Introduction

With the relentless march of modernization and industrialization, there has been a staggering surge in the release of untreated organic pollutants into wastewater. This alarming trend has exacerbated the water scarcity crisis and posed grave threats to both human well-being and ecological stability. Shockingly, reports indicate that a staggering 200,000 tons of dyes is annually released into the global environment without undergoing any treatment (Marnani and Shahbazi 2019). In China alone, a staggering 160 million cubic meters of dye wastewater is discharged into our aquatic ecosystems every year (Islam and Mostafa 2018). Organic dyes, as a category of highly hazardous pollutants frequently found in wastewater, with contact toxicity, latent pathogenicity, and carcinogenic potential, and through cumulative effects can have a significant impact on ecosystems (Khan et al. 2022; Sarkodie et al. 2023). Furthermore, the presence of chromophores or chromophore-like elements such as nitro (–NO2), azo (–N = N-), primary amine (–RNH2), and triarylmethane in dye molecules enables them to absorb light to varying degrees (Brown 1981; Christovam et al. 2022). This not only diminishes the aesthetic appeal of water bodies but also hampers the development of aquatic organisms. One such dye of concern is CR, a bis-azo dye commonly employed in industries like printing, dyeing, textiles, and paper manufacturing (Adebayo et al. 2022). CR is likewise both toxic and resistant to degradation, necessitating rigorous treatment of its industrial wastewater to meet environmental discharge standards.

To date, various wastewater treatment technologies, including biodegradation, chemical precipitation, photocatalytic degradation, and coagulation, have made significant strides in development (Abujazar et al 2022; Selvaraj et al. 2021). Nonetheless, these methods exhibit certain drawbacks, including their intricate and cumbersome procedures, substantial energy consumption, and elevated costs. In stark contrast, the conventional adsorption offers distinct advantages. Leveraging the porous structure of adsorbents and their strong affinity for target molecules, it facilitates the separation of pollutants, thereby circumventing the aforementioned limitations (Ghazvini et al. 2021). Although traditional adsorbents like activated carbon (Benmaamar et al. 2019), bentonite (Jiang et al 2022), zeolite (Imessaoudene et al. 2023), and resin (Xu et al. 2020) have been applied for dye removal in wastewater, they typically exhibit limited capacity and specificity. In pursuit of more efficient adsorbents, researchers have turned their attention toward emerging porous materials.

MOFs are fashioned through the bonding of metal ions or metal clusters with organic chains, held together by weak coordination bonds (Jeong et al. 2023; Mazloomi et al. 2019). MOFs are prized for their diverse structures, flexibility in adjustment, outstanding pore structures, generous specific surface areas, and ease of surface functionalization (Esrafili et al. 2022; Huang et al. 2023; Shi et al. 2023). MIL-101-NH2(Fe), in particular, stands out as a strong contender due to its utilization of the non-toxic transition metal Fe, which offers an abundance of adsorption active sites. Its expansive specific surface area and open cavities enhance site accessibility, rendering it exceptional in the realm of pollutant removal (Lian et al. 2019). It is noteworthy that the activity of amidogen (–NH2) MOF and the positioning of modification sites can introduce additional diffusion interactions between the ligand and the skeleton (Lim et al. 2022; Liu et al. 2022a; Pallach et al. 2021; Razavi et al. 2022). This subsequently impacts the energy landscape and pore structure of the MIL-101-NH2, ultimately altering its adsorption capabilities for the better. Whereas MOF’s remarkable water dispersibility and powder form pose a significant challenge in recyclability, leading to a substantial increase in the cost of adsorption and creating a bottleneck in its practical application. The introduction of magnetic components, especially Fe3O4, is regarded as an important means to solve such difficulty (Mahmoudian et al. 2023; You et al. 2018; Zheng et al. 2023). The complementary and compatible nature of Fe3O4 and MOF has paved the way for their widespread integration into composite structures (Cheng et al. 2023; He et al. 2023; Lee et al. 2023; Li et al. 2020; Lu et al. 2022; Wang et al. 2021b; Zheng et al. 2018). Nevertheless, it is essential to acknowledge that the magnetism of Fe3O4 is easily weakened by oxidation and lose their effectiveness, but the addition of another magnetic substance, like Co, Mn, and Ni, can create a binary magnetic oxide, bolstering its chemical stability and enhancing mass transfer efficiency (Cross et al. 2023; Wang et al. 2021a, 2019).

The novelty of this study is centered around the first-time synthesis and application of Co0.6Fe2.4O4/MIL-101-NH2 for dye industry wastewater treatment, with a particular focus on the removal of CR. A series of characterizations were conducted to attest the successful preparation of Co0.6Fe2.4O4/MIL-101-NH2. The study showed satisfactory results in the adsorption capacity, specific selectivity, and reusability of Co0.6Fe2.4O4/MIL-101-NH2 for the removal of azo dye CR. These reveal that the Co0.6Fe2.4O4/MIL-101-NH2 can provide a viable option for the removal of other organic dyes or similar compounds from wastewater. Therefore, the research objectives in this study are summarized as follows: preparing Co0.6Fe2.4O4/MIL-101-NH2 adsorbent and its characterization, exploring the effects of a series of factors on CR adsorption, systematically studying the isotherm, kinetics, and thermodynamics characteristics of adsorption of Co0.6Fe2.4O4/MIL-101-NH2 to CR, exploring the repeated use and specific selectivity of Co0.6Fe2.4O4/MIL-101-NH2, and clarifying the adsorption mechanism.

Materials and methods

Chemical and materials

FeCl3•6H2O were acquired from Shanghai McLean Biochemical Technology Co., Ltd. (Shanghai, China). Cobalt chloride hexahydrate (CoCl2•6H2O), ferrous sulfate heptahydrate (FeSO4•7H2O), oxalic acid dihydrate (H2C2O4•2H2O), ethylene glycol (EG), N, N-dimethylformamide (DMF), polyvinylpyrrolidone (PVP, K-30), and ethanol absolute (EtOH) were supplied by Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). 2-Aminoterephthalic acid (NH2-TPA), aerosol OT (AOT), and CR were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). All reagents were of analytical grade and used without further treatment. Deionized water (DI) was adopted to all aqueous solution.

