Silver Maru-fun, which is a powder made of silver round particles, was used for the comparison. One company produces 15 kinds of Maru-fun, while the other manufactures 30 kinds. The smallest powder is named No. 1, and the number increases as the diameter of the powder particles increases. According to an interview with Maki-e craftsmen, they usually use powders from No. 1 to No. 15, with No. 1 being the most used. Hence, No. 1 powder produced by both Asano Co., Ltd. (sample A) and Yoshii Co., Ltd. (sample B) was studied and compared with Wakou silver powder by Yoshii Co., Ltd. (sample C). Wakou silver powders are said to be more resistant to sulfuration than other silver powder.

Acrylic boards of the size 100 × 50 × 5 mm³ were used as substrates to create Maki-e. Silver powder was sprinkled on 80 × 50 mm² of the board surface. The powder was sprinkled densely to the maximum by the Tume-maki technique. E-Urushi, which is mixture of Suke-Urushi (reined Urushi with high transparency) and red oxide, was used to draw Maki-e pattern and bond Maki-e powder to the board. Sample preparation is schematized in Figure 1.

Sample A and B were both white in color, while Wakou silver powder was dark gray (Figure 2). The appearance of the samples is shown in Figure 3.

All sample surfaces were immersed in a dilute solution of hydrogen sulfide to accelerate silver sulfuration [15] . The Maki-e were polished by the traditional method for polishing, which is to polish the surface by hand with a polishing powder called Roiro-ko and rape-seed oil. To verify the effectiveness of the polishing procedure, residual gloss after the sulfuration test and polishing was calculated with respect to the gloss of the untreated surface with the following equation:

[16] (1)

Figure 4 compares the gloss of the samples, and Figure 5 compares their color coordinates. Sample B had the highest gloss, followed by A and C in decreasing order; A had the highest value of C*, followed by B and C in decreasing order; and B had the highest L* value, followed by C and A in decreasing order.

Sample A had moderate gloss and the highest C* value, however L* value was the lowest. Sample B had the highest gloss and L* value, and moderate C* value. Sample C had the highest gloss and the lowest C*.

The lower C* of sample C is due to the dark gray color of the powder used for this Maki-e. Besides, the difference in A and B about appearance was shown numerically as it has been pointed out which was mentioned in introduction. The surfaces of the samples were observed to examine how the silver particle shape affects the final appearance of the Maki-e.

The surfaces of the samples were observed to clarify whether gloss and color depend on the powder properties or on the decorating process of Maki-e. Photomicrographs of sample surfaces are shown in Figure 6. Both average and distribution of the particle size were smaller for sample A compared to those of samples B and C. On the contrary, the powder used for samples B and C had a wider particle size distribution.

Surface texture, given by the amount of powder on the surface, was measured through the calculation of the powder occupancy rate per substrate unit area by binary coded processing (IGOR Pro 6.34A) of Figure 6. Occupancy rate of powder per unit area was 20.5% for sample A, 30.5% for sample B, and 23.0% for sample C.

This indicated that the occupancy rate decreased with decreasing particle size if it was relatively less uneven. Therefore, it is demonstrated that particle size and particle size distribution affect the appearance of the Maki-e. However, it was unclear whether the original elemental composition of the powder or the processing method of the powder that affected the particle shapes.

Figure 7 provides the EDS spectra for the samples and Figure 8 shows relative element concentrations.

Samples A and B were made of ca. 95% Ag, while sample C was made of ca. 70% Ag and of ca. 18% Pd. Oxygen and carbon were the main other elements present in all the three powders, probably as a result of deposition on powder surface. Trace amounts of Pt were detected only in samples A and B; however, this may be inadvertently included during the metal refining process. As the powder components were almost the same, it seems that the gloss and color are not affected by powder chemistry with the exception of sample C whose powder color is attributable to the presence of Pd.

Hence, it can be concluded that the gloss and color of the samples depend on the particle shape and occupan- cy rate. Shimode et al. reported that occupancy rate decreases with decreasing particle size in a study of Egg- shell Maki-e, which is a type of Maki-e powder that is made of finely crushed egg shells. They also reported that the powder occupancy rate affects the appearance of Maki-e [17] . Low particle occupancy rate implies that the occupancy rate of Urushi becomes larger. This is the reason why C* value of sample A, with lower particle occupancy rate, was higher than that of the other samples which are less affected by the deep red color of E-Urushi. Furthermore, sample B has a higher powder occupancy rate than sample A, which may have resulted in higher gloss.

Therefore, the differences in the gloss and color of the appearance of samples after Maki-e decoration are due to the effect of occupancy rate and the differences in particle shape. Despite the occupancy rate of sample C being larger than that of A, C had the lowest gloss among all the samples. This may be attributed to the original gloss of the C powder and not its occupancy rate.

The photomicrographs of the surface of the samples after sulfuration are shown in Figure 9. The residual gloss after sulfuration and polishing is shown in Table 1. The gloss of samples A and B after sulfuration decreased by ca. 50% with respect to the gloss before, while the gloss of sample C decreased only by 10%. A sulfurized film was formed on the surface of samples A and B (Figure 9), but not on sample C. It is likely that the Pd in sample C played a pivotal role in the prevention of silver sulfuration. For example, a gold-silver-palladium alloy, of the composition 12% Au, 50% Ag, and 20% Pd, is used as metal crown for dental treatment; in this case, Pd is included to prevent discoloration after sulfuration of Ag [18] .

The appearance of the samples before and after sulfuration, and after polishing is shown in Figure 10. By polishing the sample after sulfuration, the gloss of all the samples, irrespective of the shape and chemistry of the silver powder particles, increased by almost 100%, and the particles appeared the same as they did before sulfuration. This implies that polishing removes the film that was formed during sulfuration.

The craftsman who polished the samples was interviewed to obtain information on the ease of the polishing procedure for the different samples. He said that sample B required more time for polishing, while sample C was easily returned to its original condition. It is hypothesized that the gloss of sample B decreased the most and polishing of this sample took longer because of its higher occupancy rate. The powder used in sample C had stronger resistance to sulfuration than the other two powders, remedying to a defect of silver powder which has been considered to be a problem in the past. However, the expression of each powder is different and it is important to know when to use what type of powder.