This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/authorsrights
Transcript
This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies areencouraged to visit:
Revealing the structural and mechanicalcharacteristics of ovine teeth
Simona O’Briena, Amanda J. Keownb, Paul Constantinoc, Zonghan Xied,Mark B. Bushb,n
aPerth Institute of Business and Technology, Edith Cowan University, Joondalup, WA 6027, AustraliabSchool of Mechanical and Chemical Engineering, The University of Western Australia, Crawley, WA 6009, AustraliacDepartment of Biology, Marshall University, Huntington, WV, USAdSchool of Mechanical Engineering, The University of Adelaide, North Terrace, SA 5005, Australia
a r t i c l e i n f o
Article history:
Received 1 August 2013
Received in revised form
10 November 2013
Accepted 12 November 2013
Available online 19 November 2013
Keywords:
Ovine
Sheep
Molar
Tooth properties
Enamel
Dentin
Indentation
a b s t r a c t
The survival and function of dentition over the lifetime of an animal depends upon the
ability of the teeth to resist wear and chemical erosion, and to withstand occlusal loading
conditions without suffering debilitating fracture. Understanding how geometrical factors
(radius, height, enamel thickness) and mechanical properties of the dental tissues (Young's
modulus E, hardness H and toughness KIC of enamel and dentin) combine to ensure the
survival of an animal's teeth can provide great insight into the evolutionary history of the
animal and its dietary adaptation. While the geometrical factors are beginning to be
understood, the range of animals for which measurements of dental tissue properties are
available is very narrow, being restricted almost entirely to humans and other primates.
The absence of comparative data across a broader range of species makes it impossible to
draw conclusions with any certainty. The present study expands knowledge of mamma-
lian dental tissue properties by reporting the Young's modulus and hardness of ovine
(sheep) enamel and dentin measured using nano-indentation.
We found that sheep molar enamel Young's modulus and hardness are both lower than
those of human enamel, by approximately 30%, and 9% respectively, while the properties
of dentin are similar. The combination of E and H makes the ovine enamel approximately
30% more resistant to wear than human enamel, which is an imperative in ruminant
dentition. The results of this study are interpreted in terms of the ovine feeding ecology,
and the structure of the ovine molar and its occlusal surface.
& 2013 Elsevier Ltd. All rights reserved.
1. Introduction
Mammalian teeth take many different forms, shapesand sizes representing adaptations to the task of collectingand processing a wide range of foodstuffs. The ability of the
animal to satisfy its nutritional requirements over a lifetimedepends on the capacity of its teeth to function withoutsuffering debilitating damage. Primates, for example, haverounded ‘bunodont’ molars, with a tooth height to radiusratio typically in the range 1–1.5, and a relatively thick
1751-6161/$ - see front matter & 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.jmbbm.2013.11.006
j o u r n a l o f t h e m e c h a n i c a l b e h a v i o r o f b i o m e d i c a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1 7 6 – 1 8 5
Author's personal copy
enamel layer. Bunodont teeth are also found in otters, pigsand some other mammals (Janis and Fortelius, 1988;Popowics et al., 2001, 2004; DeGusta et al., 2003; Lucas,2004). The hard, stiff enamel provides the structural resis-tance to withstand high bite forces, and the tooth shapeprovides a crushing and grinding capability to process avariety of foodstuffs, from tough, soft foods (like raw meator fibrous fruits) to hard, brittle foods (such as nuts). Theenamel is brittle and susceptible to fracture when overloaded,however the tooth structure and the mechanical properties ofthe enamel act to contain damage. The mechanics of suchfracture is now well documented (Lucas et al., 2008; Chaiet al., 2009, 2011; Lawn and Lee, 2009; Constantino et al., 2010,2011; Lee et al., 2010; Barani et al., 2011; Keown et al., 2012).
Understanding how geometrical factors (radius, height,enamel thickness) and mechanical properties of the toothcomponents (Young's modulus, hardness and toughness)affect the load bearing capacity and wear resistance canprovide great insight into the evolutionary history of theanimal and its dietary adaptation (Janis and Fortelius, 1988;Lucas et al., 2008; Constantino et al., 2011, 2012). Simple butpowerful relationships have been developed for estimatingfracture loads in primate molar teeth, which may be used toinfer bite loads (Constantino et al., 2010; Barani et al., 2011,2012; Chai et al., 2011; Keown et al., 2012). These relationshipshighlight the importance of tooth radius and enamel thick-ness as factors in resisting longitudinal cracking.
The mechanical properties of dental tissues also play arole in supporting the required bite loads, while resistingfracture. In the case of primates, it has been found that themodulus and hardness of enamel are remarkably uniformacross species (Constantino et al., 2012). Changes in loadbearing capacity have therefore largely been achieved byvarying tooth size (radius) and enamel thickness. For exam-ple, both the gorilla and orangutan have enamel and dentinwith properties similar to those of humans, yet the bite loadcapability is considerably higher. The higher occlusal loadsare sustained by the considerably larger radius of the teeth.
The California sea otter also possesses a bunodont molarshape, but uses its teeth primarily to break down shellfish.Preliminary tests indicate that otter enamel may be less stiffand less hard than human enamel, but with higher tough-ness (Constantino et al., 2011). Furthermore, the averageenamel thickness is half that of humans and the first molarradius is 50% greater. Interestingly, this combination ofproperties and geometry has produced a structure capableof withstanding bite loads similar to that of humans. Whilethe differences in material properties between otters andprimates may be a result of phylogeny rather than function,the combination of geometry and properties provides effec-tively similar capabilities for survival.
