3.1. Characterization
Figure 1
The crystal structure of g–C3N4, raw zeolite, g–C3N4-zeolite and P and B doped g–C3N4-zeolite composites was characterized by XRD(Fig. 1). Figure 1 showed diffraction peaks of the g–C3N4 crystal faces at 13.2° (100) and 27.3°(002) (JCPDS 87 − 1526)(Cao et al. 2015). The diffraction peak at 13.2° corresponds to tri-s-triazine units, which corresponds to the typical (002) plane of conjugated aromatic systems(Wang et al. 2008). In contrast, the characteristic XRD peaks of the raw zeolite are 2θ = 11.3° (200), 14.06°(102), 21.9° (400), 26.3° (312), 29.7° (044), and 30.3°(062)(JCPDS 89–1421)(Nakamura et al. 1992; Jesudoss et al. 2017; Devarayapalli et al. 2020).
After 500 oC calcination process, XRD patterns of the g–C3N4-zeolite composites showed that the internal lattice structure of the zeolite was preserved and the crystalline degree increased significantly. It can also be stated that the characteristic peaks of g–C3N4 coincide with the characteristic peaks of the raw zeolite in the g–C3N4-zeolite composite. Remarkably, the intensity of the peaks became stronger after the g–C3N4-zeolite composite interacted with boric acid and phosphoric acid. The mean crystal size of raw zeolite, g–C3N4-zeolite, and P and B doped g–C3N4-zeolite composites was determined using the Debye–Scherrer equation. The results obtained are given in Table 1.
Table 1
XRD peak parameters for the –C3N4, raw zeolite, g–C3N4-zeolite and P and B doped g–C3N4-zeolite composites
|
Raw zeolite
|
g-C3N4 zeolite
|
P and B doped –g-C3N4 zeolite
|
|
Position
|
FWHM
|
Crystallite size(nm)
|
Position
|
FWHM
|
Crystallite size(nm)
|
Position
|
FWHM
|
Crystallite size(nm)
|
(200)
|
11.53°
|
17.08
|
0.49
|
11.20°
|
6.61
|
1.26
|
11.19
|
3.47
|
2.40
|
(102)
|
14.06°
|
17.35
|
0.48
|
13.09°
|
7.52
|
1.11
|
13.06
|
3.56
|
2.35
|
(400)
|
21.90
|
17.11
|
0.49
|
22.48
|
4.83
|
1.75
|
22.47
|
2.55
|
3.32
|
(312)
|
26.30
|
15.08
|
0.57
|
26.14
|
7.39
|
1.15
|
26.06
|
3.52
|
2.42
|
(044)
|
29.77
|
17.19
|
0.50
|
28.18
|
7.75
|
1.10
|
28.16
|
3.56
|
2.40
|
(062)
|
30.30
|
16.84
|
0.51
|
30.20
|
7.85
|
1.10
|
30.14
|
3.61
|
2.38
|
The mean crystal size of raw zeolite, g–C3N4-zeolite, and P and B doped g–C3N4-zeolite composites were 0.505, 1.161 and 2.46 nm, respectively.
Figure 2
The surface morphology of g–C3N4(a), raw zeolite(b), g–C3N4-zeolite (c, d) and P and B doped g–C3N4-zeolite (e, f) composites were analyzed using SEM (Fig. 2). In contrast, it can be seen in Fig. 2a that g–C3N4 is relatively more homogeneous and porous. It has been determined that the raw zeolite has a layered structure (Fig. 2b). As shown in Fig. 2(c, d), it can be stated that after the zeolite is mixed with g–C3N4 and subjected to the calcination process, the porous structure seen in raw zeolite and g–C3N4 decreases and a more homogeneous structure emerges. In addition, in the g–C3N4-zeolite composite sample, a structure more similar to zeolite grains emerged instead of grains corresponding to g–C3N4. This may be due to the excellent incorporation of the zeolite into the g–C3N4 matrix.
It can be seen from SEM micrographs of the P and B doped g–C3N4-zeolite composite (Figs. 2(e) and (f)) that the particles are relatively rough and have a good dispersion with a porous surface compared to the g–C3N4 matrix of the zeolite.
Thus, this relatively rough and porous surface will create a higher contact surface area and will further encourage the adsorption of molecules, which can positively affect the catalytic activity.
Figure 3
Elemental composition of g–C3N4, g–C3N4-zeolite and P and B doped g–C3N4-zeolite composites was obtained by EDX analysis (Fig. 3(a)). O, C, N, B, P, Al and Si elements were observed for the P and B doped g–C3N4-zeolite composite. EDX atomic percentage elemental analysis results of g–C3N4-zeolite and P and B doped g–C3N4-zeolite composites are given in Table 2. EDX analyzes prove that the atomic percentage of oxygen groups is over 50% in both g–C3N4-zeolite and P and B doped g–C3N4-zeolite composites. It also confirms the incorporation of B and P atoms into the g–C3N4-zeolite structure with boric acid and phosphoric acid, respectively.
