Characterizing 2D Materials with AFM-IR Spectroscopy

Due to their unique properties for applications in battery technology, semiconductors, photovoltaics, and a number of other areas, 2D materials research is a very important emerging field.

2D materials have been characterized by a number of microscopy and nanoscale methods to obtain a better understanding of the nature of their properties. Nanoscale FTIR methods extend this characterization with vital optical and chemical data at the nanoscale.

The Anasys nanoIR3-s system supplies two complementary nanoscale FTIR methods for scattering-scanning near-field optical microscopy (s-SNOM), spectroscopy and near-field imaging, and photothermal AFM-Infrared (AFM-IR), including resonance-enhanced AFM-IR and Tapping AFM-IR.

Independent of other complex optical properties of the tip and sample, AFM-IR absorption spectra are direct measurements of sample absorption. So, the spectra correlate extremely well to conventional bulk transmission IR.

Scattering SNOM is a method that examines the chemical and optical properties of the material surface, providing correlation to reflection-based spectroscopy techniques. Atomic force microscopy (AFM) based methods generate nanoscale information on the thermal, mechanical, and electrical properties of these materials.

Traditionally, s-SNOM imaging techniques have been used to characterize 2D materials for chemical and optical information. However, with the development of applications such as nanopatterning and functionalization, AFM-IR is better at supplying valuable insights and unique information about materials, and thus accelerating innovative research and capabilities.

When combined, these complementary methods supply new insights into the nanoscale chemical and complex optical properties of 2D materials with resolutions of 10 nm, orders of magnitude below the diffraction limit of conventional IR spectroscopy.

This article discusses employing these complementary methods for the characterization of a variety of 2D materials, including hexagonal boron nitride, graphene, nanoantennae and semiconductor materials.

Extending s-SNOM into Nanoscale FTIR Spectroscopy

The principles of operation for s-SNOM and photothermal AFM-IR methods can be seen in Figures 1 and 2. A recent advancement in AFM-IR is tapping AFM-IR, which delivers higher resolution chemical imaging and extends AFM-IR spectroscopy to a wider range of applications.

Figure 1. Principles of operation for IR s-SNOM.

Figure 2. Principles of operation for resonance-enhanced AFM-IR and tapping AFM-IR.

The extension of this method to nanoscale FTIR spectroscopy across the broadest available mid-IR range has been permitted via Recent Bruker developments in s-SNOM technology. The modes of operation available with the nanoIR3-s Broadband laser can be seen in Figure 3.

Figure 3. nanoIR broadband modes.

The system supplies two modes in a single source:

• The imaging mode provides the largest available imaging range (670 cm^-1 to >2000 cm^-1), eliminating the need to buy QCL lasers for imaging
• The spectroscopy mode provides the largest spectral range for s-SNOM in a single laser source (670 cm^-1 to >4000 cm^-1)

Complementary Nanoscale IR Techniques

The nanoIR3-s can gather IR spectra and nanoscale images by utilizing two separate near-field spectroscopy methods: photothermal AFM-IR and s-SNOM.

These complementary methods provide nanoscale chemical analysis, in addition to thermal, optical, electrical, and mechanical mapping with spatial resolution down to a few nanometers for both hard and soft matter applications.

The precise chemical identification of infrared spectroscopy and the nanoscale capabilities of AFM are combined in nanoscale IR spectroscopy in order to chemically identify sample components with a chemical spatial resolution down to 10 nm with monolayer sensitivity, breaking the diffraction limit by >100x.

Independent of other complex optical properties of the tip and sample, AFM-IR absorption spectra are direct measurements of sample absorption. So the spectra correlate extremely well to that of conventional bulk transmission IR.

Imaging of Phonons and Plasmons

Due to their high spatial confinement, surface phonon polaritons (SPhPs) and surface plasmon polaritons (SPPs) in 2D materials can open up new pathways for enhanced light-matter interaction, subwavelength metamaterials, super lenses, and other novel photonic devices.

In-situ characterization of these polaritonic excitations across different applications calls for a versatile spectroscopy and optical imaging tool with nanometer spatial resolution. s-SNOM provides a unique way to selectively excite and locally detect electronic and vibrational resonances in real space via a non-invasive near-field light-matter interaction.

