Molecular Beam Epitaxy from Research to Manufacturing

  • PDF / 4,788,147 Bytes
  • 8 Pages / 576 x 777.6 pts Page_size
  • 51 Downloads / 214 Views

DOWNLOAD

REPORT


Introduction Tonight I will talk about molecular beam epitaxy (MBE) from research to manufacturing. First I will discuss the introduction of MBE in the early 1970s and the exciting achievements made with it. I will conclude with some new directions for MBE. First let us review this technology. Through MBE, materials like semiconducting materials, metals, and insulating materials are grown, atom layer by atom layer. Figure 1 shows a stainless steel MBE chamber, pumped to a pressure of approximately 10 10 torr, with liquidnitrogen-cooled shrouds to further condense the water vapor in the vacuum system. To grow gallium arsenide (GaAs), we mount a substrate in the center where it continuously rotates to give us the uniformity we need, and it is heated to about 580° or 600°C. The effusion cells are filled with pure Ga, pure As, and doping elements such as silicon for n-type doping, and then germanium or beryllium for p-type doping. Important in this MBE system are the in situ monitoring techniques. The system contains a reflection high energy electron diffraction (RHEED) apparatus, producing an electron beam with a grazing angle to the substrate of about one degree. The diffracted electrons are pro-

MRS BULLETIN/APRIL 1995

HEED GUN

LIQUID NITROGEN COOLED SHROUDS

MAIN SHUTTER

EFFUSION CELL PORTS

ROTATING SUBSTRATE HOLDER IONIZATION GAUGE GATE VALVE

SAMPLE EXCHANGE LOAD LOCK

/ EFFUSION CELL SHUTTERS FLUORESCENT SCREEN

1—1

JJ / * /

r—

' i VIEW PORT

—SiIll

1

TO VARIABLE SPEED MOTOR AND SUBSTRATE HEATER SUPPLY

Figure 1. Top view of a molecular beam epitaxy (MBE) system used for the growth of multilayered structures.

Figure 2. Schematic of the arrangement of dangling bonds on the (a) (001) and (b) (001) faces of GaAs before surface reconstruction.

jected on a fluorescent screen. Through the diffraction pattern, we can look at the surface as it is cleaned by desorption of the oxide before we deposit and grow semiconducting materials. Figure 2 illustrates a gallium arsenide crystal. It is seen here that GaAs is formed with alternative layers of Ga and As atoms in the [001] direction; the surface may be terminated in either Ga or As atoms. Notice also that the dangling bonds of these As atoms are 90° from those of Ga atoms. This gives us some important information: The reconstructed surface structures of As layers or Ga layers would be 90° from each other (2 X 4 or 4 X 2). Surface structures stabilized with either type of atom may be produced by controlling the arrival rates of As and Ga atoms or the substrate temperature during epitaxial growth. Figure 3 shows RHEED patterns (left) from a (001) GaAs surface and corresponding electron micrographs (right) of Pt-shadowed C replicas of the surface. The apparently highly polished surface of a GaAs wafer appears to be quite rough when examined by an electron microscope (Figure 3a). When we study this surface with RHEED, because the electron beam impinging on the surface is at a grazing angle, it can penetrate the surface asperities and produce a transmission-diffrac