In vitro marginal fit of three all-ceramic crown systems

In-Sung Yeo, DDS, MSD,a Jae-Ho Yang, DDS, MSD, PhD,b and Jai-Bong Lee, DDS, MSD, PhDc
Department of Prosthodontics, College of Dentistry, Seoul National University, Seoul, Korea
Statement of problem. Studies on marginal discrepancies of single restorations using various systems and
materials have resulted in statistical inferences that are ambiguous because of small sample sizes and limited
numbers of measurements per specimen.
Purpose. The purpose of this study was to compare the marginal adaptation of single anterior restorations made
using different systems.
Material and methods. The in vitro marginal discrepancies of 3 different all-ceramic crown systems (Celay
In-Ceram, conventional In-Ceram, and IPS Empress 2 layering technique), and a control group of metal ceramic
restorations were evaluated and compared by measuring the gap dimension between the crowns and the prepared
tooth at the marginal opening. The crowns were made for 1 extracted maxillary central incisor prepared with a
1-mm shoulder margin and 6-degree tapered walls by milling. Thirty crowns per system were fabricated. Crown
measurements were recorded with an optical microscope, with an accuracy of 0.1 m, at 50 points spaced
approximately 400 m along the circumferential margin. The criterion of 120 m was used as the maximum
clinically acceptable marginal gap. Mean gap dimensions and standard deviations were calculated for marginal
opening. The data were analyzed with a 1-way analysis of variance (.05).
Results. Mean gap dimensions and standard deviations at the marginal opening for the incisor crowns were 87
34 m for control, 83 33 m for Celay In-Ceram, 112 55 m for conventional In-Ceram, and 46 16
m for the IPS Empress 2 layering technique. Significant differences were found among the crown groups
(P.05). Compared with the control group, the IPS Empress 2 group had significantly smaller marginal
discrepancies (P.05), and the conventional In-Ceram group exhibited significantly greater marginal discrepancies (P.05). There was no significant difference between the Celay In-Ceram and the control group.
Conclusion. Within the limitations of this study, the marginal discrepancies were all within the clinically
acceptable standard set at 120 m. However, the IPS Empress 2 system showed the smallest and most homogeneous gap dimension, whereas the conventional In-Ceram system presented the largest and more variable gap
dimension compared with the metal ceramic (control) restoration. (J Prosthet Dent 2003;90:459-64.)

CLINICAL IMPLICATIONS
This in vitro study demonstrated that the marginal discrepancies of 3 all-ceramic systems tested
were within the clinically acceptable standard of 120 m.

arginal fit is a very important aspect for fixed

restorations because a large marginal opening allows
more plaque accumulation, gingival sulcular fluid flow,
and bone loss, resulting in microleakage, recurrent caries, and periodontal disease.1,2 Many studies have evaluated the marginal discrepancies of single restorations
fabricated using various systems and materials.3-21 In
most studies, however, statistical inferences were difficult to make because of relatively small sample sizes and
insufficient measurements per specimen. In the study of
Beschnidt et al,3 the number of specimens used for the
statistics of marginal gap in each group was 2 to 10.
Some authors4 selected 5 specimens for each group,
whereas others chose 6,5,6 8,7 or 12.9 Many authors10-21
a

Graduate Student and Resident, Department of Prosthodontics.

Professor, Department of Prosthodontics.
c
Assistant Professor, Department of Prosthodontics.
b

NOVEMBER 2003

designed studies using 10 specimens per group. If the

sample size is small, a nonparametric method should be
used for data analysis because it is difficult to assume that
the nature of the underlying distribution of the population is normal, and the power of nonparametric tests is
relatively low in comparison with parametric tests.22 The
results and statistical inferences drawn from such a small
sample size may be questionable. In many studies, 4
measurements per specimen5-7,9,11,13,19 have been performed (for example, labial, mesial, distal, lingual).
However, 4 measurements are not representative of the
marginal gap in 1 specimen. According to Groten et
al,23 50 measurements are required to obtain clinically
relevant information about gap size regardless of
whether the measurement sites were selected in a systematic or random manner. As previously mentioned,
many studies have used smaller sample sizes. An adequate number of specimens is required to obtain reliable
THE JOURNAL OF PROSTHETIC DENTISTRY 459

