First-principles calculations and modeling of 2D mica nanosheets
In an earlier study15, we carried out first-principles DFT calculations on the structural and electronic properties of muscovite-type (KAl3Si3O10(OH)2) 2D mica nanosheets with a controlled molecular thickness. In the 2D mica nanosheets with different numbers of layers, an abnormal bandgap narrowing was observed, contrary to well-known quantum size effects. Decreasing the number of layers resulted in a reduced bandgap energy from 7 to 2.5 eV, proposing a novel approach to preparing 2D materials with smaller bandgaps, which are needed to prepare devices from 2D heterostructures.
Since mica nanosheets are exfoliated from the host crystallites, they may present different structural properties due to the fact that a large proportion of the atoms are now situated on the surface, as opposed to being confined within the bulk. Accordingly, the structural distortions in the mica nanosheets may contribute to their counterintuitive band narrowing observed in an earlier study15 and change their stability as well. In order to investigate and compare the relative stabilities of single- and few-layered mica nanosheets, we performed a series of calculations on single-layered (1L), double-layered (2L), and triple-layered (3L) mica nanosheets as model systems. We calculated the surface energy of mica nanosheets to quantify the stability by using the following equation:
where A is the area of the slab supercell surface, Eslab is the energy of the slab supercell, Ebulk is the bulk energy per atom, and N is the number of atoms in the slab supercell. Interestingly, we found that 2L mica is 36 and 22 meV less stable than 1L and 3L mica, respectively. The stability characterized by the surface energy can be assumed to be derived from two factors, structural and electronic:
where the Estructural component originates from the local atomic environment at the surfaces (i.e., from the presence of dangling bonds, chemical potentials, etc.), while Eelectronic comes from quantum electronic effects (i.e., quantum confinement, etc.). In our case, all single-, double-, and triple-layered mica structures have the same kind of K+ termination; hence, we can rule out the Estructural factor. Therefore, the preferable stability of odd-number-layered mica sheets should originate from the electronic effects. To support this hypothesis, we have developed a core-shielding model for the description of few-layered mica. According to this model, we treat the mica structure as a sequence of charged positive and negative layers. The positive layer corresponds to the K+ region, while the negative layer corresponds to a tetrahedral–octahedral–tetrahedral sandwich region (Fig. 1 in Ref. 15). The positive and negative regions interact with each other according to a general Coulomb-like equation:
In single-layered mica nanosheets, there is a stable electromagnetic attraction between the surface (Q1) and core (Q2) charges. However, in double-layered mica, there is a positive charge in the core region. This shielded charge in the core repulses the positive K+ ion layer, making double-layered mica unstable. The equation above can be re-written for the shielded charge as follows:
To verify this hypothesis, we calculated the Bader charges on all atoms in single-, double-, and triple-layered mica. A negative value of Bader charge stands for acceptance of electrons to yield charge, while a positive value represents donation of electrons to give charge. We found that the surface K atoms in all cases possess a similar charge of +0.89|e|, indicating that they are almost completely ionized. We found that the charge of the core region has a negative sign, giving the attractive electrostatic interaction between the K+ ions and the core. However, we found that the charge of the core region is 0.004|e| smaller in magnitude in the double-layered mica as compared to single-layered mica (Q2). This suggests that the K interlayer inside double-layered mica creates a shielding effect for the core charge. As a result, we expect a much weaker interaction in the double-layered mica and, hence, less stability. The charge of the core region in triple-layered mica is close to that of single-layered mica. On this basis, we may expand our conclusion further and suggest that mica with an even number of layers would be less stable than mica with an odd number of layers.
Population of even- or odd-numbered mica nanosheets
For the purpose of proving the theoretical calculation that predicts that odd-numbered mica layers are more stable than even-numbered ones, we have attempted to measure the population of each layer in the product of exfoliated mica nanosheets. In order to identify the number of layers in the exfoliated mica product, Raman analysis was used. As is shown in Fig. 2, a and b, the results of optical microscopy and Raman analysis, respectively, show that the peak intensity in the Raman spectra of Fig. 2b is evidently dependent on the number of layers. By converting the peak intensity into an image, one can visualize the population of the mica nanosheets in terms of the number of layers, which is often used in determining the number of layers in 2D materials.
