Stalin
Joseph
a,
Devaraju M.
Kempaiah
*a,
Mercy
Benzigar
a,
Arun V.
Baskar
a,
Siddulu N.
Talapaneni
*a,
Sung Hwa
Jhung
b,
Dae-Hwan
Park
a and
Ajayan
Vinu
*a
aFuture Industries Institute (FII), University of South Australia (UniSA), Mawson Lakes Campus, Mawson Lakes, SA 5095, Australia. E-mail: Devaraju.MurukanahallyKempaiah@unisa.edu.au; siddulunaidu@gmail.com; Ajayan.Vinu@newcastle.edu.au
bDepartment of Chemistry, Green-Nano Materials Research Center, Kyungpook National University, Daegu 41566, Republic of Korea
First published on 20th September 2017
In this work, we report a simple and versatile method for the preparation of mesoporous carbon nitrides (MCNs) functionalized with highly dispersed chromium oxide nanoparticles by using a metal organic framework, MIL-100(Cr), as a template and aminoguanidine hydrochloride (AG) as a high nitrogen content single molecular precursor. We are able to synthesise these metal oxide functionalized MCN materials with single step carbonization but without using any toxic template removal process using HF or NaOH. The absence of a washing procedure with toxic acid also allows the incorporation of a large amount of metal oxide particles inside the porous channels of MCNs. The obtained MCN materials exhibit a high specific surface area and a large pore volume. The AG to template ratios are varied to control the amine functional groups and the textural parameters including the specific surface area and pore volume. It is found that the AG to template ratio of 1.5 is the best condition to obtain MCNs with a specific surface area of 1294 m2 g−1, which is the highest value reported so far for MCN-based materials. FT-IR and XPS results reveal that the prepared materials contain free NH2 groups within the CN network which help to anchor metal oxide nanoparticles and provide highly dispersed basic sites. These functionalized MCN materials are also used as adsorbents for CO2 capture. Among the materials studied, the MCN with the highest specific surface area shows the largest CO2 adsorption capacity (16.8 mmol g−1) which is much higher than those of MCN materials prepared from SBA-15 and KIT-6, activated carbon, MIL-100(Cr), SBA-15, and multiwalled carbon nanotubes. This high adsorption capacity is mainly due to the strong synergistic effect between the MCN with high specific surface area and highly dispersed metal oxide nanoparticles.
MCNs are generally synthesized by the nanohard templating approach using mesoporous silicas as templates.12 The properties of the MCN materials can be controlled with a simple adjustment of the structure and pore diameter of the templates and the nature of the nitrogen precursors. With this approach, several MCNs with a large surface area, tuneable pore diameters, and different nitrogen contents have been prepared and used for various other applications including catalysis, sensing, and energy storage and conversion.13–17 For example, since the first report of MCNs by Vinu et al. in 2005,18 various templating approaches such as hard nanocasting and soft self-assembly have been employed for the synthesis of different MCN materials.18–22 Among them, the nanocasting method was found to be an effective way to prepare highly ordered MCN materials and various porous silicas with different structures such as SBA-15, SBA-16,23 KIT-6, and IBN-4 have been widely used as templates.19–21,24–26 This silica templating process generally includes three main steps: (i) infiltration of precursors into the silica template; (ii) polymerization and cross-linking of the carbon and nitrogen precursors into the MCN material at high temperature; and (iii) removal of the silica template with a toxic washing procedure using HF. However, these traditional porous silica templates are expensive and complicated as they involve highly toxic substances such as HF to remove the silica, which limits their use in multifunctional applications with large scale production. By replacing the inorganic templates with organic templates, various steps including the toxic HF treatment, washing with expensive solvents such as ethanol and long synthesis time could be avoided. This simple and single step approach could remove the hurdle in the path of the commercialization of MCNs and help to produce MCN materials in a safe and cost effective way.
Metal–organic frameworks (MOFs) have been considered as a new kind of porous material and also attracted much attention owing to their high adsorption capacities and tuneable pore diameters which make them suitable for various potential applications such as gas separation, storage, catalysis and drug delivery.27–31 The unlimited combinations of organic and inorganic building blocks have offered enormous flexibility to MOFs with high chemical and thermal stability. Among these MOFs, MIL-100(Cr)32 with a chemical composition of Cr3O(F/OH)(H2O)2[C6H3(CO2)3]2·nH2O has great potential due to its large surface area up to 3340 m2 g−1, high pore volume (1.16 cm3 g−1), accessible microporous windows of ca. 0.5 and 0.9 nm, unsaturated chromium sites, unexpectedly high thermal stability (above 300 °C), and rigid structural integrity even after prolonged immersion in water.33 These exciting properties make it an excellent organic templating agent for the fabrication of mesoporous materials including MCNs. However, most of the research on MIL-100(Cr) focused mainly on the structural functionalization and applications in selective adsorption, separation, storage, and catalysis.34,35 By utilising this unique material as a template, not only mesoporous materials could be generated but also with ultra-fine and highly dispersed metal oxide nanoparticles in the pore channels of the mesoporous materials. However, so far, no research has been conducted on using this material as a template for the preparation of MCNs.