THE JOURNAL OF PROSTHETIC DENTISTRY

YEO, YANG, AND LEE

Fig. 1. Prepared tooth embedded in autopolymerizing resin.

statistical inferences, and at least 50 measurements are

needed to determine the marginal opening in each specimen.
The demand for esthetic dental restorations has made
the all-ceramic crown a popular form of restoration.
Two important factors for all-ceramic restoration are
strength and marginal adaptation, and if all-ceramic restorations are to be successful, they must satisfy the clinical requirements in both respects. Crowns fabricated
with all-ceramic systems may use different techniques.
Conventional In-Ceram (VITA Zahnfabrik, Bad Sackingen, Germany) uses the slip casting technique.14,19
Slip-cast alumina is partially sintered initially in a furnace
and then infiltrated with liquid glass in a second firing
process. Approximately 14 hours (10 hours of sintering
and 4 hours of glass infiltration) are needed in the furnace. This procedure makes an alumina core and porcelain is built onto this core to produce the restoration.
The IPS Empress 2 layering technique (Ivoclar Vivadent, Schaan, Liechtenstein) uses a heat-press method
to produce a high-strength core.3,24 The core material is
primarily lithium-disilicate glass. As for In-Ceram, feldspathic porcelain is built on to the core to produce the
definitive crown. Copy-milling technologies are used for
making a Celay In-Ceram crown (Mikrona Technologie
AG, Spreitenbach, Switzerland).14 Some authors reported that the copy-milling was less accurate compared
with other methods.25,26 However, copy-milled Celay
In-Ceram single crowns have been shown to be biologically acceptable and to have relatively homogenous
marginal gaps.3,10 The Celay In-Ceram system has 2
important advantages over conventional In-Ceram. The
Celay In-Ceram system has a higher fracture strength
(about 10%) than slip casting In-Ceram.27 In addition,
the Celay system uses alumina blanks that have already
460

Fig. 2. Die used for definitive restoration.

been sintered, eliminating the 10-hour sinter-firing procedure that is necessary for the conventional technique.14 The Celay core requires 40 minutes for glass
infiltration, and therefore this system saves time.14
Regarding margin design, Shearer et al28 showed no
significant difference between chamfer and shoulder
margins in the fit of In-Ceram crowns. Syu et al29 reported no significant differences for marginal gaps
among metal ceramic crowns with shoulder, shoulderbevel, and chamfer finish lines. However, Goodacre et
al30 recommended a shoulder margin design for all-ceramic restorations.
Variation exists regarding what constitutes a clinically
acceptable margin.31 McLean and von Fraunhofer32
proposed that a restoration would be successful if marginal gaps and cement thicknesses of less than 120 m
could be achieved. This criterion has been cited in some
articles.4,5,19 The purpose of this study was to compare
the marginal adaptation of single anterior restorations
made using the 3 previously mentioned all-ceramic systems (Celay In-Ceram, Conventional In-Ceram, and
the IPS Empress 2 layering technique) with metal-ceramic single restorations, using an optical microscope
with image processing.

MATERIAL AND METHODS

One extracted maxillary central incisor without caries
was cleaned and embedded in an autopolymerizing resin
VOLUME 90 NUMBER 5

YEO, YANG, AND LEE

THE JOURNAL OF PROSTHETIC DENTISTRY

Fig. 3. Demonstration of computer aided measurement of marginal fit.

