~ ~ ~~~~

Thermomechanical investigation of poly(methylmethacry1ate) Fontaining an organobismuth radiopacifying additive

H. Ralph Rawls,l,* Randall J. Granier,' Johannes Smid,2 and Israel Cabasso2 'University of Texas Health Science Center, Sun Antonio, Texas 78284-7890; 2State University of New York, College of Environmental Science and Forest y, Polymer Research Institute, Syracuse, New York 13210

Previously we demonstrated the feasibility of using up to 24% triphenylbismuth (TPB) as a radiopaque, monomer-mis- cible additive for dental acrylic resins. In this study we exam- ined the influence of TPB on thermomechanical properties of a representative polymethylmethacrylate (PMMA) ambient- cured resin used for temporary dental crowns and bridges. TPB (OYO, 5%, 15% or 30% w/w) was dissolved in the mono- mer component, added to the powder component, and al- lowed to cure in rectangular molds. After l h they were either stored at 23°C for 23 h, or heated for 5 min at either 40°C or 50°C, and then stored for 23 h. They were then scanned from - 10" to 125°C in a dynamic mechanical thermal analyzer using the three-point bending mode of deformation at 1-Hz frequency.

The onset to the glass-transition temperature (Tg) is de- creased by 13" to 32°C by addition of TPB, while the storage modulus (E') at 25°C is either unchanged or is slightly in- creased. TPB did not interfere with the curing reaction, and

postcure heating at 40°C had no effect on either E' or T,. However, heating at 50°C generally increased Tg but had very little effect on E' throughout the 0-50°C operating tem- perature range. TPB crystals were observed to have precipi- tated at TPB levels above 8%. These crystals, dispersed throughout the PMMA, act as reinforcing fillers. This rein- forcement can account for the lack of a decrease in E', as would be expected if TPB had a plasticizing effect below T,. However, even at 5%, a concentration at which all the TPB remains dissolved in the solid polymer, no decrease in E' was observed. This implies that TPB exerts an antiplasticizing effect at temperatures below 50°C, possibly by occupying free volume among the polymer chains.

It is concluded that TPB, in amounts adequate to impart diagnostic levels of radiopacity, is unlikely to adversely affect the clinical utility of PMMA-based dental acrylic resins. 0 1996 John Wiley & Sons, Inc.

INTRODUCTION

Removable dental appliances such as dentures have traditionally been fabricated from radiolucent acrylic polymers. The inability to radiographically image this material when accidentally ingested, aspirated, or im- pacted during injury poses significant problems for health-care providers. The difficulty in locating aspi- rated dental foreign bodies, for instance, has contrib- uted to death rates as high as lo%.'

A variety of materials, including barium and bis- muth halides, have fallen short as radiopacity additives because of their tendency to change the physical prop- erties of acrylic polymers. Mastication places consider- able stress on dental appliances. Consequently dental materials must display impact strength and resistance

*To whom correspondence should be addressed at The University of Texas, Health Science Center at San Antonio, Div. of Biomaterials, 7703 Floyd Curl Drive, San Antonio, TX 78284-7890.

to mechanical fatigue. Since dental acrylic resins typi- cally have low impact strength and poor resistance to mechanical fatigue, a desirable radiopaque additive should not further degrade the material. We have shown that triphenylbismuth (TPB), unlike other addi- tives, is miscible and forms a single phase with many polymers.* With polymethylmethacrylate (PMMA) di- agnostic levels of radiopacity (2 radiopacity of Al) are attained with 14% TPB. Up to 24% w/w TPB can be incorporated without unacceptably altering critical physical, mechanical, and biocompatibility proper- ties.3 TPB, however, is known to suppress the glass- transition temperature, as observed by differential scanning calorimetry.* This implies that TPB is a plasti- cizer for PMMA and should soften the resin below its glass-transition temperature, in the operating tempera- ture range of 20-50°C. Instead we found that with up to 20% TPB, transverse flexural strength decreases only slightly and there are moderate increases in brittleness?

