Kinetic and Mechanistic Study on Catalytic Decomposition of Hydrogen Peroxide on Carbon-Nanodots/Graphitic Carbon Nitride Composite

The metal-free CDots/g-C3N4 composite, normally used as the photocatalyst in H2 generation and organic degradation, can also be applied as an environmental catalyst by in-situ production of strong oxidant hydroxyl radical (HO·) via catalytic decomposition of hydrogen peroxide (H2O2) without light irradiation. In this work, CDots/g-C3N4 composite was synthesized via an electrochemical method preparing CDots followed by the thermal polymerization of urea. Transmission electron microscopy (TEM), X-Ray diffraction (XRD), Fourier Transform Infrared (FTIR), N2 adsorption/desorption isotherm and pore width distribution were carried out for characterization. The intrinsic catalytic performance, including kinetics and thermodynamic, was studied in terms of catalytic decomposition of H2O2 without light irradiation. The second-order rate constant of the reaction was calculated to be (1.42 ± 0.07) × 10−9 m·s−1 and the activation energy was calculated to be (29.05 ± 0.80) kJ·mol−1. Tris(hydroxymethyl) aminomethane (Tris) was selected to probe the produced HO· during the decomposing of H2O2 as well as to buffer the pH of the solution. The composite was shown to be base-catalyzed and the optimal performance was achieved at pH 8.0. A detailed mechanism involving the adsorb-catalyze double reaction site was proposed. Overall, CDots/g-C3N4 composite can be further applied in advanced oxidation technology in the presence of H2O2 and the instinct dynamics and the mechanism can be referred to further applications in related fields.


Introduction
Advanced oxidation technology (AOT) is one of the most effective and economical approaches dealing with non-biodegradable organic pollutants (NBDOPs) in water, such as dyestuffs, pesticides, pharmaceutical and personal care products (PPCPs), synthetic chemicals and leachate of landfills [1][2][3][4][5].In typical AOTs, different strategies like chemical, photochemical, sonochemical and electrochemical pathways, are employed to produce intermediate active oxidant radicals [1,[6][7][8].With an oxidation potential of 2.7 eV and nanosecond-level life time, hydroxyl radical (HO•) is one of the most typical radicals, which can decompose NBDOPs non-selectively, forming CO 2 , H 2 O, inorganic ions or other biodegradable molecules [9][10][11].It is worthy to note that the degradation of NBDOPs and the generation of HO• take place simultaneously [12].Thus, the core process of various AOTs is to improve the yield of HO•, which mainly leads to the decomposition of NBDOPs.
The concentration of instantaneous HO• can be hardly determined directly but can be determined indirectly by probes like Rhodamine B [13], terephthalic acid [14], dimethyl sulfoxide [15], phenylalanine [16] and Tris(hydroxymethyl)aminomethane (Tris) [17].Among the various probes, Tris can be applied in both homogeneous and heterogeneous systems as HO• scavenger and pH buffer at the same time [18].Hydroxyl radical captures hydrogen atom from Tris, producing formaldehyde (CH 2 O) and other compounds.Since the produced CH 2 O can be quantified by the modified Hantzsch method [19], the concentration of HO• can be indirectly quantified.The detailed mechanism of the reaction between HO• and Tris was reported, involving the effects of O 2 and pH [17].
H 2 O 2 is one of the most common sources of HO• in the presence of metal salt solution, carbon-based species, metal or metal oxide via Fenton/Fenton-like reaction, electron-transfer mechanism or catalytic decomposition on the solid-liquid interface [18,[20][21][22][23][24].The well-known Fenton/Fenton-like reaction may occur in both homogeneous and heterogeneous system according to many works [18,[20][21][22][23]. Similar to the Fenton reaction, HO• and HO 2 • can be formed on the surface of carbon-based catalysts via the electron-transfer mechanism due to the donor-acceptor properties of the carbon surface.The redox cycle is necessary to keep the production of HO• and HO 2 • species [24].The catalytic decomposition of H 2 O 2 on the surface of metal or metal oxides has also been studied to some extent in recent years, including Fe, W, Cu, UO 2 , ZrO 2 , CuO, CuO 2 and so on [18,22,[25][26][27].It is known that HO• and HO 2 • will be formed as intermediates during the decomposition of H 2 O 2 while the disproportion reaction of HO 2 • ends up with H 2 O 2 and O 2 .The disproportionation may also occur in the Fenton/Fenton-like reaction and the reaction between H 2 O 2 and carbon-based catalyst.From previous work, it is known that the reaction between scavenger and HO• will affect the production of O 2 [28].
Despite its high efficiency and effectiveness, the application of classic Fenton reaction faces the disadvantages of strict pH restrictions, iron precipitation and the cost for catalyst recycling [29][30][31].The formation rate of HO• is strongly dependent on the pH value while the oxidation potential of HO• declines as the pH increases [31,32].Furthermore, the generation of HO• is directly limited by the formation of iron sludge in alkaline condition [30].Since iron precipitation remains the bottleneck of classic iron-based Fenton reaction, non-ferrous heterogeneous catalysts with multiple oxidation states and redox stability (Ce, Cu, Mn and Ru) [11] and transition metal substituted iron oxide (Cr, Co and Ti) [32], have been developed for the replacement.Nevertheless, the abovementioned metal materials still face the drawbacks of high cost, high toxicity and/or environmental unfriendliness.Hence, a number of metal-free catalysts have been developed for the generation and/or decomposition of H 2 O 2 regarding to their high earth abundance, good biocompatibility and environment-friendly properties, including graphene [6], carbon nanotubes [33], activated carbon fibers [34], graphitic carbon nitride (g-C 3 N 4 ) [35][36][37][38] and carbon nanodots (CDots) [39].
As a metal-free polymer semiconductor material with suitable band gap and band position, g-C 3 N 4 has embodied its research value in the field of H 2 production, CO 2 reduction, selective oxidation of alcohols and pollutant degradation [40][41][42][43].The combination of CDots and g-C 3 N 4 was firstly introduced in 2015 by J. Liu and her co-workers for water splitting, solving the chock point that g-C 3 N 4 is poisoned by in-situ generated H 2 O 2 in hydrogen evolution [44].H 2 O can be catalytically split into H 2 O 2 and H 2 by g-C 3 N 4 in the presence of photo irradiation.However, with the two-dimensional structure and large accessible area on the surface of g-C 3 N 4 , the in-situ generated hydrogen peroxides are strongly bonded and difficult to remove, which leads to the poisoning of catalyst, thereby limiting the yield of H 2 [45][46][47].CDots was introduced to solve this problem by decomposing the bonded H 2 O 2 on the surface of g-C 3 N 4 into H 2 O and O 2 , thereby remitting the poisoning of g-C 3 N 4 .It is known that intermediate HO• will be formed via electron-transfer on the surface of carbon-based catalyst [24].Inspired from these, it can be hypothesized that the CDots/g-C 3 N 4 composite can be used as a catalyst providing promising yield of HO• via decomposing adsorbed H 2 O 2 on the surface of g-C 3 N 4 by embedded CDots.To the best of our knowledge, the kinetics and mechanism of catalytic decomposition of H 2 O 2 on CDots/g-C 3 N 4 composite has been rarely studied.
In this work, CDots/g-C 3 N 4 composite was synthesized via an electrochemical method followed by a thermal polymerization process.The obtained composites were characterized by TEM, FTIR, Brunauer-Emmett-Teller (B.E.T) and XRD.The catalytic performance of CDots/g-C 3 N 4 composite for H 2 O 2 decomposition was also investigated.The second-order reaction rate constant of H 2 O 2 decomposition and reaction activation energy were obtained by varying the dosage of composite and temperature.Furthermore, a detailed mechanism involving the adsorb-catalyze double reaction sites was proposed.

