Preparation of Phthalocyanine Immobilized Bacterial Cellulose Nanocomposites for Decoloration of Dye Wastewater: Key Role of Spacers

: We report the preparation of a series of spacer-incorporated, tetra-amino cobalt (II) phthalocyanine (CoPc)-immobilized bacterial cellulose (BC) functional nanocomposites (CoPc@s-BC). Four kinds of ﬂexible spacers with different lengths—diethylenetriamine (DT), triethylenetetramine (TT), tetraethylenepentamine (TP) and pentaethylenehexamine (PH)—were covalently attached onto pre-oxidized BC for the synthesis of the spacer-attached BC, and the attached spacers’ contents were carefully quantiﬁed. Using glutaraldehyde as a cross-linker, the CoPc catalyst was covalently immobilized onto the spacer-attached BC, and the immobilization steps were optimized by monitoring both the residual spacer contents and the resulting immobilized CoPc. All of the functionalization processes were characterized and conﬁrmed by X-ray photoelectron spectroscopy (XPS). The series of spacer-incorporated, CoPc-immobilized BC nanocomposites, CoPc@s-BC, were used for the decoloration of dye wastewater. Both the adsorption capacity and adsorption rate were increased after the incorporation of spacers. When H 2 O 2 was employed as an oxidant, dye molecules were catalytically oxidized with these nanocomposites. Electron paramagnetic resonance (EPR) spin-trapping results showed that the highly reactive hydroxyl radical ( · OH) was involved in the catalytic oxidation process. The spacer length had a direct effect on the catalytic efﬁciency of CoPc@s-BC—the decoloration rate for CoPc@TP-BC was as high as 41 µ mol · min − 1 · g − 1 , which was more than 50% higher than that without spacer.


Introduction
Nanocellulose has spawned increasing interest from broad fields, because this naturally occurring nanomaterial combines the advantages of cellulose, such as being hydrophilic, environmentally benign and having an easily tunable surface, with prominent features of nanosized materials, such as having a very high surface area to volume ratio and considerable modification possibility [1]. The preparation of nanocellulose ranges from "top-down" processing by isolation of natural cellulose to "bottom-up" processing by culture medium with certain bacteria [2,3]. Thethe nanocellulose produced directly from bacteria is well known as bacterial cellulose (BC). Compared with "top-down" produced nanocellulose, BC is highly pure (hemicellulose-and lignin-free) and possesses unique and sophisticated three-dimensional, porous network structures which have been widely used in many fields. Potential applications of BC include mechanically reinforcement nanofillers [4,5], medical products [6], wound treatments [7], bio-inspired nanomaterials [8], and reaction templates [9]. In particular, its well-defined 3D network structures, high surface-area-to-volume-ratio, high accessibility, remarkable functionality, excellent mechanical properties and sustainability suggest

Preparation and Characterization of CoPc@s-BC
To prepare CoPc@s-BC, the following steps were carried out: (1) 20 mg of pure BC was incubated with a 30 mmol/L NaIO4 solution and reacted for 8 h. The resulting oxidized BC was thoroughly washed with ultrapure water and dried at 50 °C for 12 h. The content of aldehyde groups on oxidized BC was determined according to the literatures with minor modifications [42,43]. (2) The dried, oxidized BC was submerged into a spacer solution (DT or TT or TP or PH) and shaken on a temperature-controlled shaker. The content of the attached spacer was indirectly determined by measuring the number of amino groups on the BC substrate using a ninhydrin-based monitoring system, as described in the literature [30]. Ninhydrin reagent can react with primary amines to form a colored complex which is soluble in C2H5OH/H2O mixed solvent and highly conjugated, with a strong absorption at ca. 570 nm. Two hundred milliliters of ninhydrin reagent and 100 μL of H2O were added to 1 mg of spacer-attached BC and heated in a 100 °C water bath for 1 h. Five milliliters of C2H5OH/H2O (50:50, v/v) mixed solvent was added, and the mixture was analyzed on a UV-vis absorption spectrometer (UV-2450) at the wavelength of maximum absorbance: 570 nm. (3) The spacer-attached BC was submerged into 12 mL of glutaraldehyde solution and shaken at 25 °C for 2 h. The glutaraldehyde-activated BC was washed 3 times with ultrapure water and then dried at 50 °C for 12 h. The residual attached spacer content after glutaraldehyde activation was determined with the above-mentioned ninhydrin method. (4) CoPc@s-BC was harvested by submerging the glutaraldehyde-activated BC in a 2 × 10 −2 mol/L CoPc solution and reacted for 2 h at 25 °C. The product was washed 3 times with dimethylformamide, 3 times with ultrapure water and then dried at 50 °C for 12 h. The content of immobilized CoPc was calculated according to the cobalt content on CoPc@s-BC, measured by atomic absorption spectrometry (Thermo solar M6, Thermo Fisher, Waltham, MA, USA). All of the modification and functionalization processes are schematically shown in Figure 2. The elemental compositions and chemical bonding of BC, oxidized BC, TP-