A complete understanding of tooth form and functiontherefore requires knowledge of how tooth geometry influ-ences its capacity to withstand load, but interpreted in thelight of the specific material properties of the tooth compo-nents. Nonetheless, the range of animal species for whichmechanical properties of dentin and enamel has beenreported is remarkably narrow. Significant information isnow available on the properties of tooth tissues in humans(Cuy et al., 2002; Bajaj and Arola, 2009; Constantino et al.,
2012), a range of other primates (Lee et al., 2010; Constantinoet al., 2012) and the California sea otter (Constantino et al.,2011). The only ruminant for which mechanical properties ofdental tissue have been reported is the cow (Ang et al., 2010;Bechtle et al., 2010a, 2010b, 2010c).
Most studies have produced measurements of modulusand hardness, generally obtained by micro-indentation. Mea-surements of toughness are more problematic and lesscommon. Micro-indentation can be used to obtain point-by-point measurements of toughness (Anstis et al., 1981; Imbeniet al., 2005; Constantino et al., 2011). However, perhaps themost comprehensive study of toughness gradients in humanenamel was undertaken using compact tension specimenscut from human molar enamel (Bajaj and Arola, 2009). Asimilar approach has been used to measure toughness ofbovine incisor enamel, by applying bending to a notchedmicro-beam cut from the enamel (Ang et al., 2010; Bechtleet al., 2010a, 2010b, 2010c). The modulus of the enamel wasalso measured as an adjunct to determining the toughness,and was obtained using beam bending or uniaxial compres-sion techniques applied to a segment of the enamel. Thesemethods provide a measure of the average modulus of thebulk material, which can differ considerably from the localproperty measured by indentation. As a result, these bovineenamel modulus measurements cannot be compared directlywith the indentation results from other animals, renderingcross species comparisons problematic.
The present study expands knowledge of non-primatedental tissue properties by measuring elastic modulusand hardness of ovine enamel and dentin using nano-indentation. We compare the measured properties with thoseof other animals and discuss the biological implications ofthe observed properties.
2. The structure and function of theovine (sheep) molar
The ovine molar tooth has a columnar structure, consisting oftwo or three lobes fused together to form one tooth. Aphotograph of a typical molar used in this study is shownin Fig. 1. A diagram of the cross section is given in Fig. 2a anda photograph of an actual cross section in Fig. 3.
The enamel not only encloses the dentin, but also pene-trates the dentin body to form internal enamel walls sepa-rated by a thick layer of cementum (Every et al., 1998). Theresult is a complex cutting and grinding occlusal surfacemade up of substances with differing properties (enamel,dentin and cementum) and featuring sharp projections of theharder enamel (‘shearing crests’). The average lobe radiusmeasured on the molar teeth used in this study was 4.1 mmand the average exterior enamel thickness was 0.58 mm. Theenamel was notably thicker on the buccal side of the tooth(0.73 mm) compared to the lingual side (0.45 mm). The thininternal enamel wall on the buccal side of the cementumcavity had an average thickness of just 0.1 mm. The wall onthe lingual side on the cavity was 0.53 mm thick.
Sheep and other ruminant animals, such as cattle anddeer, have a much more specialized diet (grasses) than mostprimates, yet they lack the enzymes to break down cellulose.
j o u r n a l o f t h e m e c h a n i c a l b e h a v i o r o f b i o m e d i c a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1 7 6 – 1 8 5 177
The plant material must be divided into small particles toexpose the cellulose to gut microbes that are capable ofconverting cellulose into products that the ruminant canutilize (Jurgens et al., 2012). This is achieved through repeatedprocessing of the food bolus (‘cud’) by regurgitation from thegut to the mouth when the animal is resting (‘cudding’). Thebite load does not need to be high. Instead, the cud issubdivided by repeated slicing and grinding. Such contactprocesses incur high rates of tooth wear through tooth-to-tooth contact (attrition) and the action of abrasive plant
material, such as phytoliths, and ingested sand particles.However, wear is a necessary element of optimal toothfunction. Wear on the occlusal surfaces produces sharpcutting edges that are required to process the food, but thewear rate must not be so high that the functional life of thetooth is shorter than that required for the animal to matureand reproduce.
In response to this high wear rate, the ruminant molar hastaken a ‘hypsodont’ form—the adult tooth is tall (height/radius410), with most of the tooth height residing initiallybeneath the gum line in the jaw. As the tooth wears, thetooth is continuously pushed out of the bone to maintain anapproximately constant exposed tooth height, thus helpingto preserve proper occlusion between the upper and lowerteeth (Every et al., 1998). The process of tooth ejection fromthe bone appears to be triggered by the lessening of contactforces between the opposing teeth, as would occur when theteeth wear, and is probably driven by bone formation (Wiseet al., 2007). A similar phenomenon is evident in humans andother animals when a molar has been lost to decay orother damage—the opposing molar may erupt partly fromits socket.
Unlike bunodont teeth found in primates, longitudinalfracture of enamel is not a limiting factor in the function ofovine teeth. In fact, the elongate form of the hypsodont toothhas been shown to provide protection against such fracture(Barani et al., 2012). Instead, sheep tooth functionality islimited by the rate of wear. In view of this observation, someinteresting questions can be posed: What are the mechanical
Fig. 1 – The ovine second molar, extracted from the rightside of the mandible. Note the double lobed structure, theabsence of an enamel crown and the projected sharp edgeson the remaining enamel at the occlusal surface. The brokenline indicates a typical vertical cutting plane used to sectionthe tooth.
lingualside
buccalside
Enamel Dentin
enamel
dentin
cementum
OES EDJ
a
b
c
d
e
f
g
h
0.5 mm
Fig. 2 – Indentation geometry. (a) Approximate locations ofthe indentation clusters on a tooth cross-section, labelled a–h. (b) A typical cluster of indentations ranging from the outerenamel surface (OES) across the enamel–dentin junction(EDJ) and into the dentin.