Figure 4. (a) AFM height image shows homogeneous hBN surface with different layers on a Si substrate; (b) s-SNOM amplitude shows strong interference fringes due to propagating SPhP along the surface on hBN; (c) s-SNOM phase shows a different phase signal with layer thickness; and (d) nano FTIR spectra of hBN.

This method is demonstrated by imaging the SPhPs of hexagonal boron nitride (hBN) as illustrated in Figure 4. Amplitude and phase near-field optical images supply complementary information for thorough characterization of the polaritonic resonances. Over 90 ° phase shift of SPhPs are seen on hBN, indicating strong light-matter coupling.

The SPPs of graphene can also be examined via the nanoIR3-s system, similar to the visualization of SPhPs in hBN. The standing wave of an SPP on a graphene wedge can be seen in Figure 5.

The spatial resolution of s-SNOM is generally limited only by the end radius of the AFM probe, allowing the s-SNOM method to measure cross sections of the SPP down to around 8 nm.

Figure 5. Imaging of surface plasmon polariton on a graphene wedge: (a) s-SNOM amplitude; (b) s-SNOM phase with a line cross section of the SPP standing wave; (c) s-SNOM phase line profile; and (d) cross section of a standing wave showing 10 nm resolution.

Functionalized 2D Materials Characterization with AFM-IR

Chemical functionalization of 2D materials is a key step in realizing their full potential in a large scope of applications. Nanoscale FTIR spectroscopy meets the increasing requirements for non-destructive, unambiguous, identification of chemical groups.

This is in addition to its ability to map the distribution on such materials using nanoscale spatial resolution and at monolayer thicknesses. Researchers at the University of Manchester's National Graphene Center have established that photothermal AFM-IR can analyze single-layer reduced graphene oxide flakes that have been non-covalently functionalized with sulfonated pyrenes.¹

Their research established that AFM-IR distinguished between the different pyrene moieties and mapping the sulfonate groups on a 1.7 nm functionalized monolayer of reduced graphene oxide with 32 nm spatial resolution, as shown in Figure 6.

AFM-IR was also found to be sensitive to minute alterations in the sulfonate absorption spectra arising from surface and chemical effects. It was able to differentiate between individual functionalizing molecules, even on materials with anisotropic thermal conductivity.

Figure 6. (a) Single-flake AFM-IR spectra of i) PBSrGO, ii) PCNBS-rGO and, iii) U-rGO; and (b) average AFM-IR spectra of i) PBS-rGO, ii) PCNBS-rGO, and iii) U-rGO, dashed lines highlighting the 1084 and 1036 cm^-1 intensities used for further chemical mapping studies.

Nanoscale Patterning of 2D Materials and Photothermal Effects

The controlled nanoscale patterning of 2D materials is a promising approach for engineering the thermal, optoelectronic, and mechanical characteristics of these materials in order to achieve novel functionalities and devices. Researchers at Harvard University performed high-resolution patterning of hBN.²

Using this nanofabrication approach, a number of structures were made, including a 35 nm pitch grating and individual structure sizes down to 20 nm. Later measurements by photothermal and scanning near-field optical microscopy was able to calculate the resulting near-field absorption and scattering of the nanostructures.

A large photothermal expansion of nanostructured hBN, which is dependent on the height-to-width aspect ratio of the nanostructures, was revealed in these measurements. This effect is due to the large anisotropy of the thermal expansion coefficients of hBN and the nanostructuring implemented.

In other van der Waals materials with large anisotropy, the photothermal expansion should also be present, and further research can lead to applications such as nanomechanical switches driven by light.

Figure 7. Photothermal infrared spectroscopy and imaging: (a) photothermal spectra taken from four different locations along the thinner region of the hBN grating, showing a strong SiO₂ absorption peak at 1084 cm^-1; (b) photothermal spectra taken from four different locations along the thicker region of the hBN grating, showing a strong and broad hBN absorption peak at 1368 cm^-1; (c) AFM image of the top right corner of the hBN grating at 1368 cm^-1; and (d) photothermal image of the same region at 1600 cm^-1. No photothermal contrast was detected, indicating no absorption or mechanical expansion.

