TiO <subscript>2</subscript> /CTS/ATP adsorbent modification and its application in adsorption-ultrafiltration process for dye wastewater purification.

Industrial dyeing produces highly polluting wastewater and threatens the environment. Effective treatment of dyeing wastewater is a crucial step to prevent toxic chemicals from entering receiving waters. This study aimed to assess a modified attapulgite (ATP)-based adsorbent for dyeing wastewater purification by introducing chitosan (CTS) and titanium dioxide (TiO₂) into the adsorbent material named TiO₂/CTS/ATP. It was found that after modification, the adsorbent showed a lower hydrophilicity, as demonstrated by an increase in the water contact angle from 9.1° to 42°, which could reduce the water adsorption tendency and potentially facilitate contaminants adherence. The modification also led to a significantly increased specific surface area of 79.111 m²/g from 3.791 m²/g and exhibited more uniform and smaller particle size (reduced from 3.99 to 2.52 μm). When the TiO₂/CTS/ATP adsorbent was applied to the adsorption of Congo red solution, the adsorption efficiency was observed to reach to 97.6% at the dosage of 0.5 g/L. Furthermore, the combination of adsorption and ultrafiltration was able to achieve 99% Congo red removal. Adsorption pretreatment prior to the ultrafiltration also enabled to reduce membrane fouling, increased the reversible membrane fouling, and resulted in a considerably lower transmembrane pressure as compared with the direct ultrafiltration filtration system.

Keywords: TiO2/CTS/ATP adsorbent; Ultrafiltration; Membrane fouling; Dyeing wastewater; Modified attapulgite

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s11356-021-13933-3.

Over the past few decades, rapid population growth and accelerated globalization have led to a marked upsurge in the production of drugs, food, cosmetics, textiles, etc. (Guerranti et al. [15]; Wei et al. [38]). The discharge of untreated wastewater containing these products into surface waters is known to severely damage the ecological environment (Samsami et al. [31]). As one of the highly polluting industrial effluents, dye wastewater features a high chromatic and toxic level, low biodegradability, and significant mutagenicity and carcinogenicity (Selvaraj et al. [32]). It has been reported that more than 7 × 105 tons of dyes are produced each year and about 10% originated from textile and related industries (Shahanaz et al. [34]).Therefore, it is important to develop reliable technologies for remediating textile dying wastewater. To date, multiple treatment options, such as membrane filtration (Abdel-Karim et al. [1]), electrochemical method (Nidheesh et al. [24]), biological process (Paz et al. [28]), photocatalysis (Wang et al. [37]), and adsorption (Xiaoduo et al. [40]), have been studied. Among these methods, the adsorption method has been widely considered for dying wastewater treatment due to low cost, high pollutant removal efficiency, uncomplicated design, and operation (Abdulhameed et al. [3]; Jawad et al. [18]).

Various adsorbents have been used to eliminate dyes from water (Hassan et al. [17]). Attapulgite (ATP), a typical low-cost adsorbent material with abundant reserves on the Earth, has attracted increased interest (Pan et al. [27]; Wang et al. [36]). ATP is a mineral with unique layered chain structure, mainly composed of hydrous magnesium aluminum silicate, and is widely used as an adsorbent in wastewater treatment due to its considerably porous structure, high specific surface area, and excellent physico-chemical properties (Zhang et al. [42]). However, impurities are commonly encountered in natural ATP, which increase the tendency of pore blockage, decrease the specific surface area, and thus negatively affect the adsorption performance (Wang et al. [36]). Therefore, modifications are often critical in order to achieve enhanced water treatment efficiency. For instance, chitosan@ATP composite was prepared by a self-assembly method, and this modified ATP adsorbent greatly improved the removal efficiency of uranium pollutants (Pan et al. [27]). The unique structure of CS is attributed to the presence of reactive amino(–NH2) and hydroxyl(–OH) groups which are considered active adsorption sites for unlimited types of organic and inorganic water pollutants (Abdulhameed et al. [2]; Abdulhameed et al. [3]). In another study, Zhou et al. (Zhou et al. [44]) adopted a hydrothermal carbonization method to synthesize a chitosan@ATP composite and successfully used it to remove methylene blue from wastewater.

