New in vitro SPF Evaluation Method for Hydrophilic Sunscreen Samples

: A new method was developed for the in vitro sun protection factor (SPF) evaluation of sunscreen samples. A new type of substrate, a hydroxyalkyl cellulose-coated plate, was also prepared specifically for hydrophilic samples. This new substrate was required because hydrophilic samples would be unlikely to wet the surface of the standard cosmetic PMMA UV evaluation plate. A super-hydrophilic quartz plate was prepared by corona-discharge treatment before an aqueous solution of hydroxyalkyl cellulose was spread on it. A flat and uniform hydroxyalkyl cellulose film was subsequently formed through the evaporation of water. Special care was taken to inhibit the generation of spatial non-uniformity. Six hydrophilic sunscreen samples with in vivo SPF values of 56, 55, 52, 25, 15, and 4, were then applied to the prepared hydroxyalkyl cellulose-coated plate, as well as a super-hydrophilic quartz plate and a flat hydrophobic PMMA plate. The thicknesses of the applied layers were determined using a wheel-shaped wet film thickness gauge immediately after the application, and UV transmission was measured using an SPF analyzer. The value of in vitro SPF was calculated from the UV absorbance and the thickness of the layer. For two out of the six samples, PMMA plate could not be available, as the samples were unable to wet the PMMA surface. Relatively small differences were observed between the in vitro SPF values when the super-hydrophilic and hydroxyalkyl cellulose-coated plates were used. Samples exhibiting higher in vivo SPF were also associated with higher in vitro SPF values, although a linear relationship was not observed. In contrast to the super-hydrophilic plate whose half-life of the super-hydrophilicity is only approximately five days, the hydrophilicity of the hydroxyalkyl cellulose-coated plate scarcely varied during six months of storage. Finally, a simplified evaluation method was also proposed. The validity of the method was verified through a ring test where three researchers employed this method in different laboratories at three independent organizations.


Introduction
To evaluate the sun protection factor SPF of sunscreens, traditional in vivo methods require UV light irradiation to sunscreen layers applied on the back of human study participants 1 . The development of an alternative in vitro evaluation method, where the recommended amount of sunscreen is applied onto a substrate, rather than a human subject, has therefore been expected 2 9 . In vitro UV protection efficacies of sunscreens are evaluated em-when in vivo SPF measurement is performed, the head of the optical fiber used for the UV light irradiation touches the skin surface, resulting in a smooth surface that does not represent the ridges and furrows of the skin Fig. 1 . The same effect is observed even where a flat quartz plate is used as the substrate. Here, a process referred to as viscous fingering 17 20 results in the formation of a stripe pattern in the applied layer of sunscreen, which leads to variation in the values of the in vitro UV protection efficiencies 21 23 . It is therefore important that any spatial inhomogeneity effects in the applied layer of sunscreen samples be eliminated to ensure accurate in vitro evaluation of the UV protection efficiencies.
Another problem associated with the roughened cosmetic standard PMMA UV evaluation plate is that hydrophilic samples, such as O/W emulsion sunscreens, cannot wet the hydrophobic PMMA surface. The hydrophilic sunscreens have, however, become popular due to their smooth and light texture 24 26 . In addition, it was determined that the hydrophilic O/W emulsion sunscreens can also not wet the surface of quartz plates 27,28 . Here, dewetting refers to the rupture of a thin liquid film on a solid substrate thereby generating holes. This occurs when the spreading coefficient, S, changes from positive to negative values 29 33 . The spreading coefficient can be calculated by S Γ SG Γ SL Γ LG , where Γ SG , Γ SL , and Γ LG are the interfacial tensions of the solid-gas, solid-liquid, and liquid-gas interfaces, respec-tively. A super-hydrophilic plate was successfully prepared by corona-discharge or plasma treatment of a quartz plate, and the hydrophilic nature of this plate ensured that it could be used for the in vitro evaluation of UV protection efficacies of hydrophilic samples 27,28 . Employing the sessile drop test to measure the contact angle of a drop of water on the plate, which in turn is used as an index for the hydrophilicity, indicated that contact angles became zero by these treatments. The change was, however, not permanent, and results indicated it returned to its initial value at an exponential rate with the half-life being only approximately five days. This prompted us to develop a new type of substrate that would be suitable to any type of sunscreen.
In the present study, research focused on the establishment of a new in vitro SPF evaluation method for hydrophilic sunscreen samples. A new type of hydrophilic substrate, a hydroxyalkyl cellulose-coated plate, was successfully prepared. For in vitro SPF evaluation, six hydrophilic sunscreen samples were applied to the hydroxyalkyl cellulose-coated plate as well as to the super-hydrophilic quartz plate and hydrophobic PMMA plate. Experiments were repeated independently, to verify the accuracy and reproducibility of the in vitro SPF evaluation method. In addition, a simplified evaluation method designed to reduce the number of experiments, and time and labor requirements, was also proposed. Finally, to verify the validity of this method, three researchers carried out ring tests at laboratories in three different organizations.