Buccalside
Lingualside
Enamel
Dentin
Cementumcavity
Enamel
Dentin
1 mm
0.5 mm
0.5 mm
Fig. 3 – Optical microscopy image of the top section of thesheep second molar, sectioned along the buccal to lingualplane indicated on Fig. 1.
j o u r n a l o f t h e m e c h a n i c a l b e h a v i o r o f b i o m e d i c a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1 7 6 – 1 8 5178
properties of the ovine tooth components and how do theycompare with primate tooth tissues (including human)? Asthe ovine tooth does not need to withstand high occlusalloads, is this reflected in its relative properties? Is the ovineenamel better able than primate enamel to resist wear?
3. Materials and methods
3.1. Sample collection and preparation
Ovine teeth were collected postmortem from an adult researchanimal (Ovis orientalis aries var. ‘merino’) at the Royal PerthHospital, Western Australia, with animal ethics committeeapproval for tissue reuse. A fresh mandible was removed within30min of the death of the animal. The mandible was frozen forone week, after which the molar teeth were extracted, cleaned ofsoft tissue and stored at 4 1C in HBSS (Hanks’ balanced saltsolution—SIGMA-Aldrich Co., St. Louis, USA) with the addition of0.02% thymol crystals to prevent demineralization (Habelitzet al., 2002) and bacterial growth (White et al., 1994). Themajorityof tests were performed on the teeth of this animal (Animal 1).For purposes of comparison, tooth tissue samples were alsocollected from a second animal (Animal 2) on a later occasion.Animal 2 was from a different group of sheep, but raised on thesame diet and was approximately the same age as Animal 1.
Teeth for microscopy and general handling were set in aspecially designed resin/putty construct to provide a base formanipulation in their hydrated state. Epoxy resin and hardener(EpoFix, Struers A/C, Copenhagen, Denmark) were mixed at aratio of 15 to 2 by volume to form cylindrical blocks (30mmdiameter, 10mm height) in plastic moulding cups (FixiForm,Struers). A hole of approximately 15mm diameter was cut alongthe centreline of the cured block to form a hollow cylinder thatwas then partly filled with aquatic putty (Selleys Knead-It Aqua,Selleys, Australia). This putty is able to polymerise in thepresence of water. The root of the tooth was mounted in theputty and the whole structure immersed in the HBSS/thymolsolution at 4 1C, and left for at least 60min to ensure completepolymerisation of the putty.
The right side second and third molars of Animal 1 andthe third molar of Animal 2 were selected for indentationinvestigation. The teeth were sectioned using a precision saw(Isomet 1000, Buehler Ltd., Lake Bluff, IL, USA) along a buccal-to-lingual line that intersected each cusp at its highest point(Fig. 1). Four different sections were selected for polishing andnano-indentation, specifically:
Specimen 1: The mesial side of the mesial lobe on thesecond molar, Animal 1.Specimen 2: The distal side of the mesial lobe of the thirdmolar, Animal 1.Specimen 3: The mesial side of the mesial lobe of the thirdmolar, Animal 1.Specimen 4: The mesial side of the mesial lobe of the thirdmolar, Animal 2.
Specimens 2 and 3 were extensively tested at severallocations on the polished section to allow comparison of
the properties at different positions on a given tooth section,while Specimens 1 and 4 provide the opportunity to assessdifferences between teeth within the same animal or acrossanimals. The approximate locations of the indentation sitesare illustrated in Fig. 2a.
After sectioning, the surface was ground with Struerssilicon carbide (SiC) papers of 320 and 500 grit and polishedwith papers of 1200 and 4000 grit. The final polishing of thetooth specimens was carried out using a soft polishing clothwith water (MD NapR, Struers, Copenhagen, Denmark). Aftereach grinding and polishing step, the surface was examinedusing an optical microscope (Trinocular Metallurgical Micro-scope, Brunel SP-200-XM, Brunel Microscopes Ltd, UK). Oncesatisfactorily polished, a second cut, parallel to the first, wasmade at a depth of 5 mm beneath the polished surface toproduce a 5 mm thick sample to fit into the nano-indentationdevice. HBSS was used as a coolant during cutting. Thespecimen was then stored in HBSS/thymol solution at 4 1Cprior to testing.
Three human third molar teeth extracted from patientsaged 20–30 years were also processed in a similar way,sectioned along the mesial to distal plane, to allow a compar-ison to be made between the structure and properties ofovine and human molars. The teeth were extracted forclinical reasons and on the treating dentist's advice at aprivate dental practice and collected for this study in accor-dance with a protocol approved by the Edith Cowan Uni-versity Human Research Ethics Review Committee and withthe informed consent of the patients.
3.2. Nanoindentation
The specimen was placed in a holder that would allow fortesting in a wet environment (as before, HBSS with theaddition of 0.02% thymol crystals). A depth-sensing indenta-tion system (Ultra-Micro Indentation System, UMIS-2000,CSIRO, Australia) was used in conjunction with a Berkovichindenter. The contact area of the tip as a function ofpenetration depth was first calibrated on fused silica (Oliverand Pharr, 1992, 2004). The indentation studies on the teethwere carried out in rows of 5 indents with an interval of50 mm between indents and rows (Fig. 2b). Indentationstraversed from the outer enamel surface (OES) to theenamel–dentin junction (EDJ) and into the dentin. The prop-erties reported at a given distance from the OES are theaverage of the values resulting from the 5 indents in a givenrow. In this way, variations in measured properties resultingfrom differences in the local microstructure at the indenta-tion sites are minimised.
The tests were run under controlled load conditions.A maximum load of 400 mN was applied in 8 increments.Following each increment there were 10 decrements. In thisway, the Young's modulus and hardness were determined ateach increment and partial decrement (Oliver and Pharr,1992, 2004), allowing the properties to be described as afunction of indentation depth.
