Identifying Chemistries with 10 nm Spatial Resolution Using the Inspire™ from Bruker

 

Table of Content

Introduction
Highest Resolution Chemical Imaging: Resolving the Phase Separation in Block Copolymers
Inspire’s Unique Chemical Imaging Technology
Extracting Richer Spectral Information from the Infrared Chemical Fingerprint
Opening the Door to New Research Opportunities with PeakForce Tappin
Conclusion

Introduction

Infrared (IR) spectroscopy is a standard optical technique that is widely used for material identification. Polymer blends such as liquid crystals or organic photovoltaics have unique material properties that can be optimized by correlating nanoscale inhomogeneities and material distribution with macroscopic performance. While IR spectroscopy is a commonly used technique, its spatial resolution is largely restricted by diffraction to about half the utilized wavelength, which is usually 5µm.

Scattering scanning near-field optical microscopy (s-SNOM) is a non-destructive, surface-sensitive technique based on atomic force microscopy (AFM) that effectively overcomes IR spectroscopy’s diffraction limit and allows chemical mapping with 10nm spatial resolution. This innovative method is used in various applications including imaging of local strain, nano-antenna modes, graphene plasmons, biomaterials, phase-transitions, and boron nitride phonon polaritons. This article describes the various techniques used on polymers.

Highest Resolution Chemical Imaging: Resolving the Phase Separation in Block Copolymers

In this analysis, a block copolymer is taken as a representative system having nanoscale chemical inhomogeneity. The panel (A) of Figure 1 shows a topography image of the spin-cast, 50nm thin film of the poly(styrene-block-methyl methacrylate) (PS-b-PMMA). However, this image is not sufficient to address the question as to whether both constituents PMMA and PS exist together in isolated phases or whether they are homogeneously mixed. In this scenario, topography data is generally acquired through the TappingMode without causing any damage to the sample which otherwise is not possible if contact mode is used, particularly in the case of biomaterials, polymers, and other soft samples. In Figure 1, it can be seen that at 1725cm-1 wavelength, the C=O bond in PMMA is probed through its resonantly driven stretch vibration. The silicon substrate acts as a reference material, and the green dotted line highlights the line profile position presented in (C) and (D). There is no correlation between the height in (C) and the absorption in (D), demonstrating that the absorption image does not map topography artifacts, but instead only maps the nanoscale chemistry.

Figure 1. (A) Topography of a spin-cast, 50nm thin PS-b-PMMA block copolymer film on Si substrate obtained in TappingMode AFM. (B) Corresponding nanoscale absorption at the third harmonic obtained with Bruker’s Inspire at 1725cm-1. The step in absorption shown at a position of 260nm in the height (C) and absorption (D) has a width of 15nm. This represents a measure of the obtained spatial resolution that is limited by the tip radius of <25nm.

Bruker’s Inspire system uses PeakForce Tapping that can quantitatively determine key sample characteristics, like work function, conductivity, or modulus. However, when it comes to imaging, the PeakForce Tapping mode is preferred for fragile samples such as cells and DNA. Though chemistries can be detected on the nanoscale using nanoelectrical or nanomechanical properties, alternative methods are generally used because these properties are not usually known. IR spectroscopy is an optical method that provides useful chemical data about the composition of materials. Hence this method is suitable for materials having special fingerprints in the IR spectral region. However, area-averaging, diffraction-limited far- field methods cannot be used in this case because of the less than 30nm length scale defined by the grains in Figure 1(A). Here, the s- SNOM technique provides a suitable option as it enables three orders of magnitude enhancement in terms of 10nm spatial resolution over far-field IR spectroscopy.

The quasi-lamellar PMMA domains shown in Figure 1(B) were obtained by using s-SNOM imaging on the PS-b- PMMA-block copolymer sample. In this image, the distribution of the carbonyl C=O bond can be seen clearly, and its stretching vibration was resonantly studied at 1725cm-1. Although PMMA contains the C=O bond, the PS chains lack this bond. The topography of panel (A) does not match with the distribution of PMMA. This can be observed from the line profiles obtained along the green dotted line in Figure 1(B) and illustrated in Figure 1(C) and (D) for height and absorption, respectively. As a result, the absorption channel rendered by Bruker’s Inspire system does not probe the topography artifacts and simply analyzes the chemical composition.

