Nanowires Characterization

Get Complete Project Material File(s) Now! »

Experimental Procedures

A method for producing large quantities of high aspect ratio-oriented semiconductor nanowires in a safe and scalable method has been developed. This section describes in detail the process for producing both the source preforms as well as the conditions necessary for drawing those preforms to target fiber sizes for integration in current fiber systems. Analysis of the optical and electrical properties of these nanowire arrays was conducted to understand and evaluate potential uses of these nanowires.

Semiconductor NW Fiber Production

Preform Fabrication

The fabrication of preforms is the first step in producing optical fibers and in this case consists of the production of the powders for the core region and the production of the glass cladding material. The cladding material ranges from directly using a stock purchased tube for fibers on the lathe to multiple over cladding steps to making a custom thick-walled tube for use on the tower. These initial preforms allowed fibers to be directly produced on the lathe but suffered from size and speed limitation of the lathe that provided a finite limitation on draw down and thus the length and size of the nanowires produced. These lathe produced fibers were capable of producing some nanowires but the majority of the semiconductor features were above the target 100nm diameter size and extremely inconsistent fibers in both length and diameter resulted. To overcome these limitations fiber production was shifted to the draw tower, which allows for much larger preforms and much higher draw speeds. While the size of the preforms differs greatly depending on the target diameter of the fiber and the method of drawing, the production of the core material is constant across all preforms and only differs in the compositional content.
The production of the core material is common to all the preforms and so is covered first, and then the differences in cladding regions are covered afterwards.
In addition to these powder-filled preforms a secondary type of preform was also developed. This preform consists of a stacked tube approach and allows for more precise control of location of semiconductor features as well as combining semiconductor features with traditional optical fiber cores and materials. The stacked approach can be achieved by taking thicker sections of fiber produced in the range of 1-2mm in diameter and arranging them inside a glass tube with a doped core region. These rods can consist of single silicon features similar to those produced in other work done by our group[4] or of rods that contain multiple features produced in this work.

Core Material

In all preforms, the core was composed of a mixed powder of fused silica and the desired semiconductor powders. Due to the limited purities in commercially available powders of target sizes, all powders used were produced in the lab. Powders were produced in the lab by grinding using an alumina mortar and pestle. Sieving was used to control and limit the particle sizes in the preforms. Four different sieves were used to control the sizes of particle used, No. 100, 120, 325, and 635. These meshes correspond to particle sizes of 150, 125, 45, and 20 microns respectively. The source materials and sizes used for the various preforms are summarized in Table 0-1.
The sizes were chosen to provide particles that would draw down to nano size based on the draw down ratios possible with both the lathe as well as the draw tower. The smallest size of 635 mesh was also the smallest commercially available standard sieve size. The larger 100 and 120 mesh sizes were then selected to provide particles sizes that provided interstitial spaces the size of the smaller 20-micron particles.
These powders were used to make a variety of compositions of semiconductor-glass powders. Compositions were investigated ranging from 5-30% wt. semiconductor in glass and mixed using a rolling jar mill for 1 hour.

Cladding Material

In all preforms, the outer tube used was GE Type 214 Fused Quartz tubing. This tubing was used, as it is suitable purity for use in optical applications. Initial preforms used on the lathe were produced using stock sizes of tubes available from the supplier. For the larger preforms necessary to be used on the draw tower, the cladding regions were produced in the lab. The glass lathe used to produce these preforms as well as fibers is shown below in Figure 0-1.
These tubes were produced by using a series of close-fitting tubes bought from the manufacturer and then over cladding the tubes to the desired sizes. This process was done on the glass lathe and is completed in the following series of steps.
Seal the end of the smaller inner tube.
Mount the smaller tube to the tailstock chuck.
Mount the larger overclad tube to the headstock chuck.
Using the lathe controls insert the smaller tube into the desired length of the preform being overclad.
Attach vacuum line to the end of the larger tube on the headstock side.
Heat the open end of the larger tube near the tailstock end allowing it to naturally consolidate and collapse onto the smaller tube.
Once the outer tube has collapsed and attached to the smaller tube around the entire circumference the vacuum is then turned on.
The torch is then slowly traversed (5-30mm/min) down the length of preform collapsing the outer tube onto the inner tube.
The use of the vacuum allows the process to be done much quicker than waiting for the natural collapse as well as allowing for a uniform clean interface between the two tubes. This process is then repeated continuing until the desired size of the preform is reached. For most of the preforms used for the draw tower, this process is done 2-3 times. This process allows for virtually any size of preform to be created and provides very repeatable sizes with similar tolerances in sizes to those received directly from the manufacturer. Table 0-2 below summarizes a list of tubes used to make the preforms.
This set of tubes creates a preform with an outer diameter approximately 23mm. This large outer diameter allows for much larger draw down ratios and thus the ability to create significantly smaller features in the fiber than those of smaller preforms.
The final step in creating the preforms is to fill the core region with the desired target powder composition. When inserting the powders into the tubes a static charging problem was observed between the powders and the glass tubes. This problem made powder insertion difficult and was eliminated by using a funnel made from aluminum foil. Continually tapping of the preform was done to aid the powder in settling, provide the highest packing possible, and limit the void spaces present. After the tube is filled to the desired level, a small amount of fused silica wool is inserted in the tubes to act as a filter and limit any powder movement during the vacuum assisted drawing process. Figure 0-2 below shows several preforms produced both for lathe drawing as well as for tower drawing. All of these preforms are shown after the drawing process.