Nanocontamination of Graphene

The exceptional electrical and mechanical characteristics of graphene depend on maintaining the overall conjugated structure of the sheet. As shown in Figure 8, the nanoIR3-s system can assess the quality of exfoliated graphene obtained by various techniques easily.

Contamination, which is not easily recognizable in the AFM height image, is visible in the s-SNOM reflection image. In addition, contrast in the s-SNOM reflection image differs with the amount of graphene layers present, revealing nanocontamination on the sample.

Figure 8. (a) AFM height image of exfoliated graphene; and (b) s-SNOM reflection image, showing nanocontamination.

Characterizing Nanoantenna Resonance

Ranging from sensing to energy conversion, the applications of nanoantennae are extremely diverse. The capability to calculate and tune the resonance structures of these antennas is crucial to the construction of reliable and accurate devices.

As they permit packing of a large number of individual antennas in a compact area, arrays of nanoantennas are common. An AFM topography image of an antenna array made up of single bar antennas in addition to coupled antennas is shown in Figure 9a.

The contact point to the antennas is a crucial consideration to attain optimum energy transfer efficiency when fabricating antenna arrays. s-SNOM imaging enables the simple detection of the antenna resonance hot spots, plus the ideal contact point.

The s-SNOM amplitude and phase image of a single bar antenna contained within the array is shown in Figure 9b. The dipole antenna resonance is seen with 11 µm excitation (note the ~180 ° phase change seen at dipole resonance).

Figure 9. (a) AFM height image of assembled antenna array; (b) s-SNOM phase; and (c) s-SNOM amplitude images of an antenna dipole.

Further to the capability to gather high-resolution images of optical phenomenon, the nanoIR3-s permits the user to spectrally probe nanoscale surface features. Figure 10 illustrates the AFM-IR spectra gathered on single rod and coupled antenna, and the antenna resonance can be determined clearly at 910 cm^-1, which is in agreement with theoretical predictions.

Figure 10. AFM-IR spectrum collected on single rod and coupled antenna; the peak at 910 cm^-1 corresponds to the antenna resonance of the single rod antenna, while the peak at 1100 cm^-1 shows the Si-O mode shared by both antennas.

Characterizing the Effects of Polarized Light on Metasurface Chirality with s-SNOM and AFM-IR

The combination of the complementary nanoscale s-SNOM and AFM-IR imaging methods has recently been employed to examine the role of chirality in the origins of circular dichroism in 2D nanoscale materials for the first time.

Chiral molecules are a variety of molecules that have a non-superimposable mirror image. These mirror images of chiral molecules are sometimes known as right handed and left handed, but because of the vector nature of light they can be both right and left circularly polarized.

Also known as metasurfaces, fully two-dimensional (2D) metamaterials made up of planar-chiral plasmonic metamolecules that are only nanometers thick have exhibited chiral dichroism in transmission (CDT).

Theoretical calculations show that this surprising effect is reliant on finite nonradiative (ohmic) losses of the metasurface. This theoretical prediction has only now been experimentally verified, due to the challenge of measuring non-radiative loss on the nanoscale.

In order to map the optical energy distribution when the structures were exposed to RCP and LCP IR radiation, scattering SNOM was employed, whilst AFM-IR was employed to detect the drastically different ohmic heating seen under RCP and LCP radiation.³

This study determined conclusively for the first time that the circular dichroism observed in 2D metasurfaces can be credited to handedness-dependent ohmic heating, as shown in Figure 11.

Figure 11. Experimentally measured AFM cantilever deflection amplitudes. The cantilever deflection is directly proportional to temperature increase in the sample during the laser pulse; this confirms that the magnitude and spatial distribution of the Ohmic heating of a chiral 2D metasurface markedly depends on the handedness of light.³

Analysis of Carbon Nanotubes using Nanoscale IR

The AFM-IR method works by detecting the thermal expansion of a material, induced by the absorption of infrared illumination. The thermal expansion of a material depends on a number of factors, including the thickness and the coefficient of thermal expansion of the material.