Preparation and Characterization of CoPc@s-BC
To prepare CoPc@s-BC, the following steps were carried out: (1) 20 mg of pure BC was incubated with a 30 mmol/L NaIO 4 solution and reacted for 8 h. The resulting oxidized BC was thoroughly washed with ultrapure water and dried at 50 • C for 12 h. The content of aldehyde groups on oxidized BC was determined according to the literatures with minor modifications [42,43]. (2) The dried, oxidized BC was submerged into a spacer solution (DT or TT or TP or PH) and shaken on a temperature-controlled shaker. The content of the attached spacer was indirectly determined by measuring the number of amino groups on the BC substrate using a ninhydrin-based monitoring system, as described in the literature [30]. Ninhydrin reagent can react with primary amines to form a colored complex which is soluble in C 2 H 5 OH/H 2 O mixed solvent and highly conjugated, with a strong absorption at ca. 570 nm. Two hundred milliliters of ninhydrin reagent and 100 µL of H 2 O were added to 1 mg of spacer-attached BC and heated in a 100 • C water bath for 1 h. Five milliliters of C 2 H 5 OH/H 2 O (50:50, v/v) mixed solvent was added, and the mixture was analyzed on a UV-vis absorption spectrometer (UV-2450) at the wavelength of maximum absorbance: 570 nm.
(3) The spacer-attached BC was submerged into 12 mL of glutaraldehyde solution and shaken at 25 • C for 2 h. The glutaraldehyde-activated BC was washed 3 times with ultrapure water and then dried at 50 • C for 12 h. The residual attached spacer content after glutaraldehyde activation was determined with the above-mentioned ninhydrin method. (4) CoPc@s-BC was harvested by submerging the glutaraldehyde-activated BC in a 2 × 10 −2 mol/L CoPc solution and reacted for 2 h at 25 • C. The product was washed 3 times with dimethylformamide, 3 times with ultrapure water and then dried at 50 • C for 12 h. The content of immobilized CoPc was calculated according to the cobalt content on CoPc@s-BC, measured by atomic absorption spectrometry (Thermo solar M6, Thermo Fisher, Waltham, MA, USA). All of the modification and functionalization processes are schematically shown in Figure 2. The elemental compositions and chemical bonding of BC, oxidized BC, TP-attached BC, glutaraldehyde activated BC and CoPc@TP-BC were analyzed by X-ray photoelectron spectroscopy (XPS). XPS spectra were recorded on a Kratos Axis Ultra XPS system with Al (mono) Kα irradiation (hν = 1486.6 eV). The binding energy peaks of the XPS spectra were calibrated by placing the principal C 1s binding energy peak at 284.6 eV.

Adsorption and Catalytic Oxidation Decoloration
The adsorption of reactive red X-3B was carried out in a glass flask sealed in a water bath at 50 °C. One milligram of CoPc@s-BC was added to 5 mL of reactive red X-3B dye wastewater (100 μmol/L, pH = 2). The catalytic oxidation was initiated by adding 8 mM H2O2 to the CoPc@s-BC containing the reactive red X-3B solution. At given time intervals, the samples were analyzed immediately on a UV-vis absorption spectrometer (UV-2450) at the wavelength of maximum absorbance: 539 nm. The decoloration of reactive red X-3B was expressed as the change in the (C0-C)/C0 value, where C0 is the initial concentration of the dye, and C is the residual concentration of the dye. The amount of adsorbed dye was calculated as follows: The adsorption rate of CoPc@s-BC was derived from the slope of the adsorbed dye-adsorption time curves (within 90 min). The decoloration rate of CoPc@s-BC was calculated using Formula (2): 100 mol / L 5 10 L 90% Decoloration rate ( mol min g ) = t min 1 10 g C 630 g / mol , where t is the time taken for decoloration of 90% of reactive red X-3B, and CCoPc is the immobilized CoPc content of CoPc@s-BC. The EPR signal of radical spin-trapped by DMPO was detected with a Bruker-A300 X-band EPR spectrometer (Bruker, Karlsruhe, Germany).

