3.4.1 Savannah S200
Savannah-S200 ALD system from Cambridge NanoTech Inc. has been used for the de-position of amorphous AlOxas shown in figure 23.
Figure 23.Savannah S200 ALD system [Cambridge NanoTech, 2004]
This Savannah system enables digital control of the thin films by growing single layers of films at a time from the nano scale to the micro scale [Cambridge NanoTech, 2004]. The intuitive Graphical User Interface (GUI) enables the complete control of all key system parameters. The system is controlled via a PC. Recipes are used to cycle the system through a series of steps from heating the components to flowing gases to pulsing ALD valves.
Figure 24.Savannah S200 front view [Cambridge NanoTech, 2004]
The ALD valves pulse each precursor into the carrier gas (N2) through the manifold into the reactor where it meets the sample to be coated. Inner and outer heating elements are being used for heating up the reactor. The main chamber has a recessed grove for substrate placement in the main deposition zone. The precursor inlet carries the material into the chamber and the outlet is where it is exhausted.
The pumping exhaust is controlled by the stop valve assembly, which is a heated assembly with a vacuum gauge teed off to one side. A heated trap is designed to fully capture excess or unused precursor and protect the pump from exposure to the organometallic precursor pulse, thus extending the usable life of especially the pump.
The Electronic control box (E-box) controls the I/O on the Savannah ALD system, thus all the valves, heaters, pressure gauges, ALD valves and other electrical options. The E-box is controlled by the system software that runs on the PC as illustrated in figure 25.
The computer and control software communicate with the E-box via a USB connection located at the back of the Savannah system.
Figure 25.Savannah S200 GUI software overview [Cambridge NanoTech, 2004]
3.4.2 Principle of AlOx Formation
Atomic layer deposition is a thin film fabrication method based on successive, self-limiting reactions of gaseous precursors with a solid surface. Trimethylaluminium (TMA or Al(CH3)3 serves as the alumina precursor, whereas H2O is used as the co-reactant.
ALD mechanism can be summarized as consisting of two half cycles, where in the first half cycle, the alumina precursor is introduced and the co-reactant H2O is introduced in the second half cycle. The generic ALD reaction steps are illustrated below in figure 26.
Step 1 & 2: Introduction and adsorption of the alumina precursor, TMA to the QD’s surface.
The electrophile alumina precursor, TMA stored in a high-pressure cylinder is introduced to the low-pressure reaction chamber by opening the cyclinder valve. The amount of
precursor introduced to the chamber is controlled by the vavle opening time, named as the pulse time,t1. The precursor then reacts with the elcetron rich QD surface (e.g., –OH, -COOH, -NH2etc.) liberating methane. The reaction is self-limiting as the precursor does not react with adsorbed alumina.
Figure 26. ALD Mechanism. Schematic diagram of a single ALD cycle with TMA and water precursors on Si substrate.
Step 3: Removal of the un-reacted/excess TMA precursor and reaction products.
The excess or the un-reacted TMA and methane produced from the reaction are purged out of the reaction chamber by an inert gas such as N2. This time is defined as the purge timet2.
Step 4 & 5: Introduction and adsorption of the co-reactant H2O to the surface.
The co-reactant water is introduced to the reaction chamber, where water reacts with methyl groups on the already deposited aluminium atoms forming Al-O-Al bridges, as well as new hydroxyl groups, thus making the surface electron rich again for the accep-tance of the next layer of alumina. The amount of co-reactant introduced to the chamber is controlled by the water pulse time,t3. This step further produces methane.
Step 6: Removal of the un-reacted/excess co-reactant, H2O and reaction products.
The excess co-reactant and methane are purged out of the reaction chamber with a purging time, t4, completing a single ALD cycle and creating a bridged alumina structure on the surface. The process begins again with the introduction of alumina precursor and co-reactant. Atomic layers are built up one after the other.
3.4.3 System Operation
Two primary modes of operation are available for ALD deposition.
1. Continuous Mode: The carrier gas (N2) flow is continuous during the pulsing of precursor and co-reactant into the reaction chamber. Additionally, pumping is al-ways ON, indicating that the stop valve is open.
2. Exposure Mode: Stop valve is closed during the pulsing of the precursors but opened for pumping in between pulses.
The following steps are followed for carrying out the ALD process using the software as in figure 25.
Step 1: Loading a substrate
Upon verification of all setup steps, the system temperature and degassing of all precursor bottles as well as lines, the fabricated QD on a substrate, using one of the fabricating techniques mentioned in section 3.3, is placed on to the heated plate by lifting the lid as in figure 23.
Step 2: Bring the system to vacuum
The ’Pump Reactor’ button is pressed to bring the reaction chamber under vacuum as on the ’Function Panel’ illustrated in figure 25.
Step 3: Run a recipe
Load an already created recipe or create a new recipe as required for the run. A recipe runs a series of steps as shown in figure 27 (left). The ’Recipe Panel’ (figure 25) also displays the status of the recipe as the duration for depositing a certain number of layers of the precursor (figure 27 (right)) that follows a series of steps as illustrated in figure 26.
Step 4: Substrate removal upon completion of ALD
Upon completion of the ALD process, ’Vent’ button in the ’Function Panel’ is pressed to bring the system to atmospheric pressure. Lid is lifted and the substrate is removed.
Figure 27.Recipe table (left), Recipe status (right)