reading

Chapter 1: Easy and Large-Scale Synthesis of Carbon Nanotube-Based Adsorbents for the Removal of Arsenic and Organic Pollutants from Aqueous Solutions

Chapter 2: Potentialities of Graphene-Based Nanomaterials for Wastewater Treatment

Chapter 3: Photocatalytic Activity of Nanocarbon-TiO₂ Composites with Gold Nanoparticles for the Degradation of Water Pollutants

Chapter 4: Carbon Nanomaterials for Chromium (VI) Removal from Aqueous Solution

Chapter 5: Nano-Carbons from Pollutant Soot: A Cleaner Approach toward Clean Environment

Chapter 6: First-Principles Computational Design of Graphene for Gas Detection

Chapter 8: Flocculation Performances of Polymers and Nanomaterials for the Treatment of Industrial Wastewaters

8.4 Synthesis of Nanomaterials-Based Flocculants and Utilisation in the Removal of Pollutants

Chapter 10: A Perspective of the Application of Magnetic Nanocomposites and Nanogels as Heavy Metal Sorbents for Water Purification

10.4 Heavy Metal Removal from Aqueous Solutions Using Magnetic Nanomaterials and Nanogels

Chapter 12: Adsorption of Metallic Ions Cd²⁺, Pb²⁺, and Cr³⁺ from Water Samples Using Brazil Nut Shell as a Low-Cost Biosorbent

Chapter 14: Treatment of Reactive Dyes from Water and Wastewater through Chitosan and its Derivatives

Chapter 15: Natural Algal-Based Processes as Smart Approach for Wastewater Treatment

Publishers at Scrivener
Martin Scrivener(martin@scrivenerpublishing.com)
Phillip Carmical (pcarmical@scrivenerpublishing.com)

Co-published by John Wiley & Sons, Inc. Hoboken, New Jersey, and Scrivener Publishing LLC, Salem, Massachusetts.
Published simultaneously in Canada.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permission.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com.

For more information about Scrivener products please visit www.scrivenerpublishing.com.

Smart materials have been a thrust area to the researchers in the development of new materials that lead to create new tools and techniques, which will help in the development of advance technology. At the nano size, smart materials often take on unique and sometimes unexpected properties. This means that at the nanoscale, materials can be “tuned” to build faster, lighter, stronger and more efficient devices and systems. “Smart materials” have been extensively used in a variety of applications due to the change in the characteristics of the materials with small variation on stimuli. They are also known as responsive materials. Smart materials change their properties abruptly in response to small changes in the environmental conditions such as pH, temperature, electric and magnetic fields. Due to the versatility of such characteristics, these materials are highly applicable in the area of materials science, engineering, sensors and environmental applications. Besides, such materials are applied to develop newer composites, ceramics, chiral materials, liquid crystals, conducting polymers, hydrogels, nanocomposites and biomaterials. These smart materials are highly suitable for environmental remediation.

Water used for drinking and household needs must have good taste and no odour and be harmless to human health as well as the livestock. Clean water is always a need, which often calls for a cheap and efficient water purification system. There are several technologies and have been utilized for the water treatment process. Smart materials have been used to develop more cost-effective and high-performance water treatment systems as well as instant and continuous ways to monitor water quality. Smart materials in water research have been extensively utilized for the treatment, remediation and pollution prevention. Smart materials can maintain the long-term water quality, availability and viability of water resource. Thus, water via smart materials can be reused, recycled and desalinized, and it can detect the biological and chemical contamination as well as whether the source is from municipal, industrial or man-made waste.

The present book describes the smart materials for waste water application and it will be highly beneficial to the researchers working in the area of materials science, engineering, environmental science, water research and waste water applications. Chapters included in the book have been differentiated in three sections: first section includes the various “carbon nanomaterials” with a focus on use of carbon at nanoscale applied for waste water research. Second section involves “synthetic nanomaterials” for pollutants removal. The third section includes “biopolymeric nanomaterials” where the authors have used the natural polymers matrices in a composite and nanocomposite material for waste treatment. The potential researchers working in the area will benefit from the fundamental concepts, advanced approaches and application of various smart materials towards waste water treatment described in the book. The book also provides a platform for all researchers to carry out advanced research as well as to delve into the background in the area. The book also covers recent advancement in the area and prospects about the future research and development of smart materials for the waste water applications.

Part 1

CARBON NANOMATERIALS

Chapter 1

Easy and Large-Scale Synthesis of Carbon Nanotube-Based Adsorbents for the Removal of Arsenic and Organic Pollutants from Aqueous Solutions

¹College of Chemistry and Environmental Engineering, Shanghai Institute of Technology, Shanghai, P. R. of China

²State Key Laboratory of Pollution Control and Resource Reuse, College of Environmental Science and Engineering, Tongji University, Shanghai, P. R. of China

Abstract

The as-prepared carbon nanotubes (APCNTs) synthesized by chemical vapor deposition method usually contained carbon nanotubes (CNTs) and quantities of iron nanoparticles (INP)–encapsulated carbon shells. The traditional research mainly focuses on how to remove the INPs using various chemical and physical purification methods. In this chapter, we have synthesized many kinds of CNTs-based adsorbents based on the aforementioned iron/carbon APCNT composites without purification, which can be used for the removal of arsenic and organic pollutants from aqueous solutions with excellent adsorption properties. This synthesis method is applicable to as-prepared single-walled CNTs and multi-walled CNTs containing metal catalytic particles (e.g., Fe, Co, Ni), and the resulting material may find direct applications in environment, energy storage, catalysis, and many other areas. Results of this work are of great significance for large-scale practical applications of APCNTs without purification.