Preparation and Characterization of CoPc@s-BC
CoPc@s-BC nanocomposites were prepared by covalent immobilization of CoPc onto the nanofibers after BC had been oxidized with NaIO4; this was spacer-attached with a series of flexible spacers and activated with glutaraldehyde. All the functionalization reactions were monitored by Xray photoelectron spectrum (XPS), with the preparation of CoPc@TP-BC as a typical representative (Figure 3), and the chemical compositions of the samples during the series of reactions were calculated and are shown in Table 1. For the as-prepared pure BC, the two characteristic peaks at 284.6 eV and 530.6 eV were ascribed to the binding energies of C 1s and O 1s, respectively ( Figure  3a). No significant changes were found after NaIO4 oxidation (Figure 3b). Upon the TP attachment

Adsorption and Catalytic Oxidation Decoloration
The adsorption of reactive red X-3B was carried out in a glass flask sealed in a water bath at 50 • C. One milligram of CoPc@s-BC was added to 5 mL of reactive red X-3B dye wastewater (100 µmol/L, pH = 2). The catalytic oxidation was initiated by adding 8 mM H 2 O 2 to the CoPc@s-BC containing the reactive red X-3B solution. At given time intervals, the samples were analyzed immediately on a UV-vis absorption spectrometer (UV-2450) at the wavelength of maximum absorbance: 539 nm. The decoloration of reactive red X-3B was expressed as the change in the (C 0 − C)/C 0 value, where C 0 is the initial concentration of the dye, and C is the residual concentration of the dye. The amount of adsorbed dye was calculated as follows: The adsorption rate of CoPc@s-BC was derived from the slope of the adsorbed dye-adsorption time curves (within 90 min). The decoloration rate of CoPc@s-BC was calculated using Formula (2): where t is the time taken for decoloration of 90% of reactive red X-3B, and C CoPc is the immobilized CoPc content of CoPc@s-BC. The EPR signal of radical spin-trapped by DMPO was detected with a Bruker-A300 X-band EPR spectrometer (Bruker, Karlsruhe, Germany).

Preparation and Characterization of CoPc@s-BC
CoPc@s-BC nanocomposites were prepared by covalent immobilization of CoPc onto the nanofibers after BC had been oxidized with NaIO 4 ; this was spacer-attached with a series of flexible spacers and activated with glutaraldehyde. All the functionalization reactions were monitored by  Table 1. For the as-prepared pure BC, the two characteristic peaks at 284.6 eV and 530.6 eV were ascribed to the binding energies of C 1s and O 1s, respectively ( Figure 3a). No significant changes were found after NaIO 4 oxidation (Figure 3b). Upon the TP attachment process, an additional peak at ca. 400 eV was detected, which corresponds to the characteristic peak of N 1s, implying the successful attachment of the TP spacer onto the oxidized BC ( Figure 3c). The glutaraldehyde activation step did not obviously change the relative fractions of C, O and N elements ( Figure 3d and Table 1). After the CoPc functionalization process, a marked increase (from 6.91% to 19%) of N 1s peak was found ( Figure 3e and Table 1); Furthermore, two new characteristic peaks at 779.6 eV and 795.3 eV were also detected (Figure 3e, inset). These two peaks were assigned to the binding energies of Co 2p 3/2 and Co 2p 1/2 , respectively. These results verify the success of all of the functionalization processes and the immobilization of CoPc on BC. process, an additional peak at ca. 400 eV was detected, which corresponds to the characteristic peak of N 1s, implying the successful attachment of the TP spacer onto the oxidized BC ( Figure 3c). The glutaraldehyde activation step did not obviously change the relative fractions of C, O and N elements ( Figure 3d and Table 1). After the CoPc functionalization process, a marked increase (from 6.91% to 19%) of N 1s peak was found ( Figure 3e and Table 1); Furthermore, two new characteristic peaks at 779.6 eV and 795.3 eV were also detected ( Figure 3e, inset). These two peaks were assigned to the binding energies of Co 2p3/2 and Co 2p1/2, respectively. These results verify the success of all of the functionalization processes and the immobilization of CoPc on BC.

Optimization of CoPc Immobilization Conditions
For real applications, the optimum CoPc immobilization content on the BC substrate is desired to enhance its catalytic reaction efficiency without increasing the cost. The immobilized CoPc content directly depends on the amount of attached spacer; hence, the reaction condition for spacer attachment should be carefully monitored prior to CoPc immobilization.
The attachment levels of different spacers on the oxidized BC substrate as a function of reaction time is illustrated in Figure 4. The attached amount of all spacers increased almost directly proportionally with time, from 0 to 4 h, suggesting that the spacer attachment rate was initially constant. It is interesting that there appears to be a direct correlation between the attachment level and the spacer length, with longer spacers having lower amounts of spacer attached. The difference in attachment level increased with time. After reacting for 10 h, the attached amounts of TT, TP, and PH spacers were ~2.2 mmol/g, ~2 mmol/g and ~1.7 mmol/g, which were ca. 10%, 15% and 30% lower than that of the DT spacer (~2.4 mmol/g), respectively. Further prolongation of the reaction time did not consistently increase the attached spacer content on oxidized BC. Considering both the

Optimization of CoPc Immobilization Conditions
For real applications, the optimum CoPc immobilization content on the BC substrate is desired to enhance its catalytic reaction efficiency without increasing the cost. The immobilized CoPc content directly depends on the amount of attached spacer; hence, the reaction condition for spacer attachment should be carefully monitored prior to CoPc immobilization.