Morphology of the Catalyst
The obtained CDots/g-C 3 N 4 composite was prepared via an electrochemical method followed by a thermal polymerization process.To confirm the modification of CDots on g-C 3 N 4 , FTIR spectra and XRD patterns of pure g-C 3 N 4 and CDots/g-C 3 N 4 composite were obtained and exhibited in Figure 1A In this work, CDots/g-C3N4 composite was synthesized via an electrochemical method followed by a thermal polymerization process.The obtained composites were characterized by TEM, FTIR, Brunauer-Emmett-Teller (B.E.T) and XRD.The catalytic performance of the CDots/g-C3N4 composite for the H2O2 decomposition was also investigated.The second-order reaction rate constant of H2O2 decomposition and reaction activation energy were obtained by varying the dosage of composite and temperature.Furthermore, a detailed mechanism involving the adsorb-catalyze double reaction sites was proposed.

Morphology of the Catalyst
The obtained CDots/g-C3N4 composite was prepared via an electrochemical method followed by a thermal polymerization process.To confirm the modification of CDots on g-C3N4, FTIR spectra and XRD patterns of pure g-C3N4 and CDots/g-C3N4 composite were obtained and exhibited in Figure 1A,B.The influence of CDots modification on the specific surface area was investigated by the B.E.T. method with isothermal adsorption and desorption of high purity nitrogen.The N2 adsorptiondesorption isotherms and pore size distributions of g-C3N4 and CDots/g-C3N4 composite are shown in Figure 1C.The TEM images of CDots/g-C3N4 are shown in Figure 1D,E.As can be seen in Figure 1A, the sharp peak for g-C3N4 at 810 cm −1 is attributed to stretching vibration bond of tri-s-triazine [48].Vibration peaks between 1200-1650 cm −1 corresponds to the typical stretching modes of CN heterocycles [49].A wider band can be seen at 3100-3300 cm −1 , which belongs to the stretching vibration modes for the unreacted -NH [50].The same characteristic peaks are observed in CDots/g-C3N4 composite and the peak at 1405 cm −1 can be seen as the indicator for the coupling of CDots and g-C3N4 [48].
XRD patterns of g-C3N4 and CDots/g-C3N4 are displayed in Figure 1B.The main diffraction peaks observed at 12.9° and 27.5° in both g-C3N4 and CDots/g-C3N4 composite are indexed to the (100) peak of the in-plane structure of tri-s-triszine unit and (002) crystal facets of the inter-layer stacking of aromatic segments [51,52].The two patterns fit well with graphite carbon nitride (JCPDS 87-1526) and no significant difference is observed, implying the low content of CDots in CDots/g-C3N4 composite.However, it is remarkable that the difference in relative intensity, together with the shift observed in the (002) peak location from 27.51° for g-C3N4 to 27.59° for CDots/g-C3N4, can be seen as the evidence of CDots introduction [45].As can be seen in Figure 1C, the introduction of CDots in CDots/g-C3N4 leads to a 20% increase (from 120.92 to 145.24 m 2 /g) in specific surface area, which favors the decomposition of H2O2.
Figure 1D clearly shows the two-dimensional structure of g-C3N4 together with the embedding of CDots (the white circles).The close look of the CDots embedded in g-C3N4 matrix is given in Figure 1E.The CDots are non-uniformly distributed, ranging from 2 to 10 nm, which is in line with previous studies [45,51,53].
From the results and analysis above, it can be confirmed that CDots have been successfully decorated in g-C3N4 and the inlay of CDots brings an improvement in specific surface area of the composite.

