Removal of Modern Ceramics

Andrew Spath, DDS, FAGD; and Colby Smith, DDS, FAGD

With increasing frequency, patients are presenting to dental offices requiring the removal and replacement of high-strength ceramic restorative materials such as lithium disilicate, full-contour zirconia, and porcelain-veneered zirconia. Removal of these materials can be time consuming and expensive, with the potential for negative repercussions for the patient.

The average life expectancy of crowns can vary greatly based on multiple factors. Because these materials have been on the market for a relatively limited time, no long-term survival studies (ie, greater than 12 years) are available specific to lithium disilicate and zirconia. However, the use of these new-generation high-strength ceramic materials is mounting rapidly due to lower cost and their impressive esthetic qualities. This is prompting more dentists and laboratories to switch from porcelain-fused-to-metal (PFM) crowns to all-ceramic restorations.¹

The modes of failure for crowns include recurrent caries, cementation error, endodontic access, iatrogenic dentistry, implant-screw loosening, esthetic failure, restorative-material failure, and occlusal issues. Preparation design, the choice of whether to cement or bond, and restorative-material selection are paramount factors for retention.² When considering options for high-strength ceramic materials, obtaining the necessary retention while also allowing for safe removal of the material in the future is crucial. Preparation height, remaining enamel, and the understanding of patient risk factors are important decision-making elements when choosing between adhesively and cohesively retained restorations.

This article will address the challenges faced during removal of adhesively retained ceramics based on current research and the most effective methods to reduce risk. It will also discuss techniques for preparation for cohesively retained restorations and selection of luting agents aimed at minimizing risk during future removal, as well as the latest removal techniques.

Challenges in Removal of Lithium Disilicate and Zirconia

The removal of lithium disilicate and zirconia can be time-consuming and expensive, and often may result in iatrogenic side effects. Soft-tissue issues include damage to the gingiva and mucosa and possible burns from heat generation of electric handpieces.³ Hard-tissue risks include difficulty in removal of unwanted tooth structure on the axial and pulpal walls, heat generation leading to pulpal necrosis, tooth fracture from use of a crown spreader, and damage to adjacent teeth.⁴ To limit these negative outcomes, the latest technology and contemporary research will be discussed to help clinicians increase predictability, safety, and efficiency addressing this issue.

Cementation Protocols for Lithium Disilicate

A common misconception is that because lithium disilicate is a glass-ceramic, lithium-disilicate crowns must be cemented with self-etching or dual-cure composite cement. However, with manufacturer-recommended thickness (1.5 mm for occlusal of full-coverage posterior restorations) and appropriate resistance and retention form, this material is extremely strong (360 MPa to 400 MPa flexural strength in its monolithic form)⁵ and can be cemented with many luting agents, including resin-modified glass ionomer (RMGI), without compromise to important clinical outcomes. Recently Ivoclar Vivadent, a manufacturer of lithium disilicate, released data supporting 1 mm of occlusal monolithic thickness in the posterior as acceptable. This data indicates the mean biaxial flexural strength of IPS e.max Press to be 473 mPA and IPS e.max CAD to be 536 mPA, making for an average of 506 mPA after testing 3633 batches over 11 years. It is the recommendation of the authors that at minimal thickness a dual-cure resin or self-etching resin cement be used.

Moreover, lithium-disilicate specimens produced the highest values when compared with monolithic systems. “With a breaking load of 6000 N, this material is not only capable of withstanding the physiological forces in the posterior region, which typically range from 300 to 1000 N, but also offers sufficient additional strength to tolerate undesirable overloads.”⁶ Studies have shown that “the cementation mode did not significantly influence the occurrence of complications” when comparing RMGI and self-etching resin cement after 9 years of service.⁷

It is, therefore, recommended that lithium-disilicate full-coverage restorations involving preparations with 4 to 10 degrees of taper⁸ and 4 mm of axial wall height be retained cohesively with a cement such as a RMGI. This strategy does not compromise the strength or retention of the crown and offers a distinct advantage should the crown require removal. Clinicians should take an approach that will not only provide adequate retention but also facilitate removal when and if it becomes necessary. Excessive removal of tooth structure due to the inability to discriminate between crown and tooth with translucent resin cement is one of the most common challenges when removing lithium disilicate (Figure 1). The use of a RMGI allows for safer future removal through sufficient but not excessive adhesion and clear demarcation of the cement line.