Materials
Six sunscreen samples whose in vivo SPF have already been evaluated were supplied by NIKKOL Group Nikko Chemicals Co., Ltd., 1-6 Table 1 , for the in vitro SPF measurements required for this study. Ingredients contained in each sample were values given in parentheses are their wt : 1: Water 51.562 , ethylhexyl methoxycinnamate 7.5 , butylene glycol 5.4925 , diethylamino hydroxybenzoyl

Preparation of hydroxyalkyl cellulose-coated HCC plate
Super-hydrophilic plates were prepared by using coronadischarge treatment on a flat square quartz plate, as described in our previous research 27,28 . Following the preparation, an aqueous solution of hydroxyalkyl cellulose was spread on the super-hydrophilic plate. During evaporation of water from this aqueous solution, various drying patterns were observed Fig. 2 . These can be explained by Marangoni convection, which is usually generated during the evaporation of volatile solvents from solutions spread on solid substrates. This causes the formation of surface spatial non-uniformity such as coffee rings 34 37 and Marangoni contraction 38 40 . To avoid such surface defects and ensure the formation of a flat and uniform film of hydroxyalkyl cellulose, preparation was performed according to our patented process 41 . The flatness of the HCC plate was determined using the surface profile measuring system Dektak XT .

Application of sample sunscreens and evaluation of their in vitro SPF values
The in vitro SPF evaluation method proposed in this study includes the following steps: i application of the sunscreen sample, ii determination of the thickness of the sunscreen layer, iii determination of the UV transmission of the sunscreen layer, and finally iv calculation of the UV absorbance of a 10 µm thickness sunscreen layer and in vitro SPF Fig. 3 . Six sunscreen samples, 1-6, were applied to super-hydrophilic SHP , hydroxyalkyl cellulosecoated HCC , and PMMA PMMA plates, respectively. To inhibit the spontaneous formation of patterns related to viscous fingering during the application of sunscreen samples, the four-sided applicator used previously 21 23, 27 was employed again. The gap of the applicator was 10 µm, and the application velocity was 5 mm s 1 . Since some amount of sample adhered to the bottom face of the applicator during the application, the thickness of the applied layer was not the same as the applicator gap. A wheelshaped wet film thickness gauge was therefore used to determine the thickness of the sample layer immediately following the application. Where the thickness was determined to be lower than 5 µm, the application was repeated. The UV absorbance of the applied layer was measured using an SPF analyzer SPF-290AS performed from 290 nm to 400 nm in 1 nm increment. These results were then converted to UV absorbance values associated with 10 µm thick sample layers. According to Ref. 11, in vitro SPF values should be calculated according to equation 1 : where E λ , I λ , and A λ are the erythema action spectrum, the intensity associated with sun light, and the UV absorbance at each wavelength, respectively.  method, a ring test was conducted between three researchers in different laboratories, and at three independent organizations: Keio University, NIKKOL Group Cosmos Technical Center Co., Ltd., and Kuroda Consulting Inc., respectively.

