Structural Analyses
In order to realize whether the designed composite has been prepared successfully or not, FT-IR spectroscopy was utilized. Given by Fig. 1a, the FT-IR spectra of intact forms of GO, MIL-100(Fe), and GO/MIL-100(Fe) hybrid are observed. First of all, there is a broad signal detected at 3000–3500 cm− 1 range in the spectra of three prepared samples, showing the hydroxyl groups belong to absorbed water molecules 34. In the spectrum of MIL-100(Fe), two absorption signals at 1590 and 1390 cm− 1 correspond to the asymmetric and symmetric stretching modes, respectively, that could highlight the presence of the carboxylate motifs related to BTC ligands 35, 36. As well, the characteristic signals at 1590 cm-1 and 767 cm-1 are correlated with the vibration frequencies of C = C and C-H bonds, indicating the presence of a benzene ring as the ligand's core 37. The spectrum of GO showed that GO was totally reduced by solvothermal treatment since there was no absorption associated with –OH, C = O, or strained C-O-C (epoxy) modes 38. As evidenced in Fig. 1a, the asymmetric stretching of O = C–O stretching and C–H deformation oscillations were blocked in the uppermost line, which was more relevant with the GO adding. These records showed the successful decoration of GO nano-sheets merged with Fe-MOF nanoparticles 39. Since the low concentration of graphene oxide has been used in the composite, its correlated peaks have become weakened in the spectrum of GO/MIL-100(Fe).
Via a short glance at Fig. 1b, the main differences in PXRD patterns of pure GO, pure MIL-100(Fe), and synthesized GO-MIL-100(Fe) nanocomposite are clearly detectable. The pink-colored XRD pattern of MIL-100(Fe) matches the simulated one reported in previous literature, showing the complete formation of the Fe-based MOF. In the following, it is evident that almost all the characteristic peaks of MIL-100(Fe) have appeared in the pattern of the prepared composite, while no distinguishable broad peak related to GO (at about 9.3o on the black line) is observed. Indeed, under ultrasonication, GO nanosheets become more exfoliated and smaller in size, and more importantly, restricted concertation of GO is used in the preparation of the composite.
Gaining information about the surface area and porosity of the prepared samples is done by Brunauer–Emmett–Teller (BET) analysis. The BET plot of MIL-100(Fe) illustrated the combination of both type I and IV isotherms and huge uptakes of nitrogen gas (N2) at low proportional pressure (P/P0 < 0.1), which indicates its mesoporous nature formed on a micro-scale (Fig. 1c) 40. On the other hand, GO-MIL-100(Fe) showed type II isotherm, indicating the non-porous structure of the prepared composite owing to the introduction of GO nanosheets as macropores in the composite (Fig. 1c). Furthermore, according to IUPAC classification, the isotherm of the developed composite showed an H3-type hysteresis loop that means the existence of sandwich-like pores in the 2-D network. In more detail, MIL-100 octahedrons composite was desirably trapped among GO nanosheets 41, 42. Figure 1c further showed enlarged meso-cages. Both isotherm observation and statistic data (Table 1) showed that what is called the specific surface area has become lower in GO-MIL-100(Fe) compared with primary MIL-100(Fe), which now led to the formation of meso-macropores. The pore texture analysis illustrated that the in-situ growth method might make the GO-MIL-100(Fe) composite prone to extra application beyond the synthesis; It may be able to show enough adsorption performance, as well.
Table 1
Specific surface areas and micropore volumes of three studied species: GO, MIL-100(Fe), and GO-MIL-100(Fe)
Sample | SBET [m2g− 1] |
GO | 83 |
MIL-100(Fe) | 1121.76 |
GO-MIL-100(Fe) | 191.84 |
The stability of the developed GO-MIL-100(Fe) composite against temperature was screened via TG analysis. As displayed in Fig. 1d, the composite showed two main weight loss stages in the 25–600 oC temperature range. The first weight loss occurred below 150 oC, attributed to the exit of solvent molecules from the pores through evaporation. Notably, at 150 to nearly 300 oC, almost no weight loss was observed, which indicates the applicability of the composite up to 300 oC. In the second step, a sharp weight loss happened, starting at 300 oC, and a majority of this weight loss can be due to the degradation of organic ligands laid into Fe-based MOF, which is matched with the weight loss trend of MIL-100(Fe) in literature 43. Nevertheless, a slight amount of this weight loss is related to the decomposition of functional groups that happened in the GO structure, which may be because of their low content in the composite structure as well as the interactions of these functional groups embedded in GO with the open metal sites in MIL-100(Fe) 44.
