Bonding of a light-curing glass-ionomer cement to dental amalgam Y.E.Y. Aboush R.J. Elderton Department of Conservative Dentistry University of Bristol Lower Maudlin Street Bristol BS1 2LY, England United Kindgom Received April 4, 1990 Accepted August 21, 1990 Dent Mater 7:130-132, April, 1991
Abstract-In the clinical situation, the need may arise for placement of a glass. ionomer cement over an existing amalgam restoration. This study assessed the tensile bond strength of a recently developed light-curing glass ionomer (Vitrabond) to dental amalgam (Dispersalloy), with and without the use of Scotchbond dual cure as an intermediary. Amalgam adherend specimens were prepared, then aged in water at 37°0 for seven days. Immediately before being bonded, the amalgam surfaces were finished flat on 600-grit paper. Forty specimens were used for bonding in this condition, and another 40 were covered with a thin layer of Scotchbond, which was light-cured for 10 s. The glass-ionomer was applied to the adherend surface in two increments, each light-cured for 30 s. After being bonded, half the specimens were stored in water at 37°0, while half were stored in an environment of 95 _+ 5% RH at 37°0. The 24-hour tensile bond strenglhs, in MPa, were: for specimens stored in water, without Scotchbond 8.4 _ 1.2, with Scotchbond 4.7 __ 1.3; and for specimens stored in 95 __ 5% RH, without Scotchbond 9.2 ± 2.1, with Scotchbond 4.6 ± 1.5. The data were further analyzed by the Weibull distribution function. It was concluded that a strong reliable bond can be achieved between Vitrabond and set Dispersalloy, and that the use of Scotchbond as an intermediary is contra-indicated.
G
lass-ionomer cements are being used increasingly in dentistry for a variety of purposes. The material is currently accepted by many practitioners for replacing missing tooth substance, luting of pre-formed restorations, lining under amalgam or resin composite restorations, and fissure sealing (Croll, 1990; McLean, 1988; Mount, 1990). In addition, the ability of glass-ionomer cements to bond to set amalgam has several clinical applications: addition to incomplete or damaged amalgam cores intended to support crowns or other laboratory-made restorations, "longterm" temporary restorations adjacent to existing amalgam restorations, and the repair of fractured cusps of teeth containing amalgam restorations. In these situations, the glass ionomer may be used on its own, or it may act as an intermediary for bonding composite to the amalgam in a fashion similar to that described for attaching composite to dentin (McLean et al., 1985). Other possible situations in which a glass ionomer might be beneficial if it were bonded to amalgam include: the repair of ditched amalgam restorations, and as a fissure s e a l a n t adjacent to an existing amalgam restoration. Previous work has shown that the traditional chemical-curing glassionomer cements form strong adhesive bonds with set amalgam (Aboush and Jenkins, 1989). However, the setting characteristics of the chemical-curing glass ionomers (Smith, 1990; Wilson and McLean, 1988) have been a limitation for clinicians. This is one reason why the introduction of light-curing glass ionomers, with a "command set", constitutes a major step forward, since their use reduces the valuable clinical time involved and thereby minimizes the effect of any early moisture contamination. Furthermore, the light-curing glass ionomers are claimed to
130 ABOUSH & ELDERTON/BONDING OF GLASS IONOMER TO AMALGAM
possess enhanced physical and mechanical properties, as compared with their chemical-curing counterparts. The aim of this study was to assess the tensile bond strength of a light-curing glass-ionomer cement to set dental amalgam, with and without the use of an intermediary agent. MATERIALS AND METHODS The materials used were: a lightcuring glass-ionomer liner/base (Vitrabond, 3M Dental Products Division, St. Paul, MN, USA); a dualcure dental adhesive (Scotchbond, 3M Dental Products Division, St. Paul, MN, USA); and a dispersed-phase amalgam alloy (Dispersalloy, Johnson & Johnson, East Windsor, N J, U S A - 1 spill, regular set). All the materials were dispensed, mixed, and cured according to the manufacturers' instructions. Dispersalloy capsules were mixed for eight s with a mechanical mixer (Silamat, Vivadent, Schaan/Liechtenstein). Scotchbond and Vitrabond were lightcured by means of a Visilux 2 light source (3M Dental Products Division, St. Paul, MN, USA). By use of a clinical technique, 80 amalgam adherend specimens were prepared by the packing of Dispersalloy, in excess, into holes (2 mm deep and 5 mm in diameter) which had been drilled in poly(methyl methacrylate) blocks. The specimens were then allowed to age in distilled water at 37°C for seven days. Immediately before being bonded, the amalgam surfaces were finished flat on wet 600-grit silicon carbide paper, washed with distilled water, and dried with oil-free air delivered from a chip syringe. Forty specimens were then used for bonding in this condition, while 40 were first covered with a thin layer of Scotchbond, which was light-cured for 10 s. Bond test specimens were prepared by use of the assembly shown
