a novel avalanche-free single photon detector - (BISOL), Northwestern ...
TuA2.1 (Invited) 11:00 am – 11:30 am
A NOVEL AVALANCHE-FREE SINGLE PHOTON DETECTOR H. Mohseni (Senior Member), O.G. Memis (Member), and S.C. Kong Northwestern University, Department of Electrical Engineering and Computer Science 2145 Sheridan Road, Evanston, IL 60208-3118 USA Email: [email protected] Abstract A novel single photon infrared detector is presented that is capable of operating at room temperature in principle. Unlike avalanche, this method produces no excess noise, and can potentially cover wavelengths from UV to mid-infrared. produce a significant crosstalk in an arrayed SPD5. In parallel, multiplication of both carriers in avalanche process leads to a high excess noise, especially for compound semiconductors, where the ionization rate of electrons and holes are similar. Finally, the super-exponential relation between the current and voltage of an APD device near breakdown has hindered realization of large APD arrays with a high uniformity. Here we present a novel avalanche-free single photon detector based on carrier focalization and augmentation. The principle of operation of the device is shown in Fig.1 schematically.
I. Background Single photon detectors (SPD) provide the ultimate limit of such detection in term of sensitivity, and are considered the enabling devices in a wide range of medical, commercial, defense, and research applications. Unfortunately, the existing SPDs have important drawbacks that limit the performance of many systems in the above fields. In particular, these drawbacks have prevented demonstration of the much-needed high-performance SPDs beyond the visible wavelength, and high-performance arrays of SPDs. Many applications can drastically benefit from twodimensional imaging arrays of SPD, and SPDs that can operate in the longer wavelengths. Some examples are infrared tomography1, quantum imaging2, infrared nondestructive inspection3, and remote detection of buried landmines4. Although most of the works on compact solid-state SPDs have been focused on Geiger mode avalanche photodetectors (APD), we believe that there are inherent drawbacks in the avalanche process that hinder realization of short and midinfrared SPDs, as well as high-performance 2D SPD arrays. These inherent physical limitations are resulted from two major avalanche process properties: highly energetic carriers are required, and both carriers are ionized. High electric field required for high energy carriers lead to a high tunneling rate and high dark current. Also, energetic carriers produce a lot of trapped charges and severely increase the after-pulsing. The energetic (hot) carriers that are required for avalanche process can also produce photons. In fact, avalanche detectors are known to produce “photon flashes” that are three to four orders of magnitude brighter than the incoming beam. The produced photons can severely interfere with the other components of the system in a single element SPD, and
0-7803-9558-1/06/$20.00 ©2006 IEEE
Carriers
Photons
(a)
(b) Figure 1. The schematics of the novel device, showing (a) lateral focalization and (b) vertical transport.
When an electron-hole pair is generated through optical excitation, the nano-injector attracts the hole and traps it in a very small volume. This creates a high concentration that reduces the potential barrier and leads to the injection of many electrons to the absorption layer. The structure features short
163
boundaries in the form of ohmic or Schottky contacts (i.e. Dirichlet boundary conditions) imposed infinite recombination and created stress on the domains they are attached to. The heterostructure of the device resulted in spikes in the carrier concentrations, which made current continuity harder to sustain. The most effective method to reduce these effects was to build up the simulation in steps. Starting with a less constrained model (i.e. without nonlinear effects or forced boundaries), then using the solution as initial condition turned out to have better convergence.
transit time for electrons and high lifetime for holes in the nano-injector, which accounts for the high gain. Furthermore, unlike quantum dot based detectors, the thick absorption layer and efficient collection of holes results in high quantum efficiency. To improve carrier injection, the device is utilizing type-II band alignment in the lattice matched InP/GaAsSb/InGaAs material system. Hot electron injection is easily achieved in this system6, and is used to enhance the device gain. Unlike other low dimensional detectors, this device contains a large absorption medium to increase quantum efficiency. Thus, the base (GaAsSb) and emitter (InP) regions are different in radius than the collector (InGaAs), which is optimized to be the main absorption medium. Such a large collector region, designed for a high photon-electron interaction, can also route the optically generated holes towards the nano-injector for a high amplification. The small volume of the nano-injector produces an ultra-low capacitance that leads to a large change of potential even with entrapment of a single hole. The change of potential reduces the barrier between the emitter and collector, and produces a large electron injection. Our simulation results show a gain of ~104 is easily achievable. The structure is also designed to reduce recombination of injected electrons with the photo-generated holes by separating the electron and hole current paths.