Instrumentations

The Fourier transform infrared (FT-IR) spectral was recorded on a Thermo Scientific Nicolet 5700 (America) with KBr pellets in the 4000–500 cm−1 region. Powder X-ray diffraction (PXRD) patterns were acquired by Rigaku Mini Flex 600 (Japan) diffractometer at 40 kV among 3.0–80.0 (2θ) with V-filtered Cr Ka radiation. The size and morphology of materials were observed on a Zeiss Gemini SEM 500 (Germany) field emission scanning electron microscope (FESEM) equipped with an Oxford Ultim Max 65 (Britain) energy dispersive X-ray spectroscope (EDS). The thermal gravimetric analysis (TGA) curves were obtained by a NETZSCH TG209 F1 Libra (Germany) thermogravimetric analyzer. The test was carried out in the nitrogen environment and with a 20 mL•min−1 rate (room temperature–800 °C, 10 °C•min−1). The BET test was determined by Micromeritics 3FLEX (America) through N2 adsorption. Magnetic properties of the samples were characterized by the Lake Shore 7404 (America) vibrating sample magnetometer (VSM) among ± 2 T at room temperature. The zeta potential of samples was measured by a Malvern Zerasizer 3690 (Britain) zeta potential analyzer. The Fe leaching experiment was tested by PQ-MS (Germany) inductively coupled plasma mass spectrometry (ICP-MS). Characterization of the elements contained in the samples was caught by an Thermo Fisher Scientific Escalab250Xi (America) X-ray spectrometer (XPS).

Synthesis of Co0.6Fe2.4O4

Magnetic nanoparticles (Co0.6Fe2.4O4) were prepared by conventional co-precipitation method (Zeng et al. 2014), as Fig. 1 a. Firstly, AOT (1.78 g) was dissolved in a mixture of EG (20 mL) and DI (20 mL) to make a homogeneous solution, and two identical above solutions were prepared and denoted as liquid A and liquid B, respectively. Subsequently, CoCl2•6H2O (0.19 g) and FeSO4•7H2O (0.89 g) were added to liquid A, and oxalic acid (0.50 g) was dispersed to liquid B. Immediately after mixing liquid A with liquid B and stirring the mixed solution at 25 °C for 6 h, the reaction was washed with EtOH and DI three times, and the orange precipitate was collected by centrifugation and dried under vacuum at 50 °C for 12 h. Finally, the obtained light yellowish powder was heated in a tube furnace at 1 ℃•min−1 to 400 °C for 2 h to obtain black magnetic Co0.6Fe2.4O4 particles.

Fig. 1
figure 1

Schematic diagram of preparation for Co0.6Fe2.4O4 (a) and Co0.6Fe2.4O4/MIL-101-NH2 (b)

Full size image

Preparation of MIL-101-NH2

MIL-101-NH2 was synthesized using solvothermal method as reported previously. NH2-TPA (0.23 g) was firstly dissolved in DMF (10 mL). Simultaneously, FeCl3•6H2O (0.68 g) was scattered in another DMF (10 mL). The two DMF solutions were mixed into a Teflon-lined stainless steel autoclave and reacted in an oven at 110 °C for 20 h. After cooling, the precipitate was washed with DMF and EtOH several times, centrifuged, and dried to obtain brown MIL-101-NH2.

Preparation of Co0.6Fe2.404/MIL-101-NH2

As seen in Fig. 1 b, the surfactant PVP (0.40 g) was firstly dissolved in DMF (10 mL), then magnetic nanoparticles (0.17 g) needed to be appended to the aforesaid DMF solution and sonicated for 1 h at room temperature before the addition of quantitative NH2-TPA. The following procedure was the same as the preparation of MIL-101-NH2.

Batch adsorption experiment

Batch adsorption experiments were carried out under continuous mechanical stirring at a speed of 300 rpm and protected from light. Three replicate experiments were performed and the average values were taken to reduce the error. The effects of adsorbent dosage, time, CR concentrations, pH, and interfering ions were analyzed by measuring the removal of CR by the adsorbents. After adsorption, the solution was filtered through a 0.45-μm microporous membrane and the concentration of CR in the filtrate was measured using ultraviolet (UV-3600) at the maximum absorption wavelength (λmax = 497 nm). The pH of the CR solution was adjusted adopting 0.1 M HCl or NaOH solution. The removal rate (r, %) and adsorption capacity (Qt, mg•g−1) were calculated by Eqs. (1) and (2).

$$r= \frac{{C}_{t}-{C}_{0}}{{C}_{0}}*100 \%$$
(1)
$${Q}_{t}= \frac{\left({C}_{0}-{C}_{t}\right)*V}{m}$$
(2)

where C0 (mg•L−1) and Ct (mg•L−1) are the initial and final concentrations of CR before and after certain adsorption time, mg•L−1; V (L) is the volume of CR solution; m (g) is the mass of adsorbent.

Results and discussion

Characterization

Figure 2 a shows the XRD spectra of Co0.6Fe2.4O4, MIL-101-NH2, and Co0.6Fe2.4O4/MIL-101-NH2. The formation of magnetic cobalt-iron bimetallic oxides with good crystallinity is demonstrated by the diffraction peaks of 2θ at 18.44° (1 1 1), 30.34° (2 2 0), 35.66° (3 1 1), 37.26° (2 2 2), 43.28° (4 0 0), 53.64° (4 2 2), 57.26° (5 1 1), 62.82° (4 4 0), 65.86° (5 3 1), 71.14° (6 2 0), 74.28° (5 3 3), 75.44° (6 2 2), and 79.24° (4 4 4). These peaks agree with the typical spinel CoFe2O4 (JCPDS No. 22–1086). Based on prior reports in the literature (Li et al. 2017; Song et al. 2021), the high crystallinity diffraction peaks of MIL-101-NH2 (Fe) correspond with the distinctive diffraction peaks situated at 2θ = 9.36°, 10.26°, 16.68°, 18.72°, and 20.65°. The effective fabrication of the composite Co0.6Fe2.4O4/MIL-101-NH2 is justified by the synchronized emergence of the distinctive peaks of both components on the XRD spectrum of magnetic MOF.

