XRD analysis

The XRD patterns of NO₃-LDH (Fig. 1) exhibited the characteristic reflection peaks of LDH at 11.52 and 23.40°2θ, corresponding to the (003) and (006) reflection planes, respectively (Reference Gomes, Mir, Sampaio, Bastos, Tedim, Maia, Rocha and FerreiraGomes et al., 2020). According to Bragg’s equation, the d ₀₀₃ spacing of LDH was calculated to be 0.7629 nm, which is in good agreement with previous reports (0.78 nm, Reference Zhang, Luan, Gao, Li, Li and WuZhang et al., 2017). The characteristic reflection peaks of LDH and the characteristic (001) reflection peak of Mnt (d ₀₀₁ = 1.46 nm) were observed in all composites, and the degree of crystallinity could be seen from the sharpness of the reflection peaks; it changed with different mass ratios of LDH/Mnt. When the mass ratio of LDH/Mnt was 1:1 or 1:2, the composite crystallized well. Both LDH and Mnt remained in their crystalline states in all composites. Compared with the XRD patterns of LDH and Mnt alone, the lesser intensity of the composites was caused by the mixture of the two. The position of the characteristic reflection peak of LDH shifted to the left (to ~10.82°2θ) in the synthesized intercalated composite, indicating that the interplanar spacing of the LDH/Mnt composites increases due to the self-assembly of LDH and Mnt (Reference Bakr, Mostafa and SultanBakr et al., 2016). In addition, the characteristic peak of LDH/Mnt-1 (~21.65°2θ) corresponded to the later analysis of TEM images. The composition of LDH and Mnt was not a regular layer-by-layer assembly but a random assembly. However, the LDH/Mnt composites in the present work were used mainly to provide their interlayer ions, so random assembly of layers had no effect on the ion-exchange property.

Fig. 1 XRD patterns of LDH, Mnt, and LDH/Mnt composite materials with various mass ratios of Mnt (LDH/Mnt-0.33, LDH/Mnt-0.5, LDH/Mnt-1, LDH/Mnt-2, LDH/Mnt-3)

Laser Raman spectroscopy analysis

In the Raman spectrum of LDH (Fig. 2), the characteristic peak of nitrate in the triclinic system occurred at 1048 cm^–1, while in the composites, the characteristic peak of nitrate in the rhombic crystal system at 1066 cm^–1 could be seen (Reference Jentzsch, Kampe, Ciobotă, Rösch and PoppJentzsch et al., 2013), indicating that NO₃-LDH was prepared successfully. In the Raman spectrum of Mnt, the strongest bands occurred at 147 cm^–1, which was also reflected in all the composite material samples. The occurrence of this characteristic peak was attributed to the vibrations of the distorted MO₆ octahedron with S₆ symmetry where M is the octahedral cation (Reference Frost and RintoulFrost & Rintoul, 1996). Other weaker bands were observed at 262 and 702 cm^–1. The former is attributed to the vibrations of the H–O–H triangle of C_2V symmetry and the latter to the ν₁(a₁) mode of the SiO₄ tetrahedra (Vaculíková et al., Reference Vaculíková, Plevová and Ritz2019).

Fig. 2 Laser Raman spectra of LDH, Mnt, and LDH/Mnt composite materials

N₂-adsorption/desorption analysis

The N₂-adsorption/desorption isotherms of the NO₃-LDH, Mnt, and LDH/Mnt composites (Fig. 3) revealed that all the samples belonged to type IV adsorption according to the IUPAC classification, which meant the NO₃-LDH, Mnt, and LDH/Mnt composites belonged to a mesoporous-type material characterized by a H₃-type hysteresis loop (Sing, Reference Sing1985, Reference Sing1989). The mesoporous type indicated the presence of aggregates of plate-like particles giving rise to slit-shaped pores with non-uniform size and shape (Reference Lecloux and PirardLecloux & Pirard, 1979). The isotherm showed that the more LDH components in the composite, the larger the pore volume and BET specific surface area. The pore-size distribution of the NO₃-LDH, Mnt, and LDH/Mnt composites were calculated from Barrett-Joyner-Halenda (BJH) methods (Fig. 4). The BJH pore-size is calculated as follows:

(4)$V=\frac{4}{3}\pi r^{3}N/Av$

where V is the pore volume, r is the pore size, N is the number of adsorbent molecules, A is Avogadro constant, and v is the molecular volume of adsorbent. Due to the large range of pore-size distribution, logarithmic coordinates were used for analysis, where dV/dlog(d) represented the pore volume. The surface areas of the NO₃-LDH, Mnt, and LDH/Mnt composites were calculated from Brunauer-Emmett-Teller (BET) methods (Table. 2). The data obtained from BJH indicated that the mean pore size of the samples was fitted well to the size range of layered materials and revealed that the pore-size distribution of the LDH was much wider than that of Mnt. However, these large pores disappeared, and the pores were highly uniform after formation of the composite.