Table 2
EDX elemental analysis for g–C3N4−zeolite, and P and B doped g–C3N4−zeolite composites.
|
C%
|
Al%
|
N%
|
O%
|
Si%
|
P%
|
B%
|
g–C3N4−zeolite
|
3.32
|
4.83
|
0.69
|
67.11
|
21.90
|
0
|
0
|
P and B doped g–C3N4−zeolite
|
5.60
|
4.24
|
2.87
|
58.20
|
27.80
|
0.90
|
0.08
|
Figure 4
The functional groups of g–C3N4, raw zeolite, g–C3N4-zeolite and P and B doped g–C3N4-zeolite composites were analyzed by FTIR spectroscopy. Figure 4a shows the FTIR spectra of the g–C3N4 sample in the 400–4000 cm-1 range, Figure 4b shows the FTIR spectra of the raw zeolite and g–C3N4-zeolite sample in the 400–4000 cm-1 range, and Figure 4c shows the FTIR spectra of the g–C3N4-zeolite and P and B doped g–C3N4-zeolite composites sample in the 2000–500 cm-1 range. The peak found at 3165 cm-1 for raw zeolite is known as the O–H bonds of adsorbed OH molecules(Zhao et al. 2021). The peaks in the range of 850–900 cm-1 and 1000–1100 cm-1 for the raw zeolite are attributed to SiOH bending vibrations and strain and Si-O-Si vibrations, respectively(Fernández-Jiménez and Palomo 2005). Also, the peaks at 1070 cm-1 and 1200 cm-1 indicate the presence of Si-O Si vibrations. For pure g–C3N4 sample, absorption band s-triazine ring vibrations at 810 cm−1, peaks around 900–1800 cm-1 indicate C–N(–C)–C or C–NH–C vibrations(Zhu et al. 2018) and the broad peak at 3000–3600 cm−1 indicates NH stretching vibration(Tian et al. 2014). Figure 4 In the FTIR spectrum of the zeolite calcined at 500 oC, there is dehydration of the OH groups. The zeolite crystal structure can be characterized by bands around 545 and 445 cm-1 (Zhang et al. 2012). Compared to g–C3N4-zeolite, new bands in the spectrum for P and B doped g–C3N4-zeolite in the 1300-1600 cm-1 range can be assigned to the absorption of stretch vibration of oxygen groups. The peak appearing at 1385 cm-1 in the FTIR spectrum obtained in the region of 1300 to 2000 cm-1 for the P and B doped sample can be assigned to the boron atom(Balyan et al. 2020).Also, bands at 1400 cm-1 and 1519 cm-1 can be assigned to the P=O and P-OH group, respectively, proving that H3PO4 doping has been successfully performed(Ghiaci et al. 2007).
Figure 5
In Fig. 5, the signals of the elements C, N and O were detected for the spectrum of g-C3N4 at binding energies of 285, 395 and 529 eV. For the g-C3N4 -zeolite and P and B doped g-C3N4-zeolite samples, signals corresponding to the elements Al, Si, C, N, P, B and O were detected. This indicated the inclusion of P and B atoms in both the g-C3N4-zeolite composite material and the g-C3N4 -zeolite structure. High resolution XPS spectra for Al, Si 2p, B1s, O1s, P2P, C 1s and N 1s elements were investigated for the P and B doped g-C3N4-zeolite sample and the results are shown in Fig. 5a-g. The peaks of the elements have been shown to vary significantly.
Table 3 shows the percentages of elemental analysis by XPS analysis of g-C3N4-zeolite and P and B doped g-C3N4-zeolite catalysts. The Al and Si contents of g-C3N4-zeolite catalyst treated with boric acid and phosphoric acid were decreased compared to g-C3N4-zeolite. However, a high reduction of about 50% in alumina content was observed.
Table 3
XPS elemental analysis for g–C3N4, g–C3N4−zeolite, and P and B doped g–C3N4−zeolite composites.
|
C%
|
Al%
|
N%
|
O%
|
Si%
|
P%
|
B%
|
g–C3N4−
|
39.85
|
0
|
57.30
|
2.85
|
0
|
0
|
0
|
g–C3N4−zeolite
|
2.23
|
6.05
|
2.48
|
64.33
|
23.75
|
0
|
0
|
P and B doped g–C3N4−zeolite
|
9.71
|
3.00
|
11.21
|
53.30
|
20.23
|
0.70
|
0.54
|
In the high resolution spectrum of C1s (Fig. 6(a)), the peak at 284.04 eV is due to sp3 C–C carbon. The peaks of the CN units at 286.0 eV and 287.50 eV are attributable to C–NH2 and N = C–N, respectively(Wei et al. 2018; Zhao et al. 2018).