When a self-etching resin cement and RMGI cement were compared on preparations with ideal resistance and retention form, it was found that the main difference between the cement types was the mode of failure when excessive retention forces were applied. Most self-etching resin-cemented lithium-disilicate crowns failed due to fracture, while the glass ionomer-cemented crowns could be removed yet were still intact.⁹ This is evidence of a potential danger when removing these crowns with traditional techniques such as using a crown spreader. If this danger during removal can be minimized while still using a cement of sufficient strength adequate for restoration retention, then a weaker cement should be considered.

Clinical judgment should be used to assess whether the risk for eventual removal of the restoration will outweigh the risk for the restoration becoming displaced. Re-cementing a restoration is preferable rather than removing and replacing a fractured restoration. In cases in which adequate retention is not present and remaining enamel is insufficient for a bonded restoration, the clinician should consider using self-etching resin cement for increased retention (Figure 2).

To reiterate, for full-coverage restorations, if adequate retention and resistance features are present, many cements can be used, including RMGI. In cases with minimal retention, self-etching resin cements are recommended. Inlays and onlays that are adhesively retained should, of course, be bonded with a dual-cure resin cement utilizing a complete etch technique.

Utilizing Diagnostic Data Preoperatively

When planning the effective and safe removal of restorations, the type of restorative material must be determined. Radiographically, zirconia is highly radiopaque and resembles a gold crown, while lithium disilicate is radiolucent resembling tooth structure. Because the monocrystalline material of zirconia is not bondable, it can be assumed that zirconia crowns are cohesively cemented. With lithium disilicate, this is not as easy to determine preoperatively. If dental records describing the cementation protocol are not available, it should be assumed that all restorations are bonded until otherwise determined, through either depth cuts or evaluation of the underlying tooth structure on the radiograph. At this point, a decision on removal technique between laser and bur should be made.

Removal Using Lasers

Since the early 1990s, erbium lasers have been used experimentally for debonding orthodontic ceramic brackets.^10-18 Laser application significantly decreased the required load for debonding porcelain laminate veneers—by nine-fold—when compared to control groups.¹⁸

In Figure 3, an Er:YAG laser (Waterlase iPlus, Biolase, biolase.com) was used to remove full-coverage porcelain crowns. One minute of ablation was performed on the facial and 1 minute on the lingual from a distance of approximately 1 mm using a scanning motion. This was done with the MZ6 tip (Biolase) at 5 watts, 15 Hz (adjustable 5 Hz to 100 Hz), water at 80%, air at 80%, and 600 mJs. In one study on extracted teeth the removal time for a veneer averaged 106 seconds.¹⁹

In the present clinical case, after using the laser one author (AS) utilized a Wedelstaedt chisel under the facial margin, applying slight hand pressure in the incisal direction. Most of the crowns were fully intact when removed (Figure 4). No tooth structure was damaged, and a novice user easily removed each crown in less than 3 minutes (Figure 5). The cement remained on the natural tooth (Figure 6), indicating that the bond was broken between the silanated layer of the crown and the cement.²⁰ This process can take longer depending on the thickness of the material. If necessary, the restoration thickness can be reduced with a bur prior to laser application.

Another option for removing thicker restorations is to increase the pulse range, as measured in hertz. During laser treatment the tooth may temporarily become slightly dark, which does not indicate damage to the tooth. In a study, light microscopy revealed no damage to the tooth due to the laser debonding of the veneer.²⁰ The use of an Er:YAG laser results in a linear increase in the pulpal temperature but does not cause a temperature increase above the limit considered safe for the pulp vitality.^21,22

The debonding of veneers is accomplished by the H2O/O2 absorption band, which coincides with the emission wavelength of an Er:YAG laser. All-ceramic restorations are unable to absorb that wavelength, and the laser energy is therefore transmitted through the veneer, ablating the bonding cement. Most veneers can be removed whole, which is particularly helpful in the event of a cementation error. Flexural strength appears to be the key feature to maintaining the integrity of the veneer upon laser removal,²² thus a lithium-disilicate veneer would be expected to be more resilient during this type of removal. The cost of an Er:YAG laser does, however, create a financial barrier, which is not an issue with bur removal.

