Using AFM-Raman to understand defects in two-dimensional semiconductor materials

Sanju Gupta

What if you could create a new type of semiconductor that could be more efficient in terms of absorbing the energy of the sun?

That’s the challenge facing Sanju Gupta, Ph.D., an Associate Research Professor at Penn State University, Materials Science & Engineering, and an educator in the Physics department plus a visiting scientist at The NanoScience and Technology Center, at the University of Central Florida.

To do it, she is introducing defects in the base materials that hopefully mimic what’s inherent in the manufacturing process, to see how the materials perform and what properties these defects possess.

Gupta works with graphene and MoS2, or Molybdenum disulfide, along with the stacking of these layers, forming heterostructures in a sandwich arrangement. The graphene is placed on top of gold or beneath a layer of MoS2 as a “nanospacer” to enhance certain properties. These layers are only each an atom thick and must be measured on the nanoscale. It is one of the semiconductor materials Gupta and her colleagues have studied at UCF. And that’s where AFM-Raman comes in.

“That's one of my areas of expertise for the last 10 years,” she said. “I was obsessed with them.”

Transition metal dichalcogenides (TMDCs) are becoming a class of low-dimensional materials that might be used for future nanoelectronics and optoelectronic devices. Two-dimensional (2D) MoS2 has properties that make it valuable for these applications. Potential exists for next-generation transistors with high on-off ratio, photodetectors, biosensors, and other properties, especially when combined with graphene.

Structural defects in these 2D crystals vary from zero-dimensional to one- and two-dimensional, Gupta writes in one of her studies. [1]

Defects are considered adverse, so attaining defect-free MoS2 is behind the push to understand various modes of degradation.[1,2] Two-dimensional materials are entirely surface since it is atomically thin, and its physical properties are sensitive to external treatments and processing.

Gupta’s goal is a “defect-assisted engineering paradigm” to harness these defects by means of external control (oxygen plasma and electron beam)[1,2]other than doping.

Defects also determine the optoelectronic properties in 2D TMDC semiconductors and govern device performance.

The goal

Therefore, the strategy of defect engineering is eliminating unfavorable defects, and introducing beneficial defects, which is useful for emergent physical, chemical and electrochemical functionalities. [1,3] For solar cells, the industry standard has been silicon. But Gupta et. al., are looking into new, different types or combinations of 2D materials that can possibly be more efficient, and better than silicon, for the sake of use in solar cells.

But the specific applications of her research might only be discovered five or 10 years from now when the properties of these materials are known.

The advantage to these new low-dimensional materials is you can make it thicker (more layers) or thinner (monolayer), use different types of materials and change the stack sequence. Research focuses on the properties these combinations possess. You can combine their properties. You can alter the properties and make it absorb the light across a broad spectrum from the Sun more efficiently, and therefore, you can convert solar energy into electrical energy with high efficiency. That’s the goal of some of these studies. For this purpose, she, and colleagues studies a stack of MoS2 surrounded by layers of graphene, which is atom thick layer of carbon arranged in a honeycomb lattice. Then they also deliberately try to induce defects in the layers of these MoS2 materials. And so, they look to see what is happening to the material because of these changes as a function of a number of layers. How is it impacting the performance of that semiconductor? That's what they're characterizing. And then maybe this will help to improve the application or develop the application in the future.

Once she induces the defect, she studies it using a sensitive AFM-Raman instrument to study what changes are happening due to those induced defects. Properties and the performance of the semiconductor may change favorably or unfavorably. Each defect may produce different properties changes, electrical or other.

The characteristic Raman peaks change for the materials based on the length of plasma or irradiation exposure, which tells them how this material reacts to the plasma or bombardment of electrons. The damage is greater with plasma and more irradiation. Dissociation is happening, as is oxidation due to environmental exposure, and then ring formation.

Two important papers by Gupta’s group differ in the type of irradiation. In one, electrons are used. In another, oxygen plasma is used. HORIBA technology helps her characterize the nature and types of the defect down to a nano-scale level.


Gupta and her associates addressed these questions by using Atomic Force Microscopy (AFM)-based techniques ― Kelvin probe force microscopy (KPFM), and tip-enhanced Raman spectroscopy (TERS). Gupta’s work was performed on a HORIBA AFM LabRam HR EVOlution Confocal Raman Microscope.

The LabRAM HR EVOlution Raman microscopes are ideally suited for both micro and macro measurements and offer advanced confocal imaging capabilities in 2D and 3D. The true confocal Raman microscope enables the most detailed images and analyses to be obtained with speed and reproducibility. This configuration has AFM capabilities.

The research congealed as Gupta had associates in central Florida familiar with HORIBA’s instrumentation, and Gupta herself is well versed in Raman spectroscopy with a variety of technologically relevant materials.

Gupta was interested in TERS at the time and thought it would be a perfect opportunity for her research.