The Young's modulus and hardness measurements werefound to be dependent on the indentation depth, ht. Specifi-cally, both properties declined with increasing indentationdepth. This depth effect is well known and may result from
j o u r n a l o f t h e m e c h a n i c a l b e h a v i o r o f b i o m e d i c a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1 7 6 – 1 8 5 179
Author's personal copy
the hierarchical structure of biological materials (Bechtleet al., 2010a) and/or evolution of microstructure with increas-ing indentation depth (Zhou and Hsiung, 2007). Previousstudies on teeth conducted by indentation reported resultsat an indentation depth of 0.4 mm (Cuy et al., 2002;Constantino et al., 2011, 2012). For consistency with thoseprevious studies, we have also reported results for ht¼0.4 mm.
3.3. Wear parameter
The relative resistance to abrasion by plastic deformationmechanisms (mild wear) and by surface micro-cracking(severe wear) can be inferred using parameters derived fromthe basic material properties, Young's modulus, hardnessand toughness: E, H and KIC. Both H and E have been shown tohave an important influence on wear behaviour of materials(Oberle, 1951; Leyland and Matthews, 2000; Zok and Miserez,2007). The elastic strain to failure ranking parameter, H/E,also referred to as a plasticity index or mild wear resistanceparameter, can be used to assess the resistance to surfacematerial loss through plastic deformation produced by abra-sion by hard particles (Leyland and Matthews, 2000; Roy et al.,2010). This is likely to be an important mode of wear in ovineteeth due to the transverse grinding movements of ovinejaws. More severe wear associated with surface micro-cracking and chip formation occurs where there is significantocclusal contact loading. The relevant parameter is KIC
4 /(HE2)(Miserez et al., 2008; Weaver et al., 2010). Specifically, thecritical load to produce cracks at the point of indentation by ahard particle is proportional to this parameter.
Although these parameters cannot replace direct quanti-tative measurements of wear resistance, they may be used toinfer the relative resistance to wear of materials with similarmicrostructure. The larger the value of the parameter, thegreater the potential resistance to wear. In the absence ofmeasured toughness of ovine enamel, we must rely on theelastic strain to failure ranking, H/E.
3.4. Optical microscopy
Detailed images of ovine tooth microstructure were collectedusing an optical microscope (Olympus PMG3). High magnifi-cation images of small sections of the surface were combinedto produce a detailed composite picture of the tooth section.The sample used for optical microscopy was the distal side ofthe mesial lobe on the second right molar. Similar imageswere also collected from a human molar for comparativepurposes.
4. Results
4.1. Optical microscopy
Composite images of a typical ovine molar are presented inFig. 3, illustrating the columnar nature of the tooth. Forcomparative purposes, a similar illustration is presented inFig. 4 for a human molar, although the cross-sectional cut inthis case is on the mesial-to-distal plane.
The enamel layers are clearly visible in both figures. Theenamel in the human molar is significantly thicker, havinga maximum thickness of 2.3 mm at the occlusal surfacecompared with the maximum of 0.76 mm on the buccal wallof the ovine tooth. The enamel walls are relatively uniform inthickness along the length of the ovine tooth, and extend allthe way to the root, which is embedded within the bone ofthe jaw (not shown in Fig. 3). In contrast, the enamel on thehuman molar thins toward the base of the tooth and ends atthe gum line. Most of the tooth below this point consists ofdentin covered by a thin layer of the bone-like cementum, towhich the periodontal ligament attaches.
The cementum cavity at the centre of the ovine toothis visible in Fig. 3. The soft cementum was removed by thepolishing process, leaving an empty cavity. Pulp chamberswere also observed toward the base of each lobe (not visiblein Fig. 3).
Small flaws, or ‘tufts’, can be observed in human molarenamel, originating at the EDJ and extending a short distanceinto the enamel. Examples of tufts can be seen in the detail ofFig. 4. Tufts are incipient flaws that form part of the naturalstructure of the tooth (Osborn, 1969; Lucas, 2004). No evidenceof similar structural flaws was found in the sheep teeth.
4.2. Young's modulus and indentation hardness
The Young's modulus measured across the thickness of theenamel and some distance into the dentin in Specimens
Enamel
Dentin
Tufts
1 mm
Mesialside Distal
side
0.1 mm
0.2 mm
Fig. 4 – Optical microscopy image of a human molar,sectioned along the mesial to distal plane.
j o u r n a l o f t h e m e c h a n i c a l b e h a v i o r o f b i o m e d i c a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1 7 6 – 1 8 5180
Author's personal copy
1 and 2 is shown in Fig. 5. Results from Specimens 3 and 4 arepresented in Fig. 6. The majority of traces were conducted onSpecimens 2 and 3, in each of the regions a-h indicated inFig. 2, with representative traces conducted on Specimen 1(one trace) and Specimen 4 (two traces) for comparison. Somevariations in the modulus are evident between indentationtraces, however this is typical of indentation studies on toothenamel, where local variations in microstructure and surfacefinish affect the measurement obtained from a single indentand introduce scatter into the results. For this reason theaverage of several traces is usually presented. Similar studieson human teeth (Cuy et al., 2002) have attributed variationacross the enamel section and between positions on a giventooth to changes in chemical composition and degrees ofmineralisation. We have provided the individual traces inthis instance in order to identify any consistent trendsassociated with position on the tooth, between teeth orbetween animals.
A consistent decline of E in the enamel with increasingdistance from the OES is evident in all specimens (Figs. 5and 6). A clear transition in properties occurs across the EDJ.While most specimens exhibited a continuing decline of Ewithin the dentin, some did not. However, in these speci-mens, after the initial rise E proceeded to decline again atdeeper locations.
Figs. 5 and 6 do not suggest a strong association betweenthe measured property and the position on the tooth.Although Fig. 5 does suggest that the modulus at positionsclose to the gum (blue curves) tend to be higher than otherpositions (green and orange curves), this is not reflected inFig. 6. Likewise, we see no consistent trend between the
lingual to buccal locations on the tooth. Furthermore, thetrace on a second tooth in the same animal (Animal 1,Specimen 1, Fig. 5) falls within the scatter of the results fromthe other tooth (Specimen 2). Equally, the traces from thetooth on the second animal (Animal 2, Specimen 4) fall withinthe scatter of Specimen 3 (Fig. 6).