Fig. 2
figure 2

XRD images (a), FT-IR images (b), and Raman patterns (c) of Co0.6Fe2.4O4, MIL-101-NH2, and Co0.6Fe2.4O4/MIL-101-NH2; TG and DTG (d) curves of the precursor of magnetic oxide

Full size image

The functional groups of the prepared materials were qualitatively analyzed by FT-IR spectroscopy (Fig. 2 b). For Co0.6Fe2.4O4, the broad peaks near 595 cm−1 and 1635 cm−1 are related to the stretching vibration of Fe–O, Co–O, and C = O bond (Olusegun and Mohallem 2020). The peaks of Fe–O and Co–O of Co0.6Fe2.4O4/MIL-101-NH2 are red-shifted (to 617 cm−1), which indicates that the two components are successfully compounded. In the spectrogram of MIL-101-NH2, the peak near 1492 cm−1 belongs to the vibrational mode of C = C in the benzene ring, the peaks at 1617 cm−1 and 769 cm−1 are connected with the in-plane and out-of-plane bending vibration of N–H bond, the wavenumbers of 1337 cm−1 and 1258 c−−1 are corresponding to the stretching vibration of C–N bond, and the peaks observed at 1581 cm−1 and 1431 cm−1 are the results of the antisymmetric and symmetric stretching vibrations of COO–, respectively (Zhang et al. 2020). The two distinctive COO– peaks experience a blue shift as the outcome of the compositing procedure. The O–H stretching vibration associated with the large peak around 3417 cm−1 is mostly due to the water that the sample has absorbed.

Raman spectra of the three materials were recorded and compared to demonstrate the material structure changes when MIL-101-NH2 was combined with magnetic particles (Fig. 2 c). In MIL-101-NH2, sharp peaks around 1768, 1442, 1426, 1260, 801, and 685 cm−1 are indicative of the distinctive metal terephthalate vibrations (Song et al. 2021). The 688 and 479 cm−1 in Co0.6Fe2.4O4 agree with the characteristic vibrations of Co–O and Fe–O, respectively (Zhao et al. 2017). Characteristic peaks of two components emerged in Co0.6Fe2.4O4/ MIL-101-NH2 and some of them shift to a certain extent, implying that the components had successfully undergone recombination.

As indicated in the literature (Zeng et al. 2014), the chemical formulation of the precursor for the magnetic oxide can be represented as Co0.2Fe0.8C2O4•2H2O. The composition of the magnetic oxide was substantiated through TGA analysis of the precursor. As delineated in Fig. 2 d, prior to attaining a temperature of 400 °C, the precursor undergoes two distinct stages of mass reduction, registering at 20.12% and 39.17%, respectively. These correspond to two distinct processes: the release of crystalline water and the carbonization of Co0.2Fe0.8C2O4•2H2O. It is from these measurements that we can deduce the chemical formula of the resultant magnetic material to be Co0.6Fe2.4O4. It is noteworthy that the weight loss rate in the second stage exhibits some deviation from the theoretical value, potentially attributed to the volatilization of residual solvent molecules (Zeng et al. 2014).

The microstructure and particle size of the materials were meticulously examined via SEM. It is readily apparent that the synthesized Co0.6Fe2.4O4 exhibited stick-like particle morphology, as illustrated in Fig. 3 a. The MOF particles displayed a distinctive octahedral structure, characterized by a uniform surface and particle size, as depicted in Fig. 3 b. These findings align harmoniously with previously report (Song et al. 2021). It is clear from a thorough look of the magnetic MOF’s SEM pictures (Fig. 3 c, d) that the spindle-like surface was scattered with large numbers of granular nanoparticles. The cause of this particular morphology may be that certain magnetic particles encapsulated within the MIL-101-NH2 shell affected the crystalline nucleation of MOF. Consequently, the Co0.6Fe2.4O4/MIL-101-NH2 takes on a spindle-shaped polyhedral form and is a roughly five-fold enlargement in sizes (width: 664.86 nm; length: 2153.13 nm). The solvothermal process also wielded an impact on the microscopic structure and size of Co0.6Fe2.4O4, leading to the conversion of magnetic particles, i.e., from previous stick shape to substantial nanoparticles dispersed on the MOF’s surface. Figure 3 e–k provides a comprehensive view of the EDS images of the magnetic MOF. By scrutinizing the distribution of three key elements—N, Fe, and Cothe morphological transformations of Co0.6Fe2.4O4 and MIL-101-NH2 within the magnetic MOF, along with the distribution of Co0.6Fe2.4O4 within the MOF shell and on its surface, were further affirmed. All of these results clearly show that Co0.6Fe2.4O4/MIL-101-NH2 is not only a physical mixing of Co0.6Fe2.4O4 and MIL-101-NH2 distribution diagram that the pore sizes of them are mostly distributed between 2 and 10 nm (MIL-101-NH2 and Co0.6Fe2.4O4/MIL-101-NH2) and 5 and 20 nm (Co0.6Fe2.4O4), further confirming the mesoporous structure of the sample. The BET specific surface areas of MIL-101-NH2, Co0.6Fe2.4O4, and Co0.6Fe2.4O4/MIL-101-NH2 are 142.159, 44.796, and 240.485 m2•g−1, respectively. The increase of specific surface area after recombination indicates that the porous structure of the composite components and the uniform dispersion of the components during the compounding process result in a large specific surface area of Co0.6Fe2.4O4/MIL-101-NH2.

Fig. 3
figure 3

SEM images of Co0.6Fe2.4O4 (a), MIL-101-NH2 (b), and Co0.6Fe2.4O4/MIL-101-NH2 (d, e); EDS (e) and elemental mapping (f–j) of Co0.6Fe2.4O4/MIL-101-NH2

Full size image

As displayed in Fig. 4 a, the isotherms of the three materials all show typical type II curves in the range of relative pressure (P•P0−1) 0–1.0, which is a typical feature of mesoporous materials and suggests strong adsorption advantages (Liu et al. 2023). It can be seen from the pore size distribution diagram that the pore sizes of them are mostly distributed between 2 and 10 nm (MIL-101-NH2 and Co0.6Fe2.4O4/MIL-101-NH2) and 5 and 20 nm (Co0.6Fe2.4O4), further confirming the mesoporous structure of the sample. The BET specific surface areas of MIL-101-NH2, Co0.6Fe2.4O4, and Co0.6Fe2.4O4/MIL-101-NH2 are 142.159, 44.796, and 240.485 m2•g−1 respectively. The increase of specific surface area after recombination indicates that the porous structure of the composite components and the uniform dispersion of the components during the compounding process result in a large specific surface area of Co0.6Fe2.4O4/MIL-101-NH2.