In order to discover the origin of this apparent con- flict in properties, we employed dynamic mechanical

Journal of Biomedical Materials Research, Vol. 31, 339-343 (1996) 0 1996 John Wiley & Sons, Inc. CCC 0021-9304/96/030339-05

340 JXAWLS ET AL.

thermal analysis (DMA) to further investigate the effect of TPB on PMMA. DMA is used to measure the storage modulus, E‘, and loss modulus, E”, as a function of temperature and rate of deformation. E’ determines inherent rigidity and depends on a material’s ability to store mechanical energy. The loss modulus, E”, has a strong influence on toughness and is dependent on a material’s ability to absorb and dissipate mechanical energy. The ratio tan 6 = E”/E’ defines the “dissipation factor,” tangent 6, which is a viscoelastic property and is dependent on the rate of relaxation in a material. Thus, tan 6 is a measure of the relative importance of both viscous and elastic behavior in a polymer. From these data the glass-transition temperature, T,, can be derived. T, is the temperature at which cooperative segmental motion occurs for amorphous polymers and is directly linked to viscoelastic behavior. At tempera- tures below T,, polymers remain rigid in the vicinity of T,; they exhibit plasticlike behavior, and at some point above T, they become rubbery and/or begin to flow. T, is usually reported as the high-temperature peak of tan 6. However, some argue that the onset of loss in E’, which occurs at temperatures lower than the tan 6 peak, signals the beginning of cooperative molecular motion and is a more useful measure of T,.6 For many practical applications the concept of soften- ing or deflection point, the temperature at which a material will begin to sag under a given load, is also useful .

Some soluble compounds disrupt interchain organi- zation and hence increase segmental motion, thereby decreasing the glass-transition temperature. Usually this also increases segmental mobility at temperatures below T,, softening the amorphous structure and re- ducing rigidity (i.e., E’ ) over a wide temperature range. This is the well-known plasticizing effect.

In previous studies TPB was incorporated only into the PMMA microbeads in solid solution, to which methylmethacrylate (MMA) monomer was added to form a curable, doughy resin. The resin materials used were designed for making removable dental appli- ances for long-term use. In the present study a model PMMA resin system, designed for fabricating tempo- rary dental crowns and bridges, was used. The TPB was added exclusively to the MMA monomer compo- nent. We then examined the effect of increasing the level of TPB on T,, on the deflection point, and on the storage modulus at 25°C (Er25C).

EXPERIMENTAL

Materials

PMMA specimens containing TPB were prepared using a two-component powder /monomer, ambient-

cured acrylic resin, TempriteTM (Great Lakes Orthodon- tics Ltd., Tonawanda, NY). This resin is supplied for the preparation of temporary crown and bridge dental restorations. The powder component of Tempritem contains PMMA with inert pigments, is plasticized with diethylphthalate (determined by IR analysis), and is coated with benzoyl peroxide for free-radical initia- tion. Its weight-average molecular weight is reported to be approximately 450,000 (personal communication from the manufacturer). The monomer component is MMA together with inhibitors and an aromatic amine accelerator to promote ambient-curing. The formula- tion is adjusted to provide a hardening time of about 15 min after mixing.

Specimen preparation

In a series of pilot experiments it was determined that a 1.2 : 1 weight ratio of PMMA powder/monomer provides an optimum balance among degree-of-mono- mer conversion, working time, and time required for initial hardening (about 15 min). Thus, specimens were mixed in a 1.2: 1 ratio with variable amounts of TPB. TPB readily dissolves in pure MMA monomer to as much as 70% by eight.^ In the commercial monomer formulation, TPB dissolved completely at 5%. At and above 15%, however, a slightly cloudy suspension was formed. At the 30% level a few undissolved particles were clearly visible. The TempriteTM PMMA powder was added to the TPB/monomer mixtures and stirred until evenly distributed. After the mixture had reached a syruplike consistency, it was poured into a 1 X 3 X 20-mm brass mold resting on a glass slab. A glass plate was placed over the mold and the entire assembly was compressed with a 2.5-pound weight. It was neces- sary to maintain constant pressure on the top glass plate once it was in place. Otherwise, release of tension caused air bubbles to be drawn into the mold cavity, and air pockets to form in the specimen. After checking for unwanted air bubbles, the specimen was allowed to harden for 45 min. Specimens with obvious porosity were discarded.

