Corresponding author

Bettina Lotsch

Max Planck Institute for Solid State Research

References

1.
Wang, X.; Maeda, K.; Thomas, A.; Takanabe, K.; Xin, G.; Carlsson, J.M.; Domen, K.; Antonietti, M.
A metal-free polymeric photocatalyst for hydrogen production from water under visible light
2.
Schwinghammer, K.; Tuffy, B.; Mesch, M.B.; Wirnhier, E.; Martineau, C.; Taulelle, F.; Schnick, W.; Senker, J.; Lotsch, B.V.
Triazine-based carbon nitrides for visible-light-driven hydrogen evolution
3.
Schwinghammer, K.; Mesch, M.B.; Duppel, V.; Ziegler, C.; Senker, J.; Lotsch, B.V.
Crystalline carbon nitride nanosheets for improved visible-light hydrogen evolution

In collaboration with:

M.B. Mesch, J. Senker (Inorganic Chemistry III, University of Bayreuth)

Research Group "Nanochemistry"

Crystalline carbon nitride nanosheets for improved visible-light hydrogen evolution

Authors

K. Schwinghammer, M. B. Mesch, V. Duppel, Ch. Ziegler, J. Senker, B. V. Lotsch

Departments

Research Group "Nanochemistry"

We present the one-step aqueous exfoliation of the layered carbon nitride poly(triazine imide), which leads to colloidally stable suspensions of ultrathin crystalline nanosheets of 1–2 nm in height. Successful delamination was verified by AFM and TEM. The exfoliation process was investigated by ssNMR of 15N-enriched PTI nanosheets. The nanosheets show a significantly enhanced photocatalytic activity for hydrogen evolution from water as compared to their bulk counterpart, which is testament to the importance of morphology control in photocatalytic water splitting.

Introduction

In light of the gradual exhaustion of fossil resources, the exploration of renewable energy sources is of key importance. Since the disposal and storage of nuclear waste is considered problematic, clean energy resources such as biomass, wind, water, geothermal and solar power are in high demand. In fact, hydrogen is an exemplary zero-emission energy carrier with high calorific density, which can be considered a "green" source of energy when obtained by photocatalytic water splitting. Photocatalysts for light-induced hydrogen evolution have long been composed exclusively of inorganic and to some extend scarce and toxic materials, presenting a bottleneck with respect to their large-scale utilization. In 2009, a metal-free carbon nitride called melon was introduced as an environmentally friendly and stable non-metal photocatalyst [1]. The moderate efficiency of carbon nitride photocatalysts can be enhanced by band gap engineering through doping with heteroatoms as well as organic molecules, by hybridization to counteract charge recombination, and through morphology control to increase the surface area and exposed active sites.

<strong><strong>Fig. 1: </strong></strong>Idealized PTI structure (lithium / chloride intercalation omitted for clarity) viewed along the slightly tilted <em>b</em>-axis (a) and the <em>c</em>-axis (b). AFM image of exfoliated PTI nanosheets deposited on a Si/SiO<sub>2</sub> wafer (c), the corresponding height profile of the blue line drawn across the image (inset) and the magnification of a hexagonal prismatic shaped PTI crystallite (d).<br />Adapted with permission from Ref. [3]. Copyright (2014) American Chemical Society. Zoom Image
Fig. 1: Idealized PTI structure (lithium / chloride intercalation omitted for clarity) viewed along the slightly tilted b-axis (a) and the c-axis (b). AFM image of exfoliated PTI nanosheets deposited on a Si/SiO2 wafer (c), the corresponding height profile of the blue line drawn across the image (inset) and the magnification of a hexagonal prismatic shaped PTI crystallite (d).
Adapted with permission from Ref. [3]. Copyright (2014) American Chemical Society. [less]

Recently, we have introduced the first two-dimensional (2D) carbon nitride – poly(triazine imide) (PTI) – as a polymeric photocatalyst, which rivals the commonly used amorphous melon in terms of photocatalytic efficiency [2]. In contrast to melon, which is built up of one-dimensional chains of heptazine units, PTI is a crystalline carbon nitride with an interconnected 2D carbon nitride backbone, which lends itself as an excellent model system to explore both structure, morphology and photocatalytic activity as a function of the exfoliation state. PTI/Li+Cl- is synthesized in an ionothermal reaction starting from dicyandiamide in a LiCl/KCl salt melt. The structure of PTI (Fig. 1(a)–(b)) is composed of layers of imide-bridged triazine units, which are stacked in an AB-type fashion to form a 2D network with channels running along the c-axis and filled with halide ions (Cl­– or Br). The charge compensating lithium ions are situated in structural pores within the layers. It has already been shown that layered materials such as graphite and boron nitride can be exfoliated into nanosheets when the enthalpy of mixing is minimized, which holds true when the surface energies of the solvent and the nanosheets match. For carbon nitrides water appears to be an ideal solvent considering their similar surface energies and polar character, which can cause swelling and exfoliation due to favourable hydrogen bond formation. Here we demonstrate the one-step synthesis of crystalline PTI nanosheets by aqueous exfoliation without the need for toxic solvents, preintercalation steps, or additives [3]. This method leads to highly crystalline nanosheets of 1-2 nm in height, which are significantly more active towards visible-light driven water reduction as compared to bulk crystalline PTI/Li+Cl.     