Property of hydroxyalkyl cellulose coated plate
Since the gap of the applicator used to apply the samples was 10 µm, the degree of surface roughness of the substrate should be smaller than the µm level. With the exception of a couple of small cracks and bumps, the surface roughness of the HCC plate prepared in this study was in fact found to be below this level Fig. 4 . It was therefore considered an appropriate substrate for in vitro SPF evaluation. Employing the sessile drop method, the contact angle of a drop of water on the HCC plate was measured to evaluate the hydrophilicity of the surface. This was found to be 44 2 . The water of the sessile drop test was then absorbed gradually into the hydroxyalkyl cellulose film through osmosis. Ultimately, these contact angle results remained essentially constant during six months of storage at both room temperature 25 , and when refrigerated Fig.  5 . This constituted a dramatic improvement in the stability of the hydrophilicity of the substrate, when compared to the SHP plate, whose half-life typically only approximately

Determination of in vitro SPF and its comparison with
in vivo SPF To carry out the in vitro SPF evaluation, the UV absor-bance of each applied sunscreen layer was measured for all sunscreen -substrate combinations. Measurements were performed at discrete wavelengths integer values , and results were corrected correspond to a 10 µm thick layer. Results showed that the surface of the substrate had an  impact on the UV absorbance measured for sample sunscreen layers Fig. 6 . This was especially evident in the case of sample 1 and 3, when applied to the PMMA plate.
Here these samples exhibited much lower UV absorbance and in vitro SPF than when applied to the other substrates. It is postulated that this lower UV absorbance is because samples 1 and 3 did not wet the surface of the PMMA plate Fig. 7 . While PMMA was thus shown not to be an appropriate material to ensure the formation of the required flat and thin sunscreen layer for samples 1 and 3, no dewetting was observed for any of the samples when applied to the SHP and HCC plates. Here, sunscreen samples that exhibited relatively high in vivo SPF also tended to show relatively high in vitro SPF values Fig. 6 . Both SHP and HCC plates were thus found to be appropriate substrates for the in vitro SPF evaluation of these hydrophilic sunscreens.

Accuracy and reproducibility of the in vitro SPF evaluation method
Measurements were repeated for four times during the evaluation of the accuracy and reproducibility of the method Fig. 8 . The average, standard deviation, and coefficient of variation values were calculated Table 2 . As the coefficient of variation was less than 0.20 in the majority of the experiments 13 out of 18 , and more than 0.33 in only two, it may be concluded that the accuracy and reproducibility of the method are fit for purpose.
In the case of the measurements of samples 1 and 3 using the PMMA plate, in vitro SPF was very low in all four experiments. As mentioned previously, dewetting of the layer was observed in all experiments in these cases, thus confirming that the PMMA plate was not appropriate for the in vitro SPF evaluation of samples 1 and 3. In the case of using the SHP and HCC plates, sunscreen samples that exhibited higher in vivo SPF tended to show higher in vitro SPF as well. For all sunscreen samples tested, no significant differences were observed between the in vitro SPF values determined using either the SHP or HCC plates. Both the SHP plate developed in our previous study 27,28 and the HCC plate developed in this study would therefore be suitable for the in vitro SPF evaluation of hydrophilic sunscreen samples. In the case of the SHP plate, however, it is recommended that either corona-discharge or plasma treatment should be performed just before the measurements, since the contact angle of sessile drop of water, an index for the hydrophilicity, exponentially returned to the initial value following the treatments, indicating a half-life of only approximately five days. In contrast to this, hydrophilicity of the HCC plate scarcely varied during six months of storage. Therefore, the HCC plate can be regarded as a feasible apparatus for commercial in vitro SPF evaluation. However, a linear relationship between in vivo SPF and in vitro SPF values was not ob- served. Samples exhibiting in vivo SPF ranged in 10 to 30 were associated with lower in vitro SPF values, while those exhibiting in vivo SPF values higher than 50 were associated with higher in vitro SPF values.