The attachment levels of different spacers on the oxidized BC substrate as a function of reaction time is illustrated in Figure 4. The attached amount of all spacers increased almost directly proportionally with time, from 0 to 4 h, suggesting that the spacer attachment rate was initially constant. It is interesting that there appears to be a direct correlation between the attachment level and the spacer length, with longer spacers having lower amounts of spacer attached. The difference in attachment level increased with time. After reacting for 10 h, the attached amounts of TT, TP, and PH spacers werẽ 2.2 mmol/g,~2 mmol/g and~1.7 mmol/g, which were ca. 10%, 15% and 30% lower than that of the DT spacer (~2.4 mmol/g), respectively. Further prolongation of the reaction time did not consistently increase the attached spacer content on oxidized BC. Considering both the attachment level and the attachment efficiency, the optimum attachment time was set to 10 h.
Our previous work illustrated that under optimum oxidation conditions, oxidized BC contains 14.13% (wt./wt.) of aldehyde groups (ca. 5 mmol/g, slightly lower than the theoretical maximum content of 18.1% (wt./wt.)) [25], while the amount of attached spacer was lower than 2.5 mmol/g; that is, less than half of the aldehyde groups were effectively utilized for the attachment of spacers. The relatively low-efficiency usage of aldehyde groups can be ascribed to two reasons. The first reason is the difficulty of producing heterogeneous reactions and the incompleteness characteristic of macromolecular reactions. The second reason may be attributed, at least partly, to the steric hindrance effect; that is, when the amount of attached spacer increased to a certain level, the already fixed spacers prevented further progress of the attachment reaction. It is reasonable to assume that the most obvious steric hindrance effect arises from PH, i.e., the longest spacer. Given the above factors, the attachment levels for all of the spacers were much smaller than the theoretical value, and the PH spacer had the lowest attached content. The spacer-attached BC substrates were further activated with glutaraldehyde, a classic bifunctional chemical crosslinker, for the immobilization of the CoPc catalyst. The immobilized CoPc contents were affected by many factors, such as the contents of the attached spacers and the activation levels. Figure 5a demonstrates the influence of initial glutaraldehyde concentration on the immobilized CoPc content for DT-attached BC. The immobilized CoPc content increased in association with the increase of glutaraldehyde concentration. It is reasonable to assume that with sufficient activation reagents, more binding sites were provided for CoPc immobilization. Meanwhile, the amount of DT spacer decreased drastically, which statistically represents the successfully activation of glutaraldehyde for CoPc immobilization. Two percent glutaraldehyde was required for the maximum immobilization of CoPc; accordingly, the content of immobilized CoPc was as high as ca. 550 μmol/g, and the residual DT spacer was 0.38 mmol/g. A further increase in the activation reagent did not lead to the disappearance of the DT spacer (even when the glutaraldehyde concentration reached 5%), which may be ascribed to the incompleteness of the macromolecular reaction. Similar results were found for the immobilization of CoPc onto TT-attached BC (Figure 5b), TP-attached BC (Figure 5c) and PH-attached BC (Figure 5d). Our previous work illustrated that under optimum oxidation conditions, oxidized BC contains 14.13% (wt./wt.) of aldehyde groups (ca. 5 mmol/g, slightly lower than the theoretical maximum content of 18.1% (wt./wt.)) [25], while the amount of attached spacer was lower than 2.5 mmol/g; that is, less than half of the aldehyde groups were effectively utilized for the attachment of spacers. The relatively low-efficiency usage of aldehyde groups can be ascribed to two reasons. The first reason is the difficulty of producing heterogeneous reactions and the incompleteness characteristic of macromolecular reactions. The second reason may be attributed, at least partly, to the steric hindrance effect; that is, when the amount of attached spacer increased to a certain level, the already fixed spacers prevented further progress of the attachment reaction. It is reasonable to assume that the most obvious steric hindrance effect arises from PH, i.e., the longest spacer. Given the above factors, the attachment levels for all of the spacers were much smaller than the theoretical value, and the PH spacer had the lowest attached content.