Considering the example shown here, the AFM probe coated with PtIr had a nominal tip radius of less than 25nm. This probe is needed for the s-SNOM technique. The size of the tip facilitates a spatial resolution in chemical mapping of less than 15nm. This value can be predicted from the step width in the line profile shown in Figure 1(D) at the 260nm position in an area that has only minimum topography changes. Therefore, this factor rules out any crosstalk with topography changes. The <15nm resolution is typically obtained without using any special tip or without any sample preparations. It should be noted that when compared to the contact mode, the TappingMode prevents sample or tip damage, thereby ensuring highly reproducible results throughout multiple scans.

Inspire’s Unique Chemical Imaging Technology

In Bruker’s Inspire system, the technology that allows s-SNOM is built on Bruker’s MultiMode® AFM platform, which is integrated with PeakForce Tapping. The AFM includes an optics box that illuminates the AFM tip and enables further analysis of the backscattered light, as shown in Figure 2.

Figure 2. Bruker’s fully integrated Inspire system combining PeakForce Tapping technology with chemical identification at the nanoscale, based on s-SNOM.

Phase extraction and near-field amplitude are obtained in Figure 3(A) using an asymmetric Michelson interferometer. In this interferometer, IR light from a laser source, for instance, a continuous wave quantum cascade laser, is divided and one part of the laser is directed towards the end of a metallized AFM tip. A MCT detector is used to detect the backscattered light from the tip where it impedes the reference light. Through two- phase homodyne detection, phase-sensitive measurements can be made to acquire near-field absorption and reflection information with nanoscale spatial resolution. In two-phase homodyne detection, near-field signals are obtained with the reference mirror swapped between two positions. Subsequently, Fourier transformation of the time-dependent signals enables true near-field extraction and background suppression when detecting at higher harmonics of the tip oscillation frequency Ω.

Figure 3. (A) Schematic of the s-SNOM implementation in Inspire based on an asymmetric Michelson interferometer. This shows light from an IR laser source being divided into a reference beam, which is reflected from a sample beam and a motorized mirror. The beam is directed towards an AFM tip. (B) Schematic of the AFM tip illustrating the confinement of near-field signals Enf to small tip-sample distances compared to background signals Eb. (C) TappingMode operation results in harmonic modulation of the unwanted background scattering signal while the strongly nonlinear decay of the near-field signal with height above the sample results in anharmonic signals.

Simultaneously, the incident IR laser light gets polarized along the AFM tip to merge this tip with the electric field. Here, the AFM tip serves as a nanoscale antenna, which gets polarized. The metal-coated AFM tip allows a lightning-rod effect for the incident light, that is, in concentration and field-enhancement at the top of the tip with field enhancement factors of several orders of 10s in the IR region.

On contacting a sample, the decaying evanescent fields at the tip’s apex lead to local sample polarization. This polarization is measured by the optical constants of the material. Subsequently, the sample polarization reflects back on the tip polarization, and changes the polarization as radiation into the far-field. To put it simply, the light is dispersed by the AFM tip without any change in the photon energy. The radiation, thus scattered, includes data about the local dielectric characteristics of the sample held under the AFM tip. As shown in Figure 1, the spatial resolution is shown in first order by the tip’s radius, normally in the range of 10 to 20nm. A focusing element such as a lens collects the radiation, which is then studied in a Michelson-type interferometer. Over time, high spatial resolution was seen further than the far-field diffraction limit acquired in s-SNOM. This effect was due to the evanescent fields that were limited to the apex of the tip through the lightning-rod effect of the metal-coated tip. Subsequently, the exponential decay of the near-field signal within the probed material confines the depth resolution to several tens of nanometers for media that are optically transparent, and thus makes s-SNOM a non-destructive, surface-sensitive method. A relatively larger background scattering signal, which hides the backscattered near- field signal, lacks the preferred sample-specific localized data. The diffraction-limited size of the IR focus is the source of the intense background scattering, and this IR focus illuminates the AFM tip with a spot size of several tens of micrometer, which has to be compared to the standard tip measuring 10µm long. Accordingly, strong background signals are induced by the tip, the cantilever, and the scattering of the sample. In spite of its larger magnitude when compared to the near-field signal, suppression of the near-field extraction and background turns out to be quite easy.

