Nanocolumns\ Nanowires fabrication literature review

---

This is in a series of literature reviews on InGaN solar cells, which supported the comprehensive review by D.V.P. McLaughlin & J.M. Pearce, "Progress in Indium Gallium Nitride Materials for Solar Photovoltaic Energy Conversion"Metallurgical and Materials Transactions A 44(4) pp. 1947-1954 (2013). open access Others: InGaN solar cells| InGaN PV| InGaN materials| InGan LEDs| Nanocolumns and nanowires| Optical modeling of thin film microstructure| Misc.

---

InGaN based LED and LASER devices.[edit | edit source]

The Controlled Growth of GaN Nanowires[1][edit | edit source]

Abstract: This paper reports a scalable process for the growth of high-quality GaN nanowires and uniform nanowire arrays in which the position and diameter of each nanowire is precisely controlled. The approach is based on conventional metalorganic chemical vapor deposition using regular precursors and requires no additional metal catalyst. The location, orientation, and diameter of each GaN nanowire are controlled using a thin, selective growth mask that is patterned by interferometric lithography. It was found that use of a pulsed MOCVD process allowed the nanowire diameter to remain constant after the nanowires had emerged from the selective growth mask. Vertical GaN nanowire growth rates in excess of 2 μm/h were measured, while remarkably the diameter of each nanowire remained constant over the entire (micrometer) length of the nanowires. The paper reports transmission electron microscopy and photoluminescence data.

• Reports a scalable process for the growth of high-quality GaN nanowires and uniform nanowire arrays in which the position and diameter of each nanowire is precisely controlled based on conventional metal-organic chemical vapor deposition (MOCVD) using trimethygallium (TMGa) and ammonia (NH3) and requires no additional metal catalyst.
• The location, orientation, and diameter of each GaN nanowire are controlled using a thin, selective-growth mask that is patterned by interferometric lithography.
• This method enables to fabricate high quality symmetrical hexagonal sidewall facets over the entire micrometer length of nanowires.
• This method was carried out in two phases, first phase used continous MOCVD till nanowires emerged out of mask and second phase was pulse MOCVD to fabricate nanowires emerged out of mask at the end of phase one, using pulsed MOCVD in second phase maintained the quality of nanowires and were highly alligned and ordered as desired with respect to the mask.
• diameter of the nanowire remains constant as it emerges from the growth mask, confirming from XTEM images that the nanowire diameter is indeed controlled by the diameter of the growth mask aperture.
• XTEM images show no threading dislocations (TDs) in the GaN nanowires, even though TDs were observed in the planar GaN film beneath the growth mask. This was found to be the case for GaN nanowire growth on all substrates, including growth on silicon (111).
• recorded band-edge PL peak intensity for the nanowire sample was 100 times greater than that measured for a 5 micron planar GaN film and more than 200 times greater than that measured for a 0.6 micron planar GaN film. Much of this intensity increase is undoubtedly due to the geometry of a nonplanar nanowire sample, where the input coupling of the PL pump beam and the out-coupling of the resulting PL will both be increased significantly.

InGaN nanorod arrays grown by molecular beam epitaxy: Growth mechanism structural and optical properties[2][edit | edit source]

Abstract: Vertically c-axis-aligned InGaN nanorod arrays were synthesized on c-plane sapphire substrates by radio-frequency molecular beam epitaxy. In situ reflection high-energy electron diffraction was used to monitor the growth process. X-ray diffraction, transmission electron microscopy, field-emission scanning electron microscope, and photoluminescence were used to investigate the structural and optical properties of the nanorods. The growth mechanism was studied and a growth model was proposed based on the experimental data. A red shift of photoluminescence spectrum of InGaN nanorods with increasing growth time was found and attributed to the partial release of stress in the InGaN nanorods.

