Effectiveness of intraoral suction systems and aspirating tips for evacuation of aluminum oxide particles during use of air-abrasion

Aim: The purpose of this study was to compare the effectiveness of a high-volume evacuation and a conventional intraoral suction system and aspirating tips for capturing aluminum oxide particles during use of an air-abrasion device. Methods: A phantom head was fixed at the dental chair head with secured a metallic device with 5 horizontal shafts, corresponding to operator's clockrelated working positions, and one vertical shaft to simulate the operator's nasal cavity. Petri plates were fixed to the shafts at distances of 20, 40 and 60 cm from the center of the oral cavity of the phantom head to collect the aluminum oxide particles spread over during air abrasion. The dust was aspirated with two types of suction tips used with both suction systems: a conventional saliva ejector and a saliva ejector customized by the adaptation of a 55-mm-diameter funnel. Results: The amount of particles showed that the greatest abrasive particle deposition occurred at a distance of 20 cm from the center of the oral cavity of the phantom head at 9 o'clock operatory position with the conventional saliva ejector attached to high-volume evacuation system. Conclusions: The greatest deposition of aluminum oxide particles occurred at the shortest distance between the operator and the center of the oral cavity, while using the high-volume evacuation system associated to the conventional suction tip.

Silica coating on alumina ceramic: comparison of three chairside air-abrasion devices working for different times and distances.

This study evaluated, by scanning electron microscope (SEM) and EDS, the effect of different strategies for silica coating (sandblasters, time and distance) of a glass-infiltrated ceramic (In-Ceram Alumina). Forty-one ceramic blocks were produced. For comparison of the three air-abrasion devices, 15 ceramic samples were divided in three groups (N.=5): Bioart, Microetcher and Ronvig (air-abrasion parameters: 20 s at a distance of 10 mm). For evaluation of the time and distance factors, ceramic samples (N.=5) were allocated in groups considering three applied times (5 s, 13 s and 20 s) and two distances (10 mm and 20 mm), using the Ronvig device. In a control sample, no surface treatment was performed. After that, the micro-morphologic analyzes of the ceramic surfaces were made using SEM. EDS analyzes were carried out to detect the % of silica on representative ceramic surface. ANOVA and Tukey tests were used to analyze the results. One-way ANOVA showed the silica deposition was different for different devices (P=0.0054). The Ronvig device promoted the highest silica coating compared to the other devices (Tukey test). Two-way ANOVA showed the distance and time factors did not affect significantly the silica deposition (application time and distance showed no statistical difference). The Ronvig device provided the most effective silica deposition on glass-infiltrated alumina ceramic surface and the studied time and distance for air-abrasion did not affect the silica coating.

Influence of air abrasion tips and operation modes on enamel-cutting characteristics

Objective: To assess the influence of air abrasion tips and system operation modes on enamel cutting. Methods: Forty bovine teeth were abraded with the air abrasion system Mach 4.1 for 10 and 15 seconds, employing conventional and sonic tips of 0.45-mm inner diameter and a 90° angle, and 27.5-μm aluminum oxide at 5.51 bar air pressure in continuous and pulsed modes. The width and depth of the resulting cuts were measured in SEM. Results: The multivariate analysis of variances revealed that, compared to the sonic tip, the conventional tip produced shallower cuts independent of the operation mode and the application period. Conclusions: The cutting patterns observed in this study suggest that the pulsed mode produced deeper cuts when both the conventional and sonic tips were used, and that the sonic tip cut more dental tissue than the conventional one.

Air abrasion before and/or after zirconia sintering: surface characterization, flexural strength, and resin cement bond strength