Kinetic Study
The effect of CDots content in the catalyst has been investigated in several previous works proving that a certain amount of CDots can enhance the catalytic properties of the catalysts while an excessive loading may work opposite [45,51,53].Thus, in this work the CDots/g-C3N4 composite was fabricated with a fixed fraction of CDots (1.26 wt.%) selected by preliminary experiments.It is known that the surface reaction is dominating in the present heterogeneous system, therefore surface area to solution volume ratio (SA/V) is used to represent the dosage of the composites other than the mass concentration, which has been applied in many reported works [17,18,22,27,28,[54][55][56].The SA/V value is obtained by combining the mass concentration (g/L) with specific surface area (m 2 /g) and can be normalized to m −1 .As can be seen in Figure 1A, the sharp peak for g-C 3 N 4 at 810 cm −1 is attributed to stretching vibration bond of tri-s-triazine [48].Vibration peaks between 1200-1650 cm −1 corresponds to the typical stretching modes of CN heterocycles [49].A wider band can be seen at 3100-3300 cm −1 , which belongs to the stretching vibration modes for the unreacted -NH [50].The same characteristic peaks are observed in CDots/g-C 3 N 4 composite and the peak at 1405 cm −1 can be seen as the indicator for the coupling of CDots and g-C 3 N 4 [48].
XRD patterns of g-C 3 N 4 and CDots/g-C 3 N 4 are displayed in Figure 1B.The main diffraction peaks observed at 12.9 • and 27.5 • in both g-C 3 N 4 and CDots/g-C 3 N 4 composite are indexed to the (100) peak of the in-plane structure of tri-s-triszine unit and (002) crystal facets of the inter-layer stacking of aromatic segments [51,52].The two patterns fit well with graphitic carbon nitride (JCPDS 87-1526) and no significant difference is observed, implying the low content of CDots in CDots/g-C 3 N 4 composite.However, it is remarkable that the difference in relative intensity, together with the shift observed in the (002) peak location from 27.51 • for g-C 3 N 4 to 27.59 • for CDots/g-C 3 N 4 , can be seen as the evidence of CDots introduction [45].As can be seen in Figure 1C, the introduction of CDots in CDots/g-C 3 N 4 leads to a 20% increase (from 120.92 to 145.24 m 2 /g) in specific surface area, which favors the decomposition of H 2 O 2 .
Figure 1D clearly shows the two-dimensional structure of g-C 3 N 4 together with the embedding of CDots (the white circles).The close look of the CDots embedded in g-C 3 N 4 matrix is given in Figure 1E.The CDots are non-uniformly distributed, ranging from 2 to 10 nm, which is in line with previous studies [45,51,53].
From the results and analysis above, it can be confirmed that CDots have been successfully decorated in g-C 3 N 4 and the inlay of CDots brings an improvement in specific surface area of the composite.

