Hybrid 2D Black Phosphorus/Polymer Materials: New Platforms for Device Fabrication

Hybrid materials, containing a 2D filler embedded in a polymeric matrix, are an interesting platform for several applications, because of the variety of properties that the filler can impart to the polymer matrix when dispersed at the nanoscale. Moreover, novel properties could arise from the interaction between the two. Mostly the bulk properties of these materials have been studied so far, especially focusing on how the filler changes the polymeric matrix properties. Here we propose a complete change of perspective by using the hybrid nanocomposite material as a platform suitable to engineer the properties of the filler and to exploit its potential in the fabrication of devices. As a proof of concept of the versatility and potentiality of the new method, we applied this approach to prepare black phosphorus nanocomposites through its dispersion in poly (methyl methacrylate). Black phosphorus is a very interesting 2D material, whose application have so far been limited by its very high reactivity to oxygen and water. In this respect, we show that electronic-grade black phosphorus flakes, already embedded in a protecting matrix since their exfoliation from the bulk material, are endowed with significant increased stability, and can be further processed into devices without degrading their properties.

: bP nanosheets embedded in PMMA, on a Si/SiO 2 substrate. The sketch is not to scale, so proportions are not respected. The SiO 2 layer is 300 nm thick, while the PMMA/bP layer is about 50 nm thick, and the PMMA layer 300 nm.
103000 Da, respectively), indicating an effective role taken by the 2D black phosphorus during the polymerization process. In particular, the higher molecular weight suggests the growing of macromolecules in a "confined space" and thus near or between the bP layers. 27 We measured, for the hybrid material, a glass transition temperature T g of 121.9 • C, while for the reference sample T g is 120.6 • C. From the TGA curve we extracted T onset and T infl , determined as the temperature of intercept of tangents before and after the degradation step. For the nanocomposite we measured a T onset of degradation of 270 • C and T infl of 293 • C -372 • C (meaning the temperatures of maximum rate of degradation steps), while for the reference blank sample we measured 272 • C and 287 • C-381 • C, respectively. These measurements of thermal properties show a very similar behavior for the nanocomposite and the blank polymer sample. This, in turn, demonstrates a very good dispersion of the filler and suggests the formation of an interpenetrated phase, without any distinguishable effect on the bulk thermal properties.
Then, 51.4 mg of the hybrid were dissolved under dry nitrogen in 2 mL of anisole and closed in a vial, before being spin-coated in ambient condition in a cleanroom environement on SiO 2 /Si pre-patterned substrates, with Boron-doped Silicon and 300 nm of thermal SiO 2 , as shown in Fig. 1. In order to obtain a thin, homogeneous film, the spinner was set at 5000 revolutions per min, for 1 min. Then the samples were heated in air to 120 • C to remove the solvent. On test samples, the resist thickness was measured with a stylus profilometer and was found to be approximately 50 nm. After this first process, the samples were analyzed by both optical microscopy and Raman spectroscopy to identify suitable nanosheets for fabrication. Nanosheet identification is facilitated by a marker grid previously patterned on the SiO 2 /Si substrate by standard optical lithography. In Fig. 2

Device fabrication
The selected nanosheets were then processed by standard fabrication techniques. Thus, a 300 nm thick layer of bare commercial PMMA was spun on top of the thin layer of PMMA/bP nanocomposite, in order to have a sufficient PMMA thickness for a standard lithography and an easy lift-off process. Then electron beam lithography was performed. The possibility to achieve such a thin nanocomposite layer, if compared with the PMMA thickness commonly used for EBL, is a further important advantage of the proposed approach, since bP flakes are thus confined close to SiO 2 surface. After the lithographic process, the sample was developed and then underwent a very mild oxygen plasma treatment in vacuum, to remove resist residues and to allow for better ohmic contacts to the 2D material. 29

Electrical transport measurements
In Fig. 3 (a) the source-drain current-voltage (I sd − V sd ) curve of the device at zero gate voltage (V g ) at room temperature is shown. The linear trend confirms an Ohmic behavior.
The resistance obtained by fitting the I sd − V sd curve is 14.8 kΩ, compatible with exfoliated black phosphorus. 18 After a slow cool down in vacuum to 4.2 K, the resistance was measured again and was found to have increased by one order of magnitude, to 108 kΩ. By applying a back gate voltage, carrier concentration was modulated. The device shows a clear p-type semiconducting behavior, as expected for black phosphorus. 7 We evaluate the field-effect mobility from the trans-conductance, as 30,31 where dI sd /dV g is the slope of the curve shown in Fig. 3

Further characterization
As shown in Fig. 4 (a) and (b), SEM and AFM studies were carried out on a different device after the removal of the MMA(8.5)MAA copolymer/PMMA bilayer. 33 Inspection of the AFM line profile shown in Fig. 4 (c) clearly shows that the central part of the flake, not clamped by gold, was accidentally removed during the cleaning with acetone. The upper and lower parts of the structure show height values between 50 nm and 100 nm (Fig. 4 (d)). This, together with morphological information collected on similar materials discussed in Passaglia et al.,27 suggests that the nanosheet is indeed an aggregate of smaller bP flakes, held together by thin polymer layers. On the other hand, our electrical transport measurements show that this material is comparable to electronic-grade liquid phase exfoliated bP material 32 and suggest that bP in our hybrid material is of high quality. It can be expected that any improvement in preparation conditions will lead to even better transport properties, especially in terms of mobility.