Fig. 4
figure 4

Adsorption–desorption isotherms of N2 at 77.35 K, inset: pore size distribution curve (a), VSM curves (b), TGA analysis (c), and zeta potential (d) of Co0.6Fe2.4O4, MIL-101-NH2, and Co0.6Fe2.4O4/MIL-101-NH2

Full size image

The VSM curves of the synthesized samples and their respective parameters are depicted in Fig. 4 b and Table 1. It is discernible that the magnetic parameters, including saturation magnetization strength (Ms), remanent magnetization strength (Mr), and coercivity (Hc) of Co0.6Fe2.4O4, decrease following its combination with MIL-101-NH2. This decline can be attributed to the increase in non-collinear spin structures on the surface induced by the composite, along with the alteration in the distribution of magnetic cations in the surface layer of the magnetic particles due to the MOF coating (Wu et al. 2013). The VSM curve of Co0.6Fe2.4O4/MIL-101-NH2 exhibits noticeable hysteresis phenomena, with relatively small Hc values, and the hysteresis loops become narrower, steeper, and longer, signifying that the synthesized magnetic adsorbents all adhere to the characteristics of typical soft magnetic materials (Kim and Zhao 2022). Remarkably, after undergoing five cycles, the magnetic parameters of Co0.6Fe2.4O4/MIL-101-NH2 displayed only a slight decrease and remained responsive to quick removal by an applied magnetic field within a mere 30 s.

Table 1 Magnetic parameters of Co0.6Fe2.4O4, Co0.6Fe2.4O4/MIL-101-NH2, and Co0.6Fe2.4O4/MIL-101-NH2 (after five adsorption–desorption cycle) determined at room temperature
Full size table

TGA is used to analyze the thermal stability of the adsorbents. Analyzing the weight loss profile of MIL-101-NH2 in Fig. 4 c, it can be divided into five distinct stages: 30–148 °C (9.48%); 148–326 °C (16.12%); 326–520 °C (24.49%); 520–616 °C (10.41%), and 616–1000 °C (16.03%). These five stages correspond to the volatilization of adsorbed moisture and residual DMF in the material, the pyrolysis of the organic groups on the adsorbent, the decomposition of the MOF framework, and the fractional decomposition of the ligand chains, respectively (Zheng et al. 2018). Co0.6Fe2.4O4/MIL-101-NH2 exhibits a weight loss curve akin to MIL-101-NH2. However, the addition of highly stable magnetic particles enhances the thermal stability of Co0.6Fe2.4O4/MIL-101-NH2, causing a substantially lower weight loss rate than that of MIL-101-NH2.

The zeta potential analysis of the three materials is portrayed in Fig. 4 d. Generally, the zeta potential of the adsorbents shifts from positive to negative with increasing pH, with the point of zero charge (pHpzc) measuring at 3.11 (Co0.6Fe2.4O4), 5.85 (Co0.6Fe2.4O4/MIL-101-NH2), and 9.11 (MIL-101-NH2), respectively. The inclusion of Co0.6Fe2.4O4 significantly lowers the pHpzc of MIL-101-NH2. The zeta potential of Co0.6Fe2.4O4/MIL-101-NH2 shows the variation characteristics of the two component materials, which further grounds for the in-situ growth of Co0.6Fe2.4O4 on MIL-101-NH2.

Effect of the dosage of sorbents

To achieve adsorbent concentrations between 0.1 and 1.6 g•L−1, a 50 mL solution of CR with a concentration of 100 mol•L−1 was treated with different dosages of the three adsorbents, ranging from 5 to 80 mg. The objective of this section is to evaluate the adsorption performance of different adsorbents and determine the ideal amount of additional adsorbent that strikes a compromise between high efficiency and affordability. Figure 5 a provides a graphic representation of the results. At an adsorbent dose of 0.1 g•L−1, Co0.6Fe2.4O4/MIL-101-NH2 exhibited a CR removal rate of 60.10%. As the adsorbent dosage increased, the removal rate initially decreased to 47.24% before gradually ascending. The reduction in removal rate in this initial phase may be attributed to limitations in interfacial diffusion resulting from the heightened adsorbent dosage (Weber and Morris 1963). When the adsorbent dose increased from 1.2 to 1.4 g•L−1, the removal rate sharply rose from 72.07 to 98.81%. However, beyond 1.6 g•L−1, the removal rate experienced only marginal improvement, as adsorption approached a state of equilibrium. Consequently, considering both removal rate and economic feasibility, 0.1 g•L−1 was selected as the appropriate adsorbent quantity for subsequent adsorption experiments. As the adsorbent dosage rose, the equilibrium adsorption capacity of all three adsorbents generally decreased. This behavior can be explained by the fact that a significant portion of adsorption sites remain unsaturated even at equilibrium, even when the original dye concentration and volume stay unchanged. Moreover, too much adsorbent has a tendency to aggregate, which prevents dye molecules from diffusing and exposes the site, thereby partially obstructing adsorption (Tee et al. 2022). Additionally, it is worth noting that the maximum adsorption capacity of Co0.6Fe2.4O4/MIL-101-NH2 surpasses that of Co0.6Fe2.4O4 and MIL-101-NH2. However, when the adsorbent quantity exceeds 0.6 g•L−1, the adsorption capacity of MIL-101-NH2 slightly surpasses that of Co0.6Fe2.4O4/MIL-101-NH2. This phenomenon may be attributed to the magnetic properties of the latter, which are more prone to agglomeration and site concealment when the adsorbent quantity becomes excessive.