Keywords: Magnetic carbon nanotubes, arsenic, organic pollutants, adsorption

1.1 Introduction

Magnetic carbon nanotubes (MCNTs) are of intense research interests because of their valuable applications in many areas such as magnetic data storage, magnetic field screening, and signal transmission [1]. Synthetic methods of MCNTs include chemical and physical techniques. For instance, magnetic nanoparticles (MNPs), such as cobalt, iron, or nickel and their oxide NPs, can be encapsulated by carbon nanotubes (CNTs) [2, 3]. In addition, MNPs could be deposited on the external surface of CNTs [4, 5]. However, existing synthesis methods may have the following critical disadvantages: firstly, the as-prepared carbon nanotubes (APCNTs) are usually purified using strong acids to remove metal particles and carbonaceous byproducts [6], and then MNPs are loaded on the wall of purified CNTs. As a result, the synthesis process is expensive and time consuming with a low yield. Secondly, uncovered MNPs may agglomerate when a magnetic field is applied. Thirdly, bare MNPs could be oxidized in air or erode under acidic conditions [7]. These issues may ultimately hinder widespread practical applications of the MCNTs composite. In recent years, CNTs could be produced in ton-scale quantities per year with high quality. However, APCNTs often contain a large fraction of impurities, including small catalytic metal particles and carbonaceous byproducts such as fullerenes, amorphous, or graphitic carbon particles. The current research direction in this area mainly focuses on the purification of APCNTs through physical separation [8], gas-phase oxidation [9], and liquid-phase oxidation [6], aiming at applications of purified CNTs. However, these purification processes are complex, time consuming, and environmentally unfriendly. Hence, the existing approach is suitable for fundamental research but not for large-scale applications of MCNTs.

To overcome the aforementioned issues, we herein report several new-typed methods to produce MCNTs using APCNTs. We show that MCNTs can be well dispersed in water with excellent magnetic properties. This facile synthesis method has the following advantages: firstly, metal nanoparticles in the APCNTs can be utilized directly without any purification treatment; secondly, the carbon shells provide an effective barrier against oxidation, acid dissolution, and movement of MNPs and thus ensure a long-term stability of MNPs. MCNTs were used as adsorbents for the removal of environmental pollutants in aqueous solutions, and arsenic and organic pollutants were chosen as target pollutants. MCNTs exhibit excellent adsorption and magnetic separation properties. After adsorption, the MCNTs adsorbents could be effectively and immediately separated using a magnet, which reduces potential risks of CNTs as another source of environmental contaminant. Therefore, MCNTs can be used as a promising magnetic adsorbent for the removal of arsenic and organic pollutants from aqueous solutions.

1.2 Removal of Arsenic from Aqueous Solution

1.2.1 Activated Magnetic Carbon Nanotube

1.2.1.1 Synthesis Method

The APCNTs were prepared using a chemical vapor deposition(CVD) method [10]. Ethanol was used as the carbon feedstock, ferrocene was used as the catalyst, and thiophene was used as the growth promoter. Argon flow was introduced in the quartz tube in order to eliminate oxygen from the reaction chamber. The ethanol solution dissolved with ferrocene and thiophene was supplied by an electronic squirming pump and sprayed through a nozzle with an argon flow. After several hours of pyrolysis, the supply of ethanol was terminated, and the APCNTs were collected from a collecting unit connected to the quartz tube.

The Magnetic iron oxide/CNTs(MI/CNTs) composites were prepared by an alkali-activated method using APCNTs. The APCNTs were prepared by the catalytic chemical vapor deposition method [10]. In a typical synthesis, APCNTs and KOH powder were mixed in a stainless steel vessel in an inter-gas atmosphere. The weight ratio of KOH to APCNTs was 1:4. The APCNTs and KOH powder were mixed for 10 min using a mortar, which resulted in a uniformly powder mixture. The mixture was then heated to 1023 K for 1 h under flowing argon in a horizontal tube furnace, washed in the deionized water, and then dried.

1.2.1.2 Characterization of Adsorbents

Figure 1.1a displays the transmission electron microscopy (TEM) images of APCNTs; the diameter of APCNTs is about 20–30 nm and the length is about 1 μm. After the activation treatment, the structure of APCNTs has been clearly modified, and the length of MI/CNTs is obviously shortened. Furthermore, part of the hollow tubular structure is destroyed, large quantities of defects are produced, and many flaky apertures are generated on the surface, as shown in Figure 1.1b. The X-ray diffraction (XRD) patterns of MI/CNT hybrids indicated that the MI/CNTs were a mixture of two/three phases: γ-Fe₂O₃/Fe₃O₄ and CNTs. Well-resolved diffraction peaks reveal the good crystallinity of γ-Fe₂O₃/Fe₃O₄ specimens; no peaks corresponding with impurities were detected. Peaks of C with relatively high intensity and symmetry are clearly observed in Figure 1.1c. This observation suggests the graphite structure remained, even after strong activation reaction; therefore, we may conclude that MI/CNT heterostructures were formed using the KOH activation method.

Figure 1.1 TEM images of (a and b) APCNTs and MI/CNTs, (c) XRD patterns, and (d) Raman spectrum of MI/CNTs.

