Applications of Epitaxial lift off technology for II-VI semiconductors

Semiconductor structures containing different materials are grown as thin film multilayers by techniques such as molecular beam epitaxy (MBE). MBE produces layers with excellent control of thickness but is limited to total thicknesses of just a few microns. In addition, growth takes place on a substrate, which is a highly crystalline template of a material such as gallium arsenide. After growth, the thin film layer remains bonded to the substrate.
However, if one of the layers deposited is a so-called sacrificial layer soluble in a solvent (such as a weak acid) then all the layers deposited on top of it can be removed from the substrate. This process is called epitaxial lift off (ELO) and is advantageous in applications where the substrate is either not required or even hinders the operation of the device. Often using ELO means that the substrate can be recycled, which can reduce operating costs. An additional use for ELO is that the layers can be assembled into complex structures with many different types of materials. ELO layers can be transferred to intermediate flexible plastic substrates and patterned before assembly, so very complex structures can be produced.
II-VI semiconductors are materials with a number of very useful properties, for example bandgaps ranging from 0 to 5eV. Other II-VI semiconductors have useful magnetic properties, for example some (e.g. CrS) are ferromagnets and others (e.g. MnS) are antiferromagnets. At Heriot-Watt University (HWU), we developed ELO for II-VI compounds using MgS sacrificial layers. The original method could only be used on small sample sizes (3mm square) but demonstrated many useful applications. Within the last few months we have developed a number of breakthroughs in II-VI ELO which show it has much more potential. In particular, we can remove pieces several square cm in size using a flexible plastic carrier. An additional very useful property is that when two ELO layers touch they will combine together, or stack, with the adhesion between layers so strong that they cannot be separated without breaking them.
This proposal aims to develop this technology in 3 ways. First, we will show that ELO is easily extended to whole semiconductor wafers, and ELO layers can be transferred on flexible plastic carriers and patterned into small components. The components can be transferred again (stamped) to a final destination. All of this will be done with high (~100%) yield.
Second, we will demonstrate the advantages of II-VI ELO by assembling 5 different demonstrator devices requested by our colleagues at HWU. We will supply these for evaluation as part of their own on-going research programmes. The devices include two types of sensors (temperature, and electric or magnetic fields), an optical diode, which only allows light propagation in one direction, a frequency doubler and a photonic bandgap structure. These structures are very difficult to produce by normal thin film growth techniques, but are easily produced by stacking ELO layers.
The final strand of the programme develops the potential of ELO in different ways. The ability to move electrons or holes between ELO and adjacent layers would increase the number of applications: for example allowing us in future to develop photovoltaics or detectors. We will measure the electrical transport properties across ELO junctions between ZnSe and different materials and if possible modify them with different surface treatments.
One surface treatment developed at HWU protects the II-VI layer surface after growth against contamination. At HWU it has worked for several months. We aim to show that it can be used to transport HWU ELO layers to City College, New York and show that it is possible to combine materials which are not available in the same MBE system and make ELO available to other groups.

The proposal describes how samples can be moved between centers to produce devices using ELO. The impact of this for the II-VI growth community will be very large for several reasons. First, it enables complex structures to be deconstructed into their functional units which can be grown and optimized individually. Second, it allows laboratories to specialize in the materials they grow best and work with other growth centers to provide added functionality to devices. This will mean that individual growth centers instead of working independently will form cooperative groups. This will be a major change in culture and working practices for all material suppliers.

The proposal gives examples of five demonstrator devices which will be supplied to projects at HWU in widely differing areas four of which are already funded. In all cases, successful operation of the device demonstrates impact within to a particular device community arising from simpler manufacture, improved performance or functionality not obtainable via other production methods. Two devices intended for temperature and electric/magnetic field sensor applications already have industrial partners and we would expect that successful demonstrations would lead to funding for device optimization and rapid utilization which could occur within a short timescale of 1-2 years. Development of the other devices will probably be slower, over 3-5 years but again there are identified applications areas for these devices. The device furthest from exploitation is the optical diode. Further development here will be required to demonstrate optical isolators, followed by optical transistors.

Subsequent impact for ELO and II-VI integration in devices requires that more applications be identified. In the longer term this will require that II-VI ELO technology be publicized to the widest possible audience, and we will do this. In the short term, however we note that at HWU there are already another 2-3 photonics applications which could not be included in the this proposal.

A final part of this last section concerned the investigation of other properties of ELO layers, in particular current transport through junctions. This is the most speculative part of the proposal, as nothing is currently known about the properties. However, if this part proves successful, then future development will involve the investigation of current transport devices such as sensors and photovoltaics.

There are also educational impacts arising from ELO. As a semiconductor processing step ELO is very visual and can, in some structures, produce visible results in a few minutes. It therefore works well in scientific demonstrations and naturally leads into discussions of related topics such as semiconductor growth, optoelectronic and photonic devices etc. It is a natural topic to introduce the public to semiconductor research and devices.

The project will also have a major impact on training of the future generation of research staff. All stages it will generate a large number of samples which will find their way into various student research projects at all levels. Semiconductor samples have a long useful life even after the conclusion of a funded program and based on our previous experience we estimate that over the next decade samples grown on this program would feature in research topics for perhaps 10-15 PhD students and be used in undergraduate project work for an estimated 20-30 final year projects within HWU.