Removal Using Conventional Means

Ample evidence suggests that bur selection can have a substantial effect on cutting efficiency, heat generation, and crack propagation in ceramic restorations.^23-25 When attempting to preserve fracture resistance, as in the case of occlusal adjustment or endodontic access, a 30-mm-grit or finer diamond has been recommended as the instrument of choice because it does not dramatically increase the risk for fracture.²⁷ However, when attempting total restoration removal, preservation of fracture resistance is not a goal and is logically counterproductive. In such a case, a coarse (125 mm) or extra-coarse (150 mm) diamond would be preferable due to a more efficient cutting rate. This improved cutting efficiency is tempered by the increased heat generation potential with extra-coarse grit diamonds, which can be mitigated with light touch and copious water irrigation23 (Figure 7). The use of electric handpieces, which offer higher torque and smoother cutting, has also been proposed to improve cutting efficiency and minimize heat generation.²⁵

Recently, several burs have been introduced specifically for use on high-strength ceramics. These feature various improvements such as more durable adhesion methods or surface coatings to prevent diamond loss and minimize debris buildup between diamond crystals. Anecdotal reports and the authors’ experiences suggest that these burs appear more resilient when removing lithium disilicate or composite,²⁶ though independent scientific study in this area is lacking. These specialized burs are generally limited to two grits of either coarse or fine, and three shapes of round, football, and tapered chamfer. The authors suggest that whether using single-use, traditional, or specialized burs, a fine-grit round bur is optimal for endodontic or implant screw access, a fine football for occlusal adjustment, and a coarse tapered chamfer diamond for crown removal. As with comparable traditional burs, these specialized diamonds should be used with light touch and copious water spray to prevent debris accumulation and associated heat generation. Generally, these burs are two to three times more costly than traditional diamonds, so practitioners must decide if the cost-benefit ratio favors their use versus traditional coarse diamond burs.

The technique proposed for removal of high-strength ceramic crowns luted with traditional cements is similar to the process for removal of metal or metal-ceramic crowns. Sectioning the crown to the cement line in a buccal-lingual direction produces two pieces with minimal resistance form to withstand the mesial-distal force from a crown separator (Figure 8 and Figure 9). Often, complete sectioning is not necessary, as this force is capable of propagating a fracture along the remaining ceramic (Figure 10). When this occurs, a straight restorative chisel may be placed at the cement line with controlled force directed along the axial wall (Figure 11). This concentration of force can often serve to induce adhesive or cohesive failure of the luting agent and dislodgement of a fragment of the restoration.

If the clinician suspects a ceramic is adhesively bonded to enamel, the ceramic should be removed entirely with a coarse diamond. Using a crown remover to separate the ceramic–enamel interface can often produce a high risk for tooth fracture due to the high bond strength.

Use of Zirconium-Oxide Restorations in Contemporary Dentistry

Without question, full-contour zirconia restorations have become a staple of many restorative practices. By some estimates these account for nearly half of all indirectly fabricated restorations placed today.²⁸ The high flexural strength of zirconia (1100 MPa to 1400 MPa) affords the clinician additional confidence, especially in cases of posterior teeth with short clinical crowns. These short-crown situations often lead clinicians to accept compromises in occlusal material thickness and/or axial wall height required for retention and resistance when using traditional luting cements. Currently, limited shades and translucencies are available, generally relegating these restorations to posterior areas that would previously have been restored with metal castings or metal-ceramics.

The widespread acceptance of zirconia restorations is understandable, because zirconia remains the only commercially viable ceramic material with a flexural strength rivaling that of metal-ceramics. But how strong is strong enough? A growing body of long-term data demonstrates the high clinical survival rate of bonded and traditionally luted lithium-disilicate restorations,²⁹ yet lithium disilicate possesses a flexural strength of only 350 MPa to 400 MPa, one-third that of zirconia. Because lithium disilicate contains etchable glass, it can be predictably adhesively bonded, augmenting insufficient mechanical retention. Zirconia’s monocrystalline structure does not contain silica, and is, therefore, not etchable.

