Finishing of MGC Dicor Material

Finishing ofMGC Dicor

Material

Ann C. Shearer, MSc, BDS, FDSRCS Ed'

Robert P. Kusy. BS, MS, PhD"

lohn Q. Whittey, BS"

Harald O. Heymann, DDS, MEd'"

Nairn H.E. Wiison, MSc, PhD, BDS, FDS,DRD RCS Ed'

The occljsal surfaces of inlays and onlays milled by the CEREC system areessentially flat, and surface morphology must be produced with rotaryinstruments. Six methods of finishing a machinable glass-ceramic material(Dicor MCC¡ were investigated using laser specular reflectance to assess therelative surface smoothness. The results indicated that all the methodsproduced a similarly smooth surface, Int ¡ Prosthodont 1994,7:167-173.

T he CEREC system (Siemens Dental, Charlotte,NC) is a mobile, chairside CAD/CAM (com-

puter-aided design/computer-aided manufacture)device designed to produce ceramic inlays, onlays,and veneers. An optical impression of the prepara-tion is made, and the required restoration is pro-duce'd from a single premanufactured dentalceramic block-' The ceramic blocks currentlyavailable are a feldspathic porcelain {VitablocsMark II, Vita Zahnfabrik, Bad Sackingen, Cermany)and a machinable glass-ceramic material (DicorMCC, LD Caulk, Dentsply, Milford, DE), The exter-nal surfaces of the restoration are contoured andfinished after placement, as these surfaces are notmilled by the CEREC system. The occlusal surfaceproduced is essentially flat, and surface morphol-ogy must be carved using rotary instruments,Kotschy- has described a diamond bur with con-cave surfaces to facilitate reproduction of the fis-sure pattern and convex surfaces,^

Several methods have been suggested for finish-ing CEREC restorations to optimise the surfacefinish. Mormann et aP and Smith and Cardwell"'

'Department of Restorative Dentistry, University Dental Hos-pital of Manchester.

"Dental Research Center, University of North Carolina."'Department of Operative Dentistry. University of North

Carolina.

Reprint requests: A.C. Shearer, Department of RestorativeDentistry, University Dentai Hospital of Manchester, HigherCambridge Street. Manchester Ml5 6FH, United Kingdom.

recommend the use of microfine diamonds fol-lowed by resin composite finishing discs, Nash'also recommends diamond points, but followed bydiamond pasle, Stoll and Stachniss^ added 30-flutecarbide burs to this sequence after the use of dia-mond points and before polishing using the dia-mond paste, Kunzelmann and Hickel" describedthe use of aluminum oxide discs and diamondpaste, Martin and ]ednakiewicz' studied severalmethods of finishing Vita-CEREC porcelain using asurfometer to assess surface roughness. They con-cluded that the best surface finish was achievedusing, in sequence, fine diamonds. Baker Cursonburs (tungsten carbide stones), and diamond pol-ishing paste. Bur kits are now available commer-cially for preparing, contouring, and finishingCEREC restorations (Carolina CAD/CAM ceramicinlay kit, Brasseler, Savannah, CA),

The purpose of this study was to evaluate differ-ent techniques advocated for finishing CERECrestorations made from Dicor MCC blocks. Laserspecular reflectance (LSR) was used to assess therelative surface smoothness and to compare thefinish achieved with that of conventional dentalporcelain.

Materials and Methods

The finishing sequences investigated are shownin Table 1, These sequences were selected toreflect methods advocated in the literature and fin-ishing systems marketed specifically for use with

á7, fliimber 2, 1994 167 l of Prostiiodonlics

d MGC Dicor Materia)

Table 1 Finishing SequencesMethod A30 biade tungsten carbideShofu porceiain points (abrasive impregnatedsilicone rubber) standard, uitra, ultra II

Truiuster diamond pasteMethod BTwo Slriper MFS microfine diamonds80)im, 50|im, 15|j.mShofu porcelain points (standard, ultra, ultra li)Truiuster diamond pasteMetriod C30 blade tungsten carbideShcfu porcelain points (standard, ultra, ultra li)Two Striper diamond poiisriirig systemMethod DBaker Curson burTruiuster diamond pasteMethod ETwo Striper MFS microfine diamonds80M.m, 40!im, 15M.m3M Soflex discs (medium, fine, extra fire)Method F30 biade tungsten carbideShofu porcelain points (standard, ultra, ultra il)Tnjiuster diamond paste

