PTIR (Tapping AFM-IR) vs. PiFM: Comparing AFM‑IR Techniques

Background

Since the invention of atomic force microscopy (AFM), researchers have sought technologies that would bring conventional chemical analysis techniques like infrared spectroscopy to a much smaller spatial volume. Currently, there are a few competing nano-IR techniques which aim to offer these abilities.

Photothermal induced resonance (PTIR) was an early technique to combine infrared (IR) spectroscopy with AFM. Along with s-SNOM (also known as apertureless NSOM), PTIR has provided nanoscale IR analysis from as early as 2005 [1]. In 2010, photo-induced force microscopy (PiFM) was introduced as a third AFM-IR technique based on non-contact AFM. Today, PTIR (sometimes called Tapping AFM-IR or tapping AFM-IR+ by various manufacturers) and PiFM are the two most technically advanced AFM-IR techniques commercially available. However, there are some important technical differences that make PiFM more desirable.

Contact mode PTIR

The first generation of PTIR was based on contact mode AFM which is where a cantilever with a relatively low force constant is in contact with the sample. The sample was deposited onto a ZnSe (or other similarly IR transparent material) prism and illuminated from below with an IR laser via total internal reflection [1]. The laser was pulsed at a low frequency (~1 kHz) to create an evanescent wave onto the sample. The sample would then absorb the light and expand proportionally with the absorption strength. This pulsing expansion was measured by monitoring the AFM cantilever’s response.

In this approach, the spatial resolution was not determined by the tip radius alone, but also by the thickness of the sample via mechanical coupling to surrounding material. Therefore, there existed a tradeoff between the spatial resolution and the signal-to-noise ratio (SNR). Thicker samples offered better SNR but poorer spatial resolution due to a larger bulge of thermally excited material. Samples needed to be ~100 nm thick to achieve a reasonable SNR. This requirement set the spatial resolution of contact mode PTIR to about 100 nm. For context, modern nano-IR techniques can achieve resolutions of at least 10 nm, and PiFM can achieve resolutions of ~5 nm.

Principles of PiFM

In 2010, photo-induced force microscopy was demonstrated by Dr. Kumar Wickramasinghe. This technique is based on non-contact AFM, and it relies on tip-sample force interactions. PiFM revolutionized the AFM-IR market space because it immediately had better spatial resolution than PTIR, and it didn’t have any of the experimental difficulties associated with nano-FTIR which is a descendant of s-SNOM. PiFM also works on both organic and inorganic samples due to how the force interactions work in non-contact mode. PiFM accomplishes all this in a few different ways, which will become apparent once the underlying mechanisms are understood.

In PiFM, the sample is excited from above using a tunable IR laser source focused onto a metal coated AFM tip. This avoids the tedious process of having to mount samples on an IR-transparent prism. However, it has another key benefit: tip enhanced optical illumination. By shining the light onto a metal coated tip, the tip itself creates a highly local enhanced field. This not only helps boost signal but also increases the resolution.

PiFM also avoids tip and sample damage by operating the AFM in non-contact mode. The cantilever is dithered with a small free amplitude, typically around 1 nm. The setpoint is also kept relatively high meaning that tip-sample engagement is reached when only a small reduction in the free air amplitude is detected. This means that the tip is kept free from contamination, and damage to soft samples is largely eliminated.

The physical mechanisms underlying PiFM offer some important signal processing enhancements over PTIR. To make a photo-induced force (PiF) measurement, the optical excitation of the sample is detected via a force interaction. Laser illumination of the metal coated AFM tip locally excites the material near the surface of the sample. Depending on the material below the tip, there are at least three underlying mechanisms thought to contribute to the signal: an attractive dipole-dipole interaction between the sample and the AFM tip, van der Waals-mediated thermal expansion forces, and Opto-mechanical damping. Each of these highly local mechanisms affect the tip in a subtly different way, but they can all be detected using the same multi-modal AFM techniques which allow PiFM to be both extremely sensitive and extremely high resolution.

There are two primary detection modes for PiFM, both of which offer Q-enhanced signals: direct drive detection and sideband bimodal™ detection. Both detection modes can be used to take spectra and images.

Direct drive detection operates by modulating the sample excitation at a resonance frequency of the cantilever. Direct drive PiFM can use either the first or second free-air resonance of the cantilever. Forces generated by this modulated light directly stimulate the cantilever at the chosen resonance, and so the PiF signal is amplified by the quality factor of the cantilever at that mode which is typically 400–600. In this way, the strength of the PiF directly excites the cantilever. Besides the higher quality factor of the non-contact mechanical resonance mode, variations in the resonant frequency due to tip-sample interactions are easily managed with PiFM because it operates in a non-contact AFM mode.