• Initially RHEED pattern was streaky, indicating a two-dimensional (2D) growth with a smooth surface at the initial stage. As growth proceeded, one observed mixed RHEED pattern including spots and lines, but the lines were still streaky. After deposition for 13 min, circular spotty RHEED pattern was observed; indicating a three-dimensional (3D) island growth, circular spotty RHEED pattern lasted for 30 min.
• Afterwards the circular spotty RHEED pattern broadens along the direction parallel to the surface; this is typically the RHEED pattern of nanorods. Such a pattern suggested that the nanorods nucleated on the rough InGaN layer, grew up while narrowed, nanorods were single crystalline and collectively grew along the c-axis.
• misfit accommodation can take place through the onset of islanding and generating surface roughness in strained epitaxial systems.
• growth-mode transition from layer-by-layer (2D) growth to 3D islanding occurs quite abruptly when a certain critical strain was reached. In mismatched InGaN-sapphire system, according to RHEED observations, strained InGaN layer started growing nearly pseudomorphical to sapphire and presented relatively smooth surface, when layer thickness exceeded a certain critical layer thickness (CLT) of about 30 nm, strained InGaN layer started to relax and 3D islands began to form.
• For layer thicknesses well above CLT, the surface became very rough and was dominated by large 3D islands.
• formation of InGaN nanorod array can be roughly divided into following stages: Firstly, InGaN thin film grew pseudomorphically on c-plane sapphire, and then 3D islands formed when layer exceeded the CLT. InGaN nanorods started to nucleate on rough InGaN layer and then grew up. At last, the InGaN nanorods coalescenced.
• increased crystal growth times improves crystal quality.
• band gap of strained InGaN films is larger than that of the relaxed due to the compressive stresses
• Such a stress in InGaN nanorods could not be eliminated until length exceeded several micrometers

Formation of InGaN nanorods with indium mole fractions by hydride vapor phase epitaxy[3][edit | edit source]

Abstract: This work demonstrates the formation of InGaN nanorod arrays with indium mole fractions by hydride vapor phase epitaxy. The nanorods grown on (0001) sapphire substrates are preferentially oriented in the c-axis direction. We found that the In mole fractions in the nanorods were linearly increased at x < 0.1. However, In mole fractions were slightly increased at x ≥ 0.1 and then were gradually saturated at x = 0.2. CL spectra show strong emissions from 380 nm (x = 0.04, 3.26 eV) to 470 nm (x = 0.2, 2.64 eV) at room temperature.

• nanorods have the potential for negligible non-radiative recombination loss, and thus efficiency of down-conversion is much higher than bulk InGaN/GaN.
• average diameter and length were 70 nm and 2 μm, respectively. Control of InGaN nanorod's diameter and length were achieved by adjusting growth temperature and growth time, respectively.
• Increasing HCl gas flow rate for reacting In metal, that is In precursor, from 10 sccm to 100 sccm, In mole fractions (x) were increased from 0.04 to 0.20 in the nanorods.
• When HCl reactant gas flow rate was small (≤25 sccm), In mole fractions were linearly increased with HCl gas flow rate. As HCl reactant gas flow rate was increased, however, In mole fractions in InGaN nanorods were slightly increased and then these were saturated.

Selective area metalorganic molecular-beam epitaxy of GaN and the growth of luminescent microcolumns on Si/SiO2[4][edit | edit source]

Abstract: Author(s) demonstrate the selective area growth of gallium nitride on patterned Si(111)/GaN/SiO2 wafers by metalorganic molecular beam epitaxy using triethyl gallium as a Ga source. Author(s) show that such selective area deposition may be used to grow isolated microcolumns of GaN with lateral dimensions of tens of nanometers on Si/SiO2 wafers. Via high resolution cathodoluminescence imaging Author(s) show that such microcolumn structures are highly luminescent inspite of a large surface to volume ratio, indicating that nonradiative recombination at free surfaces is not a significant issue in this system.

• While selectivity is observed, porous growth morphology is clear as well. This is a consequence of high temperatures required for selective area growth by MOMBE, where significant thermal etching can occur. Such porous morphology is not observed for layers grown at ~ 750 °C.
• temperature requirements for a dense microstructure over a large ( > 1 um) length scale and that for selective area growth of GaN are in conflict with one another for MOMBE growth. high temperatures required for selective area growth result in a loss of integrity of structure and author(s) conclusion therefore is that such selective area growth of GaN by MOMBE is unsuitable for large area device growth.
• it's an attractive technique for growing small ( < 500 nm) structures of GaN, where lateral length scales are smaller than length scale at which porosity is observed, since it does offer benefit of a relatively low temperature at which selective area growth occurs compared to MOCVD growth.
• This ensures compatibility with substrates such as Si where interfacial reactions between the Si and nitride limit usable temperatures to about 1000–1050 °C.
• The CL lateral resolution is limited by excitation volume of optically active part of specimen and minority carrier diffusion length, depending upon the specimen and measurement details.
• qualitatively computed upper-bound of non-radiative recombination centers.