The spacer-attached BC substrates were further activated with glutaraldehyde, a classic bifunctional chemical crosslinker, for the immobilization of the CoPc catalyst. The immobilized CoPc contents were affected by many factors, such as the contents of the attached spacers and the activation levels. Figure 5a demonstrates the influence of initial glutaraldehyde concentration on the immobilized CoPc content for DT-attached BC. The immobilized CoPc content increased in association with the increase of glutaraldehyde concentration. It is reasonable to assume that with sufficient activation reagents, more binding sites were provided for CoPc immobilization. Meanwhile, the amount of DT spacer decreased drastically, which statistically represents the successfully activation of glutaraldehyde for CoPc immobilization. Two percent glutaraldehyde was required for the maximum immobilization of CoPc; accordingly, the content of immobilized CoPc was as high as ca. 550 µmol/g, and the residual DT spacer was 0.38 mmol/g. A further increase in the activation reagent did not lead to the disappearance of the DT spacer (even when the glutaraldehyde concentration reached 5%), which may be ascribed to the incompleteness of the macromolecular reaction. Similar results were found for the immobilization of CoPc onto TT-attached BC (Figure 5b)

Adsorption and Catalytic Oxidation Performance of CoPc@s-BC
The prepared series of CoPc@s-BC were aimed at being functional nanocomposites for the decoloration of dye wastewater. The total decoloration includes the adsorption of dye molecules and the subsequent catalytic oxidation process [40]. The performances of the prepared CoPc@s-BC nanocomposites were firstly evaluated based on their adsorption behavior towards reactive red X-3B dye wastewater ( Figure 6). For pure BC, only 0.05 μmol of the dye molecules were adsorbed after a dynamic equilibrium, reached in 240 min. CoPc@BC had a much higher adsorption capacity than pure BC-0.2 μmol of dye was adsorbed by CoPc@BC under the same experimental conditions. The adsorption capacities of all the spacer-attached nanocomposites were higher when compared with CoPc@BC. After dynamic equilibrium was reached, the amounts of the adsorbed dye by CoPc@DT-BC, CoPc@TT-BC, CoPc@TP-BC and CoPc@PH-BC were 37%, 54%, 67% and 17% higher than that of CoPc@BC, respectively. As illustrated in Figure 5, the contents of immobilized CoPc for CoPc@DT-BC, CoPc@TT-BC, CoPc@TP-BC and CoPc@PH-BC were 27%, 21%, 15% and 7% higher than that of CoPc@BC, respectively. By comparing these statistics, it is easy to deduce that the enhancement of adsorption capacity was mainly caused by the incorporation of spacers. The incorporation of spacers caused CoPc to immobilize onto BC in a more dispersed way, which effectively increased the number of contact opportunities between CoPc and the dye molecules; furthermore, the diffusion limitation of dye molecules to CoPc was effectively reduced. These effects were correlated with the spacer length. CoPc@TP-BC has a longer spacer than CoPc@DT-BC and CoPc@TT-BC, and its adsorption capacity was much higher. However, it is possible that when the spacer was too long, a CoPc might have attached onto BC through multiple-bonding, and the peripheral amino groups on CoPc (which was necessary for proton acceptance) [40] may have been excessively consumed; thus the adsorption capacity may have been reduced to some extent. The CoPc@PH-BC has the longest incorporated spacer, while its adsorption capacity was relatively lower than others.

Adsorption and Catalytic Oxidation Performance of CoPc@s-BC
The prepared series of CoPc@s-BC were aimed at being functional nanocomposites for the decoloration of dye wastewater. The total decoloration includes the adsorption of dye molecules and the subsequent catalytic oxidation process [40]. The performances of the prepared CoPc@s-BC nanocomposites were firstly evaluated based on their adsorption behavior towards reactive red X-3B dye wastewater ( Figure 6). For pure BC, only 0.05 µmol of the dye molecules were adsorbed after a dynamic equilibrium, reached in 240 min. CoPc@BC had a much higher adsorption capacity than pure BC-0.2 µmol of dye was adsorbed by CoPc@BC under the same experimental conditions. The adsorption capacities of all the spacer-attached nanocomposites were higher when compared with CoPc@BC. After dynamic equilibrium was reached, the amounts of the adsorbed dye by CoPc@DT-BC, CoPc@TT-BC, CoPc@TP-BC and CoPc@PH-BC were 37%, 54%, 67% and 17% higher than that of CoPc@BC, respectively. As illustrated in Figure 5, the contents of immobilized CoPc for CoPc@DT-BC, CoPc@TT-BC, CoPc@TP-BC