The purpose of this in vitro study was to evaluate the effect of air-abrasion/zirconia sintering order on the yttria partially stabilized tetragonal zirconia polycrystal (Y-TZP) surface characterization (roughness, morphology, and phase transformation), flexural strength (FS), and shear bond strength (SBS) to a resin cement. Y-TZP specimens were air abraded with 50-μm Al2O3 particles after (AS), before (BS), or before and after zirconia sintering (BAS). For roughness (Ra), 30 block specimens (12×12×3.0 mm; n=10) had their surfaces analyzed by a profilometer. Next, on the air-abraded surfaces of these specimens, composite resin discs (n=30) were bonded with RelyX ARC. The bonded specimens were stored for 24 hours in distilled water at 37°C before shear testing. Failure mode was determined with a stereomicroscope (20×). The surface morphology (n=2) was evaluated by SEM (500×). For the four-point flexural strength test (EMIC DL2000), 39 bar-shaped specimens (20×4.0×1.2 mm; n=13) were air abraded according to the three conditions proposed, and an additional group (nonabraded) was evaluated (n=13). The quantitative analysis of phase transformation (n=1) was completed with Rietveld refinement with X-ray diffraction data. Ra (μm) and SBS (MPa) data were analyzed by one-way analysis of variance (ANOVA) and the Tukey test (α=0.05). Pearson correlation analysis was used to determine if there was a correlation between roughness and SBS. For FS (MPa) data, one-way ANOVA and the Dunnett C-test (α=0.05) were used. The air-abrasion/zirconia sintering order influenced significantly (p<0.001) Ra, SBS, and FS. The BS and AS groups presented the highest (1.3 μm) and the lowest (0.7 μm) Ra. The highest SBS (7.0 MPa) was exhibited by the BAS group, followed by the AS group (5.4 MPa) and finally by the BS group (2.6 MPa). All groups presented 100% adhesive failure. A weak correlation (r=−0.45, p<0.05) was found between roughness and SBS. The air-abrasion/zirconia sintering order provided differences in the surface morphology. The nonabraded (926.8 MPa) and BS (816.3 MPa) groups exhibited statistically similar FS values but lower values than the AS (1249.1 MPa) and BAS (1181.4 MPa) groups, with no significant difference between them. The nonabraded, AS, BS, and BAS groups exhibited, respectively, percentages of monoclinic phase of 0.0 wt%, 12.2 wt%, 0.0 wt%, and 8.6 wt%. The rougher surface provided by the air-abrasion before zirconia sintering may have impaired the bonding with the resin cement. The morphological patterns were consistent with the surface roughness. Considering the short-term SBS and FS, the BAS group exhibited the best performance. Air abrasion, regardless of its performance order, provides tetragonal to monoclinic transformation, while sintering tends to zero the monoclinic phase content.

Removal of enamel caries with an air abrasion powder

This study compared the efficiency of air abrasion on enamel caries with selective enamel powder (SEP) or with alumina powder and a negative and positive control group. Ninety-three extracted molars with non-cavitated incipient enamel lesions were selected. After embedding the roots in resin, each lesion was sectioned perpendicular to the surface and photographed. Each lesion was classified microscopically as having or not having dentin involvement. The lesions were distributed into four groups with an equal number of enamel caries with or without dentin involvement. Each group was treated differently: Group 1 had SEP abrasion, Group 2 had alumina abrasion, Group 3 had sodium bicarbonate abrasion (negative control) and Group 4 had bur treatment (positive control). The surface was rephotographed after treatment. Superimposition of the photographs identified areas of "correct-excavation," "under-excavation" and "over-excavation." There were no statistical differences between lesions treated with or without dentin involvement for Groups 2 through 4. However, in the SEP group, all measured areas were significantly influenced by dentin involvement. In pairwise comparisons, no statistical differences were found between the alumina and bur groups. The SEP group, however, showed statistically significant differences for each area compared to the alumina group in enamel caries without dentin involvement. SEP performed as well as alumina and bur in lesions with dentin involvement. SEP is different in its ablative properties toward caries with dentin involvement or no dentin involvement. In terms of dental treatment, SEP seems to have a diagnostic potential for enamel lesions before operative intervention in patients with high caries risk.

Effect of air-particle abrasion protocols on the biaxial flexural strength, surface characteristics and phase transformation of zirconia after cyclic loading