Single-layer graphene and single-walled carbon nanotubes (CNTs), as well as other 1D and 2D materials, are roughly 1-2 nm thick and also posses a slow coefficient of thermal expansion.

The nature of these 1D and 2D samples can make characterization a challenge. By positioning a thin layer of polymeric material below CNT and graphene samples, a two orders of magnitude growth in AFM-IR signal intensity is seen.^{4, 5}

The heat produced is transferred to the thin polymer, which has a notably higher coefficient of thermal expansion.  As the thin sample absorbs the incident IR radiation, it expands. The finite element analysis model employed to simulate the effects of polymer thickness on the thermal expansion and temperature changes is shown in Figure 12.

Figure 12. (a) Temperature rise (ΔT) and expansion (ΔZ) as a function of polymer thickness beneath the sample; temperature rise with no polymer (b) and with polymer (c) beneath the sample; vertical thermomechanical expansion with no polymer (d) and with polymer (e) beneath the sample.

By studying a range of CNTs deposited on top of a layer of 150 nm thick polystyrene on a ZnSe prism, the model was verified. A region of the polymer substrate was taken away before CNT deposition to make sure there was a region of CNT with no polymer underneath.

The IR chemical image gathered at 4000 cm^-1 exhibits clear signal from the CNT in the region that is supported by polystyrene, whilst no signal is observed where the polymer substrate has been removed (see Figure 13).

It has been proposed that the differing AFM-IR signal from different CNTs is due to the difference between semiconducting and metallic tubes. The AFM-IR imaging of graphene on top of a 106 nm thick layer of PMMA is illustrated in Figure 13c.

The image in Figure 13c shows the extension of this method to monolayer 2D materials. The amplification of the AFM-IR signal by a thin layer of polymer heightens the signal intensity by two orders of magnitude.

Figure 13. (a) AFM topography imaging of CNTs deposited on a polystyrene substrate; (b) IR chemical mapping at 4000 cm^-1 showing absorption by CNTs; and (c) IR chemical mapping image of monolayer graphene captured at 4000 cm^-1.

This new method permits AFM-IR characterization of 1 nm thick 1D and 2D materials that was previously impossible. In the future, this dramatic signal enhancement may be applied to multiple applications, including a variety of 1D and 2D materials and ultrathin biologicals.

Investigating Exothermic Peaks of Polyethylene Using nanoTA and LCR

Polyethylene (PE) is one of the most widely used polymers, and it is employed in a number of industries, including applications in 2D materials. To alter the thermal, mechanical, and electrical properties of PE, metallic and graphite particles and other metallic fillers are commnly added.

hBN has shown promise as a filler in recent years, because of its thermal conductivity, high mechanical strength, and insulating characteristics. Researchers at Sichuan University employed Lorentz contact resonance (LCR) and nanothermal analysis (nanoTA) to characterize this influence of hBN particles on the melting behavior of PE.⁵

As shown in Figure 14a and b, LCR imaging can clearly reveal regions of high hBN concentration on the surface. Then nanoTA was employed to calculate the softening temperature of multiple regions of the material.

For areas of the PE sample near hBN aggregates, an increase was seen in the transition temperature of 4-8°C when compared to areas without hBN, as illustrated in Figure 14.

The accuracy of this method was demonstrated when compared to traditional DSC analysis, with the bulk transition temperature within the standard deviation of nanoTA values, as seen in Figure 14d.

Combined with DSC analysis, these results demonstrate that the meso-phase of the PE forms near hBN particles during crystallization, which prompts a weak exothermic peak that was unexplained prior to this study.

Figure 14 also shows nanoTA measurement carried out on the hBN particles directly, for which no thermal transition was identified at temperatures up to 400°C.

Figure 14. (a) LCR-AFM height image; (b) AFM mechanical image (using LCR) of the PE/BN composites, showing boron nitride clusters in the areas A,D and E; (c) local thermal analysis data of the assigned positions were obtained by nanoTA, comparing the melting temperatures of PE and BN; and (d) DSC from the PE/BN composites (heating rate of 2°C min^-1).