and CoPc@PH-BC were 27%, 21%, 15% and 7% higher than that of CoPc@BC, respectively. By comparing these statistics, it is easy to deduce that the enhancement of adsorption capacity was mainly caused by the incorporation of spacers. The incorporation of spacers caused CoPc to immobilize onto BC in a more dispersed way, which effectively increased the number of contact opportunities between CoPc and the dye molecules; furthermore, the diffusion limitation of dye molecules to CoPc was effectively reduced. These effects were correlated with the spacer length. CoPc@TP-BC has a longer spacer than CoPc@DT-BC and CoPc@TT-BC, and its adsorption capacity was much higher. However, it is possible that when the spacer was too long, a CoPc might have attached onto BC through multiple-bonding, and the peripheral amino groups on CoPc (which was necessary for proton acceptance) [40] may have been excessively consumed; thus the adsorption capacity may have been reduced to some extent. The CoPc@PH-BC has the longest incorporated spacer, while its adsorption capacity was relatively lower than others. To determine the adsorption rate of CoPc@s-BC, the adsorption curves in Figure 6 were regressed using a linear function (within 90 min); the parameters, including the slope and correlation ratio, are presented in Table 2. The amount of adsorbed dye as a function of adsorption time showed a linear trend; all correlation ratios (r) for the CoPc@s-BC nanocomposites were >0.99. The incorporation of spacers greatly facilitated the contact between the CoPc and the dye molecules-all of the CoPc@s-BC had higher adsorption rates than CoPc@BC. The adsorption rate of CoPc@TP-BC reached 0.385 μmol/min, which was more than two times higher than that of CoPc@BC (0.172 μmol/min). One of the main concerns for this study was to evaluate the effects of the spacers, especially the spacer length, on the catalytic activity of the CoPc catalyst. All series of CoPc@s-BC were employed for the catalytic oxidation of reactive red X-3B dye wastewater, with H2O2 as an oxidant. The investigation of the influence of spacers on the catalytic oxidation rate revealed similar results to that of adsorption behavior. A control experiment was firstly conducted. H2O2 alone showed almost no decoloration activity-it was difficult to decompose reactive red X-3B with H2O2 without CoPc@s-BC ( Figure 7). All of the prepared CoPc@s-BC nanocomposites are able to catalytic decolorize dye wastewater, and the decoloration rate was correlation with the spacer length (Figure 7, inset). The typical optical image changes in reactive red X-3B solution during the decoloration process are shown in Figure 8. For CoPc@BC, the decoloration rate was ca. 27 μmol·min −1 ·g −1 . With the incorporation of the spacer, enhanced catalytic activity of CoPc catalytst was observed. Under the same reaction conditions, the decoloration rate of CoPc@DT-BC was ca. 8% higher than that of CoPc@BC. Much higher catalytic activity of CoPc was obtained with the incorporation of a longer spacer length. The To determine the adsorption rate of CoPc@s-BC, the adsorption curves in Figure 6 were regressed using a linear function (within 90 min); the parameters, including the slope and correlation ratio, are presented in Table 2. The amount of adsorbed dye as a function of adsorption time showed a linear trend; all correlation ratios (r) for the CoPc@s-BC nanocomposites were >0.99. The incorporation of spacers greatly facilitated the contact between the CoPc and the dye molecules-all of the CoPc@s-BC had higher adsorption rates than CoPc@BC. The adsorption rate of CoPc@TP-BC reached 0.385 µmol/min, which was more than two times higher than that of CoPc@BC (0.172 µmol/min). One of the main concerns for this study was to evaluate the effects of the spacers, especially the spacer length, on the catalytic activity of the CoPc catalyst. All series of CoPc@s-BC were employed for the catalytic oxidation of reactive red X-3B dye wastewater, with H 2 O 2 as an oxidant. The investigation of the influence of spacers on the catalytic oxidation rate revealed similar results to that of adsorption behavior. A control experiment was firstly conducted. H 2 O 2 alone showed almost no decoloration activity-it was difficult to decompose reactive red X-3B with H 2 O 2 without CoPc@s-BC (Figure 7). All of the prepared CoPc@s-BC nanocomposites are able to catalytic decolorize dye wastewater, and the decoloration rate was correlation with the spacer length (Figure 7, inset). The typical optical image changes in reactive red X-3B solution during the decoloration process are shown in Figure 8.