Other than convergence issues, accurate modeling of surfaces needed special attention: The nano-injector region was extremely sensitive to any surface effect due to its high surface-to-volume ratio. In addition to this sensitivity, finite element method had problems with boundaries: forced boundaries sometimes created leaks in nearby neutral boundaries (i.e. Neumann boundary condition) and current integration along forced or internal boundaries turned out inaccurate, both of which needed improvement and was eventually addressed.
III. Results Simulation results show a high gain and low dark current, which correspond to a high SNR in this device. The current components within the device are also as expected: the electron injection is mainly under the nano-injector, and hence recombination with photo-generated holes is minimized (Fig. 3). This will increase the internal quantum efficiency of the device by preventing recombination of the holes prior to their arrival to the nano-injector. Current-voltage characteristics of a 5 µm wide device at different optical powers are shown in Fig. 4. The device is capable of detecting very weak optical powers even at the thermoelectric (TE) accessible temperature of T=250 ºK, while demonstrating a respectable internal gain of above 4.5x103 with a ~0.5 volt bias. The current-voltage characteristic of the device shows a semilinear relation. This feature, combined with the sub-volt operation, makes this device very attractive for large imaging array applications.
II. Modeling A simulation tool was created using Comsol Multiphysics, which uses Finite Element Method (FEM) to simulate the physical environment with a given set of equations. The model incorporates the Poisson’s equation along with the diffusiondrift equations using appropriate boundary conditions. To improve performance and lower the complexity, the simulation of a radial cross-section is performed instead of the whole 3dimensional volume (Fig.2). The model incorporates several nonlinear effects such as incomplete ionization, surface recombination, impact ionization, nonlinear mobility, and hot electron effects.
Figure 2. The radial cross section used in FEM simulation (left) and the energy band structure of the modeled device (right). Figure 3. Simulated electron injection density (arrow) and electrostatic potential (color) of the device at T=300 K and under a bias value of ~0.5 volts.
Stability of the model was a problem, as the device possesses a highly nonlinear dynamics. Consequently, there were convergence issues which needed to be addressed: forced
164
IV.
-11
2.5
x 10
T=250K D=5 µm d=100nm
Current (A)
2 1.5
We have designed and simulated a novel single photon detector. It incorporates a nano-injector region on top of a large absorption layer. The device has several features to reduce the noise level while maintaining a high gain. Most importantly, the device is not based on energetic (hot) carriers and it produces gain for only one of the carriers (holes here). Furthermore, we created a three-dimensional non-linear simulation tool for accurate modeling of the proposed detector. The model confirms the predicted properties of the device; namely low dark current, high gain at a sub-volt bias, and a semi-linear current-voltage relation. Also, the model enables us to optimize the device geometry, composition, and doping. The optimized epitaxial layer structure and device geometry, provided by the simulation tool, were used for epitaxial growth and processing mask design. Fabrication of the devices is currently underway, and the evaluation of the detector will be done in the near future.
Popt=1 fW
λ=1700nm
1 0.5
Popt=0
0 -0.5
0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Bias (Volt)
1
Figure 4. Current-voltage characteristics of a 5 µm wide device at T=250 K and different optical powers. The device shows an internal gain of ~4000.
The current-voltage characteristic of the device shows a semilinear relation. This feature, combined with the sub-volt operation, makes this device very attractive for large imaging array applications.
V. Acknowledgement This work has been partially supported by DARPA grant number HR0011-05-1-0058.
A high dynamic range is another important feature of this device. Fig. 5 shows the gain of the device versus the optical power density. The device shows very small gain reduction over more than 50 dB change of optical power.
