- Email: [email protected]
Using bearing area parameters to quantify early erosive tooth surface changes in enamel: A pilot study
journal of dentistry 41 (2013) 1060–1067
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.intl.elsevierhealth.com/journals/jden
Using bearing area parameters to quantify early erosive tooth surface changes in enamel: A pilot study James Field a,*, Matthew German b, Paula Waterhouse c a
Restorative Dentistry, Newcastle University School of Dental Sciences, United Kingdom Dental Materials Science, Newcastle University School of Dental Sciences, United Kingdom c Child Dental Health, Newcastle University School of Dental Sciences, United Kingdom b
article info
abstract
Article history:
Objectives: Most in vitro studies investigate the erosive process using relatively simple
Received 29 April 2013
roughness parameters such as roughness average (Ra). In isolation, Ra may misrepresent
Received in revised form
the surface features. Further, few studies report baseline surface characteristics after
18 July 2013
sample preparation. This study aimed to test the hypothesis that measuring the bearing
Accepted 22 August 2013
area parameters in addition to Ra may be useful when qualifying the surface of enamel at baseline and after an erosive challenge. The null hypothesis for this study was that the bearing area parameters provide no more useful information than Ra alone, when qualifying
Keywords:
the surface of enamel at baseline and after an erosive challenge.
Erosion
Methods: Enamel slabs (n = 20) were prepared from human (n = 2) and bovine (n = 4) incisor
Roughness
teeth and polished with 0.05 mm paste. Roughness average (Ra) and bearing parameters (MR1,
Bearing parameters
MR2, Rpk, Rk, Rvk) were used to record baseline characteristics. Specimens were subjected to
Bovine enamel
erosion with 1% citric acid solution for 1 min. Profilometric characteristics were recorded post-
Human enamel
erosion, along with the maximum height changes within the profile. T-tests were carried out in
In vitro
order to compare baseline surface characteristics between tissue types. Post-erosion, analysis of variance (ANOVA) was used to test the effects of tissue type (bovine or human). Results: There was no significant difference in Ra between human and bovine incisor enamel at baseline (human 0.11 mm, bovine 0.12 mm P > 0.05), and no significant difference was observed post-erosion (human 0.23 mm, bovine 0.20 mm P > 0.05). There were significant differences in bearing parameters at baseline and post-erosion that were not identified by the Ra measurement alone. Conclusions: The results suggest that if Ra alone is measured, important differences in surface characteristics may be missed. The null hypothesis is rejected, and the recommendation is made that bearing parameters are included within profile measurements in order to further triangulate the results of surface analysis studies. Clinical relevance: In isolation, Ra may misrepresent the surface features of a profile. These results have shown that the bearing parameters are an important and informative set of measurements. The recommendation is made that bearing parameters are included within profile measurements at baseline and post-erosion, in order to further triangulate the results of surface analysis studies. # 2013 Elsevier Ltd. All rights reserved.
* Corresponding author at: Room D5.014, Level 5, School of Dental Sciences, Newcastle University, Framlington Place, Newcastle NE2 4BW, United Kingdom. Tel.: +44 191 222 8515. E-mail address: [email protected] (J. Field). 0300-5712/$ – see front matter # 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jdent.2013.08.015
journal of dentistry 41 (2013) 1060–1067
1.
Introduction
This study follows a review of methods to measure surface change on dental hard tissues in vitro.1 The review concluded that profilometry (and in particular roughness average, Ra) is the main reported measurement within dental surface studies. The review highlighted the potential drawbacks of reporting Ra alone and also proposed that other parameters could be reported from the same set of profilometric data that would allow a further and more meaningful description of the surface quality. These parameters are known as bearing parameters. The main drawbacks with Ra are that the value contains no information about the textural characteristics of a profile, the likelihood of future wear or wear-resistance, the rate of future wear or the potential of a surface to retain fluids/lubricants. This makes the qualitative assessment of the surface relatively limited and empirical and only relays information about fixed heights at regular intervals. In 1933, Abbott and Firestone defined the bearing area fraction.2 This fraction or ‘ratio’ represents the cumulative distribution of the lengths of individual plateaux that would result if the surface were abraded down to a level plane at that height. The curve is normalised by the total assessment length. If this ratio was plotted against sample height, the bearing area curve results (Fig. 1). A variety of sources describe how to interpret the bearing curve.3–6 The total height of the curve is known as the roughness total. The width of the curve represents the percentage of material at different profile amplitudes. Five parameters are determined from the curve – peak roughness (Rpk), core roughness (Rk), valley roughness (Rvk), material ratio of peaks (MR1) and material ratio of troughs (MR2). These bearing parameters are defined using German standard DIN 4776 and more recently ISO 13565-2. The Rpk value represents the height of the material peaks – the area that will come into contact with an opposing surface first; in engineering it indicates the amount of material that will wear away during early contact of the machined surface. The Rk value represents the height of the material core – the area that will support most of the contact load (in engineering
1061
this represents the longer-term running surface). The Rvk value represents the depth of the material troughs/valleys – the area that will retain fluid or lubricant. These are known as the reduced peak height, core roughness and reduced trough depth respectively. The curve can also relate Rpk and Rvk to a proportion of the material’s surface (Figs. 1 and 2). The latter are known as the material ratios – MR1 relates to the proportion of peaks within the sample, and MR2 relates to the proportion of troughs. It has been suggested that the bearing curve is better for assessing surfaces that are subjected to several different machining processes than Ra.3,7,8 One process may remove the peaks but then a finer texture may be superimposed onto the resulting plateaux by a subsequent process; the deep valleys may remain unaffected. The resulting surfaces are known as ‘multi-stratified’.9 Characteristically, these surfaces are negatively skewed which makes it difficult for a parameter such as Ra to represent them effectively. In relation to tooth tissue, the bearing curve could also be used to measure the surface effects of a synergistic process, like that of erosion and abrasion. It may be possible using the curve, to explain the bi-phasic pattern of tooth surface loss which has been previously identified and described as ‘running-in wear’ and ‘steady-state wear’.10 The bearing curve has been used to assess engineered surfaces in medicine, particularly that of femoral stem wear.11 Its use to monitor and qualify labial erosion in vivo (via silicone models) was piloted in 199712 and, more recently, it has been used to describe the native enamel surface.13 The latter study showed the potential use of the bearing curve in further qualifying the tooth surface; however data were presented in tabular form only. There may have been more value in displaying the bearing area curves for visual comparison. This study aimed to test the hypothesis that measuring the bearing area parameters would allow further qualification of the surface effects on human and bovine enamel at baseline and after an erosive challenge. The null hypothesis for this study was that the bearing area parameters provide no more useful information than Ra alone, when qualifying the surface of enamel at baseline and after an erosive challenge.
Fig. 1 – Quantifying parameters on the bearing curve. Rpk, Rk and Rvk values relate to peak, core and trough average roughness values. MR1 and MR2 relate to the percentage of the surface occupied by peaks or troughs respectively.
