Reactions in Glass-Ionomer Cements: IV. Effect of Chelating Comonomers on Setting Behavior A. D. WILSON, S. CRISP, and A. J. FERNER Department of Industry, Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, SEI 9NQ, England The effect of the addition of chelating comonomers, to the polyacid liquids, on the setting characteristics of the glass-ionomer cement system is described. Certain acid chelating agents significantly improve the setting behavior of the system. The prototype glass-ionomer cement invented by Wilson and Kentl-3 and Kent, Lewis, and Wilson,4 ASPA I, had the disadvantages of a limited working time and a comparatively slow rate of surface hardening. Although usable as a dental cement, its "progressive" setting characteristic could have proved to be a handicap in general dentistry. This present article reports on work leading to an improved formulation, which has been the subject of a two-year clinical trial reported by McLean and Wilson.5 The reaction between polyacrylic acid (PAA) and an ion-leachable glass has been elucidated by Crisp and Wilson6,7 and Crisp et al.8 It is shown simply in Figure 1, a, where side reactions are eliminated. The initial stage is the ionization of PAA which leads to a change of configuration in the polymer from a coiled to a linear form. The protons produced by ionization attack the basic glass to release metal ions. The majority of these cations are divalent or trivalent and are bound by the ionized polymer that forms salt bridges; a gel phase is precipitated and the cement sets hard. The initial ionization of the acid and unwinding of the polymer chains is probably responsible for the slowish set and surface hardening of the cement. In an earlier article, the availability of fluoride present in the glass was found to have Received for publication October 15, 1974. Accepted for publication July 28, 1975.
considerable influence on working time, an effect that was attributed to the formation of strong aluminum fluoride complexes.8 Aluminum is probably transported from the ionleachable glass to the polyelectrolyte phase as a fluoride complex and as such may not be immediately available for cross-linking polyanion chains. The effect is to prolong working time. It was inferred from this observation that the addition of chelating agents to the system would prove advantageous since they would facilitate the extraction of metal ions from the glass and combine with them in the period when the polymer was ionizing and unwinding. This chelation would have the effect of withholding metal ions from the polyanion chains in the early stages, preventing premature gelatinization and thus prolonging working time. Ultimately, the complexed metal could react with the polymer in either its ionized or unionized form (Fig 1, b, reaction iia) or the free metal ion (hydrated) could react as in Figure 1, a. The overall effect would be to produce a more concerted cementing reaction. Acidic chelating agents would have particular advantages; they would considerably aid the extraction of cations from the glass and at the same time suppress the ionization and unwinding of polyanion chains prolonging working time and at the same time ensure a build-up of metal ions in solution for cross-linking. With these considerations in mind, the effects of a number of chelating agents on the setting behavior of the cement were examined. The dentist is interested in the following three characteristics of setting behavior of a cement: the amount of time available for manipulation, setting time, and postset hardening. These factors that characterize the 489
Downloaded from jdr.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 4, 2015 For personal use only. No other uses without permission.
490
WILSON, CRISP, AND FERNER
J Dent Res May-June 1976
a) In the absence of a chelating agent [Polymer] 5- [Polyanion]X. + xH (coiled) ( inear) M oxide >Mn + > [M.Po(yanion](x-n)M + [Polyanion]x(gel salt)
b)In the presence of a chelatingagent H2L > M0(-2)++ 2H+ H L + M n'
(i)
(ii) Giii)
Wi
+ [Polymer] . [M.Polyanion](X n- +nH" +L2 (fla) L (n --2)+ + [Polyanion]xM ML(n > [ML.Polyanion] ( n) - L(2c (i b) x-(x-
ML(n-2)+
ML(n-2)+
+
[Polyanion] x
>[M. L.Poyonion) (x -n +2) -
(iic)
FIG 1.-Representation of glass-ionomer cement forming reaction: a, in absence of chelating agents; b, in presence of chelating agents.
early phase of the cement-forming reaction are conventionally measured in the laboratory by penetrometer or indentation methods. Such methods do not necessarily give a correct clinical indication, but at least serve for comparative purposes. The measurement of setting time is straightforward and is specified for dental silicate cements in the International Standards Organization Recommendation I.S.O. no. R1565. Working time is not clearly defined and it is estimated by different methods by various workers. In this laboratory, a Gillmore needle with a 28 gm load is used at 23 C and 50%O relative humidity; the end point requires individual operator judgment. Surface hardness can be measured on a superficial indentation tester.a The characterization of the setting of a cement by measuring a working and a setting time is both arbitrary and incomplete. The setting of a cement is a progressive process where there is a continuous, but not necessarily uniform, increase in the stiffness of a cement paste with time. In no sense are there definite points corresponding to a working and a setting time. What is needed, especially in the context of developing improved cement formulation, is an instrument for the continuous monitoring of setting. Such an instrument exists in the oscillating a Wallace Microindentation Tester, H. W. Wallace and Co., Ltd., Croydon, Eng.
rheometer originally developed by Bovis, Harrington, and Wilson9 for use with composites and later applied by Plant, Jones, and Wilson'0 to dental cements. This instrument gives a record of stiffness against time, a rheogram, that is qualitatively more informative than the conventional penetrometer methods. This instrument was adopted in this work as a screening test for setting characteristics. For the purposes of tabulation, a working and a setting time were taken from these rheograms, using appropriate definitions. However, the main feature of interest was the maximum rate of set, which could also be readily obtained from these rheograms. Materials and Methods The cement powder, an ion-leachable glass designated G200, was prepared by fusing a mixture of silica and alumina in a flux of calcium, aluminum and sodium fluorides, and aluminum phosphate. The melt was shock cooled, and the opal glass formed was ground to a fine powder capable of passing through a sieve with a mesh opening of 45 micrometers.4 The PAA solution was prepared by the aqueous polymerization of acrylic acid, using ammonium persulfate as initiator and propan-2-ol as the chain transfer agent. The final product had a weight average molecular weight of 23,000 and was concentrated to
Downloaded from jdr.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 4, 2015 For personal use only. No other uses without permission.