Here we describe a silica free, simple and direct synthesis of MCN materials functionalized with chromium oxide nanoparticles and with a large surface area using MIL-100(Cr) as a hard template via a simple thermo-polymerisation of aminoguanidine as a single molecular C and N precursor. To the best of our knowledge, this is the first report on the preparation of MCN materials using a microporous MOF as a sacrificial template along with the incorporation of chromium oxide onto the surface of the MCN materials. The advantages of using porous MIL-100(Cr) as a template are the presence of nanoscale cavities and an open channel structure which could be beneficial in infiltrating the small CN precursor molecules into its open channels to obtain MCN with a high surface area and large pore volume. The prepared MCN exhibits much higher CO2 adsorption capacity than activated carbon, MWCNTs, mesoporous silica, the parent template MIL-100(Cr) and MCN prepared from the SBA-15 template. This simple but effective method can be utilized for the preparation of various MCN nanostructures from different MOFs.
Fig. 1 (A) Wide angle XRD patterns of Cr-MCN-10-X and the template MIL-100(Cr), (B) and (D) HRSEM images of Cr-MCN-10-1.5 and (C) energy dispersive X-ray spectroscopy (EDS) patterns of Cr-MCN-10-X. |
The HRSEM images of the Cr-MCN-10-1.5 are shown in Fig. 1B and D. The Cr-MCN-10-1.5 possesses highly uniform spherical shaped particles that are highly interconnected and similar to those of the template. The size of the particles is in the range of 100 to 200 nm. It is interesting to note that the macropores are formed between the particles whose sizes are in the range of a few to tens of micrometers. This may be due to the decomposition of the loosely held CN polymers that are decomposed at a high temperature, leaving large mesopores. The energy dispersive X-ray spectroscopy (EDS) pattern of the Cr-MCN-10-1.5 exhibits four sharp peaks which correspond to C, N, O and Cr, revealing that the product is composed of carbon nitride and chromium oxide nanoparticles (Fig. 1C). From the intensity of the Cr peaks, it can be concluded that most of the Cr from the template is transferred to the final Cr-MCN-10 samples as chromium oxide nanoparticles. It is assumed that the Cr atoms reacted with the oxygen of the carboxyl groups in the templates to form the chromium oxide particles. The absence of the chlorine peak in the EDS pattern confirms that the precursor has completely undergone polymerization to form CN polymers and the hydrogen chloride from the precursor is completely removed during the high temperature carbonization process.
It is interesting to note that the intensity of the N peak in the EDS pattern increases concomitantly with the decrease in the intensity of the Cr peak as the amount of AG in the synthesis mixture is increased. This reveals that the content of chromium in the samples prepared with a high amount of AG is lower as compared to that of Cr-MCN-10-1.5. The N/C ratio calculated from the EDS is 0.95–0.46 which is lower than that obtained from the mesoporous silica-based templating procedure (1.33 to 1.9), suggesting that carbon from the template takes part in the reaction.39 This is again confirmed by the fact that the N/C ratio increases with increasing the amount of the precursor in the synthesis mixture. It should also be noted that the particles are agglomerated when the loading of AG is increased in the synthesis mixture (Fig. 1S†). This might be attributed to the formation of CN polymers around the template particles which tend to interact with the neighbouring particles during the polymerization process.
The porosity of the Cr-MCN-10-1.5 is further investigated by high resolution transmission electron microscopy (HRTEM). The HRTEM images reveal that the material is mainly composed of small and slightly disordered mesopores that are quite interconnected and in the size range of 2 to 3 nm (Fig. 2A and B). The elemental mapping of Cr-MCN-10-1.5 further confirms the presence of C, N, O and Cr in the sample. A homogenous distribution of these elements throughout the sample further reveals that the sample is highly uniform and chromium oxide nanoparticles are evenly distributed in the porous channels of the MCN matrix. A similar mesostructure and homogeneous distribution of elements are also observed for other Cr-MCN-10 samples (Fig. 2S to 6S†).