The SEM measurements were carried out to evaluate the facial features of GO and the fabricated GO-MIL-100(Fe) composite depicted in Fig. 2a-b. This figure shows that GO sheets have a spiral and crinkled morphology, while MIL-100(Fe) nanoparticles show a unique octahedral morphology with random connection to the surface of GO sheets. Energy dispersive x-ray spectroscopy (EDX) analysis is applied to identify the elemental composition of GO-MIL-100(Fe), where the presence of C, O, and Fe was certain, as well as confirming its chemical purity (Fig. 2c).
TEM imaging could clarify the porous and layered structure of GO-MIL-100(Fe) composite (Fig. 2d). It is observed that after the settlement of MIL-100(Fe) nanoparticles on the sheets of graphene oxide, the morphology of the obtained composite has not changed dramatically and it remained as well as before combination (thread-like pattern).
UV-Vis diffuse reflectance spectroscopy (DRS) technique and the related Kubelka Munk plot of the developed GO-MIL-100(Fe) have been shown in Fig. 3a-b. In Fig. 3a., the UV-Vis DRS spectrum of GO-MIL-100(Fe) shows several absorption signals at UV region which mostly is related to the п → п* electron transfer in terephthalic acid, indicating the existence of Fe-based complexed network within the composite. However, the intensity of visible light absorption in the visible light range noticeably is lower than the one in UV region, maybe due to the presence of a slight amount of GO in the composite structure. Moreover, in order to calculate the band gap of the prepared composite, the following equation reported by Butler was used 45:
αhν = A(hν -Eg)n/2
in which Eg, α, h, ν are named band gap energy, absorption coefficient, Planck’s constant and light frequency, respectively. The value of n also depends on the type of transition (n = 1 for direct transition and n = 4 for indirect transition). As Fig. 3b displays, Eg of the GO-MIL-100(Fe) composite is obtained by drawing the plot of (αhν)2 based on energy (hν), giving the value of 2.65eV.
Catalytic Activity Study
We have examined whether the innovative catalyst can handle multi-stage reactions efficiently by using benzyl alcohols in a photocatalytic oxidative functionalization reaction using air bubbling (Fig. 4).
Every model reaction is up to examine the capability of a reagent or any motivator (catalyst), which, here, we aimed at a commonplace oxidation coupled with the Knoevenagel condensation. The effect of the GO ratio in the presence of 5 mg of composites in EtOH at the definite time was the first assessment. Based on the obtained results, it was understood that among the prepared composites, the one with the ratio of 50% GO yielded the highest efficiency compared to other composites. In fact, from this ratio on, the more increase in the content of GO reduces the efficiency of reaction maybe because more catalytic sites are covered by extra GO sheets.
Further optimizations were performed to investigate the effect of some parameters such as amount of the catalyst, and time of reaction, temperature and solvent. As illustrated in Table 2, no product was obtained in the absence of both catalyst and visible light (Table 2, Entry 1). In addition, the constitutive discrete components of the composite and Fe(NO3)3.9H2O were surveyed (Table 2, Entries 2–4) giving the desired product in low yield.
All in all, for this part, the model reaction received support from GO-MIL-100(Fe), showing an increased yield of 57% (Table 2, Entry 5).
In order to narrow down the optimization process even further, the reaction was investigated in various solvents, including EtOH, H2O, CH3CN, DMF, THF, and n-hexane (Table 2, Entries 6–11). However, all these solvents were inappropriate for this reaction. Thus, the model reaction was performed in various DESs (Table 2, Entries 12–14). It was found that the mixture of Glycerol/K2CO3 (2:1) led to the highest yield compared to other DESs usually seen in the chemical literature. To probe the effect of the amount of catalyst _at a 5-by-5 pace of increase_ 5 mg was the closest optimum amount of added catalyst to the actual number (Table 2, Entry 5); though, the higher amount of catalyst was also individually surveyed and showed that the higher mass of catalyst up to 40 mg had no positive effect on the fact that how efficient the referred reaction progresses (Table 2, Entries 15–18), which ultimately brought us back to the earlier optimal mass, 5 mg.
