Dental Materials Journal 2020; 39(2): 272–277

Anisotropic ultimate strength and microscopic fracture patterns during tensile

testing in the dentin–enamel junction region
Yoshihiro YAMADA, Toshiko INOUE, Makoto SAITO, Fumio NISHIMURA and Takashi MIYAZAKI

Department of Conservative Dentistry, Division of Biomaterials and Engineering, Showa University School of Dentistry, 1-5-8 Hatanodai, Shinagawa-
ku, Tokyo 142-8555, Japan
Corresponding author, Toshiko INOUE; E-mail: [email protected]

The dentin–enamel junction (DEJ), which is the interface between the outer tooth enamel and the underlying dentin, has unique
biomechanical properties that allow the two different materials to function together without fracture. The aim of this study was to
investigate the anisotropic ultimate tensile strength of the DEJ. Twenty dumbbell-shaped samples were prepared from 10 extracted
human molar teeth for tensile strength testing. The maximum tensile forces at fracture were measured in the direction parallel (10
samples) and perpendicular (10 samples) to the DEJ. The mean tensile strength of the samples was significantly higher under parallel
forces (30.9±3.3 MPa) than under perpendicular forces (20.6±4.8 MPa; p<0.05). Our findings confirm the structural anisotropy of the
DEJ.

Keywords: Tooth, Dentin–enamel junction, Dentin, Enamel, Tensile strength

tensile tests. Information is also limited regarding the

INTRODUCTION
anisotropic UTS of the DEJ. No studies have analyzed
The mechanical properties of teeth provide useful the anisotropic strength at the DEJ. It is known that
information in clinical dentistry. Teeth are constituted enamel and dentin are highly anisotropic structures15-19).
of a crown and a root or roots as shown in Fig. 1A. In It is rare to see enamel that does not separate from
both parts, dentin is the most abundant mineralized dentin, and chipping of enamel is common in mammals
tissue by both weight and volume (Fig. 1B). Additionally, in general20,21). Thus, it is important to identify the
dentin is covered by enamel as shown in Fig. 1B. The anisotropic UTS of the DEJ to understand the structure
dentin–enamel junction (DEJ) is the interface between of the DEJ. Hence, the purpose of this study was to
the enamel and dentin (Fig. 1B) and is designed to allow investigate the anisotropic UTS of the DEJ structure
these two different materials to work together without and to observe microscopic fracture patterns during
fracture during mastication. The DEJ may play an tensile testing in the DEJ.
important role in inhibiting catastrophic tooth fracture.
The mechanical properties of the DEJ were first MATERIALS AND METHODS
recognized from micro-hardness profiles; enamel and
dentin near the DEJ recorded the lowest hardness Materials
values1). The strain measured across this zone when a Ten extracted, caries-free human third molars were
compressive load was applied longitudinally to the tooth obtained from patients aged 26–35 years, with informed
axis was shown to decrease from the DEJ towards the consent. The molars were extracted because of
central dentin2). Rasmussen found it is difficult to induce pericoronitis. The teeth were stored in Hank’s balanced
tooth fracture at the DEJ3). Pioch and Staehle measured saline solution (HBSS) at 4°C, and were used within 1
the shear strength of the DEJ, but fracture areas were month of their extraction according to previous reports12-14).
mainly in the dentin and never exactly at the DEJ4). Lin The protocol of this study was approved by the Ethics
et al. investigated cracks in the enamel, and found that Committee of Showa University School of Dentistry.
the cracks ran past the DEJ and into the dentin, but
did not run perpendicular to the interface5). Fracture Preparation of specimens
toughness has also been investigated6), and several The 10 molars were sectioned midway along the
investigations of the DEJ have measured hardness mesiodistal plane along the long axis of the tooth using
and fracture toughness with microindentation and a diamond saw (Isomet, Beuhler, Lake Bluff, IL, USA).
nanoindentation tests7-9). Two dentin slabs, approximately 1 mm thick, were
However, little information is available about the cut from each tooth (Fig. 1C). A rectangular block
ultimate tensile strength (UTS) of the DEJ, and previous (3.0×3.0×1.0 mm) was harvested from each dentin
studies have examined the UTS of the DEJ in only one slab (Fig. 1D). Then, dumbbell-shaped specimens
direction10,11). Tensile tests have reported useful methods (central portion; 1.0×1.0×1.5 mm, Fig. 1E) from each
for easily identifying defects; and it has been found that rectangular block were prepared using a profiling
structural features caused the failure12-14). It is possible machine. Dumbbell-shaped specimens were useful for
to identify structurally weak regions in the DEJ using identifying defects easily12-14,19). Further, we employed