Kinetic Study
The effect of CDots content in the catalyst has been investigated in several previous works proving that a certain amount of CDots can enhance the catalytic properties of the catalysts while an excessive loading may work opposite [45,51,53].Thus, in this work the CDots/g-C 3 N 4 composite was fabricated with a fixed fraction of CDots (1.26 wt.%) selected by preliminary experiments.It is known that the surface reaction is dominating in the present heterogeneous system, therefore surface area to solution volume ratio (SA/V) is used to represent the dosage of the composites other than the mass concentration, which has been applied in many reported works [17,18,22,27,28,[54][55][56].The SA/V value is obtained by combining the mass concentration (g/L) with specific surface area (m 2 /g) and can be normalized to m −1 .
To verify the synergistic effect of CDots and g-C 3 N 4 in H 2 O 2 decomposition, five samples were prepared according to the proportion of CDots and g-C 3 N 4 in the composite.Sample 5 is the CDots/g-C 3 N 4 composite with SA/V of 4 × 10 5 m −1 .The pure g-C 3 N 4 with the same SA/V value was identified as Sample 2. The single-component CDots solution containing the equivalent amount of CDots as Sample 5 (13.4 mg/L) was named as Sample 1 and it was obtained by directly diluting the originally prepared CDots solution.The physical mixture of Sample 1 and 2 is identified as Sample 3. Additionally, the originally prepared CDots solution with a high concentration (133 mg/L) was named as Sample 4.
Detailed descriptions of the samples are listed in Table 1.It should be noted that sample 1, 3 and 5 has equivalent amount of CDots and sample 3 is a simple mixture of CDots and g-C 3 N 4 while sample 5 is the composite.The catalytic decomposition of H 2 O 2 by each sample was investigated under the same experimental condition where initial concentration of of each case is plotted against reaction time respectively (shown in Figure 2).To verify the synergistic effect of CDots and g-C3N4 in H2O2 decomposition, five samples were prepared according to the proportion of CDots and g-C3N4 in the composite.Sample 5 is the CDots/g-C3N4 composite with SA/V of 4 × 10 5 m −1 .The pure g-C3N4 with the same SA/V value was identified as Sample 2. The single-component CDots solution containing the equivalent amount of CDots as Sample 5 (13.4 mg/L) was named as Sample 1 and it was obtained by directly diluting the originally prepared CDots solution.The physical mixture of Sample 1 and 2 is identified as Sample 3. Additionally, the originally prepared CDots solution with a high concentration (133 mg/L) was named as Sample 4.
Detailed descriptions of the samples are listed in Table 1.It should be noted that sample 1, 3 and 5 has equivalent amount of CDots and sample 3 is simply mixture of CDots and g-C3N4 while sample 5 is the composite.The catalytic decomposition of H2O2 by each sample was investigated under the same experimental condition where initial concentration of H2O2 ([H2O2]0) is 0.5 mM and the temperature is 298K.Normalized concentration of H2O2 ([H2O2]/[H2O2]0) of each case is plotted against reaction time respectively (shown in Figure 2).As shown in Figure 2, Sample 4 (the originally prepared CDots, 133 mg/L) incurs a graded decline in H2O2 concentration in darkness, demonstrating the inherent catalytic property of CDots for H2O2 decomposition.However, the catalytic performance becomes faint after diluting CDots to a 10.2% concentration when comparing sample 1 and 4, indicating that such catalysis process is strongly dependent on the applied CDots concentration, which is in accordance with reported work [57].Sample 2 (pure g-C3N4) also incurs slight decline in H2O2, which can be attributed to the adsorption of g-C3N4 and catalytic decomposition by carbon-based material through the delocalization of electrons on the surface [24].However, the decomposition of H2O2 catalyzed by the pure g-C3N4 should not be considered as the main process in the system with CDots/g-C3N4 composite.It is remarkable that the consumption rate of H2O2 for sample 5 is much larger than sample 3, implying the thermal polymerization process gives rise to the remarkable synergy and the proximity between the CDots and the adsorption sites of H2O2 in g-C3N4 is necessary for the high efficiency of H2O2 decomposition [44].As shown in Figure 2, Sample 4 (the originally prepared CDots, 133 mg/L) incurs a gradely decline in H 2 O 2 concentration in darkness, demonstrating the inherent catalytic property of CDots for H 2 O 2 decomposition.However, the catalytic performance becomes faint after diluting CDots to a 10.2% concentration when comparing sample 1 and 4, indicating that such catalysis process is strongly dependent on the applied CDots concentration, which is in accordance with reported work [57].Sample 2 (pure g-C 3 N 4 ) also incurs slight decline in H 2 O 2 , which can be attributed to the adsorption of g-C 3 N 4 and catalytic decomposition by carbon-based material through the delocalization of electrons on the surface [24].However, the decomposition of H 2 O 2 catalyzed by the pure g-C 3 N 4 should not be considered as the main process in the system with CDots/g-C 3 N 4 composite.It is remarkable that the consumption rate of H 2 O 2 for sample 5 is much larger than sample 3, implying the thermal polymerization process gives rise to the remarkable synergy and the proximity between the CDots and the adsorption sites of H 2 O 2 in g-C 3 N 4 is necessary for the high efficiency of H 2 O 2 decomposition [44].
It has been previously reported that [18,54], the catalytic decomposition of H 2 O 2 in the heterogeneous system follows pseudo first-order kinetics with respect to H 2 O 2 when solid is excess to H 2 O 2 and the reaction rate equation can be described as where where k 2 denotes the second-order reaction rate constant, SA denotes the surface area of the CDots/g-C 3 N 4 and V is the volume of the reaction solution.The term SA/V has been applied to denote the catalyst concentration in a number of studies regarding the heterogeneous catalysis system [18,22,27,54].
According to the preliminary experiments, the lower limit of SA/V is 3.2 × 10 5 m −1 after which it is excess to the fixed initial H 2 O 2 concentration (0.5 mM).A series of experiments were carried out by varying the dosage of catalyst (SA/V) from 3.2 to 6.4 × 10 5 m −1 under the same condition at 298 K to explore the kinetics of the present system.The logarithm of normalized H 2 O 2 concentration is plotted as a function of reaction time (Figure 3A) and the slope of the linearly fitted curve of these plots (k 1 ) is plotted against SA/V accordingly (Figure 3B).
It has been previously reported that [18,54], the catalytic decomposition of H2O2 in the heterogeneous system follows pseudo first order kinetics with respect to H2O2 when solid is excess to H2O2 and the reaction rate equation can be described as k H O , which can be integrated as where k1 is the pseudo first order rate constant at a given temperature and dosage of the solid, t is the reaction time, [H2O2] is the concentration of H2O2 at a time and [H2O2]0 is the concentration of H2O2 at t = 0.When the solid catalyst is excess to H2O2, the second order rate constant in the system can be determined by studying the pseudo first order rate constant (k1) as a function of SA/V (surface area of solid to volume of solution).The second order rate expression is given as where k2 denotes the second-order reaction rate constant, SA denotes the surface area of the CDots/g-C3N4 and V is the volume of the reaction solution.The term SA/V has been applied to denote the catalyst concentration in a number of studies regarding the heterogeneous catalysis system [18,22,27,54].
According to the preliminary experiments, the lower limit of SA/V is 3.2 × 10 5 m −1 after which it is excess to the fix initial H2O2 concentration (0.5 mM).A series of experiments were carried out by varying the dosage of catalyst (SA/V) from 3.2 to 6.4 × 10 5 m −1 under the same condition at 298 K to explore the kinetics of the present system.The logarithm of normalized H2O2 concentration is plotted as a function reaction time (Figure 3A) and the slope of the linearly fitted curve of these plots (k1) is plotted against SA/V accordingly (Figure 3B).From Figure 3A, it can be seen that all the ln([H 2 O 2 ]/[H 2 O 2 ] 0 ) plots are linearly fitted with reaction time which indicates it follows pseudo first-order kinetics at given dosage of the composite and the slopes of the fitted curves are denoted as k 1 .In addition, it is clearly that the observed decomposition rate of H 2 O 2 increases with increasing the dosage of the composite.The key parameters of the fitted curves, including SA/V, slopes (k 1 ), standard deviation and R 2 , are listed in Table 2.The obtained k 1 values from Table 2 were plotted in Figure 3B against SA/V.As can be seen from Figure 3B, k 1 is linearly correlated to SA/V in the range of 3.2-6.4× 10 5 m −1 and the slope of the fitted curve is calculated as (1.42 ± 0.07) × 10 −9 m•s −1 which can be denoted as the overall second-order rate constant.This value is far from the rate constant of a diffusion controlled reaction in the order of 10 −5 m•s −1 but still higher than some metal oxide catalysts like ZrO 2 ((2.39 ± 0.09) × 10 −10 m•s −1 ), CuO ((1.23 ± 0.06) × 10 −9 m•s −1 ) and Gd 2 O 3 ((9.4± 1.0) × 10 −10 m•s −1 ) [18,54].
Generally, the first-order rate constant k 1 is strongly related to the reaction temperature according to the Arrhenius equation: where E a denotes the activation energy for the reaction, R is the gas constant, T is the absolute temperature and A is the pre-exponential factor.The logarithm of k 1 obtained by plotting ln([H 2 O 2 ]/[H 2 O 2 ] 0 ) against T (shown in Figure 4A) is plotted as a function of 1/T in Figure 4B so as to calculate E a .From Figure 3A, it can be seen that all the ln([H2O2]/[H2O2]0) plots are linearly fitted with reaction time which indicates it follows pseudo first order kinetics at given dosage of the composite and the slopes of the fitted curves are denoted as k1.In addition, it is clearly that the observed decomposition rate of H2O2 increases with increasing the dosage of the composite.The key parameters of the fitted curves, including SA/V, slopes (k1), standard deviation and R 2 , are listed in Table 2.The obtained k1 values from Table 2 were plotted in Figure 3B against SA/V.As can be seen from Figure 3B, k1 is linearly correlated to SA/V in the range of 3.2-6.4× 10 5 m −1 and the slope of the fitted curve is calculated as (1.42 ± 0.07) × 10 −9 m•s −1 which can be denoted as the overall second-order rate constant.This value is far from the rate constant of a diffusion controlled reaction in the order of 10 −5 m•s −1 but still higher than some metal oxide catalysts like ZrO2 ((2.39 ± 0.09) × 10 −10 m•s −1 ), CuO ((1.23 ± 0.06) × 10 −9 m•s −1 ) and Gd2O3 ((9.4 ± 1.0) × 10 −10 m•s −1 ) [18,54].
The key parameters of the fitted curves, including T, pseudo-first order reaction rate constants (k1), standard deviation and R 2 , are listed in Table 3.