1062
journal of dentistry 41 (2013) 1060–1067
Fig. 2 – Comparison of two different surfaces with similar Ra using the bearing curve. Surface A has significantly more, deeper troughs (higher Rvk and higher MR2). Surface A therefore has a much greater pooling/lubricative potential. Surface B has significantly more, higher peaks (higher Rpk and higher MR1) and is therefore likely to suffer significantly more early surface loss in the future.
2.
Materials and methods
2.1.
Specimen collection and preparation
Extracted human lower permanent incisor teeth were collected from the adult Dental Emergency Clinic at Newcastle Dental Hospital, UK between February 2008 and February 2011. The teeth were stored in a 1% (w/v) aqueous solution of sodium p-toluenesulfonylchloramide hydrate (Chloramine-T; Sigma–Aldrich, Dorset, UK) solution and suitable teeth, showing no signs of coronal caries or tooth surface loss were entered into the Newcastle Tissue Bank (Human Tissue Act license number 12534), stored at 4 8C. Bovine permanent incisor teeth were harvested on two occasions – March 2010 and December 2010 from the same abattoir – Linden Foods, Burradon, Cramlington (Registered Plant Number 2056, Food Standards Agency, Department of Environment, Food and Rural Affairs). The cattle were male beef Shorthorn cattle and were aged approximately 18–20 months.
4 bovine and 2 human incisor crowns were sectioned from the root of the tooth and prepared with two equidistant and parallel undercut reference grooves across the labial surface (mesio-distally) using a pear-shaped high-speed diamond bur (Hi-Di1 medium diamond bur ISO 237/012; Dentsply, Addlestone, UK). The crowns were then positioned into individual casting moulds (Buehler, Du¨sseldorf, Germany) with the labial surface facing upwards and the cemento-enamel junction perpendicular to the base. They were held in place with sticky wax (Kemdent; Swindon, UK) and cast in acrylic resin (Bonda; Bondaglass-Voss, Beckenham, UK). Once set, the casts were removed from the moulds. The labial surface was ground flat at grit size 80 (Metaserv rotary pregrinder C200/RB; Metallurgical services, Ohio, USA) until the majority of the labial enamel surface was exposed. The sample was then further polished at successive grit sizes up to 1200 and then finished for 25 s (anticlockwise rotation at light pressure) with a 0.05 mm polishing paste (Metaserv universal polisher C200/5V; Metallurgical Services, Ohio, USA). The samples were rinsed with a balanced salt solution (HBSS; Life Technologies, Paisley, UK), and inspected under the light microscope (Apophot; Nikon, Surrey, UK) for visible defects and machine marks. The polished labial surfaces were then sectioned coronally using a low-speed water-cooled diamond wheel saw (Testbourne 650 CE; South Bay Technologies, California, USA) to produce reference slabs at least 1.5 mm wide. Some slabs became damaged during the sectioning process and when possible these were re-polished. Otherwise they were discarded. Specimens were stored in HBSS at 4 8C until required. Four samples were obtained from each of the four bovine crowns, and two samples from each of the two human crowns. 20 samples were prepared in total. The baseline surfaces were then profiled using a stylus profilometer and its associated software (Mitutoyo Surftest SV-2000 and Surfpak-SV V1.600; Mitutoyo, Halifax, UK). The instrument range was 800 mm with a contact force of 4 mN. The stylus was a diamond cone tip held at 908 to the surface, with a 5 mm radius. Average roughness values, and bearing area parameters (Rk, Rvk, Rpk, MR1 and MR2) were recorded twice for each sample 0.5 mm apart. Each evaluation length
Table 1 – Mean roughness and bearing parameters of human (H) and bovine (B) enamel at baseline and post-erosion. Standard deviations are within brackets. Differing superscripts (a/b) denotes a significant baseline/erosion effect. Values in bold denote a significant tissue effect. Roughness average (Ra) (mm)
H
B a
Peak roughness (Rpk) (mm)
H a
B a
Core roughness (Rk) (mm)
H a
a
B
Valley roughness (Rvk) (mm)
H a
B a
a
Material ratio of peaks MR1 (%)
Material ratio of troughs MR2 (%)
H
H
a
Baseline
0.11 (0.02)
0.12 (0.11)
0.20 (0.03)
0.16 (0.08)
0.31 (0.07)
0.33 (0.05)
0.36 (0.10)
0.30 (0.09)
21 (3)
Eroded
0.23b (0.08)
0.2b (0.1)
0.50b (0.05)
0.43b (0.05)
0.21b (0.003)
0.12b (0.01)
0.59b (0.05)
0.54b (0.10)
12b (1)
B a
a
7 (2)
92 (3)
8.01b (0.80)
93a (3)
B 81 (9)
Maximum height change within the eroded profile (mm) H
B
0.7 (0.2)
0.41 (0.20)
a
71b (2)
journal of dentistry 41 (2013) 1060–1067
included 5 readings with a 0.3 mm cut-off (1.5 mm total evaluation length, starting within the body of the left hand reference layer) and was analysed through a Gaussian filter.
fresh solution was used for each sample. After treatment the samples were rinsed with fresh HBSS and profiled again.
2.3. 2.2.
1063
Analysis
Sample exposure procedures
Citric acid solution was formulated at 1% (w/v) and the pH was measured (Thermo Orion 4 Star; Fisher Scientific, Leicestershire, UK) at pH 2.2. The solutions were warmed using a water bath to 30 8C. Enamel samples were immersed for 1 min without stirring and
All statistical analyses were carried out using Sigmaplot for Windows Version 11.0, Build 11.0.1.80 (Systat software, Chicago, USA). T-tests were carried out in order to compare baseline surface characteristics between tissue types. On the occasions that normality tests were failed, the Mann Whitney rank sum test was used (median values reported).
Fig. 3 – Typical stylus profilometry profile (Ra 150 mm) and bearing curve (MR1 11%, MR2 89%) for baseline lapped bovine enamel.
1064
journal of dentistry 41 (2013) 1060–1067
Fig. 4 – Typical stylus profilometry profile (Ra 129 mm) and bearing curve (material ratio 1, MR1 9.5%, material ratio 2, MR2 90%) for eroded human enamel (1% citric acid for 1 min).
Post-erosion, analysis of variance (ANOVA) was used to test the effects of tissue type (bovine or human).
significant differences were found between species in terms of MR1 and MR2 (P < 0.001; both cases); human enamel displayed significantly more peaks and fewer troughs than bovine enamel.
3.
Results
3.2.
3.1.
Baseline
Post-erosion parameters are shown in Table 1. A typical stylus profile and bearing curve for human enamel are shown in Fig. 4. Eroded roughness average (Ra) values were significantly different to baseline values (P < 0.001); becoming rougher post-erosion. Ra was not significantly affected individually by tissue type (P = 0.085).