in Fig. 1. This apparatus enabled a truncated cone of material, 2.5 mm in diameter at the bond surface and 3.5 mm in height, to be bonded to the adherend. Before the test pieces were assembled, the acetal disc which housed the adhesive was sprayed with PTFE dry film lubricant. One level scoop of Vitrabond powder was mixed with one drop of the liquid for 15 s. The mixed cement was applied to the amalgam surface through the hole in the acetal disc. So that the Vitrabond would be completely cured, the mix was applied in two increments, each light-cured for 30 s. The bonded assemblies were left undisturbed on the bench for 10 rain before being dismantled. Half the bond test specimens were then stored for 24 h in water at 37°C, while half were stored in an environment of 95 _- 5% RH at 37°C obtained by suspension of the specimens above the water level in an enclosed water bath+ With the test rig used as shown in Fig. 2, each test specimen was loaded in tension to failure, on a universal testing machine (Instron, Instron Ltd, Bucks, UK), at a cross-head speed of 2 ram/rain. On completion of the bond test, the fracture surfaces were examined under a light microscope (X 40) so that the mode of bond failure could be assessed. RESULTS The bond strength was calculated from the load required to cause frac-
ture of the bond divided by the area of the exposed adherend surface. The data obtained were analyzed by the Weibull distribution function (Table and Fig. 3), as described by McCabe and Walls (1986). The bonds achieved without the use of an intermediary were superior to those obtained when the Scotchbond intermediary was used. This was evident from the position of the curves on the stress axes in Fig. 3. Also, the higher Weibull moduli recorded for the groups in which no intermediary was used (Table) indicate that the bonds achieved were more reliable. Separately for the intermediary and the non-intermediary groups of specimens, there was little difference in the mean bond strength values or the normalizing parameters (characteristic strengths) (Table) between the specimens stored in water at 37°C and those stored in an environment of 95 _ 5% RH at 37°C. However, the higher gradients of the upper pair of curves in Fig. 3, as compared with the lower pair, indicated that w a t e r s t o r a g e of the specimens produced more reproducible results. Among the 40 specimens for which no intermediary was used, bond failure was of the adhesive/cohesive type for 34 specimens; two of those stored in water and four stored at 95 -+ 5% RH failed adhesively. F o r the specimens bonded via
Scotchbond, bond failure occurred mostly through the Scotchbondamalgam interface; only three specimens of the group that was stored in w a t e r exhibited adhesive/cohesive failure (parts of the Scotchbond layer remained attached to the amalgam).
DISCUSSION A pilot study showed that the bond strengths of Vitrabond to enamel and dentin were of the order of 13 and 5 MPa, respectively. The presently achieved bond strength values of Vitrabond to set amalgam (of around 9 MPa) lie between these two values. This, combined with the relatively high Weibull modulus obtained and the adhesive/cohesive fracture of the bond in most of the specimens, indicates that, in the clinical situation, the bonding of Vitrabond to amalgam could prove to be a reliable procedure. The r e s u l t s show t h a t using Scotchbond as an intermediary to bond Vitrabond to set amalgam is both disadvantageous and unnecessary. Examination of the mode of bond failure demonstrated that the Scotchbond-amalgam interface was the weak link in the joint.
Fracture Probability 0.8
./
o.6
.+/ ;:
0.4
/
~
4y
4:'
I TO UNIVERSAL JOINT 0.2
@@
2
.3 PLATE 0 .INE
IUM GUIDE DISC COATED "FE LUBRICANT
3
MOULD
9
12
15
Fracture Probability
3C COATED LUBRICANT )
6
Tensile Stress (Mpa)
~IOMER
o.s-
:
~,M 3ED rLIC MOULD
/
/.