This study evaluated the effect of air-particle abrasion protocols on the biaxial flexural strength, surface characteristics and phase transformation of zirconia after cyclic loading. Disc-shaped zirconia specimens (Ø: 15mm, thickness: 1.2mm) (N=32) were submitted to one of the air-particle abrasion protocols (n=8 per group): (a) 50μm Al2O3 particles, (b) 110μm Al2O3 particles coated with silica (Rocatec Plus), (c) 30μm Al2O3 particles coated with silica (CoJet Sand) for 20s at 2.8bar pressure. Control group received no air-abrasion. All specimens were initially cyclic loaded (×20,000, 50N, 1Hz) in water at 37°C and then subjected to biaxial flexural strength testing where the conditioned surface was under tension. Zirconia surfaces were characterized and roughness was measured with 3D surface profilometer. Phase transformation from tetragonal to monoclinic was determined by Raman spectroscopy. The relative amount of transformed monoclinic zirconia (FM) and transformed zone depth (TZD) were measured using XRD. The data (MPa) were analyzed using ANOVA, Tukey's tests and Weibull modulus (m) were calculated for each group (95% CI). The biaxial flexural strength (MPa) of CoJet treated group (1266.3±158A) was not significantly different than that of Rocatec Plus group (1179±216.4A,B) but was significantly higher than the other groups (Control: 942.3±74.6C; 50μm Al2O3: 915.2±185.7B,C). Weibull modulus was higher for control (m=13.79) than those of other groups (m=4.95, m=5.64, m=9.13 for group a, b and c, respectively). Surface roughness (Ra) was the highest with 50μm Al2O3 (0.261μm) than those of other groups (0.15-0.195μm). After all air-abrasion protocols, FM increased (15.02%-19.25%) compared to control group (11.12%). TZD also showed increase after air-abrasion protocols (0.83-1.07μm) compared to control group (0.59μm). Air-abrasion protocols increased the roughness and monoclinic phase but in turn abrasion with 30μm Al2O3 particles coated with silica has increased the biaxial flexural strength of the tested zirconia. © 2013 Elsevier Ltd.

Air-particle abrasion on zirconia ceramic using different protocols: Effects on biaxial flexural strength after cyclic loading, phase transformation and surface topography

This study evaluated the effect of different air-particle abrasion protocols on the biaxial flexural strength and structural stability of zirconia ceramics. Zirconia ceramic specimens (ISO 6872) (Lava, 3M ESPE) were obtained (N=336). The specimens (N=118, n=20 per group) were randomly assigned to one of the air-abrasion protocols: Gr1: Control (as-sintered); Gr2: 50 μm Al2O3 (2.5 bar); Gr3: 50 μm Al2O3 (3.5 bar); Gr4: 110 μm Al2O3(2.5 bar); Gr5: 110 μm Al2O3 (3.5 bar); Gr6: 30 μm SiO2 (2.5 bar) (CoJet); Gr7: 30 μm SiO2(3.5 bar); Gr8: 110 μm SiO2 (2.5 bar) (Rocatec Plus); and Gr9: 110 μm SiO2 (3.5 bar) (duration: 20 s, distance: 10 mm). While half of the specimens were tested immediately, the other half was subjected to cyclic loading in water (100,000 cycles; 50 N, 4 Hz, 37 °°C) prior to biaxial flexural strength test (ISO 6872). Phase transformation (t→m), relative amount of transformed monoclinic zirconia (FM), transformed zone depth (TZD) and surface roughness were measured. Particle type (p=0.2746), pressure (p=0.5084) and cyclic loading (p=0.1610) did not influence the flexural strength. Except for the air-abraded group with 110 μm Al2O3 at 3.5 bar, all air-abrasion protocols increased the biaxial flexural strength (MPa) (Controlnon-aged: 1030±153, Controlaged: 1138±138; Experimentalnon-aged: 1307±184-1554±124; Experimentalaged: 1308±118-1451±135) in both non-aged and aged conditions, respectively. Surface roughness (Ra) was the highest with 110 μm Al2O3(0.84 μm. FM values ranged from 0% to 27.21%, higher value for the Rocatec Plus (110 μm SiO2) and 110 μm Al2O3 groups at 3.5 bar pressure. TZD ranged between 0 and 1.43 μm, with the highest values for Rocatec Plus and 110 μm Al2O3 groups at 3.5 bar pressure. © 2013 Elsevier Ltd.

Influence of air-particle deposition protocols on the surface topography and adhesion of resin cement to zirconia

