d e n t a l m a t e r i a l s 2 6 ( 2 0 1 0 ) 100104

available at www.sciencedirect.com

journal homepage: www.intl.elsevierhealth.com/journals/dema

Micro-indentation fracture behavior of human enamel

Sanosh Kunjalukkal Padmanabhan a,b , Avinash Balakrishnan a,b , Min-Cheol Chu b , Taik Nam Kim a , Seong Jai Cho b,
a b

Department of Information and Electronic Materials Engineering, Paichai University, Daejeon 302735, Republic of Korea Division of Advanced Technology, Korea Research Institute of Standards & Science, Yuseong Gu, Daejeon 305340, Republic of Korea

a r t i c l e
Article history:

i n f o

a b s t r a c t
Objective. The purpose of this study was to determine the crack resistance behavior (KR ) of human enamel in relation to its microstructure. Methods. Human molar teeth were precision cut, polished and tested using Vickers microindentation at different loads ranging from 0.98 to 9.8 N. Five indentation load levels were considered, 20 indentation cracks for each load level were introduced on the surface of the test specimen (10 indentations per tooth) and their variability was evaluated using Weibull statistics and an empirical model. Scanning electron microscopy (SEM) and atomic force

Received 30 March 2009 Received in revised form 21 June 2009 Accepted 27 July 2009

Keywords: Human enamel Crack resistance Crack depth Crack bridges AFM Fracture toughness Hardness Dental ceramics Vickers indentation Lamellar rod Palmqvist crack

microscopy (AFM) were used to analyze the crack morphology and propagation mechanisms involved. Results. The results showed that enamel exhibited increasing cracking resistance (KR ) with increasing load. It was found that the crack propagation mainly depended on the location and the microstructure it encountered. SEM showed the formation of crack bridges and crack deection near the indentation crack tip. The crack mode was of Palmqvist type even at larger loads of 9.8 N. This was mainly attributed to the large process zone created by the interwoven lamellar rod like microstructure exhibited by the enamel surface. Signicance. This study shows that there are still considerable prospects for improving dental ceramics and for mimicking the enamel structure developed by nature. 2009 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.

1.

Introduction

Human enamel (HE) is one of the unique natural substances which still cannot be substituted effectively by articial restorative materials [13]. It is the hardest part of the body but still has good toughness, despite fairly poor working conditions, such as widely ranging loads, reciprocating movements, temperature shocks and possible acid attacks [4,5]. HE is composed of 96% hydroxyapatite (HA) by weight and the remaining 4% consists of water and organic material (mainly protein) [6]. Hence from a materials science point of view,

HE could be considered as a functionally graded composite material with a mineralized HA matrix and organic reinforcements. Although many biomechanical studies have focused on understanding the fracture mechanism in HE, it has still not been fully characterized. There are 2 main reasons for this (i) HE possesses complex microstructural components such as enamel rods and dental tubules, which are involved in fracture mechanisms [7,8] and (ii) structurally HE is very small and the specimens used for testing are even smaller. Although studies have reported the fracture toughness of the entire human tooth using exural tests (three point bend tests) [911], notched compact-tension test [12] and chevron-

Corresponding author. Tel.: +82 42 868 5388. E-mail addresses: the krecian@yahoo.com (T.N. Kim), sjcho@kriss.re.kr (S.J. Cho). 0109-5641/$ see front matter 2009 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.dental.2009.07.015

d e n t a l m a t e r i a l s 2 6 ( 2 0 1 0 ) 100104

101

notched short-bar [13] test; the values obtained for the whole teeth may not be representative of the HE surface. Moreover fracture toughness obtained using the above techniques could be related only to crack length and not to the crack location. For instance, micro-indentation studies have shown that fracture toughness values of HE range between 0.7 and 1.27 MPa m1/2 [14], dependent on the site where indentation is performed. This meant that that the crack resistance of HE is dependent not only on crack length but also on crack location. Hence it is important to understand such mechanisms as it would help the clinical treatments of teeth and guide the design and development of tooth-like restoratives that imitate the structure and properties of HE. The main objective of the present study was to understand the mechanism underlying the crack resistance of HE. The data analyzed using Weibull statistics show that although the interaction of the crack with the microstructure decreases with the increasing load, there is an increase in the crack resistance (KR ) which is attributed to crack bridging phenomena.