Analyzing Thermal Conductivity of Graphene Sheets with SThM

Due to its high thermal conductivity and potential in optoelectronics, graphene has been a focus of recent studies. Scanning thermal microscopy (SThM) characterizes thermal conductivity of 2D materials, as it exhibits high sensitivity in resistance detection between the sample and the probe.

These high spatial resolutions take away ambiguity in the detection of the source of a sample’s electrical capabilities, making SThM a reliable technique for monitoring a sample temperature, in addition to its thermal conductivity.

Researchers at Durham University and Lancaster University employed SThM to study thermal conductivity on multilayer and single graphene sheets.⁶ Graphene was deposited on an Si/SiO₂ substrate with prepatterned trenches, with images taken from both graphene suspended over the trench and supported by the substrate.

It was discovered that adding to the amount of supported graphene layers resulted in a clear decline in thermal resistance. A key finding was that the thermal conductance of both multilayer and bilayer graphene suspended over the trench was more than that of the supported layer.

This discovery is contrary to the expectations that conduction from the graphene to the substrate would generate more heat dissipation. As the mean free path of thermal phonons in graphene is a lot larger than the height of the trench, it is thought that ballistic acoustic phonons from the SThM tip are the principal source of heat transfer, with 90% reaching the trench in the ballistic regime.

A graphene bulge that was still suspended over the trench displayed similar properties, ruling out experimental differences, such as SThM contact area, as the reason for such behavior.

These calculations concluded that when compared to single-layer graphene, three-layer graphene had around 68% of the thermal conductance. Lastly, thermal mapping of border regions between supported graphene layers reveals that the thermal transition region has a width of 50-100 nm, verifying theoretical calculations for the mean free path.

Figure 15. (a) SThM image of supported graphene, showing varied thicknesses throughout the sample; and (b) measured contact thermal resistance as a function of the number of graphene layers, showing reduction in thermal resistance as the number of layers increases.

Conclusions

Traditionally, 2D materials have been characterized using s-SNOM imaging methods for chemical analysis and optical characterization. 2D materials characterization is still evolving with the development of applications such as nanopatterning and functionalization, and the understanding of a more complete set of material properties, such as absorptive and radiative material properties.

Recent developments in near-field spectroscopy, for example the utilization of resonance-enhanced and tapping AFM-IR techniques and s-SNOM-based nanoscale FTIR spectroscopy, have extended the characterization possibilities for this class of materials to deliver unique and complementary information, speeding up learning and new discoveries.

The nanoIR3-s Broadband instrumentation helps these advances, and supplies the most comprehensive set of capabilities available today for 2D materials characterization.

References

1. Bartlam C et al., "Nanoscale infrared identification and mapping of chemical functional groups on graphene," Carbon 139 (2018): 317-24.
2. Loepz et al., "Large photothermal effect in sub-40 nm h-BN nanostructures patterned via high resolution ion beam," DOI: 10.1002/smll.201800072.
3. Khanikaev AB, Arju N, Fan Z, Purtseladze D, Lu F, Lee J, Sarriugarte P, Schnell M, Hillenbrand R, Belkin MA, and Shvets G., "Experimental demonstration of the microscopic origin of circular dichroism in two-dimensional metamaterials," Nature Communications (2017).
4. Rosenberger MR, Wang MC, Xie X, Rogers JA, N S, and K WP, "Measuring individual carbon nanotubes and single graphene sheets using atomic force microscope infrared spectroscopy," Nanotechnology (2017).
5. Zhang X, Wu H, Guo S, and Wang Y, "Understanding in crystallization of polyethylene: the role of boron nitride (BN) particles," Royal Social of Chemistry Advances 121 (2015): 99585-100407.
6. Pumarol ME, Rosamond MC, Tovee P, Petty MC, Zeze DA, Falko V, and Kolosov OV, "Direct Nanoscale Imaging of Ballistic and Diffusive Thermal Transport in Graphene Nanostructures," Nano Letters 12 (2012): 2906-2911.

This information has been sourced, reviewed and adapted from materials provided by Bruker Nano Surfaces. For more information on this source, please visit Bruker Nano Surfaces.