Additionally, titanium dioxide (TiO2) is commonly considered in adsorption processes and is often combined with clay for the treatment of dye wastewater (Djellabi et al. [11]). This combination can not only enhance the light absorption capacity of TiO2 but also increase the specific surface area and the dye adsorption capacity (Setthaya et al. [33]). Previous research has demonstrated that kaolin/TiO2 (Wongso et al. [39]), montmorillonite/TiO2 (Djellabi et al. [11]), and zeolite/TiO2 (Setthaya et al. [33]) have the ability to effectively adsorb dye from wastewater. Combining chitosan with TiO2 is also beneficial; the composite can not only reduce the risk of agglomeration but also expand the optical reaction range of TiO2 from the ultraviolet region to the visible region, thus enabling TiO2 to play a better role in water treatment (Karthikeyan et al. [19]).

Solid-liquid separation is an important process after adsorption, and it affects numerous aspects of the treatment process, such as operating cost and recycling of adsorbents and adsorbates. Among all the separation techniques, ultrafiltration (UF) is often a preferred option due to its high solid removal efficiency, operational simplicity, and low energy consumption (Chen et al. [7]), and ultrafiltration is widely used in the treatment of dye (Lin et al. [21]). Nevertheless, wastewater treatment using a standalone ultrafiltration is usually not able to achieve the removal of some low molecular weight dissolved pollutants, and severe fouling is likely to occur (Chen et al. [7]). Combining ultrafiltration and an adsorption pretreatment process can effectively solve these problems (Hammami et al. [16]). For example, during dyeing wastewater treatment, adding an adsorbent in an ultrafiltration unit is expected to enhance the removal of some pigment molecules that are not rejected by the membrane. In addition, the adsorbent retained on the membrane surface can form a filtrate cake to prevent foulants adhering to the membrane (Dong et al. [12]; Hammami et al. [16]).

In this study, an adsorbent material is prepared via modifying ATP with TiO2 and chitosan and is termed TiO2/CTS/ATP. The adsorbent is subjected to comprehensive characterization prior to being applied to remove Congo red. Then, the TiO2/CTS/ATP adsorbent is combined with the ultrafiltration membrane to treat a synthetic dye wastewater. Both treatment efficiency and membrane performance will be examined to demonstrate the benefits of combining the TiO2/CTS/ATP adsorption coupled with ultrafiltration process for dyeing wastewater treatment

Attapulgite (ATP) was purchased from the Xuyi Botu Attapulgite Co. Ltd., Jiangsu, China. Terabutyl titanate (C16H36O4Ti2), nitric acid (HNO3), and acetic acid (CH3COOH) were all obtained from Honghui Reagent Co. Ltd., Hunan, China. Chitosan (C6H11NO4X2), hydrochloric acid (HCl), and absolute ethanol (C2H5OH) were provided by Qingdao BZ Oligo Biotech Co. Ltd., China. Congo red (C32H22N6Na2O6S2) dye was acquired from Kemi Reagent Co. Ltd., Tianjin, China. All the chemicals used were of analytical grade and without further treatment. Deionized water was used throughout the experimental work to prepare solutions and synthetic wastewater.

Congo red solution was prepared by dissolving Congo red into deionized water, and the concentration was 100 mg/L. The selected content of Congo red solution was related to the content of dye wastewater (Zhang et al. [43]). A synthetic dyeing wastewater was produced according to the characteristics of the secondary effluent at one printing and dyeing plant in Guangzhou. The water quality indexes of the actual dyeing wastewater are shown in Table 1. It can be seen from the table that the COD concentration of dyeing wastewater after biological treatment was still higher than first class effluent standard of China's "Discharge Standard of Water Pollutants for Dyeing and Finishing of Textile Industry" (GB4287-1992).

Table 1 The characteristic of dyeing wastewater

As shown in Fig. 1, 100-mL HCl at a concentration of 3 mol/L was used to treat 25-g ATP. The mixture was shaken for 3 h at a constant temperature of 25 °C and then centrifuged, washed, and dried for 12 h.

