*2.3. Characterization*

The phase compositions of diverse coatings were examined through an X-ray diffractometer (XRD) (D/max-rB, RICOH, Tokyo, Japan) with a Cu Ka source, and the accelerating voltage and applied current were 40 kV and 30 mA, respectively. In addition, the radiation emitted by the sample surface was detected by a Fourier transform infrared (FT-IR) spectrometer (JASCO FT/IR-6100, JASCO, Toykp, Japan). Meanwhile, scanning electron microscopy (SEM, JSM-6700F, JEOL, Japan) was employed to observe the microstructure and morphology of the MAO coatings, whereas the element compositions on the coating were analyzed by an energy-dispersive spectrometer (EDS, Oxford, UK) combined with SEM. Further, the coating thicknesses generated under different conditions were measured using an eddy current coating thickness measurement gauge (CTG-10, Time Company, Beijing, China). To be specific, the thicknesses at 10 different sites on the coating surface were measured to calculate and record the average. Additionally, the coating roughness was tested using a roughness tester (TR-3200, Time Company, Beijing, China, vertical resolution of 0.01 μm). The micro-hardness of the coating was evaluated using the HVS-100 micro-hardness tester (TMVS-1, TIMES Group, Beijing, China) with a load of 100 g for 10 s. In addition, the tribological behaviors of the coatings were evaluated using the ball-on-disk tester (UMT-Tribolab, BRUKER, Bremen, Germany) under dry sliding conditions. Typically, balls of GCr15 with a diameter of 10 mm and a hardness of HRC 60 were used as the counterface materials. The normal load was 5 N, and the linear sliding speed was 0.01 m/s. All tests

were run under the laboratory conditions (temperature of 25 ◦C and relative humidity of 50%) for 30 min each. The friction coefficient was recorded on a computer during each test. The wear loss was weighed using an electronic balance, and the wear rate (k) was calculated according to the following Formula (1).

$$k = \frac{\pi \bullet D \left[ \arcsin(\frac{L\_u}{2r}) r^2 - L\_u(\frac{\sqrt{4r^2 - L\_u}^2}{4}) \right]}{P \times S} \tag{1}$$

where *r* stands for the radius of the corundum ball (mm), *D* represents the diameter of the wear track (mm), *Lu* indicates the width of the wear track (mm), *S* is the sliding distance (m), and *P* is the applied normal load (N).

#### **3. Results and Discussion**

#### *3.1. Thickness and Roughness of the MAO Coating*

The tribological performance of the MAO coating was affected by its thickness and roughness; therefore, the impacts of cellulose content on the coating thickness and roughness were examined, as shown in Figure 2.

**Figure 2.** Effects of cellulose content on the thickness and roughness of the MAO coating.

It can be observed from Figure 2 that with the increase in cellulose concentration, the thickness of the MAO coating increased, whereas its roughness decreased. At the cellulose content of 0.75 g/L, the coating thickness and roughness reached 32.1 μm microns and 0.66 μm, respectively. However, further increase in the cellulose concentration showed no obvious improvement in the thickness and roughness of the MAO coating, which was mainly ascribed to the low cellulose content. Specifically, it contained the multiple-hydroxy, and a double electric layer was formed during the electrochemical process, which attracted Al3+ contiguous to the substrate surface [38], resulting in the increase in the MAO coating thickness. In addition, cellulose participated in the coating formation by filling in the microcracks and micropores, even cross-linking with Al3+ in the coating, as observed in Figure 1. Moreover, the excellent polymer plasticity contributed to reducing the quantity and size of microcracks and micropores, thus decreasing the MAO coating roughness. Nonetheless, the Al3+ escaping from the substrate was limited by the electrochemical parameters, and further increase in the cellulose contents showed no significant improvement of the coating thickness and roughness when most of the Al3+ ions were attracted by the cellulose. Additionally, the electrolyte became inhomogeneous after over 24 h of preservation, so the optimal cellulose content was determined to be 0.75 g/L.

#### *3.2. Microstructure of the MAO Coating*

The surfaces and cross-section microstructures of the MAO coatings at different cellulose contents were observed through SEM. The results are shown in Figures 3 and 4, respectively. As observed from Figure 3, the increase in cellulose content led to the decreased size of the microcracks and micropores, while it increased the quantity of the micropores.

**Figure 3.** Microsurface of the MAO coating at cellulose concentrations of (**a**)0g/L, (**b**) 0.25 g/L, (**c**) 0.50 g/L, and (**d**) 0.75 g/L.