The Effect of pH
To investigate the mechanism of the present system containing H2O2 and CDots/g-C3N4 composite, it is significant to study the pH effect as well as quantify the in-situ produced hydroxyl radicals.Due to its scavenging capacity against HO• and pH buffering ability, Tris is chosen to carry out the mechanistic study.The pKa and the buffering range of Tris are 8.07 and 7.0-9.0respectively, so the pH values were selected within this range.The decline of H2O2 together with the production of CH2O against the reaction time with different pH are exhibited in Figure 5.It can be seen from Figure 4B, ln(k 1 ) is linearly dependent on 1/T and the slope of the fitted curve is obtained.Based on the slope, the activation energy of the reaction is calculated to be (29.05± 0.80) kJ•mol −1 , which is to some extent lower than a series of metal oxides developed before, including ZrO 2 ((33 ± 1) kJ•mol −1 ), TiO 2 ((37 ± 1) kJ•mol −1 ), Y 2 O 3 ((47 ± 5) kJ•mol −1 ), Fe 2 O 3 ((47 ± 1) kJ•mol −1 ), CuO ((76 ± 1) kJ•mol −1 ), CeO 2 ((40 ± 1) kJ•mol −1 ), Gd 2 O 3 ((63 ± 1) kJ•mol −1 ) and HfO 2 ((60 ± 1) kJ•mol −1 ) [18,54,58].
The key parameters of the fitted curves, including T, pseudo first-order reaction rate constants (k 1 ), standard deviation and R 2 , are listed in Table 3.