High-speed (0,2 MPa) dry

Slow-speed (0,07 MPa) wet

High-speed wet

Kerr/Sybron, Romulus, Ml

Shofu, MenIo Parit, CABrasseler, Savannah, GA

Premier, Norristown, PA

Slow-speed wet

High-speed drySlew-speed wet

High-speed wet

High-speed wetModerate-speed (0.1 MPA) dry

High-speed drySiow-speed wet

ShofuBrasseler

KerrStiofuPremier

Ash, London, UKBrasseler

Premier3M,St, Paul, MN

BrasselerShofuBrasseler

CEREC ceramic blocks. All burs were used accord-ing to the manufacturer's instrtjctions. Eighteendark shade Dicor MGC blocks were used. To simu-late the contouring of the occlusal and approximalsurfaces, which are not fully milled by the CERECsystem, a diamond bur (100 (im) was applied to allthe blocks. The burs were held in a friction grip,high-speed handpiece (Mid-west Quiet-air ,Midwest div, Sybron, Des Plaines, IL) and used atmoderate speed and pressure (0,1 MPa) withcoolant. Each bur was applied to the surface for 30seconds in one direction using light, intermittentpressure.

Each of the six finishing sequences shown inTable 1 was applied to three blocks. To ensureequal treatment of the blocks, the burs wereapplied to the surface in one direction only, usinglight, intermittent pressure. Diamond and tungstencarbide burs were used until the ceramic surfaceappeared uniform and no further improvement wasapparent. Diamond pastes were applied using asoft bristle brush (Robinson bristle brush, Brasseler.Savannah, CA) for 20 seconds at moderate speedand pressure (0,1 mPa), One operator (ACS) pre-pared all the samples.

After an initial alcohol wipe using lens tissuepaper, each block was evaluated by LSR usinga helium-neon laser (Metrologie Instruments,Bellmawr, N|) with a 0,6328-)im (W wavelengthand 0,6-mm beam diameler. The relative smooth-ness was determined by measuring the reflectedpower of each sample (I.) in microwatts (|J.W) and

comparing it wifh fhe intensity of the incidentbeam (In), A photodefector with a 2-mm-diameteraperlure was used to detect the reflected light.Prior to each run, the initial power of the laser wasverified by placing the photodetector coaxial to thebeam. The value I. was determined from Ihe maxi-mum value on the spectral map of the reflectedlight beam as the photodetector was rotatedthrough scan angles (6) of ±10 degrees. Laser spec-tra were generated for incident angles (a) of 76,78, 80, 82, and 84 degrees. Four scans of eachsample were made by rotating the blocks throughangles (y) 0, 90, 180, and 270 degrees (Fig 1), Thiswas carried out to assess whether the direction ofbur application influenced the resulf.

The roof mean squared (RMS) optical roughness[Go] is determined in units of |im by plotting thelogio of the relative specular power {lA,} against thecosine squared of the angle of incidence (cos-a).Erom the slope of each line that equaled (8,276oJX)-, the absolute magnitude of each o« was cal-culated (Eig 2), The gentler the slope of the plof,tbe smoother the surface and lower the a» relativelo another.

Data Analysis

Eor each position and sample, fhe correlationcoefficient and statistical significance were deter-mined after regressing five incident angles (a) on alogiü (lyiü) versus cos'a plot. Thereafter, the Oo val-ues of tbe 12 position/sample combinations of each

I of Proithodonlii 168

Finishmg oí MGC Dicor Material

Fig 1 Representation ot thelaser scans at a = 80 degrees,obtained on a block ot DicorMGC prepared by mettiod C (y =0, 90,180, and 270 degrees).

150

, 100

50

+10 0 -10 0 +10Scan Angle, e n 200

II7=0°¿100

90"—f

0 -10 0 +10Scan Angle, Gf) 2OO Scan AngiB,

+10 0 -10 0 +10Scan Angle, e n

Fig 2 Plot ot the log relative power against cos'a (76degrees < a < 84 degrees) tor three blocks of Dicor preparedby method C (y = 0 degrees).

method were displayed using the box plot as ananalysis graph." This statistical approach is superi-or to any summary graph that shows only meansand standard deviations (SD), because the box plotmakes no assumption about whether the data dis-tribution is Gaussian, Poisson, Weibull, etc. In fact,information regarding tbe central tendency, range,and dispersion of tbe data can be gleaned fromthese corresponding observations: (a) whether themedian line is located in the middle of the box; (b)whetber the box that bounds 50% of the data islarge; and (c) whether the fence (vertical line] that

represents 1.5 times the box is large and whetherany outliers (asterisks) that represent from 1.5 to3.0 times the box are present. In addition to thebox plot, Student's t tests were used to comparethe six methods, while pairwise (tests were used tocompare the methods for each position after thesamples for each position were averaged. AnANOVA was used to assess the main effects ofmethod and position as weil as their interaction.