REACTIONS IN GLASS-IONOMER CEMENTS: IV
Vol 55 No. 3
is little difference in the working times obtained from the rheograms by either of these construction methods. Finally, the envelope of the trace approaches the abscissa and becomes linear and sometimes parallel to it (point Q) which is taken as the set. In practice this cannot be assigned very accurately. It corresponds to the setting time for traces at 37 C, and for traces done at 23 C, it can be regarded as the room temperature setting time. The time when the amplitude was reduced to 50% of its initial value was measured. An attempt was made to measure the maximum rate of setting that was considered to be an important parameter. The maximum slope of the curved portion of the envelope was measured ( (XZ) / (XY) on Fig 2). This was expressed as the percentage reduction of the initial amplitude per mi'nute, that is, 100 (XZ) / (AB) (XY) to give a measure of the maximum rate of setting. OTHER MEASUREMENTS.-Certain other measurements were made on selected cements. Compressive strength (24 hours), solubility and disintegration, and setting time (37 C) were determined by standard specification tests.11 Working time was measured at 23 C using a Gillmore needle but with a lower (28 gm) load than the 454 gm load in the standard instrument but with the same needle diameter of 1.05 mm. The end of working time was deemed to be that instance when the needle failed to penetrate 1 mm of cement paste. Tensile strength (24 hours) was measured by the diametral compressive loading of disks (8-mm diameter X 4-mm height). Early surface hardness was measured using
50% w/w by vacuum distillation.4 The various chelating agents were of the best quality laboratory grades and used without further purification. They were generally added to the PAA solution, in the ratio of 5:100 parts by weight, and tumbled until dissolution was complete. When the chelating agent had a limited solubility, saturated solutions or slurries were used. Cements were mixed, for one minute, on a glass block in a room controlled at 23 + I C and 50 + 5% relative humidity, and the powder-liquid ratio recorded. RHEOMETER TRAcES.-Between 1.2 and 1.6 gm of paste was transferred to the platen of the oscillating rheometer, maintained at 23 C in most experiments, and traces recorded on an ultraviolet recorder. Using the rheometer trace, a number of setting characteristics were tabulated. The instances when the amplitude had decreased to 95 (see references 9, 10) and 50% of its original value were recorded. The measurement of setting parameters is shown in Figure 2. The curve BMQP represents the envelope of the trace. The time axis lies along the abscissa whereas the ordinate gives the amplitude of the oscillations. There is an initial straight portion of the envelope parallel to the abscissa followed by a curve as the cement thickens. The point M is at the point of time at which the amplitude of the oscillation is reduced to 95% of the original value, the definition of working time given by Bovis, Harrington, and Wilson.9 An alternative method of estimating working time is to extend the initial straight portion of the curve to point N where the tangent of the second part intersects. There
L__ II
N
I\ I\ I\
II\
0
A
X
491
X
a
F
FIG 2.-Geometrical constructions on rheometric trace: point 0, start of mix; point A, completion of mix; points M and N, working time measured from 0; and Q, gel point. Downloaded from jdr.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 4, 2015 For personal use only. No other uses without permission.
492
J Dent Res May-June 1976
WILSON, CRISP, AND FERNER
cement disks 20 mm in diameter X 1.5-mm thick. The cement paste was packed into a split-ring mold and, two minutes from the start of mixing pressed between two, polythene-lined, flat 2-kg weights maintained at 37 C in an oven. Twelve minutes later, the sample was removed from the mold and placed on the anvil of a hardness machine.12b A standard Vickers diamond was applied, with a load of 1 gm, to the cement surface for a zero reading and then at precisely 15 minutes from the start of mix an additional 300 gm was applied for 15 seconds and the depth of indentation recorded in units of 10-5 inches, as the indentation- number.c
Results Rheograms of all experimental cements were recorded, and selected examples are shown in Figure 3. The setting characteristics of the whole range of experimental cements taken from their rheograms are given b Wallace Hardness Machine, H. W. Wallace and Co. Ltd., Croydon, Eng. c Wallace Indentation Number, H. W. Wallace and Co. Ltd., Croydon, Eng.
in Table 1. Physical and mechanical properties of important variants are given in Table 2. All rheograms (Fig 3) are characterized by (1) an initial period where the amplitude remains constant at about the maximum value and the cement paste remains workable; (2) an intermediate period where the rate of decrease of the amplitude reaches a maximum as the cement is setting; and (3) a final period where the cement sets and a small but constant amplitude is attained. Working times, maximum rate of set, and setting times (23 C) taken from the rheograms are given in Table I for all cements. There is little difference in working times obtained from these rheograms by either construction method. Working times measured by the modified Gillmore needle method agree with those taken from the rheograms' (95%) points (Tables 1, 2). Working time is an inexact parameter and is somewhat arbitrarily defined in laboratory tests. The maximum slope of the curved portion of the rheogram envelope was taken to repre-
t =1 2 min
FIG 3.-Rheometric traces of: (a), glass-ionomer-I cement; (b), ASPA cement plus 5% urea; (c), ASPA cement plus 5% tartaric acid; (d), 5% citric acid; (e), 1% acetylacetone; and (f), saturated solution of salicylic acid, each at P/L 4.0 gm/ml and 23 C. Downloaded from jdr.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 4, 2015 For personal use only. No other uses without permission.
