Spin coating has been used for decades as a method to deposit thin films on flat substrates. The machine used to spin coat substrates is called a spin coater. The process that is used, usually involves a substrate rotating at medium to high speeds (200-2000 rpm-rotations per minute) and the drop by drop deposition of the material to be coated The substrate is usually held with the help of powerful vacuum pumps or specifically designed substrate holders.
However, the process of attempting to create a film of a certain thickness is a tedious one that involves multiple stages. After the film has been created one has to characterize the film thickness using techniques such as cross-sectional SEM, polarized light ellipsometry, and surface profilometry, which further may provide erroneous results due to film’s physical character considerations.
First Prototype of Smart Spin Coater made in 2017 ©LabX Scientific Pvt. Ltd.
Once the films are characterized and film thickness and uniformity are known, the entire process has to be repeated at a faster/ slower rate for a longer/shorter duration, with varying the rate of acceleration and rate of deceleration, in order to get desirable quality films. The film thickness is dependent on a variety of factors such as rate of acceleration of substrate, rotational speed, the rate of drying of film, viscosity and surface tension of liquid and substrate-liquid interface and total duration of spin coating. With such a large number of factors affecting film thickness, uniformity, and overall film quality, it is not a surprise that a substantial amount of time and precious materials are used to optimize film thickness. Therefore there appears to be a need to improve upon the process of spin coating, so as to use minimal time and materials to get the best results.
As a first step, we performed a series of experiments to evaluate its performance under wired power. Here, we use an Arduino Uno microcontroller board as the ‘brain’ of the device to take the input and control the output ports such as controlling the motor speed and displaying real-time-parameters on a TFT (Thin Film Transistor type) screen. Speed control of the dc motor is achieved using PWM (Pulse Width Modulation) which is a standard industrial technique to control the averaged dc power output from modulated digital pulses. Although this technique is fairly easy to implement, it poses a challenge that the relationship between the PWM value (varies from 0-255) and RPM of the dc-motor is difficult to establish as it is known to be non-linear. This challenge was solved by the use of Tachometer. RPM v/s PWM graph was plotted and the relationship was obtained from the trendline of the resulting graph. The exponential formula so derived, gave satisfactory control on RPM.
Recent Prototype made in March 2018 ©LabX Scientific Pvt. Ltd.
A major part of the development of the ‘Smart Spin Coater’ is the establishment of a relationship between the film thickness and the transmitted intensity of the incident light. This is essential as most of the nanofilms are not opaque and transmit a sufficient amount of light to be detected by the optical detectors. The optical detectors used here are four 5.2 mm LDRs (Light Dependent Resistors) and as the name suggests, their resistant varies according to incident light (resistance decreases with increase in intensity due to increase in the number of free charge carriers). The primary reason for the use of LDRs instead of photodiodes is the slow response time of LDRs ( typically in ms) as compared to photodiodes (typically in the nanosecond). This helps by reducing the amount of noise in the analog signal from the optical sensors due to low-frequency from LDRs.
The relationship between Transmitted intensity and the analog signal is established by plotting them for multiple samples and observing the trendline. The number of samples used for plotting trendline essentially represents the numbers of data points in the plot and more number of samples means a larger set of points which improves the trendline and gives a better relationship. So, one can say that the device gets better with feedback on the films and it could be argued that it’s ‘learning’ which associates the concept of machine learning with the Smart Spin Coater.
Project Status: Provisionally Patented