The behaviour of the measured hardness values wassimilar to that of the modulus. i.e. no significant differenceswere associated with position on the tooth, between teeth orbetween animals. We have plotted the averaged values of themeasured modulus (Fig. 7) and hardness (Fig. 8), showing theaverages on each specimen (except Specimen 1) and theaverage of all measurements. It is this last curve that canbe compared with previously reported average values forother animals. Also shown on the figures is the standarddeviation of values around that mean position.
Fig. 5 – The Young's modulus (E) of ovine enamel and dentinfrom Specimens 1 (empty symbols) and 2 (filled symbols).The graphs show the variation of measured properties fromthe outer enamel surface (OES) across the enamel–dentinjunction (EDJ) and into the dentin. The distance is nomalisedrelative to the enamel thickness. The square symbolscorrespond to traces located on the lingual side of the tooth(Fig. 2a, regions a–c), triangles refer to the internal enamelwall (d and e) and circles refer to the buccal side (f and h).The green traces are at locations close to the occlusal surface(a, d, f), orange corresponds to mid locations (b, e, g) and blueto the lower locations (c and h). (For interpretation of thereferences to color in this figure legend, the reader isreferred to the web version of this article.)
Fig. 6 – The Young's modulus (E) of ovine enamel and dentinfrom Specimens 3 (filled symbols) and 4 (empty symbols).The graphs show the variation of measured properties fromthe outer enamel surface (OES) across the enamel–dentinjunction (EDJ) and into the dentin. The distance is nomalisedrelative to the enamel thickness. See caption to Fig. 5 forexplanation of the symbols and colours.
Fig. 7 – The Young's modulus (E) averaged over eachspecimen (grey curves, specimens identified by numbers)and over all specimens (dashed heavy black curve).The dashed light lines indicate the standard deviationaround the all-specimen curve.
j o u r n a l o f t h e m e c h a n i c a l b e h a v i o r o f b i o m e d i c a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1 7 6 – 1 8 5 181
Author's personal copy
4.3. Wear parameter
Values of the mild wear parameter (H/E) are given in Table 1.Although the measured Young's modulus and hardness ofovine enamel are both lower than those of human enamel, byapproximately 30% and 9% respectively, the wear parameterratio is 32% higher in the ovine, suggesting that the ovineenamel potentially exhibits enhanced resistance to mildabrasive wear.
5. Discussion
5.1. Observations on microstructure
The inset images of Figs. 3 and 4 suggest that the micro-structure of the ovine enamel and dentin are both super-ficially similar to human teeth. A decussation pattern isevident in the enamel, indicated by the light and darkbanding (‘Hunter-Schreger bands’) present in both the ovine(Fig. 3) and human enamel (Fig. 4). Tubules are clearly presentin the ovine dentin, and appear to change orientation acrossthe width of the tooth. These features echo the microstruc-ture of the teeth of humans and other mammals (Lucas,2004).
Enamel consists of mineralized rods, bound together by aweaker protein interlayer (Lucas, 2004). In human molars therods are arranged in bundles and aligned to span the thick-ness of the enamel, intertwining to form a complex three-
dimensional structure (decussation), with more pronounceddisorientation near the EDJ. The local toughness of theenamel increases from the outer surface to the EDJ (Bajajand Arola, 2009) which in part can be attributed to decussa-tion. Decussation patterns are also evident in the enamel ofother primates, the sea otter (Constantino et al., 2011) andcattle (Bechtle et al., 2010b). The decussation in ovine enamelof Fig. 3 appears to be more evenly distributed across theenamel thickness compared with human enamel (Fig. 4),suggesting that the ovine enamel may exhibit a more uni-form resistance to crack propagation across its thickness.Although fracture behavior of ovine enamel remains to bestudied, it is reasonable to expect that the structure of bovineenamel will be similar. Bechtle et al. (2010b, 2010c) found thetoughness of bovine enamel to be significantly higher thanthat of human enamel, by as much as a factor of two. Theyalso found that bovine enamel was slightly tougher thanbovine dentin. Interestingly, if the toughness of ovine enamelis also found to be higher than that of human enamel, thenthe severe wear parameter referred to in Section 4.3 will becorrespondingly higher, indicating that the enamel is moreresistant to both mild abrasive and severe fracture relatedwear modes. It might therefore be postulated that increasedtoughness and reduced hardness and modulus in ruminantenamel relative to that of primates may be directly linked tothe need for wear resistance, given the nature of the diet andthe way it is processed by the animal (cudding).
The apparent absence of tufts in the ovine enamel isinteresting. They were not observed in any of the teethinvestigated in this study. In human teeth tufts appear tobe small fissures formed during the growth of the tooth,subsequently filled with mineralised matter, but nonethelessresulting in a weakness in the enamel structure near the EDJ.Tufts have also been observed in the enamel of otherprimates and the sea otter, animals that must apply highocclusal loads (Constantino et al., 2011). There is evidencethat tufts can act as sites from which cracks begin to growand extend across the enamel thickness under sufficientlyhigh occlusal loading (Chai et al., 2009). However, the pre-sence of multiple cracks can result in ‘shielding’, whereby thetotal strain is distributed over the crack system and the stressat any single crack tip is reduced. Higher load is then neededto drive a crack to failure. While an individual tuft representsa local weakness in the enamel, an array of tufts can act tocontain or delay catastrophic damage. Tufts may then beadaptive against failure by catastrophic fracture. The ovinemolar tooth is not subjected to high occlusal loading, and itselongated structure provides natural protection against thegrowth of longitudinal cracks (Barani et al., 2012). Theabsence of tufts in ovine teeth may simply be because theyare not needed to protect against crack propagation. The roleof tufts and the implications of their presence or absence invarious species remains to be fully investigated.