Sideband bimodal™ detection aims to suppress background signals and increase spatial resolution by taking advantage of frequency mixing. Instead of modulating the excitation laser at a cantilever resonance, the laser is modulated at a frequency that when mixed with either the first or second mechanical resonance of the cantilever will produce a difference frequency (sideband) at the other cantilever resonance.

Typically, the laser modulation frequency, f_m, is chosen such that f₁ = f₂ − f_m where f₁ and f₂ are the first and second resonance modes of the cantilever, respectively. The cantilever is usually driven at f₂, which is used to detect the AFM topography, while the laser is modulated at f_m. Because the sample is excited at f_m, and the strength of the force interaction is dependent on the tip-sample distance modulated at f₂, frequency mixing produces a photo-induced force sideband component at f₁. This means the signal detected at f₁ depends on the force gradient of the PiF, and not just the strength of the PiF itself. Therefore, the resolution of the system is enhanced. An intuitive description would be to say that by forcing the strength of the signal to be dependent on the tip-sample spacing, the effect is going to be strongest and more localized immediately under the apex of the tip, which increases resolution.

By having both detection modes available, PiFM offers great flexibility when taking measurements. The stronger distance dependence of sideband detection results in increased surface sensitivity. By comparing to direct detection, one can infer results about the surface vs bulk chemical composition of a sample. Just like direct drive, the sideband detection mode signal is enhanced via the oscillation of the cantilever with a quality factor of between 400–600. Lock-in detection at the resonant frequency produces results with incredible sensitivity on both organic and inorganic materials. In fact, sideband detection can measure single-molecule-level materials with 5 nm lateral resolution regardless of what substrate the sample is on.

Between operating in a non-contact AFM mode, the significant Q-enhancement factors to amplify the PiF signals, the variety of detection schemes between direct drive detection and sideband bimodal™ detection, and the superior spatial resolution, PiFM completely changed what was possible with AFM-IR instruments.

Revised versions of PTIR

Resonance enhanced PTIR (resonance enhanced AFM-IR)

After the introduction of PiFM, resonance enhanced PTIR (RE PTIR) was introduced in 2014 to make significant improvements to the sample preparation and measurement requirements of contact mode PTIR [2]. First, the excitation laser is pulsed at the contact resonance of the cantilever. Just like PiFM, exciting the sample at this frequency amplifies the signal by the quality factor (Q) of this resonance. However, the Q factor in contact mode is much lower at about 50 than it is for PiFM. Also, this resonance frequency will be unstable since small differences in the way the tip lands on the material means that the contact resonant frequency, and therefore the Q enhancement factor, changes each time a measurement is taken. Even still, this technique enhances the SNR of the thermal expansion measurement when compared to contact mode PTIR.

Another improvement is that the sample is excited by the laser from above which eliminates the tedious mounting of the sample onto the ZnSe prism. This also gains the advantage of using a tip-enhanced beam, just like PiFM.

Due to the enhanced SNR, samples thinner than 100 nm can be measured. By depositing the sample onto a metal substrate (such as gold), even a monolayer sample can be successfully analyzed with a spatial resolution of ~20 nm, which is the typical radius of curvature for a metal coated AFM tip. RE PTIR produces nanoscale IR spectra that are well correlated with bulk FTIR measurements and is therefore still the recommended version of PTIR for acquiring nano-IR spectra.

Even with all these improvements, RE PTIR still has some notable shortcomings. First, the thermal expansion signal gets coupled to material differences in the sample since the contact resonance is less stable. Also, the technique is somewhat constrained by the thermal properties of the sample. RE PTIR works best on samples that have a large coefficient of thermal expansion which limits its capabilities to primarily work on organic molecules. Additionally, because RE PTIR relies on the cantilever to contact the heated sample, there is significant potential for the tip to become contaminated during analysis or for soft samples to be damaged. In sum, while this was a significant improvement over the first generation of PTIR, RE PTIR requires vigilance to acquire good data, and it cannot match the resolution or sensitivity offered by PiFM.

Tapping mode PTIR (tapping AFM-IR)

Released in 2017, tapping mode PTIR (TM PTIR) was developed to try to achieve similar sensitivities and spatial resolutions as PiFM sideband bimodal™ detection using a thermal expansion measurement [3]. By operating the AFM in a hard tapping mode, sample damage is reduced from the constant contact mode of RE PTIR. TM PTIR also adopts a heterodyne excitation that is similar to PiFM sideband bimodal™ detection [4] which causes much confusion to potential users of these techniques.

In heterodyne excitation, the mixing scheme is nearly identical as for sideband bimodal™ PiFM: the laser is pulsed at f_m where f₁ + f_m = f₂. In this case, f₁ and f₂ are the first and second tapping mode resonances of the cantilever. In TM PTIR f₁ is typically used to drive the cantilever so that frequency is mixed with f_m and the sum frequency at f₂ is used to measure a Q-enhanced IR signal. Since f_m is not a cantilever resonance, the thermal expansion signal would not be enhanced by any quality factor unless there is frequency mixing.