Baseline parameters are shown in Table 1. A typical profile and bearing area curve for bovine enamel are shown in Fig. 3. At baseline, no significant differences were found between the human and bovine samples in terms of their Ra (P = 0.176), Rk (P = 0.280), Rvk (P = 0.066) or Rpk (P = 0.177). However, very
Post-erosion
journal of dentistry 41 (2013) 1060–1067
Eroded core roughness (Rk) values were significantly different to baseline values (P = 0.007), becoming significantly less rough post-erosion. Rk was significantly affected by tissue (P < 0.001; Rk mean human 0.21, bovine 0.12). Eroded peak roughness (Rpk) values were significantly different to baseline values (P < 0.001), becoming significantly more rough post-erosion. Rpk was significantly affected by tissue (P < 0.001; Rpk mean human 0.50, bovine 0.43). Eroded valley roughness (Rvk) values were significantly different to baseline values (P < 0.001), becoming significantly more rough post-erosion. Rvk was not significantly affected by tissue type (P = 0.679; Rvk mean human 0.59, bovine 0.54). The proportion of eroded profile peaks (MR1) were significantly different to baseline values (P < 0.001). MR1 was significantly affected by tissue (P < 0.001; MR1 mean human 12, bovine 8). The proportion of bovine eroded profile troughs (MR2) were significantly different to baseline values (P = 0.013), with significantly more troughs post-erosion. MR2 was significantly affected by tissue (P < 0.001; MR2 mean human 93, bovine 71). The maximum height changes of the eroded profiles were significantly affected by tissue (P = 0.028; mean human 0.7 vs. bovine 0.41).
4.
Discussion
This study used a novel method of reporting profilometric measurements, in order to describe differences in surface characteristics between tissue types at baseline and posterosion. The findings suggest that if Ra alone is measured, subtle differences in surface characteristics may be missed. Indeed surface changes may be able to be modelled more effectively if these extra baseline characteristics are recorded.
4.1.
Baseline measurements
When polished with a 0.05 mm paste, the resultant surfaces of human and bovine enamel showed roughness average values of 0.113 mm and 0.120 mm respectively. These values were not significantly different to one another and this may be due to the small sample size used. However no published studies to date report a direct comparison between human and bovine incisor teeth after baseline preparation, and this area undoubtedly requires further study. Few studies report baseline human enamel roughness. Quartarone14 reported a range of control values for human enamel of between 0.05 mm and 0.120 mm, and although Rq was reported rather than Ra, the upper limits of the reported values are similar to the findings from this work; unfortunately the study did not detail any particular technique for enamel preparation and so further comparison is not possible. Higher control values have been reported by Ren15 who reported baseline Ra between 1.20 mm and 1.40 mm after polishing buccal and lingual surfaces of human third molars with 0.3 mm paste. Azrak16 also recorded similar Ra values (median 1.57 mm) for the labial surfaces of human lower incisors; the enamel surfaces were prepared using a pumice slurry and then polished with silicon carbide and aluminium oxide discs (to unreported particle and grit sizes). Given that pumices can contain particles up to 50 mm in diameter,17 it is
1065
unlikely that any amount of fine polishing will remove surface defects obtained due to treatment with pumice slurry, and this may explain the significantly higher roughness values obtained. Even fewer studies report bovine baseline enamel roughness; Gerbo18 reported values between 3.4 mm and 3.8 mm for bovine incisors but this was of the native enamel surface only. No surface preparation had been carried out. Fuji19 reported baseline Ra values for bovine incisors of between 4.7 nm and 6.4 nm. These values are significantly lower than those obtained in this study. The enamel samples were polished with a 0.25 mm paste and then cleaned ultrasonically to remove paste debris before measurement with a stylus tip of radius 2.5 mm; it is unknown whether the stylus tip was able to accurately measure roughness in the nanometre range, and it may be the case that the values reported do not relate to the surface finish. Although the paste used in this study was of a smaller particle size, it is unlikely that this would be responsible for the significantly smoother surface finish; it has already been demonstrated that grit size does not correlate well with prismatic enamel roughness.20 Further, it is unknown what effect the ultrasonic instrumentation had on the enamel surface; shearing of enamel peaks may have significantly reduced the surface roughness average, and this would warrant further investigation. Fuji also measured roughness average values using focus variation 3D microscopy, which gave higher Ra values than the reported profilometry values, but still significantly lower than expected (15–17 nm). Although in the current investigation the majority of baseline roughness parameters were not significantly different between human and bovine tissue, two parameters were significantly different. There were significantly more peaks within the human samples (higher MR1 value) and significantly more troughs within the bovine samples (lower MR2 value). To date these parameters have not been used to compare the lapped enamel surface and it is unknown to what degree the underlying enamel morphology affected the lapping process. It is proposed that a profile with an increased number of isolated profile peaks would be more susceptible to early erosive or abrasive challenges; an increased number of isolated peaks may increase the surface area that comes into immediate contact with an erosive solution. These peaks may be relatively weak and under-supported increasing the chances of structural failure or loss. It is also proposed that a profile with an increased number of profile troughs would be more likely to retain or trap fluid within the lower third of the profile. With an erosive challenge, this may serve to protect the surface if the trapped solution becomes supersaturated and is unable to be replaced. Troughs may also have implications for retaining remineralising solutions such as mouthwashes, or increasing the lubricative potential and exacerbating abrasive forces. These theoretical proposals undoubtedly require further study.
4.2.
Post-erosion
At the nano-level, an increase in roughness post-erosion may be explained by the erosive challenge preferentially ‘punching out’ the centre of the apatite crystals.21 On a micro-scale,
1066
journal of dentistry 41 (2013) 1060–1067
others describe the resultant ‘honeycomb’ pattern typically seen in human enamel,22,23 resulting from early erosion of prism and prism junctions leaving behind steep ridges of inter-prismatic enamel.24,25 Despite few studies that compare baseline to eroded profiles, these findings are in keeping with other published work; Quartarone14 reported a ‘remarkable’ increase in enamel roughness after acidic exposure (over 500%) and although the current study demonstrated a much more modest increase in roughness (around 100%) the total submersion times were significantly shorter (1 min vs. 280 min) in order to simulate surface changes rather than gross surface loss. In this study, the eroded roughness average of human and bovine enamel was not significantly different. This mirrors the baseline trend. Despite similar roughness averages, human samples exhibited a greater profile depth range after the erosive challenge. This effect may be directly linked to the increased number of isolated profile peaks at baseline. This finding is in keeping with recent work which reported that human enamel loss was significantly greater than bovine enamel loss when exposed to citric acid (pH 3.2) for up to 5 min.26 This difference may also be due to the tissue-specific proportions of prismatic to interprismatic enamel; it has already been reported that bovine tissue contains a greater proportion of interprismatic enamel than human tissue, and this interprismatic enamel is more resistant to erosion than prismatic.24 The conclusion that bovine enamel has the potential to be more resistant to erosion than human enamel, mirrors the results from the current study. A number of studies contradict these findings,27,28 suggesting that bovine enamel is more susceptible to erosion than human. The acidic challenges used however were more aggressive (at greater concentrations and lower pH, for longer periods of time). Indeed the work by White et al.26 has also shown that as the acid-exposure time increases beyond 5 min, bovine enamel then begins to display a greater surface loss than human enamel. This effect may be due to the increased proportion of troughs within the bovine enamel at baseline, which present as the erosive challenge becomes more aggressive. Less aggressive erosive challenges also result in more contrast in relief between prismatic and aprismatic enamel,29 and so the behaviour of bovine and human enamel may change significantly depending on the type of erosive insult. This has implications when attempting to compare results from different research groups or even between similar experiments that utilise disparate erosive solutions and/or erosive regimes; ultimately it adds further complexity to whether bovine enamel is a suitable model for erosion on human teeth, and the results from this pilot experiment suggest that it is not. Despite the eroded Ra values being similar between tissues there were, again, significant differences in bearing parameters between the human and bovine samples. These followed the baseline trends however post-erosion, the peak and valley roughness were also increased, whilst the core roughness showed an overall decrease. These results may suggests that the eroded surface is more susceptible to further early surface loss than at baseline, but that perhaps the surface shows a degree of resilience for further moderate abrasive challenges.