After the 45-min initial curing time, specimens were either: (1) allowed to continue curing at room tempera- ture (23°C) for 24 h, (2) heated for 5 min at 40°C and then stored at room temperature for 23 h, or (3) heated for 5 min at 50°C and then stored at room temperature for 23 h. Since it required about 15 min to measure a specimen’s dimensions, place it in the instrument, and begin the analysis, a total of either 1 h or 24 h passed between mixing and beginning a DMA scan. Postcure heating was done to simulate the common practice of heating ambient-cured dental acrylics in hot tap water in order to accelerate the final cure.

The faces of the specimens were then flattened with 600-grit emery paper, leaving dimensions approxi-

THERMOMECHANICAL PROPERTIES OF PMMA 341

mately 20 X 3 X 1 mm. If the analysis could not be performed immediately after the 24-h time periods, the specimens were stored in liquid nitrogen until the analysis was carried out. Separate experiments demon- strated that liquid N2 storage effectively stopped any further time-related changes (e.g., continued conver- sion) without otherwise altering the specimens.

quite good (R > 0.92), except for the T, data at 23°C (R = 0.67) and 40°C (R = 0.82). In separate experiments we found that crystals of TPB could be observed in the cured specimens, beginning at 8%. As TPB was increased to as high as 40%, both the number and size of these crystals increased. The presence of TPB crystals had little to no effect on appearance.

Analysis procedures

Forty-eight specimens were subjected to a 3-point bending, DMA temperature scan using a DMA-7 ther- mal analyzer (Perkin Elmer, Norwalk, CT). Each speci- men was mounted on a 15-mm knife-edge platform and chilled to -10°C in a furnace enclosed in a cooling jacket filled with a mixture of isopropyl alcohol and solid carbon dioxide. Static and dynamic forces impressed on the samples were 90 mN and 70 mN, respectively. A helium atmosphere was used to purge the sample chamber for the duration of each run. The dynamic deformation force was impressed at a fre- quency of 1 Hz. Each sample was heated at 10"C/min until its deflection point was reached (90-125°C). At this point it took on a leathery consistency, would no longer support the applied stress, and slumped to the bottom of the test fixture. The changes in storage mod- ulus, loss modulus, and tangent delta were automati- cally recorded as a function of temperature by the DMA instrument and used to determine EfZsC (storage modulus in the operating temperature range), the tem- perature at the beginning of the T, range (T, onset) and the deflection point. The temperature at the begin- ning of a two-decade drop in E', which coincides with a peak in the loss modulus, E , was used to define the T, onset. The onset of T, was reported, rather than the high-temperature tan 6 peak (which is the usual measure of T,) because the deflection point occurs be- fore the tan 6 peak is reached. Above the deflection point the permanent deformation of the specimen pre- vents any further measurements in the 3-point bend- ing mode.

RESULTS

The results are shown in Table I and Figures 1-4. Figures 1 and 2 show that TPB depresses both the glass-transition temperature and the softening point. This can also be seen in Figures 3 and 4, which show typical DMA scans. The effects of TPB and postcure heating on the modulus at 25°C can be seen in Table I. The data in Figures 1 and 2 are fitted with second- degree polynomial curves as a means of visualizing the overall trends. Although there is no theoretical basis for a polynomial function, the fit is generally

DISCUSSION AND CONCLUSIONS

The results shown in Table I and Figures 1-4 indi- cate that, as expected, TPB decreases both the glass- transition onset and the softening point. These tem- peratures are inverse functions of TPB concentration. Postcure heating at 50°C offsets these reductions some- what, illustrating the plasticizing effect of unconverted monomer. In this system MMA is the only monomer. Its plasticizing effect could be removed either through evaporation (boiling point = 1OO.YC) or by being forced to react at elevated temperatuers. Probably both mechanisms are involved. Heating for 5 min at 40°C produced only a slight and insignificant elevation of T, and deflection temperatures for any level of TPB. However, significant increases were produced by heat- ing to 50°C with 5% and 15% TPB. The results shown in Figures 1 and 2 indicate that, above about 15%, the effects of added TPB and postcure heating at 50°C are greatly reduced, but that the trends of decreasing T, and softening point continue.