The Effect of pH
To investigate the mechanism of the present system containing H 2 O 2 and CDots/g-C 3 N 4 composite, it is significant to study the pH effect as well as quantify the in-situ produced hydroxyl radicals.Due to its scavenging capacity against HO• and pH buffering ability, Tris is chosen to carry out the mechanistic study.The pKa and the buffering range of Tris are 8.07 and 7.0-9.0respectively, so the pH values were selected within this range.The decline of H 2 O 2 together with the production of CH 2 O against the reaction time with different pH are exhibited in Figure 5.It can be clearly seen in Figure 5 that the decomposition rate of H2O2 increases with pH increases in the whole range.However, the formation of CH2O shows a different trend.The evolution of CH2O remains relatively low in neutral condition (pH 7), then it starts to accelerate until pH 8, after which the formation rate of CH2O declines as pH increases.This indicates that the formation of CH2O and probably the production of HO• are alkaline favored while the production is to some degree related to the pKa of Tris [54].
To figure out whether the production of HO• is also pH dependent, it is necessary to introduce the yield (Y) of CH2O formed by HO• and Tris.Y is defined by the equation: where [HO•] is the production of HO• and [CH2O] is the accumulated CH2O in H2O2 decomposition experiment.According to a previous study using γ-radiation in homogeneous system [17], the yield (Y) of CH2O increases from 25% to 51% as increasing pH from 7.0 to 9.0.Provided that the yield (Y) in heterogeneous system is consistent with that in homogeneous system, the production of HO• in H2O2 decomposition on CDots/g-C3N4 composite can be estimated by this value together with the final concentration of CH2O.The results are shown in Figure 6.According to a previous study using γ-radiation in homogeneous system [17], the yield (Y) of CH 2 O increases from 25% to 51% as increasing pH from 7.0 to 9.0.Provided that the yield (Y) in heterogeneous system is consistent with that in homogeneous system, the production of HO• in H 2 O 2 decomposition on CDots/g-C 3 N 4 composite can be estimated by this value together with the final concentration of CH 2 O.The results are shown in Figure 6.It can be clearly seen in Figure 5 that the decomposition rate of H2O2 increases with pH increases in the whole range.However, the formation of CH2O shows a different trend.The evolution of CH2O remains relatively low in neutral condition (pH 7), then it starts to accelerate until pH 8, after which the formation rate of CH2O declines as pH increases.This indicates that the formation of CH2O and probably the production of HO• are alkaline favored while the production is to some degree related to the pKa of Tris [54].
To figure out whether the production of HO• is also pH dependent, it is necessary to introduce the yield (Y) of CH2O formed by HO• and Tris.Y is defined by the equation: where [HO•] is the production of HO• and [CH2O] is the accumulated CH2O in H2O2 decomposition experiment.According to a previous study using γ-radiation in homogeneous system [17], the yield (Y) of CH2O increases from 25% to 51% as increasing pH from 7.0 to 9.0.Provided that the yield (Y) in heterogeneous system is consistent with that in homogeneous system, the production of HO• in H2O2 decomposition on CDots/g-C3N4 composite can be estimated by this value together with the final concentration of CH2O.The results are shown in Figure 6.As appeared in Figure 6, the plots of estimated production of HO• exhibit the similar tendency as that of final production of CH 2 O and it still peaks at pH 8.0 with the maximum concentration of 2.10 mM ([HO•] = [CH 2 O]/Y where Y = 37.5% at pH 8.0).This means the stoichiometry between H 2 O 2 and CH 2 O is approximately 1:0.158 and 42.1% of H 2 O 2 ([H 2 O 2 ] 0 = 5 mM, Tris is excess to H 2 O 2 ) end up with HO• at the optimal pH.The efficiency of H 2 O 2 consumption towards HO• is much higher as compared to that of the H 2 O 2 /ZrO 2 /Tris system (13.4% at pH 8.0) [17].Therefore, it can be concluded that the production of HO• also pH-dependent and there is an optimal pH which may have something to do with the Tris [22].
Based on the results in present work, the intrinsic chemical catalytic properties of the synthesized CDots/g-C 3 N 4 composite, other than the photocatalytic properties, have been revealed to some extent.The hypothesis in the introduction section could be confirmed as following: firstly, as demonstrated in Figure 2, the pure CDots synthesized via an electrochemical pathway showed excellent catalytic ability, in line with the literature [59]; secondly, similar as the reported property of g-C 3 N 4 [60], Figure 2 also shows that the concentration of H 2 O 2 in the solution decreases slightly in the presence of the pure g-C 3 N 4 which indicates g-C 3 N 4 does provide sufficient reaction site for H 2 O 2 to adsorb; thirdly, by analyzing the results in Figures 2, 3, 5 and 6, it is known that hydroxyl radical can be formed during the decomposition of H 2 O 2 catalyzed by the CDots/g-C 3 N 4 composite in the heterogeneous system as hypothesized in the former section.Besides the strong affinity of g-C 3 N 4 towards H 2 O 2 and the catalytic property of CDots against the adsorbed H 2 O 2 [47], the delocalization of the electrons on the surface of g-C 3 N 4 may also leads to the decomposition of H 2 O 2 via electron-transfer mechanism [24].In conclusion, the synthesized CDots/g-C 3 N 4 composite exhibits the synergy of adsorption of H 2 O 2 and delocalization of electrons on g-C 3 N 4 and catalytic decomposition of H 2 O 2 by g-C 3 N 4 and CDots producing hydroxyl radicals.
Based on the results and discussion above, the mechanism of catalytic decomposition of H 2 O 2 in the heterogeneous system with CDots/g-C 3 N 4 composite is proposed and illustrated in Figure 7.As appeared in Figure 6, the plots of estimated production of HO• exhibit the similar tendency as that of final production of CH2O and it still peaks at pH 8.0 with the maximum concentration of 2.10 mM ([HO•] = [CH2O]/Y where Y = 37.5% at pH 8).This means the stoichiometry between H2O2 and CH2O is approximately 1:0.158 and 42.1% of H2O2 ([H2O2]0 = 5 mM, Tris is excess to H2O2) end up with HO• at the optimal pH.The efficiency of H2O2 consumption towards HO• is much higher as compared to that of the H2O2/ZrO2/Tris system (13.4% at pH 8) [17].Therefore, it can be concluded that the production of HO• is also pH-dependent and there is an optimal pH which may have something to do with the Tris [22].
Based on the results in present work, the intrinsic chemical catalytic properties of the synthesized CDots/g-C3N4 composite, other than the photocatalytic properties, have been revealed to some extent.The hypothesis in the introduction section could be confirmed as following: firstly, as demonstrated in Figure 2, the pure CDots synthesized via an electrochemical pathway showed excellent catalytic ability, in line with the literature [59]; secondly, similar as the reported property of g-C3N4 [60], Figure 2 also shows that the concentration of H2O2 in the solution decreases slightly in the presence of the pure g-C3N4 which indicates g-C3N4 does provide sufficient reaction site for H2O2 to adsorb; thirdly, by analyzing the results in Figures 2, 3, 5 and 6, it is known that hydroxyl radical can be formed during the decomposition of H2O2 catalyzed by the CDots/g-C3N4 composite in the heterogeneous system as hypothesized in the former section.Besides the strong affinity of g-C3N4 towards H2O2 and the catalytic property of CDots against the adsorbed H2O2 [47], the delocalization of the electrons on the surface of g-C3N4 may also leads to the decomposition of H2O2 via electrontransfer mechanism [24].In conclusion, the synthesized CDots/g-C3N4 composite exhibits the synergy of adsorption of H2O2 and delocalization of electrons on g-C3N4 and catalytic decomposition of H2O2 by g-C3N4 and CDots producing hydroxyl radicals.
Based on the results and discussion above, the mechanism of catalytic decomposition of H2O2 in the heterogeneous system with CDots/g-C3N4 composite is proposed and illustrated in Figure 7. Key of the mechanism is the so-called adsorb-catalyze double reaction sites.With plenty of accessible adsorption sites on the surface of g-C3N4, CDots/g-C3N4 composite shows high selective adsorption ability towards aqueous H2O2.From previous works studying similar heterogeneous system with H2O2 and solid catalyst [18,22,54], it is known that H2O2 concentration exhibits an initial drop indicating the adsorption on the surface of catalyst, after which the H2O2 decomposition obeys pseudo first-order kinetic when the surface reaches equilibrium state.As can be seen in Figure 3A, the H2O2 concentration follows the similar trend and kinetics.Hence, it can be demonstrated that the adsorption of H2O2 is also dominating in the initial short period.Tris was introduced in the present work as a probe of hydroxyl radical and pH buffer.It is known that Tris can be partially oxidized to CH2O and other byproducts and the ratio between the concentration of hydroxyl radicals and formed Key of the mechanism is the so-called adsorb-catalyze double reaction sites.With plenty of accessible adsorption sites on the surface of g-C 3 N 4 , CDots/g-C 3 N 4 composite shows high selective adsorption ability towards aqueous H 2 O 2 .From previous works studying similar heterogeneous system with H 2 O 2 and solid catalyst [18,22,54], it is known that H 2 O 2 concentration exhibits an initial drop indicating the adsorption on the surface of catalyst, after which the H 2 O 2 decomposition obeys pseudo first-order kinetic when the surface reaches equilibrium state.As can be seen in Figure 3A, the H 2 O 2 concentration follows the similar trend and kinetics.Hence, it can be demonstrated that the adsorption of H 2 O 2 is also dominating in the initial short period.Tris was introduced in the present work as a probe of hydroxyl radical and pH buffer.It is known that Tris can be partially oxidized to CH 2 O and other byproducts and the ratio between the concentration of hydroxyl radicals and formed CH 2 O is relatively fixed under given condition (pH and dissolved oxygen concentration) [17,27].Therefore, the formation of CH 2 O can be used to probe the formed hydroxyl radicals in the heterogeneous system containing H 2 O 2 and CDots/g-C 3 N 4 composite.It is in line with previous works that [17,18,22,27], CH 2 O formation is reflected by the decomposition of H 2 O 2 and is pH-dependent (Figures 5 and 6).Hence, it can be deduced that after the initial period, the adsorbed H 2 O 2 on the surface of the CDots/g-C 3 N 4 composite reaches an equilibrium state and the decomposition of H 2 O 2 catalyzed by embedded CDots on the surface sites turns the dominant role.During this procedure, large quantities of HO• was produced, exhibiting strong oxidation ability towards scavengers like Tris.It should be noted that the production of HO• is strongly pH dependent.To sum up, the CDots/g-C 3 N 4 composite shows synergetic effect on the decomposition of H 2 O 2 via adsorb-catalyze double reaction sites and more importantly, is proved to be a promising catalyst for the degradation of NBDOPs since it is a metal-free pathway of producing HO• efficiently.