Graph: Fig. 1 Material preparation

As shown in Fig. 1, CTS of different masses (0.08 g, 0.16 g, 0.32 g) were dissolved into 100-mL CH3COOH (2.0%), and 16-g acidified ATP was then added into the solution. Subsequently, the mixture was stirred for 30 min, dried, and grinded to obtain 0.5%, 1.0%, and 2.0% (w/w) CTS/ATP.

As shown in Fig. 1, 10-mL terabutyl titanate was added slowly into 30-mL absolute ethanol. Then, the solution was subjected to magnetic stirring and 60 °C water bath for 15 min. During the process, 2-mL CH3COOH was added to control the hydrolytic rate. The solution was then mixed with 8-g acidified ATP or 8-g CTS/ATP (1.0%) and a buffer solution composed of 0.5-mL HNO3, 10-mL ethanol, as well as 5-mL distilled water. The mixture was stirred for 30 min and then dried and grinded to get TiO2/ATP or TiO2/CTS/ATP powder.

As shown in Fig. 1, 0.08-g CTS was dissolved in 100-mL CH3COOH (2%), and then 8-g TiO2/ATP was added. Subsequently, the mixture was stirred for 30 min, dried, and grinded to obtain CTS/TiO2/ATP.

A 0.05-g acidified ATP, 0.5% CTS/ATP, 1.0% CTS/ATP, 2.0% CTS/ATP, TiO2/ATP, TiO2/CTS/ATP, and CTS/TiO2/ATP were added into 7 beakers containing 100-mL Congo red solution (100 mg/L) with a pH of 7 at 25 °C, respectively. The experiments were run for 60 min on a plate shaker at a constant shaking speed of 200 r/min. After the experiments, supernatant samples were collected using centrifugation and then analyzed in a by ultraviolet spectrophotometer at a wavelength of 498 nm (see "Analytical methods" section for details).

In this experiment, certain amounts (0.01, 0.03, 0.05, 0.07, and 0.1 g) of TiO2/CTS/ATP were added in 5 beakers with 100-mL Congo red solution (100 mg/L); the experimental condition and operation are the same as the "Adsorption experiments with different types of modified ATP as variable" section.

The experiment was carried out by adding 0.05 g TiO2/CTS/ATP into 9 beakers containing 100-mL Congo red solution (100 mg/L) with the adjustment of pH to 3, 4, 5, 6, 7, 8, 9, 10, and 11, respectively. The pH of the raw water was adjusted to the initial target pH by the addition of dilute 0.1-M HCl or NaOH as necessary. The experimental conditions were the same as above.

The flat membrane ultrafiltration rig was mainly composed of an ultrafiltration vessel, a circular plate membrane, peristaltic pump, and pressure acquisition and control device (Fig. 2). The disc flat sheet membrane (ϕ76mm, PVDF), with the nominal molecular weight cut-off value of 100 kDa, was placed at the bottom of the 200-mL ultrafiltration vessel. The pressure acquisition device and peristaltic pump were connected at the outlet of the ultrafiltration vessel. In this experiment, a constant flux (J = 120 L·m-2·h-1) of dead-end filtering was adopted, and the pressure acquisition device automatically recorded the data in a computer every 10 s. The transmembrane pressure data recorded could reflect the fouling of ultrafiltration membrane under the constant flux condition.

Graph: Fig. 2. Experimental device diagram of adsortion and ultrafiltration

The combined process started with treating the synthetic wastewater with TiO2/CTS/ATP (1.5 g/L) adsorption for 1 h at a shaking speed of 200 r/min. The mixture was fed into the ultrafiltration vessel at a constant flux of J = 120 L·m-2·h-1, and the filtration process continuously operated for 30 min (Fig. 2).