For CoPc@BC, the decoloration rate was ca. 27 µmol·min −1 ·g −1 . With the incorporation of the spacer, enhanced catalytic activity of CoPc catalytst was observed. Under the same reaction conditions, the decoloration rate of CoPc@DT-BC was ca. 8% higher than that of CoPc@BC. Much higher catalytic activity of CoPc was obtained with the incorporation of a longer spacer length. The decoloration rates of CoPc@TT-BC and CoPc@TP-BC were ca. 33 µmol·min −1 ·g −1 and 41 µmol·min −1 ·g −1 , which were more than 20% and 50% higher than that of CoPc@BC, respectively. This result is in good accordance with our principle idea-that the incorporation of a flexible spacer is able to keep the CoPc catalyst at a preferred distance from the solid substrate, which, to some extent, mitigates the adverse effects of diffusion limitation and steric hindrance, increases the reaction homogeneity and enhances the accessibility and mobility of the CoPc catalyst. These effects are more likely to occur with longer spacer lengths. However, when the spacer is too long, it is possible that the CoPc molecules covalently immobilize through multiple-bonding, which restricts the free movement of the CoPc and thus, decreases its catalytic activity to some extent. The decoloration rate for CoPc@PH-BC was ca. 33 µmol·min −1 ·g −1 , which was higher than CoPc@BC but lower than CoPc@TP-BC. to keep the CoPc catalyst at a preferred distance from the solid substrate, which, to some extent, mitigates the adverse effects of diffusion limitation and steric hindrance, increases the reaction homogeneity and enhances the accessibility and mobility of the CoPc catalyst. These effects are more likely to occur with longer spacer lengths. However, when the spacer is too long, it is possible that the CoPc molecules covalently immobilize through multiple-bonding, which restricts the free movement of the CoPc and thus, decreases its catalytic activity to some extent. The decoloration rate for CoPc@PH-BC was ca. 33 μmol·min −1 ·g −1 , which was higher than CoPc@BC but lower than CoPc@TP-BC.  To sum up, the most preferable spacer length to achieve the maximum decoloration efficiency of CoPc@s-BC was derived from the TP spacer. Under optimum immobilization conditions, 2 mmol/g of attached TP together with 495 μmol/g of immobilized CoPc were obtained, and the decoloration rate for the resulting CoPc@TP-BC reached 41 μmol·min −1 ·g −1 . These results are comparable with related literature. Firstly, flexible spacer-attached BCs were employed to support the CoPc catalyst, which permitted a much higher amount of immobilized CoPc than other materials [44,45]. Secondly, the incorporation of the spacers increased both the adsorption rate and the adsorption capacity of the nanocomposites [46]. Moreover, the incorporation of the spacers (especially for TP) effectively increased the accessibility and reaction homogeneity of the CoPc catalyst, and thus, its catalytic to keep the CoPc catalyst at a preferred distance from the solid substrate, which, to some extent, mitigates the adverse effects of diffusion limitation and steric hindrance, increases the reaction homogeneity and enhances the accessibility and mobility of the CoPc catalyst. These effects are more likely to occur with longer spacer lengths. However, when the spacer is too long, it is possible that the CoPc molecules covalently immobilize through multiple-bonding, which restricts the free movement of the CoPc and thus, decreases its catalytic activity to some extent. The decoloration rate for CoPc@PH-BC was ca. 33 μmol·min −1 ·g −1 , which was higher than CoPc@BC but lower than CoPc@TP-BC.  To sum up, the most preferable spacer length to achieve the maximum decoloration efficiency of CoPc@s-BC was derived from the TP spacer. Under optimum immobilization conditions, 2 mmol/g of attached TP together with 495 μmol/g of immobilized CoPc were obtained, and the decoloration rate for the resulting CoPc@TP-BC reached 41 μmol·min −1 ·g −1 . These results are comparable with related literature. Firstly, flexible spacer-attached BCs were employed to support the CoPc catalyst, which permitted a much higher amount of immobilized CoPc than other materials [44,45]. Secondly, the incorporation of the spacers increased both the adsorption rate and the adsorption capacity of the nanocomposites [46]. Moreover, the incorporation of the spacers (especially for TP) effectively To sum up, the most preferable spacer length to achieve the maximum decoloration efficiency of CoPc@s-BC was derived from the TP spacer. Under optimum immobilization conditions, 2 mmol/g of attached TP together with 495 µmol/g of immobilized CoPc were obtained, and the decoloration rate for the resulting CoPc@TP-BC reached 41 µmol·min −1 ·g −1 . These results are comparable with related literature. Firstly, flexible spacer-attached BCs were employed to support the CoPc catalyst, which permitted a much higher amount of immobilized CoPc than other materials [44,45]. Secondly, the incorporation of the spacers increased both the adsorption rate and the adsorption capacity of the nanocomposites [46]. Moreover, the incorporation of the spacers (especially for TP) effectively increased the accessibility and reaction homogeneity of the CoPc catalyst, and thus, its catalytic activity was greatly improved compared to that without the spacer [23].