The Raman spectrum of MI/CNTs is shown in Figure 1.1d. For MI/CNTs, the remaining peaks at 224 and 285 cm^–1 are assigned to the A1g and Eg modes of α-Fe₂O₃, the peak at 397 cm^–1 is assigned to the T2g modes of λ-Fe₂O₃, and the peak at 667 cm^–1 is assigned to the A1g modes of Fe₃O₄ [11]. The results indicate that magnetic iron oxide in MI/CNTs may be a mixture phase composed of α-Fe₂O₃, λ-Fe₂O₃, and Fe₃O₄. The G peak at 1585 cm^–1 is related to E2g graphite mode [12–14]. The D-line at ~1345 cm^–1 is induced by defective structures. The intensity ratio of the G and D peaks (IG/ID) is an indicator for estimating the structure quality of CNTs, as shown in Figure 1.1d, which suggests the structure of the CNTs was destroyed after the KOH activation process.

In Figure 1.2a, the thermogravimetric analysis (TGA) of the MI/CNTs exhibits two main weight-loss regions. MI/CNTs are considerably stable and show a little weight loss close to 5% below 200 °C in the first region, which can be attributed to the evaporation of adsorbed water and the elimination of carboxylic groups and hydroxyl groups on the MI/CNTs. The rapid weight-loss region can be due to the oxidation of CNTs. It is clearly seen that the main thermal events temperature was at ~500 °C; however, the thermal events temperature is so high that MI/CNTs could readily meet the application needs of adsorbent in water treatment.

Figure 1.2 (a) Thermal analysis and (b) adsorption/desorption isotherms of N₂ and pore distribution of MI/CNTs.

The specific surface area (SSA) and pore size characterization of MI/CNTs were performed by nitrogen (77.4 K) adsorption/desorption experiments with density functional theory (DFT) methods [15], as shown in Figure 1.2b. The MI/CNTs had a high SSA of ~662.1 m²/g (calculated in the linear relative pressure range from 0.1 to 0.3). The SSA of MI/CNTs is drastically increased by ~5 times than APCNTs; such increases correspond with a decrease in mean pore diameter from ~11.03 to ~2.26 nm (BJH). After alkalis activation treatment, not only the tube tip was opened, but also large quantities of new micropore structures with small sizes were produced. This implies that the MI/CNTs possess more small pores after the present activation treatment and thus could lead to a higher SSA. The detailed features of the pore distribution analysis are presented in Table 1.1.

Figure 1.2b displays the results of the cumulative pore volume (PV) and pore size analysis from nitrogen adsorption by applying a hybrid Non-local Density Functional Theory (NLDFT) kernel, assuming a slitshape pore for the micropores and a cylindrical pore for the mesopores. The obtained pore size/PV distribution indicates this MI/CNTs sample is distinctive from APCNTs. Compared with APCNTs, the total PV of MI/CNTs decreased due to the disappearance of maro pores. The meso-PV and micro-PV of MI/CNTs improved by almost ~1.48 and ~5.73 times than that of APCNTs. The stronger peak of pore distribution of MI/CNTs exists at ~2 nm, which indicates the presence of plenty of micropores after the alkalis activation treatment.

The composition of MI/CNTs was further determined by XPS, as shown in Figure 1.3. Typical XPS survey scans of the MI/CNTs are shown in Figure 1.3a. Figure 1.3b is the principal deconvoluted component of the C1s region recorded for the MI/CNTs. We can see that the strongest peak at 284.6 eV is assigned to double-bonding carbons for CNTs and results from non-functionalized carbon. The peak at the binding energy of about 285.1 eV is a consequence of single-bonding carbon for CNTs [16]. The O1s spectra consist of three peaks that are assigned to Fe–O (529.8 eV), C=O (531.2 eV), and C–O (533.0 eV) bonds [17], which suggests the introduction of new functional groups and the iron oxide nanoparticle loading on the surfaces of MI/CNTs. The Fe2p spectrum (Figure 1.3c) shows two broad peaks with satellite peaks at 711.4 and 724.5 eV, representing Fe2p3/2 and Fe2p1/2, respectively. The presence of these chemical bonds demonstrates that iron oxide nanoparticles formed on the surface of MI/CNTs.

Figure 1.3 X-ray photoelectron spectroscopy (XPS) survey scans of the MI/CNTs (a) the C1s deconvolution of MI/CNTs (b), for Fe2p region of MI/CNTs (c), and the O1s deconvolution of MI/CNTs (d).

The magnetization properties of MI/CNTs were investigated at room temperature by measuring magnetization curves (Figure 1.4). The saturation magnetization (Ms) of MI/CNTs is 27.2 emu·g^–1 for MI/CNTs (magnetic field=+10 kOe), indicating that MI/CNTs have a high magnetism. The loop of MI/CNTs exhibits very low coercive field (40 Oe) and remanence values (0.76 emu·g^–1), indicating that MI/CNTs hybrids are very close to behaving as superaramagnets at room temperature, which can be beneficial to the reuse without reunite for magnetization. After magnetic separation, the concentration of residual CNTs in an aqueous solution was estimated using a UV–visible absorption-based approach [18, 19].

Figure 1.4 Hysteresis loop of MI/CNTs and the digital photograph of MI/CNTs with magnetic separation.

The MI/CNT powders can be well dispersed in water during adsorption, so the removal of arsenic by MI/CNTs was found to be rapid at the initial period (in the first 1 min) and then became slower (1–10 min). The rate of removal reached a plateau after ~10 min of the experiment, as shown in Figure 1.5.

Figure 1.5 Kinetic curves of arsenic removal on MI/CNTs (As(V)/As(III) concentration=2 mg/L, MI/CNTs=0.2 g/L).