Membrane fouling calculation is calculated using the model of resistance-in-series as shown in Eq. (1) (Cheng et al. [8]):

Graph

where TMP is the transmembrane pressure (Pa); m is the dynamic viscosity (Pa·s); J is the permeate flux (m/s); and Rt, Rm, Rir, and Rr are the total hydraulic fouling resistance, intrinsic membrane resistance, hydraulic irreversible fouling resistance, and hydraulic reversible fouling resistance (m-1), respectively.

pH was measured using a Starter 3100 pH meter. The level of Congo red in the supernatant was determined by measuring the ultraviolet absorbance at 498 nm wavelength in an ultraviolet spectrophotometer using distilled water as reference. The concentration and removal of Congo red are then calculated using Eq. (2) and Eq. (3). In Eq. (2), A is the ultraviolet absorbance data, and C denotes the concentration of Congo red, In Eq. (3), C0 denotes the initial concentration of Congo red, and Ce denotes the final concentration of Congo red.

2 $$A=0.0201C+0.0308$$

Graph

3 $$R=\frac{\left(C_0-C_e\right)}{C_e}\times 100\%$$

Graph

Contact angle measuring instrument (OCA15Pro) was used to analysis the wettability of the adsorbents. Distilled water was used as the reagent, and the contact angle was measured by the hanging drop method. For particle size analysis, a small amount of adsorbent was initially subjected to ultrasonic dispersion in ultra-pure water for 20 min, and a portion of the upper liquid was then transferred to laser particle size analyzer (Mastersizer2000, Malvern Instruments Ltd., UK) for analysis. Surface Area and Porosity Analyzer (ASAP2460) was used to measure the specific surface area of adsorbent samples. Before testing, the sample was vacuumed at 100 °C for about 4 h. N2 was used as adsorption molecules for the tests.

It can be seen from Fig. 3 a, b and c that modifying ATP achieved considerably enhanced Congo red removal. Among 0.5%, 1.0%, and 2.0% CTS/ATP, the 1.0% had the best Congo red adsorption efficiency (Fig. 3a). The lower Congo red removal at the lower CTS concentration (0.5%) was likely attributed to the lower availability of adsorption-related functional groups, such as amino and hydroxyl groups (Abdulhameed et al. [2]; Abdulhameed et al. [3]). On the other hand, the higher CTS proportion (2%) could lead to increased viscosity of solution, inhibiting the uniform dispersion of solid particles and resulting in the agglomeration of CTS/ATP which reduced the absorption efficiency of Congo red. Therefore, CTS with a mass fraction of 1.0% was used in the later trials. Figure 3b shows the importance of the sequence of adding CTS and TiO2 when preparing the adsorbents, and it was found that TiO2/CTS/ATP (for which CTS was added before TiO2) achieved better Congo red removal than CTS/ TiO2/ATP (for which TiO2 was added before CTS). It was likely that CTS was positively charged after protonation and it performed ion exchange with ATP, resulting in the inner-sphere complexes (Pan et al. [27]). As a result, the interval between layers increased, so more TiO2 particles could be absorbed. However, in the case of TiO2 being initially loaded, agglomeration occurred to a greater extent, limiting the adsorption and adhesion of CTS (Salameh et al. [30]). According to Fig. 3c, ATP modified with both CTS and TiO2 was significantly more effective in Congo red removal in comparison with the ATP modified by CTS or TiO2 alone.

Graph: Fig. 3 Comparison of adsorption properties under a different chitosan contents, b different load sequences of CTS and TiO2, and c varied modification conditions.

The N2 adsorption-desorption isotherms of ATP and TiO2/CTS/ATP are shown in Fig. 4a. According to the IUPAC classification, the adsorption process was a type II isotherm, reflecting a typical multilayer adsorption occurring on non-porous or macro porous adsorbent as the adsorption equilibrium was reached (Nodehi et al. [26]), the amount of gas adsorbed by TiO2/CTS/ATP was significantly higher than that of ATP. According to the BET equation (Peng et al. [29]), modification could increase the surface area, and the specific surface area of TiO2/CTS/ATP increased by 20.8 times, from 3.791 to 79.111 m2/g.

Graph: Fig. 4. N2 adsorption-desorption isotherms of ATP (a-1) and TiO2/CTS/ATP (a-2); pore distribution of ATP and TiO2/CTS/ATP (b); particle size distribution of ATP, CTS/ATP, TiO2/ ATP, and TiO2/CTS/ATP (c); and contact angle photograph of ATP (d-1), CTS/ATP (d-2), and TiO2/CTS/ATP (d-3)

Figure 4b shows that the pore size distribution of ATP was relatively uniform, with the main peak at 0–2 nm. By contrast, the pore size distribution of TiO2/CTS/ATP exhibited double peaks, with the dominant peak at 120–140 nm and the secondary peak at 0–2 nm. The finding suggests that ATP modification created considerably larger pores that were expected to enhance the adsorption capacity of the material (Liu et al. [22]).

