Deep-UV excitation with oblique epi-illumination

For the last several years, I’ve been working on a project to make spectrally-encoded beads using luminescent lanthanide nanophosphors [1] [2]. We use the nanophosphors to make unique spectral fingerprints for different beads by varying the concentration of different lanthanide emitters with distinct emission spectra. In particular, the nanophosphors we use are ytrrium vanadate nanocrystals doped with lanthanide emitters such as europium or dysprosium. We use lanthanide nanophosphors, rather than other fluorophores because they have narrow emission lines, are photostable, and are chemically stable. However, they have one major drawback: their excitation maximum is at 280 nm. This wavelength is so short that it is not transmitted by glass or conventional optics; instead you must use fused silica or special plastics like cyclic olefin (co)polymer to get substantial transmission. This means that conventional epi-illumination (through the objective) cannot be used to excite our samples. While there are objectives optimized for transmission of such short wavelengths, they are very expensive. Instead, for our work to date, we have used transmitted light illumination to excite our samples. However, this is relatively low brightness, illuminates a small field of view, and uses an expensive arc lamp source.

The deep-UV illuminator mounted on a 4x / 0.2 NA objective.

To try and improve on this light source, I designed an epi-illuminator for 280 nm illumination of our samples. Rather than illuminating through the objective, it consists of six deep-UV LEDs aimed at the sample. The LEDs are 280 nm Optan LEDs from Crystal IS with a ball lens to produce a narrow beam of light. Each emits ~ 3-4 mW of light, for a total of ~ 24 mW at the sample. They are mounted in a 3D-printed mount designed to aim each LED at the focal point of the lens. Clean up filters are mounted in front of each LED. These are 300 / 80 nm bandpass filters from Semrock, custom cut to 9mm diameter.

Autocad Inventor sketch of a cross-section through the LED mount. Revolving this forms the base of the mount.

Designing the LED mount was an interesting challenge. I needed to mount the LEDs so that they would be aimed at the focal point of the objective, while ensuring that they cleared the objective and allowed sufficient clearance above the illuminator to allow for mounting the sample. I learned a lot about parametric modeling in Autocad Inventor while designing it. You can see the sketch that forms the base of the mount to the right. The constraints allow adjusting the clearances while ensuring that the LEDs are aimed at the focal point.

The other challenge in designing this illuminator was that the LEDs are not very efficient. They produce ~ 4 mW of light but require 100 mW of electrical power. This means that the other 96 mW of power are dissipated as heat. It is recommended to heat sink the LEDs by mounting them to a metal-core circuit board. However, there is insufficient space to mount a large enough circuit board behind each LED. Instead, the LEDs are water cooled by a stainless tube clamped behind each LED. This is connected to a water pump that recirculates water from a 1 liter reservoir. An Arduino controls both the LEDs and the water pump; the pump comes on whenever the LEDs are turned on and stays on for 15 seconds after they are turned off to dissipate any residual heat. To switch the 12V supply required for both the LEDs and the water pump, the digital outputs from the Arduino drive MOSFET switches. The Arduino firmware is a modified version of the standard Micro-manager Arduino firmware, allowing the illuminator to be controlled from Micro-manager.

Although we’ve primarily tested this illuminator with UV LEDs, we’ve also tried imaging GFP with a 470 nm LED mounted in the same way. This works well and could be used for building compact fluorescence illumination systems.

A full parts list, drawings for the 3D printed parts, and the Arduino sketch are available on Github.


  1. R.E. Gerver, R. Gómez-Sjöberg, B.C. Baxter, K.S. Thorn, P.M. Fordyce, C.A. Diaz-Botia, B.A. Helms, and J.L. DeRisi, "Programmable microfluidic synthesis of spectrally encoded microspheres", Lab Chip, vol. 12, pp. 4716-4723, 2012.
  2. H.Q. Nguyen, B.C. Baxter, K. Brower, C.A. Diaz-Botia, J.L. DeRisi, P.M. Fordyce, and K.S. Thorn, "Programmable Microfluidic Synthesis of Over One Thousand Uniquely Identifiable Spectral Codes", Advanced Optical Materials, vol. 5, pp. 1600548, 2016.

4 thoughts on “Deep-UV excitation with oblique epi-illumination

    • Not sure; we haven’t compared them. We have used full spectrum Xe arc lamps in the past; the LED system is more powerful but we have more issues with illumination uniformity.

    • Sadly, we mostly used it for some unpublished work, and never quite got it working where we were happy with it before I left. So I’m not sure if any images from it will ever be published.

Comments are closed.