References 1
Y. Shao, R. W. Silverman, R. Farrell, L. Cirignano, R. Grazioso, K. S. Shah, G. Visser, M. Clajus, T. O. Tummer, and S. R. Cherry, “Design studies of a high resolution PET detector using APD arrays,” IEEE Trans. Nucl. Sci., Vol. 47, p. 1051, 2000. 2 A. F. Abouraddy, B. E. A. Saleh, A.V. Sergienko, and M.C. Teich, “Role of Entanglement in Two-Photon Imaging,” Phys. Rev. Lett., Vol. 87, p. 123602, 2001. 3 M.R. Clark, D.M. McCann, M.C. Forde, “Application of infrared thermography to the non-destructive testing of concrete and masonry bridges,” NDT & E International, Vol. 36, p. 265, 2003. 4 B.H.P Maathuis and J.L Van Genderen,” A review of satellite and airborne sensors for remote sensing based detection of minefields and landmines,” International J. of Remote Sensing, Vol. 25, p. 5201, 2004. 5 C. Kurtsiefer; P. Zarda; S. Mayer; and H. Weinfurter “The breakdown flash of silicon avalanche photodiodes--back door for eavesdropper attacks?,” J. of Modern Optics, Vol. 48, p. 2039, 2001. 6 J. Hu, X. G. Xu, J. A. H. Stotz, S. P. Watkins, A. E. Curzon, M. L. W. Thewalt, N. Matine, C. R. Bolognesi, “Type II photoluminescence and conduction band offsets of GaAsSb/InGaAs and GaAsSb/InP heterostructures grown by metalorganic vapor phase epitaxy”, Appl. Phys. Lett. Vol. 73, p. 2799, 1998.
7000 6500
Vbias=0.45 V
6000 5500
Gain
Conclusion
5000 4500 4000 3500 3000 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Optical Power (W/m2)
Figure 5. Device gain versus the optical power density at a bias of 0.45 volts.
The simulation results show that a slight increase of the radius of the nano-injector is quite effective in further separation of electron and hole currents. The simulation also confirms that the routing of the carriers towards the narrow nano-injector region reduces the radial diffusion length compared to the axial diffusion length. This reduction was expected due to the congestion of carriers around the center of the device.
165
A NOVEL AVALANCHE-FREE SINGLE PHOTON DETECTOR H. Mohseni (Senior Member), O.G. Memis (Member), and S.C. Kong Northwestern University, Department of Electrical Engineering and Computer Science 2145 Sheridan Road, Evanston, IL 60208-3118 USA Email: [email protected] Abstract A novel single photon infrared detector is presented that is capable of operating at room temperature in principle. Unlike avalanche, this method produces no excess noise, and can potentially cover wavelengths from UV to mid-infrared. produce a significant crosstalk in an arrayed SPD5. In parallel, multiplication of both carriers in avalanche process leads to a high excess noise, especially for compound semiconductors, where the ionization rate of electrons and holes are similar. Finally, the super-exponential relation between the current and voltage of an APD device near breakdown has hindered realization of large APD arrays with a high uniformity. Here we present a novel avalanche-free single photon detector based on carrier focalization and augmentation. The principle of operation of the device is shown in Fig.1 schematically.
I. Background Single photon detectors (SPD) provide the ultimate limit of such detection in term of sensitivity, and are considered the enabling devices in a wide range of medical, commercial, defense, and research applications. Unfortunately, the existing SPDs have important drawbacks that limit the performance of many systems in the above fields. In particular, these drawbacks have prevented demonstration of the much-needed high-performance SPDs beyond the visible wavelength, and high-performance arrays of SPDs. Many applications can drastically benefit from twodimensional imaging arrays of SPD, and SPDs that can operate in the longer wavelengths. Some examples are infrared tomography1, quantum imaging2, infrared nondestructive inspection3, and remote detection of buried landmines4. Although most of the works on compact solid-state SPDs have been focused on Geiger mode avalanche photodetectors (APD), we believe that there are inherent drawbacks in the avalanche process that hinder realization of short and midinfrared SPDs, as well as high-performance 2D SPD arrays. These inherent physical limitations are resulted from two major avalanche process properties: highly energetic carriers are required, and both carriers are ionized. High electric field required for high energy carriers lead to a high tunneling rate and high dark current. Also, energetic carriers produce a lot of trapped charges and severely increase the after-pulsing. The energetic (hot) carriers that are required for avalanche process can also produce photons. In fact, avalanche detectors are known to produce “photon flashes” that are three to four orders of magnitude brighter than the incoming beam. The produced photons can severely interfere with the other components of the system in a single element SPD, and
0-7803-9558-1/06/$20.00 ©2006 IEEE
Carriers
Photons
(a)
(b) Figure 1. The schematics of the novel device, showing (a) lateral focalization and (b) vertical transport.