tion corners were measured immediately after unloading. The crack resistance was calculated using the following equation [15,16]: KR = E HV
1/2

P C3/2

(1)

where KR is the crack resistance, P is the applied indentation load, C is the indentation crack length, E and HV are the elastic modulus and the hardness values, respectively, is a dimensionless constant dependent only on the indenter geometry and has an empirical value of 0.016 0.004. The KR data were further analyzed with the conventional ranking method [17] to yield the Weibull parameters, knowledge of which would lead to complete characterization of the statistical properties of the measured KR data. This was done by arranging the data in ascending order. The ith result in the set of n = 20 data was assigned a cumulative probability of occurrence, Pi , which was calculated with: Pi = i n+1 (2)

2.

Methods
The measured KR and Pi , were then analyzed, by using a simple, least-square regression method, according to the alternative form of the well-known two-parameter Weibull distribution equation. ln ln 1 (1 Pi ) = m ln KR m ln K0 (3)

The test specimens of teeth (molar/wisdom teeth) were obtained from Dental College, Daejeon Health Science Center, Korea. The specimens were soaked in boiling water for approximately 1 min to remove the presence of any tartar/calculus (yellow stains) formation and disinfect bacteria which are commonly observed on the surface of HE and rinsed in 0.5 M KOH to remove the presence of any organic substances. The specimens were ultrasonically rinsed in acetone and dried. The top surface of the teeth was precision cut using a modied water-cooled diamond saw and was hot mounted in a resin. The mounted specimens were subsequently polished carefully using a cloth bonded diamond wheel with successively ner diamond pastes (up to 3 m) to yield a mirror-like surface suitable for indentation. Vickers indentations were made at 0.98, 1.96, 2.94, 4.9 and 9.8 N loads on the polished surfaces of the test specimens using a commercial hardness tester (Matsuzawa Co., Ltd., Japan (DVK-25)). For each of these loads four samples were used and on each sample ve indents were made, i.e., a total of 20 indentations for each load were studied on the micropolished surfaces of the test specimens. The average radial-median crack lengths produced at the indenta-

where m and K0 are the Weibull modulus and the scale parameter, respectively. Microstructural studies of the indentation cracks on specimens were carried out with a scanning electron microscope (SEM) and atomic force microscope (AFM).

3.

Results

The typical mode of indentation cracking on the HE surface is shown in Fig. 1a. Classic radial-median cracks were observed at the indentation corners along with secondary multiple cracks. The crack paths appeared zigzagged and showed the formation of crack bridges (Fig. 1b). Semi-circular crack patterns were also observed near the crack tip. AFM analysis of the etched surface (Fig. 1c) showed a highly roughened sur-

Fig. 1 SEM images of (a) indentation on the surface of the polished HE, arrows show secondary multiple cracks, (b) crack propagation near the indentation crack tip, A and B show the crack bridge and crack deection, respectively and (c) 3D AFM analysis of the etched surface.

102

d e n t a l m a t e r i a l s 2 6 ( 2 0 1 0 ) 100104

Fig. 2 Crack resistance as a function of indentation crack length under different loads. Dashed lines represent the t for Eq. (4).

Fig. 4 Plot showing transition in cracking mode from Palmqvist to radial-median as a function of indentation load P and c/a ratio.

model proposed by Ramachandran and Shetty [18]. KR = K (K K0 ) exp c (4)

Fig. 3 Weibull plots for the KR data measured at each load level.

where K0 , K and are adjustable parameters in Eq. (4). Here K0 and K represent the fracture toughness and maximum crack resistance of the material, respectively. The dash lines in Fig. 2 represent exponential curve of Eq. (4). The values of corresponding parameters K0 , K and , obtained by iterative regression were found to be 1.02 MPa m1/2 , 1.34 MPa m1/2 and 45.4 m. The parameter is related directly to the range of crack extension over which toughening effect should develop and saturate. The ratio of crack length (c) to diagonal length of indentation (a) i.e., c/a showed HE exhibiting a Palmqvist type of crack system in the load range of 0.989.8 N (Fig. 4). AFM analyses conrmed this Palmqvist type of crack mode and showed that indentation crack depth varied non-linearly with the load (Fig. 5).

face indicating projecting rod and interwoven structure on the HE. Fig. 2 shows the measured crack resistance, KR , plotted as a function of indentation crack length. The increasing tendency of the crack resistance with increasing crack length was observed. Fig. 3 shows the Weibull plots for the KR data measured at each load level. The resultant Weibull parameters, m and K0 have been listed in Table 1. It was also observed that m value increased (i.e. lower scattering in data) with increasing applied load. Further, the mechanism behind the increasing crack resistance with load was analyzed using an empirical

4.

