Since stylus profilometry applies a force on the sample surface, it is logical to hypothesize that the profilometer penetrates the surface of the enamel softened by acid solutions. The aims of the present study were, therefore, to test the hypothesis that surface profilometry measurements of eroded enamel alter the surface of the enamel, to quantify the potential effect of the surface alteration (scratches) on the measured values of enamel erosion by atomic force microscopy and to compare the values of enamel loss caused by erosion as measured by profilometry and non-contact confocal laser scanning microscopy (CLSM).
Enamel samples, cut from unerupted human third molars were treated with Volvic Mineral Water and citric acid solutions of different pH values. The enamel material loss was measured by two different contact profilometers and a reflection mode CLSM. The scratches depth was analyzed by atomic force microscopy.
Our study demonstrated that the tip of the profilometer penetrated the surface of eroded enamel during the profilometry measurements, leading to clearly visible surface scratches on the enamel samples. The profilometers created surface scratches of a depth ranging from 57.6 (47.1) nm to 577.1 (157.6) nm on the surface of the eroded enamel and led, therefore, to a larger measured value of erosion. It was shown that the depth of the scratches depends on the pH value, the erosion time and the profilometer used.
With few exceptions profilometers deliver reliable values of erosive enamel material loss, although they create surface scratches on eroded enamel. Reflection mode CLSM is a non-tactile, fast and precise method for analyzing enamel erosion quantitatively in vitro.
Human dental enamel is one of the hardest biological materials in the body. While fluorapatite is highly resistant against acidic dissolution, hydroxyapatite can be dissolved by several acids commonly found in beverages. Beverages like soft drinks frequently contain citric acid or phosphoric acid. The chemical dissolution of dental enamel without the involvement of bacteria has been defined as enamel erosion .
It has been proposed that dental erosion progresses in three different stages of enamel erosion . The first stage of erosion is called initial demineralization. Acids which dissolve Ca 2+ and PO 4 3− ions soften the enamel surfaces significantly . At this stage the enamel is susceptible to further damage, as different studies have shown . The softened enamel collapses in the second stage of enamel erosion because of further exposure to acids and mechanical stress caused by tooth brushing, mastication or bruxism . As a result, the collapsed enamel is removed in the third stage, the underlying unaffected enamel is exposed and the enamel erosion process starts over . Hence, enamel loss should increase with prolonged erosion times.
In general, acidic solutions with low pH values are able to dissolve higher concentrations of calcium and phosphate ions from dental enamel . Therefore, low pH value (pH 2–4) solutions should increase enamel softening and enamel loss of up to several micrometers . Several studies showed that softened enamel is very susceptible to scratching . It was found that the enamel loss caused by tooth brushing depends on the duration of time between enamel erosion and tooth brushing . Softened enamel is highly unstable and can be easily removed by short and relatively gentle physical action . Tooth brushing on eroded enamel leads to minor changes in surface morphology and mechanical properties .
It has been suggested that profilometry measurements may affect the surface eroded enamel . During scanning, the profilometer applies a force onto the enamel surface. Depending on this force and the contact area of the profilometer tip there is a compressive stress which may plastically deform the affected enamel surface. If the hardness of the softened enamel is too low, the enamel could collapse and the profilometer tip would, therefore, create a linear scratch on the surface, since the profilometer scans across the enamel surface. The scratches potentially created by the profilometer should be visible as straight lines on the eroded enamel surface. Optical microscopy would only allow qualitative observations of the scratches. It is necessary to use a quantitative technique such as atomic force microscopy to measure the depth of the scratches created on the enamel surface.
Profilometry, also called surfometry is a well-established technique often applied in surface science and dental research . The profilometer can be used to measure the contour of the surface, the profile and roughness, quantitatively creating a path-height diagram. It uses a probe consisting of a stylus with a sharp diamond tip which scans a line on the surface of the sample . To the best of our knowledge, it was used for the first time in dental research in 1972 to study the abrasion of dentin caused by different toothpastes . In a study, three different instruments for measuring enamel erosion – quantitative light-induced fluorescence (QLF), transverse microradiography (TMR) and optical surface profilometry – were evaluated. The authors stated disadvantages, such as the destructive nature of most methods, like for example contact profilometry. However, they neither quantified the damage of eroded enamel nor did they cite studies which corroborate this hypothesis . To our best knowledge no study has so far quantified how profilometry measurements affect or change the surface of eroded enamel.
Another method for analyzing eroded enamel surfaces is confocal laser scanning microscopy (CLSM). This method produces images of a sample surface by scanning the surface with a laser and by using the principle of confocal imaging. Thus, it is possible to measure the height difference between an eroded enamel surface and a given reference area. The effect of erosion or abrasion on dentin has been studied by confocal laser scanning microscopy previously. So far only a few studies describe the use of optical profilometry, laser scanning microscopy or white light microscopy for investigating enamel erosion . Two different studies have been carried out using CLSM for detecting dental enamel erosion either qualitatively or by quantitative measurements .