Objectives: This study evaluated the influence of air-particle abrasion protocols on the surface roughness (SR) of zirconia and the shear bond strength (SBS) of dual-polymerized resin cement to this ceramic. Materials and methods. Sintered zirconia blocks (n = 115) (Lava, 3M ESPE) were embedded in acrylic resin and polished. The specimens were divided according to the 'particle type' (Al: 110 mu m Al2O3; Si: 110 mu m SiO2) and 'pressure' factors (2.5 or 3.5 bar) (n = 3 per group): (a) Control (no air-abrasion); (b) Al2.5; (c) Si2.5; (d) Al3.5; (e) Si3.5. SR (Ra) was measured 3-times from each specimen after 20 s of air-abrasion (distance: 10 mm) using a digital optical profilometer. Surface topography was evaluated under SEM analyses. For the SBS test, 'particle type', 'pressure' and 'thermocycling' (TC) factors were considered (n = 10; n = 10 per group): Control (no air-abrasion); Al2.5; Si2.5; Al3.5; Si3.5; Control(TC); Al2.5(TC); Si2.5(TC); Al3.5(TC); Si3.5(TC). After silane application, resin cement (Panavia F2.0) was bonded and polymerized. Specimens were thermocycled (6.000 cycles, 5-55 degrees C) and subjected to SBS (1 mm/min). Data were analyzed using ANOVA, Tukey's and Dunnett tests (5%). Results. 'Particle' (p = 0.0001) and 'pressure' (p = 0.0001) factors significantly affected the SR. All protocols significantly increased the SR (Al2.5: 0.45 +/- 0.02; Si2.5: 0.39 +/- 0.01; Al3.5: 0.80 +/- 0.01; Si3.5: 0.64 +/- 0.01 mu m) compared to the control group (0.16 +/- 0.01 mu m). For SBS, only 'particle' factor significantly affected the results (p = 0.015). The SiO2 groups presented significantly higher SBS results than Al2O3 (Al2.5: 4.78 +/- 1.86; Si2.5: 7.17 +/- 2.62; Al3.5: 4.97 +/- 3.74; Si3.5: 9.14 +/- 4.09 MPa) and the control group (3.67 +/- 3.0 MPa). All TC specimens presented spontaneous debondings. SEM analysis showed that Al2O3 created damage in zirconia in the form of grooves, different from those observed with SiO2 groups. Conclusions. Air-abrasion with 110 mu m Al2O3 resulted in higher roughness, but air-abrasion protocols with SiO2 promoted better adhesion.

Can Application of Universal Primers Alone Be a Substitute for Airborne-Particle Abrasion to Improve Adhesion of Resin Cement to Zirconia?

Purpose: To determine whether universal primers alone can deliver similar levels of adhesion of resin cement to zirconia ceramic when compared to their application in conjunction with airborne-particle abrasion.Materials and Methods: Sintered zirconia blocks (N = 160) (Lava, 3M ESPE), (5.25 x 5.25 x 3 mm(3)) were embedded in acrylic resin, polished, and randomly distributed into 16 groups (n = 10 per group), according to the factors "universal primer" (8 levels) and "air-particle abrasion" (2 levels): 1. ctr: control, without application of a universal primer; 2. AP: Alloy Primer; 3. MP: Monobond Plus; 4. MZP: Metal Zirconia Primer; 5. MZ: MZ Primer; 6. Sg: Signum Zirconia Bond; 7. SbU: Singlebond Universal; 8. ZP: Z Prime Plus. The universal primers were also used after air abrasion (A) of zirconia to form the following 8 groups: Ctr-A, AP-A, MP-A, MZP-A, MZ-A, Sg-A, SbU-A, and ZP-A. After ultrasonic cleaning, air abrasion was performed using Al2O3 particles (110 mu m, 2.5 bar, 20 s at 10 mm) in a chairside air-abrasion device. After ultrasonic cleaning again, universal primers were applied according to each manufacturer's recommendation. The resin cement (RelyX ARC, 3M ESPE) was built up incrementally and photo-polymerized on the zirconia surface using a silicone mold (empty set = 3.5, height = 3 mm). All specimens were stored in distilled water (60 days at 37 degrees C) and then subjected to shear bond strength testing (SBS) in a universal testing machine (1 mm/min). On a separate set of zirconia specimens, contact angle measurements were made using the sessile drop technique with a goniometer after the application of universal primers on control and air-abraded zirconia surfaces. Data (MPa) were analyzed using one-way ANOVA, Tukey's test, and Student's t-test (alpha = 0.05).Results: When universal primers were used alone, SbU presented significantly higher mean SBS (19.5 +/- 5.8) that did the other primers (0 to 9.9 +/- 6.6) (p = 0.001). When air abraded, the groups AP-A (14.1 +/- 6.1), MP-A (15.9 +/- 5.4), ZP-A (16.9 +/- 7.3), SG-A (19.1 +/- 2.1), SbU-A (12 +/- 1.5) showed significant differences (p = 0.03). Adhesive performance of all universal primers was enhanced after air abrasion, with the exception of the SbU and MZ primers. After air abrasion, contact angle measurements were lower for the each primer (without air abrasion: 28.9 to 83.9; with air abrasion: 27.1 to 63.0), except for MZP.Conclusion: Air abrasion with 110 mu m Al2O3 followed by universal primer application increased the bond strength of tested resin cement to zirconia, with the exception of SbU and MZ.