Ring test for the establishment of new in vitro SPF evaluation method
For the standard test methods, it is very important that the method can achieve accurate results with suitable reproducibility when employed in different laboratories. Not only should instructions be clear, but the method should ideally also not be too time and labor intensive. To establish this, ring tests are consequently typically carried out under similar conditions, and at the different organizations, to evaluate a method s performance. Here, three researchers at different laboratories, and in different organizations, participated in a ring test to verify the validity of the simplified method, where the number of experiments was reduced from twelve to six. Experiments discussed in section 3.1 and 3.2 were performed by Researcher 1, while Researcher 2 and 3 carried out the additional identical experiments at the different laboratories.
The first step of the proposed simplified new in vitro SPF evaluation method, which was aimed at reducing time and labor requirements, is choosing the most suitable substrate. Here, all three researchers selected the HCC plate as the most appropriate substrate for the in vitro SPF measurements for samples 3 and 5, while two researchers also selected it for samples 4 and 6 Table 3 . Two researchers selected the SPH plate for sample 1. For these five samples, researchers therefore displayed a similar approach to substrate selection. However, in the case of sample 2, the individual researchers each selected different substrate. In this instance, the selection of the substrate would not have been critical, as in vitro SPF values obtained by Researcher 1, during four measurements on each of the three substrates, did not show significant differences Fig. 4 . Important to note is that none of the three researchers selected the PMMA plate for samples 1 and 3. Due to the dewetting associated with these samples, this plate would have been an unsuitable choice.
During the evaluation of the average and standard deviation associated with the method, the second and third steps of the method were the repeated an additional three times experiments under identical conditions. The most appropriate substrate was used here, while the highest in vitro SPF values were excluded from calculations Fig. 9 . Except for sample 1 as evaluated by Researcher 2 and sample 4 as evaluated by Researcher 3, the coefficients of variation were relatively low. Similar trends were observed between the in vivo SPF and in vitro SPF results reported by all three researchers. Here, the samples that exhibited in vivo SPF ranging from 10 to 30 were also associated with lower in vitro SPF values, while samples that exhibit-

Conclusion
In the present study, a new method was developed for the in vitro SPF evaluation of sunscreens. The new method consisted of four steps, i application of sunscreen sample, ii determination of the thickness of the sunscreen layer, iii determination of UV transmission of the sunscreen layer, and finally iv calculation of UV absorbance associated with a 10 µm thick sunscreen layer and in vitro SPF. Six hydrophilic sunscreen samples were evaluated, along with the three types of substrates: super-hydrophilic SHP , hydroxyalkyl cellulose-coated HCC , and PMMA PMMA plate. The SHP plate was prepared by coronadischarge treatment on a flat square quartz plate, according to our previous research 27,28 , while the HCC plate preparation method was developed in this study. In the preparation of the latter, an aqueous solution of hydroxyalkyl cellulose was spread on an SHP plate before the water was evaporated. Here, special care had to be exercised to prevent the generation of Marangoni convection, which could lead to a spatially inhomogeneous hydroxyalkyl cellulose film coating. In the case of the HCC plate, contact Fig. 9 Comparison of the results of the ring test of the simplified new in vitro SPF evaluation method proposed in this study by three researchers. In addition to Researcher 1 who carried out the experiments for obtaining data shown in Fig. 6 and 8, Researcher 2 and 3 participated in the ring test at laboratories in different organizations. Blue, green, and red colors of the plot indicate that the researcher selected SHP, HCC, and PMMA for the three more times measurements, respectively. Table 4 Average AVG , standard deviation STD , and coefficient of variation CV of in vitro SPF of six sunscreen samples measured by three researchers in different laboratories.
Since the data of AVG and STD are the value rounding off one decimal place, CV does not exactly consistent with the value of quotient of STD divided by AVG shown in this table.
angle results obtained from the sessile drop test, indicated that hydrophilicity scarcely varied during six months of storage. From a commercial point of view, this would indicate that this plate is incredibly stable. The PMMA plate was, however, found not to be appropriate for two of the six sunscreen samples evaluated. This was attributed to dewetting of the applied sunscreen layer. In contrast, both SHP and HCC plates were suitable for the in vitro SPF evaluation of all six sunscreen samples. Here, sunscreen samples exhibiting higher in vivo SPF values tended to show higher in vitro SPF as well. Large differences were not observed between the in vitro SPF values determined when the SHP and HCC plates were employed. Relatively high levels of accuracy and reproducibility for the experiments were confirmed with the coefficients of variation for four measurements being less than 0.20 in 13 out of 18 experiments, and more than 0.33 in only two experiments.
A simplified evaluation method that reduced the number of experiments from twelve to six was also proposed to reduce the time and labor for the researchers. To verify the validity of the simplified method, three researchers at laboratories in three different organizations carried out a ring test. Similar trends were observed between the three sets of results obtained. No linear relationship did, however, exist between the in vivo SPF and in vitro SPF results. It is postulated that the lack of physiological evidence that in vivo SPF increases linearly with the increase of the reciprocal of UV transmittance may explain the deviation from a linear relationship between in vivo SPF and in vitro SPF values observed here.