Final step of yield maximization took place around temperature of reaction, the model reaction was performed at different temperatures and the highest yield of desired (dehydrated) product, 2-benzylidenemalononitrile, was obtained at room temperature (r.t.) (Table 2, Entries 19–20).
Table 2
Optimization track of the catalysts used in the PAOF.
Entry | Catalyst | Catalyst (mg) | Solvent | Temp. (°C) | Yield (%) |
1 | - | - | EtOH | rt | - |
2 | GO | 5 | EtOH | rt | 18 |
3 | MIL-100(Fe) | 5 | EtOH | rt | 12 |
4 | Fe (NO3)3.9H2O | 5 | EtOH | rt | 15 |
5 | GO/MIL-100(Fe | 5 | EtOH | rt | 57 |
6 | GO/MIL-100(Fe) | 5 | n-Hexane | rt | 32 |
7 | GO/MIL-100(Fe) | 5 | Acetonitrile | rt | 18 |
8 | GO/MIL-100(Fe) | 5 | MeOH | rt | 17 |
9 | GO/MIL-100(Fe) | 5 | H2O | rt | 13 |
10 | GO/MIL-100(Fe) | 5 | THF | rt | 47 |
11 | GO/MIL-100(Fe) | 5 | DMF | rt | 27 |
12 | GO/MIL-100(Fe) | 5 | Urea/ChCl (2:1) | rt | 62 |
13 | GO/MIL-100(Fe) | 5 | Glycerol/ChCl (2:1) | rt | 57 |
14 | GO/MIL-100(Fe) | 5 | Glycerol/K2CO3 (2:1) | rt | 91 |
15 | GO/MIL-100(Fe) | 10 | Glycerol/K2CO3 (2:1) | rt | 71 |
16 | GO/MIL-100(Fe) | 20 | Glycerol/K2CO3 (2:1) | rt | 73 |
17 | GO/MIL-100(Fe) | 30 | Glycerol/K2CO3 (2:1) | rt | 72 |
18 | GO/MIL-100(Fe) | 40 | Glycerol/K2CO3 (2:1) | rt | 72 |
19 | GO/MIL-100(Fe) | 5 | Glycerol/K2CO3 (2:1) | 50 | 91 |
20 | GO/MIL-100(Fe) | 5 | Glycerol/K2CO3 (2:1) | 70 | 93 |
N.B.: According to the TGA analysis, GO-MIL-100(Fe) composite with 38.5/61.5 (w%/w%) ratio of graphene oxide to MIL-100(Fe) was selected as the most efficient catalyst in photocatalytic condensation reaction at r.t. in (Glycerol/K2CO3 (2:1) (DES) as reaction medium). |
The introduction of the effect of light (visible) is the last station for this work that was analyzed. The photocatalytic activity of the synthesized composite was screened on various benzyl alcohol derivatives, which have been outlined in Table 3 (Under the given condition*). Indeed, both electron donors and electron-withdrawing derivatives were employed, and above 85% yields were obtained after 420 min (= 7 h). As can be seen in Table 3, electron-withdrawing groups such as -NO2 (or partially -Cl) provide much more product (Table 3, entry 4a), whereas, for benzyl alcohols, in which an electron-pumping functional group is attached to its aromatic ring, related products with lower yields resulted (Table 3, entries 1a, 2a, 3a, and 5a). Generally, the more electron can be found at the benzylic site, the faster it proceeds.
Table 3
Light-excited preparation of various 2-benzylidenemalononitrile derivatives in the presence of GO-MIL-100(Fe).
|
Assigned Name
|
1a
|
2a
|
3a
|
4a
|
5a
|
Product
|
|
|
|
|
|
Yield (%)
|
97
|
90
|
93
|
98
|
89
|
*Reaction conditions: Substrates (1mmol), Glycerol/K2CO3 (2:1) (5 mL), 5 mg of the catalyst at room temperature under visible light irradiation for 420 min.
