Final Report Summary - NANOSPID2 (Nanowires for single photon detection and spin memory devices)
The excellent properties of silicon make it the material of choice for transport devices whereas most III-V semiconductor compounds are light emitters and therefore suited for optoelectronics. Combining both Si and III-V semiconductors would lead to important advancements in many different applications from nano to micro-optoelectronics. However, the combination of both classes of materials is challenging because of technological issues, mainly related to lattice mismatch. Nanowires enable to bridge the two distinct worlds of silicon and III-V materials because their small section area releases elastically the strain to the lateral sides of the wire. The NANOSPID2 project targets the fabrication of nanowires containing Si and III-V for applications aimed at single-photon detection and single-electron manipulations. Nanowires are one-dimensional materials, i.e. their cross-section is of the nanometre scale whereas their length is of the order of micrometres. This nanowire morphology leads to novel and unique physical properties that can be exploited in future devices. The advancements of the project can be summarised into different sections: nanofabrication, nanowire growth, and nanodevices.
Nanowires are grown by the vapour-liquid-solid (VLS) mechanism using gold nanoparticle as catalysts. As a first step, we spatially organise these nanoparticles on the substrate by nanoimpression and electron beam lithography. The developed process allows growing nearly defect-free arrays of (heterostructured) III-V nanowires. Soft nano-imprint lithography is used to pattern gold particle arrays on full 2 inch substrates. After lift-off organic residues remain on the surface, which induces the growth of additional undesired nanowires. We show that cleaning of the samples before growth with piranha solution in combination with a thermal anneal at 550 degrees of Celsius for InP and 700 degrees of Celsius for GaP results in uniform nanowire arrays with 1 % variation in nanowire length, and without undesired extra nanowires. We also used this chemical procedure for e-beam lithography and we therefore consider it to be a generic process. We demonstrated the importance of size and interdistance of the gold particles to control the nanowires morphology. In the case of GaP / Si nanowires, there is an optimal gold particle diameter at around 60 nm for which the yield of straight Si nanowire segments can be grown on GaP stems. We also observe an increase (>15 %) of the Si growth rate with decreasing interdistance for a constant diameter which we attribute to the synergetic effect.
Another nanofabrication result concerns the fabrication of Si nanotubes with unique crystal structures such as twinning superlattice or wurtzite phase. Control of crystal structure and the formation of a twinning superlattice have been shown for III-V nanowires but never for Si. We demonstrate the possibility to transfer of a twinning superlattice with the zinc blende crystal structure and the wurtzite crystal structure from a gallium phosphide core wire to an epitaxially grown silicon shell. By a selective chemical etch we remove the GaP core, we obtain arrays of free-standing Si nanotubes with the desired crystal structure.
The main work of the Marie Curie project concerns the growth of hybrid Si / III-V nanowires. The challenge here concerns the fabrication of perfect interface between the III-V semiconductor and Si in the axis of the nanowire, and vice versa. During growth by VLS, Au catalyses the decomposition of the precursor molecules from the gas phase and absorbs the elements, forming a liquid alloy. When supersaturation is reached in the alloy, atoms are expelled to form a crystal (either III-V or Si) below the liquid alloy. We aim at combining different classes of semiconductors, which are normally grown from their corresponding Au-based alloys, such as Au-Ga or Au-Si. We see here all the challenge of switching from III-V to Si during growth. In our work, we demonstrate the importance of controlling the Au catalyst composition and its surface chemistry. The surface chemistry of the Au particle can be controlled by switching on and off the flows of precursors, especially the group V materials which cover the surface of the catalyst particle during III-V semiconductors nanowire growth. By switching off the group V precursor flow, group V a removed from the catalyst surface which is then prepared for Si incorporation. By this technique we demonstrate clean and defect free interfaces between III-V and Si semiconductors in arrays of hybrid Si / III-V nanowires. We realise hybrid Si / GaAs nanowires with GaAs segments of good optical quality.