Instrumentation
The morphology and microstructure of samples were observed by JET-2100F (JEOL, Wuhan, China) transmission electron microscope (TEM).The Fourier transform infrared spectroscopy (FTIR) of the samples were recorded by Nicolet iS5 (Thermo Fisher Scientific, Wuhan, China) FTIR spectrometer with KBr pellets.The specific surface area of CDots/g-C 3 N 4 composite and pure g-C 3 N 4 were determined by the Brunauer-Emmett-Teller (B.E.T) method via isothermal adsorption and desorption of high purity nitrogen using a TriStar II 3020 (Micromeritics, Wuhan, China) instrument.X-ray diffraction (XRD) patterns were recorded with D8 advance (Bruker, Wuhan, China) diffractometer using Bragg-Brentano geometry in the 2θ angle from 10 • to 40 • and Cu Kα irradiation (λ = 1.54 Å).The samples were weighted to ± 10 −4 g in a ME104E (Mettler Toledo, Wuhan, China) microbalance.UV/Vis spectra were collected by V-5600 (METASH, Wuhan, China) and UV-5500PC (METASH, Wuhan, China) spectrophotometer.The pH of reaction solution was measured by PHS-3C (YOKE, Wuhan, China) pH meter with an accuracy of ± 0.01 pH units.

Reagents and Experiments
All the solutions used in this study were prepared using deionized water.Preparation of the catalyst: CDots were synthesized via an electrochemical method based on previous reported work [59].In a typical preparation process, two graphite rods were insert parallel into 300 mL ultrapure water as electrodes with a separation of 7.5 cm and 4 cm depth under water.60 V static potentials were applied to the rods by a direct-current (DC).After electrolyzing for 120 h, the anode graphite rod corroded and a dark brown solution was formed.The solution was filtered with slow-speed quantitative filter paper and then centrifuged at 10,000 rpm for 10 min.Finally, the soluble CDots was obtained and the concentration can be quantified by drying and weighting.
A thermal polymerization method was applied for the synthesis of pure g-C 3 N 4 [61].Typically, 40 g urea (CAS[57-13-6], 99%, Sinopharm, Wuhan, China) was dissolved in 40 mL ultrapure water in a quartz crucible, then heated to 550 • C with the rate of 7 • C/min in a muffle furnace and kept at 550 • C for 2 h.After naturally cooling down to room temperature, the resultant yellow product was collected and ground into powder to obtain pure g-C 3 N 4 .CDots/g-C 3 N 4 composite was synthesized via in-situ thermal polymerization [51].Following the same procedure, 40 g urea was dissolved in 40 [17].The decomposition experiments were carried out in 100 mM Tris solution with fixed quantities of CDots/g-C 3 N 4 (SA/V = 4 × 10 5 m −1 , SA and V stands for the surface area of solid and the volume of solution) and H 2 O 2 (5 mM).The pH values of solution were selected within the valid buffering range of Tris, namely pH 7.0-9.0for investigating the effect of pH.The produced CH 2 O was quantitatively determined by a modified Hantzsch method, where CH 2 O reacts with AAA in the presence of NH 4 Ac to form a dihydropyridine derivative with a maximum absorbance wavelength at 368 nm [19].The calibration curve where the absorbance of produced dihydropyridine derivative was plotted as a function of CH 2 O concentration with linear correlation was obtained at 368 nm in the range of 0.04-1.3mM CH 2 O for the conversion of absorbance to CH 2 O concentration.The experimental error in the determination of CH 2 O was less than 2%.