REACTIONS IN GLASS-IONOMER CEMENTS: IV
Vol 55 No. 3
493
TABLE 1 SETTING CHARACrERISTICS OF VARIOUS ASPA CEMENT SYSTEMS Cement Type Addition to Liquid ...
Hydroxy acids 5% Tartaric acid 5% Tartaric acid 5% Dihydroxytartaric acid 5% Citric acid 10% Citric acid Hydroxybenzoic acids 5% Suspension of salicylic acid Saturated solution of salicylic acid 2% 2,6-dihydroxybenzoic acid di Ketones 1% Acetylacetone 2% Acetylacetone 1% Aluminum triacetylacetonate 1% Aluminum triacetylacetonate Complexing agents Saturated solution sequestric acid 5% Slurry sequestric add 5% Nitrilotriacetic acid 5% Ethanolamine Miscellaneous 5% Urea 5% Polyglycol
P/L Ratio (gm/ml-)
Gel Time
Stiffening Rate
(mirn)
95 %
50%
3.5 4.0
1.8 2.4
3.6 3.6
9 8
30 34
3.5 4.0 1.0 4.0 3.5
2.0 1.9 2.0 1.9 2.8
2.9 2.5 3.3 2.9 3.8
6 5 8 6 9
87 32 68 51
4.0
2.5
2.7
10
33
4.0
1.6
2.7
7
49
4.0
2.1
3.5
9
37
4.0 4.0
2.3 2.3
3.5 3.6
9 8
38 33
4.0
2.0
2.9
6
54
3.7
2.0
3.0
9
46
4.0 3.5 4.0 3.5
2.3 2A 2.2 3.5
3.6 3.9 3.7 8.8
8
9 19
36 31 34 9
3.5 4.0
2.2 1.9
5.6 3.2
15 8
(min)
8
48
16
TAkBLE 2
sent the maximum rate of setting. The set-
PHYSICAL PROPERT'IES OF SELECTED ASPA C]EMENTS
ting points (measured from the rheograms) at 23 C are of limited clinical significance. They are of some comparative value and give an indication of sharpness of set. The setting characteristics of any particular glasscement system are affected by the ionomer powder-liquid ratio of the mix, and in making comparisons of data in Table 1, cements of equal powder-liquid ratio should be selected. An increase in this parameter increases the setting rate and tends to reduce the setting time (23 C) . Additives to the liquid can be classified by their effect on the maximum rate of setting. They fall into three groups: those with little effect, namely dihydroxytartaric acid, acetylacetone, sequestric acid, nitrilotriacetic acid, and polyglycol; those that reduce it, namely urea and ethanolamine; and those that increase it, namely salicylic acid, aluminum
5% Tar- 5%o Citric Liquid Additive
None taric Acid
Powder-liquid ratio 3.0 3.0 (gm/ml) 3.0 Working time (min) 2.5 2 6 4 Setting time (min) Wallace indentation number (15 min) 600 120 Compressive strength 181 165 (24 hr, N/mm) Tensile strength 13.9 13.7 (24 hr, N/mm2) Solubility and disintegration (%, 24 hr) 0.47 0.47 Modified Gillmore ne!edle method.
3.0
0
Working Time (min)
Acid 4Q0 4.0 2 3.75
347
192
0.54
Downloaded from jdr.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 4, 2015 For personal use only. No other uses without permission.
494
J Dent Res May-June 1976
WILSON, CRISP, AND FERNER
triacetylacetonate, citric acid, and tartaric acid.
