Advanced Photocathode Development
Technical Design Report
2009-2010
By
Klaus Attenkofer, James Buckley, Zeke Insepov, Zikri Yusof
Abstract......
Overview......
Background and Significance......
Multi-Alkali Cathodes
III-V Semiconductor Cathodes
III-V Nano-Structured Semiconductor Cathodes
The Scientific Program......
Multi-Alkali Cathodes
III-V Semiconductor Cathodes
III-V-Based Nano-Structured Cathodes
Interface between photocathode and window
Characterization......
Growth related characterization
Structural information of the surface
Chemical composition and electronic properties (Surface and doping profile)
Characterization of time response and energy distribution of electrons
Morphology
Simulation and Theory......
The Organization and Infrastructure......
Collaborations and Argonne Activities......
Importance to Other Communities......
External resources......
Minimum Requirements on Additional Infrastructures......
Combination of Activation/Growth and Characterization: The Cathode Center......
The Required Man-Power......
Scientific staffing:......
Creation and Maintenance of Infrastructure......
The Milestones......
Year one:......
Overall goals:......
Growth......
Characterization......
Theory/Simulation......
Year two: Overall goals:......
Year three: Overall goals:......
Abstract
This technical design report describes the scientific and engineering program of the photocathode development program at Argonne National Laboratory, Washington University, and University of Illinois, Urbana/Champaign. This includes an overview of the needs of the large area picosecond detector program, the proposed approach, and the goals. It also describes the instrumental and human resources which are needed to achieve the goals in a timely manner. Additionally we provide preliminary milestones.
The approach, described in this TDR, is radically different from conventional cathode development. It is based on an understanding of the correlation between microscopic properties of complex heterogeneous materials and the macroscopic behavior. By combining state of the art growth facilities, microscopic and macroscopic characterization tools, abinitio calculations and simulations we hope to develop a rational design concept for photocathodes.
This version will be followed by a final version within 1 month which will also describe the efforts at space sciences laboratory which are mainly focused on the engineering aspects of a large 8”x8” multi-alkali photocathode.
Property / Multi-Alkali / GaAs-based / GaN-basedPhotocathode
Properties / Wavelength response (typical) / 150nm-500nm / 450nm-850nm / 100nm-350nm
Typical efficiency / 20% / 20% / 30-40%
Maximum efficiency / 50% / 60% / 80%
Wavelength tunability / low / large / Very high
Dark current / ~100cps/cm2 / ~10000cps/cm2 / ~100cps/cm2
Growth properties / Single crystal substrate / no / yes / yes
Easy scalable / No / yes / yes
Large production volume possible / No / Yes / Yes
Prefabrication possible / No / Yes / Yes
Temperature sensitive / High / Medium / Medium
Existing Industry / No (besides night vision / small area) / Yes (foundries available) / Yes (foundries available)
Basic Physics / Good understanding / No / Yes / Yes
Microscopic understanding of growth / No / Yes / Yes
2-D Fabrication tools / No / Yes / Yes
3-D Fabrication tools / No / Yes / Some
Theoretical description / No / Yes / Yes
Band-structure engineering / No / Yes / Yes
Table 1:This table tries to provide an overview on the various properties of cathode materials and the possibility to use them in a large production volume application. None of the materials are suited for the LAPPD-project, either because of their cathode behavior or because of issues correlated with industrial production requirement. This TDR describes approaches to overcome the problematic areas for each of the three cathode materials. The III-V materials will allow to produce nano structured materials also described in this document.
Narrative
Overview
Since the discovery of the photo effect over hundred years ago scientists had used photocathodes to convert photons into electrons as the primary step of photon detection. A perfect cathode will show very high absorption in the spectral range of interest, a high electron yield, low thermal noise, low reflection losses, and fast time response. The optimum photocathode has also to meet a wide variety of engineering, process and materials compatibility, and lifetime requirements. To a large degree these properties will determine the demands on the vacuum system, the complexity of the mechanical detector assembly, and therefore the cost efficiency of the full detection system.
Even if much work was already done and many very successful detection systems for various spectral ranges were built, the large area detector project brings new challenges for the photocathode research and production. The relative large area of the individual building blocks, the required low production costs, and the large production volume of the final product will require novel concepts of photocathode production. Especially the large fluctuation of production volume will demand technologies, which widely utilizes coating technologies compatible to industrial foundries with minimal customization of machinery and processes.
The photocathode activity at Argonne National Laboratory will focus on systematic research, which will illuminate the correlation of cathode functionality and materials properties. The goal of these investigations is to develop design rules for complex two- or three-dimensional hetero-structures utilizing fully the potential of novel materials, and state-of-the-art and scaleable growth and fabrication technologies. Under the leadership of Argonne National Laboratory, the University of Illinois Urbana Champaign, the University of Illinois Chicago, Washington University, and the Space Science Laboratory of the University Berkeley will adress the following four challenges imposed by the large area detector project:
- Development of coatings or bonding strategies to allow large flexibility in the choice of the window material. The interlayer will also assure the necessary conductivity to avoid charging effects in a large area photocathode and reduction of reflection losses.