Discussion

It seems that the toughening effect in the HE arose mainly from the enamel lamellar microstructure which had a signicant effect on the mode of crack propagation. The semi-circular crack paths on the HE surface indicate that cracks propagate around the intercepting enamel rods (Fig. 1a). Similar phenomena were also observed by other researchers [11,14]. These observations suggest that it is more difcult for cracks to cut through the enamel rods and hence fol-

Table 1 The values of: half diagonal indent a, crack length c, Vickers hardness HV , Weibull modulus m and K0 values for the KR data measured at each load level. Load (N)
0.98 1.96 2.94 4.9 9.8

a ( m)
11.8 16.2 20.7 26.3 37.7 0.003 0.0025 0.0022 0.0025 0.0012

c ( m)
22 29 37 54 79 1.4 1.1 0.8 0.65 0.4

HV (GPa)
3.33 3.51 3.22 3.31 3.22 0.0030 0.0022 0.0021 0.0023 0.0013

m
13 15 25 29 32

K0 (MPa m1/2 )
0.85 0.87 0.88 0.94 0.98

d e n t a l m a t e r i a l s 2 6 ( 2 0 1 0 ) 100104

103

Palmqvist type. The presence of a larger process zone in the HE, due to interwoven lamellar rods, seems to restrict the propagation of deeper cracks within the sample even at higher loads. This was indicative in AFM analyses where the crack depth is shallow and does not vary linearly with increasing indentation loads (Fig. 5). This means that crack resistance in the HE is not only dependent on the surface microstructure but is also an inherent structural property.

5.

Conclusions

Fig. 5 Plot of indentation crack depth (average of 5 crack depth values) measured at each load level.

low the weaker rod boundaries resulting in crack bridges (see Fig. 1b). This microstructural effect was quite prominent in KR values at lower loads, where the interaction between crack and microstructure appears to be non-uniform. Studies have shown that the enamel rods have varied diameters, are not straight, and are in some cases woven [19,20]. Therefore, even after polishing it is difcult to obtain a at section exactly parallel or perpendicular to the enamel rods. Hence an anisotropic microstructure exists on the HE. However beyond a threshold load, the probability of a crack to interact with the microstructural features having the largest resistance to fracture also increases and, consequently, the scattering of the KR values was small. Analysis using the Ramachandran and Shetty [18] empirical model showed that in the presence of the crack bridging mechanism and deection in HE, one can expect the formation of a large process zone behind the crack tip. The K0 values of HE obtained by Weibull modulus (Table 1) and from Eq. (4) for HE were in agreement with the fracture toughness values reported as 0.77 0.05 MPa m1/2 [8]. The hardness of the HE (3.3 GPa) was found to be 6373% less compared to alumina (12 GPa) and zirconia (9 GPa) dental ceramics. These hardness values are indicative that these dental ceramics, even though tougher, should not be too hard or else the abrasive wear of the opposing natural tooth will be too severe. The dependence of crack resistance on the surface cracks during indentation can be described by two different crack systems [21]: (a) radial-median or half-penny crack, and (b) the Palmqvist crack. The radial-median mode derives from subsurface median cracks that initiate along the edges of the pyramidal indentation and extend deep into the material in a semi-circular manner perpendicular to the surface (hence it is known as a half-penny-shaped crack). The Palmqvist crack morphology is characterized by much shallower cracks emanating from the corners of the Vickers indentation [22]. In brittle bulk materials, such as glasses, ceramics and composites, the possibility of a transition in cracking mode from Palmqvist to radial-median can be determined as a function of indentation load and c/a ratio (where c is the total crack length and a is the indentation diagonal half-length). Usually, if c/a 2.5, then the crack system is termed as radial-median mode and if c/a < 2.5 then the crack system is referred as the