+if* +
0.2
:~
+
:;/
0
3
Fig. 1. Diagrammatic illustration of the assembly used for preparation of the bond test specimens.
Fig. 2. Test rig used for measurement of tensile bond strength.
6
9
12
Tensile Stress (MPa) Fig. 3. Plots showing the probability of adhesion failure against tensile stress. * = Scotchbond intermediary, + = No intermediary. The superimposed curves represent the best leastsquares fit of the experimental points to the Weibull distribution. (Upper) Specimens stored in water. (Lower) Specimens stored In 95 - 5% RH.
Dental Materials~April 1991
131
TABLE WEIBULL ANALYSISFOR VITRABOND/DISPERSALLOYTENSILE BOND STRENGTHS
Test Group* No intermediary-stored in water No intermediary-stored in 95% RH Scotchbond intermediary-stored in water Scotchbond intermediary-stored in 95% RH *There were 20 specimens in each group.
The results of this study are somewhat different from those observed in a previous study when chemicalcuring glass-ionomer cements were used (Aboush and Jenkins, 1989). The chemical-curing glass ionomers gave slightly lower mean bond strength values (7 to 8 MPa), and the bonds fractured cohesively through the cement. This indicates that the bond between these chemical-curing glass ionomers and amalgam was stronger than their cohesive strength, and that they have a lower cohesive strength than Vitrabond. The finding that comparable mean bond strength values were achieved with storage of bonded specimens in water or at 95 _+ 5% RH suggests that early moisture contamin a t i o n of t h e s e t ( l i g h t - c u r e d ) Vitrabond had no obvious deleterious effect on its adhesive characteristics. The more reproducible results obtained for the specimens that were stored in water imply that, under the experimental con-
Mean Bond Strength in MPa (SD) 8.42 (1.24) 9.24 (2.11) 4.71 (1.31) 4.61 (1.45)
Normalizing Parameter 8.97 10.11 5.22 5.22
Weibull Modulus (SE) 7.13 (0.24) 4.56 (0.32) 3.66 (0.15) 2.82 (0.15)
ditions of this work, water was a more consistent storage medium than 95 + 5% RH. Again, this highlights the importance of the development of a s t a n d a r d t e s t m e t h o d for u s e in d e n t a l b o n d strength tests. In a n o t h e r s t u d y (Aboush and Elderton, 1990) that investigated the bonding of amalgam triturate to set Vitrabond, and in which a fresh mix of uncured Vitrabond was used as an intermediary, a mechanical bond with a mean bond strength value of only 5 MPa was achieved. The fact that a value of 9 MPa was achieved in the present study for Vitrabond bonded to set amalgam indicates that bonding between Vitrabond and amalgam depends upon whether or not the amalgam has set completely. REFERENCES ABOUSH,Y. E. Y. and ELDERTON,R. J. (1990): Bonding of Dental Amalgam to a Light Cure Glass-ionomer Liner/
132 ABOUSH & ELDERTON/BONDING OF GLASS IONOMER TO AMALGAM
Stress for 1% Chance of Fracture in MPa 4.70 3.69 1.49 1.02
Stress for 90% Chance of Fracture in MPa 10.08 12.14 6.56 7.02
Base, J Dent Res 69: 989, Abstr. No. 277. ABOUSH, Y.E.Y. and JENKINS, C.B.G. (1989): The Bonding of Glass-ionomer Cements to Dental Amalgam, Br Dent J 166: 255-257. CROLL, T.P. (1990): Glass Ionomers for Infants, Children, and Adolescents, J A m Dent Assoc 120: 65-68. MCCABE, J.F. and WALLS, A.W.G. (1986): The Treatment of Results for Tensile Bond Strength Testing, J Dent 14: 165-168. MCLEAN, J.W. (1988): Glass-ionomer Cements, Br Dent J 164: 293-300. MCLEAN, J.W.; PowIs, D.R.; PROSSER, H.J.; and WILSON,A.D. (1985): The Use of Glass-ionomer Cements in Bonding Composite Resins to Dentine, Br Dent J 158: 410-414. MOUNT,G.C. (1990): An Atlas of Glassionomer Cements, London: Martin Dunitz, pp. 1-103. SMITH, D.C. (1990): Composition and Characteristics of Glass Ionomer Cements, J A m Dent Assoc 120: 20-22. WILSON,&D. and McLEAN, J.W. (1988): Glass-ionomer Cement, Chicago: Quintessence Publishing Co., pp. 4356.