“The HORIBA (LabRAM HR EVOlution) system was a wonderful system because that system with AFM capability is practically like a workhorse,” she said. “It can do enormous things; it can do practically anything because everything is local. Everything is right there. You can measure the Raman and optical spectra as well as the electronic properties like conductive KPFM. Alternatively, it can measure many properties and images on the same area multimodality. That’s the one word that I would use for the system. It's the multimodality of it. And then you are doing everything right there on the same spot as your material. So, there is no ambiguity, there's no question anybody will ask that.”


Gupta and her team manipulated the exfoliated 2D MoS2 layers by defect engineering induced by medium power oxygen plasma treatment for different time intervals ― she also used electron beam irradiation to create the defects. Then she investigates them for comparison using Kelvin probe force microscopy and tip-enhanced Raman spectroscopy (TERS), as well as micro-Raman and photoluminescence (PL) spectroscopy.

Confocal Raman spectroscopy identifies the vibrational properties as benchmarks, but at a micron level.

“We want to study at a nanoscale to gain an in-depth understanding because these defects are being created at the atomic level. So, we wanted to study a large area at a nanoscale, and that's exactly what we use the TERS (Tip-Enhanced Raman Scattering) for. And then we wanted to correlate the vibrational properties, and then optical properties using photoluminescence, for instance.”

Gupta, we also wanted to correlate with the electronic properties in the same area that occurred in concurrence with the Kelvin Probe Force Microscopy.

“Depending upon the TERS tip size, we can control, or sort of control our pixel size, and then we can go down to about 10 to 15 nanometers. And so, within that scale, which is not possible with the micro-Raman or micro-photoluminescence, but utilizing TERS, we could then correlate the results with the electronic properties.”

That’s one of the major achievements of these studies which Gupta and her associates carried out in Florida

Her latest research was carried out as a visiting scholar and adjunct faculty at the NanoScience and Technology Center (NSTC) and in the Department of Physics and Astronomy at the University of Central Florida-Orlando.

Gupta’s associates included Saiful Khondaker, Ph.D., and Ammon Johnston, a graduate student at the University of Central Florida.

“Dr. Gupta and Mr. Johnston contributed equally to this work in a synergistic manner, covering all aspects spanning fabrication, processing, characterization, and analyses. Dr. Khondaker also provided insights and the resources utilized in this work which was certainly invaluable,” she said.

Ammon Johnston

Saiful Khondaker

The current research work is driven by three questions:

  1. How oxygen plasma-induced (and electron beam) structural defects are distributed and what new chemical phases are introduced that modify the physical and chemical properties?
  2. What is the fate of 2D MoS2 nanolayers under plasma including lattice distortion and work function affecting the Fermi level?
  3. How does near-field Raman spectroscopy (TERS) implement in gaining insights into the spatial heterogeneities, defects, and phase identification down to the nanometer scale?
  4. How do structural defects correlate with optical and electronic and optical properties at the nanoscale?

Conclusions to these works were as follows:[1-3]

There are multiple implications to this comprehensive work as they pertain to identifying nature of and quantifying sulfur vacancies, establishing chemical phase-work function (microstructure-electronic property) diagram serving as a fingerprint that clearly differentiates the distribution of structural defects and doping at the nanoscale, both related to the local transformations of MoS2 to MoOx species, useful for energy sciences (heterogeneous electrocatalysis and hydrogen generation).[1]

“Our experimental findings showed (1) surface structuration, such as the formation of ring-like topology (“nano-sculpting”) with prolonged e-beam exposure; the associated mechanism is described as the Knotek–Feibelman mechanism led by sputtering and dissociation of ions following ionization (radiolysis), and (2) the generation of unsaturated bonds, single- or multi-vacancies (S and Mo) leading to agglomeration into extended defects due to their migration. Other events include the evolution of lattice disorder and etching depending upon the exposure time inducing the appearance of forbidden Raman bands and increasing the surface WF.”[2]

In another study, suppressing PL quenching via the introduction of graphene “nanospacer” between MoS2 and substrate for optoelectronic light-energy harvesting applications showed promise.[3]


Correlated KPFM and TERS imaging to elucidate defect-induced inhomogeneities in oxygen plasma treated 2D MoS2 nanosheets. Sanju Gupta, Ammon Johnston, and Saiful Khondaker. Journal of Applied Physics, April 26, 2022

2 Surface and Physical Properties Modifications of Electron Beam‑Irradiated Monolayer MoS2‑Au Heterointerface at Nanoscale. Sanju Gupta, Ammon Johnston, and Saiful Khondaker. Journal of Electronic Materials, November 10, 2022

3 Optoelectronic Properties of MoS2/Graphene Heterostructures Prepared by Dry Transfer for Light‑Induced Energy Applications. Sanju Gupta, Ammon Johnston, and Saiful Khondaker. Journal of Electronic Materials, April 22, 2022

Follow Us