1.2.1.3 Adsorption Properties

The MI/CNTs’ adsorption isotherms for removing As(V) and As(III) are shown in Figure 1.6, which indicates that APCNTs have a smaller adsorption capacity of As(V)/As(III), due to poor interaction between CNTs and arsenic pollutants. After the KOH activation treatment, a larger number of iron oxide nanoparticles were decorated on the surface of MI/CNTs, and the resulting adsorption capacity increased significantly, which suggests that the iron oxides contributed to the increase of adsorption capacity.

Figure 1.6 Equilibrium adsorption isotherms of As(V) and As(III) on APCNTs and MI/CNTs.

The equilibrium adsorption of arsenic on MI/CNTs was analyzed using the Langmuir [20] and Freundlich [21] isotherm models. Figure 1.7 shows the isotherms based on the experimental data, and the parameters obtained from linear regression using adsorption models are shown in Table 1.2. Based on the determination coefficient (R²), Langmuir model fits the experimental data better than the Freundlich model. The applicability of the Langmuir isotherm suggests that specific homogenous sites within the adsorbent are involved [22]. The computed maximum monolayer capacities have wonderful values of ~9.74 mg/g for As(V) and ~8.13 mg/g for As(III) onto the MI/CNTs, which are also higher than those of previously reported adsorbents [23–26]. These results suggest that MI/CNTs have great potential for As(V) and As(III) removal. The Dubinin–Radushkevich (D–R) [27] isotherm model was applied to distinguish between the physical and chemical adsorption of As(V) and As(III) on MI/CNTs, as shown in Figure 1.7c. The values of E (mean energy of adsorption) exceed 16 10 kJ·mol^–1 for As(V) and As(III), suggesting the removal process may follow chemisorption between As(V)/As(III) and MI/CNTs [28]. Surface structure information of MI/CNTs was analyzed by XPS after arsenic adsorption. The XPS spectra of As 3d are shown in Figure 1.8. The quantitative analysis of As(III)-adsorbed MI/CNTs shows 50.8% of As(III) and 49.2% of As(V) on the sorbent surface; however, only As(V) exists on the As(V)-adsorbed MI/CNTs. This result indicates solid-state oxidation between arsenate and arsenite on the surface of MI/CNTs. It has been reported that the arsenic adsorption mechanism involves electrostatic attraction [29] and surface complexation [30, 31] between the arsenic species and iron oxides in solution. Arsenate surface complexes possibly exist in three forms of binuclear, bidentate, and monodentate complexes through the ligand exchange reaction [32]. After arsenic adsorption, the atomic ratio of Fe on the surface decreased from 6.58% to 5.35% for As(III) and 5.86% for As(V) with the increase of As atomic ratio, indicating that Fe atoms were overlaid by the adsorbed arsenic. The higher adsorption capacity of MI/CNTs can be attributed to the following reasons: first, a large number of Fe₂O₃ nanoparticles formed on the surface of MI/CNTs, and iron oxide adsorbents have demonstrated superior adsorption performance [33–35]. Second, oxygen-containing functional groups of MI/CNTs would improve hydrophilicity and dispersibility in aqueous solutions. For As(V)/As(III) with considerable solubility, a better dispersion of MI/CNTs in water will increase the available adsorption sites, which may be favorable for the aqueous-phase adsorption [36]. Third, the increased SSA and meso-PV/micro-PV would provide more adsorption sites for As(V)/As(III) [37]; therefore, the adsorption properties obviously increased after the KOH activation treatment.

Figure 1.7 (a) Langmuir, (b) Freundlich, and (c) D–R isotherms for As(V) and As(III) adsorption onto MI/CNTs.

Figure 1.8 As 3d XPS spectra of MI/CNTs after As(III) (a) and As(V) (b) adsorption.

Table 1.2 Langmuir, Freundlich, and Dubinin–Radushkevich isotherms parameters of As(V) and As(III) adsorption on MI/CNTs system.

1.2.2 Sulfhydryl-Functionalized Magnetic Carbon Nanotube

1.2.2.1 Synthesis Method

Briefly, APCNTs were firstly treated by a two-step heat treatment. In the first step, the APCNTs were heated in the air at 400 °C for 60 min to destroy the carbon cages and oxidize the iron nanoparticles (INPs). Air was introduced into the quartz tube at a slow rate to provide oxygen continuously. In the second step, the hybrids obtained from step one were heated at 850 °C for 60 min under Ar gas protection to remove the rest carbon by the redox reaction between C and Fe₂O₃ through which Fe₂O₃ was reduced to Fe, FeO, or Fe₃O₄ by the encapsulating C.

Before the Glutathione (GSH) functionalization, the FeO_x/CNTs hybrids were oxidized by NaClO solution. Five hundred milligrams FeO_x/CNTs hybrids were suspended in the 70% antiformin solution of 200 mL, and then sonicated for 2 h in an ultrasonic bath. The suspension was stirred at 80 °C for 5 h and cooled to room temperature. The oxidized FeO_x/CNTs hybrids (OMI-CNTs) were washed with deionized water for three times, and then dried under vacuum at 80 °C overnight.

Subsequently, 400 mg OMI-CNTs was dispersed by sonication and magnetic stirring in 200 mL N,N-dimethylformamide (DMF) for 1 h. Then 1.5 g GSH was added into the suspension under stirring for 1 h. Afterward, 10 mg of coupling agent 2-(7-aza-1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (N-HATU) was added, and the mixture was transferred to a water bath (60 °C) with stirring and reflux condensation for 6 h. The product was washed with EtOH and deionized water for three times, respectively, and dried under vacuum at 80 °C for 12 h.