Synthesis and structural characterization

Crystalline PTI/Li+Cl was suspended in water (2 mg mL–1) and sonicated for 15 h at room temperature. The dispersion was centrifuged at 3000 rpm to remove aggregates (Precipitate I), yielding a homogeneous suspension of PTI nanosheets (Suspension 1). To further separate the dispersion according to the degree of exfoliation, Suspension 1 was further centrifuged at 5000 rpm (Precipitate II) to obtain a nanosheet suspension with yet a higher degree of exfoliation (Suspension 2; 0.2 mg mL–1). In order to compare the suspended nanosheets with their restacked form, Supernatant 2 was precipitated by centrifugation at 25000 rpm (Precipitate III). The PTI nanosheet suspension (Supernatant 2) is stable for months based on the high negative surface charge of PTI (zeta potential of –54.0 mV at pH 10.5), which is caused by the dynamic lithium-proton exchange of the bridging imide moieties. Whereas the carbon-to-nitrogen ratio (0.62) of PTI nanosheets is similar to that of bulk PTI (0.64), a decrease in lithium content was observed through inductively coupled plasma (ICP) measurements and X-ray photoelectron spectroscopy (XPS). Interestingly, the surface charge is minimal at pH 5.4 and reversed at lower pH with a zeta potential of +30 mV at pH 2, suggesting an overall amphoteric character of the PTI backbone.

<p><strong>Fig. 2:</strong> TEM image of exfoliated ultrathin PTI nanosheets (a) and the SAED pattern of the marked nanoplatelet (inset); (b) higher magnification of a PTI nanosheet edge viewed along [001] and simulation (JEMS; &Delta;<em>f</em> = +50&nbsp;nm, <em>t</em> = 2.70 nm; inset). <sup>1</sup>H MAS NMR spectra for Precipitates I-III and bulk PTI, respectively (c). <sup>1</sup>H&ndash;<sup>1</sup>H proton driven spin diffusion spectrum recorded for Precipitate III (d).<br />Adapted with permission from Ref. [3]. Copyright (2014) American Chemical Society.</p> Zoom Image

Fig. 2: TEM image of exfoliated ultrathin PTI nanosheets (a) and the SAED pattern of the marked nanoplatelet (inset); (b) higher magnification of a PTI nanosheet edge viewed along [001] and simulation (JEMS; Δf = +50 nm, t = 2.70 nm; inset). 1H MAS NMR spectra for Precipitates I-III and bulk PTI, respectively (c). 1H–1H proton driven spin diffusion spectrum recorded for Precipitate III (d).
Adapted with permission from Ref. [3]. Copyright (2014) American Chemical Society.

[less]

 Analysis of the as-prepared PTI nanosheets (Supernatant 2) by atomic force microscopy (AFM) revealed exfoliated hexagonal prismatic crystallites with lateral sizes of less than 100 nm and a height of 1-2 nm, indicating that the exfoliated nanosheets are composed of only a few carbon nitride layers, taking into account a water shell likely surrounding the nanosheets (Fig. 1 (c)–(d)). Investigations by transmission electron microscopy (TEM) confirmed the retained hexagonal shape of the crystallites, which are well isolated and spread out across the substrate (Fig. 2 (a)–(b)). Selected area electron diffraction (SAED) patterns are in full agreement with the expected hexagonal symmetry of a PTI layer. The various precipitates of the centrifugation process pertaining to different stages of exfoliation (Precipitate I–III) show reflections which are consistent with bulk PTI, again confirming that the parent structure is retained in the nanosheets. Since the exfoliation process does not lead to a complete removal of the intercalated ions, centrifugation of the nanosheets leads to restacking of the crystallites without apparent turbostratic disorder. XPS and infrared (IR) spectroscopy further confirm the identity of the nanosheets. The characteristic IR as well as XPS spectrum of the PTI nanosheets is largely reminiscent of those of the bulk material apart from slight broadening and band shifts due to hydration of the sheets.