This will undoubtedly require further investigation to tease out the importance of these surface differences and how they predispose the enamel surface to further surface loss. A technique allowing visual comparison of the enamel surfaces, such as scanning electron microscopy, may prove useful in this respect.
5.
Conclusions
This pilot study has shown that when lapped with 0.05 mm paste, there was no significant difference in roughness average between human and bovine incisor enamel at baseline. Further, no significant difference in roughness average between tissues was observed post-erosion. There were, however, significant differences in bearing parameters between tissues both at baseline and post-erosion. The null hypothesis is rejected, recognising that these parameters are more sensitive to early surface change than Ra alone. Important differences in surface characteristics between the tissues were identified, and this study has shown evidence of how these differences may predispose the enamel surfaces to subsequent surface loss (either by abrasion or further erosion); this will undoubtedly require further study.
Acknowledgements The authors would like to thank Linden Foods for access to their abattoir in Burradon. JF, MG and PW conceived and designed the experiments. JF performed the experiments. JF analysed the data. JF, MG and PW wrote the paper.
references
1. Field JC, Waterhouse PJ, German MJ. Quantifying and qualifying surface changes on dental hard tissues in vitro. Journal of Dentistry 2010;38:182–90. 2. Abbott EJ, Firestone FA. Specifying surface quality. Mechanical Engineering 1933;59:569–72. 3. Pawlus P, Grabon W. The method of truncation parameters measurement from material ratio curve. Precision Engineering 2008;32:342–7. 4. Schma¨hling J, Hamprecht FA. Generalizing the Abbott– Firestone curve by two new surface descriptors. Wear 2007;262:1360–71. 5. International Standards Organisation. In: Standard I, editor. Geometrical product specifications (GPS) – surface texture: profile method; surfaces having stratified functional properties. Part 2. Height characterisation using the linear material ratio curve. 1996. 6. Petropoulos GP, Torrance AA, Pandazaras CN. Abbott curves characteristics of turned surfaces. International Journal of Machine Tools and Manufacture 2003;43:237–43. 7. Bacˇova´ V, Draganovska´ D. Analyses of the quality of blasted surfaces. Materials Science 2004;40:125–31. 8. Torrance AA. A simple datum for measurement of the Abbott of a profile and its first derivative. Tribology International 1997;30:239–44. 9. Thomas TR. Rough surfaces. 2nd ed. London: Imperial College Press; 1999.
journal of dentistry 41 (2013) 1060–1067
10. Kaidonis JA, Richards LC, Townsend GC, Tansley GD. Wear of human enamel: a quantitative in vitro assessment. Journal of Dental Research 1998;77:1983–90. 11. Howell JR, Blunt CD, Hooper RM, Lee AJC, Ling RSM. In vivo surface wear mechanisms of femoral components of cemented total hip arthroplasties. The Journal of Arthroplasty 2004;19:88. 12. Whitehead SA, Lo LY, Watts DC, Wilson NH. Changes of surface texture of enamel in vivo. Journal of Oral Rehabilitation 1997;24:449–53. 13. Las Casas EB, Bastos FS, Godoy GCD, Buono VTL. Enamel wear and surface roughness characterization using 3D profilometry. Tribology International 2008;41:1232–6. 14. Quartarone E, Mustarelli P, Poggio C, Lombardini M. Surface kinetic roughening caused by dental erosion: An atomic force microscopy study. Journal of Applied Physics 2008;103(10):104702. 15. Ren Y-F, Amin A, Malmstrom H. Effects of tooth whitening and orange juice on surface properties of dental enamel. Journal of Dentistry 2009;37:424–31. 16. Azrak B, Callaway A, Kurth P, Willershausen B. Influence of bleaching agents on surface roughness of sound or eroded dental enamel specimens. Journal of Esthetic and Restorative Dentistry 2010;22:391–9. 17. Lutz F, Sener B, Imfeld T, Barbakow F, Schu pbach P. Selfadjusting abrasiveness: a new technology for prophylaxis pastes. Quintessence International 1993;24:53–63. 18. Gerbo LR, Lacefield WR, Barnes CM, Russell CM. Enamel roughness after air-powder polishing. American Journal of Dentistry 1993;6:96–8. 19. Fujii M, Kitasako Y, Sadr A, Tagami J. Roughness and pH changes of enamel surface induced by soft drinks in vitroapplications of stylus profilometry, focus variation
20.
21.
22.
23.
24. 25.
26.
27.
28.
29.
1067
3D scanning microscopy and micro pH sensor. Dental Materials Journal 2011;30:404–10. Maas MC. Enamel structure and microwear: an experimental study of the response of enamel to shearing force. American Journal of Physical Anthropolology 1991;85:31–49. Arends J, Christoffersen J, Christoffersen MR, Ogaard B, Dijkman AG, Jongebloed WL. Rate and mechanism of enamel demineralization in situ. Caries Research 1992;26:18–21. Meurman JH, Ten Cate JM. Pathogenesis and modifying factors of dental erosion. European Journal of Oral Sciences 1996;104:199–206. Shulin W. Human enamel structure studied by high resolution electron microscopy. Electron Microscopy Reviews 1989;2:1–16. Xiao F, Zheng J, Zhou ZR, Qian F. In-vitro erosion of human tooth enamel. Tribology 2009;29:163–7. Eisenburger M. Degree of mineral loss in softened human enamel after acid erosion measured by chemical analysis. Journal of Dentistry 2009;37:491–4. White AJ, Yorath C, ten Hengel V, Leary SD, Huysmans MC, Barbour ME. Human and bovine enamel erosion under ‘single-drink’ conditions. European Journal of Oral Sciences 2010;118:604–9. Attin T, Wegehaupt F, Gries D, Wiegand A. The potential of deciduous and permanent bovine enamel as a substitute for deciduous and permanent human enamel: erosion-abrasion experiments. Journal of Dentistry 2007;35:773–7. Featherstone JD, Mellberg JR. Relative rates of progress of artificial carious lesions in bovine, ovine and human enamel. Caries Research 1981;15:109–14. Meurman JH, Frank RM. Progression and surface ultrastructure of in vitro caused erosive lesions in human and bovine enamel. Caries Research 1991;25:81–7.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.intl.elsevierhealth.com/journals/jden
Using bearing area parameters to quantify early erosive tooth surface changes in enamel: A pilot study James Field a,*, Matthew German b, Paula Waterhouse c a
Restorative Dentistry, Newcastle University School of Dental Sciences, United Kingdom Dental Materials Science, Newcastle University School of Dental Sciences, United Kingdom c Child Dental Health, Newcastle University School of Dental Sciences, United Kingdom b
article info
abstract
Article history:
Objectives: Most in vitro studies investigate the erosive process using relatively simple
Received 29 April 2013
roughness parameters such as roughness average (Ra). In isolation, Ra may misrepresent
Received in revised form
the surface features. Further, few studies report baseline surface characteristics after
18 July 2013
sample preparation. This study aimed to test the hypothesis that measuring the bearing
Accepted 22 August 2013
area parameters in addition to Ra may be useful when qualifying the surface of enamel at baseline and after an erosive challenge. The null hypothesis for this study was that the bearing area parameters provide no more useful information than Ra alone, when qualifying
Keywords:
the surface of enamel at baseline and after an erosive challenge.