1.2.2.2 Characterization of Adsorbents

Scheme 1.1 illustrates the synthesis process of GSH-functionalized MI-CNTs. Figure 1.1a displays the TEM images of APCNTs. It was found that entangled CNTs bundles were mixed with a high density of iron catalyst particles, as indicated by the black dots. High-resolution TEM (HRTEM) (the inset of Figure 1.1a) image reveals that the INPs are covered with carbon cages or shells with diameters ranging from 2 to 5 nm, and these cages consist of several graphitic layers. After the heat treatment, the entangled state of APCNTs was hardly changed. However, the carbon cages over the INPs have been completely destroyed, which has been previously reported [6, 10]. This is beneficial for the contact between As(III) and the INPs. Furthermore, the size of the INPs obviously increased after heat treatment, which is due to the crystallization and growth of the nanoparticles at high temperature. After GSH functionalization, no obvious change was observed in the morphology of GMI-CNTs compared with MI-CNTs, as shown in Figures 1.1b and 1.S1. The lattice interplanar spacings of 0.286 and 0.242 nm correspond to the (220) and the (222) planes of Fe₃O₄, respectively, revealing the Fe₃O₄ NPs are polycrystalline. The interplanar spacings of 0.221 and 0.252 nm are characteristics of the (113) plane of α-Fe₂O₃ and the (311) plane of γ-Fe₂O₃, respectively.

Scheme 1.1 Illustration of Glutathione functionalized magnetic carbon nanotube (GMI-CNT) synthesis.

Figure 1.9d shows the XRD patterns of APCNTs, (magnetic carbon nanotubes)MI-CNTs, (oxidized magnetic carbon naotubes)OMI-CNTs, and GMI-CNTs. The peaks associated with the mixture of zero-valent Fe, γ-Fe₂O₃/Fe₃O₄ appeared after the heat treatment. Well-resolved diffraction peaks reveal the good crystallinity of Fe, γ-Fe₂O₃/Fe₃O₄. The peaks of C with a relatively high intensity and symmetry are clearly observed, which suggests that the graphite structure remained even after heat treatment. Therefore, we can conclude that MI-CNT heterostructures were formed after the heat treatment. In our study, APCNTs were firstly treated by a two-step heat treatment. In the first step, the APCNTs were heated in the air at 400 °C for 60 min to destroy the carbon cages and oxidize INPs: Fe + O₂→fFe₂O₃. In the second step, the hybrids obtained from step one were heated at 850 °C for 60 min under Ar gas protection to remove the rest carbon by the redox reaction between C and Fe₂ O₃: C + Fe₂O₃ → Fe +CO_x (Scheme 1.2). After the oxidation step, the relative intensity of C peaks decreased, indicating large quantities of defects were generated during the oxidation process. Moreover, the intensity of Fe peaks also obviously decreased because the zero-valent Fe was partially oxidized to α-Fe₂O₃, which was confirmed by the appearance of new peaks located at 33.152°, 49.496°, and 54.089° in the XRD pattern of OMI-CNTs. This result is consistent with the aforementioned HRTEM analysis, and the iron of GMI-CNTs can be identified as FeO_x (x=0, 4/3, 3/2) (Figure 1.9b). The transformation of Fe states during the synthesis process is illustrated in Scheme 1.2. The XRD patterns of GMI-CNTs indicate that the functionalization process did not change the component of INPs on the surface of CNTs and would not inhibit the reaction between INPs and As(III).

Figure 1.9 TEM images of APCNTs and GMI-CNTs (a and b), HRTEM image of INPs on GMI-CNTs (c), and XRD patterns of prepared samples (d). The inset of (a) shows INPs in APCNTs covered with carbon cages.

Scheme 1.2 The transformation of Fe species during the synthesis process.

XPS was employed to analyze the surface chemical composition, as shown in Figure 1.10. Typical survey scans of GMI-CNTs and OMI-CNTs are shown in Figure 1.10a. After functionalization, the new peaks of S and N appeared. For C 1s spectra in Figure 1.10b, the peak of typical graphitic carbon attributed to C 1s spectra was found at 284.6 eV. Other three peaks located at 285.4, 286.8, and 288.7 eV are assigned to C–N, C–O, and C=O, respectively. The S 2p peak was deconvoluted into two separate peaks at 163.8 and 165.1 eV, contributing to the –SH groups (Figure 1.10c) [38]. The peak located at ~400.2 eV corresponds to the –NH– groups (Figure 1.2d) [39]. The O 1s spectra consist of three peaks at 530.1, 531.5, and 532.9 eV, which are assigned to Fe–O, C=O, and O–C=O, respectively [17]. A broad peak at 718.0 eV represents overlapping components for oxidized iron and zero-valent iron [40]. The Fe 2p spectra show two broad peaks of Fe 2p3/2 and Fe 2p1/2 with satellite peaks at ~711.0 and ~724.8 eV, respectively (Figure 1.11). The two peaks at 712.9 and 726.7 eV correspond to Fe₂O₃, while peaks at 710.8 and 724.3 eV represent Fe₃O₄. The surface analysis demonstrated the iron phase in GMI-CNTs was the mixture of zero-valent iron and iron oxides which is in accordance with the TEM and XRD analyses and confirmed that GSH molecules have been successfully grafted on the OMI-CNTs.

