Plasma-Assisted Molecular Beam Epitaxy Growth of Mg₃N₂ and Zn₃N₂ Thin Films

In this video, we show how to grow epitaxial films of magnesium nitrite and zinc nitrite by plasma-assisted molecular beam epitaxy, or MBE for short. Magnesium nitrite and zinc nitrite are II-V compound semiconductor materials. This is a relatively unexplored class of semiconductors.

They have the anti-bixbyite crystal structure, which has 80 atoms in the conventional cube unit cell. The films are grown in a VG V80 MBE system. The horizontal chamber on the left is the preparation chamber and the round chamber on the right is the growth chamber where the film growth takes place.

The sample entry lock, located at the left-hand end of the preparation chamber. The best substrate that we have found for growing epitaxial magnesium nitrite and zinc nitrite is 100-oriented single crystal magnesium oxide. The one centimeter square substrates are first placed on a sapphire wafer sample carrier with polished side up and annealed for nine hours at 1, 000 degrees C.High temperature annealing removes the carbon from the surface and reconstructs the surface crystal structure of the magnesium oxide single crystal substrates.

After annealing, the samples are rinsed in deionized water, boiled in acetone for 30 minutes to remove any organic carbon contamination from handling, then they are rinsed again in methanol and blown dry with nitrogen. The first step in the MBE growth is to turn on the cooling water for the fusion cells and the cryo-shroud in the growth chamber. Then we turn on the growth monitoring laser, the RHEED power supply, the RF plasma generator power supply, and the quartz crystal micro balance system.

Magnesium oxide substrates are mounted on three-inch diameter molybdenum sample holders with tungsten spring clips. The first step in loading the samples into the MBE is to turn off the turbo pump and vent the fast entry lock. The sample holder cassette is removed from the fast entry lock and a new sample loaded into the cassette and the cassette is put back into the fast entry lock.

The turbo pump is used to evacuate the fast entry lock. So we usually de-gas the substrate in the fast entry lock at 100 degree, Celsius degree, for 30 minutes. And then, transfer it to preparation chamber for de-gassing at 400 Celsius degree for five hours.

The de-gas sample holder is transferred by a trolley mechanism into the growth chamber where it is loaded into the sample manipulator. The sample is out-gassed in the manipulator at 750 degrees C for 30 minutes. Make sure the cooling water is turned on in the cryo shroud to avoid overheating.

In the case of magnesium nitrite growth, the temperature of the substrate is ramped down to 330 degrees. The growth chamber pressure should now be below 10 to the minus eight Torr. The voltage on the reflection high energy electron diffraction gun, or RHEED for short, is slowly increased to 15 kilovolts and the filament heater current is set at one and a half amps.

The substrate holder is rotated until the electron diffraction pattern shows alignment with the principle crystal graphic axis of the substrate and a clear single-crystal electron diffraction pattern is visible. Standard group three type diffusion cells or low temperature diffusion cells are used for magnesium and zinc. The crucibles were loaded with 15 grams and 25 grams of high purity magnesium and zinc shot, respectively.

The zinc and magnesium source fusion cells are out-gassed at 250 degrees for one hour with their shutters closed. Normally this is done before loading the substrate into the manipulator. After the substrate is loaded, we heat the zinc fusion cell up to 350 degrees C and the magnesium cell to 390 degrees C.The fusion cells are allowed to stabilize for 10 minutes at their operating temperatures before opening the shutters.

The retractable quartz crystal monitor is positioned in front of the substrate inside the chamber. Make sure that the substrate is fully covered by the detector, so that no metal is deposited on the substrate. Input the density of the metal into the quartz crystal monitor controller, so that the controller can read out the thickness of the deposited metal on the quartz crystal sensor.

In order to calibrate the flux, we open the shutter on one of the metal sources and allow the metal flux from one of the infusion cell to deposit on the sensor. The thickness measured by the controller will increase linearly with time as the metal builds up on the sensor. By fitting a straight line to the thickness as a function of time, we obtain an accurate measurement of the metal flux.

Once the flux measurements are completed, close the shutters on the infusion cells and retract the quartz crystal monitor detector from the front of the sample holder. This graph shows the temperature dependence of a flux that the metal source is measured with a quartz crystal monitor. The straight lines are fixed to an Arrhenius relation.