It was also observed that crystals appear in the cured specimens above about 8%, even though up to about 70% TPB is soluble in solid solution with PMMA. The 70% w/w solubility of TPB in PMMA reported by Delaviz et aL8 was determined in pure PMMA that had been polymerized under near-equilibrium conditions. The precipitation of TPB that we observed well below saturation is probably due to a combination of rapid solidification of specimens and to the presence of pig- ments and/or other minor components in the commer- cial formulation. The ErZsC results shown in Table I appear to rise and fall in response to TPB concentration. For any postcure temperature EfZsC is higher for all TPB resins compared with PMMA alone. However, only ErZsC for 15% TPB and 50°C postcure heating is signifi- cantly greater than Ef25C at 0% and 30% TPB. Thus, compared with PMMA alone, there is no decrease in storage modulus in the vicinity of 25°C due to addition

The TPB crystals embedded in cured resin were ob- served to increase in number and size as the TPB con- centration in the monomer was increased. These pre- cipitated TPB crystals, dispersed throughout the PMMA, could act as reinforcing fillers. This would account for the increases in E' observed for TPB con- centrations > 8%. However, even at 5%, a concentra-

of UP to 30% TPB.

342

TABLE I Effect of TPB and Postcure Heating on Thermomechanical Properties of PMMA

RAWLS ET AL.

Postcure E' at 25°C T, Onset Deflection Point TPB (Yo) Temperature ("C) No. ( S E , GPa)* (?SE, O C ) * TE, = 0 (LSE, 'C)*

0 23 3 2.9 (50.1) a 79 (2 6) a,b 113 (56) a,b 40 2 2.8 (20.0) a,b 78 (59) a,b 114 (22) a,b 50 3 2.6 (20.1) b 97 (tl) a 125 ( 2 0 ) a

5 23 6 3.0 (50.1) a 75 (23) b 104 (54) b 40 6 3.1 (20.2) a 82 (51) a,b 110 (51) b 50 6 2.7 (20.1) a,b 86 (55) a 126 (22) a

15 23 4 2.8 (20.1) a,b 66 (22) c 92 (54) c 40 4 2.9 (20.1) a,b 68 ( 5 1) b,c 95 (51) c

30 23 5 3.4 (50.1) a 76 (21) b 97 (21) c 50 3 3.2 (50.1) a 73 (22) b 100 (20) b,c

50 9 2.9 (20.1) b 65 (1-3) c 85 (23) d

*Values with the same letter are not significantly different at p < 0.05 according to analysis of variance followed by either Fisher's partial least significant difference test (0-15% TPB) or Student's unpaired t test (30% TPB). Statistics were calculated using StatViewTM, version 11, software (Abacus Concepts, Inc., Berkeley, CA).

"M 50 0 10 20 30

Triphenyl Bismuth (YO, w/w)

Figure 1. Effect of TPB on T, of PMMA postcure heated 5 min at 23"C, 40T, and 50°C. (0, 50°C; ., 40°C; A, 23°C. Error bars are f standard error. R = 0.67 for 23"C, 0.82 for 40"C, and 0.99 for 50°C.)

n I- - 1 90-

i 80 I I 1 I

0 10 20 30

Triphenyl Bismuth (%, w/w)

Figure 2. Effect of TPB on softening point of PMMA post- cure heated 5 min at 23T, 40°C, and 50°C. (O,5O0C; ., 40°C; A, 23°C. Error bars are 1- standard error. R = 0.93 for 23T, 0.98 for 40°C, and 0.97 for 50°C.)

tion at which all the TPB remains in solid solution, no decrease in E' was observed in the vicinity of 25°C. As shown in Figures 3 and 4, from -10°C to about 5OoC, the value of E' is not significantly affected by the level of added TPB. Above 5OoC, E' decreases with increasing TPB concentration.

These results imply an antiplasticizing effect for dis- solved TPB at temperatures below 50"C, in combina- tion with a dispersed phase-reinforcing effect from TPB crystals. An antiplasticizing effect could come about by TPB occupying free volume among the polymer chains. When the temperature is raised above the melt- ing point of TPB, 78"C, all of the crystals probably dissolve and the reinforcing effect is lost. This may account for the observation (Table I) that the deflection temperature (softening point) never falls below about 85°C for any postcure heating temperature or any level of added TPB.

