Process of photolithography

The fundamental of photolithography process is to put a layer of Photo-Sensitive Resistor on the wafer surface, usually by spinning. We call this step as PR (photo resistor) Coating.

Then by using the parallel light source to irradiate through the retile with defined patterns on this Photo-Sensitive Resistor proceeds selective exposure. The light wavelength is significant to the line pitch. The wavelength is smaller; the line pitch can go narrower more easily. After exposed, the patterns on reticle are completely transferred to the photo resistor on wafer surface. We call this step as Exposure.

By the different types of photo resistors react to the light, there are two types of photo resistors. One is positive photo resistor; the other is negative photo resistor. What means positive and negative? The positive photo resistor represents that after exposure, the photo resistor can be vanish out with the area where reacted with the light by solvent developing is called positive photo resistor. On the other hand, the negative photo resistor is that when after exposure, the photo resistor is been vanish out with the area where are not reacted with the

light by solvent developing is called negative photo resistor. (See Fig. 1.1). Basically speaking, negative photo resistor is cheaper, but the resolution is worse than positive photo resistor.

Opposite, the positive photo resistor is more expensive, but the resolution is better than negative photo resistor. Especially in near years, under seventy nanometer process, the photo resistor used for ArF (Argon fluoride) LASER is hundreds thousand NT dollars a gallon.

The major steps of photolithography process are (see Fig. 1.2):

*Reference [1],[2],[3]

1.3.1 Primer:

Because of the silicon will oxidate with oxygen to form the silicon oxidation layer. And the chemistry characteristic of silicon oxidation layer is hydrophilic, so water may stick on the silicon oxidation layer. Nevertheless, the chemistry characteristic of positive photo resistor is hydrophobic. Therefore, it is necessary to put HMDS (Hexamethyldisilazane) on the silicon oxidation layer for increasing the adhesion between positive photo resistor and silicon oxidation layer. Due to the silicon oxidation layer is hydrophilic, steams will stick to the wafer surface. If we did not remove the wafer molecule from the surface, it is hard to make combination with HMDS and wafer surface. So before HMDS priming, it has to bake with temp about centigrade 150 for tens of seconds. We call the HMDS coating step as Priming, HMDS is the Primer.

*Reference [1],[2],[3]

1.3.2 Resistor Coating:

Nowadays photo resistor coating is to vacuum the wafer on the chuck of the resistor coater, then dispense the resistor on the wafer and spin the resistor uniformly on the whole wafer with suitable spin speed, may be one to five thousand rpm (round per minute). The spin speed depends on the requirement of the process and the characteristics of the resistor.

Although positive and negative photo resistor can be used for photolithography, but the organic solvent possesses the Swelling Effect, it is difficult to the control when the line pitch is narrower than one micrometer. So, for line pitch controlling, positive resistor much conforms to the process demand.

*Reference [1],[2],[3]

1.3.3 Soft Bake

In the process of resistor coating, when we have put the photo resistor on the wafer, it needs the second baking. We call this soft bake or pre-bake. The purpose of the soft bake or pre-bake is to drive out the most solvent of the photo resistor, besides transferring the photo resistor from liquid state to solid state. The soft bake can also increase the adhesion for the photo resistor on the wafer. After soft bake, the thickness of the photo resistor thin film might have the shrinkage about ten to twenty percentages. The baking time and temperature degree is determinate by the photo resistor types and the process requirement. If under bake, no matter lower temperature or not enough time might cause the resistor peel from the wafer. It can also influent the resolution of the pattern. Over bake might make the photo resistor get gather too early, and cause the photo resistor’s sensitivity down.

*Reference [1],[2],[3]

1.3.4 Alignment and Exposure

It is no doubt that alignment and exposure are the most important steps of the photolithography for whole IC manufacture processes. This step determinates that if we can transfer the designed patterns on the reticle to the photo resistor on the wafer successfully. The exposure process is so like to make a picture with a camera. Patterns on reticle expose to the photo resistor on the wafer just like expose the images on the camera’s negative. The difference is the resolution of exposure system of the machine that we called stepper or

scanner is much superior to the camera. And this is why the scanner or stepper is so expensive, especially the scanner. The name of stepper is come out with a machine repeats and step.

Because the step can only expose one small area of the wafer once, so we need to repeat this step so many times until the whole wafer is exposed.

The alignment is executed for the overlay requirement. For mass production, the alignment time is the less the better; but for overlay quality, the alignment is the more the better. It needs to make a reasonable choice. Fig. 1.3 shows a sketch map of an alignment and exposure system.

The light source: In 2000s, the light sources for IC use are I-line (wavelength=365nm) emitted from the mercury lamp; DUV light, excimer LASER of KrF (krypton fluoride, wavelength=248nm) and excimer LASER of ArF (argon fluoride, wavelength=193nm). The wavelength is directly influent the narrowest line pitches. According to the Rayleigh criteria,

*Reference [1],[2]

We are easily able to understand the relationship between the line pitch and the wavelength.

This is why nowadays photolithography process improvements are finding out a short wavelength with all developers’ strength.

*Reference [1],[2]

1.3.5 PEB

This step has different meanings to difference generation of photo resistors. The classification of the photo resistor we discussed before is positive and negative. By using different light source, the photo resistors are different. We use a CAR (chemical amplifier resistor) for DUV light sources such as KrF and ArF. Because of the excimer light power is smaller than I-line. It needs the chemical method to enhance the exposure light reacts with the photo resistor.

So, for I-line the PEB (Post Exposure Bake) is to minimize the standing wave effect of photo resistor. The standing wave effect is that when exposure light reflected from the interface of photo resistor-substrate, it produces interference with the injection exposure light and performs constructive and destructive interferences at different depth of photo resistor.

The standing wave effect will perform over exposure area and under exposure area in the photo resistor layer. The main purpose of I-line photo resistor is to use the heat to make the photo resistor reflow for decreasing the dispatches of the photo resistor.

For excimer LASER CAR photo resistor, this thermal process provides the activation energy for the light acids diffuse to its suitable width for covering the insufficient of the LASER power. CAR photo resistor will release an H+ after reacting with exposure light, the H+ can make the chain effect with other non-exposure area’s photo resistor by thermal diffusion just like these non-exposure area are exposed. Then these decomposed bonding from the photo resistor will be taken out by development and the patterns are remained.

*Reference [1],[2]

1.3.6 Development

The development removes the photo resistor where we do not need. And perform the defined reticle patterns on the wafer. For positive photo resistor, the exposure part will discompose into the developer. They usually use TMAH ((CH3)4NOH) to be the positive photo resistor’s developer. And Xylene is for the negative photo resistor use. During the development process it is necessary to keep the constant temperature or the higher temperature will cause higher chemical reaction rate. It leads to over development and CD (critical dimension) loss. On the other hand, lower temperature will cause lower chemical reaction rate. It leads to under development and incomplete development. For development process temperature control is very important for developer and wafer. Different photo resistors use different developers and different develop temperature.

*Reference [1],[2]

1.3.7 PDB

The step follows development process is PDB (post exposure bake). It is also called hard bake. Hard bake is to rid of the residual solvent in the photo resistor and increases the strength of the photo resistor by further polymerization for improving the resistance from the etching process or ion implantation. And by the thermal dehydration will enhance the adhesion between photo resistor and wafer.

*Reference [1],[2]