The method that is most often used to do this is modulate the distance between the tip and the sample. This method correlates with the typical TappingMode AFM operation. On account of the instant decay of evanescent waves, the near-field signal increases nonlinearly near the surface of the sample. However, the far-field scattering only changes linearly with respect to the sample distance (As shown in Figure 3(B) and (C)). As a result, the harmonic movement of the tip in the TappingMode at the cantilever resonance frequency Ω leads to a time-domain signal, which is harmonic or unharmonic for the background or the near-field contribution, respectively. Background contamination is then effectively suppressed by demodulation and signal detection at higher harmonics nΩ, which in turn include less background signal at the expense of signal strength. Near-field signals are usually obtained either at the second or third harmonic of the cantilever resonance frequency. Here, the tapping frequency Ω is tens to hundreds of kilohertz at a normal probe oscillation amplitude of several tens of nanometer.

An asymmetric Michelson-type interferometer is used to improve the weak scattered near-field signal as well as to acquire the phase of the scattered light and the electric field. This approach provides an easy option to achieve the near-field reflection and absorption, which is the nanoscale analogue to the traditional far-field Fourier-Transform Infrared (FTIR) reflection and absorption. Figure 3(A) shows a standard configuration for this two-phase homodyne detection scheme. A 50:50 beamsplitter is used to split the IR light source. A piezo-controlled mirror included in one arm of the interferometer reflects the light to render a reference beam. This beam is subsequently directed towards the MCT detector. The AFM tip is included in the second arm of the interferometer and its backscattered light is superimposed on the MCT detector along with the reference beam. The MCT detector then determines the absolute value squared of the sum of the involved electric fields. The recorded interference signal thus obtained on the detector V includes the near-field phase and amplitude, φnf and Enf in that order, as well as the relative quantities for the background signals (Eb, φb) and the reference (Er, φr):

The above equation can be simplified to extract the preferred near-field phase and amplitude. Through signal demodulation, the last four terms are fully removed and the second term can be overlooked because Er >> Eb. The first term alone is left out. It can be seen that multiplication with the larger reference field results in the amplification of the near-field signal. A two-phase homodyne detection scheme is used to acquire near-field phase and amplitude. Under this scheme, the signal is obtained in dual positions of the reference mirror that corresponds to two reference phases φr. The positions of the mirror are altered with the s-SNOM tip placed over a non-absorbing material in the IR region of interest, like silicon or gold, which is a highly-reflecting material. However, in reality, the mirror is transformed till there is a maximum in the near-field signal, thereby corresponding to a disappearing phase variation between near-field and reference phases. The signal is fully suppressed by shifting the mirror by 1/8 of a wavelength which corresponds to a reference phase shift by pi/2. Bruker’s Inspire system is integrated with software that can be used to make these adjustments. Once the positions of the reference mirror have been defined in the depicted way on the non-absorbing material, the s-SNOM tip is placed over the sample. Now, the detector determines a signal Vrefl∝ Er Enf' cos(φnf') and Vabs∝ Er Enf' sin(φnf') corresponding to the 1st and 2nd mirror positions respectively. The signals are relative to the sample’s reflection and absorption, respectively, that is, they determine the out-of-phase and in- phase near-field components with regard to the light reflected from the reference mirror. The signals also indicate the true and imaginary area of the near-field Enf'exp(iφnf') from which it is possible to acquire the phase and amplitude. Through this method, material can be identified without the added necessity of extensive modeling of the near-field interaction, because the absorption spectra acquired overlap with FTIR spectra to a good approximation. This direct interaction is applicable in the case of weak oscillators that include most biomaterials and polymers. A positive part of the dielectric function across the material resonances defines weak oscillators. Modeling is needed to interpret the complicated near-field spectra determined on other materials and for quantitative extraction of dielectric functions. There are various models that span from those facilitating a qualitative understanding to those providing a complete quantitative description of the interaction between the tip and the sample. The Inspire system implements the simplified two-phase homodyne detection scheme, which has featured in a number of publications to demonstrate good agreement between traditional far-field FTIR data and near-field absorption data in the case of biomolecules or polymers. These aspects make s- SNOM a perfect modeling-free method that can be used in many practical applications.

Extracting Richer Spectral Information from the Infrared Chemical Fingerprint

A spin-cast PS-PMMA blend on a silicon substrate was examined to highlight the high chemical sensitivity of this method, its spectral precision, and its ability to provide data that can be easily interpreted and also relates to far-field spectra. The topography with prominent PMMA domains integrated in the PS matrix is shown in Figure 4(A). Figures 4(B) and (C) show the reflection and absorption maps respectively for different kinds of IR laser frequencies adjusted around the PMMA’s carbonyl resonance.