Conclusions
In this work, a promising catalyst CDots/g-C 3 N 4 composite for degradation of non-biodegradable organic pollutants was synthesized via a two-step pathway including electrochemical exfoliation of graphite rod preparing CDots and thermal polymerization of CDots mixed urea.Through different characterization methods and kinetic experiments, it has been confirmed that CDots embed in g-C 3 N 4 matrix and such structure accounts for the synergetic catalytic performance of the composite.Kinetics of catalytic decomposition of H 2 O 2 on CDots/g-C 3 N 4 composite were researched.The second-order rate constant (k 2 ) was measured to be (1.42 ± 0.07) × 10 −9 m•s −1 and the activation energy of the reaction was measured to be (29.05± 0.80) kJ•mol −1 under the applied conditions.The effect of pH (pH 7.0-9.0)on the production of HO• was also investigated by using Tris as a probe.It has been shown that the production of HO• is strongly alkaline dependent and the maximum reaches at pH 8.0 which is close to the pKa of Tris.A mechanism based on the adsorb-catalyze double reaction site theory has been proposed.This work implies that the photocatalyst (CDots/g-C 3 N 4 composite) for water splitting or H 2 evolution may also be applied as an alternative catalyst in degradation of non-biodegradable organic pollutants.The instinct kinetics and the mechanism can be referred to for further applications in related fields.
,B.The influence of CDots modification on the specific surface area was investigated by the B.E.T. method with isothermal adsorption and desorption of high purity nitrogen.The N 2 adsorption-desorption isotherms and pore size distributions of g-C 3 N 4 and CDots/g-C 3 N 4 composite are shown in Figure 1C.The TEM images of CDots/g-C 3 N 4 are shown in Figure 1D,E.Catalysts 2018, 8, x FOR PEER REVIEW 3 of 15

Figure 6 .
Figure 6.Final production of CH2O and estimated production of HO• as a function of pH.The final productions of CH2O were extracted from Figure 5.The estimated [HO•] = [CH2O]/Y and the yields (Y) were derived from a previous work [17].

Figure 5 .
Figure 5.The concentration of H 2 O 2 ([H 2 O 2 ] 0 = 5 mM) and CH 2 O as a function of reaction time in the presence of Tris ([Tris] 0 = 100 mM) with the catalyst dosage of 4.0 × 10 5 m −1 at 298.15K in pH from 7.0 to 9.0.It can be clearly seen in Figure 5 that the decomposition rate of H 2 O 2 increases with pH increases in the whole range.However, the formation of CH 2 O shows a different trend.The evolution of CH 2 O remains relatively low in neutral condition (pH 7.0), then it starts to accelerate until pH 8.0, after which the formation rate of CH 2 O declines as pH increases.This indicates that the formation of CH 2 O and probably the production of HO• are alkaline favored while the production is to some degree related to the pKa of Tris [54].To figure out whether the production of HO• is also pH dependent, it is necessary to introduce the yield (Y) of CH 2 O formed by HO• and Tris.Y is defined by the equation: Y = [CH 2 O]/[HO•] (4) where [HO•] is the production of HO• and [CH 2 O] is the accumulated CH 2 O in H 2 O 2 decomposition experiment.According to a previous study using γ-radiation in homogeneous system[17], the yield (Y) of CH 2 O increases from 25% to 51% as increasing pH from 7.0 to 9.0.Provided that the yield (Y) in heterogeneous system is consistent with that in homogeneous system, the production of HO• in H 2 O 2 decomposition on CDots/g-C 3 N 4 composite can be estimated by this value together with the final concentration of CH 2 O.The results are shown in Figure6.

Figure 6 .
Figure 6.Final production of CH2O and estimated production of HO• as a function of pH.The final productions of CH2O were extracted from Figure 5.The estimated [HO•] = [CH2O]/Y and the yields (Y) were derived from a previous work [17].

Figure 6 .
Figure 6.Final production of CH 2 O and estimated production of HO• as a function of pH.The final productions of CH 2 O were extracted from Figure 5.The estimated [HO•] = [CH 2 O]/Y and the yields (Y) were derived from a previous work [17].

Figure 7 .
Figure 7.The mechanism of catalytic decomposition of H 2 O 2 on CDots/g-C 3 N 4 composite.
k 1 is the pseudo first-order rate constant at a given temperature and dosage of the solid, t is the reaction time, [H 2 O 2 ] is the concentration of H 2 O 2 at a time and [H 2 O 2 ] 0 is the concentration of H 2 O 2 at t = 0.When the solid catalyst is excess to H 2 O 2 , the second-order rate constant in the system can be determined by studying the pseudo first-order rate constant (k 1 ) as a function of SA/V (surface area of solid to volume of solution).The second-order rate expression is given as

Table 2 .
The key parameters of the fitted curves with different SA/V values.

Table 2 .
The key parameters of the fitted curves with different SA/V values.

Table 3 .
The key parameters of the fitted curves with different temperatures.

Table 3 .
The key parameters of the fitted curves with different temperatures.