Fiber Drawing via Lathe

Fiber production was initially completed using the lathe to provide preliminary results and small-scale evaluation. The glass lathe provides motorized control of both the pulling tailstock as well as the torch carriage. This allows for draw and feed speeds ranging from 0-5000mm/min. Heating for the lathe is provided by an oxy-hydrogen torch with rotation to provide a uniform heating profile. Based on the draw and feed speeds capable on the lathe, the maximum diameter of preform capable of producing a fiber of the target 125 microns is 8mm. Initial cladding tubes for fiber production on the lathe were stock tubes with an OD of 8mm and an ID of 3mm. This size was chosen as it provided the thickest wall and did not exceed the max size.
Producing fibers from the small preforms directly on the lathe suffered from several limitations. The oxy-hydrogen torch heat source provided an uneven heat profile that was partially alleviated by rotation of the preform during drawing. This uneven heat source and horizontal orientation led to pooling and movement of the molten semiconductor material present. This caused fibers produced to have inconsistent sizes as well as regions with no core material present. The lathe also limited the max length of fiber produced and the draw down. Additionally, a blow out of a gaseous species in the heated preforms was frequently observed.

Fiber Drawing via Tower

The majority of fiber drawing was completed using the draw tower located at the Fiber Optic Draw Tower Facility on Prices Fork Rd. The draw tower provides considerably increased capabilities in producing fibers. The tower is 6m tall, capable of continuously producing kilometer lengths of acrylate-coated fibers and is shown below in Figure 0-3. This tower is equipped with a resistive element graphite furnace capable of rapidly heating to and operating at a maximum temperature of 2400°C. This furnace operates in an inert nitrogen atmosphere and provides a uniform heating profile for optimal fiber drawing. The tower also provides real time diameter measurements using a pair of laser micrometers to measure both the drawn fiber size as well as the fiber size after being coated with an acrylate coating. The draw tower is capable of feed speeds ranging from 1-50mm/min and draw speeds of 10-100cm/sec. These speeds allow for a much larger range of preforms to be drawn with the only limit on size being provided by the 25mm opening of the furnace.
Development of the drawing process was ongoing with the goal of producing long uniform fibers with continuous sections of nanowires present in the core region. Preforms for use on the draw tower underwent additional steps that were not needed for drawing on the lathe. The first step added a “drop” to the bottom end of the preform. The drop consists of a fused quartz rod 2-5cm in length of similar diameter as the preform being added to the end of the preform. This drop is added to provide additional mass to minimize waste of preform at startup of the draw on the tower. The additional mass allow for draws to be started quicker and at a lower temperature. This drop is added prior to filling the preform with powder to prevent any unwanted heat effects on the powder. After the preform is then filled with powder, it is returned to the lathe for a handle tube to be attached. This handle tube is a long ID 19mm OD 25mm tube attached to allow a long enough sample to be mounted on the draw tower and connected to the operating vacuum. The process for mounting and starting a draw on the tower is summarized below.
Preform is inserted into the mounting chuck at the top of the tower with at least 400mm of preform present below the chuck.
Process vacuum is then connected to the top of the preform using Teflon tape to provide a seal between steel vacuum tube and the glass preform.
Process vacuum is turned on ensuring that fiber filter is prevent powder from being removed from the preform.
Preform is then lowered into the furnace so that the drop is located just below the hot zone approximately 180mm from the top of the furnace.
The furnace is purged with nitrogen gas and maintains continually flowing nitrogen during drawing.
The furnace is then rapidly heated to draw temperature of approximately 2000°C in 30-40 minutes.
Upon reaching temperature, the weight of the drop assists in the causing drawing to begin. The bottom furnace gate is opened to allow the passage of the drop and closed back around the fiber diameter.
Once the drop is clear of the furnace it is removed from the fiber and the fiber is hand pulled down the tower passing through the pressure coating system and the UV curing lamp.
The fiber is then inserted into the drawing capstan going at its minimum speed and feed of the preform is turned on.
Temperature, draw speed, and feed speed are then controlled to reach the target fiber size and to control the tension present in the fiber during the draw.
Once target sizes are achieved acrylate coating is applied and cured to provide strength and protect the fiber.
Fiber is then fed to the spooler and continuous drawing can begin.
The primary method of controlling size on the draw tower is done by draw speed with the feed speed being left primarily at its minimum value for most draws. The draw is initially started at a higher temperature then lowered to increase the viscosity of the glass being drawn and provide a higher tension draw. The high tension is desirable to increase the force applied to the liquid semiconductor particles and aide in the formation of wires during drawing. Figure 0-4 below schematically shows the drawing process of a preform on the tower.