Erosion
Methods: Enamel slabs (n = 20) were prepared from human (n = 2) and bovine (n = 4) incisor
Roughness
teeth and polished with 0.05 mm paste. Roughness average (Ra) and bearing parameters (MR1,
Bearing parameters
MR2, Rpk, Rk, Rvk) were used to record baseline characteristics. Specimens were subjected to
Bovine enamel
erosion with 1% citric acid solution for 1 min. Profilometric characteristics were recorded post-
Human enamel
erosion, along with the maximum height changes within the profile. T-tests were carried out in
In vitro
order to compare baseline surface characteristics between tissue types. Post-erosion, analysis of variance (ANOVA) was used to test the effects of tissue type (bovine or human). Results: There was no significant difference in Ra between human and bovine incisor enamel at baseline (human 0.11 mm, bovine 0.12 mm P > 0.05), and no significant difference was observed post-erosion (human 0.23 mm, bovine 0.20 mm P > 0.05). There were significant differences in bearing parameters at baseline and post-erosion that were not identified by the Ra measurement alone. Conclusions: The results suggest that if Ra alone is measured, important differences in surface characteristics may be missed. The null hypothesis is rejected, and the recommendation is made that bearing parameters are included within profile measurements in order to further triangulate the results of surface analysis studies. Clinical relevance: In isolation, Ra may misrepresent the surface features of a profile. These results have shown that the bearing parameters are an important and informative set of measurements. The recommendation is made that bearing parameters are included within profile measurements at baseline and post-erosion, in order to further triangulate the results of surface analysis studies. # 2013 Elsevier Ltd. All rights reserved.
* Corresponding author at: Room D5.014, Level 5, School of Dental Sciences, Newcastle University, Framlington Place, Newcastle NE2 4BW, United Kingdom. Tel.: +44 191 222 8515. E-mail address: [email protected] (J. Field). 0300-5712/$ – see front matter # 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jdent.2013.08.015
journal of dentistry 41 (2013) 1060–1067
1.
Introduction
This study follows a review of methods to measure surface change on dental hard tissues in vitro.1 The review concluded that profilometry (and in particular roughness average, Ra) is the main reported measurement within dental surface studies. The review highlighted the potential drawbacks of reporting Ra alone and also proposed that other parameters could be reported from the same set of profilometric data that would allow a further and more meaningful description of the surface quality. These parameters are known as bearing parameters. The main drawbacks with Ra are that the value contains no information about the textural characteristics of a profile, the likelihood of future wear or wear-resistance, the rate of future wear or the potential of a surface to retain fluids/lubricants. This makes the qualitative assessment of the surface relatively limited and empirical and only relays information about fixed heights at regular intervals. In 1933, Abbott and Firestone defined the bearing area fraction.2 This fraction or ‘ratio’ represents the cumulative distribution of the lengths of individual plateaux that would result if the surface were abraded down to a level plane at that height. The curve is normalised by the total assessment length. If this ratio was plotted against sample height, the bearing area curve results (Fig. 1). A variety of sources describe how to interpret the bearing curve.3–6 The total height of the curve is known as the roughness total. The width of the curve represents the percentage of material at different profile amplitudes. Five parameters are determined from the curve – peak roughness (Rpk), core roughness (Rk), valley roughness (Rvk), material ratio of peaks (MR1) and material ratio of troughs (MR2). These bearing parameters are defined using German standard DIN 4776 and more recently ISO 13565-2. The Rpk value represents the height of the material peaks – the area that will come into contact with an opposing surface first; in engineering it indicates the amount of material that will wear away during early contact of the machined surface. The Rk value represents the height of the material core – the area that will support most of the contact load (in engineering
1061
this represents the longer-term running surface). The Rvk value represents the depth of the material troughs/valleys – the area that will retain fluid or lubricant. These are known as the reduced peak height, core roughness and reduced trough depth respectively. The curve can also relate Rpk and Rvk to a proportion of the material’s surface (Figs. 1 and 2). The latter are known as the material ratios – MR1 relates to the proportion of peaks within the sample, and MR2 relates to the proportion of troughs. It has been suggested that the bearing curve is better for assessing surfaces that are subjected to several different machining processes than Ra.3,7,8 One process may remove the peaks but then a finer texture may be superimposed onto the resulting plateaux by a subsequent process; the deep valleys may remain unaffected. The resulting surfaces are known as ‘multi-stratified’.9 Characteristically, these surfaces are negatively skewed which makes it difficult for a parameter such as Ra to represent them effectively. In relation to tooth tissue, the bearing curve could also be used to measure the surface effects of a synergistic process, like that of erosion and abrasion. It may be possible using the curve, to explain the bi-phasic pattern of tooth surface loss which has been previously identified and described as ‘running-in wear’ and ‘steady-state wear’.10 The bearing curve has been used to assess engineered surfaces in medicine, particularly that of femoral stem wear.11 Its use to monitor and qualify labial erosion in vivo (via silicone models) was piloted in 199712 and, more recently, it has been used to describe the native enamel surface.13 The latter study showed the potential use of the bearing curve in further qualifying the tooth surface; however data were presented in tabular form only. There may have been more value in displaying the bearing area curves for visual comparison. This study aimed to test the hypothesis that measuring the bearing area parameters would allow further qualification of the surface effects on human and bovine enamel at baseline and after an erosive challenge. The null hypothesis for this study was that the bearing area parameters provide no more useful information than Ra alone, when qualifying the surface of enamel at baseline and after an erosive challenge.
Fig. 1 – Quantifying parameters on the bearing curve. Rpk, Rk and Rvk values relate to peak, core and trough average roughness values. MR1 and MR2 relate to the percentage of the surface occupied by peaks or troughs respectively.
