Hyperspectral fluorescence microscopy requires light sources with a number of excitation wavelength bands from deep UV to visible range to excite native fluorophores of cells and tissues. Achieving adequate optical power and wavelength range in the UV is difficult because the excitation light must be channelled to the sample through the microscope which results in power loss.
We report the design of a light source for fluorescence microscopy which can combine the optical power available from multiple LEDs with the same wavelength range. The light source has the shape of a truncated cone with the exit aperture size of 9 mm which is equivalent to the aperture of a conventional mercury lamp used for fluorescence excitation. We used aluminium reflective coating for all inner surfaces of the cone with 92% reflectivity. We analysed its performance by using a ray tracing software; the efficiency of this light source found to be optimised for the diameter to length ratio of unity, and it was the highest for the smallest size (25mm). Variations of the cone’s efficiency with the positions of LEDs, inter-LED distance, and the angle of LED were determined. The efficiency of the cone depends upon the inner surface area of cone’s lid and slanting sides, and decreases with increasing area. The optimised source has the efficiency of around 28% for 60 mm diameter and length. This efficient design for multi-LEDs illumination is applicable to hyperspectral microscopy and it can be used with any other fluorescence microscopy as a retrofit.
The main objective of our work was to design a light source which should be capable to collect and illuminate light of LEDs at the smaller aperture of cone (9mm) which could be either coupled with secondary optics of a microscope or utilized independently for hyperspectral studies.
Optimized performance of cone was assessed for different substrates (diffused glass silica, Alumina, Zerodur glass, acrylic plastic) and coating surfaces (white diffused, flat white paint, standard mirror) using a simulation software. The parameters optimized for truncated cone include slanting length and Top Major R (Larger diameter of cone) which were also varied from 10 to 350 mm and 10 to 80 mm respectively. In order to see affect of LED positions on cone efficiency, the positions of LED were varied from central axis to off-axis. Similarly, interLED distance was varied from 2 mm to 6 mm to reckon its effect on the performance of cone.
The optimized Slant length (80 mm) and Top Major R (50 mm) were determined for substrates (glass zerodur or acrylic plastic) and coating surface (standard mirror). The output profile of truncated source was found non uniform, which is a typical presentation of non imaging optics problem. The maximum efficiency of cone has been found for LED at the centre and it was found decreasing as LED moves away from the central axis. Moreover, shorter the interLED distance, better is the performance of cone.
The primary optics of cone shaped light source is capable to lit visible and UV LEDs in practical design. The optimum parameters obtained through simulations could be implemented in the fabrication procedure if the reflectance of source would have been maintained upto finish level of a standard mirror.