The present study shows that human enamel exhibits increasing crack resistance with increasing loads. This property of teeth is attributed to the highly intricate interwoven lamellar microstructure which is an inherent structural property of enamel. The indentation cracks emanating showed formation of crack bridges and deection by these lamellar rods resulting in a large process zone which prevents immediate formation of deep cracks (radial-median) preventing the failure of teeth. This study shows that there is a huge prospect for the improvement of properties and microstructures of dental ceramics including a more critical attempt to mimic the structure developed by nature.

references

[1] Craig RG. Advances in biomaterials from 1957 to 1997. J Oral Rehabil 1999;26:8416. [2] Kishen A, Ramamurty U, Asundi A. Experimental studies on the nature of property gradients in the human dentine. J Biomed Mater Res 2000;51:6509. [3] Craig RG, Peyton FA. The microhardness of enamel and dentin. J Dent Res 1958;37:6618. [4] Eugeniusz S. On evaluation of wear resistance of tooth enamel and dental materials. Wear 2006;260:125661. [5] Eisenburger M, Addy M. Erosion and attrition of human enamel in vitro. Part II: inuence of time and loading. J Dent 2002;30:34952. [6] He LH, Swain MV. Inuence of environment on the mechanical behavior of human enamel. Biomaterials 2007;28:451220. [7] Stanford JW, Paffenbarger GC, Kumpula JW. Determination of some compressive properties of human enamel and dentin. J Am Dent Assoc 1958;57:48796. [8] Xu HHK, Smith DT, Jahanmir S, Romberg E, Kelly JR, Thompson VP, et al. Indentation damage and mechanical properties of human enamel and dentin. J Dent Res 1998;77:47280. [9] Bowen RL, Rodriguez MS. Tensile strength and modulus of elasticity of tooth structure and several restorative materials. J Am Dent Assoc 1962;64:37887. [10] Rasmussen ST, Patchin RE. Fracture properties of human enamel and dentin in an aqueous environment. J Dent Res 1984;63:13628. [11] Rasmussen ST, Patchin RE, Scott DB, Heuer AH. Fracture properties of human enamel and dentin. J Dent Res 1976;55:15464. [12] El Mowafy OM, Watts DC. Fracture toughness of human dentin. J Dent Res 1986;65:67781. [13] Lin CP, Douglas WH. Structureproperty relations and crack resistance at the bovine dentin-enamel junction. J Dent Res 1994;73:10728.

104

d e n t a l m a t e r i a l s 2 6 ( 2 0 1 0 ) 100104

[14] Hassan R, Caputo AA, Bunshah RF. Fracture toughness of human enamel. J Dent Res 1981;60:8207. [15] Gong JH, Guan ZD, Wu JJ. R-curve behavior of indentation cracks in silicon nitride based ceramics composites. J Non-Cryst Sol 1994;177:4653. [16] Gong J, Zhao Z, Guan Z. On the local crack resistance of Al2 O3 TiC composites evaluated by direct indentation method. J Eur Ceram Soc 2001;21:9416. [17] Schneider SJ. Engineered materials handbook, ceramics and glasses, vol. 4. USA: ASM International; 1991. p. 110. [18] Ramachandran N, Shetty DK. Rising crack-growth-resistance (R-curve) behavior of toughened alumina and silicon nitride. J Am Ceram Soc 1991;74:263441.

[19] Ten Cate AR. Oral histology: development, structure and function. St. Louis, MO: Mosby; 1989. p. 21327. [20] Sturdevant CM, Roberson TM, Heymann HO, Sturdevant JR. Operative dentistry. 3rd ed. St. Louis, MO: Mosby; 1995 [chapters 2 and 7]. [21] Ponton CB, Rawlings RD. Dependence of the Vickers indentation fracture toughness on the surface crack length. J Br Ceram Trans 1989;88:8390. [22] Bamzai KK, Kotru PN, Wanklyn BM. Fracture mechanics, crack propagation and microhardness studies on ux grown ErAlO3 single crystals. J Mater Sci Technol 2000;16: 40510.