A binding energy of 287.9 eV corresponds to NCO bindings. In the high-resolution spectrum of N1s (Fig. 6(b)), the three peaks corresponding to 396.2eV, 397.5 eV and 398.9 eV are C–N = C in aromatic CN heterocycles and C–NH2 and tertiary nitrogen N–(C) in the CN frame, respectively(Zhao et al. 2016; Devarayapalli et al. 2020).
The shift of the N 1s binding energy of the P and B doped g-C3N4-zeolite towards a lower binding energy than the g-C3N4-zeolite indicates that the electronic character and chemical environment of the composites can be changed by the doping of P and B.(Zhou et al. 2015).
For O1s, binding energies at 529.4 and 530.4 eV were associated with oxidation to oxygen. In addition, the binding energy of 530.9 eV is attributed to the water molecules adsorbed on the surface of the material.(Schoiswohl et al. 2006). The high-resolution B 1s spectrum of P and B doped g-C3N4-zeolite exhibits three peaks at ∼187.9, 189.1 and 191.0 eV, corresponding to B–N–C, B–C3 and B–O bonds, respectively (Fig. 5b) (Jang et al. 2018). The high-resolution P 2p spectrum of P and B doped g-C3N4-zeolite peaks at 130.0 and 131.1 eV were belonged to P2p3/2 (P–P) and P2p1/2 (P–P), respectively(Zhou et al. 2021). The binding energy for P2p at 133.1 eV (Fig. 3c) can be interpreted into P-N bonds, indicating that P atoms have likely replaced C atoms in the triazine rings of g-C3N4. (Zhang et al. 2013a). The successful doping of phosphorus to this hybrid sample has been proven by XPS.
Figure 6
3.2. Catalytic activities
Figure 7
For the H2-P from NaBH4-MR, the catalytic activities of g-C3N4, g-C3N4-zeolite, P doped g- g-C3N4-zeolite, B doped g-C3N4-zeolite and B and P doped g-C3N4-zeolite catalysts at 30 oC were performed and the results were presented in Fig. 7. The amount of catalyst was taken as 10 mg, the calcination temperature was 500 oC, and the ratio of g-C3N4 and zeolite samples was taken as 1:1. HGR values for g-C3N4, g-C3N4-zeolite, P doped g-C3N4-zeolite, B doped g-C3N4-zeolite and B and P doped g-C3N4-zeolite catalysts were obtained with 2550, 2628, 2944, 4250 and 6250 ml min− 1 g− 1. HGR values for the B and P doped g-C3N4-zeolite catalyst showed higher catalytic performance. The steps of H2-P from NaBH4-MR can be written as;
The dissociation of NaBH4 in CH3OH into Na + and BH4− ions can be considered as the first reaction step. Then, BH4− and CH3OH molecules are adsorbed onto the charged B and P doped g-C3N4-zeolite catalyst surface, forming an activated complex. This complex then dissociates into BH3 and H2. At the same time, the compound B(CH3O)3 is formed from the reaction with the CH3O− ion and boron, and H2 is released. This reaction continues until a total of 4H2 and NaB(OCH3)4 products are formed. (Li et al. 2012) (Ali et al. 2019). It has been stated that the new metal or oxide addition to the reaction medium can act as a Lewis acid or electrophilic site to increase the overall reaction rate. It is known that acids promote acidic species formation as Brønsted and Lewis sites on oxide surfaces (Kummert and Stumm 1980; Akdim et al. 2009). The catalytic properties of zeolites, which are important solid acid catalysts, are due to the strong acidity of the zeolite framework, which occurs when protons compensate for the negative charge of the zeolite framework due to the interaction of the Si4+ lattice with Al3+. The acidic properties and catalytic activities of the zeolite can be strongly enhanced by incorporating extra species into the zeolite framework with various modifications([CSL STYLE ERROR: reference with no printed form.]; Van Speybroeck et al. 2015; Pidko 2017).
The presence of acid sites in a catalyst and the accessibility of molecules to these active sites significantly affect the catalytic behavior. In a study on the effect of phosphorus on metal oxide surfaces, the acid properties and pore structure of the zeolite were modified with phosphorus, and the acid region distribution and strength were changed. 3 At the same time, phosphorus can cause alumina formation in the zeolite structure and/or react with extra-frame aluminum species to form aluminophosphates(Manrique et al. 2019; Lakiss et al. 2020).
Like phosphoric acid, it was stated in a study that the surface was functionalized with boric acid that strong Lewis acid sites could be formed due to the electron-deficient boron atom(Ghosh and Curthoys 1983).