manufactured block (Orthodontic resin; Densply International Inc, Milford, Del). The long axis of the tooth
was oriented perpendicular to the surface of the block.
The tooth was prepared for crown fabrication. With a
high-speed hand piece, an incisal reduction of 2 to 3 mm
and axial reduction of approximately 1 mm were prepared. The preparation was completed by milling (F2;
Degussa Korea Inc, Seoul, Korea), resulting in an approximately 1-mm shoulder margin, 6-degree tapered
angles and an approximate height of 7 mm (Fig. 1).
The data of other studies3,5,7,10 were included in a
power analysis to determine the sample size using the
statistical software (Release 8.1; SAS Institute Inc, Cary,
NC). The sample size calculation at a power of 99%
(.05) indicated that there should be 30 specimens
per group.
A preliminary impression using a stock tray was made
using irreversible hydrocolloid impression material
(Aroma fine; GC Korea Co Ltd, Seoul, Korea). A plaster
(Samwoo plaster; Samwoo Co, Ulsan, Korea) cast was
made. After using 2 sheets of baseplate wax (Modeling
wax; Pemaco Inc, St. Louis, Mo) for relief on the cast,
the 120 custom-made trays were fabricated using acrylic
resin (Quicky; Nissin Dental Products Inc, Kyoto, Japan). Definitive impressions were made with vinyl polysiloxane (Examix; GC America Inc, Alsip, Ill) using the
custom trays, and 120 definitive stone (Rhombrock;
Mitsubishi, Tokyo, Japan) dies were fabricated (Fig. 2).
One hundred twenty veneered crowns (30 crowns
per group) were fabricated. Thirty metal-ceramic (Rexillium III; JENERIC/PENTRON Inc, Wallingford,
Conn and VMK 95; VITA Zahnfabrik) crowns with a
metal margin (control), 30 copy-milled Celay In-Ceram
NOVEMBER 2003

Table I. Shapiro-Wilks normality test for marginal fit data

Metal Ceramic
Celay In-Ceram
Conventional In-Ceram
Empress 2

Statistic

df

P value

.939
.945
.953
.937

30
29
30
29

.105
.214
.305
.100

df, Degree of freedom.

crowns (Vitadur alpha; VITA Zahnfabrik), 30 In-Ceram

alumina crowns (Vitadur alpha; VITA Zahnfabrik) produced by slip casting, and 30 IPS Empress 2 crowns
using a layering technique were fabricated. Each group
of crowns was fabricated by an experienced dental technician who was accustomed to the specific system. All
crowns were fabricated according to their respective
manufacturers recommendations.
The marginal fit was evaluated by measuring the gap
between the edge of the crown and the prepared tooth
margin in an optical microscope with image processing
software (Acura 2000; INTEK PLUS, Taejon-City, Korea) at original magnification 240. The accuracy of the
optical microscope was 0.1 m. The gap was measured as the minimum distance from 1 point of the
crown edge to a line determined by least squares of
points to the tooth margin (Fig. 3). All measurements
and the least squared lines were computed by the software. Measurements were made without cementation.
The marginal gap of a crown was measured at 50 points
along the margin that were randomly selected in distances of approximately 400 m. The marginal fit of a
crown was defined as a mean value of these 50 measurements. The means and standard deviations of the mar461

THE JOURNAL OF PROSTHETIC DENTISTRY

YEO, YANG, AND LEE

Fig. 4. Normal probability plots of each group. A, Metal-ceramic, control group. B, Celay In-Ceram group. C, Conventional
In-Ceram group. D, IPS Empress 2 group.

ginal fit were rounded to 1 m level. The criterion of

120 m was used as the maximum clinically acceptable
marginal opening in this study.32
The Shapiro-Wilks test was performed to confirm
that the data of marginal gaps were normally distributed
(.05). Mean values and SDs per group were calculated, and statistical inferences among the groups were
made using 1-way analysis of variance and the Dunnett
test (.05).

RESULTS
Two marginal discrepancy values, 1 in each of the
Celay In-Ceram and Empress 2 groups, were excluded
because they were not normally distributed. The Shapiro-Wilks test confirmed that the remainder of the data
were normally distributed (Table I). Figure 4 shows
normal probability plots of each group. Therefore, a
parametric test was performed on the basis of the as462

sumption that population distributions were normal.