1062
journal of dentistry 41 (2013) 1060–1067
Fig. 2 – Comparison of two different surfaces with similar Ra using the bearing curve. Surface A has significantly more, deeper troughs (higher Rvk and higher MR2). Surface A therefore has a much greater pooling/lubricative potential. Surface B has significantly more, higher peaks (higher Rpk and higher MR1) and is therefore likely to suffer significantly more early surface loss in the future.
2.
Materials and methods
2.1.
Specimen collection and preparation
Extracted human lower permanent incisor teeth were collected from the adult Dental Emergency Clinic at Newcastle Dental Hospital, UK between February 2008 and February 2011. The teeth were stored in a 1% (w/v) aqueous solution of sodium p-toluenesulfonylchloramide hydrate (Chloramine-T; Sigma–Aldrich, Dorset, UK) solution and suitable teeth, showing no signs of coronal caries or tooth surface loss were entered into the Newcastle Tissue Bank (Human Tissue Act license number 12534), stored at 4 8C. Bovine permanent incisor teeth were harvested on two occasions – March 2010 and December 2010 from the same abattoir – Linden Foods, Burradon, Cramlington (Registered Plant Number 2056, Food Standards Agency, Department of Environment, Food and Rural Affairs). The cattle were male beef Shorthorn cattle and were aged approximately 18–20 months.
4 bovine and 2 human incisor crowns were sectioned from the root of the tooth and prepared with two equidistant and parallel undercut reference grooves across the labial surface (mesio-distally) using a pear-shaped high-speed diamond bur (Hi-Di1 medium diamond bur ISO 237/012; Dentsply, Addlestone, UK). The crowns were then positioned into individual casting moulds (Buehler, Du¨sseldorf, Germany) with the labial surface facing upwards and the cemento-enamel junction perpendicular to the base. They were held in place with sticky wax (Kemdent; Swindon, UK) and cast in acrylic resin (Bonda; Bondaglass-Voss, Beckenham, UK). Once set, the casts were removed from the moulds. The labial surface was ground flat at grit size 80 (Metaserv rotary pregrinder C200/RB; Metallurgical services, Ohio, USA) until the majority of the labial enamel surface was exposed. The sample was then further polished at successive grit sizes up to 1200 and then finished for 25 s (anticlockwise rotation at light pressure) with a 0.05 mm polishing paste (Metaserv universal polisher C200/5V; Metallurgical Services, Ohio, USA). The samples were rinsed with a balanced salt solution (HBSS; Life Technologies, Paisley, UK), and inspected under the light microscope (Apophot; Nikon, Surrey, UK) for visible defects and machine marks. The polished labial surfaces were then sectioned coronally using a low-speed water-cooled diamond wheel saw (Testbourne 650 CE; South Bay Technologies, California, USA) to produce reference slabs at least 1.5 mm wide. Some slabs became damaged during the sectioning process and when possible these were re-polished. Otherwise they were discarded. Specimens were stored in HBSS at 4 8C until required. Four samples were obtained from each of the four bovine crowns, and two samples from each of the two human crowns. 20 samples were prepared in total. The baseline surfaces were then profiled using a stylus profilometer and its associated software (Mitutoyo Surftest SV-2000 and Surfpak-SV V1.600; Mitutoyo, Halifax, UK). The instrument range was 800 mm with a contact force of 4 mN. The stylus was a diamond cone tip held at 908 to the surface, with a 5 mm radius. Average roughness values, and bearing area parameters (Rk, Rvk, Rpk, MR1 and MR2) were recorded twice for each sample 0.5 mm apart. Each evaluation length
Table 1 – Mean roughness and bearing parameters of human (H) and bovine (B) enamel at baseline and post-erosion. Standard deviations are within brackets. Differing superscripts (a/b) denotes a significant baseline/erosion effect. Values in bold denote a significant tissue effect. Roughness average (Ra) (mm)
H
B a
Peak roughness (Rpk) (mm)
H a
B a
Core roughness (Rk) (mm)
H a
a
B
Valley roughness (Rvk) (mm)
H a
B a
a
Material ratio of peaks MR1 (%)
Material ratio of troughs MR2 (%)
H
H
a
Baseline
0.11 (0.02)
0.12 (0.11)
0.20 (0.03)
0.16 (0.08)
0.31 (0.07)
0.33 (0.05)
0.36 (0.10)
0.30 (0.09)
21 (3)
Eroded
0.23b (0.08)
0.2b (0.1)
0.50b (0.05)
0.43b (0.05)
0.21b (0.003)
0.12b (0.01)
0.59b (0.05)
0.54b (0.10)
12b (1)
B a
a
7 (2)
92 (3)
8.01b (0.80)
93a (3)
B 81 (9)
Maximum height change within the eroded profile (mm) H
B
0.7 (0.2)
0.41 (0.20)
a
71b (2)
journal of dentistry 41 (2013) 1060–1067
included 5 readings with a 0.3 mm cut-off (1.5 mm total evaluation length, starting within the body of the left hand reference layer) and was analysed through a Gaussian filter.
fresh solution was used for each sample. After treatment the samples were rinsed with fresh HBSS and profiled again.
2.3. 2.2.
1063
Analysis
Sample exposure procedures
Citric acid solution was formulated at 1% (w/v) and the pH was measured (Thermo Orion 4 Star; Fisher Scientific, Leicestershire, UK) at pH 2.2. The solutions were warmed using a water bath to 30 8C. Enamel samples were immersed for 1 min without stirring and
All statistical analyses were carried out using Sigmaplot for Windows Version 11.0, Build 11.0.1.80 (Systat software, Chicago, USA). T-tests were carried out in order to compare baseline surface characteristics between tissue types. On the occasions that normality tests were failed, the Mann Whitney rank sum test was used (median values reported).
Fig. 3 – Typical stylus profilometry profile (Ra 150 mm) and bearing curve (MR1 11%, MR2 89%) for baseline lapped bovine enamel.
1064
journal of dentistry 41 (2013) 1060–1067
Fig. 4 – Typical stylus profilometry profile (Ra 129 mm) and bearing curve (material ratio 1, MR1 9.5%, material ratio 2, MR2 90%) for eroded human enamel (1% citric acid for 1 min).
Post-erosion, analysis of variance (ANOVA) was used to test the effects of tissue type (bovine or human).
significant differences were found between species in terms of MR1 and MR2 (P < 0.001; both cases); human enamel displayed significantly more peaks and fewer troughs than bovine enamel.
3.
Results
3.2.
3.1.
Baseline
Post-erosion parameters are shown in Table 1. A typical stylus profile and bearing curve for human enamel are shown in Fig. 4. Eroded roughness average (Ra) values were significantly different to baseline values (P < 0.001); becoming rougher post-erosion. Ra was not significantly affected individually by tissue type (P = 0.085).
