105
Water sorption and mechanical properties of light-cured proprietary composite tooth restorative materials S. Kalachandra and T.W. Wilson** Department of Operative Dentistry, Dental Research Center, University of North Carolina, Chapel Hill, NC 27599-7450, USA Seven light-cured proprietary composite restorative materials, P-50 (3M), P-10 (3M), P-30 (3M), FulFil (Caulk), Herculite (Kerr), Silux Plus (3M) and Silux (3M) were characterized in terms of water uptake at 37°C. For several of the systems, elastic modulus and glass transition temperature were evaluated with a dynamic mechanical analyser (Autovibron DDV-II-C). Model systems such as bis-GMA (75%) + triethylene glycol dimethacrylate (25%) bis-GMA (30%) + triethylene glycol dimethacrylate (10%) + lithium aluminium silicate (60%) bisGMA + triethylene glycol dimethacrylate (14%) + barium glass (silanated) (66%) and bisGMA + triethylene glycol dimethacrylate (14%) + zinc glass (silanated) (86%) were also studied with reference to water sorption. It was concluded that the changes in elastic modulus tend to confirm the hypothesis that the matrix-filler interface contains water. Silux and Silux Plus accommodated the greatest amount of water at the interface. Silux Plus displayed a dramatic reduction in elastic modulus c. O”C, indicating possible melting of water clusters. It appeared that the only available hole for water clusters was at the matrix-filler interface. Keywords:
Dental materials,
water sorption,
mechanical
properties
Received 4 September 1990; revised 18 October 1990; accepted 22 January 1991
Extensive studies have been made on the serious problems of the deterioration of glassy polymers in aqueous environmentsl-lo. Recent work on the mechanical properties of dental composites focused on the implications during service performance in the mouth. These composite restorations are expected to function in a complex environment characterized by aqueous fluids, alternate temperature conditions and cyclic mechanical forces. The modulus of elasticity (E’) in bending tests has been found to be appreciably lower than that of amalgams. The sensitivity in molar teeth restored with large mesio-occlusal-distal composites may be due to excessive bending under loading”. Oral fluid uptake aggravates the problem by reducing the modulus. In other recent work, studies were made of composites characterized with respect to residual double bonds. It was found that uptake of water had a marked effect on mechanical properties, reducing both yield stress and fracture toughness by as much as 30%. It was pointed out that the evidence for embrittlement due to water might be pertinent to clinical observations of the fracture of restorations”. A recent study on the mechanical properties of Plexiglass showed that the absorption of 1% water drastically reduced the fatigue life of bulk Correspondence to Dr S. Kalachandra. **Present address: Family Health Intl., Research Triangle Park, NC 27709, USA. 0 1992 Butterworth-Heinemann 0142-9612/92/020105-05
Ltd
polymerized poly(methy1 methacrylate) (PMMA) (Plexiglass)13. This paper extends previous workI on the influence of fillers on the water sorption of composites. We studied the water sorption and mechanical properties of lightcured proprietary tooth restorative materials and model systems to design composites which will take up a predictable amount of water. The role of the coupling agent in reducing the interfacial water penetration is relevant. MATERIALS
AND METHODS
Materials Seven proprietary tooth restorative materials were used as supplied: P-10 (3M), P-30 (3M), P-50 (3M), Silux (3M), Silux Plus (3M), Herculite (Kerr) and Fulfil (Caulk]. Also the following model systems were employed in the experiments: bis-GMA (75%) + triethylene glycol dimethacrylate (TEGDM) (25%), bis-GMA (30%) + TEGDM (10%) + lithium aluminium silicate (LiA1Si4010) (60%), bis-GMA + TEGDM (14%) + barium glass (SSW), bis-GMA + TEGDM (14%) + zinc glass (86%). Methods To effect photopolymerization, (6.19%) and dimethylaminoethyl
camphoroquinone methacrylate (0.12%)
Biomaterials
1992, Vol. 13 No. 2
106
Properties
of composite
