System Design of a Visible Emission and Near-Infrared Upconversion Spectrometer (VE-NUS)

System Design of a Visible Emission and Near-Infrared Upconversion Spectrometer (VE-NUS) to Implement a Silicon Detector Array in Place of Indium Gallium Arsenide Detector Array

Sean Crystal and Jason Owens

,

CREOL, College of Optics and Photonics, University of Central Florida, Orlando, Florida, 32816-2700

Abstract — The purpose of this project is to design and implement a spectrometer system that detects and measures infrared light with a silicon based detector array to create a spectrum. This can be achieved by upconverting the infrared light to visible light. The device is expected to lower the cost of infrared spectrometers while still being able to resolve infrared wavelengths ranging from 900 nm – 1500 nm. The spectrometer is partially a proof of concept design that can further be improved upon by implementing other materials which upconvert longer infrared wavelengths to visible light.

Index Terms — Fluorescence, image resolution, infrared spectra, phosphors, silicon devices.

1. Introduction

Infrared spectroscopy is an ever growing field with applications ranging from research interests in organic and inorganic chemistry to everyday uses in the food industry and forensics. Infrared spectroscopy is usually done with indium gallium arsenide (InGaAs) detector arrays since it has a bandgap corresponding to the infrared region (1000 nm – 1700 nm). Silicon detector arrays are cheaper and more readily available in comparison to InGaAs detector arrays. Silicon detectors, however, have a bandgap that corresponds mostly to the visible region (400 nm – 1200 nm). Engineering the infrared light such that it is detectable by a silicon detector array is thus an appropriategoal that we will show can be accomplished with our method. Being able to do so will lower the cost of infrared spectrometers, a feat desirable not only for sellers of this product but consumers as well.

1. System Definition

A. Device Usage

The device we are developing will be used for any transmission, absorption spectroscopy. It will be based on the basic transmission and absorption techniques.

Our device will be designed for the purpose of characterizing laser spectrums. Lasers can have very narrow simple spectrums all the way to complicated and varying outputs. Knowing this specific output can be very important. The near infrared regime is becoming more and more popular in the laser arena and being able to characterize these new developing lasers is a must.

A key reason we have chosen to use this device for this use is the complication of the upconverting materials. The key to this whole project is the success of our upconverting materials. This will be the most time consuming and scientifically complicated part but by using a more intense and coherent laser source we have more confidence in the success of our phosphors. In the future we do plan of continuing to develop this device to work with other test lights that are not as powerful or coherent.

B. Light Conversion Method

The device will use phosphorescent materials to upconvert the infrared light to visible light. Upconversion is defined by taking a light signal with a lower frequency and converting it up to light signal with a higher frequency, in this case infrared light to visible light. Frequency is proportional to energy so to prevent any loss from occurring when going from lower energy photons to higher energy photons an external pump source is needed. This counteracts any nonlinearity in the upconversion method and acts more like an energy averaging affect such that no energy is lost in our system.

The phosphorescent material is initially depleted; all the electrons in our material will start in the ground state. High energy ultraviolet photons create an electron-hole pair and excite the electrons to a higher energy level. The electrons then decay to a metastable energy level where they are trapped, this effect is known as electron trapping. Infrared photons will excite the electrons within this metastable energy level to a higher energy state, where the electrons will immediately decay to the ground state and emit visible photons. It is imperative that the phosphorescent material can be excited by infrared photons which cover the range of interest.

C. Detection Wavelength Range

The desired wavelength detection range was determined based on multiple factors between our group and our sponsor, Ocean Optics. The factors that were considered were, where do standard Silicon detectors currently cut off, what wavelengths does the commercially available NIR InGaAs detector analyze, what wavelength ranges to laser scientists care about and what wavelengths are we confident our phosphors can work at. With all this being considered the wavelength range of 900nm to 1500nm was set. While this range may cut off slightly below what Ocean Optics current NIRQuest cuts off at, this I a developing product and the range of this system is purely limited by the phosphor technology. With further exploration and research in phosphor up-conversion, such as other materials, temperature dependence or pumping mechanism, this up-converting spectrometer wavelength range can easily increase in the future.

D. Resolution

Resolution of a spectrometer is defined as the number data points collected per wavelength (usually per μm or nm) that can be differentiated from each other for an incoming unknown source of light. The resolution is dependent on a number of factors depending on the type of spectrometer and this will be explained in greater detail in the design section. With this being a prototype of technology, we believe a resolution of 5nm to 40nm is a reasonable benchmark to show the technology works efficiently.

