3.1. Particle size distribution and zeta potential
Size distribution of GO nanosheets dispersed in water (pH = 4.5), CS and CS/GO nanocomposites investigated by DLS analyses. The standard spherical particle models were used in DLS. Results reported in Fig. 1(a) show two types of peaks corresponding to 800 and 3200 nm size populations (poly dispersion index; PdI = 6.75). It is worth noting that the heights of the two peaks shown in Fig. 1(a) are related to the scattering intensities and not the amount of small and big GO particles. The low value of PDI indicates that the GO sample is constituted of only two populations, but we cannot say which is the most abundant. Our outcomes are similar to the results obtained by Liu et al. regarding the size of GO and rGO. As described earlier, the model-derived diameters are not their actual sizes because most graphene-based materials are not spherical particles. DLS results only to illustrate the size contrasts between the two materials. The size, thickness and corresponding height profile of information for GO nanosheets were observed by AFM (Fig. 1(b)). The AFM images of the GO reveal the presence of irregularly shaped sheets with uniform thickness and different lateral dimensions. Nanosheets can be observed with different sizes ranging from a few tens of nanometers to hundreds of nanometers. As shown from the height profiles recorded at different locations, the measured thickness of the GO sheets was uniform (∼1 nm), while lateral dimensions ranged from 1.6 to 6.6 µm.
The calculated average size of CS and CS/GO nanocomposite particles was about 4.5 and 6.6 µm, respectively (Fig. 2 (a-c)). The size distribution of CS/GO nanocomposite particles is more significant than that of CS polymer and GO nanosheets. It seems that GO nanosheets were isolated by CS polymer.
After mixing GO with CS solution by ultrasonication, the suspension indicated good stability for several weeks. The zeta potential values of CS and CS/GO dispersion indicate a significant effect of the negative charge of GO on CS polymer. The Zeta potential of the CS was + 42 ± 0.8 mV, and this value changed as a result of GO addition to + 40.6 ± 0.5 mV and + 36.6 ± 0.6 mV for the CS/1wt%GO and CS/2wt%GO nanocomposites respectively. Moreover, when the zeta potential index in the colloidal systems is above ± 30 mV, particles are stable in electrostatic repulsion forces [29]. Good dispersion of GO nanosheets in CS solution may be the consequence of the hydrophilic groups in GO and electrostatic interaction between the cationic CS and the GO surface's negative charge.
3.2. Characteristics Of The Nanocomposite Coatings
The GO, CS, and CS/2wt% GO coating were first evaluated by FT-IR (Fig. 3(a)). The GO spectrum indicated two adsorption peaks at 1721 cm− 1 and 1625 cm− 1, which were assigned to the C = O stretches of the carboxylic group and the distortion of the O–H bond in water, respectively. The peak appearing at 3240 cm− 1 was ascribed to the hydroxyl group on GO [38]. The FT-IR spectrum of CS shows a broad absorption band between 3600 cm− 1 and 2964 cm− 1, centered at 3260 cm− 1, due to O–H stretching vibration, N–H extension vibration, and the intermolecular H-bonds of the polysaccharide moieties [39]. A spectral band at 2870 cm− 1 is observed, corresponding to the axial stretching of C–H bonds. A peak at 1642 cm− 1 represents the acetamide group's axial stretching of C = O bonds. Another spectral band at 1540 cm− 1 is associated with the amino group's angular deformation of N–H bonds. A band at 1388 cm− 1 due to symmetrical angular deformation of CH3 and the amide III band at 1331cm− 1 are also observed. The spectral band corresponding to the polysaccharide skeleton, including vibrations of the glycoside bonds and C–O and C–O–C stretching, are also observed in the FT-IR spectrum of chitosan within 1156–655 cm− 1 [38]. In the spectra of CS/GO, the dominant peaks at 1020 cm− 1 and 1550 cm− 1 correspond to the absorbance of glucosidic bond, stretching vibration from C = O of -NHCO- and the N-H bending NH2, respectively, compared to the pure CS and GO [40].
Figure 3(b) shows the UV spectra of CS/GO nanocomposite coatings with different GO percentages. The UV spectra of CS/1wt % GO coating exhibit a specific peak at 219 nm, related to the electron transfer of π–π* from the aromatic rings of C = C bonds [41]. This absorption peak red-shifted to 223 nm by increasing the GO contents to 3wt%, resulting from overlapping π-π* GO bonds transmission and CS polymerization of n-π* bonds [41]. This result shows a successful chemical relocation of CS polymer with GO.33,34 The observed peak as a shoulder at about 297 nm is related to the transfer of n-π* from the C = O bond for CS/1wt% GO nanocomposite coating. By contrast, increasing the GO contents to 3wt%, the shoulder peak exhibits a small red shift to 305 nm, indicating an increase in the GO layers in the nanocomposite coating [42].