The particle size distribution of TiO2 before and after modification is shown in Fig. 4c. The average particle sizes of ATP, TiO2/ATP, and CTS/ATP were 3.99 μm, 12.62 μm, and 14.16 μm, respectively, indicating that particle agglomeration occurred in the system after the ATP was treated separately by TiO2 or CTS. In comparison, the TiO2/CTS/ATP adsorbent displayed the lowest average particle size of 2.52 μm. The result suggests that treating ATP with both TiO2 and CTS can not only solve the agglomeration problem but also distribute adsorbents more uniform in size (Siripatrawan and Kaewklin [35]).

Figure 4d illustrates the water contact angle of the laminates of ATP, CTS/ATP, and TiO2/CTS/ATP. ATP had the strongest hydrophilicity with a water contact angle of 9.1°, followed by CTS/ATP (11.7°). By contrast, the water contact angle of TiO2/CTS/ATP soared to 42°, and thus its hydrophobicity was improved substantially (Cieśliński and Krygier [10]; Ku et al. [20]). As shown in Fig. S1, FTIR was applied to analyze the substance structure of the TiO2/CTS/ATP adsorbent in our previous work (Nie et al. [25]). As reported, the absorption of 3600 and 3000 cm-1 indicated the existence of an O–H functional group (Chon and Cho [9]). There was an adsorption peak at 3420 cm-1 with a stretching and vibration of the O–H bond in the hydroxyl functional groups, revealing the presence of TiO2. It is clear that there existed two absorbance peaks at 1612 and 1534 cm-1, corresponding to the stretching vibration of −C=O (amide I peak) and the bending vibration of -N-H (amide II peak), demonstrating the accumulation of Ti–O, N–H, and O–H functional group (Yu et al. [41]). By comparing the main FTIR spectra (700–630 cm-1) of (a) ATP, (b) CTS/ATP, and (c) TiO2/CTS/ATP, it could be concluded that the hydrogen bonds with the hydroxyl groups of TiO2/CTS was found in ATP. As shown in Fig. S2, the morphological features (Nie et al. [25]) of the particle-like TiO2/CTS/ATP were more homogeneous in comparison with those of layered ATP and CTS/ATP, and particle-like adsorbent might be easily for flow passing and Congo red adsorption.

Therefore, the TiO2/CTS/ATP modification a hindered was anticipated to hinder the adsorption of water molecules during the treatment process, which facilitated the removal of Congo red.

It can be seen from Fig. 5a that the amount of adsorbent exhibited a great influence on the removal effect of Congo red. As the dose of TiO2/CTS/ATP increased from 0.1 to 0.5 g/L, the removal efficiency of Congo red soared from 5 to 97.6%. Further increase in adsorbent dose showed an insignificant impact on Congo red removal. Therefore, the optimal dosage under this test condition was 0.5 g/L. Mansor et al. ([23]) also reported that increasing the adsorbent dosage from 1.5 to 2.5 g/L increased the removal efficiency from 72.6 to 96% after 50 min of adsorption time using composite membrane for dye adsorption and filtration. The sorption capabilities of CR onto the TiO2/CTS/ATP first increased as the adsorption period extends. The sorption kinetics of TiO2/CTS/ATP onto CR are given in Fig. S3, and the parameters of the kinetics from the pseudo-second-order are listed in Table S1. A pseudo-second-order adsorption kinetic model (Zhou et al. [45]) can be used to elevate the processes of the adsorbates transferring from the CR solutions to TiO2/CTS/ATP. Thus, the formation of chemical bounds at the bounding sites between the adsorbates and adsorbents is regarded as the rate limiting step during adsorbing process.