Conclusions
In this work, we have proposed a new paradigm for hybrid 2D/polymer materials, focused on choosing the right polymer to exploit the properties of the 2D material at its best. We applied this approach to the preparation of a PMMA/bP nanocomposite. We showed that few-layer black phosphorus devices can be prepared from this multifunctional PMMA/bP hybrid material without the need for a glove box or any other kind of protective atmosphere, since the few-layer black phosphorus is exfoliated in MMA, which also efficiently acts as a protective coating. We showed also that few-layer bP survived the polymerization process of MMA without degrading. Furthermore, we demonstrated that the PMMA/bP nanocomposite is a suitable platform for device applications, since PMMA may be directly used as a resist for electron beam lithography. In fact, we used standard fabrication techniques to implement a simple device, and show a resistivity and carrier mobility characteristic of black phosphorus, as well as the expected p-type behaviour upon gate voltage modulation. The mobility can be further improved by optimizing sonication parameters. In summary, we have shown that the method here described constitutes a simple, low cost approach to handle sensitive 2D materials such as few-layer black phosphorus. Beyond that, the proposed approach can provide a series of innovative platforms by varying the 2D material and/or the embedding polymer, which can be tuned to provide other functionalities to the final composite.

Experimental techniques
Number average molecular weight (M n ) and weight average molecular weight (M w ) were determined using Size Exclusion Chromatography (SEC), Agilent Technologies 1200 Series.
The instrument is equipped with an Agilent degasser, an isocratic HPLC pump, an Agilent refractive index (RI) detector, and two PLgel 5 µm MiniMIX-D columns conditioned at 35 • C. Chloroform (CHCl 3 ) was used as the mobile phase at a flow rate of 0.3 mL min −1 . This system was calibrated with polystyrene standards in a range from 500 to 3 × 10 5 g mol −1 .
Samples were dissolved in CHCl 3 (2 mg mL −1 ) and filtered through a 0.20 micron syringe filter before analysis (twice in the case of hybrids). Number average molecular weight (M n ) and weight average molecular weight (M w ) were calculated using the Agilent ChemStation software.
Thermal Gravimetric Analyses (TGA) were carried out with a Seiko EXSTAR 7200 TGA/DTA by introducing about 5-8 mg of sample in an alumina sample pan of 70 µL. In a typical experiment, run was carried out at a standard rate of 10 • C/min from 30 • C to 700 • C under nitrogen flow. T onset and T infl , which are the were determined by analyzing the TGA curve (as the temperature of intercept of tangents before and after the degradation step) and DTG curve (as the maximum of the peak), respectively.
The glass transition temperature (T g ) was determined by Differential Scanning Calorimetry (DSC) using a PerkinElmer DSC4000 equipped with intracooler and interfaced with Pyris software (version 9.0.2). The range of temperatures investigated was 40-180 • C. Thermal scans were carried out on 5-10 mg samples in aluminum pans under nitrogen atmosphere.
Raman spectroscopy was performed using a Renishaw inVia system equipped with a 532 nm laser and a motorized stage for 2D mapping of samples. A laser spot size of approximately 1 µm in diameter was used. Laser power was 55 µW. We have verified that this power did not damage the bP flakes during measurements. We note that PMMA is a suitable polymer for this kind of analysis, since it is transparent in the spectral region of interest. For transport measurements, the samples were bonded using Au wire to a 16 pin dual in line chip carrier. The transport properties were measured in DC in vacuum (p < 10 −4 mbar) in a custom made insert, equipped with a diode for temperature measurement, and compatible with the used chip carriers. The leakage current though the gate was measured during V g loops and found to be negligible.
SEM imaging was performed with a Zeiss Merlin microscope with 5 kV acceleration voltage. Atomic force microscopy (AFM) measurements were performed with a Bruker Dimension Icon AFM, in pick force mode. Data analysis was performed with WSxM software. 34

Materials
In our experiments, we used bP crystals prepared according to the procedure developed by Nilges et al., 35 wherein high-purity red phosphorus (> 99.99%), tin (> 99.999%), and gold (> 99.99%) are heated in a muffle oven with a SnI 4 catalyst. The solid product was placed in a quartz tube, subjected to several evacuation-purge cycles with N 2 gas, and then sealed under vacuum. The evacuated quartz tube was heated to 406 • C at 4.2 • C/min, where it remained for 2 hours. The tube was then heated to 650 • C at 2.2 • C/min and held at this temperature for 3 days. The tube was then cooled to room temperature at 0.1 • C/min. The final product is crystalline bP with a typical size of some mm. All the other materials (polymers, reagents, solvents) are commercial and used as they are without further purification.

Acknowledgement
This work was financially supported by EC through the project PHOSFUN Phosphorene Coiai is acknowledged for helpful discussion.

Graphical TOC Entry
Novel insight in hybrid materials is given. A judicious choice of the matrix polymer allows the exploitation of the full potential of the 2D filler: poly (methyl methacrylate) nanocomposites open a novel way towards a scalable, low cost production route for few-layer black phosphorus devices.