Another key limitation of full-contour zirconia restorations is esthetics, as it is monolithic and relatively opaque. Although manufacturers continue to develop more translucent zirconia materials, these generally exhibit lower flexural strength (700 MPa to 900 MPa) and, notably, lack the same long-term clinical support as traditional opaque zirconia materials. If patient demand is swinging the pendulum in favor of more esthetic materials in the posterior, and high-strength ceramics may often require future replacement before they have complete loss of retention, the choices clinicians make today could greatly impact dentistry’s efforts to conservatively care for patients’ teeth in the future.

As previously mentioned, the monocrystalline structure of zirconia is not etchable. Though specialized cleansers, primers, and techniques exist to improve bonding to this material, the evidence does not appear to support the concept that these can compensate for inadequate mechanical retention. Therefore, in the case of inadequate clinical crown height on posterior teeth, clinicians may find themselves making the decision to provide minimal or inadequate occlusal clearance, removing all existing cervical enamel and placing crown margins deep subgingivally in order to approach minimally acceptable restoration retention form. Clinicians may attempt to supplement this inadequate or minimally adequate retention with bonding, even when factors such as moisture control and remaining enamel are unfavorable.

A potential alternative in these clinical situations is use of an adequately reduced bonded ceramic onlay restoration with more accessible enamel finish lines. Where applicable, this option has the potential benefits of being more conservative of tooth structure, providing more cleansable margin placement on enamel that allows for improved moisture control by rubber dam or other means, more predictable adhesive bonding, improved translucency and esthetics, and easier eventual removal.

Conclusions

The authors determined lasers to be the most conservative and efficient removal method for lithium disilicate and other glass-ceramics, particularly in the anterior. This technique is specific to erbium, not diode, lasers and allows bonded or traditionally cemented ceramic restorations to be removed efficiently without damaging the underlying tooth structure. However, the high cost of these laser systems may require most clinicians to continue to use rotary instruments. If this is the case, the use of a coarse-grit diamond with light touch and copious water spray is preferable, and an electric handpiece could be considered to improve efficiency and minimize heat generation (Figure 12).

If adequate retention and resistance form is present, a traditional RMGI cement would be preferable because of its moisture tolerance, easy cleanup, and more predictable removal.

About The Authors

Andrew Spath, DDS, FAGD
Lecturer
University of California
Los Angeles, California
Private Practice
Newport Beach, California
Mentor
Kois Center
Seattle, Washington

Colby Smith, DDS, FAGD
Director
AEGD Venice Dental Center
University of California
Los Angeles, California

References

1. Christensen G. The future: materials, challenges in dentistry, and education. Interview by Dr. Damon Adams. Dent Today. 2012;31(2):102-106.

2. Reason for replacement of restorations. Operative Dentistry. 2005;30(4):409-416.

3. Sarrett DC, ed. A laboratory evaluation of electric handpiece temperature and the associated risk of burns. ADA Professional Product Review. 2014;9(2):18-24.

4. Elias K, Amis AA, Setchell DJ. The magnitude of cutting forces at high speed. J Prosthet Dent. 2003;89(3):286-281.

5. Culp L, McLaren EA. Lithium disilicate: the restorative material of multiple options. Compend Contin Educ Dent. 2010;31(9):716-725.

6. Steiner M, Sasse M, Kern M. Fracture resistance of all-ceramic crown systems [abstract]. J Dent Res. 2011;90(spec iss A):Abstract 2999.

7. Gehrt M, Wolfart S, Rafai N, et al. Clinical results of lithium-disilicate crowns after up to 9 years of service. Clin Oral Investig. 2013;17(1):275-284.

8. Malone WFP, Koth DL, Cavazos E, et al. Tylman’s Theory and Practice of Fixed Prosthodontics. 8th ed. St Louis, MI: Ishiyaku Euro-America; 1989.