**Figure 4.** Cross-section of the MAO coating at cellulose concentrations of (**a**)0g/L, (**b**) 0.25 g/L, (**c**) 0.50 g/L, and (**d**) 0.75 g/L.

The cross-section photograph displayed in Figure 4 proves the above findings. In addition, Figure 4 also suggests that when the cellulose concentration was 0 g/L, the coating thickness was small and the adhesion between coating and substrate was poor (seen in Figure 4a). With the increase in cellulose concentration, the coating thickness increased and the adhesion between coating and substrate was enhanced (seen in Figure 3b,c,d). The possibility of adhesive wear was reduced with the increase in the bonding force between the coating and the substrate. To further investigate the coating component, EDS was carried out. The carbon contents at different sites are presented in Figure 5 and Table 1. As displayed in Table 1, the carbon element spread all over the coating, which proved that part of the cellulose filled in the microcracks and micropores, while part of it cross-linked with the Al3+ in the coating. In addition, the cellulose content in the micropores and microcracks was higher than it was at the other sites, indicating that they were filled in by a relatively small portion of the cellulose.

**Figure 5.** Micrograph illustrating the zone of energy-dispersive spectrometer (EDS) analysis.


**Table 1.** C, O, Al element contents at different positions.

#### *3.3. Phase Structure of the Coating*

The crystalline phase compositions of the MAO coatings at the cellulose contents of 0, 0.25, 0.50, 0.75 and 1 g/<sup>L</sup> were analyzed by means of FTIR and XRD, respectively. The results are shown in Figures 6 and 7, separately. Figure 6 illustrates the infrared absorption peaks of the MAO coatings obtained under various cellulose contents. Notably, the peaks at 3406 cm<sup>−</sup><sup>1</sup> were assigned to O–H stretching vibrations, while those at 1630 cm<sup>−</sup><sup>1</sup> corresponded to C–O stretching vibrations, and those at 838 and 648 cm<sup>−</sup><sup>1</sup> were indexed to Al–O stretching vibrations. As indicated by Figure 7, the increase in the cellulose content gave rise to the enhanced characteristic peak of the cellulose and the decreased peak intensity of the alumina. Taken together, the analyzed results of the FTIR and XRD spectra proved the presence of cellulose in the MAO coating.

**Figure 6.** Fourier transform infrared (FT-IR) spectra of the MAO coatings obtained at di fferent cellulose contents.

**Figure 7.** X-ray di ffraction (XRD) spectra of the MAO coatings obtained at di fferent cellulose contents.

#### *3.4. Tribological Performances of the MAO Coatings*

Figure 8 exhibits the influences of the cellulose content on the friction coe fficient. Clearly, the friction coe fficient was significantly reduced after the aluminum alloys were treated by the MAO technology, and it slowly decreased with the further increase in the cellulose content. In addition, the friction coe fficients of most samples were maintained at fixed values when cellulose was used as the additive; however, that of the cellulose-free MAO coating was suddenly increased after 20 min. These findings revealed that the MAO surface treatment technology reduced the friction coe fficient of the aluminum alloy, and the addition of cellulose into the electrolyte was beneficial to further decrease and maintain the friction coe fficient for a long time. To examine the tribological properties of the MAO coatings obtained at di fferent cellulose contents, the micro-hardness of the MAO coatings were tested

by the micro-hardness tester, while the wear loss (the amount of material lost during the mechanical tests) and wear rate were determined through wear tests. Results are presented in Table 2.

**Figure 8.** Friction coe fficients of the MAO coatings obtained at di fferent cellulose contents.



According to Table 2, the micro-hardness of the coatings remained at about 1230 HV0.1, while the wear loss and wear rate decreased when the cellulose content was elevated from 0% to 1.0%. These results sugges<sup>t</sup> that the addition of cellulose was beneficial for improving the tribological performances of the MAO coatings.

#### *3.5. Stability of the Electrolyte*

Apart from the favorable anti-wear performance, the electrolyte stability, especially when polymer is used as the additive, is also a crustal parameter in practical industrial production. To investigate the electrolyte stability during long-term storage, the performances of the MAO coatings (such as thickness, roughness, hardness, friction coe fficient, wear loss and wear rate) obtained at di fferent electrolyte storage periods were compared, as shown in Figure 9 and Table 3. There was no obvious di fference between them, demonstrating that the electrolyte might be employed to improve the tribological performance of the MAO coating within 30 days.

**Figure 9.** Friction coe fficients of the MAO coatings under di fferent storage periods.


**Table 3.** Coating performances under di fferent storage periods.