E. Device Size

Size is a major factor in the design of this device. Ocean Optics makes field ready spectrometers that are simple to transport so compact is a must. Ocean Optics currently has a bed for there NIR spectrometer not utilizing phosphors that is 182mm x 110mm x 47mm and we plan to fit our system into this same bed. The original device will be built on optical and electrical breadboards much bigger than this but once the design is complete it will be downsized from optical bench pieces to custom mounts and electrical breadboards to PCB’s.

1. Design

The design of our system will be broken up into two main parts: the front end system and the back end system. The front end system will consist of the entrance pupil, mirrors, and Bragg grating. The back end system will consist of the phosphor coated onto KG glass to absorb any infrared light that might leak through the phosphor, LEDs and pumping circuit, detector, detecting circuit and monitor.

The entrance pupil for our system will be given by a fiber input with a half angle of 17°. The light will be incident upon a curved mirror so that it is collimated. The collimated light will then be incident upon a Bragg grating with a spacing of 300 lines/mm at the blaze angle which is 27.14°. The Bragg grating will separate out the light into its constituent components which will be incident upon another curved mirror to focus the light onto an image plane. Each wavelength of light will be separate out spatially and will be focused to a spot. The ability to differentiate between these spots is our resolution. This process is illustrated in figure 1.

Figure 1: Front end of the system. a) Simulation of front end spectrometer design. A mirror with curvature 50 mm is placed one focal length away from the fiber input end. The infrared light is then collimated and incident upon the Bragg grating at an angle of 27.14°. The infrared light is separated out and sent to a mirror with a curvature of 134 mmwhere it is reflected and focused onto the image plane. b) The point spread of the front end of the system.The different wavelengths are labeled on the right side of the image and it is shown that a resolution of 40 nm is feasible.

From the front end of our system the light travels to the back end of our system where it is then converted to visible light and detected.

We have chosen a phosphor which meets the criteria of our system and is commercially available. This is coated onto the surface of the KG glass and pumped from the sides of the glass by the ultraviolet LEDs. To prevent any leakage of ultraviolet light and thus interference with the system the light will be total internally reflected. To keep the illumination of the phosphor uniform the phosphor will only be coated on the areas of the KG glass that are uniformly illuminated by the ultraviolet light. This is shown in figure 2.

Figure 2: Simulation of KG glass filter. The purpose of the KG glass is to filter out any infrared light that may be transmitted through the phosphor while simultaneously transmitting visible light and total internally reflecting the ultraviolet light. a) KG glass dimensions with angle necessary to total internally reflect the ultraviolet light. b) Top/side view of the filter to show that there is some light that is scattered, however it can easily be blocked so that it will not interfere with the rest of the system. c) Side view of the KG glass to show where the ultraviolet LEDs will be placed and how the light will be total internally reflected. d) Top view of the KG glass to show the uniformity of the ultraviolet light. Approximately 10 mm from both the left and right corner of the glass is where the light becomes more uniform and is where the phosphor will be coated.

The ultraviolet LEDs play an important role in this system. They are necessary to charge the phosphor but they also cause a complication. When the LEDs are on the detector is saturated. We obviously cannot take a spectrum when the detector is saturated. This occurs even with the LEDs light being total internally reflected due to the evanescent wave. While this seems like a problem it is also the fundamental concept that allows us to use total internal reflection to charge the coated phosphor. They scheme being used to allow us to still get a spectrum and avoid saturation is timing the pumping and integration times of the system. The LEDs are turned on for a significant amount of time to charge the phosphor. Then the integration of the detector is started and completed before the decay time of the phosphor is reached.

To accomplish this, a specialized dual timer circuit was constructed. This circuit allows us to pulse the LED’s on at a specific time interval and when they are off a second signal is sent to the detector. A detector that allows for external triggering was necessary. The pumping circuit is shown below in figure 3.

Figure 3: LED pump/trigger circuit. a) A schematic of side one of the circuit. This side of the circuit controls one of the two sets of LEDs. b) A schematic of side two of the circuit. This side of the circuit controls the second of two sets of LEDs and sends the trigger signal to the detector.

The detector used is a silicon array detector. This system has a stationary grating therefor the only way to capture the varying spectral components is with an array vs. single detector. This detector will see the same visible signal across all its pixels. This signal is generated by the phosphor. Because the Silicon detector is sensitive up to 1100 nm, the Kg glass was implemented to prevent this detector from seeing additional signal at the lower wavelengths of this system.

As mentioned previously, this detector receives a trigger signal from the dual timer circuit. This signal is limited to 5.5V by the detector so a 5V voltage regulator was added to the circuit output. When the trigger signal reaches the detector the start of the signal cause the detector to capture an integration period and send it to the screen.