Results

The results, expressed as RMS optical roughness(a,,), are displayed for each metbod as a function ofposition and sample (Table 2). The correlationcoefficients, from which each Co was obtained,were highly significant (P < .01 ) in 57 of 72 deter-minations and not significant in 6 of 72 determina-tions. From the ANOVA, no significant interactionwas noted, but the main effects of method andposition were weakly significant at P < .05 and P =.05, respectively. From the box plot of these data(Fig 3), however, no difference is apparentbetween the methods; all the boxes that bounded50% of the data sets partially overlapped oneanother. From f tests, only two significant differ-ences were found among the 15 combinations ofmethods; between methods A and E (P< .02] andbetween methods B and F (P < .02). Since the I testpresumes an ideal Gaussian distribution and thebox plot graphically shows that the data of meth-ods B and F are skewed towards one another, the í

7, Number 2, 1994 169 al of Proítliodontics

Table 2 Summary ol RMS Optical Roughness ((r=)as a Function of Positioning

RMS i^m) +

Position (-y) Sample 1 Sample 2 Sample 3

090'180"270'0*90'18D*270"0*90'tao"270'0"90'180"270"0 '90 '180*270*0 '90°180'270'

0.2410.2560.200*0.2060.353-0.2190.1840.363-0.2060.2000.2020.2290.2250.2550.275*0.1560.1820.2220.2190.2410.2430.143"0.2260.138'-

0.3040.203'01990 3350.2260.2260.3790.2400.2280.1840.2290.1730.2600.1820.229'0.2560.097—0.2220.120-0.235-0.3290.175'"0.2700.188"*

0.2820.1710.2580 2600.2400.202'0.2170.2020.3320.1790.2090.1910149Ci.2290.2350.22601710.1320.2190.1940.2940.120"*0.1970.208

+ For the number of data points evaluated, the correlation coetticiehts were sjch that the linear regression lineswere highly sighiticant P< .01, except ', P< 05; ", P= 05;'". hot significant.

Fig 3 Box plat showing RMS (root mean squared) opticalroughness (aa) tor each finishing sequence.

test overstates the statistical significances. Whenthe six methods versus the sample average of thereplicates are evaluated pairwise for each position,only in 1 out of 60 paired methods (methods B andF at '30 degrees) is a significant difference observed

ÍP < .02). Once convinced that no consistent sig-nificant differences exist, it can be observed fromthe box plot that method C produced a generallytighter grouping with a mean comparable to thatfor methods E and F (the lowest groups] togetherwith the smallest fence Icf Data Analysis). Assum-ing that operator errors were evenly distributed,method C appears to be the most reproducible andreliable finishing sequence.

Discussion

Laser specular reflectance is a noncontactmethod of measuring relatively smooth surfacesthat has previously been used in dentistry to inves-tigate finishing of porcelain, porcelain veneers, andorthodontic arch wires.'"" In this technique, theangle and intensity of the reflected laser light iscompared with the incident angle and intensity.From these optical results, surface roughnessesmay be evaluated. Laser specular reflectanceassumes that a perfectly smooth and nonabsorbingsurface will reflect light at an angle (ß) that equalsthe incident angle (a) and with an intensity equalto that of the incident beam (Fig 4). Consequently,any loss of intensity or angular shift indicates someform of surface roughness, providing internal re-flections or refractions arc absent.

I of Pros thod on lies 170

Finishing of MCC Oicor Maten

Fig 4 The laser specularreflectance apparatus.

Shutter

Aperture

Sample

Table 3 Comparison of Porcelain Polishing Techniques by Laser Specular Reflectance

Present work*

P revio tJSwork"

Method

ABCDEFMAP

Low

13,410,930,28,3

33.121,013.76.40,1

Intensity, 1.

Range

High

21,932,432,416,452,250,727,510,80,4

Mean

18.81 9 331.312,945,637,320,9

8,B0,2

Standarddeviation (SD)

3,99.40-93,48,9

12,36,11,80,1

Sample size(n)

333333222

•Calculated from Ihe estimated I/L values of regression lines al a = 7Q' using the mean value ot I, = 548|JW (SD =9 3iiW and n = 2B), which were measured before and after tlie samples were evaluated,-Measured directly at a = 70' {Haywood et al")M = Meclianically tinislied porcelain using line and microline diamonds, 30 tlute tungslen carbide bur and diamondpolishihg paste; A = Auloglazed porcelain; P = Diamond polishing paste alone

The comparison between the results of this studyand those of a previous study that used LSR toassess the finishing of conventional porcelain isshown in Table 3,'" These results indicate thatDicor MCC finished using methods A, B, C, E, or Fis smoother than conventional porcelain that hasbeen mechanically finished, autoglazed, or fin-ished using diamond paste alone.

