For optical characterizing nano-antennas a lot of “smaller” optical devices are needed besides the necessary microscopes. Here we will cover various types of laser light sources and different cameras/detectors.
The simplest lasers are continuous wave (cw) lasers. Historically they were often gas lasers — think of a fluorescent lamp with highly reflective mirrors at its ends — but nowadays solid-state lasers are the norm. They are more efficient, compact, robust and available in every color you want. Cw lasers are usually specialized to do one task very well, e.g., they have a good mode profile, a high power, a narrow line width or a small dispersion.
We own different cw lasers: Coherent Obis 532LS, OEM 532 nm, Helium-Neon 632 nm, Coherent Obis 785LX as well as some Raman lasers which have a very narrow line width. The specialty of the Coherent Obis Coherent Obis 532LS is its good mode profile but it can also be modulated up to 50 kHz. (In contrast, the mode profile of the 785LX is not as good but it can be modulated up to 150 MHz!). The 532LS is our standard laser for conducting photoluminescence measurements not only of molecules or quantum dots but also of Gold. Yes, Gold shows some fluorescence when hitting it very hard with a laser. We use that for characterizing the resonances of our gold nano-antennas without the need of additional tricks.
Tunable Laser Sources
The next laser — or better laser system — we use is a tunable laser source. The basic principle is that you use a highly intense laser pulse which is sent through a non-linear optical fiber to generate “white light”. This white light can be thought of laser pulse with various wavelength covering the whole visible spectrum. Then a variable filter is used to select the specific wavelength you want.
For doing that we have a NKT Photonics SuperK for generating the white-light pulses and the NKT Photonics Select, which is an acousto-optical filter, to select wavelengths in the 500-900 nm regime. Using that we can scan over the resonance spectrum of an nano-antenna and measure its response.
Finally, when you want to characterize the ultra-fast response of your system to learn about possible non-linear effects, you need femto-second laser pulses. They are usually generate using a Titan-Sapphire laser that needs to be pumped with a high-power laser to get it operating. Further equipment is then needed to characterize the pulses spectrally, in time and also to optimize them.
We own a Coherent Mira tunable Titan-Sapphire laser that is pumped with a Coherent Verdi green laser. The pump provides 10 W which result to an output of 2 W behind the Mira. The pulses a roughly 100 fs long, we characterize them with an OceanOptics USB4000 fiber spectrometer (not shown) and APE Carpe autocorrelator. Afterwards they are routed through a pulse compressor to compensate for the dispersion of optical components, which increase the pulse length, further down the beam line.
Besides light source one also needs cameras and/or detectors for optical experiments.
We have some simple cameras attached to our Nikon microscope to monitor the sample from above (not shown here) or to acquire high-resolution images through the objective. For the latter two options are available: an ancient black-and-white CCD and color CMOS camera from The Imaging Source. The b/w CCD is needed for optimizing the laser focus as the CMOS Chips would not survive such intensities.
For scientific images and spectra an Electron-Multiplying Charge-Coupled Device (EMCCD) is needed. It can measure weak signals down to single-photon level and is calibrated to precisely obtain the amount of light. The trick behind EMCCDs is that they amplify the electrical signal directly on the CCD chip and that they are cooled to -80°C.
Our setup includes a Andor Ixon3 EMCCD which is attached to a Andor Shamrock 303 spectrometer. We exchanged one of the three gratings of the spectrometer with a mirror to allow low-loss imaging. This is particularly important to ensure that acquired spectra really belong to objects we want to investigate.
The final detectors in our list are Avalanche PhotoDiodes (APDs). Like EMCCDs they can detect single photons but much faster. Their temporal accuracy is in the 100-picosecond range while their dead time is ~25 ns. That makes the ideal not only for counting single photons but also for collecting the photon statistics of a quantum light source.
We own PerkinElmer (formerly EG&G, now Excelitas) SPCM-AQR single photon counting APDs and typically set them up in pairs in a Hanbury-Brown-Twiss arrangement. This allows us to easily measure the photon statistics of a nano-light source.