To further identify the existence of the GO in nanocomposite coating, Raman analysis was applied. Figure 3(c) shows the Raman spectrum of GO and the CS/2wt% GO nanocomposite. As expected, both the conspicuous and characteristic peaks of GO were observed at 1350 cm− 1 (D band, associated with the disorder in GO structure) and 1575 cm− 1 (G band, caused by the vibration of sp2 carbon atom). However, the G band for the CS/2wt% GO was shifted upward to 1596 cm− 1. The spectral shifts could be ascribed to disturbing the GO structure caused by the physical or chemical interactions between the carbon/oxide atoms of GO and the reactive sites of CS [43].
The XRD patterns of the coatings with GO different values are indicated in Fig. 3(d). The characteristic XRD peak of GO appeared at 2θ = 10.5° [44]. All nanocomposite coatings illustrate two peaks, one at 15.3º corresponding to the hydrated crystalline structure, whereas the broadened peak at about 20.3º exhibits an amorphous structure [45].
The diffraction angles of the nanocomposite coatings were similar to the CS, and the diffraction peaks corresponding to GO were not seen in the XRD peak of nanocomposite coatings. It is indicated that the addition of GO did not change the amorphous structure of CS, and GO was well exfoliated with CS in nanocomposite coatings. The diffraction intensity of CS was a little stronger than that of nanocomposite coatings at 2θ = 15.3°, implying that the crystalline degree of chitosan decreased after adding GO into CS. Also, the chemical structure of the CS in the nanocomposite coatings was changed with the increasing GO values, indicating that there was mainly physical interaction between CS and GO.
3.3. Determination Of The Optimal Coating By Taguchi Method
Table 2 shows nanocomposite coatings suggested by Taguchi's design. Using Minitab software, the Taguchi method offers the best levels for each parameter to prepare the optimal nanocomposite coating with the highest properties. Figure 4 represents the best PED parameters and levels (2:2:3:2) for nanocomposite coatings suggested by the Taguchi method (CD = 20 mAcm− 2, GO wt %=2, DC = 0.5, and pH = 5).
Table 2
The results of PED parameters suggested by Taguchi design for CS/GO nanocomposite coatings
Coating | CD (mA cm− 2) | GO content (%) | DC | pH |
1 | 10 | 1 | 0.1 | 4 |
2 | 10 | 2 | 0.2 | 5 |
3 | 10 | 3 | 0.2 | 6 |
4 | 20 | 2 | 0.5 | 4 |
5 | 20 | 2 | 0.1 | 5 |
6 | 20 | 3 | 0.5 | 6 |
3.4. Determination Of The Highest Corrosion Resistance Of Cs/go Coatings By The Taguchi Method
EIS measurements were also carried out in SBF to elucidate the corrosion protection mechanism of coatings. Figure 5 presents the Bode curves of nanocomposite coatings. The sample obtained by Taguchi factors is the optimal coating. The curves in the high-frequency range correspond to the solution resistance, while the curves in the middle-frequency range are related to the properties of passive layers. Moreover, the curves in the low-frequency range correspond to the interface between the passive layers and the Mg-2Zn surface. The impedance magnitude of the coating suggested by the Taguchi method is about 1.47 ×1010 Ω, which is the highest among all the coatings. This result shows that the coating suggested by the Taguchi method can significantly improve the corrosion resistance compared to the other coatings. Also, the slope of the curve in the middle of this coating is more significant than other coatings, indicating the capacitive properties. It is noteworthy that the Bode curve of the nanocomposite coating suggested by the Taguchi method is closer to Taguchi 5 and 6 (Table 2).
3.5. Evaluation Of Corrosion Resistance Of The Optimal Coating
The potentiodynamic polarization curves are used for comparing the corrosion resistance of optimal coating with Mg-2Zn scaffold and CS coating. Figure 6(a) indicates the polarization curves of coatings at DC = 0.5, CD = 20 mA/cm2, pH = 5, t = 20 min, T = 37 oC, and f = 1000 Hz in SBF. Table 3 lists the parameters extracted from the polarization curves using Corr. View Software. The optimal coating's corrosion potential (Ecorr) was increased. However, the corrosion current density was decreased compared to the Mg-2Zn and CS coating. Also, the corrosion rate of the optimal coating was 0.1442 mm/year, which was very low and appropriate compared to the CS coating and Mg-2Zn, with corrosion rates of 0.4058 and 2.1065 mm/year, respectively.
Figure 6(b) shows the relationship between the frequency and impedance magnitude of Mg-2Zn coatings. The bode plot denotes the higher impedance modulus at low frequency. The impedance magnitude of Mg-2Zn, CS, GO, and optimal coatings at low frequencies are about 395, 9395, 3268, and 1.47 ×1010 Ω, respectively. The impedance magnitude shows that coating with optimal conditions improves corrosion resistance. In addition, the slope of the curves in the middle-frequency range for the optimal coating is more significant than other coatings, indicating the capacitive properties.