Rapid growth and characterization of InN nanocolumns on InGaN buffer layers at a low ratio of N/In[5][edit | edit source]

Abstract: c-Axis-aligned InN nanocolumn arrays were vertically grown on 3 μm GaN epilayers with InGaN buffer layers by radio-frequency molecular beam epitaxy without any catalysts. X-ray diffraction, transmission electron microscopy, and field-emission scanning electron microscope were used to study the structural properties of the nanocolumns. It has been found that without InGaN buffers, InN films, rather than nanocolumns, were grown even at the same N/In ratio. In addition, high-quality InN nanocolumns can grow faster on InGaN buffers. The growth mechanism was discussed and the joint actions of the gas–solid and Volmer–Weber modes promote the nucleation and the growth of InN nanocolumns.

• The different growth rates of InN nano-structures may be due to the different growth modes
• 100 nm In0.05Ga0.95N and 30 nm In0.25Ga0.75N buffer layers were deposited in sequence at 600 and 530 °C, respectively, on the 3 μm GaN epilayers that were grown by metal organic chemical vapor deposition. After that, InN nanocolumns and films were grown at 470 °C with the N/In ratio equal to 1, 2, and 4, respectively. The N-flow was kept constant to maintain the pressure stable, and the In-cell temperature was varied in order to grow InN nano-structures or films with different N/In ratio.
• results of the SEM observation suggested that, at a low ratio of N/In, InN nanocolumns could not grow directly on GaN epilayers, but could grow on InGaN buffer layers.
• varied lattice constants of both a-axis and c-axis indicated indium spontaneous phase separation from InGaN buffer closed to the InGaN surface.
• it is deduced that formation of InN nanocolumn array may be divided into 3 stages: firstly, phase separation of indium from In0.25Ga0.75N buffer layer took place in the cooling process, forming InGaN with higher indium composition by the end of the growth of the buffer layer. Indium atoms easily gather into clusters. Secondly, InN preferentially nucleated on In-rich regions according to gas–solid (G–S) mode then neighboring InN islands coalescence according to Volmer–Weber (V–W) mode. Finally, the InN nanocolumns grew up on isolated InN islands according to Stranski–Krastanow (S–K) mode. Therefore, it can be presumed that InN nanocolumns preferentially grow on InN, then on indium precipitates, and finally on InGaN.

Growth of InN nanocolumns by RF-MBE[6][edit | edit source]

Abstract: InN nanocolumns were grown on (0 0 0 1) sapphire substrates by radio-frequency plasma-assisted molecular-beam epitaxy, simply with substrate nitridation, InN nucleation, and high V/III growth conditions. Here, the InN nucleation was conducted by forming In droplets on the nitridated substrates and subsequently nitriding these In droplets. Author(s) investigated the growth evolution of InN nanocolumns and the growth-temperature dependence of the morphology. X-ray rocking curves (XRC) for the (0 0 0 2) reflection of the samples basically consisted of two components: a broad peak and a very sharp one. Author(s) propose that the broad peak came from the lower part of relatively short nanocolumns and the near-interface region of relatively long nanocolumns. In contrast, the sharp peak came from the (strain-free) upper part of relatively short nanocolumns and the high-crystal-quality region apart from the interface in relatively long InN nanocolumns. The relative intensity of the latter to the former increased with growth time. The shape of the nanocolumns varied with growth temperature: nanocolumns grown at 380 and 420 °C had a taper-like appearance, but the top broadened with increasing growth temperature, becoming broader at the top than the base at 470 and 500 °C. When Author(s) grew InN nanocolumns under the same conditions, but without surface nitridation, the subsequent columns were not all along a c-axis. If In droplets were not used, then columns did not form. Thus, both the surface nitridation and In droplets were needed to form c-axis aligned nanocolumns.

• As growth time increased from 1 to 10 min, the density of nanocolumns (islands at this early stage) was increased by additional nucleation in areas between original InN islands.
• GaN nanocolumns were shown to exhibit stronger luminescence than GaN films, explained by nanocolumn's nature containing no or drastically reduced dislocations, which act as non-radiative recombination centers.
• higher V/III ratios than unity at the growing surface reduced the surface migration of In atoms.
• InN islands nucleated directly on nitridated substrates during the growth, relatively short nanocolumns also nucleated directly on nitridated substrates during the growth, and relatively long nanocolumns nucleated from the In droplets.
• InN nucleation with In droplet formation and the subsequent nitriding treatment of In droplets is essential for InN nanocolumn growth.
• lack of substrate nitridation resulted in a severe deterioration in the c-axis orientation of nanocolumns.
• InN nanocolumns had a taper-like shape when grown at temperatures below 420 degree C and thickened toward the end with increasing growth temperature.

