dental

materials
Dental Materials 15 (1999) 250–256
www.elsevier.com/locate/dental

Surface roughness and bond strength of enamel to composite

M. Jung*, L.O. Wehlen, J. Klimek
Department for Operative and Preventive Dentistry, Justus-Liebig-University, Schlangenzahl 14, D-3592 Giessen, Germany
Received 19 November 1998; accepted 18 March 1999

Abstract
Objectives: The aim of the study was to assess the influence of pretreating enamel by different rotary instruments on shear-bond strength of
composite to enamel after the acid etch technique.
Methods: Ninety human enamel specimens with transversely and longitudinally sectioned prisms were ground flat, and half of the enamel
area of each specimen was prepared with one of the eight different rotary instruments: a 120 mm diamond bur, three finishing diamonds (30,
15 and 8 mm), two tungsten carbide finishing instruments (8- and 30-fluted) and two stones (Dura green and Dura white). The other half of the
enamel area was left untreated. Enamel roughness was measured by laser-stylus profilometry. Two cylinders filled with the hybrid composite
Pertacw were adhesively fixed on each enamel surface, one on the prepared side and the other on the untreated side. After seven days of water
storage, shear-bond strength was measured with a Universal testing machine.
Results: Rz values of the treated enamel surfaces were between 4.14 and 17.75mm. The shear-bond strength data varied between 16.8 and
29.4 MPa and were similar for transversely and longitudinally orientated prisms. The different roughness data had no significant influence on
shear-bond strength of composite to enamel. Additional SEM investigation showed that the surface geometry caused by the rotary instru-
ments largely persisted after acid etching.
Significance: With regard to shear-bond strength, there was no advantage in using a special bur for finishing enamel surfaces prior to the
acid etch technique. Burs causing extreme roughness should not be used for this purpose. q 1999 Academy of Dental Materials. Published by
Elsevier Science Ltd. All rights reserved.
Keywords: Acid etch technique; Enamel; Adhesion; Surface roughness; Composite resin; Finishing

1. Introduction known about the influence of different enamel surface

The bonding of composite to enamel after acid etching is
one of the most common techniques in current operative
dentistry. This includes the placement of direct restorations
and the application of fissure sealants, as well as the
adhesive cementation of tooth-coloured inlays made of
composites or ceramics. In contrast to conventionally placed
or cemented restorations, adhesive bonding permits a mini-
mum reduction of dental hard tissue and increases the
fracture resistance of restored teeth, by reducing the cusp
deformation under load [1].
Since its introduction by Buonocore in 1955 [2], the acid
etch technique has been the subject of many investigations.
The major topics of concern among these studies have been
the type and concentration of acids [3–5], the time and
practice of rinsing [6], the influence of air drying [7] and
the effect of enamel contamination [8,9]. Presently little is
* Corresponding author. Tel.: 1 49-641-99-46173; fax: 1 49-641-99-
46169.
E-mail address: martin.jung@dentist.med.uni-giessen.de (M. Jung)
0109-5641/99/$20.00 + 0.00 q 1999 Academy of Dental Materials. Published by Elsevier Science Ltd. All rights reserved.
PII: S01 09-5641(99)0004 2-1
textures resulting from cutting with rotary instruments on
bond strength after acid etching.
Our study aimed to assess the effect of varying surface
roughness caused by rotary instrumentation on the shear-
bond strength of enamel to composite. The influence of
acid etching on machined enamel surfaces was also consid-
ered in the investigation.
2. Materials and methods
Ninety previously impacted third human molars were
used for the study. Tissue remnants were removed with a
scaler and the teeth were cleaned with a rotating brush and
pumice. Each tooth was then embedded in a chemically
curing acrylic resin (Technovitw 4071, Kulzer, Wehrheim,
Germany), with the buccal surface left uncovered. With the
help of a water-cooled tape-saw (Exakt-Trennschleifsystem,
Exakt Apparatebau, Norderstedt, Germany) each embedded
tooth was cut mesio-distally from cusp-tip to cusp-tip. The
buccal surfaces and the sectioned surfaces were ground flat
 M. Jung et al. / Dental Materials 15 (1999) 250–256
Table 1
The nine different test groups with the corresponding rotary instruments
Test group Type of rotary instrument
0 Abrasive paper disc
1 Diamond instrument
2 Finishing diamond
3 Finishing diamond
4 Finishing diamond
5 Tungsten carbide finishing instrument
6 Tungsten carbide finishing instrument
7 Dura-green stone
8 Dura-white stone
with 600 grit SiC abrasive paper discs (Leco Corpration, St.