were dissolved in a mixture of bis-GMA (75%) and TEGDM (25%). The monomer solution was contained between glass coverslips (1mm), bounded by a window frame of polytetrafluorethylene (3 X 3 cm; recess 0.1 cm thick). It was polymerized at ambient temperature under nitrogen, by exposure (5 min exposure on each side) to visible light from a dental lamp with a 0.5 cm window (Command, Kerr Sybron, MI, USA). The product had the same dimensions as the recess of the mould and was used directly for studies of water sorption*5p1”. The same method was followed in the preparation of specimens for the model systems. In these model systems, LiAlSi,O,, (60%) (Pfaltz 8r Bauer, Inc., Waterbury, CT, USA), barium glass (Silane coated) (86%) (ICI Chemical Co., USA) and silanated zinc glass (86%) (Corning) were used maintaining bis-GMA and TEGDM in the same proportions of 3:l throughout. The gross particle size of all the inorganic fillers used in model systems was determined by reflected light microscope (Olympus, Japan, model BHM). The size ranged from 10 to 30pm. Because of polymerization shrinkage, some porosity was unavoidable, but from experiments on unfilled samples, it was found that such pores did not accommodate water and had little influence on diffusion coefficients (D). Polymer contents were determined by incineration of the composites to constant weight with a Bunsen burner. Dynamic mechanical analysis (DMA) was performed with an Autovibron DDV-II-C at a strain frequency of 11 Hz, which is the standard frequency for most evaluations with an Autovibron17. Samples were scanned to +200”C. Saturated at 2.0-2.5’C/min from -120 samples for DMA were first scanned in a dry condition from -120 to +35’C; then the specimen was saturated with distilled water at 37°C; and lastly, the saturated specimen was scanned from -120 to +200“C. This allowed direct comparison of the changes occurring in a particular specimen. Samples for DMA were prepared by incrementally photopolymerizing strips with dimensions of 7.0 X 0.4 X 0.07 cm. Changes in length as a function of temperature were determined simultaneously with the dynamic mechanical data. As the sample expands, the tensioning mechanism of the Autovibron adjusts the length between the clamps. The change in length is stored in the data file with the other parameters. After the test run, plots of percentage change in length as a function of temperature are generated.
RESULTS Samples immersed in water increased in weight and reached values which are apparently constant over several days. Such stationary values are used in the present work but it should be noted that on prolonged immersion a slow drop in weight was detected, perhaps due to leaching out of soluble material. This effect may also be a minor factor where estimates of water uptake by desorption are generally greater than the values estimated by sorption [Figure 1 and Table 1),which is consistent with our previous observation in sorption studies’4v Is*1’slg. The rate of desorption and sorption was analysed with reference to the conventional solutions of Fick’s Laws of diffusion for plane sheet geometry”, Biomaterials
1992, Vol. 13 No. 2
tooth restorative
materials:
S. Kalachandra
O.11 0
200
400
600
800
and T. W. Wilson
loo0
1200
1400
Figure 1 Kinetics of water sorption of the model system at 37X, i.e. zinc glass (s) (86%) + (his-GMA + TEGDM) (14%): (0) sorption, (0) desorption.
where n/r, and M, are the water sorbed or desorbed at times t and co, respectively, and 2 1 is the thickness of the specimen, Data in both sorption and desorption conformed experimentally to Equation 2 (Figure 2). Usually two or three samples were tested, which gave satisfactory results.
M A=2 Mm
l/2
s2 (
1
(21
D = $ (slope)’ As a result of polymerization shrinkage, some porosity was unavoidable, but from experiments on unfilled PMMA, it was found that such pores did not accommodate water and had little influence on diffusion coefficients [D). Samples of dry fillers used in model systems were soaked in double-distilled water to saturation, The samples were dried at 80°C for 4 h and then incinerated to constant weight. It was seen that there was no significant change in the final weights of the samples. This indicates that there is no considerable absorption of water by the filler particles themselves. Several of the proprietary systems, both dry and water saturated, were evaluated by DMA. Typical plots of E’ and tan 6 for Silux Plus are shown in Figure Za. The percentage change in length for the above specimen is shown in Figure Zb. The change in slope on the lengthtemperature plot is taken as the glass transition temperature, Tg2*,“, The length method is in contrast to other workz3 in which the peak in tan delta is used. The length method is a one-dimensional analogue of the classical definition of TBproposed by Kauzmann’l, i.e. a change in slope with reference to volume-temperature plot. The peak in tan delta has been shown to give values of TB up to 80°C greater than those obtained from the length methodzz9 24. The length method has the additional advantage that the estimation of T8 is not frequencydependent like tan delta. Furthermore, the relatively slow heating rate (2.0-2.5”C/min), used by Wilson and
