Ten years ago, the idea to directly excite surface plasmons* using a Scanning Tunneling Microscope (STM) was floating around. The first to realize it for propagating surface plasmons were Bharadwaj, Bouhelier and Novotny in 2011 with a proof-of-principle experiment. We wanted to use this technique to study local surface plasmons and built two specific STMs for doing so.

* A plasmon is the combined oscillation of electrons in metals**. These oscillations change in their nature close to a surface and can couple to light fields. Therefore, they got a special name and are called surface plasmons and are also playing a part as localized surface plasmons in our nano-antennas.
** Or to be more precise: a coherent oscillation of free electrons.

The concept of the project was simple: One fabricates gold nanostructures on a conducting substrate and uses a STM tip to scan over it. When applying voltages around 2 V, localized surface plasmons should be excited and emit light in the visible regime. To collect and analyze it, the whole setup needs to be placed on an inverted optical microscope.

The requirements on the setup were not quite so simple:

• sub Å resolution in z direction
• good I-V conversion + electronics
• shielding from el.-magn. radiation
• mechanical stability
Δz > 100nm
• particle positioning + scanning
• thermal stability
• transparent substrate,
• thickness < 150µm
• positioning of tip into focus
• high resolution / SPPs observation
→ oil objective
Plasmon Excitation
• clean Au films / particles
• sharp tips (preferably of Au)
• small water layer
• stable current

Fortunately, there was a lot of STM know-how in my group in Dresden (esp. by Stefan Grafström + Lukas Eng), the right electronics was available and with the Zeiss Axiovert 200 as well as its piezo stage we had a good base for the STM head.

Unfortunately, we faced some quite serious stability issues: The first one had its origin in the long mechanical loop of the Axiovert and always set in when we used the immersion oil objective. The second electrical instability was due to the IV converter sitting relatively far away from the tip due to the needed long piezo tube.

In Würzburg we took a new attempt and built a more stable version with the IV converter sitting right next to the tip directly on the piezo. Furthermore, we place the STM head on a home-built inverted setup with a much tighter mechanical loop. However, we faced a more fundamental problem: When applying small voltages below 500 mV, that are typically used in STMs, everything worked like expected, but for the needed higher voltages electro-chemical interaction with the surface water layer, which is inherent in ambient conditions, set in and made it unstable.

As by the time others, especially the Dujardin group, already published some parts we wanted to cover, and our electro-optical nano-antennas were more promising, we focused more on that topic.