Fig. 2 (A) HRTEM image, (B) bright field scanning TEM image, and (C–F) elemental mapping of Cr-MCN-10-1.5, showing the distribution of carbon, nitrogen chromium and oxygen, respectively. |
Fig. 3 (A) Nitrogen adsorption–desorption isotherms and (B) NLDFT-pore size distributions of Cr-MCN-10-X samples. |
Sample | Surface area (m2 g−1) | Pore volume (cm3 g−1) | Micropore area (m2 g−1) | Pore diameter (nm) | EDS analysis (wt%) | ||||
---|---|---|---|---|---|---|---|---|---|
C | N | Cr | O | N/C | |||||
Cr-MCN-10-1.5 | 1294 | 0.42 | 771 | 1.60, 2.10 | 52.1 | 24.1 | 6.6 | 17.2 | 0.46 |
Cr-MCN-10-2.0 | 1013 | 0.29 | 641 | 1.57, 2.03 | 53.8 | 25.8 | 3.1 | 17.3 | 0.48 |
Cr-MCN-10-2.5 | 921 | 0.27 | 583 | 1.57, 2.00 | 42.4 | 40.7 | 0.8 | 16.1 | 0.95 |
Fig. 4 (A) Electron energy loss spectrum, (B) FT-IR spectrum, (C) UV-Vis spectrum and (D) bandgap data of the Cr-MCN-10-1.5 sample. |
Information on the chemical bonding between the carbon and nitrogen atoms in the Cr-MCN-10 material was also obtained from FT-IR. The FT-IR spectrum of Cr-MCN-10-1.5 is shown in Fig. 4B. The stretching mode observed in the region 1628 and 1560 cm−1 is assigned to the characteristic vibrations of aromatic CN, which is also observed for non-porous CN prepared from AG precursors. The peaks at 943 and 1370 cm−1 also represent the typical tetrazine rings.46 Another band at 1450 cm−1 is assigned to the graphitic C–N band owing to the aromatic ring in the structure.47 The stretching mode between 3150 and 3350 cm−1 is attributed to the N–H stretching of NH2 groups attached to the sp2 hybridized carbon framework, which coincides with the previous reports.12,25 These results obviously support the presence of a CN polymeric network of the carbon nitride structure.
The UV-visible diffuse reflectance spectrum of Cr-MCN-10-1.5 and its corresponding bandgap data are shown in Fig. 4C and D, respectively. Two broad peaks are observed at 250 and 320 nm and a large broad peak in the visible region (400 to 600 nm) was also observed. The peaks at lower wavelengths are attributed to the π–π* or n–π* transitions in the aromatic CN framework. It is interesting to note that the bandgap of Cr-MCN-10-1.5 is determined to be 2.2 eV from the ultraviolet-differential reflectance spectrum as shown in Fig. 3D. The pure MCN materials have the energy bandgap of 2.7 eV whereas the MCN prepared from AG using mesoporous silica as templates exhibits the bandgap of 2.25–2.49 eV. After incorporation of Cr into the CN structure, the bandgap of the material is decreased from 2.7 to 2.2 eV. The lower bandgap is highly necessary for the photocatalytic application involving water splitting or photocatalytic degradation. The presence of Cr-based nanoparticles will also provide catalytically active redox sites or more basicity to the materials. It has been reported that24 the bandgap of porous carbon nitride materials could be tuned in the range of 2.26 to 2.69 eV with a simple adjustment of the type and nature of the bonding structure and/or the nitrogen contents. It is worth noting that we have successfully prepared the Cr-MCN-10 with a much lower bandgap in the presence of chromium without any doping reagent.