Abstract-In the clinical situation, the need may arise for placement of a glass. ionomer cement over an existing amalgam restoration. This study assessed the tensile bond strength of a recently developed light-curing glass ionomer (Vitrabond) to dental amalgam (Dispersalloy), with and without the use of Scotchbond dual cure as an intermediary. Amalgam adherend specimens were prepared, then aged in water at 37°0 for seven days. Immediately before being bonded, the amalgam surfaces were finished flat on 600-grit paper. Forty specimens were used for bonding in this condition, and another 40 were covered with a thin layer of Scotchbond, which was light-cured for 10 s. The glass-ionomer was applied to the adherend surface in two increments, each light-cured for 30 s. After being bonded, half the specimens were stored in water at 37°0, while half were stored in an environment of 95 _+ 5% RH at 37°0. The 24-hour tensile bond strenglhs, in MPa, were: for specimens stored in water, without Scotchbond 8.4 _ 1.2, with Scotchbond 4.7 __ 1.3; and for specimens stored in 95 __ 5% RH, without Scotchbond 9.2 ± 2.1, with Scotchbond 4.6 ± 1.5. The data were further analyzed by the Weibull distribution function. It was concluded that a strong reliable bond can be achieved between Vitrabond and set Dispersalloy, and that the use of Scotchbond as an intermediary is contra-indicated.
G
lass-ionomer cements are being used increasingly in dentistry for a variety of purposes. The material is currently accepted by many practitioners for replacing missing tooth substance, luting of pre-formed restorations, lining under amalgam or resin composite restorations, and fissure sealing (Croll, 1990; McLean, 1988; Mount, 1990). In addition, the ability of glass-ionomer cements to bond to set amalgam has several clinical applications: addition to incomplete or damaged amalgam cores intended to support crowns or other laboratory-made restorations, "longterm" temporary restorations adjacent to existing amalgam restorations, and the repair of fractured cusps of teeth containing amalgam restorations. In these situations, the glass ionomer may be used on its own, or it may act as an intermediary for bonding composite to the amalgam in a fashion similar to that described for attaching composite to dentin (McLean et al., 1985). Other possible situations in which a glass ionomer might be beneficial if it were bonded to amalgam include: the repair of ditched amalgam restorations, and as a fissure s e a l a n t adjacent to an existing amalgam restoration. Previous work has shown that the traditional chemical-curing glassionomer cements form strong adhesive bonds with set amalgam (Aboush and Jenkins, 1989). However, the setting characteristics of the chemical-curing glass ionomers (Smith, 1990; Wilson and McLean, 1988) have been a limitation for clinicians. This is one reason why the introduction of light-curing glass ionomers, with a "command set", constitutes a major step forward, since their use reduces the valuable clinical time involved and thereby minimizes the effect of any early moisture contamination. Furthermore, the light-curing glass ionomers are claimed to
130 ABOUSH & ELDERTON/BONDING OF GLASS IONOMER TO AMALGAM
possess enhanced physical and mechanical properties, as compared with their chemical-curing counterparts. The aim of this study was to assess the tensile bond strength of a light-curing glass-ionomer cement to set dental amalgam, with and without the use of an intermediary agent. MATERIALS AND METHODS The materials used were: a lightcuring glass-ionomer liner/base (Vitrabond, 3M Dental Products Division, St. Paul, MN, USA); a dualcure dental adhesive (Scotchbond, 3M Dental Products Division, St. Paul, MN, USA); and a dispersed-phase amalgam alloy (Dispersalloy, Johnson & Johnson, East Windsor, N J, U S A - 1 spill, regular set). All the materials were dispensed, mixed, and cured according to the manufacturers' instructions. Dispersalloy capsules were mixed for eight s with a mechanical mixer (Silamat, Vivadent, Schaan/Liechtenstein). Scotchbond and Vitrabond were lightcured by means of a Visilux 2 light source (3M Dental Products Division, St. Paul, MN, USA). By use of a clinical technique, 80 amalgam adherend specimens were prepared by the packing of Dispersalloy, in excess, into holes (2 mm deep and 5 mm in diameter) which had been drilled in poly(methyl methacrylate) blocks. The specimens were then allowed to age in distilled water at 37°C for seven days. Immediately before being bonded, the amalgam surfaces were finished flat on wet 600-grit silicon carbide paper, washed with distilled water, and dried with oil-free air delivered from a chip syringe. Forty specimens were then used for bonding in this condition, while 40 were first covered with a thin layer of Scotchbond, which was light-cured for 10 s. Bond test specimens were prepared by use of the assembly shown