- Provide design concepts to optimize absorption and minimize reflection losses for a given spectral range. (default: 250-400nm)
- Develop strategies to increase photoelectron emission and minimize dark current.
- Provide structures, which are process compatible with the manufacturing of the detector itself.
We will focus on III-V and multi-alkali transmission photocathode systems. This includes major efforts in Sb-Rb-Cs and Sb-K-Cs multi-alkali systems, and GaAs-based, GaN-based cathodes with additional materials, either grown as alloys or as layered systems, to tune the wavelength responsibility.
The optimization of these cathodes is strongly dependent on the geometry in which the device will be used. In contrast to transmission cathodes, e.g. the electron emission will be on the opposite side of the light entrance, opaque cathodes are working in reflection geometry, e.g. the electron emission happens on the same surface where the light enters. In the early stage of this project we will focus on transmission cathodes; depending on the results and progress on nano-structured cathodes we will extent our activity also to opaque cathodes.
The development of nano-structured cathodes will strongly depend on the results of the III-V activities. Most design concepts will be tested in flat geometry; so that the effort can be focused on growth aspects of the nano structures.
Background and Significance
The functionality of photocathodes is often described in athree-step model which distinguishes between the photon absorption (step one), the carrier separation and transport to the surface (step two), and finally the electron emission from the surface (step three). Photon reflection losses, photo absorption, electron hole recombination, and electron emission probability have to be optimized to achieve maximum quantum efficiency for a given photon energy range. At the same time the dark current, e.g. electrons not originated by a photon and emitted from the surface have to be minimized.
The functionality of a photocathode strongly depends on the band gap of the individual layers, their doping profile, and the defect concentration yielding to a recombination of the carrier and therefore a reduction of the quantum efficiency of the device. A dominating role for efficiency and dark current is the near-surface doping profile and defect concentration. Conventional photocathodes are built from constant p-doped semiconductor materials and a strongly n-doped surface yielding to a band bending close to the surface. The work function of the cathode, e.g. the energy which is required to eject an electron from the surface, will be reduced by the amount of the band bending. The absorption and drift area of the cathode is electric- field- free due to the constant p-doping profile.
Both, multi-alkali and III-V semiconductor photocathodes are commonly used and production procedures are known. However, the large area photo detector application brings additional challenges which are mainly correlated with the homogeneity of the relative large area of 8”x8”, the potential large production volume changes, and the quality variation between different batches. Process compatibility requirementof the final assembly, like vacuum or inert gas assembly or maximal process temperature during the sealing of the detection system, will also significantly influence the choice of cathode material.
Motivated by the success of nano-technology in the area of solar cells we will also investigate if nano structures can be employed to create photocathode structures with high yield and low dark current. In the following we will discuss the pros and cons of the three cathode families.
Multi-Alkali Cathodes
In respect to multi-alkali cathodes we will focus on Sb-Rb-Cs and Sb-K-Cs (K2CsSb) systems which are employed in the wavelength range of 300-450nm. They show extraordinary good signal-to-background ratio and a typical quantum efficiency of 20-25%. Newest improvements result in commercially produced photocathodes with quantum efficiencies of 40% and more. Non-commercial cathodes were developed which show efficiency up to 60%. For conventional photocathodes detailed recipes for growth and activation are reported in literature but processing of high efficient cathodes are proprietary. The increase of efficiency is mainly contributed to reduced reflection losses by tuning the window-cathode interface and cathode thickness. Efficiencies larger than 50% can only be achieved if electron is preferentially accelerated towards the emission side for example by creating an internal electrical field.
Even if many of the technological problems are solved and the commercial production of large area cathodes is solved for phototubes there is major research needed to produce a large area cathode for multichannel plate applications. The flat window surface and the small distance between channel plates and window don’t permit the coating within the device like done in photomultiplier tubes. Instead, the cathode has to be evaporated and in some cases activated onto the window material first and finally assembled and sealed with the rest of the detector. To avoid damage of the cathode strong requirements on vacuum and maximal temperature exists which severely restrict the assembly conditions and available processes.
Current production technology will require a piece-by-piece production. The relative large size of the cathode of 8”x8” in combination with high homogeneity will require large vacuum chambers and consequently a long and expensive production cycle. High production volumes would be only possible if many of these chambers would be built. In this case, production volume changes are very hard to handle. Avoiding the need of vacuum assembly or any success to reduce the size of the chambers or allow a line-production will have significant influence on the future costs of the detectors.
III-V Semiconductor Cathodes
GaN and GaAs are the most common III-V semiconductor photo-cathodes. GaN cathodes are mainly used in the energy range of 100nm to 380nm and GaAs cathodes are employed in the range of 350nm-900nm. Both cathodes show typical minimum quantum efficiencies between 20% and 30% in the specified energy range. They are stable and tolerate relative high process temperatures even after activation (up to 400C over multiple hours). New developments of GaAs-based cathodes that focus on optimizing thickness and vertical doping profile of the cathode resulted in significant increase of quantum efficiency up to 60%. Similar increases are expected for GaN based systems. For GaN cathodes the signal to background ratio is excellent due to the large bandgap; dark currents of a few events per second and square cm are reported. However, uncooled GaAs cathodes show typically a much higher dark current. At room temperature dark counts of 15000/sec cm2 are reported. Moderate cooling to 0C will reduce this count rate to about 100/sec cm2.