The EPR spin-trapping technique is a powerful tool for the detection of short-lived, active species. Herein, the 5,5-dimethyl-1-pyrroline-N-oxide (DMPO)-trapped EPR spectra were employed to demonstrate the formation of radicals during the catalytic oxidation of reactive red X-3B with the CoPc@s-BC (CoPc@BC)/H 2 O 2 reaction system ( Figure 9). Without the existence of H 2 O 2 , no obvious EPR signal was detected. When both CoPc@BC and H 2 O 2 were added, a strong, four-line spectrum with a peak intensity of 1:2:2:1 was easily detected, which is the typical characteristic spectrum of the DMPO-·OH adducts [47]. This result indicates that ·OH was generated during the reaction and was responsible for the catalytic oxidation of reactive red X-3B. Similar results were found for the CoPc@s-BC/H 2 O 2 reaction system; the only differences were the peak intensities, which were proportional to the amount of the formed ·OH radical during the reaction. An obvious higher amount of ·OH was observed for the CoPc@TP-BC/H 2 O 2 reaction system. Therefore, the CoPc@TP-BC had the highest catalytic activity for the decoloration of reactive red X-3B dye wastewater, which is in good accordance with the results illustrated in Figure 7. to demonstrate the formation of radicals during the catalytic oxidation of reactive red X-3B with the CoPc@s-BC (CoPc@BC)/H2O2 reaction system (Figure 9). Without the existence of H2O2, no obvious EPR signal was detected. When both CoPc@BC and H2O2 were added, a strong, four-line spectrum with a peak intensity of 1:2:2:1 was easily detected, which is the typical characteristic spectrum of the DMPO-·OH adducts [47]. This result indicates that ·OH was generated during the reaction and was responsible for the catalytic oxidation of reactive red X-3B. Similar results were found for the CoPc@s-BC/H2O2 reaction system; the only differences were the peak intensities, which were proportional to the amount of the formed ·OH radical during the reaction. An obvious higher amount of ·OH was observed for the CoPc@TP-BC/H2O2 reaction system. Therefore, the CoPc@TP-BC had the highest catalytic activity for the decoloration of reactive red X-3B dye wastewater, which is in good accordance with the results illustrated in Figure 7. Based on the above experimental results, a possible catalytic reaction mechanism is proposed in Figure 10. Firstly, the reactive red X-3B dye molecules were adsorbed onto the nanocomposites from solution. Thanks to the incorporation of flexible spacers, the contact between dye molecules and the CoPc pendants became easier; therefore, both the adsorption capacity and the adsorption rate were enhanced by the spacers. Secondly, the adsorbed dyes were catalytically decolorized in-situ by the ·OH radical. Due to the presence of spacers, the free movement and the reaction homogeneity of CoPc on CoPc@s-BC were improved; thus, more ·OH is produced during the oxidation of dye molecules. These two processes cooperate in a synergistic manner and effective enhancement of the catalytic efficiency of CoPc can be reasonably expected. The decoloration was significantly associated with the spacer length of CoPc@s-BC. The most preferable spacer length arose from TP; thus, CoPc@TP-BC has the highest catalytic efficiency. Magnetic field (G)  Based on the above experimental results, a possible catalytic reaction mechanism is proposed in Figure 10. Firstly, the reactive red X-3B dye molecules were adsorbed onto the nanocomposites from solution. Thanks to the incorporation of flexible spacers, the contact between dye molecules and the CoPc pendants became easier; therefore, both the adsorption capacity and the adsorption rate were enhanced by the spacers. Secondly, the adsorbed dyes were catalytically decolorized in-situ by the ·OH radical. Due to the presence of spacers, the free movement and the reaction homogeneity of CoPc on CoPc@s-BC were improved; thus, more ·OH is produced during the oxidation of dye molecules.
These two processes cooperate in a synergistic manner and effective enhancement of the catalytic efficiency of CoPc can be reasonably expected. The decoloration was significantly associated with the spacer length of CoPc@s-BC. The most preferable spacer length arose from TP; thus, CoPc@TP-BC has the highest catalytic efficiency.

Conclusions
BCs attached with varying lengths of flexible spacers were prepared for the immobilization of CoPc and the fabrication of novel, spacer-incorporated, CoPc-immobilized CoPc@s-BC nanocomposites. These nanocomposites are promising materials for the decoloration of reactive dye wastewater. The incorporation of spacers effectively increased the decoloration efficiency of CoPc@s-BC. The most preferable spacer length for optimizing both the adsorption capacity and the catalytic oxidation efficiency was derived from the TP spacer. With the incorporation of TP, the adsorption capacity of CoPc@TP-BC was 67% higher than that of CoPc@BC, and its adsorption rate increased by more than two times. The CoPc@TP-BC was able to efficiently decolorize reactive dye wastewater with H 2 O 2 as an oxidant, and the decoloration rate was as high as 41 µmol·min −1 ·g −1 , ca. 50% higher than that of CoPc@BC. EPR spin-trapping experiments revealed that highly reactive ·OH is responsible for the catalytic oxidation reaction. The incorporation of spacers boosts the formation of ·OH and thus increases the catalytic activity of CoPc@s-BC.