Insights into morphological and electronic properties of two-dimensional materials at the nanometre-scale are of great interest to the industry due to the decreasing sizes of semiconductor devices. Complementary atomic force microscopy (AFM) techniques such as conductive AFM allow simultaneous insights into electronic and structural sample information. Conductive AFM (C-AFM) in high vacuum offers an additional improvement in the sensitivity of the electronic measurement signals due to the removal of the surface water layer.
Since the discovery of graphene in 2004, two-dimensional (2D) materials have seen a growing interest in the scientific community. Researchers have synthesised and characterised a variety of 2D materials, including the group of transition metal dichalcogenides (TMDs). TMDs attracted the attention of the semiconductor industry, because many of the compounds are direct band gap semiconductors with low dielectric constants and high charge-carrier mobilities. Due to their layer thickness of less than 1nm, TMDs qualify as promising candidates for semiconductor devices below 5nm following the trend of ever lower device sizes according to Moore’s law. However, the integration of TMDs in commercial semiconductor devices still faces some challenges. As such, the desired electronic properties often are limited to individual islands or grains and controlled layer growth, transfer and processing of high-quality TMD films remain complicated.(1,2)
Electrical atomic force microscopy (AFM
Electrical atomic force microscopy (AFM) is ideally suited to visualise local differences in the electronic properties of TMD films with a nanometre-resolution. In AFM, a nanometre-sized tip at the end of a cantilever scans sample surfaces. From the deflection of the cantilever at each point of the sample, the surface structure is then reconstructed and assembled into a topography image. Using a conductive cantilever/tip-system and establishing a mechanical contact between tip and sample facilitates C-AFM measurements with an applied bias. C-AFM allows resolving the spatial distribution of heterogeneous conductivities in addition to the sample topography. For this purpose, C-AFM records the current flow between the cantilever/tip-system and the sample at varying sample biases with the aid of a current-voltage preamplifier. Since C-AFM is a contact mode AFM method, the use of soft cantilevers with low spring constants is advantageous to minimise tip wear or sample damage.
This article compares C-AFM measurements on the TMD-material molybdenum disulfide (MoS2) in air and in a high vacuum environment of less than 1 ∙ 10-5 Torr achieved by the Park NX-Hivac AFM (Fig. 1). Generally, a thin layer of water forms on surfaces in ambient conditions, which can reduce the quality and sensitivity of the conductivity measurement in cAFM. Additionally, water induces a p-doping in MoS2, therefore limiting the electrical conductivity of the material.
Atomic force microscopy on MoS2
Fig. 2 summarises topography measurements on three MoS2 samples with different layer thicknesses: one sample with 1-2 MoS2 layers (a), one sample with 3-4 MoS2 layers (b), and another multi-layer sample with pyramidal surface structures (c). The 1-2-layer sample consists of one closed MoS2 monolayer with additional single-layered islands, featuring a step height of 0.6 nm – the layer thickness of a single monolayer. The islands represent the onset of the growth of the second layer and can be recognised in Fig. 2 as bright regions in the topography. Similarly, the 3-4-layer sample comprises a closed three-layered MoS2 film with additional single-layer islands. Figure 2d schematically shows the basic structure of the 3-4-layer sample as well as the C-AFM experimental setup for the following measurements. Here, each green layer represents one MoS2 monolayer, while the grey arrangement on the left side is the electrical contact required to apply of the sample bias during the C-AFM measurement. The multi-layered MoS2 sample differs strongly from the former in its surface structure. This sample exhibits three-dimensional (3D) pyramid-shaped features on top of a closed 3-layer MoS2 film. The formation of the pyramid-shaped features originates from a change of the growth mechanism from layer-by-layer to a 3D growth at thicknesses beyond 4 layers.
In addition to the MoS2 islands, the AFM imaged steps in the topography of the 1-2-layer and 3-4-layer samples that run diagonally across both images in Fig 2a and b. These lines originate from the terrace structure of the underlying sapphire substrate, which the AFM detected through the MoS2 films. However, the height of the steps allows a clear distinction between sapphire terraces and MoS2 steps: While the characteristic step height on the c-plane of the sapphire substrate is 0.2nm, a monolayer MoS2 has a thickness of 0.6nm. The height profile in Fig. 2e visualised the different step heights of the substrate and the MoS2 sample.
Conductivity in air and high vacuum
To demonstrate the advantageous effects of high vacuum on the quality of C-AFM, comparative measurements in air and in vacuum <1∙10-5 Torr were carried out on the same sample with the same instrument – the Park NX-Hivac AFM. While the topography images in air and high vacuum appeared largely similar (Fig. 3a and b), C-AFM data differed significantly (Fig. 3c and d). Here, the average current increased by three orders of magnitude in high vacuum compared to the measurement in air. In air, the average current was 1.4nA at 5V sample bias. In high vacuum on the other hand the average current increased to 1.1 μA with the same sample bias. The increase in average current originates from the removal of the thin layer of water that forms on the surface of materials in air due to the humidity level in ambient conditions. This layer of water is particularly problematic for MoS2 because water p-dopes the material and therefore becomes insulating. The C-AFM results agree with the results of previous studies, in which researchers showed that the on-state current and charge-carrier mobility in MoS2 semiconductor devices dropped significantly after contact with deionised water.
In addition to the overall increase in the average current, C-AFM improved in sensitivity for the measurements in a high vacuum environment. While the current image recorded in air appeared largely homogeneous without defined features, C-AFM in high vacuum was able to detect fine structures in the current channel. As such, C-AFM revealed that the current at grain boundaries and step edges of the MoS2 islands decreases. Fig. 4 shows the reduced current flow at step edges on a 1-2-layer sample that could not be resolved in air (Fig. 3c). The decrease in the detected current at the position of step edges and grain boundaries translates into a local reduction in conductivity. Thus, these topographic features act as electrostatic barriers on the charge-carrier transport and are detrimental for the implementation of MoS2 in commercial semiconductor devices. This correlative current-topography behaviour may have been missed if not for the increased sensitivity of C-AFM in high vacuum.
This study examined the morphological and electrical properties of the 2D material MoS2. The direct comparison of cAFM measurements in air and in a vacuum <1∙10-5 Torr demonstrated that the water layer on the sample surface in air considerably lowered the reliability of C-AFM. Removal of this water layer in high vacuum increased the current by three orders of magnitude. Furthermore, C-AFM in high vacuum allowed imaging of finer structures, thus enabling researchers to resolve the reduced current flow at the MoS2 step edges.
References
1. Chiappe, D. et al. Layer-controlled epitaxy of 2D semiconductors: bridging nanoscale phenomena to wafer-scale uniformity. Nanotechnology 29, 425602 (2018).
2. Ludwig, J. et al. Effects of buried grain boundaries in multilayer MoS2. Nanotechnology 30, 285705 (2019).
Jonathan Ludwig and Kristof Paredis are with IMEC. For more information visit Park Systems