X-ray photoelectron spectroscopy (XPS) analysis was also carried out for the Cr-MCN-10-1.5 and the results are shown in Fig. 5A–D. The XPS survey spectrum shows the peaks related to Cr, C, N and O (Fig. 5A). Interestingly, the intensity of the nitrogen peaks increases with increasing the amount of AG in the synthesis mixture. This indicates that the nitrogen content of the materials increases with the addition of more amount of AG. The presence of the O peak confirms the presence of a trace amount of oxygen which might have originated from the adsorbed CO2 or chromium oxide nanoparticles. This result is consistent with the data obtained from EDS analysis. The chemical bonding between carbon, nitrogen and chromium atoms has been obtained from the high resolution C 1s, N 1s and Cr 2p spectra of Cr-MCN-10-1.5 which are shown in Fig. 5. The deconvoluted C 1s spectrum shows four peaks around 288.1, 286.9, 285.7, and 284.6 eV (Fig. 5B). The large peak at the lowest energy contribution is due to the sp2 hybridized carbon in the sample which is quite different from the sample prepared from mesoporous silica as the template. The peak at 285.7 eV is attributed to the sp2 hybridized carbon bonded to nitrogen inside the aromatic ring whereas the peak at 286.9 eV is ascribed to carbon bonded with nitrogen present in the aromatic tetrazine ring (Scheme S1†).21 The highest energy contribution at 288.1 eV might have originated from the carbon bonded with the trigonal nitrogen atoms.24,42 In the N 1s XPS spectra, we can clearly observe two strong peaks centered at 398.7 and 400.2 eV (Fig. 5C). The lowest energy contribution is assigned to nitrogen atoms bonded with sp2 hybridised carbon whereas the highest energy contribution is attributed to the trigonal nitrogen phase bonded to sp2 hybridised carbon. The observed binding energies of C 1s and N 1s spectra closely matched with the reported binding energies of MCN materials.48,49 The Cr 2p spectrum reveals that the binding energies of Cr 2p3/2 and Cr 2p1/2 fall at 576.5 and 586.6 eV, respectively (Fig. 5D). These peaks are attributed to Cr3+ and Cr6+ in Cr2O3 and CrO3 nanoparticles in MCN. Both the Cr 2p3/2 and Cr 2p1/2 peaks are deconvoluted into two valence states: low-valence (Cr3+) at 576.6 eV and high-valence (Cr6+) at 579.3 eV, respectively, indicating a mixed phase of Cr2O3 and CrO3 nanoparticles.50–52 These results confirm that ultra-fine chromium oxide nanoparticles which are formed on the porous channels of MCN after the decomposition of the MIL-100(Cr) template are present in the form of chromium oxide in Cr-MCN-10.
Fig. 5 (A) XPS survey spectra of Cr-MCN-10-X samples, (B) deconvoluted C 1s spectra, (C) N 1s spectra and (D) Cr 2p spectra of the Cr-MCN-10-1.5 sample. |
In order to verify the superiority of the Cr-MCN-10 materials for the CO2 adsorption, the CO2 adsorption capacity of Cr-MCN-10-1.5 was compared with those of the template, MIL-100(Cr), activated carbon, multiwalled carbon nanotube and MCN prepared from mesoporous silica templates. It is interesting to note that the CO2 adsorption capacity of Cr-MCN-10-1.5 is much higher than those of the template MIL-100(Cr) (9.8 mmol g−1), activated carbon (3.6 mmol g−1), and other carbon nitride MCN-1-130 (14.5 mmol g−1).6,12,36,37 These results demonstrate the superior nature of the Cr-MCN-10 that was prepared in a single step without any hazardous HF washing procedure for the CO2 adsorption. The effect of adsorption temperature on the CO2 adsorption capacity was also investigated. Fig. 6C shows the CO2 adsorption isotherms of Cr-MCN-10-1.5 measured at different temperatures (0, 10 and 25 °C). As can be seen in Fig. 6C, the amount of CO2 adsorbed increases with increasing the analysis temperature and the adsorption temperature of 0 °C was found to be the best. The nature of the interaction between the adsorbate and the adsorbent was analyzed by calculating the isosteric heat of adsorption using the Clausius–Clapeyron equation. The sample exhibits a heat of adsorption of 25 kJ at low CO2 adsorption to 21 kJ as the amount of CO2 is increased. The value of heat of adsorption at higher CO2 loading is almost similar to those of MCN-1 materials but higher than those of mesoporous carbon-based adsorbents. This clearly explains the role of nitrogen functionalities and chromium oxides in the CO2 adsorption. In addition, the excellent CO2 adsorption performance is the result of weak van der Waals forces and the capillary condensation effect due to the mesoporous and microporous nature of the pores and the chemical bonding with the CO2 and specific functionalities owing to the presence of abundant Lewis base sites in the active catalyst.53
The hydrogen adsorption properties of the prepared Cr-MCN-10 samples were also investigated and the results are compared with those of the template MIL-100(Cr) (Fig. 7). It should be noted that Cr-MCN-10-1.5 registered the hydrogen adsorption capacity of 22.5 mmol g−1 which is much higher than those of other Cr-MCN-10 samples and the template MIL-100(Cr). Again, the higher hydrogen adsorption capacity of this material is attributed to the large surface area and the presence of a small amount of chromium oxide nanoparticles that offer a synergistic effect which is crucial for enhancing the amount of hydrogen adsorption. All these results clearly confirm that the prepared materials are excellent candidates for the adsorption of hydrogen or CO2 molecules and have the potential to replace the existing highly ordered porous materials owing to their ease of preparation and low cost.
Fig. 7 (A) H2 adsorption isotherms of Cr-MCN-10-X samples at zero degrees and (B) comparison of H2 adsorption isotherms of Cr-MCN-10-X samples with the template MIL-100(Cr). |
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c7ta06969c |
This journal is © The Royal Society of Chemistry 2017 |