Table II showed the mean values and SDs of the average
gap dimensions of the crown groups. The mean values
and SDs of the marginal fit were 87 m 34 m for the
control group, 83 m 33 m for the Celay In-Ceram
group, 112 m 55 m for the conventional In-Ceram
group, and 46 m 16 m for the IPS Empress 2
layering group. The coefficients of variation for each
group demonstrated that the variability of the IPS Empress 2 group was smaller and that of the conventional
In-Ceram group was larger compared with other
groups.
Significant differences in marginal discrepancies were
found among the crown groups (Table III). The result
of the Dunnett test (Table IV) showed that the IPS
Empress 2 crown had significantly better marginal fit
than the metal ceramic restoration (P.05). The conventional In-Ceram group exhibited significantly
VOLUME 90 NUMBER 5

YEO, YANG, AND LEE

THE JOURNAL OF PROSTHETIC DENTISTRY

Table II. Mean and standard deviations of marginal fit in

each of 4 groups (m)
N per group

Mean ( SD) (m)

CV

30
29
30
29

87 (34)
83 (33)
112 (55)
46 (16)

0.39
0.39
0.49
0.34

Metal ceramic
Celay In-Ceram
Conventional In-Ceram
Empress 2
CV, Coefficient of variation.

Table III. One-way ANOVA test

Sum of
squares

df

Between groups 64300.0

3
Within groups
157147.4 114
Total
221447.3 117

Mean square

F value

P value

21433.329
1378.486

15.548

.000

df, Degree of freedom.

Table IV. Result of Dunnetts test

Control

Mean difference
(m)

P value

Metal ceramic
Metal ceramic
Metal ceramic

4.8816
24.255*
41.139*

.923
.034
.000

Experimental

Celay In-Ceram
In-Ceram
Empress 2

*Mean difference is significant at .05 level.

greater marginal discrepancies than the control

(P.05). However, no significant differences were recorded between the Celay In-Ceram and the control
group at the .05 level. On the basis of the criterion of
120 m as the limit of clinical acceptability, the mean
marginal fits of all 4 systems were acceptable.

DISCUSSION
There have been many studies regarding the marginal
fit of crowns. However, no studies were found to be
similar to this study with respect to sample size and the
number of measurements per specimen. In Celay InCeram, Beschnidt et al3 reported that maxillary incisor
crowns had the marginal gap of 78 m, which was in
agreement with the results found in this study. However, Groten et al10 reported a different value of 18.3
m. Beschnidt et al3 tested the marginal fit of IPS Empress 2 maxillary incisor crowns and found the marginal
opening to be 62 m, which was similar to the result of
this study. Sulaiman et al19 and Grey et al33 evaluated
the marginal gaps of conventional In-Ceram crowns,
which were 160.66 m and 123 m, respectively. These
results were also in accordance with this study. However, Rinke et al14 reported the marginal gap of 33.5 m
in In-Ceram crowns. An explanation of the lack of agreement may be variation in the methods used by various
investigators studying marginal accuracy. Sulaiman et
NOVEMBER 2003

al19 suggested that the cause could be the use of different measuring instruments. Sample size and the number
of measurements per specimen may also have contributed to the variation.
Two variables were identified between the groups in
this experimental design; 1 was the difference between
the systems and, the other was the difference between
the skills of dental technicians who made the single restorations. It was assumed in this study that there was no
difference in the techniques of the dental technicians
because they each had fabricated more than 500 crowns
using each system and were considered experienced.
Therefore it is the authors opinion that the disparity is
due to the differences in the systems. The differences in
the standard deviations and the coefficients of variation
(Table II) of the 4 groups were assumed to be the technique sensitivity of the systems. This study showed clinically acceptable marginal discrepancy of all 4 groups
tested. However, the Celay In-Ceram and the IPS Empress 2 systems seemed to be less technique-sensitive
than the conventional In-Ceram system. The Celay
technique may be considered better than the traditional
slip casting method, considering manufacturing time,14
strength,14,27 and marginal accuracy.3,14
There were some limitations in this study. Marginal
opening was measurable in this experimental design.
However, internal fit of the crowns is impossible to measure. Although certain investigations focused on marginal fit, other researchers4,7,9 also evaluated the internal
fit of crowns. Measuring the internal fit of artificial
crowns requires cementing the crowns and sectioning
the specimens. In the case of sectioning, the number of
measurements per specimen is limited. The assumption
that all technicians produced equivalent work is another
limitation in this study, although they were skilled. Further investigations are required for developing new experimental designs to measure both the marginal and
internal fit.