Joseph, MI, USA) at 120 rpm under water cooling by use of
an automatic polisher A-250 (Jean Wirtz, Düsseldorf,
Germany). This resulted in two specimens from each
tooth: the flattened buccal surface with mainly transversely
cut enamel prisms (n ˆ 90; referred to as “transverse speci-
mens”) and the sectioned surface with mainly longitudinally
orientated enamel prisms (n ˆ 90; referred to as “longitu-
dinal specimens”). One half of each surface was used as the
test surface, the other half was left untreated and served as a
control.
Fig. 1. Schematic drawing of the specimen preparation and the positioning
of the composite cylinders on the enamel surfaces with transverse and
longitudinal prism orientation.
251
Grain size/blades Code number Producer
600 grit Leco Corp., USA
120 mm 6852314023 Brasseler, Germany
30 mm 8852314014 Brasseler, Germany
15 mm 852EF314014 Brasseler, Germany
8 mm 852UF314014 Brasseler, Germany
8-Fluted 134314014 Brasseler, Germany
30-Fluted 134UF314014 Brasseler, Germany
105 mm CA/CN1/0052 Shofu, Japan
28 mm CA/CN1/0221 Shofu, Japan
2.1. The rotary instruments
The test halves of the specimens were prepared manually
using eight different rotating instruments: a coarse diamond
with a grain size of 120 mm, three finishing diamonds, two
tungsten carbide burs and a Dura-green and a Dura-white
stone (Table 1). The diamonds and tungsten carbide burs
were used in a red-ring handpiece with friction-grip
mechanism (Intramatic Lux 2 25 LS, KaVo, Biberach,
Germany) at 120 000 rpm under three-way water-cooling.
A blue-ring handpiece Intramatic Lux 20L with the head 68
LDN (KaVo) at 20 000 rpm and two-way water-cooling
was used for the stones. Each surface was cut for 20 s;
after two preparations a new bur was selected.
The treatment with the abrasive discs and the eight differ-
ent rotating instruments resulted in nine test groups. Ten of
the transverse specimens and ten of the longitudinal speci-
mens were randomly assigned to each group.
2.2. Quantitative surface analysis
Quantitative surface analysis values were determined by
profilometry for two specimens of each group. The laser
stylus pick-up Focodyn with a focus diameter of 1 mm
(Rodenstock, Munich, Germany) was used in combination
with the processing unit S8P (Mahr, Göttingen, Germany).
Because of the limited enamel thickness of the sectioned
surfaces, profilometry had to be performed under different
conditions for the longitudinal and transverse specimens.
The transverse length (LT) was 1.75 mm for transverse and
1.39 mm for longitudinal specimens; the cut-off (RC-filter)
was set at 0.25 and 0.08 mm for transverse and longitudinal
specimens respectively.
Nine automatically generated parallel tracings were
carried out on each transverse surface, six tracings on
each longitudinal surface, resulting in a scanned surface of
1.75 × 1.75 mm 2 for transverse and 1.39 × 0.56 mm 2 for
longitudinal specimens. The mean depth of roughness
among five adjacent spaces (Rz) was recorded.
2.3. Shear-bond strength tests
The test and control areas on each surface were etched
252 M. Jung et al. / Dental Materials 15 (1999) 250–256
Fig. 2. The positioning of the enamel specimen with the composite cylinder
and the wedge-shaped crosshead before shearing.
with 37% phosphoric acid (Etchant Gel, Espe, Seefeld,
Germany) for 40 s, rinsed with tap-water for 20 s and
dried for 10 s. Pertacw Universal Bond (Batch No.
743410/05, Espe) was applied and air-thinned. Plastic
cylinders with an inner diameter of 1.004 mm and a
height of 3 mm were filled with Pertacw Hybrid Unifil
(Batch No. 004/0010, Espe). Two cylinders were placed
on each specimen, one on the test-side, and one on the
control-side (Fig. 1). Light curing was done for 40 s
with the polymerization unit Optilux 400 (Demetron
Research Corporation, Danbury, USA). Finally, the plastic
coverings of the cured composite cylinders were removed
with a scalpel.