in Fig. 1. This apparatus enabled a truncated cone of material, 2.5 mm in diameter at the bond surface and 3.5 mm in height, to be bonded to the adherend. Before the test pieces were assembled, the acetal disc which housed the adhesive was sprayed with PTFE dry film lubricant. One level scoop of Vitrabond powder was mixed with one drop of the liquid for 15 s. The mixed cement was applied to the amalgam surface through the hole in the acetal disc. So that the Vitrabond would be completely cured, the mix was applied in two increments, each light-cured for 30 s. The bonded assemblies were left undisturbed on the bench for 10 rain before being dismantled. Half the bond test specimens were then stored for 24 h in water at 37°C, while half were stored in an environment of 95 _- 5% RH at 37°C obtained by suspension of the specimens above the water level in an enclosed water bath+ With the test rig used as shown in Fig. 2, each test specimen was loaded in tension to failure, on a universal testing machine (Instron, Instron Ltd, Bucks, UK), at a cross-head speed of 2 ram/rain. On completion of the bond test, the fracture surfaces were examined under a light microscope (X 40) so that the mode of bond failure could be assessed. RESULTS The bond strength was calculated from the load required to cause frac-
ture of the bond divided by the area of the exposed adherend surface. The data obtained were analyzed by the Weibull distribution function (Table and Fig. 3), as described by McCabe and Walls (1986). The bonds achieved without the use of an intermediary were superior to those obtained when the Scotchbond intermediary was used. This was evident from the position of the curves on the stress axes in Fig. 3. Also, the higher Weibull moduli recorded for the groups in which no intermediary was used (Table) indicate that the bonds achieved were more reliable. Separately for the intermediary and the non-intermediary groups of specimens, there was little difference in the mean bond strength values or the normalizing parameters (characteristic strengths) (Table) between the specimens stored in water at 37°C and those stored in an environment of 95 _ 5% RH at 37°C. However, the higher gradients of the upper pair of curves in Fig. 3, as compared with the lower pair, indicated that w a t e r s t o r a g e of the specimens produced more reproducible results. Among the 40 specimens for which no intermediary was used, bond failure was of the adhesive/cohesive type for 34 specimens; two of those stored in water and four stored at 95 -+ 5% RH failed adhesively. F o r the specimens bonded via
Scotchbond, bond failure occurred mostly through the Scotchbondamalgam interface; only three specimens of the group that was stored in w a t e r exhibited adhesive/cohesive failure (parts of the Scotchbond layer remained attached to the amalgam).
DISCUSSION A pilot study showed that the bond strengths of Vitrabond to enamel and dentin were of the order of 13 and 5 MPa, respectively. The presently achieved bond strength values of Vitrabond to set amalgam (of around 9 MPa) lie between these two values. This, combined with the relatively high Weibull modulus obtained and the adhesive/cohesive fracture of the bond in most of the specimens, indicates that, in the clinical situation, the bonding of Vitrabond to amalgam could prove to be a reliable procedure. The r e s u l t s show t h a t using Scotchbond as an intermediary to bond Vitrabond to set amalgam is both disadvantageous and unnecessary. Examination of the mode of bond failure demonstrated that the Scotchbond-amalgam interface was the weak link in the joint.
Fracture Probability 0.8
./
o.6
.+/ ;:
0.4
/
~
4y
4:'
I TO UNIVERSAL JOINT 0.2
@@
2
.3 PLATE 0 .INE
IUM GUIDE DISC COATED "FE LUBRICANT
3
MOULD
9
12
15
Fracture Probability
3C COATED LUBRICANT )
6
Tensile Stress (Mpa)
~IOMER
o.s-
:
~,M 3ED rLIC MOULD
/
/.
+if* +
0.2
:~
+
:;/
0
3
Fig. 1. Diagrammatic illustration of the assembly used for preparation of the bond test specimens.