9. Mobilio N, Fasiol A, Mollica F, Catapano S. Effect of different luting agents on the retention of lithium disilicate ceramic crowns. Materials. 2015;8(4):1604-1611.

10. Tocchio RM, Williams PT, Mayer FJ, Standing KG. Laser debonding of ceramic orthodontic brackets. Am J Orthod Dentofacial Orthop. 1993;103(2):155-162.

11. Strobl K, Bahns TL, Willham L, et al. Laser-aided debonding of orthodontic ceramic brackets. Am J Orthod Dentofacial Orthop. 1992;101(2):152-158.

12. Obata A, Tsumura T, Niwa K, et al. Super pulse CO2 laser for bracket bonding and debonding. Eur J Orthod. 1999;21(2):193-198.

13. Oztoprak MO, Nalbantgil D, Erdem AS, et al. Debonding of ceramic brackets by a new scanning laser method. Am J Orthod Dentofacial Orthop. 2010;138(2):195-200.

14. Nalbantgil D, Oztoprak MO, Tozlu M, Arun T. Effects of different application durations of ER:YAG laser on intrapulpal temperature change during debonding. Lasers Med Sci. 2011;26(6):735-740.

15. Dostalova T, Jelinkova H, Sulc J, et al. Ceramic bracket debonding by Tm:YAP laser irradiation. Photomed Laser Surg. 2011;29(7):477-484.

16. Feldon PJ, Murray PE, Burch JG, et al. Diode laser debonding of ceramic brackets. Am J

17. Tehranchi A, Fekrazad R, Zafar M, et al. Evaluation of the effects of CO2 laser on debonding of orthodontics porcelain brackets vs. the conventional method. Lasers Med Sci. 2011;26(5):563-567.

18. Sarp AS, Gülsoy M. Ceramic bracket debonding with ytterbium fiber laser. Lasers Med Sci. 2011;26(5):577-584.

19. Oztoprak MO, Tozlu M, Iseri U, et al. Effects of different application durations of scanning laser method on debonding strength of laminate veneers.Lasers Med Sci. 2012;27(4):713-716.

20. Morford CK, Buu NC, Rechmann BM, et al. Er:YAG laser debonding of porcelain veneers. Lasers Surg Med. 2011;43(10):965-974.

21. Nalbantgil D, Oztoprak MO, Tozlu M, Arun T. Effects of different application durations of ER:YAG laser on intrapulpal temperature change during debonding. Lasers Med Sci. 2011;26(6):735-740.

22. Correa-Afonso AM, Pecora JD, Palma-Dibb RG. Influence of pulse repetition rate on temperature rise and working time during composite filling removal with the Er:YAG laser. Photomed Laser Surg. 2008;26(3):221-225.

23. Nakamura K, Katsuda Y, Ankyu S, et al. Cutting efficiency of diamond burs operated with electric high-speed dental handpiece on zirconia. Eur J Oral Sci. 2015. doi:10.1111/eos.12211.

24. Siegel SC, von Fraunhofer JA. Cutting efficiency of three diamond bur grit sizes. J Am Dent Assoc. 2000;131(12):1706-1710.

25. Ottl P, Lauer HC. Temperature response in the pulpal chamber during ultrahigh-speed tooth preparation with diamond burs of different grit. J Prosthet Dent. 1998;80(1):12-19.

26. Ercoli C, Rotella M, Funkenbusch PD, et al. In vitro comparison of the cutting efficiency and temperature production of ten different rotary cutting instruments. Part II: electric handpiece and comparison with turbine. J Prosthet Dent. 2009;101(5):319-331.

27. Giordano R, Cima M, Pober R. Effect of surface finish on the flexural strength of feldspathic and aluminous dental ceramics. Int J Prosthodont. 1995;8(4):311-319.

28. LMT 2014 Digital Technology Survey. Lab Management Today. https://s3.amazonaws.com/voxle.net/lmt/assets/53692da71d505.pdf. 2014. Accessed September 20, 2016.

29. Simeone P, Gracis S. Eleven-year retrospective survival study of 275 veneered lithium disilicate single crowns. Int J Periodontics Restorative Dent. 2015;35(5):685-694.