After seven days of water storage at 378C, shear-bond
strength tests were performed by mounting the specimens
Fig. 3. Roughness (Rz) of enamel surfaces with transversely and longitudinally orientated prisms after use of different rotary instruments (each bar represents
the mean and standard deviation of 18 parallel tracings on two enamel surfaces for transverse and 12 tracings on two enamel surfaces for longitudinal
specimens: FD—finishing diamond; TCFI—tungsten carbide finishing instrument).
in a Zwick Universal Testing machine (Zwick of America
Inc, East Windsor, USA) using a wedge-shaped rod at a
crosshead speed of 1 mm/min (Fig. 2). The maximum
load at failure was recorded in N and then converted
to MPa.
Statistical analysis to determine the effect of rotary treat-
ment of the nine test groups and of the prism orientation on
shear-bond strength data was carried out with 3-way
ANOVA.
2.4. SEM evaluation
In order to assess the influence of etching on the surface
geometry of the enamel after cutting, 18 additional teeth
were prepared as described above, resulting in 18 transverse
and 18 longitudinal specimens. Following instrumentation,
half of the treated surfaces were coated with nail varnish
while the other half were etched with 37% phosphoric acid
for 40 s (Etchant Gel, Espe). After water rinsing and drying
of the etched area, the nail varnish was removed with acet-
one.
The specimens were sputtered with gold (SCD 040,
Balzers Union, Liechtenstein) and examined under SEM
(PSEM 500, Philips, Eindhoven, Netherlands) at a tension
of 25 kV.
3. Results
3.1. Quantitative surface analysis
The cutting of the specimens with eight different rotating
instruments yielded roughness values ranging from an Rz of
4.14 mm for the smoothest to 17.75 mm for the roughest
enamel surface (Fig. 3). By the application of the 600 grit
abrasive paper discs, the specimens were polished to an Rz
of 4.05 mm for transversely and 6.08 mm for longitudinally
sectioned prisms. Only the use of the 30-fluted tungsten
 M. Jung et al. / Dental Materials 15 (1999) 250–256 253

Fig. 4. Shear-bond strength …x ^ SD† of enamel surfaces with transversely sectioned prisms after the use of different rotary instruments (n ˆ 10 for each group:
FD—finishing diamond; TCFI—tungsten carbide finishing instrument).

carbide finishing instrument achieved a further reduction of
roughness to 5.59 mm on longitudinally sectioned prism
surfaces. All the other rotating instruments caused a rough-
ening of the enamel surfaces to various degrees. The rough-
est surface was produced by the coarse diamond (Rz ˆ
17.75 mm for transverse and 11.77 mm for longitudinal
sections). The three different diamond finishing burs led to
successively smaller roughness values. The Dura-green and
Dura-white stones as well as the 8-fluted tungsten carbide
bur yielded Rz data, which were similar to those of the 8 and
15 mm finishing diamonds.
When comparing the two types of sectioned surfaces, it
became evident that the Rz values of the transversely
sectioned prisms were of a much wider range (from 4.14
to 17.75 mm) than those of the longitudinally sectioned
prisms (from 5.59 to 11.77 mm).
The surfaces with transversely sectioned enamel
prisms could be finished to lower Rz values than the
corresponding specimens with a longitudinal prism
orientation.
3.2. Shear-bond strength tests
Both transversely and longitudinally sectioned prisms
exhibited no statistically significant difference between
groups regarding their shear-bond strength values (p ˆ
0.644). A comparison between the four diamond instrument
groups revealed only a slight tendency towards higher
shear-bond strength data with increasing roughness (Figs.
4 and 5). The use of the tungsten carbide instruments
produced similar shear-bond strength values for the long-
itudinally sectioned prisms; for transversely sectioned
prisms, shear-bond strenth reached 25.39 MPa after the
use of the 30-fluted instrument and 15.05 MPa after use of
the 8-fluted carbide finishing bur. For both types of prism
orientation, the shear-bond strength data obtained after the
use of Dura-white stones were similar. The Dura-green
stones yielded 27.83 MPa on longitudinally and
16.81 MPa on transversely sectioned prisms. As mentioned
above, these differences were not statistically significant.