The results of this study indicate that any of thesix sequences used will give an acceptable finish

to a machinable glass-ceramic material (DicorMGCl- This agrees with the finding of the study byMartin and )ednakiewicz/ who compared sevensequences for finishing Vita-CEREC porcelain andfound several systems produced a fine surface fin-ish. These workers used a surfometer, whereas inthis study, the porcelain surfaces were investigatedusing specular reflectance, a method best suited forstudying relatively smooth surfaces.

The surface produced on the Dicor MCC also

á•7, Number 2. 1994 171 I of Frosthodonlit

iinfi ai MCC Dic:or Material

compares well with finished conventional por-celain and is smoother than that of autoglazedporcelain,'"

Although none of the finishing sequences stud-ied consistently gave statistically different results interms of surface roughness (Eig 3), method C {30-flute tungsten carbide bur, Shofu porcelain points.Two Striper diamond polishing system) gave atighter group of data with limited spread. MethodC differed from method A only in the final compo-nent. In method A, Truluster diamond paste wasused, whereas in method C the Two Striper dia-mond polishing system was the final component.The diamond polishing system consists of two dia-mond crystal gels, with particle sizes of 4 to 6 (imand < 0.5 ¡am, used sequentially. These resuitswould suggest that this system may give a betterfinish than conventional diamond paste.

Methods E and F gave the lowest a« values.Method E (Two Striper MFS microfine diamonds,Soflex discs) is the method recommended byMormann,' Discs are often considered to be moresuitable for finishing smooth exiernal surfaces ofteeth, whereas porcelain points and diamond pasteon a brush or cup (as in method F) may be moreappropriate for finishing tbe fissures of occlusalsurfaces.

Method F differs from method A only in themanufacturer of the 30-blade tungsten carbide bur.Although no statistical difference exists betweenthe finishing sequences, method A was measuredas producing the roughest surface and method Fthe smoothest. Either there is a difference in thesurface geometry of the burs or this indicates thelimitations in the finishing method.

Method D (Baker Curson bur, Truluster diamondpaste) is similar to the regime that Martin andJednakiewicz" found to give the finest surface toVita-CEREC porcelain (fine diamond [25 fim),Baker Curson bur, diamond paste). Baker Cursonburs have the advantage that they can be shaped tothe desired profile and resurfaced as required byrunning the bur in an air turbine handpiece againsta fine diamond point.

Method B (Two Striper MES diamonds, Shofuporcelain points, Truluster diamond paste) involvesmore bur changes than any of the other regimesused, and this may prove time-consuming-

DeLong et al,'- Jacobi et al, ' ' and Seghi et a l "have suggested that porcelain finished with abra-sive systems is smoother and results in less wear toopposing tooth structure than glazed porcelain. AsCEREC restorations are unglazed and mechanicallyfinished, this process would appear not to con-traindicate their clinical use. The concern about

unglazed restorations has mainly been the wear ofthe opposing enamel. The results of this and otherstudies would seem to suggest that this should notbe of concern if a suitable finishing sequence isfollowed. However, the effect of trying to finish acontoured restoration may be different from finish-ing a block of porcelain. In the clinical situation,restorations attempt to mimic anatomic form withpits and fissures. These areas may not be reachedby rotary instruments, and the porcelain would befinished with paste alone.

Small differences between finishing techniquesmay be of little consequence clinically, as all thesurfaces produced may look the same following ashort period of function.

Ideally, the finishing regime should have, as aminimum criterion, a systematic decrease in theparticle size of the abrasive systems used. As longas this guideline is followed, the choice of finishinginstruments may be one of personal preference.

Conclusions

Six methods of finishing Dicor MCC blocks wereevaluated. Within the limitations of the small sam-ple size and Ihe use of samples having flat sur-faces, the following conclusions may be made:

1, No significant difference was found betweenthe surface finishes obtained using any of thesix methods,

2, All the finishing sequences used produced asmooth surface,

3, Dicor MGC finished using any of tbe six meth-ods is smoother than conventional porcelainthat has been mechanically finished, auto-glazed, or finished using diamond paste alone.

Acknowledgments

The authors wisii to thank the Dental Research Center at theUniversity oi North Carolina for the use of the laser specuiarreflectance apparatus.