READ  Cell volume measurements of dendritic cells (DCs) in collagen gel

Gaseous Products during Drawing

During drawing of fibers on both the lathe and the draw tower, gas formation was observed. This gas formation is unable to be completely removed by the vacuum process during drawing and leads to the drawing of capillary tubes and ultimately blowout of the preform. This gas formation was very detrimental to production of nanowires and slowed down the drawing process by interrupting draws with blowouts. The thicker walled preforms reduced the occurrence of blowouts but still allowed for large sections of fiber to consist of capillary tubes with no core material present.
Two possible causes of the gas formation were determined to be formation of volatile monoxides of Silicon and Germanium and carbon contamination present in the powders. Formation of the monoxides is possible when silicon and germanium are heated in oxygen deficient environments and sublimates below the melting temperatures of the semiconductors. This should have minimal effect on the blowout, as the gas is present below the softening point of the silica tube and should no longer be produced once the semiconductors have reached their molten states. This leaves the likely culprit of gas production to be carbon contamination. This is further confirmed by the presence of a yellowish-brown residue observed in the capillary tube sections produced that is consistent with carbon contamination observed in other fiber drawing projects. In work on photonic crystal fiber preforms carbon contamination is removed from the glass tubes via burnout with torch. This leads to the development of a furnace treatment to remove the carbon from the powders prior to their placement in preforms. A furnace treatment of 760°C for 2 hours in air was used to eliminate carbon contamination of powders. The powders were spread out in fused quartz boats into thin layers to aid in providing the surface area to remove the contaminant. The temperature of 760°C was determined based on common treatments used for carbon burnout in ceramic materials as well as being above the recommended oxidation temperature of carbon in ASTM7542-15. Preforms produced with powders post heat treatment virtually eliminated blowout and completely removed the yellow contamination and hollow regions present in the fibers.

Characterization of NW Containing Fibers

Characterization of these fibers containing nanowire features was performed using a variety of methods to fully understand the morphology created and how they relate to the starting sizes of the powders. The electrical and optical properties of these fibers were evaluated by measuring the transmission of light both broad spectrum and laser light of various wavelengths. The electrical properties were evaluated using a DC power source and picoammeter. Specific relations between the optical and electrical properties were evaluated by looking for any light generated by electroluminescence as well as looking for potential photocurrent generated by these samples.

Characterization of NW Sizes

Characterization of the NW arrays was done using both optical microscopy as well electron microscopy. Optical microscopy was performed using both transmitted as well as reflected light techniques on an Olympus BX51 microscope. Samples for optical microscopy were cleaved on both ends using an optical fiber cleaver to provide a flat smooth surface on both ends. The samples were cleaved to a length of approximately 1 cm. This length was chosen to provide an easy to handle length that could be mounted and allow the transmission light source to be coupled into the bottom of the fiber. The optical microscope was equipped with 100W halogen light sources in both the transmission and reflected directions. Additionally, the microscope was equipped with a mercury halide lamp in reflection to provide a higher intensity source as well as providing strong UV peaks for fluorescence work. Optical microscopy was performed to provide an overview of the entire fiber structure as well as to produce side views of the features.
Scanning Electron Microscopy (SEM) was used to provide high-resolution images of the samples. Two different electron microscopes were used in this work to provide the most information. The Leo 1550 field emission SEM was used as the primary imaging tool due to its in-lens detector, which is capable of providing excellent resolution down to the 2-5nm range. The FEI Quanta 600 FEG environmental SEM was used to analyze samples using backscatter electrons to provide contrast based on atomic number. The FEI scope is also equipped with a higher sensitivity detector for performing Energy Dispersive Spectroscopy (EDS). EDS is used to provide detailed chemical information of a sample with a resolution in the range of 1% atomic. While EDS is limited to a minimum analysis region of approximately one cubic micron, which is considerably larger than the features produced, it is possible to get chemical analysis by removing part or the entire glass matrix surrounding the nanowires.

Table of Contents
General Audience Abstract 
Table of Contents 
Table of Figures 
List of Tables 
1.1 Properties of Silicon and Germanium
1.2 Semiconductor Fiber Production
1.3 Powder Packing
1.4 Etch Chemistries
Experimental Procedures 
1.5 Semiconductor NW Fiber Production
1.6 Characterization of NW Containing Fibers
Results and Discussion 
1.7 Nanowires Characterization
1.8 Analysis of Stress Created by Semiconductor Drawing
1.9 Light Transmission through NW Fibers
1.10 Electrical Characterization of Fibers
Future Work 
Synthesis and Characterization of Si, Ge, and SixGe1-x Nanowires by Fiber Drawing

Related Posts