Figure 1.10 XPS survey scans of the GMI-CNTs (a), the core peaks of C 1s (b), S 2p (c), and N1s (d) on the surface of GMI-CNTs.

Figure 1.11 XPS Fe2p spectra of GMI-CNTs.

The Fourier transform infrared (FT-IR) spectrum of OMI-CNTs and GMI-CNTs provides further evidence of the successful graft of GSH on OMI-CNTs (Figure 1.12a). The peak at ~3300 cm^–1 corresponds to the O–H stretching vibration of adsorbed water or some other O–H containing groups, such as carboxyl [41]. The band at 1717 cm^–1 attributes to the stretching vibrations of C=O of the carboxyl groups, which confirms the formation of carboxyl groups after the oxidation step [25, 42]. The peaks at 1654 and 1575 cm^–1 indicate the formation of secondary amide on the OMI-CNTs resulting from the functionalization [43]. The strong peak at 568 cm^–1 corresponds to the stretching vibration of large quantities of Fe–O. The FT-IR spectra also confirm that the GSH molecules are covalently bonded to OMI-CNTs.

Figure 1.12 FT-IR spectra (a) of OMI-CNTs and GMI-CNTs and TG-DSC (b) of GMI-CNTs.

The TGA indicated that the GMI-CNTs exhibit three main weight-loss peaks (Figure 1.12b). The total weight loss of GMI-CNTs is approximately 45% before 550 °C, indicating the loading ratio of INPs on GMI-CNTs reaches around 55%, which is much higher than many other iron oxide/CNTs based composites. The result is consistent with the strong Fe–O peak in the aforementioned FT-IR spectrum. A slight weight loss close to ~6% occurred below ~180 °C, which is due to the evaporation of adsorbed water and the elimination of carboxylic groups and hydroxyl groups on the GMI-CNTs [44, 45]. The second stage weight loss observed between ~180 and 400 °C is associated with the thermal decomposition of GSH on the OMI-CNTs [46, 47]. The rapid weight-loss region between ~400 and ~540 °C is attributed to the oxidation of CNTs. The TGA indicated that the stability of GMI-CNTs can meet the application needs of adsorbents in water treatment.

The SSA and pore parameters of GMI-CNTs were measured by nitrogen (77.4 K) adsorption/desorption experiments (Figure 1.13a). The SSA of MI-CNTs (299.4 m²·g^–1) significantly increased by ~2.6 times, corresponding with the decrease in the average pore size from 11.03 to 5.01 nm (BJH). The micropore volumes of APCNTs calculated by the NLDFT kernel before and after heat treatment were both close to 0, indicating most surface area was attributed to mesopores. The meso-PV and micro-PV of MI-CNTs (0.778 cc·g^–1) slightly changed compared with that of APCNTs (0.897 cc·g^–1). However, the pore size becomes smaller and more centralized, indicating the heat treatment can not only remove the carbon cages over the INPs but also produce much smaller and more uniform micropore/mesopore structures, which may finally result in the increase of SSA. After the functionalization, the SSA (139.9 m²·g^–1) and micro-PV/meso-PV (0.501 cc·g^–1) of GMI-CNTs decreased compared with MI-CNTs, whereas the pore size was still uniformly distributed with an average pore size of 5.0 nm, which is beneficial for adsorption of pollutants. Although the SSA and meso-PV/micro-PV decreased after the functionalization, adsorption capacity of GMI-CNTs for As(III) is much higher than that of MI-CNTs, indicating the sulfhydryl groups of GSH play a very important role in the enhancement of adsorption capacity.

Figure 1.13 (a) N₂ adsorption/desorption isotherms and pore size distribution (inset) of GMI-CNTs and (b) hysteresis loop of GMI-CNTs. The inset of (b) is the digital photograph of GMI-CNTs dispersed in water (1) and separated with magnetic separation (2).

The magnetization properties of GMI-CNTs were investigated at room temperature by measuring magnetization curves, as shown in Figure 1.13b. The magnetization properties investigation showed the Ms of GMI-CNTs is 27.3 emu·g^–1 with relatively low coercive force and remanence of 104.9 Oe and 2.4 emu·g^–1, which can be beneficial for the reuse without reunite for magnetization. It can be found from the insets that the As(III)-loaded GMI-CNTs can be easily separated from water by using a magnet (the inset of Figure 1.13b). The concentration of residual CNTs in an aqueous solution was estimated nearly 0 g·L^–1 by a UV–visible absorption-based approach [45].

1.2.2.3 Adsorption Properties

The amounts of adsorbed As(III) versus aqueous-phase concentration are plotted as adsorption isotherms in Figure 1.14a. The isotherms data were fitted by two commonly used models, Langmuir (Figure 1.14b) and Freundlich model (Figure 1.14c), in a linear way. It was observed that GMI-CNTs exhibit significantly enhanced adsorption performance (19.12 mg·g^–1) for As(III) compared with OMI-CNTs (9.39 mg·g^–1). The Langmuir and Freundlich models were employed to fit the experimental data. The Freundlich model describes the equilibrium on heterogeneous surfaces, while the Langmuir model assumes all adsorption occurs through the same mechanism on a homogeneous surface. Aforementioned TEM, XRD, and XPS analyses have confirmed the Fe element exists as multiphase on the surface of OMI-CNTs and GMI-CNTs. After the GSH functionalization, the GMI-CNTs become more heterogeneous in terms of microstructure than unmodified OMI-CNTs. Thus, we speculate that the adsorption mechanism of As(III) on GMI-CNTs may involve not only the complexation with Fe species but also the surface complexation with sulfhydryl groups carried by GSH. A smaller intensity parameter 1/n (0.214) of GMI-CNTs implies the stronger adsorption bond and a more heterogeneous surface compared with OMI-CNTs, while a larger capacity parameter K_F (12.56) of GMI-CNTs reveals a higher adsorption capacity [48, 49]. The results clearly demonstrate the improved adsorption performance of GMI-CNTs in comparison with OMI-CNTs. The adsorption capacity is also much higher than that of other carbon-based composites reported previously (Table 1.3).

