir-able, since the lower isocyanate concentrationresulted in a polymer with higher elongation.The ratio of 1.0 would have been acceptable,while the formations with OH/NCO ratios of0.9 had insufficient elongation. The materialswith OH/NCO ratios of 1.1 had lowerstrengths, but not significantly below thosesamples with OH/NCO ratios of 1.0. There-fore, 1.1 was selected as the preferred OH/NCO ratio.

The crosslink density for the poly(etherurethane) was selected by reviewing the elon-gation data in Figure 2, the stress-elongationcurves in Figure 4, and the equilibrium datain Figure 5. A minimum molecular weightper crosslink of approximately 6,000 was nec-essary for adequate elongation and strength.Mc values exceeding 13,000, however, resultedin higher than desirable values of modulus.Therefore, final formulations were selectedwhich resulted in OH/NCO ratios of 1.1 andmolecular weights per crosslink between 6,000and 13,000. Two specific polymers were pre-pared for further testing. These polymers werebased on 1',4-butanediol/trimethylol propaneratios of 80/20 and 70/30, which resulted inelastomers with Mc values of 12,900 and 8,600,respectively. Results of additional testing willbe reported in subsequent articles.

Conclusions

A polyurethane elastomer based on analiphatic diisocyanate and a polyether macro-glycol can provide the necessary tensile prop-erties for a satisfactory maxillofacial prostheticmaterial. An optimum balance of tensile prop-erties for the poly(ether urethane) was obtainedwith a molecular weight per crosslink between8,600 and 12,900 gm/mole and a 10% stoich-iometric excess of the curing agents. The re-sulting polymer had a tensile strength of 11MN/M2, 100% modulus of 2MN/m2, andultimate elongation of 500%. Satisfactoryproperties were obtained for the poly ( etherurethane) without the use of plasticizers orstabilizers.

J Dent Res April 1978


References

1. BULBULIAN, A.H.: Facial Prosthetics,Springfield: Charles C Thomas, 1973.

2. CASTLEBERRY, D.J.: Materials for ExternalProstheses; The Choices and the Challenge,Symposium on Dental Biomaterials Re-search Priorities, NIDR, Des Plaines, Il, Aug1973.

3. FIRTELL, D.N., and BARTLETT, S.O.: Maxil-lofacial Prosthetics: Reproducible Fabrica-tion, J Prosth Dent, 22:247-252, 1969.

4. LEPLEY, J.B., and STRAUSS, P.W.: Func-tional and Cosmetic Restoration of a MassiveOrofacial Defect, J Prosth Dent, 30: 635-636, 1973.

5. CHALIAN, V.A.; DRANE, J.B.; and STAN-DISH, S.M.: Maxillofacial Prosthetics: Mul-tidisciplinary Practice, Baltimore: Williamsand Wilkins Co., 1971, p 96.

6. LEWIS, D.H.; MILLER, D.R; and COWSAR,D.R.: Development of Improved Materialsfor Extraoral Maxillofacial Prostheses,P.H.S., N.I.D.R. Contract NOL-DE-42435,June 1976.

7. ROBESON, L.M.; SAUNDERS, W.J.; CHOW,S.; and MATZNER, M.: Development of Im-proved Materials for Extraoral Prostheses,P.H.S., N.I.D.R. Contract NOL-DE-42436,November 1976.

8. LYMANN, D.J.; KWAN-GETT, C.; ZWART,H.H.J.; BLAND, A.; EASTWOOD, N.; KAWAI,J.; and KOLFF, W.J.: The Development andImplantation of a Polyurethane Hemispheri-cal Artificial Heart, Trans Am Soc Artif In-tern Organs, 17:456-463, 1971.

9. KOLFF, T.; BURKETT, G.; and FEYEN, J.:Copoly(ether-urethanes) as Materials forArtificial Hearts, Biomater Med DevicesArtif Organs 1:669-679, 1973.

10. GARDNER, D.L.: Stability of Heparin Sul-fur-35 Attached to a Modified PolyurethaneVascular Prosthetic, Trans Am Soc ArtifIntern Organs, 15: 7-14, 1969.