Effect of metal conditioners on the adhesive bonding of resin cements to cast titanium

Objective: To assess the effect of metal conditioners on the bond strength between resin cements and cast titanium. Method and Materials: Commercially pure titanium (99.56%) was cast using an arc casting machine. Surfaces were finished with 400-grit silicon carbide paper followed by air abrasion with 50-mu m aluminum oxide. A piece of double-coated tape with a 4-mm circular hole was then positioned on the metal surface to control the area of the bond. The prepared surfaces were then divided into 4 groups (n=10): G1, unprimed Panavia F; G2, Alloy Primer-Panavia F; G3, unprimed Bistite DC; G4, Metaltite-Bistite DC. Forty minutes after insertion of the resin cements, the specimens were detached from the mold and stored in water at 37 C for 24 hours. Shear bond strength was performed in a testing machine (MTS 810) at a crosshead speed of 0.5 mm/min. Data were analyzed using ANOVA and Tukey's test with a .05 significance level. The fractured surfaces were observed through an optical microscope at 10x magnification. Results: the G1 group demonstrated significantly higher shear bond strength (17.95 MPa) than the other groups. G3 (13.79 MPa) and G4 (12.98 MPa) showed similar mean values to each other and were statistically superior to G2 (9.31 MPa). Debonded surfaces generally presented adhesive failure between metal surfaces and resin cements. Conclusion: While the Metaltite conditioner did not influence the bond strength of the Bistite DC cement, the Alloy Primer conditioner significantly decreased the mean bond strength of the Panavia F cement.

Effect of conditioning methods on the microtensile bond strength of phosphate monomer-based cement on zirconia ceramic in dry and aged conditions

The objective of this study was to evaluate the durability of bond strength between a resin cement and aluminous ceramic submitted to various surface conditioning methods. Twenty-four blocks (5 X 5 X 4 mm 3) of a glass-in filtrated zirconia-alumina ceramic (inCeram Zirconia Classic) were randomly divided into three surface treatment groups: ST1-Air-abrasion with 110-mu m Al2O3 particles + silanization; ST2-Laboratory tribochemical silica coating method (110-mu m Al2O3, ilO-PM Silica) (Rocatec) + silanization; ST3-Chairside tribochemical silica coating method (30-mu m SiOx) (CoJet) + silanization. Each treated ceramic block was placed in its silicone mold with the treated surface exposed. The resin cement (Panavia F) was prepared and injected into the mold over the treated surface. Specimens were sectioned to achieve nontrimmed bar specimens (14 sp/block) that were randomly divided into two conditions: (a) Dry-microtensile test after sectioning; (b) Thermocycling (TC)-(6,000X, 5-55 degrees C) and water storage (150 days). Thus, six experimental groups were obtained (11 = 50): Gr1-ST1 + dry; Gr2-ST1 + TC. Gr3-ST2 + dry; Gr4-ST2 + TC; Gr5-ST3 + dry; Gr6ST3 + TC. After microtensile testing, the failure types were noted. ST2 (25.1 +/- 11) and ST3 (24.1 +/- 7.4) presented statistically higher bond strength (MPa) than that of STI (17.5 +/- 8) regardless of aging conditions (p < 0.0001). While Gr2 revealed the lowest results (13.3 +/- 6.4), the other groups (21.7 +/- 7.4-25. 9 +/- 9.1) showed statistically no significant differences (two-way ANOVA and Tukey's test, a 0.05). The majority of the failures were mixed (82%) followed by adhesive failures (18%). Gr2 presented significantly higher incidence of ADHESIVE failures (54%) than those of other groups (p = 0.0001). Both laboratory and chairside silica coating plus silanization showed durable bond strength. After aging, airabrasion with 110-mu m Al2O3 + silanization showed the largest decrease indicating that aging is fundamental for bond strength testing for acid-resistant Arconia ceramics in order to estimate their long-term performance in the mouth. (c) 2007 Wiley Periodicals, Inc.