Received Jan 23, 2019: Accepted Apr 14, 2019

doi:10.4012/dmj.2019-024 JOI JST.JSTAGE/dmj/2019-024
 Dent Mater J 2020; 39(2): 272–277 273

Fig. 1 Illustrations of the tooth (A) outer structure and (B) inner structure. (C) The location of dentin slabs in human
molar tooth. Schematic representations of (D) rectangular block and (E) dumbbell-shaped specimen. The location of
the tensile test specimen from a human molar tooth slab: (F) PA (subjected to tensile loading parallel to the DEJ);
and (G) PE (subjected to tensile loading perpendicular to the DEJ), where arrows show the tensile test orientation.
Schematic diagrams of dumbbell-shaped specimens: (H) PA; right side is enamel, and left side is dentin, and (I) PE;
upper side is enamel, and lower side is dentin, where arrows indicate the tensile orientation.

miniaturized dumbbell-shaped specimens for tensile were stored in HBSS to retain moisture12-14).
testing in previous study12-14,19). The specimen size and
shape affected tensile strength. Thus, these same size Tensile strength test
and shape are selected to use for this tensile study. The specimens were set on a universal testing machine
Specimens were prepared in the enamel, DEJ and (EKO Instruments, Tokyo, Japan) immediately after
dentin of the external slope in the occlusal area for two preparation. The specimens were stored in 37±0.5°C
experimental groups: PA specimens were subjected to HBSS during the tensile strength test. The maximum
tensile loading parallel to the DEJ (Fig. 1F); and PE load was used to calculate the strength. Ten specimens
specimens were subjected to tensile loading perpendicular in each experimental group (PA and PE) were tested,
to the DEJ (Fig. 1G). In other words, in PA specimens, and the mean tensile strength of the specimens was
the right side is enamel, and the left side is dentin (Fig. calculated in each group. The results were analyzed
1H); and in PE specimens, the upper side is enamel, and using ANOVA and Tukey’s multiple comparison test
the lower side is dentin (Fig. 1I). Additionally, specimens (α=0.05).
274 Dent Mater J 2020; 39(2): 272–277

Observations of the fracture surface SEM observations of enamel, dentin and the DEJ
The fractured specimens from the tensile strength tests The fractured surface is shown at low magnification in
were fixed and coated according to previous reports12-14). Fig. 4. In the PE group, the fracture occurred above the
The fractured surfaces were then observed using a DEJ; thus, only enamel is visible in the fractured surface,
scanning electron microscope (S-4700, Hitachi, Tokyo, not the DEJ or dentin (Fig. 4B). At higher magnification,
Japan). the DEJ can be seen to arrest crack propagation (Fig.
5A), and SEM images show the fibrous matrix with
RESULTS the collagen fibers and mantle dentin (Fig. 5B). Mantle

Tensile strength of the DEJ region

The tensile strengths of the two groups are shown in
Fig. 2. The mean tensile strengths were 30.9±3.3 MPa
(PA) and 20.6±4.8 MPa (PE). The tensile strength of the
parallel junctions was significantly higher than that of
the perpendicular junctions (p<0.05). Figure 3 shows
a schematic diagram of the fracture specimens after
tensile strength testing.