Fig. 5
figure 5

Effect of adsorbent dosages (a) and pH (b) on CR removal using Co0.6Fe2.4O4/MIL-101-NH2 (inset: compassion of pH before and after CR adsorption; adsorption condition: 298.15 K, 24 h); c the leaching concentration and ratio of Fe in solution with different pH of Co0.6Fe2.4O4/MIL-101-NH2

Full size image

Effect of pH

The pH value of dye solution determines the protonation degree of dye molecules and the charging condition of the adsorbent surface, which is one of the important factors for the adsorption behavior. Hereon the initial pH value of 2.15–10 was selected to explore the effect of pH value on the adsorption. From Fig. 5 b, when pH is 2.15, the composite adsorbent has the highest adsorption capacity for CR. Then with pH increases from 2 to 3, the adsorption capacity drops sharply. At pH is 5, the lowest adsorption capacity emerges, and followed by the overall augment of adsorption capacity with the increase of pH. The pHpzc of Co0.6Fe2.4O4/MIL-101-NH2 is 5.85, which means that if the pH of the solution is higher than 5.85, its surface will be negatively charged; if not, it will be positively charged. The two –NH2 and hydroxyl group (–OH) on the naphthalene ring in the CR are protonated in acidic environments, making it positively charged; in alkaline environments, the CR is present as anions as a result of their deprotonation (Csillag et al. 2023). Also the R-SO3− ionization from the co-color sulfonic acid groups (–SO3H) in CR is responsible for the greatest adsorption capacity under severe acidic circumstances (pH = 2.15), which permits the adsorption of CR on the positively charged Co0.6Fe2.4O4/MIL-101-NH2 via electrostatic attraction. Concurrently, as Fig. 5 c illustrates, 8.16% of Fe is leached from the solution in the strong acid solution (pH = 2), whereas in the weakly acidic and alkaline range, only a trace amount of Fe is leached (< 0.1%). This suggests that the leaching of iron in strong acid can liberate a portion of the adsorption site and encourage the adsorption of CR. Furthermore, The pH of all the solutions following adsorption was less than 5.33, as shown in Fig. 5 b inset. This could be attributed to the interaction between the OH– in the solution and the acid –COOH or –NH2, which is acidic because of the presence of –SO3H, which have a strong ability to absorb electrons on the surface of Co0.6Fe2.4O4/MIL-101-NH2 (Jaafar et al. 2007).

Adsorption kinetics

The pseudo-first-order kinetic model (PFO), PSO, and Elovich model (Eq. 3–5) were used to fit the correlation curve between adsorption capacity and adsorption time to evaluate the adsorption rate of the adsorbent and infer the time required to reach adsorption equilibrium. The fitting curve and related parameters are shown in Fig. 6 a–c and Table 2. For different concentrations of CR solutions, PSO has the best fitting effect, with correlation coefficients (R2) closest to 1, which means that the adsorption process is mainly a chemical adsorption process (Hoang et al. 2019). And it is more rate limited by chemisorption, which is consistent with the previously reported adsorption behavior of MIL-101 on dye molecules. Equipped relatively fast adsorption rate, Co0.6Fe2.4O4/MIL-101-NH2’s removal rate to CR can reach 90% in the first 12 h, and also almost attained adsorption equilibrium within 18 h. Additionally, the theoretical values of equilibrium adsorption capacity at different concentrations obtained by PSO are also closer to the experimental results, so the PSO model has high applicability. The intraparticle diffusion (ID) model (Eq. 6) was applied to analyze the diffusion process of CR molecules in the adsorbent. It is found that the adsorption process can be divided into three consecutive stages, and the fitting lines of each part do not pass the origin point (Fig. 6 d), so that ID is not the main rate-limiting step (Aly et al. 2014). The three stages are the diffusion of CR to the surface of the adsorbent, the intraparticle diffusion of CR into the inner pores of the adsorbent, and the final equilibrium stage (Xiong et al. 2021). The rate constants of the three stages show a decreasing trend, since with the increase of the adsorption time, the adsorption sites gained easy access are gradually occupied by CR molecules until the adsorption equilibrium (Wang and Wang 2008). The higher CR concentration presents to the larger rate constant of each stage.

$${Q}_{t} = {Q}_{e , 1} \left[1-exp\left(-{k}_{1}t\right)\right]$$
(3)
$${Q}_{t} = \frac{{k}_{2}{Q}_{e, 2}^{2}t}{1+{k}_{2}{Q}_{e,2}t}$$
(4)
$${Q}_{t} = \beta ln\left({k}_{e} \beta t\right)$$
(5)
$${Q}_{t} = {k}_{x} \sqrt{t}$$
(6)

where t (min) is the adsorption time, k1 (min−1), k2 (g•mg−1•min−1), ke (g•mg−1•min−1), and kx (mg•min−1/2•g−1) are the rate constant of PFO, PSO, Elovich, and ID kinetics, respectively; \({Q}_{e , 1}\) and \({Q}_{e 2}\) are the equilibrium adsorption capacities determined by fitting of models; \(\beta\)(g•mg−1) is the parameters related to activation energy of chemisorption.

Fig. 6
figure 6

The non-linearized kinetic models (PFO, PSO, and Evolich) of the adsorption of CR with different concentrations at 25 °C: a 100 mg•L−1, b 500 mg•L−1, and c 1000 mg•L−1; d the linearized ID kinetic equation of CR adsorption with three concentrations at 25 °C

Full size image
Table 2 Kinetic parameters for adsorption of CR onto Co0.6Fe2.4O4/MIL-101-NH2
Full size table

Adsorption isotherms

Four isotherm models (Eqs. 7–10) of Langmuir, Freundlich, Temkin, and Dubinin-Radushkevich (D-R) were used to fit the results of \({C}_{e}\) and \({Q}_{e}\) obtained under different concentrations of CR to analyze the adsorption isotherm of Co0.6Fe2.4O4/MIL-101-NH2. The Fig. 7 a–d describe the non-linear fitting results of the adsorption isotherms of CR by Co0.6Fe2.4O4/MIL-101-NH2 at 293.15, 298.15, 303.15, and 308.15 K, and the relevant parameters are shown in Table 3. The adsorption isotherm data of CR on Co0.6Fe2.4O4/MIL-101-NH2 fitted well with the four isotherm equations (R2 = 0.85–0.99), and among them Langmuir thanks to the highest fitting degree (R2 is closer to 1) is more suitable for describing the adsorption equilibrium of CR on Co0.6Fe2.4O4/MIL-101-NH2. The isotherms’ fitting results manifest that the adsorption of CR on Co0.6Fe2.4O4/MIL-101-NH2 is monolayer adsorption and the surface chemisorption occupied center stage (Valadi et al. 2020). Meanwhile, the maximum theoretical equilibrium adsorption capacity of Co0.6Fe2.4O4/MIL-101-NH2 for CR is 1756.19 mg•g−1 at 298.15 K, which reflects the excellent adsorption performance of Co0.6Fe2.4O4/MIL-101-NH2. All the RL calculated (Eq. 11) within the range of 0–1, and the higher the CR concentration with the lower the RL value (Fig. 7 e), i.e., the adsorption process is favorable at the four temperatures, and the easiness of adsorption increased with the increase of CR concentration. The higher temperature has the lower RL, so the high temperature is conducive to this adsorption (Xiong et al. 2021). The \({E}_{a}\) values obtained by the D-R model (Eq. 12) are 42.27, 27.44, 50.69, and 77.43 kJ•mol−1, respectively, all exceeding 16 kJ•mol−1, illustrating that the adsorption of CR on the composite is mainly chemical adsorption, which is consistent with the fitting results of the kinetic model.