Figure 4. (A) Topography of a PMMA domain within a PS matrix. (B) S-SNOM reflection and (C) absorption at the third harmonic. (B) and (C) show a constant reflection and vanishing absorption of the PS matrix for different infrared frequencies where the PS material with its flat, absorption-free dispersion serves as a reference material. On the contrary, the PMMA domains exhibit a distinct wavelength- dependent behavior which is extracted in panel (D) for multiple frequencies. Both absorption and reflection can be fitted with a common Lorentz oscillator model (solid lines).

The reference mirror phases in the two-phase homodyne detection scheme can be set by using the flat and non-absorbing IR response of the PS material in the specified region. As shown in Figures 4(B) and (C), the PMMA’s nanoscale absorption and reflection varies with laser frequency and can be used for tracking the distribution of PMMA. The acquired reflection and absorption profiles in Figure 4(D) are collected on the PMMA domain and have a shape typical of a Lorentzian absorption line. This has been confirmed by fitting with the standard Lorentz oscillator model. In a recent publication, it has been shown that the same repetitive imaging was used at distinct laser frequencies to sweep across the PMMA carbonyl resonance in a PS-b-PMMA block copolymer system. This block copolymer system is a sample that is analogous to the sample imaged in Figure 1. Line widths as well as absorption line center frequencies were determined with a spectral accuracy of 0.2cm-1. This precision helped overcome the intermolecular Stark shifts of the carbonyl resonance that takes place between the interface and middle of the PMMA domains owing to local electric fields, which vary considerably.

In yet another instance that highlights the high chemical sensitivity of the s-SNOM technique, the Inspire system successfully resolved the vibrational resonance of phosphate in an ultra-thin biomaterial down to 1nm thin protein nanoribbon. These experiments reinforce the fact that near-field absorption is highly reproducible and correlates with far-field information to allow material identification without the added necessity of modeling the actual near-field tip-sample interaction. The experiments also underline the excellent spectral precision and chemical sensitivity that are reliably obtained with near-field imaging. These aspects enable quantitative analysis of the chemistry of protein nanoribbon, or the physics of minute effects such as the Stark shift between polymer chains

Burker’s Inspire system was also utilized to demonstrate the extensive applications of the s-SNOM technique as well as the additional data it provides regarding sample properties. Here, the Inspire system was employed to study the biodegradable polymer of poly(3-hydroxybutyrate-co-3-hydroxyvalerate) or PHBV. Figure 5(A) shows the topography of the PS-PHBV blend acquired from spin-casting, exposing the following morphologies:

  • Disk-like structures that are integrated in a background matrix
  • A long island-like structure that has a lower height when compared to the matrix

Figure 5. (A) Topography of a PS-PHBV blend. Two morphologies can be observed: disk-like ones and an extended, island-like one at the top left, both embedded into the background matrix. (B) Nanoscale reflection and (C) absorption for different frequencies around the carbonyl vibrational mode. In panel (C) PHBV is identified from resonant absorption mapping around 1720 cm-1. PHBV is found to occur in the two different morphologies. The island-like domain in the top left additionally shows some internal structure that cannot be seen in the topography image.

Figures 5(B) and 5(C) show the nanoscale reflection and absorption, respectively. The same were taken over PHBV’s carbonyl vibrational resonance and continue to behave in a similar fashion as the PS-PMMA example shown in Figure 4. In Figure 5(C), 1720cm-1, the chemical absorption map on resonance assigns two morphologies in topography to PHBV. On resonance, boundaries that are less absorbing are seen in the island-resembling structure. This may suggest partial phase separation between PHBV and PS materials, or may possibly correspond to a local variation in degree of crystallinity within the PHBV. This shows how the s-SNOM technique can be effectively used for acquiring additional data regarding the PHBV domains that cannot be seen in the topography image alone.

Opening the Door to New Research Opportunities with PeakForce Tapping

The afore-mentioned polymer examples demonstrate the Bruker’s Inspire system effectiveness. A chemically sensitive tool for nanoscale imaging. Given that s-SNOM is a well-established AFM-based technique, it can be complemented and improved by AFM modes besides the TappingMode. PeakForce Tapping is ideal for imaging biomaterials, polymers, and other similar soft samples. In PeakForce Tapping, the distance between the tip and sample is sinusoidally modulated, and the tip-sample in contact is regulated to maintain a constant peak force.

