122 JOURNAL OF COSMETIC SCIENCE Tooth bleaching became an accepted routine treatment in dental offices in the 1970s (5). The use of liquid hydrogen peroxide (H2O 2 ) in dentistry, described as early as 1884, is currently the most common procedure (6-13). Several dentifrices containing H 2 O 2 are also currently being marketed (14). Generally, tooth bleaching has been found to be effective on discolored teeth. On the other hand, the use of such strong oxidizing agents has raised questions as to adverse side effects on tooth structure and restorative materials. Tooth sensitivity is a potential side effect of dentist-dispensed home tooth-whitening systems (15 ). In the past decade, numerous studies have evaluated the effects of peroxide­ containing bleaching agents on tooth hard tissue. Most of the studies found insignificant alterations of the enamel surface (16-24). Some researchers, however, actually observed enamel surface alterations by surface analysis techniques (25,26). Especially when pa­ tients had enamel cracks or other damage, careful treatment has been proposed to be necessary, and fluoride treatment has been recommended to lessen discomfort (27). The manufacturers of the in-home bleaching systems generally provide these products with various extras such as fluoride rinses and pre- and post-bleaching treatments. The mechanism of H202 or fluoride has been individually investigated and described as follows: H 2 Orcontaining bleaching agents remove tooth discolorations, where H202 acts as an oxygenator and an oxidant and then affects the organic phases in the dental enamel (11,28). The caries-preventive effect of fluoride has been well known for many years, and the mechanism has also been proposed (29-3 7). According to the current concepts on the action mechanism of fluoride (38-40), the caries-preventive effect is mainly attributed to the enhancement of remineralization at the interface of the tooth and oral fluids. Although the individual mechanisms of H 2 O 2 and fluoride are fairly well known, the combined use of both H 2 O 2 and fluoride complicates the action mechanism, and the problem of how these two components react all at once with the enamel surface has not been fully understood. The use of fluoride for reducing the discomfort caused by H 2 O 2 has been recommended without an understanding of the detailed mechanism. Thus, this study was undertaken to examine the mechanism of NaF added to lessen the adverse side effects of H2O2 using hydroxyapatite (HAP) as a model material of tooth enamel. X-ray photoelectron spectroscopy (XPS) studies have made it possible for the first time to observe directly the successive change in the ratio of fluoridated hydroxy­ apatite (FHAP) and calcium fluoride (CaF 2 ) formed on the surface of the HAP. We also have clarified the reaction mechanism, discovering that FHAP or CaF 2 is produced not simply by depending on the fluoride concentration as shown elsewhere (3 5 ), but also by depending on the H2O2 concentration. The information about the mechanism and the methodology should be useful in developing bleaching agents. METHODS MATERIALS Hydroxyapatite (HAP Ca/P molar ratio = 1.65 surface area = 8 m2g- 1 ), dehydrated dicalcium phosphate (DCPD), sodium fluoride (NaF), and hydrogen peroxide (H 2 O 2 , 30%) were purchased from Wako Pure Chemicals Co., Japan. Fluorapatite (FAP) crystal from Mexico was powdered, and used as XPS standard material. Calcium fluoride (CaF 2 ), purchased from Kan to Chemicals Co., Japan, was also used as XPS standard material. NaF and H 2 O 2 were diluted with distilled water to given concentrations. The powder

H202/NaF TOOTH-BLEACHING SYSTEM 123 samples of HAP and DCPD, substitute minerals for dental hard tissue, were dispersed in NaF and/or H202 aqueous solutions to investigate the reaction characteristics. A stained anterior tooth, which was offered from a forty-year-old male, was used for analysis of its amino acid composition. REACTION OF HAP AND DCPD NAF AND Hp2 SOLUTIONS The typical experimental procedure was as follows: NaF was dissolved in 20 cm3 of H 2 0 2 solutions (15%, pH 5.0 and 30%, pH 4.7) such that the NaF concentrations were 0.0526-0. 526 mol/dm - 3 , i.e., 1,000-10,000 ppm of fluorine. To these solutions, HAP or DCPD (0.5 g) was added, then stirred in a water bath at 25°C for given times. The HAP or DCPD solids were separated from the solutions by filtration (Advantec Co., filter #2), rinsed with 50 cm3 of deionized water, dried at 105°C, and then examined by X-ray diffraction (XRD) and XPS as shown in the next paragraph. The solutions were analyzed for dissolved Ca2 + and PO/- by ICP-AES (Seiko Instruments Inc., SPS 1500). CHARACTERIZATION OF REACTION PRODUCTS BY XRD AND XPS The treated HAP and DCPD samples were examined by XRD (Rigaku Co., RINT 2000) and XPS (ULVAC-PHI Co., ESCA 5100 Mgk et= 1253.6 eV). The XPS studies were carried out in order to examine the chemical states of elements in the HAP, following our previous work (36,3 7). The powder samples were dusted on polymer film tapes that were mounted on a sample holder. First wide-scan spectra, then narrow-scan spectra for all detected elements were obtained. For characterization of fluorine, an Auger parameter (41), defined as the sum of the binding energy (Eb) for the Fls line and the kinetic energy (Ek) for the F(KLL) Auger line, was utilized. The Ek of the Auger line was plotted against the Eb of the photoelectron line, where the Ek of the Auger line was calculated by subtracting energy at the corrected Auger line position on the spectrum from the photon energy (1253.6 eV), as shown in the equations below. The Auger parameter grids were drawn as a family of lines with slopes of 45 degrees, where all the points on a line had the same value for the Auger parameter because the static charges were rationally cancelled. In the present work, however, the Cls line from adventitious hydrocarbon, which nearly always appears at 284.6 eV, was used for charge correction thus, each chemical state was supposed to occupy not only a whole Auger parameter line but also a unique point on the line: Auger parameter = Eb of Fls line + Ek of F(KLL) Auger line. (Ek of F(KLL) Auger line = 1253.6 - energy at Auger line position on the spectrum.) RESULTS DISSOLUTION OF HAP AND THE FOLLOWING STRUCTURAL CHANGES IN Hp 2 SOLUTION The dissolved Ca2 + and H 2 P0 4 - ions from powdered HAP in H 2 0 2 solution (15%, pH 5.0 and 30%, pH 4.7) were analyzed results are listed in Table I. The dissolution of HAP was accelerated by H 2 0 2 addition, where the driving force was the acidity of the solution as shown, for example, by equations 1 and 2, though the differences due to the concentration of H20 2 were not significant. The pH values in the solutions changed to