Discussion RHEOLOGICAL CHARACTERISTICS.-The addition of comonomers to the PAA liquid has a striking effect on setting characteristics. The variety of shapes of the envelope curves is shown in Figure 3. Simple visual inspection demonstrates considerable differences and the drawn out rheogram of the sluggish cement (curve b, Fig 3) is easily distinguishable from that of the sharply setting cement (curve c, Fig 3). The effect of the hydroxy acids is the most significant. Both tartaric acid and citric acid considerably increase the maximum setting rate and reduce the setting time (23 C) (Table 1; Fig 3) but do not affect the working time. This represents an obvious improvement in setting characteristics. Aluminum triacetylacetonate and saliyclic acid showed a similar, but much reduced, effect. Surprisingly, dihydroxytartaric acid has little effect. Urea and ethanolamine slow down the setting reaction (Table 1; Fig 3) and adversely affecting setting characteristics. Acetylacetone, sequestric acid, nitrilotriacetic acid, and dihydroxytartaric acid, which are chelating agents, and also polyglycol, have no effect. PHYSICAL PROPERTIES.-The two most promising cement systems (those with tartaric acid and citric acid additives) were subjected to further testing. Results from this study showed that both hardened much more rapidly than the control system without additives; the tartaric acid system also proved to be considerably superior in this respect to the citric acid system. THE ROLE OF CHELATING COMONOMERS.-In
a previous communication,8 the important role of the fluoride ion in the ASPA system was established and the workability of the cement was found to improve with an increase in the rate of elution of fluoride. It was inferred that the formation of a fluoride complex (aluminum forms a range of stable complexes, for example, AlF2+ (K1 = 6.83) AIF2+ (K1 = 5.02)) would temporarily prevent metal ions binding the polyanionic chains thus preventing premature gelation. Organic chelates would be expected to play a similar role and both aluminum and calcium form chelates of the type model 1:
M
(1)
r
This model represents the complexes formed with tartaric acid, citric acid, and acetylacetonate ligands. Complexes formed with trivalent metals are more stable than those formed with divalent ones. For example, the stability constant (K1) for the aluminum-tartaric acid complex is 6.35 compared with 2.17 for the calcium complex. The addition of tartaric acid comonomer to the liquid does not prolong working time, but it increases the rate of hardening. This result is to be attribuited to the dual effect of tartaric acid, which not only tends to hold metals in solution but, by virtue of its complexing action, will extract them more rapidly from the glass powder. If these two effects cancel out, working time will be unaffected. The particular effectiveness of tartaric acid in improving the hardening rate may derive from its structure. There are a pair of ligands for chelation at each end of the molecule (model 2) so that this molecule may serve to bridge pairs of metal ions, the whole grouping linking proximate polyanionic chains. 0
H
11
0
C
M
\
0
/ M
\\ 0 (2) CH--CH
/\ /
\ O
C
11
H 0 The lack of effectiveness of acetylacetone, sequestric acid, salicylic acid, and nitrilotriacetic acid which form stable chelates with aluminum and calcium calls for some explanation. Three possible reasons may be advanced: (1) with the exception of nitrilotriacetic acid, these chelates are relatively insoluble in PAA; (2) unlike tartaric acid, these ligands are only unifunctional and cannot bridge metal ions; and (3) they are
Downloaded from jdr.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 4, 2015 For personal use only. No other uses without permission.
REACTIONS IN GLASS-IONOMER CEMENTS: IV
Vol 55 No. 3
weaker acids than tartaric acid and may not extract ions readily from glasses. The failure of dihydroxytartaric acid to be as effective as tartaric acid requires some further explanation. Dihydroxytartaric acid is not a simple hydroxy acid. Debus'3 in 1904 suggested a cyclic structure (model 3): HO-C COOH 0
H20
HO -C COOH a (3) whereas Lachman'4 in 1921 observed that it acted as a hydrated a-diketone. These structures would account for the difference in behavior from that of tartaric acid. Dihydroxytartaric acid forms insoluble salts with metals, including sodium, which would nullify any effect it might have on the cementing reaction.
Conclusions The oscillating rheometer is a valuable instrument for studying the effects of additives on the setting behavior of a cement system. Using this instrument, it was found that certails chelating comonomers, the hydroxycarboxylic acids, could improve the setting characteristics of the glass-ionomer cement system when added to the PAA solution. The acid chelates probably assign the extraction of metal ions from the glass and also tend to hold them in solution, preventing premature ion binding of the polyanion chains. The effect is to increase the rate of hardening without reducing the working time, which may indeed be slightly increased. Tartaric acid, the most effective of the comonomers, can form a chelate bridge between aluminum atoms, and this metal complex probably acts as a flexible bridge structure linking polyanion chains. This mechanism offers some steric advantages over a simple salt bridge. The authors thank the Government Chemist, Dr. H. Egan, for permission to contribute this article. This work was done on behalf of the National Research Development Council. Crown Copyright Reproduced by the Controller of Her Britannic Majesty's Stationery Office.
495
References 1. WILSON, A.D., and KENT, B.E.: The Glasslonomer Cement: A New Translucent Dental Filling Material, J Appl Chem Biotechnol 21: 313, 1971. 2. WILSON, A.D., and KENT, B.E.: A New Translucent Cement for Dentistry: The Glass lonomer Cement, Br Dent J 132: 133135, 1972. 3. WILSON, A.D., and KENT, B.E.: (Surgical Cemcnt) Br Patent No. 1,316,129, 1973. 4. KENT, B.E.; LEWis, B.G.; and WILSON, A.D.: The Properties of a Glass lonomer Cement, Br Dent J 135: 322-326, 1973. 5. McLEAN, J., and WILSON, A.D.: Fissure Sealing and Filling with an Adhesive Glasslonomer Cement, Br Dent J 136: 269-276, 1974. 6. CRISP, S., and WILSON, A.D.: Reactions in Glass lonomer Cements: I. Decomposition of the Powder, J Dent Res 53: 1408-1413, 1974. 7. CRISP, S., and WILSON, A.D.: Reactions in Glass lonomer Cements: III. The Precipitation Reaction, J Dent Res 53: 1420-1424, 1974. 8. CRISP, S.; PRINGUER, M.A.; WARDLEWORTH, D.; and WILSON, A.D.: Reactions in Glass Ionomer Cements: II. An Infra-Red Spectroscopic Study, J Dent Res 53: 1414-1419, 1974. 9. BoVIs, S.C.; HARRINGTON, E.; and WILSON, H.J.: Setting Characteristics of Composite Filling Materials, Br Dent J 131: 352-356, 1971. 10. PLANT, C.G.; JONES, I.H.; and WILSON, H.J.: Setting Characteristics of Lining and Cementing Materials, Br Dent J 133: 21-24, 1972. 11. British Standard 336511: Dental Silicate and Dental 5ilicophosphate Cement, 1969. 12. British Standard 3990:1966 Amendment Slip No. 1, in Specification for Acrylic Resin Teeth, August 7, 1968. 13. DEBUS, H.: Contributions to the History of Glyoxylic Acid, J Chem Soc 85: 1382-1403, 1904. 14. LACHMAN, A.: Dihydroxy-Tartaric Acid, J Am Chem Soc 43: 2091-2097, 1921.