This detector software interface is intended for a visible spectrometer. The spectrum acquired must be converted to the appropriate wavelengths and normalized. The normalization is needed due to the fact that the phosphor has varying sensitivity over the wavelength range specified. Without normalization the spectral intensities could not be considered accurate.

1. Testing

This report includes multiple tests that were run to characterize the phosphor. This report only includes these tests at the 1560nm wavelength, which is the furthest wavelength on our sensitivity curve. These tests run at the other wavelengths should result in better values. These tests are being conducted.

A. Emission and Linearity

To have a reliable curve for our spectrum tests need to indicate that the phosphor responds linearly to the infrared light. The visible response of the phosphor is in fact linear according to figure 4.

Figure 4: Laser light with a wavelength of 1560 nm is incident upon the phosphor material. The corresponding irradiance of both a) visible and b) infrared is measured for a linearly increasing infrared irradiance.

B. Quantum Efficiency

The quantum efficiency is calculated by equation (1).

(1)

(1)

The quantum efficiency is 0.29% for a laser wavelength of 1560 nm.

C. Decay Time

The decay time measurement is defined as the following: 90% of the peak intensity – 10% of the peak intensity. Figure 5 illustrates the data collected to measure and calculate this. The decay time was found to be 2.68 seconds.

Figure 5:Measurement for the decay time at 1560 nm. Given the definition for the decay time stated above (90% peak intensity – 10% peak intensity) the decay time is shown to be 2.68 seconds.

D. Point Spread Function (PSF)

Measurements were taken to quantify the size of the infrared laser beam width being used. This value was found to be 1.027mm.

Next measurements were taken to quantify the visible light point spread from the phosphor to the detector. This value was found to be 1.25mm.

From this it can be concluded that a point will spread by approximately 0.223 mm for a 1560 nm laser source. Assuming this value does not change drastically across the range of wavelengths of interests and remembering the simulated imaging system shown in figure 1b, a resolution of 40 nm for the system is achievable.

1. Conclusion

NIR spectrometers are a highly desirable product. The ability to convert infrared light to visible light is a relatively simple technique we describe as up-conversion. This process can allows us to convert NIR light to visible light and detect it on a visible spectrum detector which is much cheaper than a NIR detector. Normally this process would be a non-linear, however we bypass this process through a phenomenon that occurs naturally in commercially available phosphors called electron trapping. We can do this by pumping our phosphor with a high energy LED that charges the electron to its trapped level and the NIR light cause the phosphor to emit a visible photon. This process is a near linear one that bypasses the complexities and extra variables of a non-linear process.

The system definition, specifications to be met, design and testing of the constructed system were discussed. The biggest constraints that are affected from this process are resolution and conversion efficiency. Work is still being done to show that these engineering challenges can easily be overcome and improved upon. For example, other phosphors exist that can continue farther into the near – infrared range and possibly into the mid – infrared range. Testing has shown good promise under non-ideal conditions.

This system shows promise to deliver a spectral analyzing device for operation in the near-infrared at a significantly reduced cost compared to current systems. The system specifications will continue to improve with further optimization and new ideas.

1. References

[1] "Wikipedia," [Online]. Available: [Accessed July 2016].
[2] E. D. l. Rosaa, L. Diaz-Torresa, P. Salasb and R. Rodriguezc, "Visible light emission under UV and IR excitation of rare earth doped ZrO2 nanophosphor," Optical Materials, vol. 27, no. 7, pp. 1320-1325, 2004.
[3] L. Diaz-Torres, E. D. l. Rosa-Cruz, P. Salas and C. Angeles-Chavez, "Concentration enhanced red upconversion in nanocrystalline ZrO2 : Er under IR excitation.," Journal of Physics, vol. 37, no. 18, 2004.
[4] L. Xie, S. Zou, R. Zhang, J. Cai, Y. Zou, L. Cao and J. Meng, "Fluorescence enhancing under UV-NIR simultaneous-excitation in ZnS:Cu,Mn phosphors," American Institute of Physics Advances, 2012.
[5] Z. Wang and Z. Yan, "Ultraviolet and infrared photon-excited synergistic effect in Er3-doped YbF3 phosphors," Optics Letters, vol. 36, no. 22, pp. 4353-4355, 2011.
[6] A. Alhalawany, S. He, H. Hodaei, N. Johnson, D. N. Christodoulides, A. almutairi, and M. Khajavikhan, "Enhanced UV Upconversion Emission Using Plasmonic Nanocavities," inConference on Lasers and Electro-Optics, OSA Technical Digest (2016) (Optical Society of America, 2016), paper JW2A.93.