Table 3
The parameters extracted from the polarization curves
Scaffold | Corrosion potential (V) | Corrosion current density (µA/cm2) | Corrosion rate (mm /year) |
Mg-2Zn | -1.64 ± 0.02 | 97.60 ± 1.22 | 2.1065 ± 0.9245 |
CS coating | -1.59 ± 0.02 | 18.88 ± 0.63 | 0.4058 ± 0.0051 |
Optimal coating | -1.52 ± 0.02 | 6.71 ± 0.34 | 0.1442 ± 0.0027 |
3.6. Thermal Stability Analysis
TGA, DTG, and DTA were used to characterize the thermal properties of GO, CS, and optimal coating (Fig. 7 (a and b)). TGA is considered the most important method for studying the thermal stability of polymers. The peaks represent the weight loss at the specific temperature range. According to the TGA curve, there are three stages in the thermal decomposition of CS. The first stage centered at about 70°C to 100°C due to the moisture content. The second stage begins at 284°C and ends at 330°C, which shows the degradation of leading polymer chains, [46] and three-stage is related to the degradation of the lateral chains. The TGA curve indicates that the significant weight loss is about 17% which is attributed to the degradation of the polymer backbone [46]. Two more peaks centered at 284°C and 333°C for CS coating are related to the initial and final decomposition stages in correlation to DTG and DTA curves. The GO had a weight loss of almost 15% below 200°C from water evaporation and about 40% at 318°C due to the elimination of oxygen-containing functional groups [47]. Here, it can be noted that the DTA maximum temperature order is the same as DTG. In the case of the optimal coating, the second peak at 294°C is related to the decomposition of epoxy and carboxylic groups, which is 10°C higher than the second peak of CS (284°C). Since the optimal coating mainly comprises CS, its thermal decomposition diagram is similar to the CS coating.
Moreover, nanocomposite coating exhibits better thermal stability than GO and CS coatings. The weight loss at the temperature range of 284°C-333°C for CS and 94°C-333°C for CS/2wt% GO is 22% and 17%, respectively. At around 450°C, the remaining weight of GO, CS, and optimal coatings was 54%, 56%, and 61%, respectively. The reason for the low weight loss of the optimal coating is the existence of functional groups on the surface of GO that interacts with the matrix of CS. Also, proper dispersion of GO sheets in the matrix of pure CS plays an essential role in this field. These factors prevent the slip of CS polymer chains during heat treatment in the weight loss test.
3.7. Surface Morphology Of The Optimal Coating
SEM and AFM techniques characterized the morphology of CS and the optimal coating. The SEM micrographs of the CS polymer and the optimal coating are shown in Fig. 8. Porosity is seen in the SEM image of the CS biopolymer (Fig. 8(a)), whereas the CS-2wt% GO optimal coating exhibits a somewhat rough surface and less porosity (Fig. 8(b)). The dispersion and the exfoliation of GO nanosheets in the CS matrix can affect porosity. The pore size of the CS coating is approximately 100–400 µm (Fig. 8(c)). However, the optimal coating is between 50–250 µm (Fig. 8(d)). The suitable porosity is created according to the scaffold requirement for osteoblast cells' growth, adhesion, and proliferation [48].The optimal coating is integral and dense on the surface of the scaffold with a thickness of about 1.67 µm (Fig. 8(e)).
Surface roughness was inspected using AFM. This method possesses a three-dimensional ability allowing the quantification of the surface roughness. This is typically accomplished using the AFM software. Figure 9 (a and c) show three-dimensional AFM scans of the CS and optimal coating surfaces, respectively. The scans were taken on a small coating surface area directly behind the pre-existent roughness. The two-dimensional profiles were quantified using the AFM image processing software. The higher values of the maximum height of the CS and optimal coatings peaks were 0.95 and 1.73 µm, respectively (Fig. 9 (b and d)).
Compared to the CS coating, the surface of the optimal coating was rougher. This is consistent with the findings in the literature indicating that GO nanosheets can disperse in the CS matrix. Here, it is essential to highlight that the nanomaterial coatings with notable enhancement in surface roughness are always favorable for positive utilization in biological investigations [49].
3.8. In Vitro Biocompatibility Of Coatings
The MTT method was further used to quantify the proliferation of MG63 cells cultured onto the Mg-2Zn scaffold, CS, and optimal coating after three days (Fig. 10 (a)). The cell viability percentage on the Mg-2Zn scaffold, CS, and the optimal coating were 83, 95, and 98%, respectively. It is clear that the cells on optimal coating proliferated better than Mg-2Zn and CS coating, mainly due to the simultaneous effect of CS and GO nanosheets.
Suo et al. investigated the enhancement of osseointegration using HA, CS, and GO/CS/HA composite coatings on titanium fabricated by electrophoretic deposition. The results show that HA and CS coating considerably enhanced the bone marrow stromal cell (BMSCs) interactions in vitro. In addition, this GO/CS/HA coating might increment osseointegration in vivo. Consequently, GO/CS/HA may have potential applications in dental implants [50].
The cell morphology of the MG63 cultured cells on Mg-2Zn, CS, and optimal coating after 5 hours was indicated in (Fig. 10 (b, c, and d)). The spindle-shaped cells could be seen on the samples (Zones with red circles). However, the MG63 cells on optimal coating exhibited better stretch with more Filopodia extensions (Zones with blue circles) than the Mg-2Zn and CS coating. The presence of multiple Filopodia expanded coating provides solid proof that the cells are well attached to the optimal coating. Therefore, the optimal coating can be used as a suitable coating for Mg-2Zn in bone tissue engineering.