Downloaded from jdr.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 4, 2015 For personal use only. No other uses without permission.
considerable influence on working time, an effect that was attributed to the formation of strong aluminum fluoride complexes.8 Aluminum is probably transported from the ionleachable glass to the polyelectrolyte phase as a fluoride complex and as such may not be immediately available for cross-linking polyanion chains. The effect is to prolong working time. It was inferred from this observation that the addition of chelating agents to the system would prove advantageous since they would facilitate the extraction of metal ions from the glass and combine with them in the period when the polymer was ionizing and unwinding. This chelation would have the effect of withholding metal ions from the polyanion chains in the early stages, preventing premature gelatinization and thus prolonging working time. Ultimately, the complexed metal could react with the polymer in either its ionized or unionized form (Fig 1, b, reaction iia) or the free metal ion (hydrated) could react as in Figure 1, a. The overall effect would be to produce a more concerted cementing reaction. Acidic chelating agents would have particular advantages; they would considerably aid the extraction of cations from the glass and at the same time suppress the ionization and unwinding of polyanion chains prolonging working time and at the same time ensure a build-up of metal ions in solution for cross-linking. With these considerations in mind, the effects of a number of chelating agents on the setting behavior of the cement were examined. The dentist is interested in the following three characteristics of setting behavior of a cement: the amount of time available for manipulation, setting time, and postset hardening. These factors that characterize the 489
Downloaded from jdr.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 4, 2015 For personal use only. No other uses without permission.
490
WILSON, CRISP, AND FERNER
J Dent Res May-June 1976
a) In the absence of a chelating agent [Polymer] 5- [Polyanion]X. + xH (coiled) ( inear) M oxide >Mn + > [M.Po(yanion](x-n)M + [Polyanion]x(gel salt)
b)In the presence of a chelatingagent H2L > M0(-2)++ 2H+ H L + M n'
(i)
(ii) Giii)
Wi
+ [Polymer] . [M.Polyanion](X n- +nH" +L2 (fla) L (n --2)+ + [Polyanion]xM ML(n > [ML.Polyanion] ( n) - L(2c (i b) x-(x-
ML(n-2)+
ML(n-2)+
+
[Polyanion] x
>[M. L.Poyonion) (x -n +2) -
(iic)
FIG 1.-Representation of glass-ionomer cement forming reaction: a, in absence of chelating agents; b, in presence of chelating agents.
early phase of the cement-forming reaction are conventionally measured in the laboratory by penetrometer or indentation methods. Such methods do not necessarily give a correct clinical indication, but at least serve for comparative purposes. The measurement of setting time is straightforward and is specified for dental silicate cements in the International Standards Organization Recommendation I.S.O. no. R1565. Working time is not clearly defined and it is estimated by different methods by various workers. In this laboratory, a Gillmore needle with a 28 gm load is used at 23 C and 50%O relative humidity; the end point requires individual operator judgment. Surface hardness can be measured on a superficial indentation tester.a The characterization of the setting of a cement by measuring a working and a setting time is both arbitrary and incomplete. The setting of a cement is a progressive process where there is a continuous, but not necessarily uniform, increase in the stiffness of a cement paste with time. In no sense are there definite points corresponding to a working and a setting time. What is needed, especially in the context of developing improved cement formulation, is an instrument for the continuous monitoring of setting. Such an instrument exists in the oscillating a Wallace Microindentation Tester, H. W. Wallace and Co., Ltd., Croydon, Eng.
rheometer originally developed by Bovis, Harrington, and Wilson9 for use with composites and later applied by Plant, Jones, and Wilson'0 to dental cements. This instrument gives a record of stiffness against time, a rheogram, that is qualitatively more informative than the conventional penetrometer methods. This instrument was adopted in this work as a screening test for setting characteristics. For the purposes of tabulation, a working and a setting time were taken from these rheograms, using appropriate definitions. However, the main feature of interest was the maximum rate of set, which could also be readily obtained from these rheograms. Materials and Methods The cement powder, an ion-leachable glass designated G200, was prepared by fusing a mixture of silica and alumina in a flux of calcium, aluminum and sodium fluorides, and aluminum phosphate. The melt was shock cooled, and the opal glass formed was ground to a fine powder capable of passing through a sieve with a mesh opening of 45 micrometers.4 The PAA solution was prepared by the aqueous polymerization of acrylic acid, using ammonium persulfate as initiator and propan-2-ol as the chain transfer agent. The final product had a weight average molecular weight of 23,000 and was concentrated to
Downloaded from jdr.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 4, 2015 For personal use only. No other uses without permission.