Additional experiments are required in order to con- firm these possibilities. However, it can be concluded that TPB in amounts adequate to impart diagnostic levels of radiopacity, does not significantly affect the modulus within the performance temperature range

10

G ' n. 5 iu 0 1

0 01 0 25 50 75 100 125

Temperature ( " C )

Figure 3. (E') and tan 6 of PMMA postcure heated 5 min at 50°C.

Effect of TPB concentration on elastic modulus

THERMOMECHANICAL PROPERTIES OF PMMA 343

m * - d * a 4 W

C

Temperature ("C)

Figure 4. Effect of the temperature of postcure heating on elastic modulus (E') and tan 6 of PMMA with 15% TPB.

for oral uses of PMMA. This implies that practical performance properties for dental applications are not likely to be adversely affected and that clinically useful radiopaque acrylic resins can therefore be prepared using TPB. Similarly, beneficial effects are anticipated for the use of TPB as a radiopacifying additive with acrylic bone cements, catheter tubing, and other bio- medical polymer material^.^

This work was supported by NIH Grant R01-DE06179.

References

1. J. P. Newton, R. W. Abel, C. H. Lloyd, and R. Yemm, "The use of computed tomography in the detection of radiolucent denture base material in the chest," 1. Oral Rehab., 14, 193-202 (1987).

2. Y. Delaviz, Z. X. Zhang, I. Cabasso, and J. Smid, "Homo- geneous radiopaque polymers with organobismuth com- pounds," J. Appl. Polymer Sci., 40, 835-843 (1990). (a) H. R. Rawls, M. V. Marshall, L. Cardenas, H. R. Bha- gat, and I. Cabasso, "Cytotoxicity evaluation of a new radiopaque resin additive," Dent. Mater., 8, 54-59 (1992). (b) P. A. Mattie, H. R. Rawls, and I. Cabasso, "Develop- ment of a radiopaque, auto-polymerizing dental acrylic resin," J. Prosthodonitcs, 3, 213-218 (1994).

4. (a) Y. Delaviz, Z. X. Zhang, I. Cabasso, and J. Smid, "Ho- mogeneous x-ray contrast polymerorganobismuth com- posites," Polym. Prepr., Am. Chem. SOC. Div. Polym. Chem., 30,215-216 (1989). (b) Y. Delaviz, Z. X. Zhang, I. Cabasso, and J. Smid, "Homogeneous radiopaque polymers with organobismuth compounds,"J. Appl. Polym. Sci., 40,835- 843 (1990). (a) P. A. Mattie, H. R. Rawls, and I. Cabasso, "Develop- ment of a radiopaque, auto-polymerizing dental acrylic resin,"J. Prosthodontics, 3,213-218 (1994). (b) P. A. Mattie, H. R. Rawls, and I. Cabasso, "Development of a radio- paque, heat-cured dental acrylic resin," J. Prosthodontics, to appear.

6. T. W. Wilson and D. T. Turner, "Characterization of polydimethacrylates and their composites by dynamic mechanical analysis," J. Dent. Res., 66,1032-1035 (1987).

7. Y. Delaviz, Z. X. Zhang, I. Cabasso, and J. Smid, "Homo- geneous radiopaque polymers with organobismuth com- pounds," J. Appl. Polym. Sci., 40, 835-843 (1990). (a) Y. Delaviz, Z. X. Zhang, I. Cabasso, and J. Smid, "Ho- mogeneous x-ray contrast polymerorganobismuth com- posites," Polym. Prepr. Am. Chem. Sac. Div. Polym. Chem., 30,215-216 (1989). (b) Y. Delaviz, Z. X. Zhang, I. Cabasso, and J. Smid, "Homogeneous radiopaque polymers with organobismuth compounds," J. Appl. Polym. Sci., 40,835- 843 (1990).

9. I. Cabasso and J. Smid, "Bismuth compounds as compati- ble x-ray contrast polymer additives," Bull. Bismuth Inst.,

3.

5.

8.

60, 1-6 (1990).

Received May 11, 1995 Accepted October 4, 1995