There are a number of important OEM photonic applications based on the detection of low incident light levels, including flow cytometry, PET (positron emission tomography) imaging, and applied spectroscopy. In flow cytometry, a blood sample is first treated with fluorescently labeled antibodies, which preferentially adhere to different cell types. As the cells flow through the instrument in a single-file fashion, a focused laser beam excites fluorescence. The various antibody labels each have emission peaks at different points in the visible spectrum, which are separated by dichroic optical filters and detected with photodetectors. This allows several different cell types to be simultaneously counted. In this brief a novel cooled, solid-state photodetector designed to detect low light levels in flow cytometry and other applications is described.
The photomultiplier tube (PMT) has long been used as the detector of choice in this application. However, this detector has a number of limitations. For example, the quantum efficiency (QE) of available PMTs is only 1-10% at 670 nm, which limits the signal-to-noise ratio of the instrument and its ability to accurately distinguish fluorescent cells with the new probes.
Silicon photodiodes, on the other hand, can have high quantum efficiencies (>80%) throughout the visible and near-IR. But with no internal gain, the silicon PIN photodiode is ill-suited to detecting low photon fluxes.
The Large-Area Avalanche Photo-diode (LAAPD) combines the advantages of both detector types. Originally developed by Advanced Photonix engineers for high-energy physics applications, the LAAPD offers the ideal combination of extended spectral range and high QE (see Figure 1) with substantial internal gain. In an avalanche photodiode, an incoming photon creates an electron-hole pair. The combination of low-defect transmutation-doped silicon and a patented beveled-edge architecture allows a large-area device to be fabricated that can withstand applied voltages up to 2 kV without breakdown. This large reverse field causes electrons to accelerate through the doped silicon toward the device's cathode, producing an avalanche of electrons by collisional ionization. Each initial photoelectron typically results in several hundred electrons reaching the cathode. In theory, an internal gain factor of several hundred is more than sufficient to support flow cytometry and other traditional PMT applications. And with active areas as large as 16 mm in diameter, the LAAPD can replace the PMT in instruments such as flow cytometers without requiring any optical redesign.
There are several major advantages to replacing vacuum-tube technology with a solid-state alternative. Compared to the glass-enclosed PMT, the LAAPD is much smaller and orders of magnitude more rugged. The LAAPD also offers a much greater dynamic range, with a linear response over 106. In contrast the dynamic range of the typical PMT is limited to 104, principally because of charge-cloud build-up at the anode. Continuous current draw by the voltage divider circuit is another disadvantage of the PMT in instruments, such as PET scanners, that use a number of these devices. This current is eventually converted to resistive heat. Conversely, the only current flowing through an LAAPD circuit is the photocurrent - the signal.
High QE is the biggest advantage of the LAAPD in flow cytometry. In this application, a gain of several hundred is sufficient to amplify typical signals above the noise floor of subsequent electronics. Once this noise floor has been exceeded, a more important consideration is the probability of converting an incident photon into a detected charge carrier--the quantum efficiency. Simply stated, if a photon doesn't produce a photoelectron, how much that photoelectron is amplified is irrelevant.
In a typical PMT, only 10-25% of incident photons generate photoelectrons at the photocathode, and even this efficiency is achieved over only a narrow spectral range. But in a LAAPD, the QE can be as high as 90% in the visible light spectrum (see Figure 1).
The other important consideration in low-light applications such as flow cytometry is noise from the detector itself. Both PMTs and LAAPDs benefit from active cooling. In the case of the LAAPD, lowering the temperature reduces the dark current (and hence dark noise) due to thermally generated charge carriers, and also increases the gain for a given bias. But unlike the bulky PMT, the compact LAAPD can be cooled by integration of just a tiny thermoelectric cooler. Figure 2 shows the final format that has been developed for the flow cytometry application: both the LAAPD and its cooler are completely contained in a modified TO can.
To summarize, the photomultiplier had long been the only detector choice for low-light applications. As we learn how to tailor the solid-state LAAPD alternative to important applications, Advanced Photonix expects it to displace the PMT, just as surely as the transistor replaced the vacuum tube in the world of electronics.