Fig. 3 Schematic diagram of fracture specimens after

tensile strength testing: (A) PA; and (B) PE.
In specimens subjected to a parallel load (PA
group), the fracturing occurred half in the enamel
and half in the dentin (A). In specimens exposed
to a perpendicular load (PE group), all fracturing
Fig. 2 Tensile strength of the two groups. An asterisk occurred above the DEJ within the enamel and
indicates statistical significance (p<0.05). neighboring the DEJ (B).

Fig. 4 Fractured surface representation after tensile strength testing at low magnification: (A) PA; and (B) PE.
The origin of the fracture in the PA group is indicated by a black arrow (A). Crack propagation clearly ran from the
origin, and cracks can be seen near the inner dentin. Cracks were observed within the dentin, but they have been
prevented from running into the DEJ. However, cracks in the inner enamel were also observed. The inner and outer
enamel are distinguishable from each other. The inner and outer enamel differed in the orientation of their crack
propagation. In the PE group (B), the fracture origin is indicated with a white arrow.
 Dent Mater J 2020; 39(2): 272–277 275

dentin is the outer layer of dentin that is different from Prism fractures then occur obliquely and leave prisms
the bulk of dentin. Figure 6 shows SEM micrographs of in the inner enamel (arrows, Fig. 6A). However, in the
the fractured surfaces of PA specimens. Typical fractures outer enamel of the fractured surface of the specimens,
of the enamel are shown in Figure 6A (inner enamel) fractures occur preferentially along the enamel prisms
and 6B (outer enamel). The white arrows indicate (arrows, Fig. 6B). The prisms are cleaved, and porosity
cracks, which pass through the change in orientation is seen in the surface of the fractured prisms (Fig. 6C).
from inner to outer enamel; crack propagation can be Additionally, high magnification SEM micrographs of
seen running at an angle of approximately 45° to the enamel prisms show the hydroxyapatite (HAP) crystals
DEJ and being deflected by 90° in the transitional region both in the prism and interprisms (Fig. 6D). The prism
between the inner and outer enamel. Fractures occur shows parallel HAP crystallites, and the interprism also
preferentially and transversely to the enamel prisms. shows perpendicular HAP crystallites (arrows).

Fig. 5 Fractured surface representation after tensile strength testing of PA: (A) DEJ region, and (B) mantle dentin.
At this higher magnification, cracks can be seen running to the DEJ both in the enamel and dentin regions as
indicated by the arrows. The DEJ can be seen to arrest crack propagation in the dentin and prevent failure, and the
cracks in the enamel also arrest crack propagation and prevent failure (A). SEM images of the fractured surface show
the presence near the DEJ of porous fibrous reticulate intertubular dentin with collagen fibers (B).

Fig. 6 Fractured surface representation after tensile strength testing of PA: (A) lower magnification of the inner enamel,
(B) lower magnification of the outer enamel, (C, D) higher magnification of the outer enamel.
276 Dent Mater J 2020; 39(2): 272–277