5.2. Properties of ovine enamel and dentin—Biologicalimplications
In Fig. 9 we compare the measured H and E of ovine enamelwith results for a range of other species available in theliterature. Each curve represents the average behaviour taken
Fig. 8 – The hardness (H) averaged over each specimen (greycurves, specimens identified by numbers) and over allspecimens (dashed heavy black curve). The dashed lightlines indicate the standard deviation around the all-specimen curve.
Table 1 – The wear parameter values at the OES for ovineand human enamel, calculated using average values of Eand H based on N measurements.
Stiffness Hardness Wear parameter
N E (GPa) H (GPa) H/EOvine 16 80 4.3 0.054Human 3 116 4.7 0.041
j o u r n a l o f t h e m e c h a n i c a l b e h a v i o r o f b i o m e d i c a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1 7 6 – 1 8 5182
Author's personal copy
from the various reports for each species. All measurementswere obtained using indentation. The results for primates(green curves), are a representative subset of a larger group ofprimates studied by Constantino et al. (2012). The non-human primate studies were conducted on dry specimens,while the human teeth had been kept hydrated since extrac-tion and tested in the hydrated state. The degree of hydrationand age can affect specific results. In particular, the indenta-tion modulus has been observed to increase in the dry statecompared to the wet state (Cuy et al., 2002; He and Swain,2007). Nonetheless, the data on primate teeth (includinghuman) vary by only 11% around the median value for allspecies in Fig. 9).
Other results from humans are also shown in Fig. 9 (Cuyet al., 2002; Constantino et al., 2011), including the measure-ments made as part of this study. All of these studies werecarried out on teeth that had been kept in a hydrated statesince extraction, with the exception of the study by Cuy et al.(2002). The close agreement between the results for humanenamel obtained by different research groups on both wetand dry specimens (solid lines on Fig. 9) gives us someconfidence in the consistency of techniques and the relia-bility of the comparisons across species.
The modulus and the hardness of ovine enamel (Figs. 7and 8) exhibit a decline from the OES to the EDJ, in commonwith several of the other animals that have been studied.However, the magnitudes of both E and H are noticeably lessthan those measured on primate teeth (Fig. 9). On the otherhand, the magnitude of the average modulus of the dentin
close to the EDJ (approximately 20 GPa) is very similar to thatreported for humans (Lucas et al., 2008).
The properties of the ovine enamel most closely resemblethose of the sea otter, despite the vast difference in feedingecology and tooth structure between these two groups. Theseobservations raise some obvious central questions: ‘Whydoes sheep enamel exhibit gradients in properties across itsthickness?’ and ‘Why do sheep and sea otters have softerenamel than humans and other primates?’ While we cannotanswer these questions definitively, we offer some tentativediscussion and explanations below. The picture may beclarified in time as mechanical property measurements on awider range of species become available.
The reason for the gradient in properties across theenamel is not clear. Cuy et al. (2002) found the gradient inhuman enamel to be correlated with compositional changesin P2O5 and CaO, which together make up hydroxyapatite—the mineral component of enamel. The stiffer enamel nearthe OES appears to be due mainly to increased mineraldensity, but what is the purpose of this increased mineral-ization? As noted earlier, animals that eat hard foods oftenhave bunodont dentition (low, rounded cusps). In suchanimals, the harder and stiffer outer enamel may offermaximum protection against contact with hard foods, as wellas increasing stress on that food for greater likelihood offracture. At the same time, the decreased modulus at the EDJincreases compliance in that region, possibly to reducetensile forces caused by the stiffer enamel flexing on thesofter dentin during high loading events (Lucas et al., 2008).
However, this does not explain why sheep would featurethe same trend in properties, given that their teeth have notbeen adapted for eating hard foods. The presence of shearingcrests on the occlusal surface of sheep teeth means thatregular processing of hard foods will result in microfracturesand crumbling of the enamel, even if the enamel is harder inthat region compared to the EDJ. Unfortunately, we do not yetknow enough about variations in the mechanical propertiesof enamel across animal groups to determine how phylogen-etically widespread these property gradients may be. Amongprimates, Constantino et al. (2012) found gradients to beprevalent but highly variable in pattern. Some genera (e.g.Macaca) actually exhibit higher stiffness near the EDJ whileothers (e.g. Hylobates) showed no significant difference acrossthe enamel thickness. However, given that only one speci-men per taxon was tested in that study, further work isneeded to understand whether there exists intraspecificvariation in these gradients as well. The lack of comparativedata in this area makes it impossible to draw conclusionswith any certainty.
It seems reasonable to assume that sheep would needhard enamel in order to create and maintain well-definedshearing crests. But very hard materials also tend to bebrittle, and brittle edges are not well suited for this purposebecause the enamel will start to crumble even under smallloads. Crumbled edges should result in reduced herbivorousfeeding efficiency, as many plant materials will becomesnagged on the resulting jagged edges of the shearing crests.In addition, the crumbling will likely lead to a more rapidwear rate that prevents the animal from reaching sexualmaturity prior to dental senescence. Therefore, some level of
You
ng’s
Mod
ulus
, E (G
Pa)
Har
dnes
s, H
(GPa
)
00.8 1.00.60.40.20
0.8 1.00.60.40.20
20
40
60
80
100
120
140
0
2
4
6
8
EDJOESHumanGorillaChimp
GibbonTrue LemurBamboo Lemur
(Constantino et al, 2012)
HumanSea Otter Human (Cuy et al, , 2002)
Sheep Human
Normalised distance
(Constantino et al, 2011)
(Current study)
Fig. 9 – The Young's modulus (E) and hardness (H) of enameltaken from a range of animals. The distance is nomalisedrelative to the enamel thickness.
j o u r n a l o f t h e m e c h a n i c a l b e h a v i o r o f b i o m e d i c a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1 7 6 – 1 8 5 183
Author's personal copy
toughness in the enamel is desired. Increased toughnesscould keep the shearing edges intact by reducing the like-lihood that incipient cracks propagate far enough to result inchipping or spalling of the enamel.
