Nanoscale Light-Matter Interactions - Q&A

August 12, 2020 by Claudius Riek

This blog post answers questions asked by the audience during our webinar "Nanoscale light-matter interactions" hosted by Physics Today and featuring guest speakers Markus Raschke and Samuel Johnson.

The webinar was recorded and is available on demand here.

Here’s a selection of questions asked during the webinar, which we have answered retrospectively. The questions are divided between instrumentation-focused ones answered by me and s-SNOM specific questions answered by Markus and Samuel.


How many sidebands can you analyze simultaneously?

The Lock-in amplifier of Zurich Instruments can demodulate a maximum number of 8 frequencies in parallel, depending on the instrument. In the AM/FM modulation option this leads to the demodulation of up to 2 carriers with 4 sidebands.

Can the PLL be used in parallel to the lock-in amplifier in your devices?

Yes, all the options we provide in our instruments can be used in parallel. An important application for the PLL/PID is actually, to use the output of the demodulator as the error signal input for PLL. Here the implementation in one instrument is beneficial since delay times between the different parts are minimized.

What is the shortest temporal window for the boxcar averager?

The shortest Boxcar window is one sample of the 1.8 Gsa/s. Still, typically we suggest using a little longer window since this makes the measurement more robust and the Signal-noise-ratio will be affected only very little.

Is a boxcar averager always better for a small duty cycle signal?

Although the boxcar averager often is indeed the better choice for a short duty cycle signal. There is not a general answer - since the best choice of the measurement scheme depends on various parameters, e.g. signal shape, noise floor, desired precision. To find the best solution for your specific use case, please get in touch, we are happy to help with an assessment.

Do I need two devices, if I want to use a boxcar averager and a lock-in amplifier in the tandem scheme?

Because the UHFLI Lock-in Amplifier can be equipped with the Boxcar averager option only one instrument is required to perform the measurement.

What is the omega in the first stage here (of the tandem demodulation scheme)?

Omega refers to the carrier frequency of the signal - the higher frequency modulation of the signal to analyze.

In pump-probe measurements, I see you use a forward imaging scheme. However, in biological imaging applications, we usually collect a backward signal, and this signal is generally weak. So how well is an LIA can demodulate the very weak backscattered signal?

Measuring small signals is actually the very purpose of a lock-in amplifier. The weaker the signal is, the more you will benefit from the use of a Lock-in compared to just digitizing the signal. Therefore the benefit in a backscattering geometry will be huge.

When collecting multiple sideband information, would the amplitude and phase of each peak be analyzed separately?

Yes, the amplitude and phase of each sideband is measured. In fact, an independent demodulator is used for each sideband.

Nanoscale light-matter interactions

What is the highest spatial and temporal resolution you can achieve using SNOM?

To first order, the nearfield resolution is given by the radius of the tip apex. Typically, we use tips with 30 nm radius and achieve similar resolution -- however, sharper tips can provide higher resolution.

How long does it take to collect a hyperspectral image and what limits the SNR for faster acquisition?

The time it takes to collect a hyperspectral image is a function of how many spatial points are collected, the desired spectral resolution, and the SNR on the sample of interest. Typical spectra are on the order of ~1 minute and are primarily limited by laser, AFM, and detector noise.

Can you apply this to other kinds of perovskite? What other scientific questions can we address related to perovskite?

Infrared vibrational nano-imaging can be applied to any perovskite samples with organic cations, which are fairly common in many hybrid perovskites, including 3D and 2D perovskites. For example, we can detect vibrational resonance arising from methylammonium cation as well, in addition to formamidinum that we study here. In particular, methylammonium's NH stretch or bend mode is known to be sensitive to hydrogen bonding between cations and anions. We are applying this to mixed anion perovskites where anions are known to migrate upon light illumination, and we can study this intriguing phenomena from the perspective of cation-anion hydrogen bonding.

What is the importance of doing nanoscale pump-probe particularly in the mid-infrared range, related to perovskite?