is weaker than the DEJ and the dentin below the DEJ. In
DISCUSSION
clinical dentistry, enamel cracks are observed relatively
The results of this study confirm the presence of often in sound teeth28). Additionally, chipping of the
structural anisotropy in the DEJ. The DEJ is anisotropic, enamel is common in general but it is rare to see chipping
and its strength is considerably lower when stressed in the DEJ; that is, the enamel does not separate from
perpendicular to the orientation of the DEJ. the dentin at the DEJ20,21). These findings are consistent
The tensile strength of the PE group was with our study. Specialized, highly mineralized first-
approximately 20 MPa. Previous studies on the tensile formed enamel is present close to the DEJ29,30). The
strength of tooth structure along the axis of the tooth fracture toughness of the DEJ zone was investigated
have reported that the strength of enamel is 20 MPa17), using Vickers microindentation in a previous study,
and that of dentin is 40 MPa14). Another study examined which found that the crack deflection primarily occurs
the UTS using a microtensile technique11) and recorded in the first-formed enamel7). This first-formed enamel
mean UTS values of 47 MPa (parallel to the DEJ); 42 was a zone of peak hardness and radiodensity31). The
MPa (parallel to the enamel); 11 MPa (perpendicular first-formed enamel may act as a shield helping to avoid
to the enamel); and 62 MPa (superficial dentin). They catastrophic interfacial failure7).
also found that the UTS of the DEJ is closer to enamel The tensile strength of the PA group was
than to dentin, which is in agreement with our results. approximately 30 MPa. The DEJ was well bonded
However, SEM observations revealed that the fracturing with the enamel and dentin, and there was no fracture
occurred within the enamel, and never in the DEJ. at the DEJ. The surface fractures occurred from the
Reasons for the lack of a pure interfacial failure of the inner dentin. The inner dentin tubules might serve
DEJ are that the DEJ is stronger than the enamel, and as sites for crack initiation32). Previous studies on the
that the DEJ structure is complex and is able to modify tensile strength of tooth structure have reported that
crack propagation. the strength of dentin is 60 MPa14) when tested with
Several studies have examined the tensile strength the load applied transversely to the orientation of the
of the DEJ. In a study in which specimens were prepared dentinal tubules. We tested the same specimen size as
from the buccal surface at its greatest point of contour, in a previous study (approximately 1.0 mm2). Thus, the
all specimens fractured near the DEJ10). Three types strength of the dentin part was approximately 30 MPa
of fracture pattern were observed: 50% of specimens because it was half the size (approximately 0.5 mm2).
fractured in the superficial dentin; 10% fractured in the Additionally, enamel tensile strength is lower than that
enamel; and 40% fractured in the DEJ area10). These of dentin11), so when a specimen is subjected to tensile
fracture patterns differed from those in our study. loading perpendicular to the DEJ, the tensile strength
Fractures never occurred at the interface between could be dominated by the dentin (30 MPa), resulting in
the enamel and dentin. Our specimens were prepared a theoretical strength of the DEJ of 30 MPa. Our data
from the occlusal area of molars, and all occlusal DEJ indicated a value of approximately 31 MPa, in agreement
specimens fractured within the enamel. As mentioned with the experimental data.
above, the position of the sampled specimen (for Additionally, SEM images show a porous reticulate
example occlusal or cervical) may affect the properties matrix of dentin near the DEJ –the mantle dentin. This
of the tested materials. Examination of the DEJ surface zone is also thought to play an important role, functioning
has shown that both enamel and dentin are scalloped as a cushion or gasket that allows the enamel and dentin
in nature22,23). However, the scallop size and shape vary to work together33).
with tooth type; the scallops of posterior teeth are larger Enamel consists of outer enamel and inner enamel.
than those of anterior teeth24). Whittaker reported that Outer enamel is close to the tooth surface with the long
the DEJ was more scalloped in the occlusal portion of axes of the enamel rods being straight and parallel.
the molars, became less scalloped in the buccal region, Inner enamel is close to the DEJ where the enamel
and was almost flat and relatively non-scalloped in the rods are interwoven or decussated with a smooth
cervical region25). Another study found that the width of transition between both enamel types34). In this study,
the DEJ was 13 µm in the occlusal region and 6 µm at crack propagation occurred and ran at an angle of
the cervical region; the difference in the width of the DEJ approximately 45° to the DEJ in the inner enamel, and
was confirmed in these positions26). Additionally, the the cracks were deflected by 90° in the transitional region
DEJ zone is composed of different organic and inorganic between the inner and outer enamel. Three-point bending
components at the occlusal and cervical positions26). experiments were performed in a previous study35). In
It has been suggested that the DEJ scallops provide agreement with our study, the results revealed that
increased surface area and thus reduce interfacial within the enamel layer, cracks propagated at 45° to the
stress concentration27). Therefore, structural differences sample surface and were deflected by 90° in the inner
between the occlusal and cervical DEJ specimens and outer enamel. In the inner enamel, oblique crack
would have resulted in different fracture patterns. propagation was not caused by the orientation of the
The differences might also be related to differences in enamel rods. They found that in the inner enamel layer,
function, with the occlusal surfaces exposed to greater rods no longer determine the crack paths because they
loads during mastication than the cervical area26). are decussated. Furthermore, the structure from outer
This study confirmed that the enamel above the DEJ to inner enamel increased in toughness36). Decussated
 Dent Mater J 2020; 39(2): 272–277 277

