**4.3 Bonding substrate**

Dentine is quite a variable tissue. Within the tooth itself the dentine approaching the dentino-enamel junction is more highly mineralized and the area occupied by the tubules is less than that of dentine adjacent to the pulp. In addition to this, dentine should be considered as a dynamic tissue that changes due to ageing, in response to caries and restoration placement. Most changes relate to occlusion of tubules and also an increase in the mineralization of the dentine. The implication of this is that the dentine becomes slightly more difficult to etch and exposure of collagen fibrils can also be reduced, hence there is a potential for the bond to be somewhat tenuous. This is particularly the case for the highly sclerosed dentine of non-carious cervical lesions (NCCLs). Laboratory studies indicate that the hybrid layer of the dentine surface of NCCLs is thinner than that of normal dentine [29, 63]. In addition, it seems that some bonding systems do not adhere as well to this surface and show a slightly decreased bond strength [25, 35, 79].

A considerable amount of work has also been done looking at the variation of the bond to caries-affected dentine. Some of the early studies used artificial caries like lesions. However, this does not reproduce the situation that occurs in the oral cavity since caries is a process of demineralization and remineralization associated with the damage of the supporting collagen matrix [48, 53]. The increased thickness of the hybrid layer is mainly because the dentine is already partially demineralized from the caries and the action of the acid etch is therefore somewhat greater. This provides a clear basis for not etching for longer than that recommended by the manufacturer. In addition, the water content of caries-affected dentine is believed to be greater than normal dentine. This too will also have an effect on the ability of the resins to penetrate to the full depth of the demineralized dentine. In the case of cariesaffected dentine treated with chemo-mechanical caries removal solutions, there appear to be no adverse effects on the bond with a DBA [79].

However, the bond to radicular and pulp chamber dentine does seem to vary quite a lot depending on the DBA used. This perhaps provides a strong case for being careful with the selection of a DBA for these regions of the tooth. It is believed that it may be necessary to use different DBAs for different regions of the tooth, or a system needs to be selected where it has been shown to provide a reliable bond to all parts of the tooth. Another alternative is the use of GIC restorative materials when then is a deep cavity on the radicular surface of a tooth, as it is known that a reliable bond can be achieved and moisture control is not such a problem [25, 35, 79].

Filling Materials for the Caries 349

introduced in the 1950s. These composites had improved mechanical properties and good esthetics; they did not bond to tooth structure, and still exhibited significant polymerization

One way to address the polymerization shrinkage problem is to use high molecular weight monomers. In 1962 Bowen [11], while at the National Bureau of Standards, synthesized an acrylated epoxy using glycidylmethacrylate and Bisphenol A epoxy for use as a matrix for dental composite. The resulting monomer, called Bis-GMA or Bowen's resin, possessed the viscosity of honey, and therefore limited the amount of filler particles that could be incorporated. Subsequent experiments incorporated triethylene glycol dimethacrylate (TEGDMA) as a diluent to reduce the viscosity. This monomer combination worked well, and has become one of the most widely used matrix monomer combinations for dental composites to date. The structures of Bis-GMA and TEGDMA are shown in Figs. 3 and 4,

Fig. 3. The chemical structure of Bis-GMA, a resin invented by Ray Bowen. It also is referred

Fig. 4. TEGDMA. The chemical structure of triethyleneglycol dimethacrylate (TEGDMA, which is also abbreviated TEDMA and TEGMA). The structure of methyl methacrylate

shrinkage [10, 46, 55, 58].

respectively.

to as Bowen's resin.

(MMA) is shown for comparison.

### **4.4 Clinical studies**

There has been a considerable amount of work done to evaluate the success or otherwise of DBAs in clinical studies. However, one of the great problems has been that many of the DBAs have been considerably changed or a new material introduced by the time these studies are completed or published. Many of the studies have also been performed on NCCL, which means the outcomes cannot really be applied to restorations in other parts of the mouth because NCCL dentine is usually sclerosed and therefore different from that of an intracoronal cavity. However, these outcomes will provide some indication as to whether the DBA is able to achieve a durable bond under very harsh conditions. Since the early materials were introduced, the retention rates of the DBAs to sclerosed cervical dentine have steadily improved to extent that retention rates are little different from GICs [79].

With regard to clinical studies on posterior teeth restored with a DBA, there is still little evidence available. It would seem though, that clinical studies of resin composite restorations are showing evidence that when placed in the correct manner and the patient has a low caries rate, restoration survival is approaching that of amalgam [30].

When it comes to the use of DBAs, it is important to follow the manufacturers' directions carefully. Overetching can create a situation where there will potentially be a region of poorly or uninfiltrated dentine. This zone may be susceptible to acid or enzyme attack from oral bacteria, hence leading to bond failure [82].

In the case of the self-etching priming materials, this is not believed to be a problem. However, the converse problem may occur: as mentioned, the dentine or smear layer may neutralize the etching primer if the primer has a relatively high pH. The anecdotal evidence would seem to indicate that gentle agitation of these solutions may assist with the etching. However, there are no research data to support this [35, 79].