InGaN nanorods grown on (1 1 1) silicon substrate by hydride vapor phase epitaxy[7][edit | edit source]

Abstract: In this study, Author(s) report on the growth of the defect-free (dislocation-free) InGaN nanorods on (1 1 1) silicon substrate by our hydride vapor phase epitaxy (HVPE) system. Morphological and structural characterization of the InGaN nanorods by high-resolution scanning electron microscopy (HRSEM) and transmission electron microscopy (HRTEM) indicates that the nanorods are straight and preferentially oriented in the c-axis direction. Cathodoluminescence (CL) characteristic of the single InGaN nanorod shows a strong blue emission peaking at around 428 nm at room temperature.

• many threading dislocations (TDs) are produced in bulk GaN and InGaN due to lattice mismatch with the substrate and difference of thermal expansion coefficient between film and substrate, and thus they affect significantly the device performance as non-radiative recombination centers.
• growth mechanism for nanorods is completely different, and threading dislocations can be all non-existent in nanorods.
• nanorods have potential for negligible non-radiative recombination loss, and thus the efficiency of down-conversion is much higher than in bulk InGaN/GaN layer.
• Control of the InGaN nanorod's diameter and length were achieved by adjusting the growth temperature and growth time, respectively.
• A strong blue emission was observed from a single nanorod peaking at around 428 nm (≈2.9 eV) with electron beam excitation, no GaN-related emission from this nanorod was observed implying that InGaN nanorods grown by using In metal was a ternary single crystal without mixing GaN.

InGaN nanopillars grown on silicon substrate using plasma assisted molecular beam epitaxy[8][edit | edit source]

Abstract: Single crystalline and single phase InxGa1−xN nanopillars were grown spontaneously on (111) silicon substrate by plasma assisted molecular beam epitaxy. The surface morphology, structural quality, and optoelectronic properties of InGaN nanopillars were analyzed using scanning electron microscopy (SEM), energy dispersive x-ray (EDXA) analysis, high resolution x-ray diffraction (HR-XRD), and both room and low temperature photoluminescence spectra. The EDXA results showed that these nanopillars were composed of InGaN and the amount of indium incorporation in InxGa1−xN NPs could be controlled by changing the growth temperature. The room temperature and low temperature PL spectra revealed that the emission wavelength could be tuned from a blue to green luminescent region depending on the growth temperature. The wavelength tuning was attributed to a higher amount of In incorporation at a lower growth temperature which was consistent with the EDXA and HR-XRD results.

• lower substrate temperature promotes higher indium incorporation in InGaN.
• driving force for 3-D growth mode of III-nitrides on silicon is lattice mismatch strain energy and the high surface energy of nitrogen stabilized (0001) III-N surfaces. The major difference between GaN and InN nanopillars growth is substrate temperature used.
• growth of InN nanopillars, however, can be sustained at higher substrate temperature without the observation of such metallic accumulation due to the nitrogen rich condition employed. Hence InxGa1−xNNPswere grown at Tsub that was varied from 500 to 650 degree C.
• InxGa1−xN nanopillars were grown at T_sub varied from 500 to 650 ◦C. At lower T_sub of 500 and 550 ◦C, the formation of InxGa1−xN nanopillars was not observed but, on the contrary, InxGa1−xN was formed as a porous film. However, InxGa1−xN nanopillars were formed at T_sub equal to or higher than 575 ◦C.
• The InxGa1−xN nanopillars have a faceted structure, which is consistent with being a hexagonal pyramid bounded by {1-101} facets. This structure is energetically stable and has been observed in microscale feature growth.
• As the T_sub increases, the diameter of the InGaN NPs decreases while the density increases as a result of lesser coalescence at higher T_sub.
• EDX spectra confirmed that the Nanopillars were composed of InGaN.
• samples with higher density, improved crystalline quality, smaller size and different composition of nanopillars, resulted in higher PL intensity.
• author(s) observed that higher In incorporation degraded the optical quality (lower PL intensity and broadening of the spectra) of the InxGa1−xN NPs.
• optical quality of InxGa1−xN NPs improved at higher growth temperature.