To gain more insights into the local structure of the nanosheets relative to bulk PTI, we performed solid-state nuclear magnetic resonance (ssNMR) spectroscopy with 15N enriched samples of Precipitates I-III and bulk PTI (Fig. 2(c)–(d)). While being overall similar, both 15N and 13C spectra reveal slight differences in the relative signal intensities on going from the bulk to the restacked nanosheets, which can be attributed to a decrease in the lithium content caused by the sonication process. In contrast, the water content in the samples, apparent from 1H NMR spectra (6.2 and 4.6 ppm), gradually increases from Precipitate I to III. 1H–1H proton driven spin diffusion experiments point to a close vicinity of the nanosheets and water on a molecular scale, which is consistent with dense monolayers of water on the surface of the nanosheet stacks. Based on the relative proton signal intensities, we deduce that 2 to 4 PTI layers are interspersed with two water layers, which corresponds well with the height of the nanosheet stacks measured by AFM (1–2 nm). 

<strong><strong>Fig. 3:</strong> </strong>Photocatalytic activity toward water reduction of the Pt-doped Precipitates (I and II) compared to sonicated PTI and bulk PTI (a), and of the nanosheet suspensions compared to bulk PTI (b), measured in a 10&nbsp;vol% aqueous TEoA solution for 3&nbsp;h illumination with visible light (&gt; 420&nbsp;nm). Cyclic stability tests of the nanosheet suspension with MeOH as electron donor (c). The reactor was purged several times (marked with an asterisk) to avoid saturation, and MeOH was re-injected twice.&nbsp; UV/vis spectrum of bulk PTI compared to the nanosheet suspension (Supernatant 2) (d).<br />Adapted with permission from Ref. [3]. Copyright (2014) American Chemical Society. Zoom Image
Fig. 3: Photocatalytic activity toward water reduction of the Pt-doped Precipitates (I and II) compared to sonicated PTI and bulk PTI (a), and of the nanosheet suspensions compared to bulk PTI (b), measured in a 10 vol% aqueous TEoA solution for 3 h illumination with visible light (> 420 nm). Cyclic stability tests of the nanosheet suspension with MeOH as electron donor (c). The reactor was purged several times (marked with an asterisk) to avoid saturation, and MeOH was re-injected twice.  UV/vis spectrum of bulk PTI compared to the nanosheet suspension (Supernatant 2) (d).
Adapted with permission from Ref. [3]. Copyright (2014) American Chemical Society. [less]

The brown color of the PTI nanosheet suspension (Fig. 3(d)) indicates substantial absorption in the visible range of the spectrum, similar to bulk PTI. Since there are only minute changes in the absorption spectra and the determined optical band gap (2.6 eV) between the PTI nanosheets and the parent bulk material, quantum confinement effects likely do not operate. Owing to the small particle size and, thus, higher exposed surface area, the PTI nanosheets were tested for photocatalytic hydrogen evolution and compared to the water reduction efficiency of the bulk material. The photocatalytic activities (Fig. 3(a)–(b)) of the (re)dispersed photocatalysts were investigated in a water/triethanolamine (TEoA) solution under visible light illumination (> 420 nm) using 2.3 wt% of Pt nanoparticles as co-catalyst. Under these conditions, bulk PTI evolves 4.3 µmol H2 per hour, whereas partially exfoliated PTI (Precipitate II) shows an improvement in hydrogen evolution by 28% (6.4 µmol H2 h–1). Using the PTI nanosheet suspension (Supernatant 2; containing approximately 2 mg of nanosheets) instead leads to a tremendous enhancement by a factor of 18 (3.5 µmol H2 h–1) compared to the bulk material (0.2 µmol H2 h–1 when 2 mg material was tested). More highly concentrated nanosheet suspensions show an even higher photocatalytic activity, suggesting that hydrogen evolution is not yet diffusion limited under these conditions and light harvesting is not yet impeded by scattering effects. Hydrogen is steadily evolved for at least 130 h during long-term measurements when methanol was chosen as sacrificial electron donor (Fig. 3(c)). Furthermore, wavelength-dependent studies show that the apparent quantum efficiency is maximal at 400 nm.

Conclusion

Crystalline, exfoliated PTI shows the highest activity towards photocatalytic water reduction among pristine (i.e. non-doped) carbon nitrides. The importance of morphology-related effects such as surface area and, hence, active site exposure, adds to a better understanding of structure-property-activity relationships in carbon nitrides and other polymeric photocatalysts.

 
loading content