The flux approximately doubles for every 12 degree increase in the source temperature. Turn off the filament current and the high voltage on the RHEED gun to prevent damage to the filament in the presence of a high nitrogen gas pressure in the growth chamber. The next step is to start the nitrogen plasma source.

Open a gas valve on the high pressure cylinder, then slowly open the leak valve until the nitrogen pressure in the growth chamber reaches three to four times 10 to the minus five Torr. Then set the power on the 13.56 MHz RF power supply to 300 watts. The plasma is started with an igniter on the plasma source.

When the plasma has started, a bright purple glow is visible from the view port at the back of the plasma source. Adjust the control on the radio frequency matching box to minimize the reflective power as much as possible. A reflected power of less than 15 watts is good.

Focus the chopped 488 nanometer wavelength argon laser light reflected from the substrate in the growth chamber onto the silicone photo diode, so that an electrical signal can be detected by the lock-in amplifier. This is accomplished by adjusting the angle of the substrate by rotating the substrate holder around two axes and by adjusting the position of the silicone detector and focusing lens that collects the reflected light as shown in this picture. A laser line filter is used to block all the light except for the 488 nanometer light from the argon laser.

The photo diode output is measured with a lock-in amplifier and this single is proportional to the reflectivity of the surface of the substrate. Open the shutter of one of the metal sources. Record the time dependent reflectivity with a computer-controlled data logger.

The growth of an epitaxial film will produce an oscillatory reflected signal associated with thin-film optical interference between the front and back surfaces of the film. When the magnesium nitrite films are first taken out of the MBE, they are yellow, but quickly fade to a whitish color. To protect the films from oxidation and air, it is recommended that an encapsulation layer of magnesium oxide be deposited on top before taking the film out of the growth chamber to protect the film from oxidation when it is exposed to air.

This is especially important for magnesium nitrite and less critical for zinc nitrite. In order to deposit a magnesium oxide encapsulation layer, close the nitrogen gas and switch to oxygen gas and increase the pressure of the oxygen to 10 to the minus five Torr. During the growth of the capping layer, we reduce the RF power to 250 watts.

The plasma starts at a lower RF power with oxygen than with nitrogen. Once the oxygen plasma is running, open the shutter on the magnesium source and monitor the time dependent reflectivity for 10 minutes. This will produce a magnesium oxide film that is about 10 nanometers thick.

An optical reflectivity of the samples can be modeled with this equation. n2 is the index of refraction of the magnesium oxide substrate at 488 nanometers, which is equal to 1.75. Theta naught is the angle of the incident being measured with respect to the substrate normal.

And t is time during the growth process. The optical constants of the film, n1 and k1, and the growth rate are obtained by fitting the reflectivity as a function of time with the equation. The yellow square is an example of the magnesium nitrite film capped with magnesium oxide and the black square is a zinc nitrite film.

The magnesium nitrite is yellow because it has a band gap in the visible, while the zinc nitrite is black because it's band gap is the infrared. The picture on the left is the RHEED electron diffraction pattern for a bare magnesium oxide substrate with the electron beam aligned parallel to the 110 direction. The middle picture is the diffraction pattern from a zinc nitrite film and the picture on the right is from a magnesium nitrite film.

These results show that the crystal structures of the deposited films are oriented in the plane of the substrate as we would expect for epitaxial films. This shows what happens to the electron diffraction pattern when you rotate the bare magnesium oxide substrate in the sample manipulator. This graph shows the optical reflectivity as a function of time during the growth of zinc nitrite and magnesium nitrite films.

By fitting the reflectivity as a function of time to the optical model, you can extract the index of refraction, n, the extinction coefficient, k, and the growth rate, g, for the films. The reflectivity drops with time in the case of the magnesium nitrite films due to surface roughness scattering, which we modeled mathematically by a damped exponential. In this video, we have shown you how to grow epitaxial magnesium and zinc nitrite films by plasma-assisted molecular beam epitaxy.

One of our results is that measuring the optical reflectivity of the samples while they grow is a good way to determine both the growth rate and the optical constants of the film. Unfortunately, our material did not show photoluminescence, either at room temperature or at low temperature, so there is a need to make further improvements in film quality. Experiments in our lab on powder samples provide a clue for how this might be done.

Zinc nitrite powders made by reacting zinc with ammonia at high temperature do show strong photoluminescence. This suggests that using ammonia instead of nitrogen gas as the nitrogen source might be a way to make material with improved electronic properties.