Properties of composite tooth restorative materials: S. Kalachandra Table 1
107
and T.W. Wilson
Water sorption data for copolymers in model and proprietary materials
Svste m
bis-GMA (75%) f TEGDM (25%) bis-GMA (30%) + TEGDM (10%) + LiAISidO10(60%) (bis-GMA + TEGDM) 14% -t barium glass silanated (86%) (bis-GMA + TEGDM) 14% + zinc glass silanated (86%) P-50 (3M) P-10 (3M) P-30 (3M) FulFil (Caulk) Herculite (Kerr) Silux Plus (3M) Silux (3M)
% (w)
% Water uptake
copoiymer by incineration
based on copolymer
100
Diffusion coefficient D x 10Bcm2 s-l (at 37°C)
Ratio D, /D,
Sorption D,
Desorption D,
3.85
0.70
0.81
1.15
40
1.54
2.73
1.74
3.30
1.90
14
0.54
0.75
2.27
5.30
2.34
14
0.54
0.61
1.06
2.17
2.05
16 14 18 24 25 43 47
0.62 0.54 0.69 0.92 0.96 1.66 1.81
0.80 0.84 0.84 1.18 1.24 2.22 2.31
1.80 3.16 1.80 1.92 1.65 2.18 1.68
3.30 9.15 3.60 3.57 2.83 5.44 3.74
1.83 2.89 2.00 1.86 1.72 2.50 2.23
suppresses any tendency to shift the data to Turner” higher temperatures. After water sorption, E’ and Tg typically decrease (Figure 3 and Table 2). The initial Tg values were a few degrees < 0°C. The changes in Te were minimal [Table 21. The percentage change in length as a function of temperature for Silux Plus-dry in Figure 2b reveals a TBat -30°C. There is another unusual transition at 120°C evidenced by a contraction in length. The latter transition is associated with additional cure in undercured, crosslinked systems24. The additional cure of double bonds results in a volume contraction. Table 1 summarizes the data on percentage water uptake and diffusion coefficients for all the proprietary composite materials, together with model systems.
DISCUSSION both in sorption and desorption conformed experimentally to Equation 2 [Figure I). A higher rate of
Data
% Water uotake (desorption at 37°C)
desorption is similar to that reported previously in studies of PMMA alone at room temperature and interpreted as due to a dependence of the diffusion coefficient on water content”, perhaps due to clustering. Furthermore, it is believed that there is a small additional uptake of water superimposed on Fickian diffusion, due to a retarded swelling component”’ lg. It is seen from data in Table 1 that filled specimens, i.e. the model systems and the proprietary materials, adsorbed more water than expected on the basis of copolymer content (TabZe 1, columns 3 and 4). This is consistent with the observation made previously that composites take up more water than estimated on the copolymer content4-6. The most probable site for accommodation of additional water is the interface between the inorganic filler particle and the polymer matrix14. In agreement with this, the values of D for water in filled specimens were significantly greater than in unfilled specimens (Table 1, column 6). Thus the filler-matrix interface provides paths of facile diffusion, a sort of grain boundary diffusion14* 26. The enhanced values of D observed (Table 2.0 -l
10.0
_-I-@ i tand
-L
b
’
TEMPERATURE(%)
Figure 2
a, Elastic modulus (E ‘) and loss tangent (tan S) as a function of temperature for Silux Plus-dry. b, Percentage change in length as a function of temperature for Silux Plus-dry. Biomaterials
1992, Vol. 13 No. 2
108
-...-_____
Properties
of composite
-~
I, column 6) in the model systems and the proprietary materials may be attributed to the paths of facile diffusion provided at the filler-matrix interface for the transport of water. The variations in the additional water uptake at the interface perhaps depend on the nature of the filler particles, the use of a coupling agent and the method of polymerization. From the data (Table 1, column 4) it was observed that the uptake of water decreased from 3.85 to 0.61% suggesting formation of the most effective filler-matrix bond by the coupling agent. Fu~he~o~, the decrease in water uptake by the model system containing nonsilanated LiAlSi,O,, particles was not significant (Ttfble 2, column 4; 3.85 and 2.73%) when compared with the other systems in which filler particles were pretreated with a silane coupling agent (Table 2, column 4; 3.85 and 0.75%). The