Baseline parameters are shown in Table 1. A typical profile and bearing area curve for bovine enamel are shown in Fig. 3. At baseline, no significant differences were found between the human and bovine samples in terms of their Ra (P = 0.176), Rk (P = 0.280), Rvk (P = 0.066) or Rpk (P = 0.177). However, very
Post-erosion
journal of dentistry 41 (2013) 1060–1067
Eroded core roughness (Rk) values were significantly different to baseline values (P = 0.007), becoming significantly less rough post-erosion. Rk was significantly affected by tissue (P < 0.001; Rk mean human 0.21, bovine 0.12). Eroded peak roughness (Rpk) values were significantly different to baseline values (P < 0.001), becoming significantly more rough post-erosion. Rpk was significantly affected by tissue (P < 0.001; Rpk mean human 0.50, bovine 0.43). Eroded valley roughness (Rvk) values were significantly different to baseline values (P < 0.001), becoming significantly more rough post-erosion. Rvk was not significantly affected by tissue type (P = 0.679; Rvk mean human 0.59, bovine 0.54). The proportion of eroded profile peaks (MR1) were significantly different to baseline values (P < 0.001). MR1 was significantly affected by tissue (P < 0.001; MR1 mean human 12, bovine 8). The proportion of bovine eroded profile troughs (MR2) were significantly different to baseline values (P = 0.013), with significantly more troughs post-erosion. MR2 was significantly affected by tissue (P < 0.001; MR2 mean human 93, bovine 71). The maximum height changes of the eroded profiles were significantly affected by tissue (P = 0.028; mean human 0.7 vs. bovine 0.41).
4.
Discussion
This study used a novel method of reporting profilometric measurements, in order to describe differences in surface characteristics between tissue types at baseline and posterosion. The findings suggest that if Ra alone is measured, subtle differences in surface characteristics may be missed. Indeed surface changes may be able to be modelled more effectively if these extra baseline characteristics are recorded.
4.1.
Baseline measurements
When polished with a 0.05 mm paste, the resultant surfaces of human and bovine enamel showed roughness average values of 0.113 mm and 0.120 mm respectively. These values were not significantly different to one another and this may be due to the small sample size used. However no published studies to date report a direct comparison between human and bovine incisor teeth after baseline preparation, and this area undoubtedly requires further study. Few studies report baseline human enamel roughness. Quartarone14 reported a range of control values for human enamel of between 0.05 mm and 0.120 mm, and although Rq was reported rather than Ra, the upper limits of the reported values are similar to the findings from this work; unfortunately the study did not detail any particular technique for enamel preparation and so further comparison is not possible. Higher control values have been reported by Ren15 who reported baseline Ra between 1.20 mm and 1.40 mm after polishing buccal and lingual surfaces of human third molars with 0.3 mm paste. Azrak16 also recorded similar Ra values (median 1.57 mm) for the labial surfaces of human lower incisors; the enamel surfaces were prepared using a pumice slurry and then polished with silicon carbide and aluminium oxide discs (to unreported particle and grit sizes). Given that pumices can contain particles up to 50 mm in diameter,17 it is
1065
unlikely that any amount of fine polishing will remove surface defects obtained due to treatment with pumice slurry, and this may explain the significantly higher roughness values obtained. Even fewer studies report bovine baseline enamel roughness; Gerbo18 reported values between 3.4 mm and 3.8 mm for bovine incisors but this was of the native enamel surface only. No surface preparation had been carried out. Fuji19 reported baseline Ra values for bovine incisors of between 4.7 nm and 6.4 nm. These values are significantly lower than those obtained in this study. The enamel samples were polished with a 0.25 mm paste and then cleaned ultrasonically to remove paste debris before measurement with a stylus tip of radius 2.5 mm; it is unknown whether the stylus tip was able to accurately measure roughness in the nanometre range, and it may be the case that the values reported do not relate to the surface finish. Although the paste used in this study was of a smaller particle size, it is unlikely that this would be responsible for the significantly smoother surface finish; it has already been demonstrated that grit size does not correlate well with prismatic enamel roughness.20 Further, it is unknown what effect the ultrasonic instrumentation had on the enamel surface; shearing of enamel peaks may have significantly reduced the surface roughness average, and this would warrant further investigation. Fuji also measured roughness average values using focus variation 3D microscopy, which gave higher Ra values than the reported profilometry values, but still significantly lower than expected (15–17 nm). Although in the current investigation the majority of baseline roughness parameters were not significantly different between human and bovine tissue, two parameters were significantly different. There were significantly more peaks within the human samples (higher MR1 value) and significantly more troughs within the bovine samples (lower MR2 value). To date these parameters have not been used to compare the lapped enamel surface and it is unknown to what degree the underlying enamel morphology affected the lapping process. It is proposed that a profile with an increased number of isolated profile peaks would be more susceptible to early erosive or abrasive challenges; an increased number of isolated peaks may increase the surface area that comes into immediate contact with an erosive solution. These peaks may be relatively weak and under-supported increasing the chances of structural failure or loss. It is also proposed that a profile with an increased number of profile troughs would be more likely to retain or trap fluid within the lower third of the profile. With an erosive challenge, this may serve to protect the surface if the trapped solution becomes supersaturated and is unable to be replaced. Troughs may also have implications for retaining remineralising solutions such as mouthwashes, or increasing the lubricative potential and exacerbating abrasive forces. These theoretical proposals undoubtedly require further study.
4.2.
Post-erosion
At the nano-level, an increase in roughness post-erosion may be explained by the erosive challenge preferentially ‘punching out’ the centre of the apatite crystals.21 On a micro-scale,
1066
journal of dentistry 41 (2013) 1060–1067
others describe the resultant ‘honeycomb’ pattern typically seen in human enamel,22,23 resulting from early erosion of prism and prism junctions leaving behind steep ridges of inter-prismatic enamel.24,25 Despite few studies that compare baseline to eroded profiles, these findings are in keeping with other published work; Quartarone14 reported a ‘remarkable’ increase in enamel roughness after acidic exposure (over 500%) and although the current study demonstrated a much more modest increase in roughness (around 100%) the total submersion times were significantly shorter (1 min vs. 280 min) in order to simulate surface changes rather than gross surface loss. In this study, the eroded roughness average of human and bovine enamel was not significantly different. This mirrors the baseline trend. Despite similar roughness averages, human samples exhibited a greater profile depth range after the erosive challenge. This effect may be directly linked to the increased number of isolated profile peaks at baseline. This finding is in keeping with recent work which reported that human enamel loss was significantly greater than bovine enamel loss when exposed to citric acid (pH 3.2) for up to 5 min.26 This difference may also be due to the tissue-specific proportions of prismatic to interprismatic enamel; it has already been reported that bovine tissue contains a greater proportion of interprismatic enamel than human tissue, and this interprismatic enamel is more resistant to erosion than prismatic.24 The conclusion that bovine enamel has the potential to be more resistant to erosion than human enamel, mirrors the results from the current study. A number of studies contradict these findings,27,28 suggesting that bovine enamel is more susceptible to erosion than human. The acidic challenges used however were more aggressive (at greater concentrations and lower pH, for longer periods of time). Indeed the work by White et al.26 has also shown that as the acid-exposure time increases beyond 5 min, bovine enamel then begins to display a greater surface loss than human enamel. This effect may be due to the increased proportion of troughs within the bovine enamel at baseline, which present as the erosive challenge becomes more aggressive. Less aggressive erosive challenges also result in more contrast in relief between prismatic and aprismatic enamel,29 and so the behaviour of bovine and human enamel may change significantly depending on the type of erosive insult. This has implications when attempting to compare results from different research groups or even between similar experiments that utilise disparate erosive solutions and/or erosive regimes; ultimately it adds further complexity to whether bovine enamel is a suitable model for erosion on human teeth, and the results from this pilot experiment suggest that it is not. Despite the eroded Ra values being similar between tissues there were, again, significant differences in bearing parameters between the human and bovine samples. These followed the baseline trends however post-erosion, the peak and valley roughness were also increased, whilst the core roughness showed an overall decrease. These results may suggests that the eroded surface is more susceptible to further early surface loss than at baseline, but that perhaps the surface shows a degree of resilience for further moderate abrasive challenges.