TM PTIR induces signal mixing by operating in the transition between contact and noncontact to exploit a part of the van der Waals force curve with high gradient. To accomplish this, the cantilever must tap the sample hard to enter the repulsive regime of the force curve. This requires “a higher free amplitude (>10 nm typically) and/or an amplitude setpoint corresponding to a larger percentage reduction of the free air amplitude” [4]. While this could theoretically produce a nonlinear coupling coefficient that is much larger than in the attractive mode used by PiFM, it appears to not have much practical benefit as other challenges are introduced from the hard tapping. For example, the stability of the cantilever tapping mode resonant frequency will be worse, making lock-in detection of the signal more difficult. Additionally, this operation introduces a strong material-dependent tip-sample mechanical interaction, which could convolute the photo-excited thermal expansion signal. Empirically, mixing products from both sideband bimodal™ PiFM and TM PTIR are easily detectable, but PiFM offers better usability by operating in a curved part of the force curve away from contact. This is the primary difference between the two techniques.

While there may be some theoretical arguments for the measurement scheme of TM PTIR, there are no known practical improvements over PiFM. Instead, the harder contact required for TM PTIR has significant implications including the possibility for sample damage, tip contamination problems, resonant frequency instability, material dependent signal convolution, and lower SNR on inorganic samples. Furthermore, PiFM still offers higher spatial resolutions.

Summary

Today, PiFM, RE PTIR, and TM PTIR are the leading nano-IR techniques available. The primary difference between these techniques is how the excitation of the sample is detected, but this difference has some important consequences. PiFM operates in an attractive non-contact regime while PTIR-based techniques rely on repulsive contact forces. By operating in a non-contact mode, PiFM can provide non-destructive measurements on organic and inorganic materials with spatial resolutions of only 5 nm. By using both direct drive detection and sideband bimodal™ detection for imaging and spectra, PiFM can also provide more information about a sample at varying depths. Therefore, PiFM offers the most versatility and the highest spatial resolutions possible.

PTIR-based techniques use alternate physical mechanisms to provide similar results as PiFM. However, because these techniques rely on repulsive force interactions, issues such as tip contamination, sample damage, poor performance on inorganic materials, and material dependent signal convolution become a much greater concern. With these shortcomings and worse spatial resolutions, it is clear PiFM is the most capable AFM-IR technique available.

Table 1 summarizes some key traits of both PiFM and PTIR.

Specification	PiFM	PTIR
Laser Type	QCL OPO/DFG	QCL (CW/P), OPO/DFG
Focus Optics	Parabolic mirror	Parabolic mirror
Spectral Range	QCL: 760 – 1900 cm−1 OPO/DFG*: 550 – 4400 cm−1	QCL1: 900 – 1900 cm−1 QCL2: 1900 – 2600 cm−1 OPO: 2710 –3600 cm−1
AFM Operation	Non-contact/light TM	Contact/hard TM
Near-field Detection	Attractive force	Repulsive force
Background Suppression	PiFM: Heterodyne detection s-SNOM [6]: Generalized lock-in	TM PTIR: Heterodyne detection s-SNOM: Homodyne
IR Spectrum Acquisition Mode	Non-contact	Contact (RE PTIR)
IR Spectrum Quality (Organics)	Excellent	Excellent (when thickness > 50 nm)
Data Quality (Inorganics)	Excellent	Poor
Data Quality (E-field)	Excellent	Poor
Dielectric Constant Imaging	Yes	No
Spatial Resolution	~ 5 nm	~ 20 nm (RE PTIR), ~ 10 nm (TM PTIR)
Spectrum Sensitivity (Thickness)	Single-molecule-level	>~ 20 nm (RE PTIR), monolayer (if on Au)
Spectrum Acquisition Time	QCL: < 1 second OPO/DFG: < 1 second per 1000 cm−1	QCL: < second for each QCL OPO: Unknown
s-SNOM Option	Yes	Yes, for some models

References

1. A. Dazzi et al., Optics Let., 30, 2388 (2005)
2. F. Lu et al., Nature Photonics, 8, 307 (2014)
3. Prater et al., US Patent #10 ,228 ,388 B2
4. Nowak et al. Sci. Adv. 2016; 2: e1501571 (2016)
5. https://molecularvista.com/technology/pifm-and-pif-ir/scientific-principles/
6. A good discussion of the deficiency of homodyne detection compared to pseudo-heterodyne detection is given here: N. Ocelic et al., App. Phys. Let. 89, 101124 (2006) while the advantage of Generalized Lock-in over the pseudo-heterodyne is given here: https://molecularvista.com/applications/vista-ir-s/