Figure 1.14 (a) Equilibrium adsorption isotherms of As(III) on APCNTs, OMI-CNTs, and GMI-CNTs; equilibrium data of As(III) adsorption on GMI-CNTs fitted by (b) Langmuir model and (c) Freundlich model in the linear form; (d) pH effect on As(III) adsorption by GMI-CNTs.

Table 1.3 Comparison of the adsorption performance of GMI-CNTs with other adsorbents.

The effect of pH on As(III) adsorption by GMI-CNT was studied for further understanding of adsorption mechanism. As shown in Figure 1.14d, as the pH increased from 2 to 10, the As(III) adsorption capacity increased to the maximum and then decreased. The low As(III) removal efficiency at the relatively strong acid condition is due to the release of Fe/FeO_x to the solution (Figure 1.14d) and probably the hydrolysis of amide bond connecting the OMI-CNTs and GSH. The maximum adsorption of As(III) by GMI-CNTs was observed at pH ~7, where the As(III) exists predominantly as H₃AsO₃. Thus, the enhanced As(III) adsorption by GMI-CNTs could not be explained by the electrostatic interactions but explained by the binding of the non-ionic H₃AsO₃ to the sulfhydryl groups. It has been reported that As(III) reacts with the sulfhydryl groups by directly bonding with three S atoms [57]. However, the bonding between As(III) and the sulfhydryl groups is strongly pH dependent that the pKa values of As(III)-GSH are 2.10, 3.59, and 8.68 (two carboxyl groups and sulfhydryl group, respectively) [58]. Therefore, the significant decrease of As(III) uptake by GMI-CNTs in the pH range from 9 to 11 can be explained as: (a) the repulsive force between the ionized As(III) species and negatively charged mercaptides formed due to the dissociation of As(III)-GSH and (b) the competitive adsorption between the ionized As(III) species and hydroxyl ions. In the pH range from 6 to 8, As(III) can strongly bond to the ionized GSH to form complexes of [As(GSH)₃]³⁻ and [As(GSH)(OH)₂]²⁻ with releasing water molecules [58]. The results showed that GMI-CNTs would be favorably used under pH range from 5 to 8. Near-edge x-ray absorption fine structure (NEXAFS) studies of the Fe K-edge indicated that there is no phase change of Fe before and after As(III) adsorption, suggesting the Fe species are chemically stable and no redox reaction occurred in the adsorption process (Figure 1.15). The intensity and position of all peaks hardly changed, indicating As atoms were not directly bonded with Fe atoms. Instead, the As atoms and Fe atoms might be connected by O atoms to form surface complexes [59, 60]. Schematic illustration of the proposed mechanism of As(III) adsorption by GMI-CNTs is shown in Schema 1.3.

Figure 1.15 Fe K-edge NEXAFS spectra.

Scheme 1.3 Proposed mechanism of As(III) adsorption by GMI-CNTs.

1.3 Removal of Organic Pollutants from Aqueous Solution

1.3.1 Magnetic Carbon Nanotubes for Dye Removal

1.3.1.1 Synthesis Method

The APCNTs samples were oxidized using 30% NaClO (70 mL of H₂O + 30 mL NaClO) solution, with magnetic stirring at ambient temperature for 12 h. After oxidation, the mixture was filtered through a 0.45 μm millipore polycarbonate membrane and the filtered solid was washed using deionized water repeatedly until the solution pH became neutral.

1.3.1.2 Characterization of Adsorbents

The morphology of magnetic as-prepared carbon nanotubes (MAPCNTs) was characterized by HRTEM as shown in Figure 1.16. Figure 1.16a shows that entangled MAPCNTs were mixed with some metallic particles, as indicated by numerous black dots. Figure 1.16c and d reveals that the MNPs were covered with carbon cages or shells with diameters ranging from 2 to 30 nm, and these cages consisted of several graphitic layers, which were widely observed in the synthesis of CNTs by all previous methods [61, 62]. It is usually difficult to prevent metallic MNPs (Fe, Co, Ni) from oxidation under conventional experimental conditions [6]. The carbon cages would provide an effective barrier against oxidation and agglomeration of MNPs and ensure a long-term stability. The special structure of MNPs encapsulated by graphitic shells could be beneficial to the stable application of MAPCNTs under high temperature or acid conditions. The elemental composition of MAPCNTs was analyzed by energy-dispersive X-ray spectroscopy (EDX) (Figure 1.17a). The EDX results show that the MAPCNTs consisted of iron and carbon, and the Cu peaks resulted from the copper grid used to support the MAPCNTs.

Figure 1.16 HRTEM images of MAPCNTs (a and b) and iron particles embedded in graphitic shells (c and d).

Figure 1.17 Energy-dispersive X-ray spectrum (a) and room temperature magnetization curve (b) of MAPCNTs.

