Controlling the growth of semiconducting nanowires

    MIT doctoral candidate Sam Crawford has contributed to fundamental understanding of growth processes using metal seed particles.

    MIT graduate student Samuel C. Crawford, working in Associate Professor Silvija Gradečak’s Laboratory for Nanophotonics and Electronics, is lending a new understanding to growing semiconducting nanowires for solar cells, LEDs and other uses.

    Crawford’s research demonstrated control of the composition and diameter along individual nanowires composed of indium gallium nitride (InGaN)by varying the flow of gaseous precursors containing the desired materials, such as gallium, through a quartz chamber containing substrates coated with gold seed particles. “Essentially what we’re doing is changing the flows of our III and V precursors (elements from columns III and V of the periodic table) during growth in order to change the composition and morphology of the nanowires,” Crawford said. Crawford and colleagues grew nanowires with a “caterpillar” shape by alternating indium nitride and indium gallium nitride layers within the nanowire, leading to changes in diameter. (See related video.)

    Video: Caterpillar-shaped nanowires

    Video by Georg Haberfehlner (CEA-Leti) and Sam Crawford (MIT)

    MIT graduate student Sam Crawford, Associate Professor Silvija Gradečak, and colleagues grew semiconducting nanowires which exhibited a "caterpillar" shape by alternating the composition in InN/InGaN heterostructure nanowires. The team developed an experimental process for controlling the composition and diameter along individual nanowires, as well as a theoretical model to guide nanowire growth, which could lead to better solar cells, LEDs and sensors.

    Crawford, 27, a National Science Foundation graduate research fellow, first reported the nanowire synthesis results in a paper with Gradečak, Sung Keun Lim (Ph.D. ’11) and Geog Haberfelner of CEA-LETI in France, then in a subsequent paper provided a theoretical understanding of principles generally applicable to nanowire growth.

    The seed particle

    In Gradečak’s chemical vapor deposition system, nanowires are grown at temperatures from 500 to 1,000 degrees Centrigade by the “vapor-liquid-solid” mechanism in which elements from gaseous precursors alloy with a liquid seed particle and react with one another to form a solid nanowire. The seed metal helps to mediate nanowire growth but does not participate in the reaction.

    Scanning transmission electron microscopy coupled with energy dispersive X-ray spectroscopy allowed Crawford to calculate the percentages of the precursor atoms in the seed particle alloy, which confirmed that the changes in nanowire diameter resulted from changes in seed particle volume and composition. “X-rays can be emitted as a result of knocking electrons out of an atom, and so the energy of the produced X-ray depends on the atom that you’re probing, so the X-rays provide a fingerprint of the atoms in your material,” Crawford said. In collaboration with CEA-LETI, electron tomography also revealed the geometric shape, or morphology, of the nanowires.

    To be most effective, a seed metal should be able to promote nanowire growth over a large range of compositions. But that’s not something that just any metal can do. “Gold in general has been shown to be a good seed metal for a variety of different nanowire materials, but the gold is liquid over the entire range of compositions that we are working with. For other metals or different operating temperatures, the seed particle may have to undergo phase transitions from liquid to solid phases in order to change the composition and thus modulate the diameter,” Crawford said.


    Associate Professor Silvija Gradečak and doctoral candidate Sam Crawford with metal organic chemical vapor deposition system used to grow semiconducting nanowires. Crawford’s research demonstrated how to control growth of nanowires such as indium gallium nitride (InGaN) by varying the flow of gases containing the desired materials, such as gallium, through a quartz chamber with a gold seed particles on a base layer, or substrate. Photo: Denis Paiste, Materials Processing Center

    Modeling diameter modulation

    Even without changing the nanowire composition, they were able to change the diameter along individual nanowires during growth.

    The modeling paper builds on the experimental work Crawford and colleagues developed for experimentally modulating nanowire diameter, using results from diameter-modulated InN and GaN nanowires as a basis for theoretical investigation.

    “This theoretical manuscript parses out the relative contributions of volume, wetting angle, the shape of nanowire cross-section, and so forth, to the observed changes in diameter. Using our results as a basis for applying our model, we can then extract general principles for diameter modulation in different materials systems,” Crawford said.

    The world is not flat

    By modeling the process of nanowire formation, Crawford showed that nanowires, while nominally one-dimensional, have a complex nanometer-scale interface morphology between the seed and the nanowire that is either faceted or rounded at the edges. But it is unlikely to be flat, which had been a generally held assumption for about 50 years since metal-seeded nanowire growth was first discovered in 1964. “Over the past couple of years, there have been a few different studies that have suggested it might not be flat,” Crawford said. “This work goes to further support that notion that it’s not flat and actually shows from a fundamental force balance that the seed particle wouldn't even be stable if the surface was flat. A lot of the models that are used to describe phenomena of nanowire growth just assume a flat interface, and our result shows that it's not necessarily a valid assumption. This finding can have significant impacts on general understanding of nanowire growth. The more you know about how they grow, the more you know how you can manipulate growth parameters and so forth to get the structure that you want.”

    Although the findings regarding interface morphology were a consequence of Crawford’s analysis of diameter modulation, “It was an important fundamental insight into what is happening during the growth,” Gradečak said.

    Opportunity in complexity

    Research publications

    Fundamental Insights into Nanowire Diameter Modulation and the

    Liquid/Solid Interface

    Controlled Modulation of Diameter and Composition along

    Individual III−V Nitride Nanowires


    MIT News Office

    A new twist on nanowires


    Explained: Nanowires and nanotubes

    A fifth-year graduate student in materials science, Crawford expects to defend his doctoral thesis in September 2013 and receive his Ph.D. shortly after. He hopes to work on materials issues in energy technologies. His work has shed light on the complexity of nanowire growth, but that complexity also offers flexibility, Gradečak says. “You can think about this complexity as being an opportunity to modify the nanowire properties. Instead of just simply having straight one-dimensional nanowires, the fact that you do have multiple processes, one can use these different ‘knobs’ to tune the properties of the material. Only through the fundamental understanding of how these ‘knobs’ influence the nanowire properties can we gain something more than random and uncontrolled growth of nanowires,” Gradečak said.

    Until recently, researchers knew how to grow nanowires but did not understand the process well enough to control that growth. “We are just in the early stages of that understanding, so we are really hoping to contribute to the field in a significant way, and Sam played a pivotal role,” Gradečak said.

    Born in Kansas City, Kansas, Crawford grew up in Kansas City, Mo., and received his bachelor’s degree in chemical engineering at Washington University in St. Louis. Crawford expects to publish another paper demonstrating diameter modulation in a different material system based on the findings in his previously published work. “We demonstrate we can control the phase of the seed particle by controlling the initial conditions during growth, whether it prefers to be a liquid or a solid,” he said.

    The work was supported in part by the National Science Foundation and the MIT-France Seed Fund (MIT International Science and Technology Initiatives-MISTI), with use of the nanocharacterization platform (PFNC) at MINATEC Campus in Grenoble, France, and Shared Experimental Facilities provided by the MIT Center for Materials Science Engineering supported in part by the NSF MRSEC Program under award number DMR-0213282.back to newsletter

    Last modified on Thursday, 05 September 2013 15:12