Fig. 3 N₂-adsorption/desorption isotherms of NO₃-LDH, Mnt, and LDH/Mnt composite materials

Fig. 4 Barrett-Joyner-Halenda pore-size distributions for NO₃-LDH, Mnt, and LDH/Mnt composite materials

TEM analysis

The microstructure of synthesized LDH and LDH/Mnt-1 composite (Fig. 5) revealed that LDH possesses a polygonal lamellar structure (Fig. 5a). The lamellar size was relatively uniform, compact stacking was irregular, and intergranular spacing was small, because of preparation under a saturated sodium hydroxide solution, where LDH nucleated rapidly and particles grew rapidly. Meanwhile, some LDH lamellae grew crosswise under the condition of weak interactions, forming support structures to improve agglomeration, which was conducive to the stability of layered structures (Adachi-Pagano et al., Reference Adachi-Pagano, Forano and Besse2003). In Fig. 5b with greater magnification, the lattice fringe spacing of LDH was determined to be 0.21 nm, corresponding to the peak position (006) in the XRD pattern. After composite formation, the lamellar structure of Mnt and LDH flakes were observed (Fig. 5c) on the surface and the interlayer structure became loose and dispersed due to the self-assembly of LDH and Mnt layers (Reference Dong, Ma and ZhouDong et al., 2013). The lattice fringe spacing (0.41 nm) of the new layer structure occurred in the image of the composite sample LDH/Mnt-1 (Fig. 5d), corresponding to the characteristic peak of LDH/Mnt-1 (21.65°2θ) in the XRD pattern.

Fig. 5 TEM images of a,b LDH and c,d LDH/Mnt-1

TG-DSC analysis

The combined thermogravimetry and differential scanning calorimetry behavior of LDH/Mnt composites were investigated by TG-DSC tests (Fig. 6). Two important weight-loss stages appeared on the TG curves of all samples (Fig. 6a). The mass decline in the first stage was caused mainly by the desorption of surface-adsorbed water and loss of interlayer bound water, corresponding to the wide endothermic valley at ~200℃ on the DSC curve (Fig. 6b), which was consistent with the commonly known thermal characteristics of LDH (Kanezaki, Reference Kanezaki1998). According to the calculation, the mass loss of five groups of samples during the first stage was ~10%. Among them were two small thermal valleys in the TG curve of LDH/Mnt-3, i.e. at 210 and 144°C, representing the thermal characteristics of Mnt (Reference Lv, Li, Jiang, Chang and LiaoLv et al., 2015) and corresponding to the removal of interlayer water and adsorbed water. When the proportion of LDH in the composite increased, the proportion of Mnt decreased and the small endothermic peak near 144°C decreased gradually and finally disappeared. In the second weight-loss stage, mainly at ~240–480°C, the hydroxyl groups were removed from the hydroxide layer of the LDH, along with removal of the intercalated nitrate ions. The mass loss of LDH/Mnt-0.33, LDH/Mnt-0.5, LDH/Mnt-1, LDH/Mnt-2, and LDH/Mnt-3 during the second stage was 17, 12, 11, 8, and 5.53%, respectively. When the proportion of LDH in the complex increased, the mass loss during the second stage also increased due to the decomposition of nitrate ions. The mass proportion of LDH was calculated as 74.12, 62.11, 46.13, 32.92, and 22.84%, respectively, which is consistent with the theoretical proportion. Moreover, the removal of hydroxyl groups from Mnt lamellae generally occurs in the temperature range of 600–700°C, showing slight fluctuations in the curve (Reference Zou, Zhang, Wang, Xue and ChenZou et al., 2020). At >600°C, the residual nitrate ions between the layers of NO₃-LDH continued to decompose and MgAl₂O₄ started to crystallize (Reference Wu, Zuo, Dong, Xing and LuoWu et al., 2019). At ~800–953°C, the structure of Mnt disintegrated, the lattice was completely destroyed, and an amorphous structure was formed.