Finally, to demonstrate detection of single photons in our nanowires, we fabricated avalanche photodetector devices by contacting either p-i-n Si nanowires or p-i-n InP nanowires containing an InAsP quantum dot in the intrinsic region of the nanowire. While applying a reverse bias on the device, photons that have high enough energy are absorbed in the intrinsic region of the nanowire. The electron-hole pairs are then separated and multiplied by impact ionisation because of the high electric field present in the region. Such nanowire avalanche photodetectors are efficient one-dimensional channels for photon detection and are sensitive to single photons. The next step of the project is to demonstrate that our hybrid Si / III-V nanowires will not only allow the detection of single photons in the infrared range by introducing low bandgap III-V in a Si nanowire, but also open new routes in the field of quantum information processing.
Nanowires are grown by the vapour-liquid-solid (VLS) mechanism using gold nanoparticle as catalysts. As a first step, we spatially organise these nanoparticles on the substrate by nanoimpression and electron beam lithography. The developed process allows growing nearly defect-free arrays of (heterostructured) III-V nanowires. Soft nano-imprint lithography is used to pattern gold particle arrays on full 2 inch substrates. After lift-off organic residues remain on the surface, which induces the growth of additional undesired nanowires. We show that cleaning of the samples before growth with piranha solution in combination with a thermal anneal at 550 degrees of Celsius for InP and 700 degrees of Celsius for GaP results in uniform nanowire arrays with 1 % variation in nanowire length, and without undesired extra nanowires. We also used this chemical procedure for e-beam lithography and we therefore consider it to be a generic process. We demonstrated the importance of size and interdistance of the gold particles to control the nanowires morphology. In the case of GaP / Si nanowires, there is an optimal gold particle diameter at around 60 nm for which the yield of straight Si nanowire segments can be grown on GaP stems. We also observe an increase (>15 %) of the Si growth rate with decreasing interdistance for a constant diameter which we attribute to the synergetic effect.
Another nanofabrication result concerns the fabrication of Si nanotubes with unique crystal structures such as twinning superlattice or wurtzite phase. Control of crystal structure and the formation of a twinning superlattice have been shown for III-V nanowires but never for Si. We demonstrate the possibility to transfer of a twinning superlattice with the zinc blende crystal structure and the wurtzite crystal structure from a gallium phosphide core wire to an epitaxially grown silicon shell. By a selective chemical etch we remove the GaP core, we obtain arrays of free-standing Si nanotubes with the desired crystal structure.
The main work of the Marie Curie project concerns the growth of hybrid Si / III-V nanowires. The challenge here concerns the fabrication of perfect interface between the III-V semiconductor and Si in the axis of the nanowire, and vice versa. During growth by VLS, Au catalyses the decomposition of the precursor molecules from the gas phase and absorbs the elements, forming a liquid alloy. When supersaturation is reached in the alloy, atoms are expelled to form a crystal (either III-V or Si) below the liquid alloy. We aim at combining different classes of semiconductors, which are normally grown from their corresponding Au-based alloys, such as Au-Ga or Au-Si. We see here all the challenge of switching from III-V to Si during growth. In our work, we demonstrate the importance of controlling the Au catalyst composition and its surface chemistry. The surface chemistry of the Au particle can be controlled by switching on and off the flows of precursors, especially the group V materials which cover the surface of the catalyst particle during III-V semiconductors nanowire growth. By switching off the group V precursor flow, group V a removed from the catalyst surface which is then prepared for Si incorporation. By this technique we demonstrate clean and defect free interfaces between III-V and Si semiconductors in arrays of hybrid Si / III-V nanowires. We realise hybrid Si / GaAs nanowires with GaAs segments of good optical quality.
Finally, to demonstrate detection of single photons in our nanowires, we fabricated avalanche photodetector devices by contacting either p-i-n Si nanowires or p-i-n InP nanowires containing an InAsP quantum dot in the intrinsic region of the nanowire. While applying a reverse bias on the device, photons that have high enough energy are absorbed in the intrinsic region of the nanowire. The electron-hole pairs are then separated and multiplied by impact ionisation because of the high electric field present in the region. Such nanowire avalanche photodetectors are efficient one-dimensional channels for photon detection and are sensitive to single photons. The next step of the project is to demonstrate that our hybrid Si / III-V nanowires will not only allow the detection of single photons in the infrared range by introducing low bandgap III-V in a Si nanowire, but also open new routes in the field of quantum information processing.