A third method to quantify demineralization and erosion of dental tissue is atomic force microscopy (AFM), which uses the attractive and repulsive forces between the surface and a tip to detect different surface characteristics . The high resolution and the ability to measure the surface hardness by AFM nanoindentation make the AFM a very useful device for analyzing dental erosion .
The aims of the present study were (i) to test the hypothesis that surface profilometry measurements of eroded enamel alter the surface of the enamel, e.g. by penetration of the profilometer tip into the softened enamel surface, (ii) to quantify the potential effect of such surface alteration on the values of enamel erosion as measured by atomic force microscopy, and (iii) to compare the values of enamel loss caused by erosion as measured by profilometry and non-contact confocal laser scanning microscopy (CLSM).
Materials and methods
Samples were prepared from non-erupted human third molars which had been extracted for medical reasons. The teeth were disinfected in 1%-thymol solution (Merck, Darmstadt, Germany) until preparation. For cleaning, teeth were stored in sodium hypochloride (Carl Roth, Karlsruhe, Germany) for 24 h. Organic remains were removed by carefully using a dentist’s set of instruments.
The enamel surface of each tooth was sectioned by a low speed diamond saw (Bühler Isomet, Bühler, Düsseldorf, Germany). The enamel pieces, which had a size of 3 mm × 2 mm × 1.5 mm, were embedded in epoxy resin (Stycast 1266 Emerson & Cumming, ICI Westerlo, Belgium). After a polymerization time of 12 h, samples were ground with SiC paper and polished with monocrystalline diamond suspension with a particle size ranging from 6 μm to 1 μm. The samples were stored in deionized water after preparation. One half of the polished surface was covered with PVC adhesive tape (Tesapack ultra strong, Tesa, Hamburg, Germany) to create a reference area unaffected by erosion . The uncovered part of the enamel surface was subsequently cleaned with a cotton swab and ethanol. Since some glue could remain at the reference area after removing the adhesive tape, this part of the tooth surface was cleaned in the same way after enamel erosion. Fig. 1 a shows a sketch of such a sample.
Erosion of the enamel samples
From 60 teeth obtained, each tooth was cut into manageable samples. From each of these cut samples a random piece was taken from each tooth and divided into three groups (A–C), 9 samples per group. Two aqueous (deionized water) solutions of citric acid (Merck Schuchardt, Hohenbrunn, Germany) were prepared for the erosion test of the samples. Citric acid (CA) was dissolved in deionized water in two different concentrations: CA 1 was a 0.1 M CA solution with pH 2.3, whereas CA 2 was a 0.01 M CA solution with pH 3.3. The samples of the control group were treated with Volvic Mineral Water (VMW) (Volvic Naturell, Danone Waters Deutschland, Wiesbaden, Germany). The mineral water pH was 7.1. The pH values of all solutions were measured with a pH-meter (Knick pH-meter 765 Calimatic, Germany).
Samples were treated in 25 ml of CA 1 (group A), CA 2 (group B) and VMW (group C) at room temperature. To prevent an increased concentration of dissolved Ca 2+ and PO 4 3− ions at the sample surface during the erosion process, the solutions were agitated with an automatic shaker (Heidolph Polymax 2040, Heidolph Instruments, Schwabach, Germany) at 50 rpm. Three samples from each group were removed from the solution after 10, 20 and 40 min of erosion time, respectively. The sample surfaces were rinsed with deionized water for 10 s and subsequently dried with compressed air.
After removing the tape, all sample surfaces were examined using an optical light microscope (Leica MZ8, Leica Microsystems, Wetzlar, Germany) at a 5× magnification. Any adhesive remnants from the tape found on the reference area were removed carefully with ethanol and a cotton swab. All samples from each group, arranged by erosion time, were fixed in rows onto a single glass slide and stored in Petri dishes, in air, until use.
Calibration of the profilometers, the CLSM and the AFM
The profilometers, the CLSM, and the AFM were calibrated in z -direction using a standard for profilometer calibration (Carl Zeiss Jena, Jena, Germany) with rectangular trenches which have a rectangular cross-section. The height of the calibration standard profile was 4.2 (±0.1) μm. A line scan was made to check the accuracy of the profilometers, whereas the section mode for 3D images was used to calibrate the CLSM and the AFM. With these methods the height of the standard was measured with each instrument at ten different positions of the height standard. After the measurements the average height and the calibration standard deviation of the standard were calculated for each instrument.
Prior to taking measurements, a glass slide with the samples was fixed onto a mount with double-sided adhesive tape to prevent the samples from moving during the measurements. Two profilometers were used to compare the effect of two different tip radii, loads, and sensor detection technologies on the eroded enamel surfaces.
The first instrument is a profilometer with an laser interferometric transducer (Taylor Hobson FormTalysurf Series 2, Taylor Hobson, Leicester, England). The cone shaped diamond tip has a tip radius of 2.00 μm. The testing force applied to the enamel surface was 0.87 mN. Line scans of a length of approximately 4 mm were performed with the FormTalysurf Series 2. After each line scan, the sample was moved slightly to prevent a repetition of the measurements on the scan line of the previous measurement. The lateral distance between each recorded profile was approximately 50 μm. Each sample was measured three times, scanning from the reference surface area to the eroded surface area.