REACTIONS IN GLASS-IONOMER CEMENTS: IV
Vol 55 No. 3
is little difference in the working times obtained from the rheograms by either of these construction methods. Finally, the envelope of the trace approaches the abscissa and becomes linear and sometimes parallel to it (point Q) which is taken as the set. In practice this cannot be assigned very accurately. It corresponds to the setting time for traces at 37 C, and for traces done at 23 C, it can be regarded as the room temperature setting time. The time when the amplitude was reduced to 50% of its initial value was measured. An attempt was made to measure the maximum rate of setting that was considered to be an important parameter. The maximum slope of the curved portion of the envelope was measured ( (XZ) / (XY) on Fig 2). This was expressed as the percentage reduction of the initial amplitude per mi'nute, that is, 100 (XZ) / (AB) (XY) to give a measure of the maximum rate of setting. OTHER MEASUREMENTS.-Certain other measurements were made on selected cements. Compressive strength (24 hours), solubility and disintegration, and setting time (37 C) were determined by standard specification tests.11 Working time was measured at 23 C using a Gillmore needle but with a lower (28 gm) load than the 454 gm load in the standard instrument but with the same needle diameter of 1.05 mm. The end of working time was deemed to be that instance when the needle failed to penetrate 1 mm of cement paste. Tensile strength (24 hours) was measured by the diametral compressive loading of disks (8-mm diameter X 4-mm height). Early surface hardness was measured using
50% w/w by vacuum distillation.4 The various chelating agents were of the best quality laboratory grades and used without further purification. They were generally added to the PAA solution, in the ratio of 5:100 parts by weight, and tumbled until dissolution was complete. When the chelating agent had a limited solubility, saturated solutions or slurries were used. Cements were mixed, for one minute, on a glass block in a room controlled at 23 + I C and 50 + 5% relative humidity, and the powder-liquid ratio recorded. RHEOMETER TRAcES.-Between 1.2 and 1.6 gm of paste was transferred to the platen of the oscillating rheometer, maintained at 23 C in most experiments, and traces recorded on an ultraviolet recorder. Using the rheometer trace, a number of setting characteristics were tabulated. The instances when the amplitude had decreased to 95 (see references 9, 10) and 50% of its original value were recorded. The measurement of setting parameters is shown in Figure 2. The curve BMQP represents the envelope of the trace. The time axis lies along the abscissa whereas the ordinate gives the amplitude of the oscillations. There is an initial straight portion of the envelope parallel to the abscissa followed by a curve as the cement thickens. The point M is at the point of time at which the amplitude of the oscillation is reduced to 95% of the original value, the definition of working time given by Bovis, Harrington, and Wilson.9 An alternative method of estimating working time is to extend the initial straight portion of the curve to point N where the tangent of the second part intersects. There
L__ II
N
I\ I\ I\
II\
0
A
X
491
X
a
F
FIG 2.-Geometrical constructions on rheometric trace: point 0, start of mix; point A, completion of mix; points M and N, working time measured from 0; and Q, gel point. Downloaded from jdr.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 4, 2015 For personal use only. No other uses without permission.
492
J Dent Res May-June 1976
WILSON, CRISP, AND FERNER
cement disks 20 mm in diameter X 1.5-mm thick. The cement paste was packed into a split-ring mold and, two minutes from the start of mixing pressed between two, polythene-lined, flat 2-kg weights maintained at 37 C in an oven. Twelve minutes later, the sample was removed from the mold and placed on the anvil of a hardness machine.12b A standard Vickers diamond was applied, with a load of 1 gm, to the cement surface for a zero reading and then at precisely 15 minutes from the start of mix an additional 300 gm was applied for 15 seconds and the depth of indentation recorded in units of 10-5 inches, as the indentation- number.c
Results Rheograms of all experimental cements were recorded, and selected examples are shown in Figure 3. The setting characteristics of the whole range of experimental cements taken from their rheograms are given b Wallace Hardness Machine, H. W. Wallace and Co. Ltd., Croydon, Eng. c Wallace Indentation Number, H. W. Wallace and Co. Ltd., Croydon, Eng.
in Table 1. Physical and mechanical properties of important variants are given in Table 2. All rheograms (Fig 3) are characterized by (1) an initial period where the amplitude remains constant at about the maximum value and the cement paste remains workable; (2) an intermediate period where the rate of decrease of the amplitude reaches a maximum as the cement is setting; and (3) a final period where the cement sets and a small but constant amplitude is attained. Working times, maximum rate of set, and setting times (23 C) taken from the rheograms are given in Table I for all cements. There is little difference in working times obtained from these rheograms by either construction method. Working times measured by the modified Gillmore needle method agree with those taken from the rheograms' (95%) points (Tables 1, 2). Working time is an inexact parameter and is somewhat arbitrarily defined in laboratory tests. The maximum slope of the curved portion of the rheogram envelope was taken to repre-
t =1 2 min
FIG 3.-Rheometric traces of: (a), glass-ionomer-I cement; (b), ASPA cement plus 5% urea; (c), ASPA cement plus 5% tartaric acid; (d), 5% citric acid; (e), 1% acetylacetone; and (f), saturated solution of salicylic acid, each at P/L 4.0 gm/ml and 23 C. Downloaded from jdr.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 4, 2015 For personal use only. No other uses without permission.
REACTIONS IN GLASS-IONOMER CEMENTS: IV
Vol 55 No. 3
493
TABLE 1 SETTING CHARACrERISTICS OF VARIOUS ASPA CEMENT SYSTEMS Cement Type Addition to Liquid ...