While we were unable to measure toughness directly, it isoften found in reverse proportion to Young's modulus(Constantino et al., 2009). Sheep may have more compliantenamel not only because compliance is potentially advanta-geous in regard to wear resistance, but also because tough-ness is, and evolution is seldom able to maximize bothstiffness and toughness. Again, the one other herbivore forwhich reliable toughness measurements are available, cattle(Bos primigenius), produces enamel is that is considerablytougher than that of primates (Bechtle et al., 2010c). If thisis reflected in ovine enamel, then the relative resistance tosevere wear would be enhanced further.
Why then do sea otters also have softer/more compliantenamel than primates? Unlike sheep, they have bunodontteeth and are well known to process hard foods with theirpost-canines. Yet they have thin enamel, about half as thickas human enamel, which means they produce substantialtensile forces on the under surface of the enamel at the EDJ(Constantino et al., 2012). This causes incipient cracks in thisregion (tufts) to extend towards the OES as radial cracks (Chaiet al., 2009). Increasing enamel toughness and reducingstiffness (modulus) can limit the extent of radial cracklengthening in sea otter teeth during hard food consumption.Further protection from crack propagation comes from thelarge size of their first molars (Constantino et al., 2012). Thelarger tooth radius helps protect against the propagation ofmargin and median cracks around the shoulder (suprabulge)of the tooth (Lawn and Lee, 2009). However, it does notprotect against the propagation of radial cracks from theEDJ to the occlusal surface, and if not stopped, these radialcracks can weaken the tooth and provide avenues for cariousinfection. It is therefore plausible, though speculative, thatboth sheep and sea otters have evolved softer and morecompliant enamel in order to increase enamel toughness. Insheep, the tougher enamel allows for smoother shearingcrests that can process food more efficiently, while in seaotters the increased toughness helps to decrease their sus-ceptibility to dangerous radial cracks.
6. Conclusions
The Young's modulus of ovine molar enamel is approxi-mately 30% lower than that of human enamel, while thehardness is about 9% lower. On the other hand, the hardnessand modulus of the ovine dentin are similar to those ofhuman dentin. The microstructure of the tissues, whichexhibits significant decussation across the entire thicknessof the enamel, suggests that the toughness of ovine enamelshould be at least as high as for human enamel, and possiblyhigher. We have postulated that this tradeoff between highertoughness and lower hardness and modulus producesenhanced resistance to wear in the ovine enamel relative tothat of humans and other primates. This conclusion is inaccord with the postulate that the over-riding challenge tosurvival of ovine molars is wear, rather than fracture. We
therefore expect that further studies conducted on the teethof a wider range of ruminants will echo this relative differ-ence between the dental tissue properties of ruminants andprimates.
Acknowledgments
The assistance of Mr. Nicholas Grainger (University of Wes-tern Australia) in collecting the sheep teeth is gratefullyacknowledged. Funding for this work was provided by theAustralian Research Council and the National Science Foun-dation (grant no. 1118385).
r e f e r e n c e s
Ang, S., Bortel, E., Swain, M., Klocke, A., Schneider, G., 2010.Size-dependent elastic/inelastic behavior of enamel overmillimeter and nanometer length scales. Biomaterials 31,1955–1963.
Anstis, G.R., Chantikul, P., Marshall, D.B., Lawn, B.R., 1981. Acritical evaluation of indentation techniques for measuringfracture toughness: I. Direct crack measurements. J. Am.Ceram. Soc. 64, 533–538.
Bajaj, D., Arola, D.D., 2009. On the R-curve behavior of humantooth enamel. Biomaterials 30, 4037–4046.
Barani, A., Keown, A., Bush, M., Lee, J.-W., Chai, H., Lawn, B., 2011.Mechanics of longitudinal cracks in tooth enamel. ActaBiomater. 7, 2285–2292.
Barani, A., Keown, A., Bush, M., Lee, J.-W., Lawn, B., 2012. Role oftooth elongation in promoting fracture resistance. J. Mech.Behav. Biomed. Mater. 8, 37–46.
Bechtle, S., Ang, S., Schneider, G., 2010a. On the mechanicalproperties of hierarchically structured biological materials.Biomaterials 31, 6378–6385.
Bechtle, S., Fett, T., Rizzi, G., Habelitz, S., Klocke, A., Schneider,G.A., 2010b. Crack arrest within teeth at the dentinoenameljunction caused by elastic modulus mismatch. Biomaterials31, 4238–4247.
Bechtle, S., Habelitz, S., Klocke, A., Fett, T., Schneider, G.A., 2010c.The fracture behavior of dental enamel. Biomaterials 31,375–384.
Chai, H., Lee, J.-W., Constantino, P., Lucas, P., Lawn, B., 2009.Remarkable resilience of teeth. Proc. Nat. Acad. Sci. U.S.A. 106,7289–7293.
Chai, H., Lee, J.-W., Lawn, B., 2011. On the chipping and splittingof teeth. J. Mech. Behav. Biomed. Mater. 4, 315–321.
Constantino, P.J., Lee, J.J.-W., Chai, H., Zipfel, B., Ziscovici, C.,Lawn, B.R., Lucas, P.W., 2010. Tooth chipping can reveal biteforces and diets of fossil hominins. Biol. Lett. 6, 826–829.
Constantino, P.J., Lee, J.J.-W., Grbig, Y., Hartstone-Rose, A.,Talebi, M., Lawn, B.R., Lucas, P.W., 2012. The role of toothenamel mechanical properties in primate dietary adaptation.Am. J. Phys. Anthropol. 148, 171–177.