Graph: Fig. 5 Effects of doses (a) and pH (b) on Congo red sorption capacity and zeta potential (c) of TiO2/CTS/ATP adsorbent (conditions: adsorbent dosage = 1.0 g/L)

The effect of different pH on the absorption of Congo red by TiO2/CTS/ATP is shown in Fig. 5b. The removal of Congo red by TiO2/CTS/ATP remained above 95% when pH was between 3 and 8, and a higher pH (8–10) led to a slight decrease in Congo red removal. This finding can be explained by the fact that electrostatic attraction served as a major pathway for Congo red removal. In details, Congo red was negatively charged at pH of 3–10, while the TiO2/CTS/ATP adsorbent was positively charged when pH was lower than 6.26 according to the Zeta Potential (Fig. 5c). As the pH was in the range of 6.26–10.0, the sulfonate groups of Congo red readily formed hydrogen bonds with the hydroxyl groups of TiO2/CTS/ATP, thus facilitating the adsorption process (Nodehi et al. [26]). However, increasing pH beyond 10 sharply decreased the Congo red adsorption efficiency. The reason was that electrostatic repulsion became more significant (Chatterjee et al. [6]), leading to a decrease in the adsorption efficiency of Congo red. Specially, at the highest pH of 11, in addition to the strongest electrostatic repulsion, the large number of hydroxides (OH-) present in the solution would compete with Congo red for adsorption sites (Feng et al. [13]), resulting in a sharp decrease in Congo red adsorption.

It can be seen from Fig. 6 that in the case of standalone ultrafiltration treatment, the removal of Congo red from synthetic dyeing wastewater was only 22%. Severe membrane fouling was observed, with TMP rising by 51 kPa within 30 min. The resistance Rf, representing irreversible pollution, caused by adsorption or blockage accounted for 70% of the membrane fouling resistance. The filter cake resistance Rc, representing reversible fouling, was only responsible for 30% of the membrane fouling resistance. When the adsorption-ultrafiltration process was combined, the removal of Congo red was as high as 99%. Previous research also reported that a hybrid adsorption membrane process also achieved more than 97% dye removal in removing dyes from contaminated water (Alardhi et al. [5]). Compared with ultrafiltration alone, membrane fouling was considerably mitigated, as demonstrated by a minor TMP increase of 6 kPa in 30 min. In addition, pretreatment with adsorption increased the reversibility of the fouling, with Rc playing a greater role (67%) compared with Rf (33%). Dong et al. ([12]) also reported that UF process enhanced with powdered activated carbon (PAC) adsorption showed the good application potential in the dye wastewater treatment, and they emphasized that PAC enabled to form an intact filter cake and exhibited the high adsorption ability of PAC toward dye molecules.

Graph: Fig. 6 Effect of adsorption pretreatment on CR removal efficiency (a), TMP (b), and resistance distribution of membrane fouling (c). (conditions: adsorbent dosage = 1.5 g/L, J = 120 L·m-2·h-1)

Figure 7 indicates the deposition of pollutants on the membrane surface. Comparison between Fig. 7b shows an uneven sludge layer formed on the surface of the membrane, and some membrane pores on the surface were still visible. Pretreatment with TiO2/CTS/ATP before ultrafiltration led to the generation of a compact cake layer that covered almost all the pores on the membrane surface. Comparison of Fig. 7c and Fig. 7e demonstrates that the cleaned membrane of the adsorption-ultrafiltration unit had clearer pores as compared with that of direct ultrafiltration. As an acknowledgement, a critical concentration was reached in the membrane boundary layer when the drag force was larger than the thermodynamic forces leading to cake formation (Field and Pearce [14]). Due to high dosage of TiO2/CTS/ATP at 1.5 g/L, a slow aggregation might be happen within a boundary layer on an UF membrane during operation (Aimar and Bacchin [4]).