High-quality In0.47Ga0.53N/GaN heterostructure on Si(111) and its application to MSM detector[edit | edit source]

Abstract: Purpose – This paper aims to report on the use of radio frequency nitrogen plasma-assisted molecular beam epitaxy (RF-MBE) to grow high-quality n-type In0.47Ga0.53N/GaN on Si(111) substrate using AlN as a buffer layer. Design/methodology/approach – Structural analyses of the InGaN films were performed by using X-ray diffraction, atomic force microscopy, and Hall measurement. Metal-semiconductor-metal (MSM) photodiode was fabricated on the In0.47Ga0.53N/Si(111) films. Electrical analysis of the MSM photodiodes was carried out by using current-voltage (I-V) measurements. Ideality factors and Schottky barrier heights for Ni/In0.47Ga0.53N, was deduced to be 1.01 and 0.60 eV, respectively. Findings – The In0.47Ga0.53N MSM photodiode shows a sharp cut-off wavelength at 840 nm. A maximum responsivity of 0.28 A/W was achieved at 839 nm. The detector shows a little decrease in responsivity from 840 to 200 nm. The responsivity of the MSM drops by nearly two orders of magnitude across the cut-off wavelength. Originality/value – Focuses on III-nitride semiconductors, which are of interest for applications in high temperature/power electronic devices.

InGaN/GaN quantum-well nanocolumn crystals on pillared Si substrate with InN as interlayer[edit | edit source]

Abstract: Nanocolumn InGaN/GaN crystals were deposited on micropillared Si substrate by molecular beam epitaxy. Low-temperature InN was used as interlayer. With enough free space, the column crystals grew around all the surface plane of the Si pillars and formed InGaN/GaN quantum-well flower structure. The QW crystals are about 100 nm in diameter and 1.1-1.4 μm in length. Raman spectra measurement of the fower structure shows that E2 mode peak line observed at 567.28 cm–1. Photoluminescence measurement indicates a room temperature PL peak position of 620 nm and two peak positions of 404 nm and 519 nm at temperature 15 K. Hg lamp excited photoluminescence demonstrated a clear fluorescence distribution in the flower structure and much stronger emission compared with the quantum-well crystals on the flat Si substrate.

CVD growth of InGaN nanowires[9][edit | edit source]

Abstract: In this paper, the chemical vapor deposition (CVD) growth of InGaN nanowires was systematically studied. The catalyst was Au and the starting materials were Ga, In and NH3. The samples were characterized with scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDX), transmission electron spectroscopy (TEM), and X-ray diffraction (XRD), etc. The influence of the growth temperatures, Au thicknesses, gas flowrates and Ga and In amount on the morphology and properties of InGaN nanowires was investigated. It is found that 600 °C is a suitable growth temperature. On the substrate with Au thickness of 150 Å, helical InGaN nanowires are obtained. The change of NH3 partial pressure and Au thickness will result in the morphology change of the samples. An increase of Ga results in shorter InGaN nanowires while an increase of In amount will lead to longer InGaN nanowires. The morphology will also change when both the amount of In and Ga were increased or reduced without changing the ratio of Ga to In.

• growth temperature, catalyst thickness, gas flow rates, Ga and In amount are all found to have an effect on the morphology.
• suitable fabrication temperature is 600 °C.
• thickness of Au has an effect on the morphology of InGaN nanowire and no InGaN nanowires can be obtained when there is no Au on the substrates.
• On the substrate with Au thickness less than 100 Å, only straight InGaN nanowires were observed while both straight and helical InGaN nanowires were obtained on the substrates with Au thickness of 150 Å.
• Different NH3 and Ar flow rates were tried and it is found that larger NH3 partial pressure will result in InGaN nanowires with better morphology.
• Both straight and helical InGaN nanowires were obtained with larger NH3 partial pressure on substrates with Au of 150 Å.
• Different In and Ga amount were also tried and it is discovered that an increase of Ga amount results in shorter InGaN nanowires and an increase of In amount will lead to longer InGaN nanowires. When both the amount of In and Ga were increased or decreased without changing the ratio, the morphology also changed and no helical InGaN nanowires were observed when Ga and In amount were changed.