Therefore, both Lewis base sites and Lewis acid sites can be mentioned on this catalyst surface by heterojunction coupling and acid treatments. Lewis acid domains play an important role in catalytic reactions. The results obtained from XPS and EDX analyzes proved the presence of P and B atoms in the g-C3N4 and zeolite structure. Thus, on this catalyst surface, P+ and B+ centers can act as Lewis acid sites, and amine or imine groups in g-C3N4 can act as Lewis base sites(Zhou et al. 2015). In addition, the binding of more electronegative O atoms will cause P and B atoms to be positively affected in Lewis acidity. P+ and B+ centered positively charged groups can promote faster H2 formation by providing faster adsorption of ions in CH3OH to the catalyst surface. Therefore, it can be said that the obtained B and P doped zeolite-g-C3N4 plays an important role in the H2-P from NaBH4-MR.
Figure 8
The H2-P from NaBH4-MR with B and P doped zeolite-g-C3N4 catalyst (10 mg) were investigated at solution temperatures ranging from 20 to 50 oC, and the results are given in Fig. 8. As expected, the H2 production rate increased in methanol with the increase in solution temperature. For example, the HGR values obtained at temperatures between 20 and 50 oC are 2318 an 14857 ml min− 1g− 1, respectively. It can be mentioned about a 7-fold increase in the HGR value. The activation energy (Ea) for the NaBH4-MR was determined by the Arrhenius equation.
lnk = l nA − Ea/RT (7)
where Ea, k, R, A, and T denote the activation energy, the rate constant for the reaction, the gas constant, the Arrhenius constant, and the absolute temperature, respectively. As given in Fig. 9, the Ea value for the reaction catalyzed by B and P doped zeolite-g-C3N4 catalyst from the slope of the Arrhenius plot was found to be 29.39 kJ mol− 1. The Ea value obtained in Table 4 is compared with the catalytic systems used for H2-P. It can be stated that the obtained Ea value is quite reasonable. Also, the maximum HGRs of the B and P doped zeolite-g-C3N4 catalyst catalyst and other reported catalysts are listed in Table 4 (Sahiner and Sengel 2016, 2017; Sahiner et al. 2017; Saka et al. 2020; Wang et al. 2020a, b; Zhang et al. 2020; Saka 2022e, a, f, 2021f, c, e, d, b; Inger et al. 2021; Dai et al. 2021; Saka and Balbay 2021; Yang et al. 2021).
Table 4
Comparison of HGR and Ea values for B and P doped zeolite-g-C3N4 for H2-P from NaBH4-MR with the other catalyst
Catalyst
|
HGR (ml min− 1 gcat−1)
|
Ea (kJ mol− 1)
|
Reference
|
C-KOH-S-P
|
13000
|
18.94
|
(Saka 2021b)
|
S-KOH-S-P
|
18571
|
12.54
|
(Saka 2022e)
|
O doped g-C3N4
|
10800
|
36.13
|
(Saka 2022a)
|
P (TAEA-co-GDE)-HCl
|
3018
|
30.37
|
(Sahiner and Sengel 2016)
|
N-AC-N
|
16250
|
11.45
|
(Saka and Balbay 2021)
|
HNT–NH2–HCl
|
220
|
30.41
|
(Sahiner and Sengel 2017)
|
Co-Mo-P/CNTs-Ni foam
|
2640
|
49.94
|
(Wang et al. 2020b)
|
CMS-ZnCl2-Cu-B
|
4730
|
22.71
|
(Saka et al. 2020)
|
SiO2@PAA
|
5120
|
24.03
|
(Yang et al. 2021)
|
S-AC-N
|
10105
|
39.75
|
(Saka 2021d)
|
T-PEI+
|
4408
|
36.1
|
(Inger et al. 2021)
|
p(2-VP)++C6
|
1664
|
20.8
|
(Sahiner et al. 2017)
|
SP-KOH-P
|
19500
|
38.79
|
(Saka 2021f)
|
P doped g-C3N4
|
8666
|
30.29
|
(Saka 2022f)
|
g–C3N4–EDTA-H
|
7571
|
35.6
|
(Saka 2022b)
|
ZIF-67@GO-2
|
3200
|
|
(Dai et al. 2021)
|
B and P doped g-C3N4-zeolite
|
6250
|
29.39
|
This study
|
Figure 9
Reusability experiments were performed five times in succession with 10 mg of B and P doped zeolite-g-C3N4 catalyst, 10 ml of CH3OH and 0.125 g of NaBH4 at 30 °C (Figure 10). After the release of H2 gas in each experiment, it was continued with 0.125 g NaBH4 again. On first use, 100% NaBH4 conversion and H2 release were completed in 7 minutes. However, on the fifth use, the H2 release time increased to 10.5 minutes. Besides, 100% conversions were obtained for each use. The possible cause of the reduction in the rate of the reaction may be the increased viscosity of the reaction solution or the effect of the reaction by-product blocking the catalyst active sites(Gao et al. 2019).
Figure 10