$${Q}_{e} = \frac{{C}_{e} {Q}_{max, 1} {K}_{L}}{1+{C}_{e }{K}_{L}}$$
(7)
$${Q}_{e} = {K}_{F} {C}_{e}^\frac{1}{n}$$
(8)
$${Q}_{e} = \frac{RT}{{K}_{T}} ln\left(A {C}_{e}\right)$$
(9)
$${Q}_{e} = {Q}_{max, 2} exp\left\{-{K}_{D} {\left[R T ln\left(1+\frac{1}{{C}_{e}}\right)\right]}^{2}\right\}$$
(10)
$${R}_{L} = \frac{1}{1+{C}_{0 }{K}_{L}}$$
(11)
$${E}_{a}=\frac{1}{\sqrt{2 {K}_{D}}}$$
(12)

where Ce (mg•L−1) is the equilibrium concentration of CR solution; \({Q}_{e}\) (mg•g−1) is the equilibrium adsorption capacity; \({Q}_{max, 1}\) and \({Q}_{max, 2}\) (mg•g−1) are the maximum adsorption capacities calculated by Langmuir and D-R model, respectively; \({K}_{L}\) (L•mg−1), \({K}_{F}\) (mg•(1−1/n) •L1/n•g−1), \({K}_{T}\) (J•mol−1), and \({K}_{D}\) (mol2•kJ−2) are the adsorption coefficients of Langmuir, Freundlich, Temkin, and D-R isotherm, respectively; n is the Freundlich constant connected with adsorption intensity; R is molar gas constant and its value is 8.314 J•mol−1•K−1; A (L•mg−1) is adsorption equilibrium binding constant; RL is a constant relevant to the affinity between adsorbent and adsorbate; and \({E}_{a}\) (kJ•mol−1) is free energy of adsorption.

Fig. 7
figure 7

Four non-linearized isotherms models (Langmuir, Freundlich, Temkin, and D-R) of the adsorption of CR at different temperatures from 293.15 to 308.15 K (a–d); the RL values of different initial concentrations at different temperatures from 293.15 to 308.15 K (e); the linearized thermodynamic curve of CR adsorption on Co0.6Fe2.4O4/MIL-101-NH2

Full size image
Table 3 Four isotherm parameters for CR adsorption on Co0.6Fe2.4O4/MIL-101-NH2
Full size table

Adsorption thermodynamics

The adsorption behavior was studied using thermodynamics, and Gibbs free energy (ΔG0), entropy change (ΔH0), and enthalpy change (ΔS0) were gained through Eqs. (13)–(15) calculation and fitting. These aimed at determining the energy change, the order or disorder of the reaction, and the ability to proceed spontaneously of the adsorption process.

$$K={K}_{W} {K}_{L}$$
(13)
$${\Delta G}^{0}=-RT{\text{ln}}K$$
(14)
$${\Delta G}^{0}={\Delta H}^{0}-T{\Delta S}^{0}$$
(15)

where Kw (mg•mol−1) is the molar mass of CR.

The fitting results and the parameters are shown in Table 4 and Fig. 7 f and the equation fit better reaching R2 = 0.9847. The ΔG0 values for the adsorption CR using Co0.6Fe2.4O4/MIL-101-NH2 are − 22.04, − 22.47, − 23.30, and − 24.06 kJ•mol−1 at 288.15–308.15 K, respectively. All the negative ΔG0 values manifest that the adsorption process can proceed spontaneously (Munagapati and Kim 2017; Valadi et al. 2020). The effect of the transformation of functional groups on the surface chemistry of the adsorbent at high temperature causes the ΔG0 to decrease with increasing temperature, signifying that high temperature can promote the adsorption. The positive ΔH0 and ΔS0 values of Co0.6Fe2.4O4/MIL-101-NH2 manifest that the removal of CR is exothermic and that as the adsorption proceeds (Mazloomi et al. 2019). The CR and Co0.6Fe2.4O4/MIL-101-NH2 interface with an increased probability of random collisions between them may come down to the desorption of water molecules from Co0.6Fe2.4O4/MIL-101-NH2’s surface (Al Sharabati and Sabouni 2020; Gökmen and Serpen 2002). Hence, the adsorption of CR on Co0.6Fe2.4O4/MIL-101-NH2 is an endothermic, spontaneous, and orderly course (Tan et al. 2019).

Table 4 Thermodynamic parameters for CR adsorption on Co0.6Fe2.4O4/MIL-101-NH2
Full size table

Effect of co-existing ions

There are usually various anions and cations in the actual wastewater, which may have a certain effect on the adsorption of CR. The 0.1 M ionic strength for Na+, K+, Ca2+, Mg2+, Cl−, NO3−, CO32−, and SO42− were selected for batch adsorption experiments to investigate the interference of these ions. Other experimental conditions were kept consistent, and a blank group was set for control. From Fig. 8 a, the presence of Na+ and K+ somewhat suppressed the adsorption of CR, and the adsorption capacity decreased by 21.86% and 9.08%, respectively. Inversely, the equilibrium adsorption capacity of CR increased significantly when two divalent cations, Ca2+ and Mg2+, were present in the solution. These probably because Ca2+ and Mg2+ could form an intramolecular bridge with the –COO− on Co0.6Fe2.4O4/MIL-101-NH2 for better recognizing the negatively charged CR, while the divalent cations could shield off the negative charge on the surface of the adsorbent (Vermohlen et al. 2000). The higher adsorption capacity was achieved through the enhancement of electrostatic adsorption. For the anions (Fig. 8 b), the reduction of adsorption capacity comes down to their surface competition on the sorbent. Among them, CO32− has the strongest interfering ability, probably since its greatly change to the pH of the solution and then alter the adsorption (Ding et al. 2020).