Knowing more about the practical elements, having shown their importance in photocatalytic-oxidative transformations, is done in continue. Outlining the precise impact of each one led us to a clearer view (as its mechanism) of this tandem process. By eliminating the airflow, the conversion rate of benzyl alcohol to benzaldehyde also declined to an 18-percent point, where the essential role of a natural oxidizer could be boldly colored. Moreover, in another control experiment, some quenchers, including tert-butanol (TBA), ammonium oxalate (AO), potassium persulfate (K2S2O8), benzoquinone (BQ), and 2,2,6,6 Tetramethyl-1-piperidinyloxy (TEMPO) were utilized to inactivate radical scavengers including hydroxyl radical (●OH), hole (h+), electron (e−), superoxide free radicals (O2●−) and all radicals, respectively. Addition of TEMPO altered the conversion somehow it fell from 91–14%, indicating the complete quench of all radicals as they have the main role in performing the PAOF. Separately, ammonium oxalate (AO) represented a relieving effect banning excessive oxidation and reducing the reaction efficiency to 26%, indicating the significant role of h + in the reaction mechanism. Nevertheless, using K2S2O8 and TBA had no remarkable effect on the conversion rate, which shows the trivial effect supposed to have arisen from e− and ●OH on the reaction. Eventually, using BQ, just 18% product was gained, proposing the key role of O2●− in the PAOF process under visible light.
Since the GO-MIL-100(Fe) composite has been generally admitted to be a heterogeneous catalyst, hot filtration test was utilized to determine the rate of heterogeneity. In this way, the PAOF was performed catalysis by the prepared composite. Then, after 100 min, the reaction continued without the catalyst and the efficiency of the reaction was controlled by GC-FID. From then on, no changes in the yield of reaction were observed as if no catalyst leakage was there in the reaction mixture. Therefore, it was concluded that the GO-MIL-100(Fe) is totally heterogeneous and is appropriate enough to undergo the norm of the PAOF (Fig. 5a).
After ensuring the heterogeneity of GO-MIL-100(Fe), its stability and reusability were studied. First of all, after the first run of reaction, we did the catalyst re-gathering by centrifugation, washing with ethanol, and drying in an oven at 80°C. Next, this recovered solid catalyst was added to the reaction medium in the second run. This experiment passed via five repetitions, and the developed composite was quite efficient after the 5th run (Fig. 5b).
To confirm our hypothesis, the obtained retrieved catalyst from the 5th cycle was analyzed by the FT-IR and PXRD methods and as Fig. 6a-b displays, the obtained FT-IR spectrum and XRD pattern were almost like the ones had already synthesized, confirming the high integrity and stability of GO/MIL-100(Fe) nanocomposite.
Prediction of Mechanism using Observations and Mathematical Equations
Having observed the physico-chemical levers of this PAOF reaction, theoretically, the valence band (VB) and the conduction band (CB) levels of the MIL-100(Fe) were obtained using the following formulas 46:
EVB = χ - Ee + 0.5Eg (1)
ECB = EVB - Eg (2)
In these equations, EVB and ECB are the VB and CB potentials, respectively. Ee is the energy of free electrons, and Eg is the band gap energy of the semiconductor, which is 3.19 eV for MIL-100(Fe). The last parameter shown as χ is the electronegativity of any semiconductor, which is obtained through the equation below:
Χ = [χ(A)aχ(B)bχ(C)c]1/(a + b + c) (3)
In which a, b, and c stand for the number of atoms in the compounds 47. The χ value for MIL-100(Fe) is obtained equal to 4.64 eV. Thus, the ECB and EVB of MIL-100(Fe) were calculated at about − 0.43 and 2.76 eV. Summarization of the obtained results directed us to a plausible mechanism of condensation reaction has been illustrated in Fig. 7. As shown in this figure, initially, since MIL-100(Fe) has more negative CB (-1.23 eV vs. NHE) than the level of O2/O2•− (-0.33 eV vs. NHE), the photoexcited electrons on the CB of MIL-100(Fe) are transferred to the conjugated system of graphene oxide and convert adsorbed O2 molecules to O2•− radicals. This charge transfer disrupts the recombination of these photo-induced electron-hole pairs, leading to a diverge stance and increasing the photocatalytic activity of the utilized catalyst. It is worth mentioning that the oxidation of alcohols is not carried out by the produced holes in the VB of MIL-100(Fe). In a simpler way to explain, the resulting aldehyde species reacts with malononitrile, and the expected products will be there.