Hydroxy acids 5% Tartaric acid 5% Tartaric acid 5% Dihydroxytartaric acid 5% Citric acid 10% Citric acid Hydroxybenzoic acids 5% Suspension of salicylic acid Saturated solution of salicylic acid 2% 2,6-dihydroxybenzoic acid di Ketones 1% Acetylacetone 2% Acetylacetone 1% Aluminum triacetylacetonate 1% Aluminum triacetylacetonate Complexing agents Saturated solution sequestric acid 5% Slurry sequestric add 5% Nitrilotriacetic acid 5% Ethanolamine Miscellaneous 5% Urea 5% Polyglycol
P/L Ratio (gm/ml-)
Gel Time
Stiffening Rate
(mirn)
95 %
50%
3.5 4.0
1.8 2.4
3.6 3.6
9 8
30 34
3.5 4.0 1.0 4.0 3.5
2.0 1.9 2.0 1.9 2.8
2.9 2.5 3.3 2.9 3.8
6 5 8 6 9
87 32 68 51
4.0
2.5
2.7
10
33
4.0
1.6
2.7
7
49
4.0
2.1
3.5
9
37
4.0 4.0
2.3 2.3
3.5 3.6
9 8
38 33
4.0
2.0
2.9
6
54
3.7
2.0
3.0
9
46
4.0 3.5 4.0 3.5
2.3 2A 2.2 3.5
3.6 3.9 3.7 8.8
8
9 19
36 31 34 9
3.5 4.0
2.2 1.9
5.6 3.2
15 8
(min)
8
48
16
TAkBLE 2
sent the maximum rate of setting. The set-
PHYSICAL PROPERT'IES OF SELECTED ASPA C]EMENTS
ting points (measured from the rheograms) at 23 C are of limited clinical significance. They are of some comparative value and give an indication of sharpness of set. The setting characteristics of any particular glasscement system are affected by the ionomer powder-liquid ratio of the mix, and in making comparisons of data in Table 1, cements of equal powder-liquid ratio should be selected. An increase in this parameter increases the setting rate and tends to reduce the setting time (23 C) . Additives to the liquid can be classified by their effect on the maximum rate of setting. They fall into three groups: those with little effect, namely dihydroxytartaric acid, acetylacetone, sequestric acid, nitrilotriacetic acid, and polyglycol; those that reduce it, namely urea and ethanolamine; and those that increase it, namely salicylic acid, aluminum
5% Tar- 5%o Citric Liquid Additive
None taric Acid
Powder-liquid ratio 3.0 3.0 (gm/ml) 3.0 Working time (min) 2.5 2 6 4 Setting time (min) Wallace indentation number (15 min) 600 120 Compressive strength 181 165 (24 hr, N/mm) Tensile strength 13.9 13.7 (24 hr, N/mm2) Solubility and disintegration (%, 24 hr) 0.47 0.47 Modified Gillmore ne!edle method.
3.0
0
Working Time (min)
Acid 4Q0 4.0 2 3.75
347
192
0.54
Downloaded from jdr.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 4, 2015 For personal use only. No other uses without permission.
494
J Dent Res May-June 1976
WILSON, CRISP, AND FERNER
triacetylacetonate, citric acid, and tartaric acid.
Discussion RHEOLOGICAL CHARACTERISTICS.-The addition of comonomers to the PAA liquid has a striking effect on setting characteristics. The variety of shapes of the envelope curves is shown in Figure 3. Simple visual inspection demonstrates considerable differences and the drawn out rheogram of the sluggish cement (curve b, Fig 3) is easily distinguishable from that of the sharply setting cement (curve c, Fig 3). The effect of the hydroxy acids is the most significant. Both tartaric acid and citric acid considerably increase the maximum setting rate and reduce the setting time (23 C) (Table 1; Fig 3) but do not affect the working time. This represents an obvious improvement in setting characteristics. Aluminum triacetylacetonate and saliyclic acid showed a similar, but much reduced, effect. Surprisingly, dihydroxytartaric acid has little effect. Urea and ethanolamine slow down the setting reaction (Table 1; Fig 3) and adversely affecting setting characteristics. Acetylacetone, sequestric acid, nitrilotriacetic acid, and dihydroxytartaric acid, which are chelating agents, and also polyglycol, have no effect. PHYSICAL PROPERTIES.-The two most promising cement systems (those with tartaric acid and citric acid additives) were subjected to further testing. Results from this study showed that both hardened much more rapidly than the control system without additives; the tartaric acid system also proved to be considerably superior in this respect to the citric acid system. THE ROLE OF CHELATING COMONOMERS.-In
a previous communication,8 the important role of the fluoride ion in the ASPA system was established and the workability of the cement was found to improve with an increase in the rate of elution of fluoride. It was inferred that the formation of a fluoride complex (aluminum forms a range of stable complexes, for example, AlF2+ (K1 = 6.83) AIF2+ (K1 = 5.02)) would temporarily prevent metal ions binding the polyanionic chains thus preventing premature gelation. Organic chelates would be expected to play a similar role and both aluminum and calcium form chelates of the type model 1:
M
(1)
r
This model represents the complexes formed with tartaric acid, citric acid, and acetylacetonate ligands. Complexes formed with trivalent metals are more stable than those formed with divalent ones. For example, the stability constant (K1) for the aluminum-tartaric acid complex is 6.35 compared with 2.17 for the calcium complex. The addition of tartaric acid comonomer to the liquid does not prolong working time, but it increases the rate of hardening. This result is to be attribuited to the dual effect of tartaric acid, which not only tends to hold metals in solution but, by virtue of its complexing action, will extract them more rapidly from the glass powder. If these two effects cancel out, working time will be unaffected. The particular effectiveness of tartaric acid in improving the hardening rate may derive from its structure. There are a pair of ligands for chelation at each end of the molecule (model 2) so that this molecule may serve to bridge pairs of metal ions, the whole grouping linking proximate polyanionic chains. 0
H
11
0
C
M
\
0
/ M
\\ 0 (2) CH--CH
/\ /
\ O
C
11
H 0 The lack of effectiveness of acetylacetone, sequestric acid, salicylic acid, and nitrilotriacetic acid which form stable chelates with aluminum and calcium calls for some explanation. Three possible reasons may be advanced: (1) with the exception of nitrilotriacetic acid, these chelates are relatively insoluble in PAA; (2) unlike tartaric acid, these ligands are only unifunctional and cannot bridge metal ions; and (3) they are
Downloaded from jdr.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 4, 2015 For personal use only. No other uses without permission.