Fig. 8
figure 8

Effect of different cations (a) and anions (b) on CR removal using Co0.6Fe2.4O4/MIL-101-NH2 at the equilibrium time; the reusability of Co0.6Fe2.4O4/MIL-101-NH2 for CR (c); the FT-IR spectra of Co0.6Fe2.4O4/MIL-101-NH2 before and after adsorption (d)

Full size image

Reusability study

Eluents of 0.1 M NaOH and EtOH were used to pretreat the adsorbent, respectively. And it was discovered that the latter, when used as an eluent, had a greater regeneration effect. In contrast, 0.1 M NaOH had a larger CR elution impact, but the adsorbent sites were adversely affected by its alkalinity, which decreased the regenerated Co0.6Fe2.4O4/MIL-101-NH2’s adsorption performance. Therefore, in order to assess the reusability of Co0.6Fe2.4O4/MIL-101-NH2, EtOH was used as the eluent for five adsorption–desorption cycles. Figure 8 c presents the findings. It is evident that as the number of cycles increases, the adsorption performance declines, which is brought on by the uneluted CR molecules taking up some space on the site. Co0.6Fe2.4O4/MIL-101-NH2 continues to have a high capacity for adsorbing CR after five cycles, and the removal rate is around 78% of the original adsorption. Co0.6Fe2.4O4/MIL-101-NH2 has strong cycle reusability for removing CR from water, to put it briefly.

Adsorption mechanisms

In line with the electrostatic interaction theory (Gilson 1995), when the solution is barely neutral, the CR molecules can be attracted by electrostatic force to achieve the adsorption of CR because they have no charge on their surface and the positive charge for Co0.6Fe2.4O4/MIL-101-NH2. Even in the face of electrostatic repulsion, the adsorbent retains a high adsorption capacity for CR in strongly acidic and basic conditions, according to the results of the pH effect on adsorption. Stated differently, the adsorption process of CR on Co0.6Fe2.4O4/MIL-101-NH2 is a product of many processes and is impacted by mechanisms other than electrostatic interactions, such as hydrogen bonding.

The increased adsorption capacity of the adsorbent on CR under acidic and basic conditions can be explained by the –NH2 and –OH in Co0.6Fe2.4O4/MIL-101-NH2 and CR serving as H-donors or acceptors to enhance adsorption through hydrogen bonding. The direction of hydrogen bonding can also be judged by the effect of pH to adsorption (Stockton and Rubner 1997). At low pH levels, if the adsorbent’s –NH2 functions as an H-donor, it is likely to get protonated, which facilitates the uptake of protons by groups like –SO3H on CR and encourages the creation of hydrogen bonds, which in turn increases the adsorption capacity. This is consistent with the result that the adsorption capacity of CR increases with acidity so the hydrogen bonding is formed by Co0.6Fe2.4O4/MIL-101-NH2 as the donor and the CR molecule as the acceptor; at high pH, adsorption capacity will decrease if Co0.6Fe2.4O4/MIL-101-NH2’s –NH2 acts as the H-donor and readily accepts OH– in the solution; however, in alkaline conditions, adsorption performance improves, suggesting that the CR molecule is the H-donor and the adsorbent is the H-acceptor in such a scenario (Chen et al. 2022).

Figure 8 d displays the comparison of FT-IR spectra of Co0.6Fe2.4O4/MIL-101-NH2 before and after adsorption. A new peak at 1044 cm−1 corresponding to the C–N stretching vibration on CR is clearly observed after adsorption (Banerjee et al. 2016). The characteristic peaks at 585, 769, 1431, 1492, and 1617 cm−1 all show some shifts in the adsorption of CR. These mean that CR is adsorbed on the adsorbent. The peak of C = C changes from 1492 to 1497 cm−1, which represents that there is π-π interaction between CR and adsorbent (Yuan et al. 2015). And the peak of N–H offsets from 1617 to 1624 cm−1, which further confirmed the hydrogen bonding interaction between Co0.6Fe2.4O4/MIL-101-NH2 and CR (Kazim et al. 2020).

The adsorption mechanism of CR on Co0.6Fe2.4O4/MIL-101-NH2 was further verified by XPS characterization. The full spectrum before adsorption is shown in Fig. 9 a, from which the spectral peaks of five elements, C1s, O1s, N1s, Fe 2p, and Co 2p, can be clearly observed. This further supports the distribution and attachment of Co0.6Fe2.4O4 on the surface of MIL-101-NH2 from the side. The fine spectra of four elements, C, O, Fe, and Co, are shown in Fig. 9 b–e. Through the peak fitting of C 1 s, three peaks of 288.50, 286.50, and 284.80 eV can be obtained, which correspond to O–C = O (C1), O–C–O/C–N (C2), and C–C/C = C (C3), respectively (Wang et al. 2023). O1s is deconvolved to form C = O at 534.65 eV (COO– of MIL-101-NH2), C–O at 531.89 eV (metal carboxylate formed by coordination of central metal Fe3+ of MIL-101-NH2 with carboxylic acid ligand), and Co–O/Fe–O at 529.69 eV (Co0.6Fe2.4O4 in the component). The double peaks and satellite peaks of Fe3+ and Fe2+ can be fitted from the deconvolution peak of Fe 2p, which manifests that Fe exists in two valence states (II and III) in Co0.6Fe2.4O4/MIL-101-NH2 (Biesinger et al. 2011). The peaks of Co 2p at 802.81, 795.58, 789.55, 784.43, and 780.38 eV are consistent with the XPS peak of CoO and Co3O4, indicating that Co exists in the form of divalent and tervalent in Co0.6Fe2.4O4. After the adsorption of CR, the peak of S 2p appeared on the adsorbent, indicating that CR was adsorbed successfully. And it is found that the peak area ratio of C2 to C3 decreased from 0.20 to 0.14, revealing that the π-π interaction between CR and the adsorbent is weak, which is not the dominant mechanism of CR adsorption on Co0.6Fe2.4O4/MIL-101-NH2 (Yang et al. 2006). The decrease of the binding energy of the O 1 s peak near 531.89 eV and the increase of the peak area may manifest that some new chemical groups or surface defect sites interacting with CR have been formed on the surface of the adsorbent. The change of the binding energy of each peak in O1s indicates that some new chemical bonds or surface defect sites interacting with CR may be formed on the surface of the adsorbent, indicating that the hydrogen bond may be one of the dominant mechanisms. The peaks of Fe 2p and Co 2p shift to a certain extent after the adsorption of CR, suggesting that Fe and Co are the adsorption sites of the adsorbents. The nitrogen-containing groups on CR have the ability to establish coordination bonds with Fe and Co during the adsorption process, which will allow CR molecules to partially transfer the charge to the iron and cobalt atoms on Co0.6Fe2.4O4/MIL-101-NH2’s surface (Lei et al. 2020). Concurrently, the adsorption process modifies the charge state of metal species, resulting in a shift in peak area. This phenomenon is associated with the emergence of interacting metal species on the Co0.6Fe2.4O4/MIL-101-NH2’s surface. In summary, the combined effects of π-π interactions, electrostatic interactions, hydrogen bonding, and metal coordination lead to the adsorption of CR on Co0.6Fe2.4O4/MIL-101-NH2 (Fig. 10).