The second profilometer has an inductive transducer to record height measurements (Hommeltester T 1000, Hommel-Etamic, Schwenningen, Germany). The cone shaped diamond tip of this profilometer has a tip radius of 1.95 μm. The testing force applied to the enamel surface was 1.60 mN. With the Hommeltester T 1000 each sample was also measured three times, scanning from the reference surface to the eroded surface.
After the profile measurements, the enamel samples were examined with an optical microscope in order to examine if the profilometry measurements had created visible surface scratches. The height profiles of the eroded enamel surface recorded by both profilometers were analyzed by the software according to the profilometer FormTalysurf Series 2 (Taly μltra V.184.108.40.206, Taylor & Hobson, Leicester, England). The curves were filtered by a primary filter (cut off = 0.025 mm) to eliminate the signal noise. Lastly, the fitted curves were adjusted to the baseline.
Confocal laser scanning microscopy (CLSM)
The CLSM (Zeiss 510 Meta, Carl Zeiss MicroImaging, Jena, Germany) is equipped with a helium/neon-laser with a wavelength λ = 488 nm. A 20× dry lens objective lens (Zeiss NEONFLUAR 20×, 0.5 HD) with a basic field of view of 450 μm × 450 μm was used in reflection mode. The transition zone between the eroded sample area and the reference area was positioned in the center of the field of view. The eroded enamel surface was imaged by the CLSM and three height profiles were measured on the surface. The data from CLSM were analyzed by the additional software Zeiss LSM 510 Meta. The surface image was plane fitted into a 3D image ( Fig. 1 b). Noise was filtered with a 7 × 7 Gaussian filter. The profiles at three different positions were measured and exported to ASCII files for evaluation.
Atomic force microscopy (AFM)
Topographic images of the profilometer traces were recorded with an atomic force microscope (Digital Instruments Dimension 3100, Veeco Instruments, Santa Barbara, CA, USA) operating in tapping mode and using silicon cantilevers (OMCL AC160TS-W, Olympus, Tokyo, Japan). The AFM was used because of its ability to measure the depth of the surface scratches caused by the profilometers and its extremely high resolution in x -, y – and z -directions (<1 nm) . It was necessary to capture the AFM images in tapping mode to avoid damaging the softened enamel. In this AFM operation mode lateral forces applied to the enamel surface by the AFM tip are minimized by the tapping movement of the tip. After being captured, the height images were fitted by an x – y -fit, similar to the fit used for the CLSM micrographs but without filtering. For depth measurements, the “average section tool” of the AFM software (Veeco NanoScope V. 5.12r5, Digital Instruments, Santa Barbara, CA, USA) was used. With this software tool, a line A was drawn along the bottom of each scratch of the enamel and a second line B was drawn parallel to the first one on the eroded surface ( Fig. 6 c). The software integrated the surface structure within the marked area and calculated the means <SPAN role=presentation tabIndex=0 id=MathJax-Element-1-Frame class=MathJax style="POSITION: relative" data-mathml='z¯’>z¯z¯
of the height values z ( A ) and z ( B ) on the tooth surface along each line A and B . The height of the line B was set to 0 and the difference Δ z AFM of these two mean values (Eq. (1) ) was computed.
Δ z AFM = z ¯ ( B ) − z ¯ ( A )
To analyze the influence of profilometric scratches on the measured enamel loss by profilometer, the depth of the scratches Δ z AFM was subtracted from the enamel loss z profilometer measured by the profilometers (Eq. (2) ).
Δ z profilometer = z profilometer − Δ z AFM
The subtracted values Δ z profilometer were compared to the tactile measured original values of enamel loss ( z profilometer ) by a one-way ANOVA.
A statistics software (StatGraphics Centurion XV Version 15.2.00, StatPoint, Herndon, VA, USA) was used to perform all statistical analyses. To determine the influence of the used instrument, the pH value and the erosion time on the measured enamel loss, a multifactor ANOVA and a Bonferroni test were performed for the eroded samples.
The depths of the scratches on the enamel surface caused by the profilometer were also measured. A multifactor ANOVA with the factors pH value, erosion time and profilometer as well as a Bonferroni test were used to investigate the influence of these factors on the depth of the scratches on the eroded samples.
In addition, a one-way ANOVA and Bonferroni test were performed to compare the statistical significance of the original values of the enamel loss recorded by the profilometers to the corrected values.
All statistical analyses were performed with a 95% confidence interval.
The measurements from the two profilometers and the CLSM were compared to each other using a correlation test. In this test, the correlation factor ρ quantifies how similar two datasets are. If the correlation factor is 1 or −1, the datasets are identical. If the correlation factor is 0, the datasets are not similar. The correlation factor is defined by
ρ = K O V ( X , Y ) S x ⋅ S y
K O V ( X , Y ) = 1 n ⋅ ∑ i = 1 n [ ( x ¯ − x ¯ i ) ⋅ ( y ¯ − y i ) ] .