CONCLUSIONS
Within the limitations of this study, the following
conclusions were drawn:
1. Mean gap dimensions and standard deviations at
the marginal opening for the anterior single crowns were
87 34 m for the control (metal-ceramic restoration),
83 33 m for Celay In-Ceram, 112 55 m for
conventional In-Ceram, and 46 16 m for IPS Empress 2 layering technique.
2. The marginal discrepancies found in this study
were within the clinically acceptable standard (120 m).
3. The IPS Empress 2 system showed significantly
smaller (P.05) and homogeneous marginal gap than
the control. The conventional In-Ceram presented significantly larger (P.05) and variable gap dimensions
compared with the metal ceramic control. There was no
463

THE JOURNAL OF PROSTHETIC DENTISTRY

significant difference in marginal fit between the Celay

In-Ceram and the control at the .05 level.
4. IPS Empress 2 crowns demonstrated the best marginal fit in this study, compared with In-Ceram and
Celay In-Ceram.
REFERENCES
1. Sorensen JA. A rationale for comparison of plaque-retaining properties of
crown systems. J Prosthet Dent 1989;62:264-9.
2. Sorensen SE, Larsen IB, Jorgensen KD. Gingival and alveolar bone reaction
to marginal fit of subgingival crown margins. Scand J Dent Res 1986;94:
109-14.
3. Beschnidt SM, Strub JR. Evaluation of the marginal accuracy of different
all-ceramic crown systems after simulation in the artificial mouth. J Oral
Rehabil 1999;26:582-93.
4. May KB, Russell MM, Razzoog ME, Lang BR. Precision of fit: the Procera
AllCeram crown. J Prosthet Dent 1998;80:394-404.
5. Leong D, Chai J, Lautenschlager E, Gilbert J. Marginal fit of machinemilled titanium and cast titanium single crowns. Int J Prosthodont 1994;
7:440-7.
6. Schaerer P, Sato T, Wohlwend A. A comparison of the marginal fit of three
cast ceramic crown systems. J Prosthet Dent 1988;59:534-42.
7. Oruc S, Tulunoglu Y. Fit of titanium and a base metal alloy metal-ceramic
crown. J Prosthet Dent 2000;83:314-8.
8. Vahidi F, Egloff ET, Panno FV. Evaluation of marginal adaptation of
all-ceramic crowns and metal ceramic crowns. J Prosthet Dent 1991;66:
426-31.
9. Karlsson S. The fit of Procera titanium crowns. An in vitro and clinical
study. Acta Odontol Scand 1993;51:129-34.
10. Groten M, Girthofer S, Probster L. Marginal fit consistency of copy-milled
all-ceramic crowns during fabrication by light and scanning electron
microscopic analysis in vitro. J Oral Rehabil 1997;24:871-81.
11. Hung SH, Hung KS, Eick DJ, Chappell RP. Marginal fit of porcelain-fusedto-metal and two types of ceramic crown. J Prosthet Dent 1990;63:26-31.
12. Cooney JP, Richter WA, MacEntee MI. Evaluation of ceramic margins for
metal-ceramic restorations. J Prosthet Dent 1985;54:1-5.
13. Holmes JR, Sulik WD, Holland GA, Bayne SC. Marginal fit of castable
ceramic crowns. J Prosthet Dent 1992;67:594-9.
14. Rinke S, Huls A, Jahn L. Marginal accuracy and fracture strength of
conventional and copy-milled all-ceramic crowns. Int J Prosthodont 1995;
8:303-10.
15. Blackman R, Baez R, Barghi N. Marginal accuracy and geometry of cast
titanium copings. J Prosthet Dent 1992;67:435-40.
16. Keith SE, Miller BH, Woody RD, Higginbottom FL. Marginal discrepancy
of screw-retained and cemented metal-ceramic crowns on implant abutments. Int J Oral Maxillofac Implants 1999;14:369-78.
17. Morris HF. Department of Veterans Affairs Cooperative Studies Project
No. 242. Quantitative and qualitative evaluation of the marginal fit of cast
ceramic, porcelain-shoulder, and cast metal full crown margins. Participants of CSP No. 147/242. J Prosthet Dent 1992;67:198-204.
18. Tjan AH, Li T, Logan GI, Baum L. Marginal accuracy of complete crowns
made from alternative casting alloys. J Prosthet Dent 1991;66:157-64.