Fig. 9
figure 9

XPS spectra of the Co0.6Fe2.4O4/MIL-101-NH2 sorbent before and after CR adsorption: a full-scan spectra (inset: P 2p), b C 1 s, c O 1 s, d Fe 2p, and e Co 2p

Full size image
Fig. 10
figure 10

The main adsorption mechanism of CR on the Co0.6Fe2.4O4/MIL-101-NH2

Full size image

Comparison of adsorption capacity

The adsorption capacity, adsorbent dose, temperature, and adsorption equilibrium time in this work were compared with those of previously published materials for the removal of CR, as indicated in Table 5, to illustrate the superior performance of Co0.6Fe2.4O4/MIL-101-NH2. As can be observed, Co0.6Fe2.4O4/MIL-101-NH2 has a far better adsorption capacity than the majority of other publications with comparable data. Co0.6Fe2.4O4/MIL-101-NH2 only needs 0.1 g•L−1 of adsorbents to successfully remove CR, whereas some documented systems can reach high adsorption capacity only by adding more adsorbents. The primary drawback of our work is the comparatively high number of preparation steps, which will partially raise the material’s preparation costs. However, the benefit is that Co0.6Fe2.4O4/MIL-101-NH2 we created has excellent thermal stability, high recovery and reuse performance, strong selectivity, and a significant adsorption capacity for CR. Adsorbent costs can be significantly decreased. In summary, the Co0.6Fe2.4O4/MIL-101-NH2 developed in this work offers benefits for CR removal.

Table 5 Qmax of some reported adsorbents for CR
Full size table

Selectivity study

MB and MO were chosen as representative cationic and anionic dyes, respectively, to probe the adsorption selectivity of Co0.6Fe2.4O4/MIL-101-NH2. The before-and-after UV and physical diagrams of the adsorption of Co0.6Fe2.4O4/MIL-101-NH2 to the three dyes under the same conditions reveal that Co0.6Fe2.4O4/MIL-101-NH2 has the largest adsorption capacity for CR, followed by MO, which is also an anionic dye, and MB has the smallest adsorption capacity (Fig. 11 a–d). The better adsorption capacity of Co0.6Fe2.4O4/MIL-101-NH2 for the anionic dyes further indicates the contribution of electrostatic interaction in the adsorption. Co0.6Fe2.4O4/MIL-101-NH2 is generally good for selective adsorption of CR in the presence of competing dye molecules. In the binary mixture of dyes, the preferential adsorption of Co0.6Fe2.4O4/MIL-101-NH2 on CR is evident from the UV spectral peaks and the color change of the solution before and after adsorption. But the specificity adsorption capacity is relatively poor in the binary system with the same anionic dyes, probably because of the stronger competitiveness between CR and MO, resulting in the preferential occupation of some sites by MO (Xiong et al. 2021).

Fig. 11
figure 11

a Comparison of adsorption capacity of Co0.6Fe2.4O4/MIL-101-NH2 to 100 mg•L−1 CR, 100 mg•L−1 MB, and 100 mg•L−1 MO; the ultraviolet absorption scanning spectrum actual photographs and of adsorption by Co0.6Fe2.4O4/MIL-101-NH2: b CR, c MO, d MB single solution, and e 50 mg•L−1 CR/10 mg•L−1 MB, f 50 mg•L−1 CR/50 mg•L−1 MO mixture solution

Full size image

Conclusion

The magnetically separable adsorbent Co0.6Fe2.4O4/MIL-101-NH2 was synthesized and successfully applied to CR treatment. With a little amount of adsorbent (0.1 g•mL−1), an ultra-high adsorption capacity (1756.19 mg•g−1) can be attained. The produced adsorbent displays outstanding adsorption performance in a wide pH range (pH = 2.15–10) and high concentration (1000 mg•L−1). The results of isotherm fitting and kinetics show that chemical adsorption predominates when CR is adsorbed by Co0.6Fe2.4O4/MIL-101-NH2, and higher temperatures promote adsorption. The adsorption process is spontaneous, endothermic, and ordered, according to thermodynamic study. Other ions, with the exception of Ca2+ and Mg2+, will limit the adsorbent’s ability to adsorb CR, and their inhibitory effect on CR removal will go as follows: CO32−  > Na+–Cl−  > NO3−  > K+  > SO42−. It is discovered that hydrogen bonding, metal coordination, electrostatic π-π interaction, and electrostatic interaction are all involved in the adsorption mechanism of CR on Co0.6Fe2.4O4/MIL-101-NH2. In CR/MB and CR/MO binary systems, Co0.6Fe2.4O4/MIL-101-NH2 can also readily and selectively separate CR. In addition, the adsorbent is highly cycle-reusable and retains a high adsorption capacity even after five cycles. The adsorbent is an effective and versatile adsorbent with a broad range of applications for eliminating CR dye from wastewater, as demonstrated by all of the experimental data. Finding appropriate substrates to enhance the adsorbent’s mechanical capabilities could be the next topic of attention for research, given that the adsorbent’s powder form restricts its processing possibilities.