Constantino, P.J., Lee, J.J.-W., Morris, D., Lucas, P.W., Harstone-Rose, A.,Lee, W.-K., Dominy, N.J., Cunningham, A., Wagner, M., Lawn, B.R.,2011. Adaptation to hard-object feeding in sea otters andhominins. J. Hum. Evol. 61, 89–96.
Constantino, P.J., Lucas, P.W., Lee, J.J.-W., Lawn, B.R., 2009. Theinfluence of fallback foods on great ape tooth enamel. Am.J. Phys. Anthropol. 140, 653–660.
Cuy, J., Mann, A., Livi, K., Teaford, M., Weihs, T., 2002.Nanoindentation mapping of the mechanical propertiesof human molar tooth enamel. Arch. Oral Biol., 47.
j o u r n a l o f t h e m e c h a n i c a l b e h a v i o r o f b i o m e d i c a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1 7 6 – 1 8 5184
Author's personal copy
DeGusta, D., Everett, M., Milton, K., 2003. Natural selection onmolar size in a wild population of howler monkeys (Alouattapalliata). Proc. R. Soc. London, Ser. B 270, S15–17.
Every, D., Tunnicliffe, G., Every, R., 1998. Tooth-sharpeningbehaviour (thegosis) and other causes of wear on sheep teethin relation to mastication and grazing mechanisms. J. R. Soc.N. Z. 28, 169–184.
Habelitz, S., Marshall, W.G., Balooch, M., Marshall, S.J., 2002.Nanoindentation and storage of teeth. J. Biomech. 35, 995–998.
He, L., Swain, M., 2007. Influence of environment on themechanical behaviour of mature human enamel. Biomaterials28, 4512–4520.
Imbeni, V., Kruzic, J.J., Marshall, G.W., Marshall, S.J., Ritchie, R.O.,2005. The dentin–enamel junction and the fracture of humanteeth. Nat. Mater. 4, 229–232.
Janis, C., Fortelius, M., 1988. On the means whereby mammalsachieve increased functional durability of their dentitions,with special reference to limiting factors. Biol. Rev. 63,197–230.
Jurgens, M., Bregendahl, K., Coverdale, J., Hansen, S., 2012. AnimalFeeding and Nutrition, eleventh ed. Kendall/Hunt, Dubuque,Iowa.
Keown, A., Bush, M., Ford, C., Lee, J.-W., Constantino, P., Lawn, B.,2012. Fracture susceptibility of worn teeth. J. Mech. Behav.Biomed. Mater. 5, 247–256.
Lawn, B., Lee, J.-W., 2009. Analysis of fracture and deformation inteeth subjected to occlusal loading. Acta Biomater. 5, 2213–2221.
Lee, J.-W., Morris, D., Constantino, P., Smith, T., Lawn, B., 2010.Properties of tooth enamel in great apes. Acta Biomater. 6,4560–4565.
Leyland, A., Matthews, A., 2000. On the significance of the H/Eratio in wear control: a nanocomposite coating approach tooptimised tribological behaviour. Wear 246, 1–11.
Lucas, P., 2004. Dental Functional Morphology: How Teeth Work.Cambridge University Press, Cambridge, U.K.
Lucas, P.W., Constantino, P.J., Wood, B.A., Lawn, B.R., 2008. Dentalenamel as a dietary indicator in mammals. BioEssays 30,374–385.
Miserez, A., Weaver, J.C., Thurner, P.J., Aizenberg, J., Dauphin, Y.,Fratzl, P., Morse, D.E., Zok, F.W., 2008. Effects of laminate
architecture on fracture resistance of sponge biosilica: lessonsfrom nature. Adv. Funct. Mater. 18, 1241–1248.
Oberle, T.L., 1951. Properties influencing wear of metals. J. Met. 3,438 (-&).
Oliver, W.C., Pharr, G.M., 1992. An improved technique fordetermining hardness and elastic modulus using load anddisplacement sensing indentation experiments. J. Mater. Res.7, 1564–1583.
Oliver, W.C., Pharr, G.M., 2004. Measurement of hardness andelastic modulus by instrumented indentation: advances inunderstanding and refinements to methodology. J. Mater. Res.19, 3–20.
Osborn, J.W., 1969. The 3-dimensional morphology of the tufts inhuman enamel. Acta Anat. 73, 481–495.
Popowics, T., Rensberger, J., Herring, S., 2001. The fracturebehavior of human and pig molar cusps. Arch. Oral Biol. 46,1–12.
Popowics, T., Rensberger, J., Herring, S., 2004. Enamelmicrostructure and microstrain in the fracture of human andpig molar cusps. Arch. Oral Biol. 49, 595–605.
Roy, M.E., Whiteside, L.A., Xu, J., Katerberg, B.J., 2010. Diamond-like carbon coatings enhance the hardness and resilience ofbearing surfaces for use in joint arthroplasty. Acta Biomater. 6,1619–1624.
Weaver, J.C., Wang, Q., Miserez, A., Tantuccio, A., Stromberg, R.,Bozhilov, K.N., Maxwell, P., Nay, R., Heier, S.T., DiMasi, E.,Kisailus, D., 2010. Analysis of an ultra hard magneticbiomineral in chiton radular teeth. Mater. Today 13, 42–52.
White, J., Goodis, H., Marshall, S., Marshall, G., 1994. Sterilizationof teeth by gamma radiation. J. Dent. Res. 73, 1560–1567.
Wise, G., Yao, S., Henk, W., 2007. Bone formation as a potentialmotive force of tooth eruption in the rat molar. Clin. Anat. 20,632–639.
Zhou, J., Hsiung, L., 2007. Depth-dependent mechanicalproperties of enamel by nanoindentation. J. Biomed. Mater.Res. Part A 81A, 66–74.
Zok, F.W., Miserez, A., 2007. Property maps for abrasion resistanceof materials. Acta Mater. 55, 6365–6371.
j o u r n a l o f t h e m e c h a n i c a l b e h a v i o r o f b i o m e d i c a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1 7 6 – 1 8 5 185