The magnetization characteristics of MAPCNTs were investigated at room temperature (Figure 1.17b). The Ms of MAPCNTs is 3.64 emu·g^–1 (magnetic field = ±10 kOe), indicating that MAPCNTs are ferromagnetic materials, which can be attributed to the iron MNPs. From Figure 1.17b, the lower values of coercivities HC (285 Oe) show that MAPCNTs are soft magnetic materials, which could be used in transformer and inductor cores, recording heads, microwave devices, and magnetic shielding. Kuryliszyn-Kudelska examined the influence of chemical treatment on magnetic properties of APCNTs [63, 64]; a significant change of Ms was observed after chemical treatment. Therefore, proper chemical modification and heat treatment technologies could control and improve the magnetic properties and further expand the application fields of magnetic APCNTs.

The XRD patterns and Raman spectra of APCNTs and MAPCNTs are shown in Figures 1.18 and 1.19, both of which indicate that the MAPCNTs were a mixture of two phases: Fe and CNTs. No other peaks corresponding to impurities were detected.

Figure 1.18 XRD patterns of APCNTs and MAPCNTs.

Figure 1.19 Raman spectra of APCNTs and MAPCNTs.

The XRD patterns of APCNTs and MAPCNTs are shown in Figure 1.18, which indicate that the MAPCNTs were a mixture of two phases: Fe and CNTs. The diffraction peak at 2θ=26.2O is assigned to (002) plane of CNTs, and the diffraction peak at 2θ=44.6O is assigned to (110) plane of Fe and (101) plane of CNTs; the two peaks corresponding to the structure of APCNTs also exist in the XRD pattern of the MAPCNTs. No other peaks corresponding to impurities were detected after NaClO modification.

The Raman spectra of APCNTs and MAPCNTs are shown in Figure 1.19. No peaks have been found in the low frequencies (100~800 cm^–1), which indicates that no iron oxide exists in APCNTs or MAPCNTs. In APCNTs or MAPCNTs samples, because INPs are encapsulated by graphitic shells and iron has only one atom in the primitive unit cell, no optical-branch zero wave vector vibrational modes exist and the Raman peak of iron is difficult to be tested [65]. The G peak at 1,585 cm^–1 is related to E2g graphite mode [12–14]. The strong intensity of this peak indicates good graphitization of CNTs. The D peak at around 1,345 cm^–1 is induced by defective structures, which could include minor amorphous carbon and some graphite particles seen in the sample (Figure 1.16). G′ band and weak structures arising from double resonance processes are observed in the second-order region of the spectra. The intensity ratio of the G and D peaks (IG/ID) is an indicator of the structure quality of CNT sample. As shown in Figure 1.19, the higher IG/ID ratio means a higher structure quality of CNTs after the modification process. Therefore, we can see that MAPCNTs are less defective than APCNTs due to the removal of amorphous carbon after NaClO modification.

The TGA results of all samples are presented in Figure 1.20. The quantity of iron MNPs in the APCNTs can be estimated as 16.38%, which is lower than that in the MAPCNTs (18.62%). The increase of weight fraction of MNPs can be attributed to the purification of carbonaceous byproducts by NaClO solution. More importantly, the results indicate that the MNPs were retained in carbon shells and did not erode or dissolve after NaClO modification. The surface chemical properties of APCNTs and MAPCNTs were further determined by XPS. MAPCNTs sample shows a small O1s peak with an atomic content of 7.15 at.%, which is more than that of 2.88 at.% for APCNTs. This observation reveals that the oxygen-containing functional groups (carboxyls, phenolic, hydroxyls, carbonyls, etc.) were decorated on the wall of MAPCNTs after NaClO modification.

Figure 1.20 Thermal analysis curve of the APCNTs and MAPCNTs.

Next, APCNTs and MAPCNTs were dispersed in water to study the dispersion properties with a concentration of 0.5 mg·mL^–1 by ultrasonic vibration. The optical images show that MAPCNTs could be dispersed in water uniformly after ultrasonication for 15 min, and agglomeration and settling were not observed after several days. In contrast, the APCNTs could not be dispersed even after hours of ultrasonic vibration. Instead, they tended to float on the surface of water or deposit on the bottom of the glass bottle. The good dispersion of MAPCNTs in water could be attributed to hydrophilic functional groups on the surface of MAPCNTs after NaClO treatment. Compared with APCNTs, the total SSA and the PV of MAPCNTs are increased to 186.3 cm²/g and 0.53 cm³/g, which can be attributed to opening up of tube tips after the NaClO treatment.

Part 1

CARBON NANOMATERIALS

Chapter 1

Easy and Large-Scale Synthesis of Carbon Nanotube-Based Adsorbents for the Removal of Arsenic and Organic Pollutants from Aqueous Solutions

Abstract

1.1 Introduction

1.2 Removal of Arsenic from Aqueous Solution

1.2.1 Activated Magnetic Carbon Nanotube

1.2.1.1 Synthesis Method

1.2.1.2 Characterization of Adsorbents

1.2.1.3 Adsorption Properties

1.2.2 Sulfhydryl-Functionalized Magnetic Carbon Nanotube

1.2.2.1 Synthesis Method

1.2.2.2 Characterization of Adsorbents

1.2.2.3 Adsorption Properties

1.3 Removal of Organic Pollutants from Aqueous Solution

1.3.1 Magnetic Carbon Nanotubes for Dye Removal

1.3.1.1 Synthesis Method

1.3.1.2 Characterization of Adsorbents

1.3.1.3 Adsorption Properties