water uptake was significantly reduced at the interface in the systems where matrix and filler particles were effectively coupled. This is consistent with our previous observations in which 4-META considerably reduced water uptake from 17 to 5% at the interface of tribasic calcium phosphate composites14. From the data (Table 2, column 4) it was seen that the model system (bis-GMA - TEGDM) 14% + zinc glass (silanated) (86%) exhibits the least water uptake (O.SlW), suggesting that matrix and the filler particles were most effectively coupled. Variations in the water uptake observed among the seven proprietary materials may he attributed to the nature of the filler particle and the coupling agents, i.e. differences in filler-matrix bond strengths. DMA of the proprietary materials revealed an unexpected result, i.e. the decrease in Ts was minimal. However, if the matrix nominally absorbs 3.85% water, then Tg should only decrease about 10°C which is consistent with the values in Table 2”. The elastic moduli showed minimal (P-10) to drastic (Silux Plus) changes after water sorption. Qualitatively, the systems which absorbed the most water displayed the greatest decline in modulus. Furthermore, these systems have the highest percentage of organic components. A further observation is that the system exhibiting the smallest decline in modulus (P-10) was the only autocure system: all the others were visible light-cured. The changes in E’ tend to confirm the hypothesis that the matrix-filler interface contains water. The excess water that the system accommodates (Table 1, column 4 minus column 3) is greatest for Silux and Silux Plus, and these two systems have the largest decrease in modulus. Additionally, Silux Plus ~splayed a dramatic change in E’ at around 0°C (F&rue 31, indicating possible melting of water clusters. The only available hole for water clusters is at the matrix-filler interface. Table 2
Dynamic mechanical
properties
of proprietary
tooth restorative
S. K~/~~~~~~~~ and T. W. Wi/so~
materials:
11.0,
r
0
-E 10.8 -
$
10.6-
- -1 D
b: rx 10.4 R
E
w 5
s” 10.2 -
--2
10.0 -
9.8 s :: 7
7
$
o
8
8
3
Q
I -3 g
TEMPERATURE (%)
Figure3 Elastic modulus and loss tangent as a function of temperature for Silux Plus-water saturated.
CONCLUSIONS
11) The filled systems, proprietary
composites and model systems, accommodated additional water at the interface between inorganic filler particles and the matrix. (21 Changes in modulus (E’) confirm the hypothesis that the filler-matrix interface contains water. (3) Variations in the uptake of additional water in the spaces at the interface between filler particles and matrix, seem to depend on the nature of the filler particle, the use of coupling agents and the method of polymerization. (41 The uptake of water by composite filling materials is a diffusion-controlled process. The diffusion coefficients (D) seem to exhibit concentration dependence, i.e. D decreases with increasing water concentration. The values of D for water in the proprietary composite materials and the model systems were considerably larger than in unfilled specimens. This enhanced rate of diffusion may be attributed to a facile diffusion, i.e. a sort of grain boundary diffusion. (5) Uptake of water in these materials decreased in a manner of an effective coupling between filler particles and matrix. This finding is consistent with our previous observation in which water uptake by TEGDM- and PMMA-based composites was reduced significantly in the presence of Silane coupling agents. The uptake of water by composite filling materials seems to be a diffusion-controlled process. It has been shown that the kinetics of water sorption of composites conform
materials (-120
to +200°C).
E’ dry
E’ wet
X lO-‘O dyne/cm2
X tO-lo dyne/cm2
%
AE change
ATczl Change in T,“C
P-10
5.83
5.69
-2
-2
P-30 P-50 Silux Silux Plus Herculite
4.49 5.90 5.37 3.95 4.94
4.13 5.42 3.48 1.36 4.39
1:: -35 -66 -11
1; -13
System
Biomaterials
1992, Vol. 13
No. 2
Properties
of composite
tooth
restorative
materials:
S. Kalachandra
to Fick’s Law, permitting quantitative characterization by a single number, i.e. a diffusion coefficient (D). This permits exact comparisons and suggests how silane coupling agents reduce the interfacial penetration of water between filler particles and poiymeric matrix, This is important with regard to water resistance and as a molecular probe to deduce information about the interface, which is critical for the properties of composites.