11. BRASH, J.L.; FRITZINGER, B.K.; and BRUCK,S.D.: Development of Block Copolyether-Urethane Intraaortic Balloons and OtherMedical Devices, J Biomed Mater Res, 7:313-334, 1973.

12. BORETOS, J.W., and PIERCE, W.S.: Seg-mented Polyurethane: A Polyether Polymer,J Biomed Mater Res 2:121-130, 1968.

13. BORETOS, J.W.; DETMER, D.E.; and DON-ACHY, J.H.: Segmented Polyurethane: APolyether Polymer, II: Two Years Experi-ence, I Biomed Mater Res, 5: 373-387, 1971.

14. SKORNIK, W.A.; DRESSLER, D.P.; and RHI-CARD, J.K.: Adherence of Prosthetic Skin,J Biomed Mater Res, 2:447-456, 1968.

15. DRESSLER, D.P.; BARLYN, L.; and SKORNIK,W.A.: Viable Prosthetic Interface, J BiomedMater Res Symposium, 1: 169-178, 1971.

16. SHARP, W.V.; TAYLOR, B.C.; WRIGHT, J.;and FINELLI, A.F.: Experience with Nega-tively Charged Polyurethane Backed Ve-lours, J Biomed Mater Res Symposium, 1:75-81, 1971.

17. SHARP, W.V.; GARDNER, D.L.; and ANDRE-SEN, G.J.: Bioelectric Polyurethane and ItsUse in Internal Prostheses, U.S. Patent 3,512,183 (cl. 3-1; a6lf), 19 May, 1970.

18. LLEWELLYN-THOMAS, E.; WANG, P.Y.; andCANNON, J.S.: Adhesion of Synthetic Or-ganic Polymer on Soft Tissue. I. Fast SettingPolyurethane Adhesive, J Biomed Mater Res,8:35-43, 1974.

19. BERNHARD, W.F.; CUMMIN, A.S.; HARRIS,P.D.; and KENT, E.W.: New Flexible Vascu-lar Adhesive for Use in Cardiovascular Sur-gery, Circulation, 27:739-741, 1963.

20. SHERMAN, R.T.: Evaluation of Polymers asSurgical Prostheses, U.S. Clearinghouse Fed.Sci. Tech. Inform., AD 1969, AD690700.

21. PHILLIPS, R.W.: Science of Dental Materi-als, 7th ed., Philadelphia: W. B. Saunders,1973, p 239.

22. SAUNDERS, J.H., and FRISCH, K.C.: Poly-urethanes, Chemistry and Technology, Tech-nology, Parts I and II, in High Polymers V.XVI, New York: Interscience, 1962 and1964.

23. WRIGHT, P., and CUMMING, A.P.C.: SolidPolyurethane Elastomers, New York: Gor-don and Breach, 1969.

24. SAUNDERS, J.H.: Polyurethane Elastomers,in Polymer Chemistry of Synthetic Elasto-mer, Kennedy, J.P. and Tornquist, E.G.M.,ed., New York: John Wiley, 1969.

25. FRISCH, K.C.: Recent Advances in theChemistry of Polyurethanes, Rubber ChemTechnol, 45:1442-1466, 1972.

26. BONART, R.; MOBITZER, L.; and HENTZE,G.: X-ray Investigations Concerning thePhysical Structure of Crosslinking UrethaneElastomers. II. Butanediol as Chain Extend-er, J Macromol Sci Phys, 3:339-358, 1969.

27. BONART, R.; MORBITZER, L.; and MUELLER,E.H.: X-ray Investigations Concerning thePhysical Structure of Crosslinking in Ure-thane Elastomers, III. Common StructurePrinciples for Extensions with Aliphatic Di-amines and Diols, J Macromol Sci Phys, 9:447-461, 1974.

28. PIGOTT, K.A., and FRYE, B.F.: Data, JChem Eng, 5:391, 1960.

29. DIETER, C.: Personal Communication, Dept.of Polymer Science, University of Detroit,Detroit, Michigan, 1972.

30. SMITiH, T.L., and MAGNUSSON, A.B.: Diiso-cyanate-Linked Polymers. II. Mechanicaland Swelling Properties of Some Polyure-thane Elastomers, J Polym Sci, 42:391-416.1960.

Vol. 57 Nio. 4

maxillofacial prosthetics

Documents