REACTIONS IN GLASS-IONOMER CEMENTS: IV
Vol 55 No. 3
weaker acids than tartaric acid and may not extract ions readily from glasses. The failure of dihydroxytartaric acid to be as effective as tartaric acid requires some further explanation. Dihydroxytartaric acid is not a simple hydroxy acid. Debus'3 in 1904 suggested a cyclic structure (model 3): HO-C COOH 0
H20
HO -C COOH a (3) whereas Lachman'4 in 1921 observed that it acted as a hydrated a-diketone. These structures would account for the difference in behavior from that of tartaric acid. Dihydroxytartaric acid forms insoluble salts with metals, including sodium, which would nullify any effect it might have on the cementing reaction.
Conclusions The oscillating rheometer is a valuable instrument for studying the effects of additives on the setting behavior of a cement system. Using this instrument, it was found that certails chelating comonomers, the hydroxycarboxylic acids, could improve the setting characteristics of the glass-ionomer cement system when added to the PAA solution. The acid chelates probably assign the extraction of metal ions from the glass and also tend to hold them in solution, preventing premature ion binding of the polyanion chains. The effect is to increase the rate of hardening without reducing the working time, which may indeed be slightly increased. Tartaric acid, the most effective of the comonomers, can form a chelate bridge between aluminum atoms, and this metal complex probably acts as a flexible bridge structure linking polyanion chains. This mechanism offers some steric advantages over a simple salt bridge. The authors thank the Government Chemist, Dr. H. Egan, for permission to contribute this article. This work was done on behalf of the National Research Development Council. Crown Copyright Reproduced by the Controller of Her Britannic Majesty's Stationery Office.
495
References 1. WILSON, A.D., and KENT, B.E.: The Glasslonomer Cement: A New Translucent Dental Filling Material, J Appl Chem Biotechnol 21: 313, 1971. 2. WILSON, A.D., and KENT, B.E.: A New Translucent Cement for Dentistry: The Glass lonomer Cement, Br Dent J 132: 133135, 1972. 3. WILSON, A.D., and KENT, B.E.: (Surgical Cemcnt) Br Patent No. 1,316,129, 1973. 4. KENT, B.E.; LEWis, B.G.; and WILSON, A.D.: The Properties of a Glass lonomer Cement, Br Dent J 135: 322-326, 1973. 5. McLEAN, J., and WILSON, A.D.: Fissure Sealing and Filling with an Adhesive Glasslonomer Cement, Br Dent J 136: 269-276, 1974. 6. CRISP, S., and WILSON, A.D.: Reactions in Glass lonomer Cements: I. Decomposition of the Powder, J Dent Res 53: 1408-1413, 1974. 7. CRISP, S., and WILSON, A.D.: Reactions in Glass lonomer Cements: III. The Precipitation Reaction, J Dent Res 53: 1420-1424, 1974. 8. CRISP, S.; PRINGUER, M.A.; WARDLEWORTH, D.; and WILSON, A.D.: Reactions in Glass Ionomer Cements: II. An Infra-Red Spectroscopic Study, J Dent Res 53: 1414-1419, 1974. 9. BoVIs, S.C.; HARRINGTON, E.; and WILSON, H.J.: Setting Characteristics of Composite Filling Materials, Br Dent J 131: 352-356, 1971. 10. PLANT, C.G.; JONES, I.H.; and WILSON, H.J.: Setting Characteristics of Lining and Cementing Materials, Br Dent J 133: 21-24, 1972. 11. British Standard 336511: Dental Silicate and Dental 5ilicophosphate Cement, 1969. 12. British Standard 3990:1966 Amendment Slip No. 1, in Specification for Acrylic Resin Teeth, August 7, 1968. 13. DEBUS, H.: Contributions to the History of Glyoxylic Acid, J Chem Soc 85: 1382-1403, 1904. 14. LACHMAN, A.: Dihydroxy-Tartaric Acid, J Am Chem Soc 43: 2091-2097, 1921.
Downloaded from jdr.sagepub.com at UNIV OF CALIFORNIA SANTA CRUZ on April 4, 2015 For personal use only. No other uses without permission.