Fig. 6 a TG curve and b DSC curve of NO₃-LDH, Mnt and LDH/Mnt composite materials

XRF analysis

According to previous literature (Reference Yusuf, Moheb and DinariYusuf et al., 2021), the structural formula of LDH is Mg_1–x²⁺Al_x³⁺(OH)₂ (NO₃^-)_x·mH₂O and the structural formula of Mnt is E_x(Al_2–x,Mg_x)₂·{p[Si₄O₁₀]}(OH)₂·4H₂O. The observed fractions of Mg and Al in the octahedral sheet were Mg/(Mg+Al) = 83.60% and Al/(Mg+Al) = 16.40% (Table 3). Plugging the ratio into the structural formula for LDH and Mnt resulted in x=0.328, so the stoichiometries of Mg and Al in 100g LDH/Mnt-1 in the Mnt structural formula were, therefore, 1.672 and 0.328 mol, respectively. Based on the mass ratio of Fe₂O₃, CaO, Na₂O and K₂O and their relative molecular mass, The exchangable cation content, E_x(Fe, Ca, Na, K), was a total of 0.0652 mol in 100g LDH/Mnt-1. As E_x was supplied entirely by Mnt, the amount of Mnt was 0.39 mol. The relative fraction of Si in the Mnt, p, was calculated from the stoichiometry of Si and based on the amount of Mnt, resulting in a value of 2.37. After the removal of Mg and Al from Mnt, the amounts of Mg and Al in the complex were 0.2576 and 0.051 mol, respectively. According to the proportions, Mg/(Mg+Al) = 83.47% and Al/(Mg+Al) = 16.53%, therefore, the amounts of Mg and Al in the structural formula of LDH were 0.836 and 0.164 mol, respectively, where the amount of LDH was calculated to be 0.31 mol. m generally decreased with increase in x, showing a relationship of m = 0.8 – x. Hence, m = 0.636 in LDH/Mnt-1. Therefore, the molar ratio of LDH/Mnt-1 was 0.31/0.39. The formula of NO₃-LDH was Mg_0.836²⁺Al_0.164³⁺(OH)₂(NO₃^-)_0.164·0.636H₂O and the formula of Mnt was (Fe,Ca,Na,K)_1.672(Al_0.328, Mg_1.672)₂·{2.37[Si₄O₁₀]}(OH)₂·4H₂O.

Table 3 Elemental analysis of LDH/Mnt-1

Electrochemical analysis

EIS was carried out to explore the corrosion resistance behavior and mechanism of the self-assembled LDH/Mnt composites, including the Nyquist and Bode plots of carbon steel soaked in SCCP solution with and without LDH/Mnt composites (Fig. 8). The fitting parameters are summarized in Table 6.

Fig. 8 The a Nyquist plots and b Bode plots of carbon steel soaked in SCCP solution with and without NO₃-LDH, Mnt, and LDH/Mnt composite materials added after immersion

Table 6 The fitting parameters obtained from EIS with various LDH/Mnt composite materials in SCCP solution

The impedance spectrum of the control solution (Fig. 8a) consisted of a small, high-frequency capacitive ring and a low-frequency Warburg impedance, indicating corrosion of the carbon steel surface. After the addition of the LDH/Mnt composite, the Warburg impedance disappeared from the Nyquist plots and only the capacitive ring remained visible. The diameter of the capacitive ring increased significantly and the maximum phase angle became larger and wider, indicating that the LDH/Mnt composite greatly improved the impedance of carbon steel. The occurrence of disposable time constants (only one capacitive reactance arc) in the Bode diagram (Fig. 8b) was attributed to the formation of a passive film on the surface of carbon steel (Reference Wang, Xiang, Liang, Song and LiuWang et al., 2014). At this time, the high polarization resistance in the SCCP solution indicated that the carbon steel electrode was completely passivated. The diameter of the capacitive ring followed the order: LDH/Mnt-2 > LDH/Mnt-1 > LDH/Mnt-0.5 > LDH/Mnt-3 > LDH/Mnt-0.33, which was consistent with the trend of corrosion resistance. LDH/Mnt-1 and LDH/Mnt-2 rendered the optimal corrosion resistance, protecting the steel bar from corrosion caused by chloride ions.

To demonstrate the desirable correspondence between elements of a circuit and physical processes (Reference David and DriesscheDavid & Driessche, 2011), the electrochemically equivalent circuits (Fig. 9) were used to fit the EIS data, where R_s represents the ohm resistance between the working electrode and the reference electrode, R_ct denotes the interfacial charge conversion resistance, and W denotes the Warburg impedance. The total resistance of the circuit can be expressed as the following formula (Reference Javadian, Yousefi and NeshatiJavadian et al., 2013):

Fig. 9 Electrochemical equivalent circuits used to fit the EIS results in SCCP solution for a general semicircle and b Warburg type

(5)$Z=R_S+\frac{R_{\mathrm{ct}}+W}{1+\left(j,\omega \right)Z_{\mathrm{CPE}}W+\left(j,\omega \right)Z_{\mathrm{CPE}}{R}_{\mathrm{ct}}}$

Due to charge transfer, R_ct is an important parameter to judge the impedance of the carbon steel. The higher value of R_ct corresponded to the greater impedance. The constant phase element (CPE) was used widely to explain the deviation caused by the surface roughness, and its impedance can be expressed as the following formula (Reference Javadian, Yousefi and NeshatiJavadian et al., 2013; Reference Ryu, Singh, Lee, Ismail and ParkRyu et al., 2017):