The only main effect observed was in the overall

Fig. 5. Shear-bond strength …x ^ SD† of enamel surfaces with longitudinally orientated prisms after the use of different rotary instruments (n ˆ 10 for each
group: FD—finishing diamond; TCFI—tungsten carbide finishing instrument).
254 M. Jung et al. / Dental Materials 15 (1999) 250–256
Fig. 6. Application of a 120 mm diamond instrument on enamel surface
with longitudinal prism orientation (left side) and after etching with phos-
phoric acid (right side).
treatment; as the shear-bond strength values of the control
sides were significantly greater than that of the test sides
(p ˆ 0.002).
There was no significant difference between transverse
and longitudinal prism orientation on shear-bond strength
values (p ˆ 0.372).
3.3. SEM evaluation
The SEM evaluation of the transition zone between cut
and etched surfaces showed a modification of the character-
istic surface features after the application of phosphoric
acid. In most cases, the surface geometry caused by the
rotating instruments continued onto the etched surface in a
reduced intensity; the typical pattern of etched enamel was
superimposed onto the surface irregularities that resulted
from the prior use of different burs. From Fig. 6, an example
of longitudinally sectioned prisms after preparation with a
coarse diamond, it can be seen that the grooves continued on
Fig. 7. Instrumentation of transversely sectioned enamel prisms with a
30 mm finishing diamond (left side) and after etching with phosphoric
acid (right side).
the etched surface. The same effect could be seen after the
application of a 30 mm diamond finishing instrument on
transversely cut enamel prisms (Fig. 7). In some cases
however, only a few of the surface irregularities could be
recognized on the etched surface. This is depicted in Fig. 8,
at the borderline between the etched and unetched areas
after the use of a Dura-white stone.
4. Discussion
The aim of the sample preparations was to simulate the
two most common types of enamel margin quality prior to
the acid etch technique. A primarily transverse prism orien-
tation is created clinically when the margins of a class III
cavity preparation are bevelled. By flattening the buccal
surface with the abrasive paper an enamel surface was
created with a prism orientation resembling this clinical
situation. Sectioning the teeth mesio-distally resembled
the preparation of a class II cavity with no margin
bevelling and a primarily longitudinal orientation of
enamel prisms.
After cutting the specimens with different rotary instru-
ments, the surfaces were evaluated quantitatively by an opti-
cal laser pick-up system. This type of surface tracing has
some advantages over conventional, contact stylus profilo-
metry. The laser is focussed to a diameter of 1 mm, thus
permitting a precise tracing of even relatively smooth
surfaces. The diameter of a conventional stylus varies
between 4 and 20 mm. This could be too inexact for those
enamel surfaces in the present study, which were finished to
an Rz of 4 mm or less. Another advantage of the laser is that
the contact-free tracing of surfaces eliminates the possibility
of surface damage caused by direct contact with the tip of a
mechanical pick-up.
In addition to Rz, there are several other parameters for
the characterization of surface roughness. Maximum rough-
ness (Rt) and average roughness (Ra) are frequently used in
the literature [10]. It was the aim of the present study to
determine if there was a correlation between surface rough-
ness after cutting and shear-bond strength. Rz was chosen,
because this parameter gives a better indication of the actual
depth of the surface irregularities.
The profilometric evaluation of the prepared enamel
surfaces showed that the roughness varied widely, depend-
ing on the type of instrument, its grain size or number of
blades and on the prism orientation of the enamel speci-
mens. The use of diamond burs resulted in the greatest Rz
values, the tungsten carbide finishing instruments produced
the smoothest surfaces and the stones were between these
two groups.
The results are consistent with the findings of Al-Omari et
al. [11], who concluded that the smoothest surfaces on
dentine and enamel were obtained using tungsten carbide
finishing burs. A study investigating the preparation of
dental hard tissue with diamond burs of different grit size
 M. Jung et al. / Dental Materials 15 (1999) 250–256
Fig. 8. The use of a Dura-white stone on the enamel with longitudinal prism
orientation (left side) and after etching with phosphoric acid (right side).
[12] showed similar results to the present study regarding
the roughness after the use of a coarse diamond bur.
