Efficient liquid exfoliation of KP₁₅ nanowires aided by Hansen's empirical theory

Introduction

Low-dimensional materials have drawn significant attention in recent years. So far, not only new composite materials with excellent properties have been obtained by the synthesis of different materials, but also low-dimensional materials with different properties than those of bulk materials have been synthesized by physical and chemical methods. For instance, Bingjun Yang used one-dimensional graphene nanoscroll-wrapped MnO nanoparticles as anode materials to promote the rapid diffusion and electron transfer of lithium, and Rongjun Zhao prepared n-butanol gas sensors with one-dimensional In₂O₃ nanorods [1,2]. Different from 2D materials, 1D materials generally have a chain-like crystal structure and are easily exfoliated due to a weak interaction between these chains [3,4]. Therefore, those 1D materials have defect-free surfaces, high anisotropy, and carrier mobility. For example, TiS₃ nanowires obtained by mechanical stripping have a large carrier mobility of about 10000 cm²·V⁻¹·s⁻¹ [5-7]. Fibrous phosphorus is also a new one-dimensional material with high carrier mobility (308 cm²·V⁻¹·s⁻¹) and rapid response time [8-10]. These one-dimensional materials are ideal for photovoltaic and photocatalytic applications.

The KP₁₅ is considered to be a novel low-dimensional material with layered structure, high hole carrier mobility (1000 cm²·V⁻¹·s⁻¹), and highly anisotropic properties [11]. The photodetectors prepared with KP₁₅ have a fast response time and are ideal materials for photovoltaic applications [12]. Based on our previous studies, KP₁₅ is also a one-dimensional material with a defect-free surface [13,14]. This is beneficial for the development of high-performance nanodevices. Searching effective synthesis routes for nanoscale KP₁₅ has become an urgent issue. Liquid-phase exfoliation is one of the most straightforward methods to prepare low-dimensional materials at a low cost and with simple processes and high flexibility. In this case, the surface of bulk materials is peeled off or corroded by physical or chemical reactions in a liquid medium, and finally low-dimensional materials are obtained. Based on the Hansen's empirical theory, the exfoliation efficiency of low-dimensional materials can be improved by adjusting the composition and type of solutions used in the liquid-phase exfoliation [15-17]. This theory has been successfully used for improving the exfoliation efficiency in several low-dimensional materials, such as carbon, graphene, metal oxides, and fibrous phosphorus. [18].

In a previous study, we exfoliated KP₁₅ in alcohol; however, this method was still inefficient [13]. Herein, the Hansen's empirical theory was firstly introduced to improve the liquid-phase exfoliation efficiency of KP₁₅ nanowires. In addition, Hansen solubility parameters (HSPs) for KP₁₅ were also obtained in this work. By using a solution containing 50% water and 50% acetone, the exfoliation efficiency of KP₁₅ was effectively improved. Our results show that 79% of the KP₁₅ nanowires had thickness values below 50 nm and 60.9% of these nanowires had width values below 100 nm. The thinnest KP₁₅ nanowires reached 5.1 nm and had smooth boundaries. Meanwhile, a strong temperature-dependent Raman response was found in exfoliated KP₁₅ nanowires. This indicates a strong phonon–phonon coupling in KP₁₅ nanowires, which favors non-invasive temperature measurements of KP₁₅ nanodevices.

Methods

Synthesis of KP₁₅ bulks

The KP₁₅ bulks were prepared by the gas-phase transfer method. High-purity red phosphorus (1.370 g, 99.9999%) and metallic potassium (0.130 g, 97%) were mixed in a quartz tube. The temperature gradient in the quartz tube was 650 °C/400 °C and the heat treatment time was 12 h. After annealed, dark-red KP₁₅ bulks were finally obtained.

Liquid exfoliation

For the liquid-exfoliation process, 1 mg of KP₁₅ was mixed in 20 mL of solvent and ultrasonically processed at a power of 80 W in an ice bath for 6 h, followed by centrifugation at 2000 rpm for 20 min. For the samples with predetermined concentration, centrifugation was not used.