References

1. Leinfelder KF, Isenberg BP, Essig ME, A new method forgenerating ceramic restorations: A CAD-CAM systerr, J AmDent Assoc 1989:118:703-707.

; Kotschy P Bessere Ausarbeitung der naturlichen Kauílach-enanatorrie durch neue Diamentenforni, QuintezzZahnärtzl 1980:31:29-32,

3 Sioli R, Stachniss V. Adiusting and finishing the occluraimorphology of Cerac restorations using rotating instru-ments, in: Mormann WH (edi. International Symposium onComputer Restorations. State of the Art of the CeracMethod, Berlin: Quintessence, 1991: 117-125.

i ol Prosthodortii 172

Finishing of MGC Oicoi Material

4, Mormann WH, Brandestini M, Lutz F. Barbakow F, Chair.side computer-aided direct ceramic inlays QuintessenceInt l989;2O:329-339.

5, Smith BGN, Cardweil ]E. Ore visit ceramic restorationsmade at the chairside: The Cerac machine. Rest Dent198');S:6Ü-65,

6, Nash RW, Integrating CAD/CAM into the general practice,Co5metic Dent Gen Pratt 1991 ;4(9):1-4,

7, Kunzelmann KH, Hickel R. Feinpoiitur von Cerac-lniaysmit Diamantpoliersystemen. Dtsch Zahnarzti Z I59O;45:680-682.

8, Martin N, lednakiewicz N. Surface analysis of Ceracrestorations iinished by different techniques, in: MormannWH (edl. International Symposium on Computer Restor-ations. State of the Art of the Cerac Method. Beriin:Quintessence, 1991:469^79,

9, Simpson RS, Johnson TA, Amara IA. The box-plot: An

exploratory analysis graph for biomédical publications.Am Heart I 1 988;11 6:1 663-166S,

10, Haywood VB, Heymann HO, Kusy RP, Whitley |Q,Andreaus SB Polishing porceiain veneeis. An SEM andspecjiar reflectance analysis. Dent Mater 1988;4:n6-1 21,

11, Kusy RP, Whitley |Q, Mayhew MJ, Buckthai IE, Surfaceroughness of orthodontic archwires via laser spectroscopy.Angle Orthod 1988;58:33^5,

12, DeLong R, Pintado MR, Douglas WH, The wear of enamelopposing shaded ceramic restorative materials: An in vitrostudy, I Prosthet Dent ¡992;G8:42-48,

13, lacobi R, 5hillinebjig HT, Duncanson MG. A comparisonof [he abrasiveness of si \ ceramic surfaces and gold, |Prosthet Dent 1991;66:303-309,

14, Seghi RR, Rosenstiel SI-, Bauer P. Abrasion of humanenamei by different dental ceramics in vitro. I Dent Resl991;;0:221-225.

On Clinical Loading of Osseottitegrated Implants. A Methodological and Clinical Study.

The extent and consequence of force transmission through prosthet icaily loaded osseo integrated implants Losupporting bone are not fully understood. Acceptable ioad iimits are not known, and the possibiiity of stress

tt th b l d i l i f i l i t Th t i i i d b h i nd

lone test subject. The maxiilary dentition was restored using conventional fixed prosthetics, while sixImplants supported a hybrid prosthesis in the edentulous mandible. Four strain gauges were attached to fiof the implant abutment cylinders. In vitro evaluation addressed both pure axial loading as weii as theappii-cation of bending force moments, in vivo data Collection occurred during the following clinical maneuvers:II) maxima i occlusal force dunng normal closure for a 10-second duration; (21 maximal occiusal force on aforce gauge; and 13} mastication of test food samples, Resuits demonstrated differences between functionalloads recorded during the in vivo tests used, as weli as significant variabiiity between in vivo and in vitroexperimental situations. Considerable bending of the abutment cyiinders was recorded in a i most ali experi-ment conditions, including during maximai occiusal force with normal closure in or ciose to tentric reiationocciusion. Results clearly demonstrate strain gauges to be useful in the study ot clinical implant loadingand the need for the use of in vivo, not in vitro, e^iperimental conditions to ascertain valid biomechanicalinformation.

Glantz, P-O, et al. Clin Oral ImpI Res 1993;4:99-105, References: 14. Reprints; P.-O Glantz, Department of ProstheticDentistry, FaculLy of Odontology, Univeriky of Lund, S-214 21 Mälmo, Sweden -David R. Cagna, DMD, Depanmeni oiProslhodontics. The Universily oí Tejas Hej/íh Science Center at San Antonio, San Antonio. Tenas

; 7, Number 2, 1994 173