Spin coater solution
Applying uniform thin films to flat substrates can be achieved by spin coating procedure. An excess amount of a solution is applied to a substrate (manually: using a syringe, or automatically: with a dispense unit and dispense nozzles in the lid of the spin coater). The substrate is then rotated at high speed (e.g. up to 10.000 rpm) in order to spread the fluid by centrifugal force. A machine used for spin coating is called a spin coater, spin processor or simply spinner.
Rotation is continued while the fluid spins off the edges of the substrate, until the desired thickness of the film is achieved. The applied solvent is usually volatile, and simultaneously evaporates. Next to the material characteristics of the photoresist used (solid content and viscosity) the final film thickness is defined by the rotational speed. Keeping the speed constant during the film formation as well as from wafer to wafer is essential for a homogenous layer and a reproducible process.
Although different engineers count things differently, there are four distinct stages to the spin coating process.
√ Deposition of the coating fluid onto the wafer or substrate
This can be done by using a nozzle and pouring the coating solution or by spraying it onto the surface. A substantial excess of coating solution is usually applied compared to the amount that is required.
√ Acceleration of the substrate up to its final, desired, rotation speed
√ Spinning of the substrate at a constant rate; fluid viscous forces dominate
the fluid thinning behavior
√ Spinning of the substrate at a constant rate; solvent evaporation
dominates the coating thinning behavior
Spin coating is widely used in the semiconductor industry, as one of the applications of thin films, creating thin films with thicknesses below 10 nm of even high quality thickness. It is used intensively in photolithography, to deposit layers of photoresist about 1 micrometer thick. Photoresist is typically spun at 1000 to 4000 revolutions per minute for 30 to 60 seconds. Owing to the low values of thickness which can be achieved using spin coating methods, this method is often also employed in the fabrication of transparent titanium dioxide thin films on quartz or glass substrates, such thin film coatings may exhibit self-cleaning and self-sterilizing properties.
Although many polymers with a wide range of weights may be spin coated, one of the easiest and most typical polymers to spin coat is poly(methyl methacrylate), commonly known as PMMA, with a moderate molecular weight (e.g. ~120,000), dissolved in a solvent with some moderate polarity. The solvent 1, 2, 3-trichloropropane was once a traditional solvent commonly used in spin coating, but due to toxicity issues cyclohexanone is generally preferred and produces coatings of comparable quality. More polar solvents such as N,N-dimethylformamide (DMF) are also commonly used. Typically between 10 and 30% (w/w) polymer is dissolved in the solvent. Both the choice of solvent and especially the molecular weight of the polymer significantly affect the viscosity of the solution and thus the thickness of the resultant coating.
For many applications the solution is doped with an active dye or compound. In nonlinear optics, typical doping compounds include Disperse Red and DANS with concentrations up to 30% w/w (dye/polymer). The active doping compound is typically referred to as the "guest," while the inert polymer is known as the "host."
Many substrates may be coated, but in many research applications 1” square glass plates (for many applications often coated with a transparent electrode made of indium tin oxide or ITO) cut from microscope slides are commonly coated for experimental purposes. Prior to spin coating, the polymer-solvent solution must be filtered to remove dust, typically with a 0.45 micrometer filter. Although it is preferable to spin coat in a cleanroom environment (Class 10 or 100), for many R&D experiments, spin coating may be performed in a glove box or fume hood (where the gloves make it awkward to program the spin coater or to handle the substrates). The glass plate is placed upon the spin coater, cleaned successively with acetone and then methanol using lint-free swabs, and followed up with isopropanol, then coated liberally with the polymer-solvent solution by use of syringe or eye-dropper. The plate is spun in at least two stages which may be programmed into the spin coater. During the first stage, the plate is spun at a low to moderate speed 500-1000 rpm for 5-10 seconds to evenly spread the solution. The thickness of the coating is then determined and controlled during the second stage by spinning the coating at a higher speed, between 1500-3000 rpm for anywhere between a few seconds and a minute. These conditions will typically produce high quality coatings of thickness between 2 and 10 micrometers.
Once spin coating is complete, the plate is typically placed quickly onto a hot plate (heated up to somewhere around 100 ºC) for several seconds or minutes to initially evaporate solvent and solidify the coating. The slide is than baked-out for several hours, or typically overnight, in an oven or vacuum oven, at a temperature high enough to sufficiently remove the remaining solvent. Although it is not uncommon to place plates in a petri dish during one or both bake-out steps to protect films from dust, condensation of solvent on the roof of the dish may greatly affect the film smoothness and quality and thus the lid of the dish must be set ajar to allow for the evaporating solvent to escape.
Many other polymers, such as polystyrene or more polar polymers such as polysulfones or polyetherimides, may also be spin-coated. More polar polymers are typically dissolved in N,N-dimethylformamide (DMF). Some polymers are more difficult to coat than others moisture absorption due to environmental humidity can influence the result. Although strict control over the laboratory environment is the best way to improve coating quality, when this is difficult or infeasible one quick-fix to significantly improve film quality is to gently spray a dry inert gas (such as nitrogen, or preferably argon) over the sample during coating to lower the rate of solvent evaporation so that the sample does not have a chance to absorb moisture.