Measurement equipment

UV−visible spectrophotometry was performed by using a Shimadzu UV-3101PC system. Atomic force microscopy (AFM) tests were performed in a Multimode 8 system. The Raman tests were performed on a WITec alpha300 RA confocal Raman microscopy system. For the Raman tests, KP₁₅ samples were spun on SiO₂(300 nm)/Si substrates. The excitation wavelength used was 532 nm, the spot size was approx. 1 μm, and the laser power was kept below 20 μW. For low-temperature Raman measurements, a Linkam THMS600 cryostat cooled by liquid nitrogen was used to control the temperature. To prevent sample drift, SiO₂ (300 nm)/Si substrates with tested KP₁₅ samples were attached by fixtures to the Linkam THMS600 cryostat.

Results and Discussion

KP₁₅ bulks, prepared by the gas-phase-transfer method, had a flat and smooth surface shown in Figure 1a. The X-ray diffraction patterns of the synthesized KP₁₅ were both theoretically calculated and experimentally measured. The consistency between the two patterns shows that there is no impurity phase (Figure 1b), which confirms an excellent crystallization quality of the KP₁₅ bulks.

Measurement of the absorption coefficient and the Hansen solubility parameters for KP₁₅

According to the Hansen’s theory [19], the dispersed concentration C of a KP₁₅ dispersion prepared by liquid exfoliation can be expressed by Equation 1 as follows.

where δ_D is the intermolecular dispersion force, δ_H is the intermolecular hydrogen bond; δ_P is the intermolecular polar force; δ_A,D, δ_A,P, δ_A,H are the Hansen solubility parameters (HSPs) of the solute; and δ_B,D, δ_B,P, δ_B,H are the HSPs of the solvent. Therefore, to get a high concentration of KP₁₅ in dispersion, the HSPs of the solvent for the exfoliation of KP₁₅ should be close to those of KP₁₅. A weighted average method was used to calculate the HSPs of KP₁₅. The concentration of KP₁₅ was used as a weight factor for each suspension. This way, the HSPs of KP₁₅ can be expressed according to Equation 2 [19].

where δ_i,sol are the HSPs of the solvent and C is the concentration of the KP₁₅ dispersions. The Lambert–Beer law (Equation 3) was then used to measure the concentration of the KP₁₅ dispersions:

where A is the absorbance, K is the absorption coefficient of the material, b is the absorbing layer thickness (which in this work is the width of the cuvette, i.e., 1 cm), and C is the concentration of the KP₁₅ dispersions. The absorbance A and the absorption coefficient K are related to the wavelength of the incident light. To determine A and K, it is necessary to choose a specific incident wavelength. The bandgap of bulk KP₁₅ is approx. 1.75 eV [20]. However, according to our previous study, with thickness reduction of the KP₁₅ nanowires, a surface-state luminescence at 693 nm gradually dominates in the KP₁₅ nanowire [14]. This could affect light absorption properties of KP₁₅ due to its decreased size.

To avoid the generation of concentration error caused by the absorbance influence of the surface state, a wavelength (800 nm) which is far away from the bandgap of KP₁₅ bulk and surface state in the KP₁₅ nanowires was chosen. Some dispersions for which we predetermined the concentration were prepared to fit and determine the absorption coefficient K. Solutions of five different concentrations of KP₁₅ dispersions in butyrolactone were prepared by liquid exfoliation with a predetermined concentration. UV−visible absorption spectra results are shown in Figure 2. The concentration linearly varies with absorbance. The slope of this fitted linear equation is 3.86 ± 0.13. This means that the absorption coefficient of KP₁₅ is 3.86 ± 0.13 mL·mg⁻¹·cm⁻¹.

We selected 20 common solvents, including benzyl benzoate, toluene, ethyl acetate, acetone, alcohol, butyrolactone, N,N'-dimethylpropyleneurea, bromobenzene, cyclopentanone, N-dodecyl-2-pyrrolidone, glycol, vinyl acetate, hexane, isopropyl alcohol, N,N-dimethylformamide, O-phthalic dimethyl ester, dimethyl sulfoxide, N-methylpyrrolidone, water, and cyclohexanone. The HSPs of those solvents are listed in Table 1.

