Nils Otterstrom and Dallin Durfee, Department of Physics and Astronomy
Introduction
Physics laboratories all over the world depend on accurate wavelength meters to tune their lasers to desired optical frequencies. Our ion interferometry lab, for example, relies heavily on the precision of these instruments to laser cool beams of strontium ions and then split and recombine their wave functions. Unfortunately, some these devices can be extremely expensive and very cumbersome to use. A frequency comb wavelength detector, although remarkably accurate, can cost around $100,000 dollars. In our lab we employ a Michelson-Morley interferometer based wavelength meter, which costs around $8,000 dollars. Despite its relatively high accuracy, the device is extremely fragile.
However, the almost overwhelming cost and need for special care could very well be eliminated. I have worked to develop an inexpensive and accurate wavelength meter using a commercially available array of filtered photodiodes. This method of determining wavelength could allow extremely robust and relatively accurate meters to be manufactured for less than $30.
Methodology
We employ a high precision color sensor (TCS 3414) that is commercially mass-produced for HDTVs, cellular devices, and computers. The sensor’s original purpose is to determine the general color of the ambient light in order to adjust the color display of the device. However, this color sensor was not designed for measuring the exact wavelength of single frequency light.
The TCS 3414 is a 3-dollar electro-optical chip that contains an array of filtered photodiodes with an impressive 16-bit digitizer. First, we must understand how a photodiode works before exploring its application in laser spectroscopy. A photodiode is a junction of two semiconductive substances that converts light into and electrical current. Using an array of filtered photodiodes we can accurately determine the wavelength of the incident light. For example, if a red laser were incident upon the sensor, more light would be transmitted through the red filters. Hence, by comparing ratios among the four types of filtered photodiodes we can precisely determine the wavelength using and algorithm I work to develop. (Figure 1)
We soon observed a sinusoidal response with respect to both wavelength and temperature. This suggested that we had a small etalon or thin film interference. We found a 0.4 mm film of glass superimposed upon the filters that confirmed our assumption. To mitigate the sinusoidal effect from the etalon, we roughened the surface of the glass in order to randomize the reflection angles and macroscopically eliminate spatially coherent reflections.
Results
However, during this last year, we found that we could actually use the etalon to our advantage. As the etalon introduces relatively large features in the wavelength dependence, we could use these features to attain even greater precision. In conjunction with precise temperature control, we could use the steep curves introduced by the etalon in order to achieve greater sensitivity over a short wavelength range. Over 9 hours of continuous data gathering we have achieved precision of less than 0.002 nm, several times better than that of a $1,000 diffraction grating spectrometer! (Figure 2)
This year, I spent a considerable amount of time designing a surface mount circuit board and programming an external microprocessor to simultaneously collect data from four sensors and synthesize a digital to analog signal in order to control the frequency input of our 461 nm laser. I placed two FN variations and two CS variation of the color sensor on the board and then sanded half of them. 10 days ago we locked our 461 nm laser to an atomic transition of strontium and then took data on all four sensors for a total of 10 days. This data will be critical to assess the effects of package type and roughening the surface on our method’s precision.
Discussion and Conclusion
With recourse to a 3-dollar chip, external microprocessor, and powerful algorithm we can make very precise measurements of wavelength. This research has exciting applications in affordable laser spectroscopy as well as single -iber optical communications using multiple wavelengths. With the funding from the ORCA grant I presented our results at two APS meetings. At the most recent meeting, this research attracted the interest of Dr. John Hall, a world-class spectroscopist and Nobel Laureate.