Accurate control of the interaction force between the tip and sample facilitates analyzing and capturing approach and allows withdrawal of curves at individual pixels with a repetition rate of 1 to 2kHz. This process can be carried out at user-defined forces down to tens of pN. PeakForce Tapping provides a major benefit because its force feedback maintains the sample’s AFM tip. Hence, this method is suitable for imaging fragile samples with very high resolution and enables routine observation of both minor and major grooves in the DNA double helix structure. Another benefit of the PeakForce Tapping is that it is possible to obtain and assess force curves in real-time during the imaging process to achieve the sample’s mechanical properties in a quantitative manner, including the deformation, adhesion, and modulus. The various benefits provided by PeakForce Tapping can also be transferred to conductive samples, as this mode can be integrated with current measurements in PeakForce TUNA™ tunneling and potential measurements in PeakForce Kelvin probe force microscopy (KPFM™). The MultiMode 8 AFM equipped with the full PeakForce Tapping capabilities serves as the basic AFM platform for the Inspire system. As a result, the different AFM modes can be integrated with s-SNOM, thereby providing new research opportunities and enabling correlated and concurrent nanomechanical, nanoelectrical and nanochemical data acquisition. Given that the Tapping Mode has to be used in s-SNOM to suppress the background signals, s-SNOM is either implemented in interleave mode or LiftMode when integrated with PeakForce Tapping. The s-SNOM-PeakForce Tapping combination is demonstrated on a sample of low-density polyethylene (LDPE) material, which is often used to make plastic bags. Figure 6(A) shows the topography of a PS-LDPE blend.

Figure 6. (A) Topography of a PS-LDPE blend and simultaneously acquired modulus in (B) and infrared reflection in (C) at 1880cm-1. The taller disks in topography (A) are identified in (B) as the softer LDPE domains compared to the PS matrix. The reflection image in (C) confirms this assignment. Here, LDPE with its lower refractive index in the absorption-free, flat spectral region exhibits lower near-field reflection. This example highlights the correlative measurement possibilities. Furthermore, it shows that material identification by the reflection information alone is feasible.

The contrast in the modulus mapping of Figure 6(B) is rendered by different modulus of both constituents. Consequently, the harder PS material with 2GPa modulus acts as a matrix where LDPE disks with 0.1GPa modulus have been integrated. At the IR frequency of 1880cm-1 these materials do not absorb, while a distinct contrast in the concurrently acquired reflection channel is seen (Figure 6(C)). When compared to the PS matrix, LDPE with a lower refractive index has a lower near-field reflection signal. In this case, materials can be easily differentiated using the non-resonant, nanoscale reflection data. This concept is generally utilized when investigating metallic domains in insulator-to-metal phase transition materials. As a general rule, the higher nanoscale reflection helps in differentiating metals from lower-reflective insulators. However, plasmon or phonon resonances can lead to resonant signal improvement, wherein the total near-field amplitude can surpass that of metals.

Conclusion

This article demonstrates the various applications of s-SNOM and provides polymer research examples. It also underlines the major benefits offered by this technology. Bruker’s Inspire system integrates a scattering SNOM microscope and the MultiMode 8 AFM platform. s-SNOM is a standard method that allows direct acquisition of nanoscale absorption and reflection data, and provides modeling-free information that can be easily understood and compared against traditional FTIR spectra. Polymers that were phase-separated were chemically detected depending on a vibrational resonance, while the high spatial resolution of 10 to 20nm was shown on a block copolymer film. In the majority of cases, materials can be identified through mapping reflection properties alone. In the context of correlative nanomechanical and nanooptical mapping, s-SNOM combined with the AFM mode of PeakForce Tapping was also discussed in detail.

PeakForce Tapping maps the nanomechanical properties, images delicate samples, and enables either amplitude modulation or frequency KPFM. It also supports high-resolution conductivity mapping in conducting or tunneling AFM and thus protects soft samples. Here, contact mode is not needed to make conductivity measurements. With the help of Bruker’s Inspire system, nanomechanical, nanooptical, nanopotential, and nanoelectrical experiments can now be performed easily, therefore paving the way for new research applications. In polymer science studies, the Inspire system also helps correlate material distribution and functionality in new fuel cells, organic photovoltaics, conducting polymers, liquid crystal films or organic electronics.

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.

Ask A Question

Do you have a question you'd like to ask regarding this article?

Leave your feedback
Submit