enamel plays an important role in the direction of crack stress in a tooth. J Dent Res 1993; 72: 1526-1531.
extension37). Therefore, the crack resistance changes 16) Carvalho RM, Fernandes CA, Villanueva R, Wang L, Pashley
DH. Tensile strength of human dentin as a function of tubule
from outer to inner enamel. Thus, in the lifetime of a
orientation and density. J Adhes Dent 2001; 3: 309-314.
tooth, the enamel transitions are also shielded from 17) Carvalho RM, Santiago SL, Fernandes CAO, Suh BI, Pashley
tooth failure. DH. Effect of prism orientation on tensile strength. J Adhes
Dent 2000; 2: 251-257.
18) Lertchirakarn V, Palamara JE, Messer HH. Anisotropy of
CONCLUSIONS tensile strength of root dentin. J Dent Res 2001; 80: 453-456.
The mechanical properties of tooth structure are 19) Inoue T, Takahashi H, Nishimura, F. Anisotropy of tensile
strengths of bovine dentin regarding dentinal tubule
important for understanding tooth fractures. Failures
orientation and location. Dent Mater J 2002; 21: 32-43.
did not occur in the DEJ subjected to tensile loading 20) Wallace JA. Tooth chipping in the australopithecines. Nature
perpendicular to the DEJ. The enamel above the DEJ is 1973; 244: 117-118.
weaker than the DEJ and the dentin below the DEJ. 21) Constantino P, Lee JJ-W, Chai H, Zipfel B, Ziscovici C, Lawn
BR, et al. Tooth chipping can reveal the diet and bite forces of
fossil hominins. Biol Lett 2010; 6: 719-722.
ACKNOWLEDGMENTS 22) Sela J, Sela M, Lustmann J, Ulmansky M. Dentinoenamel
junction area of a resorbing permanent incisor studied by
This work was supported by a Grant-in-Aid for means of scanning electron microscopy. J Dent Res 1975; 54:
Challenging Exploratory Research and a Grant-in-Aid 110-113.
for Scientific Research (C) from the Japan Society for the 23) Oliveira CA, Bergqvist LP, Line SR. A comparative analysis
Promotion of Science. of the structure of the dentinoenamel junction in mammals. J
Oral Sci 2001; 43: 277-281.
24) Brauer DS, Marshall GW, Marshall SJ. Variations in human
REFERENCES DEJ scallop size with tooth type. J Dent 2010; 38: 597-601.
25) Whittaker DK. The enamel-dentine junction of human and
1) Craig RG, Peyton FA. The micro-hardness of enamel and
Macaca irus teeth. A light and electron microscopic study. J
dentin. J Dent Res 1958; 37: 661-668.
Anat 1978; 125, 323-335.
2) Wang R, Weiner S. Strain-structure relations in human teeth
26) Xu C, Yao X, Walker MP, Wang Y. Chemical/molecular
using Moiré findings. J Biomech 1998; 31: 135-141.
structure of the dentin-enamel junction is dependent on the
3) Rasmussen ST. Fracture properties of human teeth in
intratooth location. Calcif Tissue Int 2009; 84: 221-228.
proximity to the dentinoenamel junction. J Dent Res 1984;
27) Lin CP, Douglas WH. Structure-property relations and crack
63: 1279-1283.
resistance at the bovine dentin-enamel junction. J Dent Res
4) Pioch T, Staehle HJ. Experimental investigation of the shear
1994; 73: 1072-1078.
strengths of teeth in the region of the dentinoenamel junction.
28) Alassaad SS. Approaches to managing asymptomatic enamel
Quintessence Int 1996; 27: 711-714.
and dentin cracks. Gen Dent 2014; 62: 58-62.
5) Lin CP, Douglas WH, Erlandsen SL. Scanning electron