This will undoubtedly require further investigation to tease out the importance of these surface differences and how they predispose the enamel surface to further surface loss. A technique allowing visual comparison of the enamel surfaces, such as scanning electron microscopy, may prove useful in this respect.
5.
Conclusions
This pilot study has shown that when lapped with 0.05 mm paste, there was no significant difference in roughness average between human and bovine incisor enamel at baseline. Further, no significant difference in roughness average between tissues was observed post-erosion. There were, however, significant differences in bearing parameters between tissues both at baseline and post-erosion. The null hypothesis is rejected, recognising that these parameters are more sensitive to early surface change than Ra alone. Important differences in surface characteristics between the tissues were identified, and this study has shown evidence of how these differences may predispose the enamel surfaces to subsequent surface loss (either by abrasion or further erosion); this will undoubtedly require further study.
Acknowledgements The authors would like to thank Linden Foods for access to their abattoir in Burradon. JF, MG and PW conceived and designed the experiments. JF performed the experiments. JF analysed the data. JF, MG and PW wrote the paper.
references
1. Field JC, Waterhouse PJ, German MJ. Quantifying and qualifying surface changes on dental hard tissues in vitro. Journal of Dentistry 2010;38:182–90. 2. Abbott EJ, Firestone FA. Specifying surface quality. Mechanical Engineering 1933;59:569–72. 3. Pawlus P, Grabon W. The method of truncation parameters measurement from material ratio curve. Precision Engineering 2008;32:342–7. 4. Schma¨hling J, Hamprecht FA. Generalizing the Abbott– Firestone curve by two new surface descriptors. Wear 2007;262:1360–71. 5. International Standards Organisation. In: Standard I, editor. Geometrical product specifications (GPS) – surface texture: profile method; surfaces having stratified functional properties. Part 2. Height characterisation using the linear material ratio curve. 1996. 6. Petropoulos GP, Torrance AA, Pandazaras CN. Abbott curves characteristics of turned surfaces. International Journal of Machine Tools and Manufacture 2003;43:237–43. 7. Bacˇova´ V, Draganovska´ D. Analyses of the quality of blasted surfaces. Materials Science 2004;40:125–31. 8. Torrance AA. A simple datum for measurement of the Abbott of a profile and its first derivative. Tribology International 1997;30:239–44. 9. Thomas TR. Rough surfaces. 2nd ed. London: Imperial College Press; 1999.
journal of dentistry 41 (2013) 1060–1067
10. Kaidonis JA, Richards LC, Townsend GC, Tansley GD. Wear of human enamel: a quantitative in vitro assessment. Journal of Dental Research 1998;77:1983–90. 11. Howell JR, Blunt CD, Hooper RM, Lee AJC, Ling RSM. In vivo surface wear mechanisms of femoral components of cemented total hip arthroplasties. The Journal of Arthroplasty 2004;19:88. 12. Whitehead SA, Lo LY, Watts DC, Wilson NH. Changes of surface texture of enamel in vivo. Journal of Oral Rehabilitation 1997;24:449–53. 13. Las Casas EB, Bastos FS, Godoy GCD, Buono VTL. Enamel wear and surface roughness characterization using 3D profilometry. Tribology International 2008;41:1232–6. 14. Quartarone E, Mustarelli P, Poggio C, Lombardini M. Surface kinetic roughening caused by dental erosion: An atomic force microscopy study. Journal of Applied Physics 2008;103(10):104702. 15. Ren Y-F, Amin A, Malmstrom H. Effects of tooth whitening and orange juice on surface properties of dental enamel. Journal of Dentistry 2009;37:424–31. 16. Azrak B, Callaway A, Kurth P, Willershausen B. Influence of bleaching agents on surface roughness of sound or eroded dental enamel specimens. Journal of Esthetic and Restorative Dentistry 2010;22:391–9. 17. Lutz F, Sener B, Imfeld T, Barbakow F, Schu pbach P. Selfadjusting abrasiveness: a new technology for prophylaxis pastes. Quintessence International 1993;24:53–63. 18. Gerbo LR, Lacefield WR, Barnes CM, Russell CM. Enamel roughness after air-powder polishing. American Journal of Dentistry 1993;6:96–8. 19. Fujii M, Kitasako Y, Sadr A, Tagami J. Roughness and pH changes of enamel surface induced by soft drinks in vitroapplications of stylus profilometry, focus variation
20.
21.
22.
23.
24. 25.
26.
27.
28.
29.
1067
3D scanning microscopy and micro pH sensor. Dental Materials Journal 2011;30:404–10. Maas MC. Enamel structure and microwear: an experimental study of the response of enamel to shearing force. American Journal of Physical Anthropolology 1991;85:31–49. Arends J, Christoffersen J, Christoffersen MR, Ogaard B, Dijkman AG, Jongebloed WL. Rate and mechanism of enamel demineralization in situ. Caries Research 1992;26:18–21. Meurman JH, Ten Cate JM. Pathogenesis and modifying factors of dental erosion. European Journal of Oral Sciences 1996;104:199–206. Shulin W. Human enamel structure studied by high resolution electron microscopy. Electron Microscopy Reviews 1989;2:1–16. Xiao F, Zheng J, Zhou ZR, Qian F. In-vitro erosion of human tooth enamel. Tribology 2009;29:163–7. Eisenburger M. Degree of mineral loss in softened human enamel after acid erosion measured by chemical analysis. Journal of Dentistry 2009;37:491–4. White AJ, Yorath C, ten Hengel V, Leary SD, Huysmans MC, Barbour ME. Human and bovine enamel erosion under ‘single-drink’ conditions. European Journal of Oral Sciences 2010;118:604–9. Attin T, Wegehaupt F, Gries D, Wiegand A. The potential of deciduous and permanent bovine enamel as a substitute for deciduous and permanent human enamel: erosion-abrasion experiments. Journal of Dentistry 2007;35:773–7. Featherstone JD, Mellberg JR. Relative rates of progress of artificial carious lesions in bovine, ovine and human enamel. Caries Research 1981;15:109–14. Meurman JH, Frank RM. Progression and surface ultrastructure of in vitro caused erosive lesions in human and bovine enamel. Caries Research 1991;25:81–7.