When an electron-hole pair is generated through optical excitation, the nano-injector attracts the hole and traps it in a very small volume. This creates a high concentration that reduces the potential barrier and leads to the injection of many electrons to the absorption layer. The structure features short
163
boundaries in the form of ohmic or Schottky contacts (i.e. Dirichlet boundary conditions) imposed infinite recombination and created stress on the domains they are attached to. The heterostructure of the device resulted in spikes in the carrier concentrations, which made current continuity harder to sustain. The most effective method to reduce these effects was to build up the simulation in steps. Starting with a less constrained model (i.e. without nonlinear effects or forced boundaries), then using the solution as initial condition turned out to have better convergence.
transit time for electrons and high lifetime for holes in the nano-injector, which accounts for the high gain. Furthermore, unlike quantum dot based detectors, the thick absorption layer and efficient collection of holes results in high quantum efficiency. To improve carrier injection, the device is utilizing type-II band alignment in the lattice matched InP/GaAsSb/InGaAs material system. Hot electron injection is easily achieved in this system6, and is used to enhance the device gain. Unlike other low dimensional detectors, this device contains a large absorption medium to increase quantum efficiency. Thus, the base (GaAsSb) and emitter (InP) regions are different in radius than the collector (InGaAs), which is optimized to be the main absorption medium. Such a large collector region, designed for a high photon-electron interaction, can also route the optically generated holes towards the nano-injector for a high amplification. The small volume of the nano-injector produces an ultra-low capacitance that leads to a large change of potential even with entrapment of a single hole. The change of potential reduces the barrier between the emitter and collector, and produces a large electron injection. Our simulation results show a gain of ~104 is easily achievable. The structure is also designed to reduce recombination of injected electrons with the photo-generated holes by separating the electron and hole current paths.
Other than convergence issues, accurate modeling of surfaces needed special attention: The nano-injector region was extremely sensitive to any surface effect due to its high surface-to-volume ratio. In addition to this sensitivity, finite element method had problems with boundaries: forced boundaries sometimes created leaks in nearby neutral boundaries (i.e. Neumann boundary condition) and current integration along forced or internal boundaries turned out inaccurate, both of which needed improvement and was eventually addressed.
III. Results Simulation results show a high gain and low dark current, which correspond to a high SNR in this device. The current components within the device are also as expected: the electron injection is mainly under the nano-injector, and hence recombination with photo-generated holes is minimized (Fig. 3). This will increase the internal quantum efficiency of the device by preventing recombination of the holes prior to their arrival to the nano-injector. Current-voltage characteristics of a 5 µm wide device at different optical powers are shown in Fig. 4. The device is capable of detecting very weak optical powers even at the thermoelectric (TE) accessible temperature of T=250 ºK, while demonstrating a respectable internal gain of above 4.5x103 with a ~0.5 volt bias. The current-voltage characteristic of the device shows a semilinear relation. This feature, combined with the sub-volt operation, makes this device very attractive for large imaging array applications.
II. Modeling A simulation tool was created using Comsol Multiphysics, which uses Finite Element Method (FEM) to simulate the physical environment with a given set of equations. The model incorporates the Poisson’s equation along with the diffusiondrift equations using appropriate boundary conditions. To improve performance and lower the complexity, the simulation of a radial cross-section is performed instead of the whole 3dimensional volume (Fig.2). The model incorporates several nonlinear effects such as incomplete ionization, surface recombination, impact ionization, nonlinear mobility, and hot electron effects.