Graph: Fig. 7 Deposition maps and SEM micrographs (× 500 magnification) of membrane surface pollutants (a virgin, b cake by ultrafiltration, c cake by ultrafiltration after cleaning, d cake by adsorption-ultrafiltration, e cake by adsorption-ultrafiltration after cleaning)

As shown in Fig. 8, the comparison between the adsorption-ultrafiltration process and the controlled ultrafiltration process reveals that the levels of C and Na, which were the main element of Congo red on ultrafiltration membrane and the element on the ultrafiltration membrane without pretreatment, were higher than those measured for the adsorption-UF membrane. This finding indicates that the adsorption-ultrafiltration process alleviated the fouling of Congo red on the ultrafiltration membrane. Furthermore, the levels of O, Al, Si, and Ti on the adsorption-ultrafiltration membrane was significantly higher than its counterpart, as expected due to these four elements being the main components of TiO2/CTS/ATP (Fig. 8a). When the adsorption-ultrafiltration membrane was cleaned, the levels of these four elements decreased significantly, suggesting that these elements contributed to reversible membrane fouling (Fig. 8b).

Graph: Fig. 8 EDS analysis of membrane surface a before and b after physical cleaning

As shown in Fig. 9a, when ultrafiltration was used to remove dye molecules, membrane pores were easily blocked by dye molecules due to direct contact. To address the fouling issue, pretreatment with adsorption is considered (Fig. 8b). The adsorbent was produced by modifying ATP with CTS and TiO2. After modification, the water contact angle of ATP increased from 9.1° to 42° (Fig. 4d), indicating a decrease in hydrophilicity, which reduced the adsorption of water molecules and improved Congo red removal. The specific surface area also increased from 3.791 to 79.111 m2/g, providing more adsorption sites for pollutants. In the process of TiO2/CTS/ATP adsorption, pollutants attached with TiO2/CTS/ATP through electrostatic interaction and hydrogen bonds (Fig. 5). During filtration of pre-adsorbed wastewater, a cake layer containing abundant adsorbents was formed on the ultrafiltration membrane. When approaching the membrane, pollutants were intercepted by the cake layer without entering and blocking the membrane pore (Fig. 9b). As a result, not only enhanced treatment efficiency (99% Congo red removal) but also minimized membrane fouling and improved water permeability were achieved. Furthermore, the disposal of TiO2/CTS/ATP after UF process is advantageous because it does not require considerable maintenance or manpower, and the retained TiO2/CTS/ATP can be sent to landfill.

Graph: Fig. 9 Schematic of mechanism for ultrafiltration in dye wastewater treatment (a) and TiO2/CTS/ATP adsorbent combined with ultrafiltration in dye wastewater treatment (b)

In the current work, TiO2/CTS/ATP adsorbent was developed and applied for adsorption-ultrafiltration process for dye wastewater purification. Our finding was summarized as follows:

• TiO2/CTS/ATP composites were prepared by modifying ATP with CTS and TiO2. Characterization of TiO2/CTS/ATP showed that modification significantly increased hydrophobicity and specific surface area of the adsorbent and decreased the average particle size, which was conducive to adsorption of Congo red.
• When the dosage of TiO2/CTS/ATP was 0.5 g/L, the removal of Congo red was 97.6%. Under acidic conditions, the removal of Congo red was almost not impacted. Increasing pH over 10 was found to greatly reduce the treatment efficiency due to the strongest electrostatic repulsion.
• The removal of Congo red was notably enhanced and reached up to 99% during treatment with by the adsorption-ultrafiltration process. Membrane fouling was considerably mitigated due to adsorption. Adsorption also increased the reversibility of the fouling components.

These aforementioned benefits demonstrate that the combination of TiO2/CTS/ATP adsorption and ultrafiltration is a promising approach to dyeing wastewater treatment.

Zhihong Wang: Experimental design; Writing, original draft preparation. Zekun Wu: Sample analysis; Graph editing. Xujun Zhi: Reactor operation and sample analysis, Data curation. Wanling Tan and Tianfu Tu: Sample analysis. Jinxu Nie: Reactor design, Experimental design. Xing Du: Reactor design; Writing, reviewing; Editing. Yunlong Luo: English editing; Writing, reviewing.

This research was jointly supported by the National Natural Science Foundation of China (51808131) and Natural Science Foundation of Guangdong Province, China (2018A030310569).

All data generated is already in the manuscript.

Not applicable.

All authors read and approved the final manuscript. All authors are fully aware of this manuscript and have permission to submit the manuscript for possible publication

The authors declare no competing interests.

Graph: (DOCX 422 kb)

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.