Sucção de alta potência no controle das partículas emitidas pelo sistema de abrasão a ar em consultório odontológico

O objetivo do estudo foi avaliar a deposição de óxido de alumínio no campo operatório do cirurgião-dentista durante a utilização do sistema de abrasão a ar em consultório odontológico, bem como a efetividade da sucção de alta potência na captação desse pó. Por meio de um dispositivo para a coleta das partículas nos locais correspondentes às posições e distâncias de trabalho do profissional, dentes artificiais foram abrasionados. O sistema de sucção empregado para aspiração das partículas foi o de alta potência com sugador de saliva convencional e sugador com abertura ampliada por funil. A mensuração das partículas foi determinada pela quantidade em massa de óxido de alumínio depositada em placas de Petri. Os resultados obtidos por meio de estatística descritiva gráfica revelaram que a maior quantidade de pó se encontrava a 20 cm do operador e na posição de trabalho 9h, quando foi utilizado o sugador de saliva convencional. Uma vez comprovado que a sucção não é totalmente eficiente na aspiração do pó de óxido de alumínio, reforça-se a importância da proteção individual apropriada para o emprego seguro do sistema de abrasão a ar para pacientes e, principalmente, para os profissionais que trabalham com este tipo de tecnologia.

Repair bond strength of a resin composite to alumina-reinforced feldspathic ceramic

This study compared the microtensile bond strength of a repair resin to an alumina-reinforced feldspathic ceramic (Vitadur-alpha, Vita) after 3 surface conditioning methods: Group 1, etching with 9.6% hydrofluoric acid for 1 minute plus rinsing and drying, followed by application of silane for 5 minutes; group 2, airborne particle abrasion with 110-mm aluminum oxide using a chairside air-abrasion device followed by silane application for 5 minutes; group 3, chairside tribochemical silica coating with 30-mu m SiOx followed by silane application for 5 minutes (N = 30). Group 1 presented the highest mean bond strength (19.7 +/- 3.8 MPa), which was significantly higher than those of groups 2 (10 +/- 2.6 MPa) and 3 (10.4 +/- 4 MPa) (P <.01). Scanning electron microscope analysis of the failure modes demonstrated predominantly mixed types of failures, with adhesive and/or cohesive failures in all experimental groups.

Surface conditioning of a composite used for inlay/onlay restorations: Effect on mu TBS to resin cement

Purpose: To assess the effect of the composite surface conditioning on the microtensile bond strength of a resin cement to a composite used for inlay/onlay restorations.Materials and Methods: Forty-two blocks (6 x 6 x 4 mm) of a microfilled composite (Vita VMLC) were produced and divided into 3 groups (N = 14) by composite surface conditioning methods: Gr1 - etching with 37% phosphoric acid, washing, drying, silanization; Gr2 - air abrasion with 50-Im Al203 particles, silanization; Gr3 - chairside tribochemiCal silica coating (CoJet System), silanization. Single-Bond (one-step adhesive) was applied on the conditioned surfaces and the two resin blocks treated with the same method were cemented using RelyX ARC (dual-curing resin cement). The specimens were stored for 7 days in water at 37 degrees C and then sectioned to produce nontrimmed beam samples, which were submitted to microtensile bond strength testing (mu TBS). For statistical analysis (one-way ANOVA and Tukey's test, = 0.05), the means of the beam samples from each luted specimen were calculated (n = 7).Results: mu TBS values (MPa) of Gr2 (62.0 +/- 3.9a) and Gr3 (60.5 +/- 7.9a) were statistically similar to each other and higher than Gr1 (38.2 +/- 8.9b). The analysis of the fractured surfaces revealed that all failures occurred at the adhesive zone.Conclusion: Conditioning methods with 50-Im Al203 or tribochemical silica coating allowed bonding between resin and composite that was statistically similar and stronger than conditioning with acid etching.

The effectiveness of alumina powder on carious dentin removal

This study determined the size of aluminum oxide particles used in an air abrasion system that is able to remove carious dentin tissue with maximum preservation of sound structure. Thirty extracted and carious-free third molars were used in this study. The dentin sample was obtained by sectioning the middle of the crown longitudinal to the long axis of the tooth in a mesio-distal direction. One half of the crown corresponded to the sound dentin group (SD), while the other half was used to develop artificial caries, constituting the, carious dentin group (CD). The specimens were air abraded for 15 seconds. The SD and CD groups were each randomly divided into three subgroups (N=10) according to the particle diameter employed (27, 50 and 125 pm). The prepared cavity was perpendicularly cut in half, and the profiles of all hemi-fragments were observed using SEM microscopy. The cavity measurements were made using a modified cephalometric analysis. The 27, 50 and 125 pun aluminum oxide particles did not present selectivity in the removal of carious dentin. However, when using the air abrasive technique for carious dentin treatment, the use of 27 and 50 pun aluminum oxide particles is recommended, due to their capacity to remove less sound tissue than the 125 pun particles.