(6)$Z_{\mathrm{CPE}}={\left({Y_0\left(j,\omega \right)}^u\right)}^{-1}$

where Y ₀ represents the size of CPE, the angular frequency, for j ² = –1, j is an imaginary number, and u is the index of CPE. The value of u is related to the uniformity of the carbon steel surface, and the larger the value of u, the better the uniformity. The behavior of CPE depends on the value of index u, and the modified C _dl for its actual capacity value is calculated as follows:

(7)$C_{\mathrm{dl}}={Y_0\left({\omega }_{\max }\right)}^{u-1}$

where ${\omega }_{\max }$ is the frequency at which the imaginary part of impedance (Z _im) maintains its maximum value. Corrosion inhibition efficiency, η%, is calculated according to the following formula (Reference Cao, Dong, Zheng, Wang, Zhang, Du, Song and LinCao et al., 2017; Reference Javadian, Darbasizadeh, Yousefi, Ektefa, Dalir and KakemamJavadian et al., 2017):

(8)$\eta \%=\frac{R_{\mathrm{ct}}-R_0}{R_{\mathrm{ct}}}\times 100\%$

In the formula, $R_{\mathrm{ct}}$ and R₀ represent the interface charge conversion resistance of carbon steel in SCCP solution before and after adding the composite, respectively. The R_ct reflected the corrosion inhibition effect of the passive film on the carbon steel surface before and after the action of composite material. The larger R_ct corresponded to better corrosion resistance and greater corrosion inhibition efficiency (η%).

According to the fitting parameters, the R_ct value of the carbon steel electrode increased after the addition of LDH/Mnt composite. The R_ct values of LDH/Mnt-1 and LDH/Mnt-2 reached 1224.04 and 1269.24 W·cm², respectively, and the corresponding corrosion inhibition efficiency was found to be 94.16 and 94.36%. The u value of LDH/Mnt-2 reached its maximum of 0.8060 and the electrode surface became dense and uniform. Y ₀ reached a minimum value of 252.96×10^–5 S·cm^–2s^u. Owing to the addition of composites, LDH released the interlayer anions, and calcium cations were replaced by Mnt. When the passive film on the steel surface was subjected to corrosion, the active adsorption of chloride was generated to form a calcium nitrate passive film, forming an active anti-corrosion system and improving the corrosion resistance. In addition, although the corrosion inhibition efficiency of LDH/Mnt-1 was slightly less than that of LDH/Mnt-2, its nitrate content was also 1% less than that of LDH/Mnt-2 according to the result of TG-DSC, so the overall corrosion inhibition performance of LDH/Mnt-1 was outstanding.

FTIR analysis

The FTIR spectra of the precipitates after ion exchange of LDH/Mnt-1 with chloride and sulfate (Fig. 10) demonstrated that the characteristic peaks of the complex sample did not change. However, the intensity of characteristic peaks of nitrate decreased significantly with the increase in adsorbent concentration. The concentration of nitrate ions between LDH layers decreased after the adsorption experiment.

Fig. 10 FTIR spectra of LDH/Mnt-1 after stirring in Cl^– and SO₄^2– solutions

Analysis of the formation of passive film

After the passive film was formed on the surface of the steel bar, three layers were observed in the TEM image of the cross-sectional sample (Fig. 11), including an electron beam, a passive film, and the steel bar (Ohet al., Reference Oh, Ahn, Eom, Jung and Kwon2014). In particular, the thickness of the second layer or the passive film was ~2.31 nm. The high-resolution X-ray photoelectron spectra (XPS) of Ca 2p and O 1s (Fig. 12a) revealed that the positions located at 351.73 and 348.23 eV correspond to characteristic peaks at Ca 2p_1/2 and Ca 2p_3/2, respectively (Reference Haber, Stoch and UngierHaber et al., 1976; Reference Christie, Lee, Sutherland and WallsChristie et al., 1983). The O 1s peak in the high-resolution XPS spectra (Fig. 12b) was used to confirm the existence of calcium ions in the form of calcium nitrate. The two dominant O 1s peaks were from calcium nitrate (533.60 eV) and ferric oxide (529.50 eV) (Reference Demri and MusterDemri & Muster, 1995). The iron oxide component of the steel bar was mixed in when the surface passivation film was scraped off, but it was not a product of the experiment. The above results showed that the passive film (Fig. 11) generated on the surface is calcium nitrate.

Fig. 11 TEM image of a cross-section of the steel bar after the passive film had formed

Fig. 12 XPS spectra of the surface product covering the steel bar in SCCP solution with LDH/Mnt-1 after immersion