The production process of the cathodes can be divided into the growth process of the multilayer structure and its doping profile on one hand and the activation process on the other. The growth of the structures is industrial compatible and various foundries are available which can produce the structures on demand. The structure is typically grown on an appropriate single crystal substrate. To produce a semi-transparent cathode the substrate is etched away and the remaining film is bonded to the entrance window of the detector. Various, mostly proprietary, processes exist to transfer and bond the cathode to the glass window. However, these processes are optimized for small volume production and small areas.
The prefabricated structures can be stored in inert atmosphere and will be available at any time when needed. The cathode has to be finally cleaned and surface activated before it can be used as a cathode in a detector. To produce large areas many small elements that can be cost-efficiently produced will be stitched together.
The major challenge of III-V cathode development is to optimize materials composition for the required spectral range. Tri-component materials combination like GaAsP or InGaN will be a possibility to achieve that but the large strain caused by lattice mismatch and the alloying process itself will increase the defect density and therefore reduce the quantum efficiency, result in brittle and difficult to handle materials, and increase the dark current. Alternatively, multilayered structures can be developed where individual layers are tuned for photon absorption and electron propagation and emission.
III-V Nano-Structured Semiconductor Cathodes
Currently no nano-structured photocathodes are used or developed. However, various groups developed a wide range of fabrication tools which allow the growth of well defined three dimensional hetero structures with tailored doping profiles. The quality of the available structures ranges from epitaxial to amorphous. The morphology can be tuned to the requirements of the application and ranges from fully random to highly ordered structures.
Employing nano-technology has potentially a large benefit. The strain of layered systems with large lattice miss match is often reduced allowing a larger variety of materials combination with minimized defect density. Additionally, the three-dimensional structure can be optimized to minimize reflection losses for light which is coupled into the cathode. A cathode made from individual pillars will act like a photon trap nearly independent from the incidence angle.
Many members of the cathode community suspect that fieldemission caused by the three-dimensional structure is creating an unacceptable dark current. A feasibility test which determines the dark current is therefore the most important action item.
The Scientific Program
The purpose of this program is to understand in a quantitative way the correlation between materials properties and the functionality of the photocathode. By growing systematically various structures and characterize not only the functionality but also their electronic, composition-, and structural properties we hope to be able to develop novel design concepts for photocathodes similar to rational design concepts in related technological areas like the development of solar cells.
The strong interference of the individual fabrication aspects during the optimization process makes a basic understanding of the underlying processes very difficult. To overcome the heuristic approach of present photocathode development we will break the problem up in individual optimization problems allowing parallel efforts. To minimize the cost and efforts we also will work with 3 different sample sizes, a small size (10x10mm2) to systematically study growth and handling effects, a 33mm diameter test size for production of full small detection systems and a 8”x8” prove of principle size in the last phase of the project.
Multi-Alkali Cathodes
There are two major thrusts in our effort for the multi-alkali photocathodes: (i) the fabrication of high quality photocathodes, and (ii) the study of how various characteristics such as morphology, doping, defects, vacuum pressures, inert gas atmosphere, etc. affects the important properties of the photocathodes such as QE, lifetime, dark current levels, etc.
1. Fabrication
Fabrication of multi-alkali photocathodes will focus on two different aspects. The first will be the study on the feasibility to make large, uniform photocathode with long lifetime. This is predominantly an engineering study on the ability to scale-up the size of fabrication up to 8”×8”. The second aspect will be the fabrication of high-quality, single-crystal multi-alkali samples. The availability of such single-crystal film is crucial in the ability to study the material properties of this photocathode and how these properties can affect the physical characteristics such as QE, lifetime, etc. To the best of our knowledge, single-crystal multi-alkali photocathodes have not been extensively studied, and therefore, our ability to fabricate such single-crystal material is crucial.
While most multi-alkali photocathodes have been grown either using thermal vapor deposition or chemical vapor deposition (CVD), these techniques do not produce single-crystal samples that we require to make the necessary in-depth investigation. Instead, we plan on using the molecular-beam epitaxy (MBE) technique to grow single-crystals of multi-alkali photocathodes. The initial plan calls for this to be done on single-crystal sapphire substrates with a particular crystal orientation.
One added advantage of this technique is that it allows for the “option” of producing polycrystalline-amorphous photocathode by employing a derivative of the MBE process – the molecular-beam deposition (MBD). Such technique has been shown to produce photocathodes comparable in quality as CVD cathodes. It will allow us to also produce photocathodes that will be closer in resemblance to the ones that will actually be used for the photodetector and to subject it to various conditions. This includes a study of how its performance, such as QE, is affected by a backpressure of inert gasses. Further studies on the relationship between the photocathode material and the substrate are also planned. This includes the study of suitable conducting layer between the substrate and the photocathode to provide ground contact and prevent charging effects.