Figure 2. The radial cross section used in FEM simulation (left) and the energy band structure of the modeled device (right). Figure 3. Simulated electron injection density (arrow) and electrostatic potential (color) of the device at T=300 K and under a bias value of ~0.5 volts.
Stability of the model was a problem, as the device possesses a highly nonlinear dynamics. Consequently, there were convergence issues which needed to be addressed: forced
164
IV.
-11
2.5
x 10
T=250K D=5 µm d=100nm
Current (A)
2 1.5
We have designed and simulated a novel single photon detector. It incorporates a nano-injector region on top of a large absorption layer. The device has several features to reduce the noise level while maintaining a high gain. Most importantly, the device is not based on energetic (hot) carriers and it produces gain for only one of the carriers (holes here). Furthermore, we created a three-dimensional non-linear simulation tool for accurate modeling of the proposed detector. The model confirms the predicted properties of the device; namely low dark current, high gain at a sub-volt bias, and a semi-linear current-voltage relation. Also, the model enables us to optimize the device geometry, composition, and doping. The optimized epitaxial layer structure and device geometry, provided by the simulation tool, were used for epitaxial growth and processing mask design. Fabrication of the devices is currently underway, and the evaluation of the detector will be done in the near future.
Popt=1 fW
λ=1700nm
1 0.5
Popt=0
0 -0.5
0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Bias (Volt)
1
Figure 4. Current-voltage characteristics of a 5 µm wide device at T=250 K and different optical powers. The device shows an internal gain of ~4000.
The current-voltage characteristic of the device shows a semilinear relation. This feature, combined with the sub-volt operation, makes this device very attractive for large imaging array applications.
V. Acknowledgement This work has been partially supported by DARPA grant number HR0011-05-1-0058.
A high dynamic range is another important feature of this device. Fig. 5 shows the gain of the device versus the optical power density. The device shows very small gain reduction over more than 50 dB change of optical power.
References 1
Y. Shao, R. W. Silverman, R. Farrell, L. Cirignano, R. Grazioso, K. S. Shah, G. Visser, M. Clajus, T. O. Tummer, and S. R. Cherry, “Design studies of a high resolution PET detector using APD arrays,” IEEE Trans. Nucl. Sci., Vol. 47, p. 1051, 2000. 2 A. F. Abouraddy, B. E. A. Saleh, A.V. Sergienko, and M.C. Teich, “Role of Entanglement in Two-Photon Imaging,” Phys. Rev. Lett., Vol. 87, p. 123602, 2001. 3 M.R. Clark, D.M. McCann, M.C. Forde, “Application of infrared thermography to the non-destructive testing of concrete and masonry bridges,” NDT & E International, Vol. 36, p. 265, 2003. 4 B.H.P Maathuis and J.L Van Genderen,” A review of satellite and airborne sensors for remote sensing based detection of minefields and landmines,” International J. of Remote Sensing, Vol. 25, p. 5201, 2004. 5 C. Kurtsiefer; P. Zarda; S. Mayer; and H. Weinfurter “The breakdown flash of silicon avalanche photodiodes--back door for eavesdropper attacks?,” J. of Modern Optics, Vol. 48, p. 2039, 2001. 6 J. Hu, X. G. Xu, J. A. H. Stotz, S. P. Watkins, A. E. Curzon, M. L. W. Thewalt, N. Matine, C. R. Bolognesi, “Type II photoluminescence and conduction band offsets of GaAsSb/InGaAs and GaAsSb/InP heterostructures grown by metalorganic vapor phase epitaxy”, Appl. Phys. Lett. Vol. 73, p. 2799, 1998.
7000 6500
Vbias=0.45 V
6000 5500
Gain
Conclusion
5000 4500 4000 3500 3000 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Optical Power (W/m2)
Figure 5. Device gain versus the optical power density at a bias of 0.45 volts.
The simulation results show that a slight increase of the radius of the nano-injector is quite effective in further separation of electron and hole currents. The simulation also confirms that the routing of the carriers towards the narrow nano-injector region reduces the radial diffusion length compared to the axial diffusion length. This reduction was expected due to the congestion of carriers around the center of the device.
165