However, the Rz values caused by the 30 mm finishing
diamond were much smaller than those of the present
data. This inconsistency might be due to the use of a less
exact mechanical stylus with a tip diameter of 20 mm, which
was not able to accurately trace the small grooves created by
the finishing diamonds.
The lowest Rz values were recorded for the surfaces with
transversely sectioned enamel prisms. This could be due to
the fact, that when longitudinally orientated prisms are
prepared, a cracking and tearing out of enamel rods might
occur more easily than on the transversely sectioned
surfaces. This explanation is supported by the findings of
Xu et al. [13], who investigated enamel subsurface damage
after tooth preparation with diamonds. The cracks in the
enamel were dependent on the orientation of the prisms
and were significantly longer in the direction parallel to
the enamel rods.
The shear-bond strength data reported in this study varied
between 16.8 and 29.4 MPa. These results are consistent
with those published by several others concerning the bond-
ing between enamel and composite [5,8,14].
The use of rotary instruments on enamel prior to bonding
may have two effects. An increase in roughness results in a
greater area available for bonding and could therefore facil-
itate the wetting of the etched surface. On the other hand,
deep and narrow grooves could lead to the inclusion of air
bubbles between the enamel and the composite, thus weak-
ening the bond strength. Under the conditions of this study,
there was no correlation between surface roughness caused
by rotary instruments and shear-bond strength after acid
etching. Neither the rough nor the smooth surfaces could
improve the bond strength. SEM analysis of prepared and
etched enamel areas demonstrated that the surface irregula-
rities in most cases largely persisted after etching. This
indicates that the microroughness as a result of acid etching
was of greater importance for the quality of the adhesive
255
bonding process than the additional surface roughness
obtained after the use of rotary instruments.
Finger et al. [15] found similar results for dentin surfaces
and concluded that the type of rotary instrument has little or
no influence on bonding strength of composite to dentin. On
the other hand, Mowery Jr. et al. [16] showed that differ-
ences in dentin surface roughness caused by sandpaper discs
from 60 to 1200 grit had an influence on shear-bond strength
data.
Shear-bond strength tests however, may not be the most
suitable method for investigating the quality of adhesive
bonding. The present data show a high standard deviation,
thus indicating that several tooth-dependent factors such as
varying mineralization or ultrastructural differences in
enamel composition may have influenced the measured
shear-bond strength data.
One problem often involved in this type of study is to
achieve a complete wetting of the enamel by the composite.
It has been shown in a previous study with a similar tech-
nique for the placement of composite, that there were no
signs of trapped air bubbles or other irregularities on the
failed composite surfaces after the shear tests [17].
In the present study, the shear-bond strength values of the
controls were significantly greater than those of the test
surfaces after preparation. Evidently, any kind of rotary
instrumentation had a weakening effect on the measured
bond strength. This could be explained by the enamel
subsurface damage through crack formation after tooth
preparation [13]. Furthermore, the stones that were used at
slower speed are rarely perfectly round and concentric.
They can thereby produce a hammering effect upon the
surface on which they are used [18]. Consequently, the
structure of enamel after rotary instrumentation could be
weakened and therefore adhesive bonding could be
impaired. In contrast to the rotary instrumentation, the
application of the abrasive paper disks at low speed was a
much more gentle and moderate type of treatment, which is
unlikely to cause any kind of enamel damage.
Munechika et al. [19] reported greater tensile bond
strength for transverse than for longitudinal enamel
sections. Under the conditions of this study the enamel
prism orientation had no effect on shear-bond strength.
One possible reason for this is that the teeth were not
sectioned in the middle of the approximal surface, but rather
from cusp-tip to cusp-tip. The prism orientation might there-
fore not always have been strictly longitudinal.
5. Conclusion
Under the conditions of the present study varying enamel
roughness caused by the use of eight different rotary instru-
ments had no effect on shear-bond strength of enamel to
composite after acid etching.
The coarse diamond and the Dura-green stone should not
be used for cutting enamel prior to bonding because of the
256 M. Jung et al. / Dental Materials 15 (1999) 250–256
increased roughness obtained after the use of these instru-
ments. The resultant undermining of enamel prisms could
cause them to loosen under mechanical stress, thus weak-
ening the strength of the adhesive bonding.
On the other hand, finishing diamonds and tungsten
carbide finishing instruments are equally suited for cutting
enamel prior to the acid etch technique.