Figure 3 exhibits the concentrations of KP₁₅ dispersions exfoliated in different solvents. Cyclopentanone and butyrolactone were more suitable than the other solvents to exfoliate KP₁₅. Figure 4 shows the relationship between the HSPs of different solvents and the concentration of the KP₁₅ suspension. Based on Equation 2, the HSPs of KP₁₅ were δ_D = 17.60 MPa^1/2, δ_P = 11.19 MPa^1/2, and δ_H = 8.95 MPa^1/2. As long as the difference between the HSPs of KP₁₅ and the HSPs of a given solvent is reduced, τ can be reduced with an improved exfoliation efficiency. Figure 5 shows the concentration of KP₁₅ dispersions as a function of τ. When τ tends to zero, the concentration of the KP₁₅ dispersion reaches the maximum value, which corresponds to the results of the aforementioned equation.

Liquid exfoliation of one-dimensional KP₁₅

The HSPs obtained for KP₁₅ were δ_D = 17.60 MPa^1/2, δ_P = 11.19 MPa^1/2, and δ_H = 8.95 MPa^1/2. We chose a mixed solution containing water and acetone to exfoliate KP₁₅. The HSPs of water were δ_D = 15.8 MPa^1/2, δ_P = 8.8 MPa^1/2, and δ_H = 19.4 MPa^1/2. The HSPs of acetone were δ_D = 15.5 MPa^1/2, δ_P = 10.4 MPa^1/2, and δ_H = 7.0 MPa^1/2. The HSP range of a mixed solution of water and acetone can cover the HSPs of KP₁₅, however, both of them can be easily removed. The HSPs (δ_i) in a mixed solution containing water and acetone can be expressed by Equation 4.

where ϕ_i,comp is the volume fraction of the corresponding solvent and δ_i,comp is the HSPs of the solvent. The concentration of the KP₁₅ dispersion can be measured by the Lambert–Beer law (Equation 3). As shown in Figure 6a, by tuning the volume fraction of acetone in the mixed solution, the HSPs of the mixed solution can be close to those of KP₁₅, and the exfoliation efficiency can be clearly improved. The concentration values of the KP₁₅ suspension in the solutions were 0.0268 mg·mL⁻¹ (exfoliated in deionized water), 0.0079 mg·mL⁻¹ (acetone), and 0.0236 mg·mL⁻¹ (alcohol), respectively [13]. When the solvent mixture with a 50% volume fraction of acetone is used for stripping, the concentration of the KP₁₅ dispersion finally increases to 0.0458 mg·mL⁻¹. At this point, the parameter τ is close to the minimum value.

The Raman result for the KP₁₅ nanowires exfoliated in water−acetone mixed solution is shown in Figure 7c. At least 11 distinguishable Raman peaks located at 476.6, 453.0, 408.8, 378.3, 368.4, 354.1, 303.7, 288.5, 126.1, 114.1, and 90.7 cm⁻¹ were seen and those Raman results were similar to the Raman modes of mechanically exfoliated KP₁₅ [11]. As shown in Figure 7d, Figure 7e, and Figure 8, the thinnest KP₁₅ nanowires obtained by liquid exfoliation could reach 5.1 nm and had smooth boundaries. The thicknesses of 79% of the liquid-exfoliated KP₁₅ nanowires were below 50 nm; the widths of 60.9% of the KP₁₅ nanowires were below 100 nm. The sizes of the obtained KP₁₅ nanowires were much smaller than those obtained in our previous studies [13]. Meanwhile, a strong temperature-dependent Raman response in exfoliated KP₁₅ nanowires has been observed. That may help with non-invasive temperature measurements of KP₁₅ nanodevices (details are demonstrated in Supporting Information File 1).

Conclusion

In summary, based on the Hansen's empirical theory, the liquid phase exfoliation efficiency of KP₁₅ nanowires has been improved. The HSPs of KP₁₅ were calculated to be δ_D = 17.60 MPa^1/2, δ_P = 11.19 MPa^1/2, and δ_H = 8.95 MPa^1/2. In addition, based on the Hansen's empirical theory, the exfoliation efficiency was improved by adjusting the ratio of water and acetone. When the mixed solvents had the smallest τ, the thicknesses of 79% of liquid-exfoliated KP₁₅ nanowires were below 50 nm and the widths of 60.9% of KP₁₅ nanowires were below 100 nm. Meanwhile, a strong temperature-dependent Raman response has been found in exfoliated KP₁₅, which may help with non-invasive temperature measurements of KP₁₅ nanodevices.