The transient mid-infrared response has recently been shown to be related to polaron absorption, as recently shown by Asbury group at Penn State University. Polaron, where photoinduced charges couple with the phonon modes, is believed to be a fundamental mechanism leading to long carrier lifetime in perovskites. Photovoltaic performance is known to be spatially heterogeneous across the perovskite film, so we elucidate the origin of such heterogeneity at the most fundamental level.

How do you know the signal acquired in the pump-probe measurement is related to the polaron?

John Asburys group at Penn State University did a far-field study on a variety of perovskites and found a new peak emerges around ~1200 cm-1 with a characteristic lineshape with an extended tail to the high-frequency side, which agrees with polaron absorption well. We tune our laser close to the peak of the polaron resonance. In this study, the laser bandwidth (~100 cm-1 FWHM) is too narrow to observe the entire polaronic spectral landscape, but certainly, enough to resolve the polaron dynamics with spatio-temporal resolution. We are working on novel light sources with extremely large spectral bandwidth to nano-probe the heterogeneity in polaron size and stabilization energy.

Why is a broadband source used for s-SNOM precision spectroscopy?

Broadband light sources are useful for spectroscopy as FTIR spectroscopy is less sensitive to light source noise and is easier to reference and compare peaks within a single spectrum (Fellgett's and Jacquinot's advantages).

Do you use active feedback to stabilize the interferometer of your SNOM?

In the IR, active feedback is not typically necessary as the long wavelengths are less sensitive to small variations in the interferometer. Some schemes for single wavelength imaging have drift manifest as a slowly varying phase across the image which can then be plane subtracted to remove the drift induced artifact.

What is the full form of MCT Detector?

While different frequencies require different detectors, we typically use mid-IR photovoltaic detectors with an element size of 0.1 mm.

How can we make use of the nanoscale light analysis in astrophysics research?

SNOM can be and has been used for the analysis of, for example, meteorites.

Are there any hardships that you may face as a nano light-matter interaction researches?

As the resolution increases, the total sample volume being probed is reduced, which reduces the total signal. Work is ongoing to be able to study smaller and smaller sample volumes like, for example, single molecules.

How do you account for the thermal effects in sensitive measurements using ultrafast infrared nanoimaging?

We perform power-dependent measurements to determine if the signal we are observing is due to thermal interactions and compare our results with those obtained in the far-field to piece together what is a local effect from heating and what isn't.

Are the kinetically trapped defects structural in nature?

For lead halide perovskites, it has been increasingly more established that optoelectronic responses are heterogeneous within a spin-coated film, and such heterogeneity has in part been attributed to structural non-uniformity, as suggested in our vibrational IR nano-spectroscopy as well as micro-XRD studies.

How is the inhibition of excitons leads to an improvement in photovoltaic efficiency?

In perovskites, exciton-binding energy is very low that a major portion of the excitons is believed to split into free carriers at room temperature. This spontaneous formation of free carriers is key in the perovskite photovoltaics. A large polaron formation, which stabilizes such free carriers, likely facilitates this process.

Can you distinguish between exciting the polarons by the nearfield effect and exciting them directly be the focussed fs-pulses?

Yes, one way is to compare the carrier decay rate of the near-field pump-probe signal associated with the tip-apex to that of far-field excitation with no tip. Polarization dependence (a polarization parallel to the tip axis couples better to the tip near-field) would be also useful to distinguish the two possible excitation mechanisms.

Thank you for the nice presentation. Regarding polaron movie: What is the time scale of polaron change. Can you go below 3 ps?

We are in the Auger recombination regime with high pump fluence - this is the cause of the relatively fast decay of the polaron signal occurring less than 100 ps. The formation of polaron is believed to occur on a fast time scale less than ~100 fs - resolving such a fast process would require laser pulse much shorter than the ones we employ in this study.

What type of broadband laser source is being used and what is the pulse width and intensity?

We have used a variety of broadband light sources. Thermal, difference frequency generation on the output from OPOs and OPAs, and synchrotron light sources. Bandwidth varies from < 20 cm-1 up to > 2200 cm-1 with pulses as short as < 100 fs.