Acknowledgements
The authors would like to express their thanks to Prof Dr
H.-F. Kappert from the Dental Clinic of the Albert-
Ludwigs-University in Freiburg, Germany, for his efforts
in supporting the measurements with the Zwick Universal
Testing machine.
References
[1] A.J. McCullock, B.G.N. Smith, In vitro studies of cusp reinforcement
with adhesive restorative material, Br. Dent. J. 161 (1986) 450–452.
[2] M.G. Buonocore, A simple method of increasing the adhesion of
acrylic filling materials to enamel surfaces, J. Dent. Res. 34 (1955)
849–853.
[3] P.T. Triolo Jr., E.J. Swift Jr., A. Mudgil, A. Levine, Effects of etching
time on enamel bond strength, Am. J. Dent. 6 (1993) 302–304.
[4] W.N. Wang, C. Lin, B.D. Fang, K.T. Sun, M.G. Arvystas, Effect of
H3PO4 concentration on bond strength, Angle Orthodont. 64 (1994)
377–382.
[5] J.R. Holtan, G.P. Nystrom, R.A. Phelps, T.B. Anderson, W.S. Becker,
Influence of different etchants and etching times on shear bond
strength, Oper. Dent. 20 (1995) 94–99.
[6] J.B. Summit, D.C.N. Chan, J.O. Burgess, F.B. Dutton, Effect of air/
water rinse versus water only and of five rinse times on resin-to-
etched-enamel shear bond strength, Oper. Dent. 17 (1992) 142–151.
[7] K. Ichiki, T. Fukushima, Y. Yoshida, T. Horibe, Correlation between
air-drying duration and bond strength of composite resins to teeth, J.
Prosthet. Dent. 63 (1990) 525–529.
[8] J. Xie, J.M. Powers, R.S. McGuckin, In vitro bond strength of two
adhesives to enamel and dentin under normal and contaminated
conditions, Dent. Mater. 9 (1993) 295–299.
[9] J. Perdigao, G.E. Denehy, E.J. Swift Jr., Silica contamination of
etched dentin and enamel surfaces: a scanning electron microscopic
and bond strength study, Quintessence Int. 25 (1994) 327–333.
[10] J. Leitão, T. Hegdahl, On the measuring of roughness, Acta Odontol.
Scand. 39 (1981) 379–384.
[11] W.M. Al-Omari, C.A. Mitchell, L. Cunningham, Surface roughness
of enamel and dentine surfaces prepared with different dental burs, J.
Dent. Res. 77 (1998) 810.
[12] M. Jung, H. Pantke, Die Wirkung von Diamantschleifkörpern
unterschiedlicher Körnung auf die Zahnhartsubstanz, Dtsch. Stoma-
tol. 41 (1991) 320–322.
[13] H.H.K. Xu, J.R. Kelly, S. Jahanmir, V.P. Thompson, E.D. Rekow,
Enamel subsurface damage due to tooth preparation with diamonds, J.
Dent. Res. 76 (1997) 1698–1706.
[14] W.W. Barkmeier, R.L. Erickson, Shear bond strength of composite to
enamel and dentin using Scotchbond Multi-Purpose, Am. J. Dent. 7
(1994) 175–179.
[15] W.J. Finger, A. Manabe, B. Alker, Dentin surface roughness vs. bond
strength of dentin adhesives, Dent. Mater. 5 (1989) 319–323.
[16] A.S. Mowery Jr., M. Parker, E.L. Davis, Dentin bonding: the effect of
surface roughness on shear bond strength, Oper. Dent. 12 (1987) 91–
94.
[17] M. Jung, C. Ganss, S. Senger, Effect of Eugenol-containing temporary
cements on bond strength of composite to enamel, Oper. Dent. 23
(1998) 63–68.
[18] F. Lutz, J.C. Setcos, R.W. Phillips, New finishing instruments for
composite resins, J. Am. Dent. Assoc. 107 (1983) 575–580.
[19] T. Munechika, K. Suzuki, M. Nishiyama, M. Ohashi, K. Horie, A
comparison of the tensile bond strengths of composite resins to long-
itudinal and transverse sections of enamel prisms in human teeth, J.
Dent. Res. 63 (1984) 1079–1082.