Fig. 2. Test rig used for measurement of tensile bond strength.
6
9
12
Tensile Stress (MPa) Fig. 3. Plots showing the probability of adhesion failure against tensile stress. * = Scotchbond intermediary, + = No intermediary. The superimposed curves represent the best leastsquares fit of the experimental points to the Weibull distribution. (Upper) Specimens stored in water. (Lower) Specimens stored In 95 - 5% RH.
Dental Materials~April 1991
131
TABLE WEIBULL ANALYSISFOR VITRABOND/DISPERSALLOYTENSILE BOND STRENGTHS
Test Group* No intermediary-stored in water No intermediary-stored in 95% RH Scotchbond intermediary-stored in water Scotchbond intermediary-stored in 95% RH *There were 20 specimens in each group.
The results of this study are somewhat different from those observed in a previous study when chemicalcuring glass-ionomer cements were used (Aboush and Jenkins, 1989). The chemical-curing glass ionomers gave slightly lower mean bond strength values (7 to 8 MPa), and the bonds fractured cohesively through the cement. This indicates that the bond between these chemical-curing glass ionomers and amalgam was stronger than their cohesive strength, and that they have a lower cohesive strength than Vitrabond. The finding that comparable mean bond strength values were achieved with storage of bonded specimens in water or at 95 _+ 5% RH suggests that early moisture contamin a t i o n of t h e s e t ( l i g h t - c u r e d ) Vitrabond had no obvious deleterious effect on its adhesive characteristics. The more reproducible results obtained for the specimens that were stored in water imply that, under the experimental con-
Mean Bond Strength in MPa (SD) 8.42 (1.24) 9.24 (2.11) 4.71 (1.31) 4.61 (1.45)
Normalizing Parameter 8.97 10.11 5.22 5.22
Weibull Modulus (SE) 7.13 (0.24) 4.56 (0.32) 3.66 (0.15) 2.82 (0.15)
ditions of this work, water was a more consistent storage medium than 95 + 5% RH. Again, this highlights the importance of the development of a s t a n d a r d t e s t m e t h o d for u s e in d e n t a l b o n d strength tests. In a n o t h e r s t u d y (Aboush and Elderton, 1990) that investigated the bonding of amalgam triturate to set Vitrabond, and in which a fresh mix of uncured Vitrabond was used as an intermediary, a mechanical bond with a mean bond strength value of only 5 MPa was achieved. The fact that a value of 9 MPa was achieved in the present study for Vitrabond bonded to set amalgam indicates that bonding between Vitrabond and amalgam depends upon whether or not the amalgam has set completely. REFERENCES ABOUSH,Y. E. Y. and ELDERTON,R. J. (1990): Bonding of Dental Amalgam to a Light Cure Glass-ionomer Liner/
132 ABOUSH & ELDERTON/BONDING OF GLASS IONOMER TO AMALGAM
Stress for 1% Chance of Fracture in MPa 4.70 3.69 1.49 1.02
Stress for 90% Chance of Fracture in MPa 10.08 12.14 6.56 7.02
Base, J Dent Res 69: 989, Abstr. No. 277. ABOUSH, Y.E.Y. and JENKINS, C.B.G. (1989): The Bonding of Glass-ionomer Cements to Dental Amalgam, Br Dent J 166: 255-257. CROLL, T.P. (1990): Glass Ionomers for Infants, Children, and Adolescents, J A m Dent Assoc 120: 65-68. MCCABE, J.F. and WALLS, A.W.G. (1986): The Treatment of Results for Tensile Bond Strength Testing, J Dent 14: 165-168. MCLEAN, J.W. (1988): Glass-ionomer Cements, Br Dent J 164: 293-300. MCLEAN, J.W.; PowIs, D.R.; PROSSER, H.J.; and WILSON,A.D. (1985): The Use of Glass-ionomer Cements in Bonding Composite Resins to Dentine, Br Dent J 158: 410-414. MOUNT,G.C. (1990): An Atlas of Glassionomer Cements, London: Martin Dunitz, pp. 1-103. SMITH, D.C. (1990): Composition and Characteristics of Glass Ionomer Cements, J A m Dent Assoc 120: 20-22. WILSON,&D. and McLEAN, J.W. (1988): Glass-ionomer Cement, Chicago: Quintessence Publishing Co., pp. 4356.