12
13
15
16
18
Braden, M. and Davy, K.W.M., Water absorption characteristics of some unfilled resins, Biomaterials Braden, M., The absorption of water by acrylic resins and other materials, 1. Prosthet. Dent. 1964, 14, 301-316 Stafford, G.D. and Braden, M., Water absorption of some denture base polymers, J. Dent. Res. 1968, 47, 341 Braden, M., Causton, B.E. and Clarke, R.L., Diffusion of water in composite filling materials, f. Dent, Res. 1976,
19
20
21
65, 730-732 5
6
7
Braden, M. and Clarke, R.L., Water absorption characteristics of dental microfine composite filling materials, I. Proprietary materials, Biomaterials 1984, 5, 369-372 Braden, M., Water absorption characteristics of dental microfine composite filling materials, II. Experimental materials, Biomaterials 1984, 5, 373-375 Braden, M., The formation of composite filling materials, Operative
8 9 10 11
22
23 24
Dent. 1978, 3, 97-102
Braden, M., Rheology of composite filling material pastes, J. Dent. Res. 1977, 56, 627-630 Rowland, P.S., Water in polymers (Ed, S.P. Rowland& ACS Symposium Series 1980, Lz7 Vinson, J.F. (Ed.), Advanced composite materials Environmental effects, ASTM STP, 1978, p 658 Bryant, R.W. and Mahler, D.B., Modulus of elasticity in bending of Composite Resin and Amalgam, IADR
109
Ferracane, J.L., ~tonio, RX. and Matsumoto, H., Variables affecting the fracture toughness of dental composites, 1, Dent. Res. 1987, 66,1140-1145 Shen, J., Chen, C.C. and Saver, J.A., Fatigue of PMMA effects of molecular weight, water content and frequency, Fatigue in Polymers Conference Proceedings The Plastics and Rubber Institute, London, UK, 1983 Kalachandra, S., Influence of fillers on the water sorption of composites, Dent. Mater. 1989, 6, 283-288 Kalachandra, S. and Turner, D.T., Water sorption of polymethacrylate networks: bis-GMAITEGDM copolymers, J. Biomed. Mater. Res. 1987, 21, 329-338 Nagata, K. and Turner, D.T., Influence of $-methac~loxy-ethyl trimeliitic anhydride on composites subjected to hydrothermal cycling, I; ~iorn~. Mater. Res. 1985, 19,831-642
17
REFERENCES
1986, 7, 474-475
_.._..-...---___
Abstract No. 311,J. Dent. Res. (Special Issue D), 1985,64,
14
1
T. W. Wilson
D311
ACKNOWLEDGEMENTS This investigation was supported by United States Public Health Service (USPHS] Research Grants Nos DE06201 and DE05487 and General Research Support No. 05333 and DE05487. Our grateful thanks are due to the Caulk, Corning, Kerr, SM and ICI Companies for their generous supply of the materials used in this investigation.
and
25
26
Murayama,
T., Dynamic Mechanical Analysis of Polymeric Materials Elsevier Scientific, New York, USA, 1978 Kalachandra, S. and Turner, D.T., Water sorption of poly(methy1 methacrylate): 3-effects of plasticizers, Polymer 1987, 26,1749-1752 Kalachandra, S. and Turner, D.T., Water sorption of plasticized acrylic lining material, Dent. Mater. 1989, 5, 161-164 Crank, J., Diffusion in a plane sheet, in The Mathematics of Diffusion 2nd Edn, Ch. IV, Clarendon Press, Oxford, UK, 1975 pp. 44-68 Kauzmann, W., The nature of the glassy state and the behavior of liquids at low temperature, Chem. Rev. 1948, 3, 219-256 Wilson, T.W. and Turner, D.T., Characterization of polydimethacrylates and their composites by dynamic mechanical analysis, J. Dent. Res. 1987, 66, 1032-1035 Clarke, R.L., Dynamic mechanical thermal analysis of dental polymers, Biomaterials 1989, 10,630-833 Wilson, T.W., Radiation effects on the dynamic mechanical properties of epoxy resins and graphic fiber/ epoxy composites, PhD Thesis, North Carolina State University Raleigh, North Carolina, USA, 1986 Barrie, J.A., Water in polymers, in Diffusion in Polymers
(Eds J. Crank and G.S. Park], Ch. 8, Academic Press, London, UK, 1968, pp. 259-313 Pace, R.J. and Datyner, A., Statistical mechanical model for diffusion of simple penetrants in polymers, J. Poly. Sci. Part B: Polym. Phys. 1979,17,437-45X
Biomaterials 1992, Vol. 13 No. 2