464

YEO, YANG, AND LEE

19. Sulaiman F, Chai J, Jameson LM, Wozniak WT. A comparison of the

marginal fit of In-Ceram, IPS Empress, and Procera Crowns. Int J Prosthodont 1997;10:478-84.
20. Omar R. Scanning electron microscopy of the marginal fit of ceramometal
restorations with facially butted porcelain margins. J Prosthet Dent 1987;
58:13-9.
21. Duncan JD. The casting accuracy of nickel-chromium alloys for fixed
prostheses. J Prosthet Dent 1982;47:63-8.
22. Chao LL. Statistics: methods and analysis. 2nd ed. Tokyo: McGraw-Hill;
1974. p. 436-7.
23. Groten M, Axmann D, Probster L, Weber H. Determination of the minimum number of marginal gap measurements required for practical in vitro
testing. J Prosthet Dent 2000;83:40-9.
24. Tinschert J, Natt G, Mautsch W, Augthun M, Spiekermann H. Fracture
resistance of lithium disilicate-, alumina-, and zirconia-based three-unit
fixed partial dentures: a laboratory study. Int J Prosthodont 2001;14:231-8.
25. Molin M, Karlsson S. The fit of gold inlays and three ceramic inlay systems.
A clinical and in vitro study. Acta Odontol Scand 1993;51:201-6.
26. Trushkowsky RD. Ceramic inlay fabrication with three-dimensional copy
milling technology. Celay. Compend Contin Educ Dent 1998;19:1077-80,
1082-4, passim; quiz, 1086.
27. Hwang JW, Yang JH. Fracture strength of copy-milled and conventional
In-Ceram crowns. J Oral Rehabil 2001;28:678-83.
28. Shearer B, Gough MB, Setchell DJ. Influence of marginal configuration
and porcelain addition on the fit of In-Ceram crowns. Biomaterials 1996;
17:1891-5.
29. Syu JZ, Byrne G, Laub LW, Land MF. Influence of finish-line geometry on
the fit of crowns. Int J Prosthodont 1993;6:25-30.
30. Goodacre CJ, Campagni WV, Aquilino SA. Tooth preparations for complete crowns: an art form based on scientific principles. J Prosthet Dent
2001;85:363-76.
31. Hunter AJ, Hunter AR. Gingival margins for crowns: a review and discussion. Part II: discrepancies and configurations. J Prosthet Dent 1990;64:
636-42.
32. McLean JW, von Fraunhofer JA. The estimation of cement film thickness
by an in vivo technique. Br Dent J 1971;131:107-11.
33. Grey NJ, Piddock V, Wilson MA. In vitro comparison of conventional
crowns and a new all-ceramic system. J Dent 1993;21:47-51.
Reprint requests to:
DR JAE-HO YANG
DEPARTMENT OF PROSTHODONTICS
SEOUL NATIONAL UNIVERSITY DENTAL HOSPITAL
28 YONKON-DONG, CHONGRO-GU
SEOUL
KOREA
FAX: 82-2-760-3860
E-MAIL: jhoyang@snu.ac.kr
Copyright 2003 by The Editorial Council of The Journal of Prosthetic
Dentistry.
0022-3913/2003/$30.00 0
doi:10.1016/j.prosdent.2003.08.005

VOLUME 90 NUMBER 5