On this basis, we can make a Raman imaging, indicating the number of layers for exfoliated mica products. One example of the Raman imaging is shown in Fig. 2c, revealing the existence of few-layered mica nanosheets. The typical numbers of mica nanosheets were observed to be 1, 2, 3, and 5. We observe that mica nanosheets with an even number of layers (i.e., 2 in the present case) have been produced, in addition to those with an odd number of layers (i.e., 1, 3, and 5). Although the results confirm the existence of 2L mica nanosheets, the amount of odd-layered mica nanosheets is significantly larger than that of even-layered ones. We have estimated the population of odd and even layers from 100 Raman images, which were taken from different exfoliated mica products. The results are summarized in Fig. 2d, evidently demonstrating that the odd-numbered mica nanosheets are predominant. This direct observation of the population of mica nanosheets in terms of the number of layers is well consistent with the core-shielding model, indicating that even-layered mica nanosheets are less stable in comparison to the odd-layered ones.
Surface charge states of mica nanosheets
According to the core-shielding model, due to the alternation of charges in the core region that may lead to the alternation in dipole fields with components parallel to the surface, there may be two-type surface charges for odd/even numbered layers. That is, the surfaces of all types of layers are all positive, but the odd-numbered layers have a strong dipole due to the more negative core regions, and the even-numbered layers have a relatively weak dipole due to the less negative core regions. To verify this, we have measured the surface charge state of the 2D mica nanosheets by using Kelvin probe force microscopy (KPFM), with which one can observe the work function of surfaces at a molecular scale. The map of the work function produced by KPFM gives us information about the composition and electronic state of the local structures on the surface of a solid because the work function of a solid is determined by various surface phenomena, chemical composition, reconstruction of surfaces, and doping and band-bending of semiconductors. As is shown in Fig. 3, odd-numbered layers such as 1L and 3L mica nanosheets are more positive than even-numbered layers such as 2L mica nanosheets. What is more important is that the surface charge states for 1L and 3L are almost identical. We proved that the 3L nanosheet has a larger bandgap than the 1L and 2L ones; if the surface charge does not play a role in Fig. 3, we should see 3 types (one for each layer type) corresponding to the 3 work functions. The fact that there are two types observed in Fig. 3 is in favor of the alternating charges with the odd/even layers.
Dye degradation
For the purpose of exploiting the exfoliated mica nanosheets, we have investigated their photocatalytic properties by testing dye degradation. The photocatalytic degradation activities of the mica and TiO2 (used as a reference) are shown in Fig. 4. The ultraviolet–visible (UV–vis) absorption spectra of methylene red (MR) solutions containing TiO2 or mica nanosheets are shown in Fig. S1 (see Supplementary Fig. S1 online). Prior to the sunlight irradiation, the suspensions of mica or TiO2 and the dye-containing MR solution were stirred in the dark to establish the adsorption/desorption equilibrium between the dye and the photocatalyst.
As shown in Fig. 4, mica nanosheets demonstrate a strong absorption of the anionic MR dye with the initial absorption value of 4.05. Particularly, the absorbance of MR on the mica nanosheets decreased from 4.05 to 2.8 after 250 min of sunlight irradiation. On the other hand, MR did not degrade at all on TiO2 and its concentration remained nearly unchanged after 250 min. Dye degradation needs two steps: 1) dye adsorption on the mica nanosheet and 2) dye destruction by photoelectrons. The mica nanosheets prepared by exfoliation consist of few-layered flakes, being composed of mainly 1 and 3 layers and partly of 2 layers. According to our “core-shielding” model, the mica nanosheet structure can be treated as a sequence of charged positive and negative layers. The positive layer corresponds to the K+ region, while the negative layer corresponds to a sandwiched SiO4–AlO6–SiO4 region. Accordingly, the effective MR degradation by the exfoliated mica nanosheets is reasonable. In contrast, we do not expect significant degradation of MR on the negatively-charged TiO2 surface. Our studies on mica nanosheets suggest it to be a potential candidate that can effectively degrade acidic dyes.