29) von Ebner V. Uber die histologischen veranderungen des
microscopy of type I collagen at the dentin-enamel junction of
zahnschmelzes wahrend der erhartung, inbesondere beim
human teeth. J Histochem Cytochem 1993; 41: 381-388.
menschen. Arch Mikrosk Anat 1906; 67: 18-81.
6) Dong XD, Ruse ND. Fatigue crack propagation path across
30) von Korff K. Die entwicklung der zahnbeingrundsubstanz der
the dentinoenamel junction complex in human teeth. J
saugetiere. Arch Mikrosk Anal 1906; 67: 1-17.
Biomed Mater Res 2003; 66A: 103-109.
31) Elliott JC, Anderson P, Gao XJ, Wong FSL, Davis GR, Dowker
7) White SN, Miklus VG, Chang PP, Caputo AA, Fong H,
SEP. Application of scanning microradiography and X-ray
Sarikaya M, et al. Controlled failure mechanisms toughen
microtomography to studies of bones and teeth. J X-Ray Sci
the dentino-enamel junction zone. J Prosthet Dent 2005; 94:
Technol 1994; 4: 102-117.
330-335.
32) Staininec M, Marshall GW, Hilton JF, Pashley DH, Gansky
8) Marshall GW Jr, Balooch M, Gallagher RR, Gansky SA,
AS, Marshall SJ, et al. Ultimate tensile strength of dentin:
Marshall SJ. Mechanical properties of the dentinoenamel
Evidence for a damage mechanics approach to dentin failure.
junction: AFM studies of nanohardness, elastic modulus, and
J Biomed Mater Res (Appl Biomater) 2002; 63: 342-345.
fracture. J Biomed Mater Res 2001; 54: 87-95.
33) Zaslansky P, Friesem AA, Weiner S. Structure and mechanical
9) Imbeni V, Kruzic JJ, Marshall GW, Marshall SJ, Ritchie RO.
properties of the soft zone separating bulk dentin and enamel
The dentin-enamel junction and fracture of human teeth. Nat
in crowns of human teeth: insight into tooth function. J Struct
Mater 2005; 4: 229-232.
Biol 2006; 153: 188-199.
10) Urabe I, Nakajima S, Sano H, Tagami J. Physical properties
34) Ten Cate AR. A fine structural study of coronal and root
of the dentin-enamel junction region. Am J Dent 2000; 13:
dentinogenesis in the mouse: observations on the so-called
129-135.
‘von Korff fibers and their contribution to mantle dentine. J
11) Giannini M, Soares CJ, de Carvalho RM. Ultimate tensile
Anat 1978; 125: 183-197.
strength of tooth structures. Dent Mater 2004; 20: 322-329.
35) Bechtle S, Fett T, Rizzi G, Habelitz S, Klocke A, Schneider
12) Inoue T, Miyazaki T, Nishimura F. Tensile strength and
GA. Crack arrest within teeth at the dentinoenamel junction
durability of bovine dentin. Dent Mater J 2007; 26: 348-354.
caused by elastic modulus mismatch. Biomaterials 2010; 31:
13) Inoue T, Nishimura F, Debari K, Kou K, Miyazaki T. Fatigue
4238-4247.
and tensile properties of radicular dentin substrate. J
36) Bajaj D, Arola D. Role of prism decussation on fatigue crack
Biomech 2011; 44: 586-592.
growth and fracture of human enamel. Acta Biomater 2009;
14) Inoue T, Saito M, Yamamoto M, Nishimura F, Miyazaki T.
5: 3045-3056.
Relation between incremental lines and tensile strength of
37) Yahyazadehfar M, Bajaj D, Arola D. Hidden contributions of
coronal dentin. Dent Mater J 2012; 31: 541-548.
the enamel rods on the fracture resistance of human teeth.
15) Spears IR, van Noort R, Crompton RH, Cardew GE, Howard
Acta Biomater 2013; 9: 4806-4814.
IC. The effects of enamel anisotropy on the distribution of