NO. SC-802

SC TECHNOLOGY APPLICATION NOTE

Using INS 800 Reflectivity Measurements to:

Quantify Photochemical Conversion of Stepper/Mask and Substrate Sets

Match Photochemical Performance of Process Cells

The optical properties of substrates, photoresists, steppers and masks are not 100% identical and therefore exhibit differences in imaging characteristics and the resulting critical dimensions (CD’s) for process cells (i.e. integrated coat-bake-expose-bake-develop processes). Even slight differences in substrate layer thickness, photoresist layer thickness, lens and mask transmission and reflectivity properties and lamp spectra can cause significant changes in CD'’ and product yields.

Ideally, if the transmission intensity and reflectivity properties of the substrate, substrate layer(s), photoresist layer, stepper optics and masks are exactly replicated cell-to-cell CD replication can be achieved. Unfortunately exact replication is not entirely possible or even practical given current technology. However, it is possible to orchestrate the co-optimization of the substrate layer thickness, the photoresist layer thickness and the exposure energy to approach ideal replication. Co-optimization can be accomplished by means of thickness and reflectivity measurements.

The following discusses the fundamental principles of quantifying and controlling the optical and photochemical performance characteristics of process cells by means of thickness and reflectivity measurements.

During the exposure process, a photoresist film undergoes a chemical change. The chemical change causes a change in the absorption spectra of the photoresist film. A large change in absorption is observed in the 400-nm region for many positive tone photoresists (both I-Line and G-Line types). By monitoring a changing absorption wavelength, one can quantify the degree of photochemical conversion using Beer’s Law:

Where:
A = absorption
%T = Percent Transmission
a = absorbency constant characteristic of the material
b = optical path length
c = concentration of the absorbing species

Since most semiconductor wafer substrates are not transparent to the wavelengths of choice for monitoring purposes, reflected light measurements are used to quantify the photochemical conversion. Therefore, the relationship between the reflectivity measurements and the relative photoactive compound concentration ([PAC]) is described by Equation #2.

Where:
[PAC] = photoactive compound concentration
%R = Percent Reflectivity
z = optical path length (approx. (2)(film thickness))

Using equation # 2, the relative [PAC] may be calculated from the post-exposure reflectivity values shown in Figure 1. The relationship of [PAC] and exposure time is illustrated in Figure 2.

Figure 2: Calculated [PAC] versus Exposure Time

Neglecting the sinusoidal effects of the dynamic change in refractive index, a simple empirical model may be used to estimate the {PAC} for any given exposure. Equation #3 was used to generate the fitted model illustrated in figure 2 above.

Concurrent with the change in absorbency occurring during the exposure process, there is also a change in the refractive index of the exposed area. As a result, there is a corresponding dynamic change in the optical path length. These phenomena are manifested as the sinusiodal pattern evident in the reflectivity signal and derived [PAC] values shown in Figures 1 and 2.

It should be noted that the use of the pre-exposure and post-exposure reflectivity values to quantify relative [PAC] accounts for the functional and optical properties of the substrate, substrate layers, photoresist layer, stepper optics, lamp spectra and the mask simultaneously.

Changes in substrate reflectivity, topography, substrate layer thickness and/or photoresist thickness result in what could be termed a change in the exposure requirement for the process. In effect, these changes cause a change in the post-exposure [PAC] and thus, the resulting CD’s.

The measurements, which are necessary to adequately characterize the entire process, include:

1. Substrate reflectivity
2. Substrate layer thickness
3. Photoresist thickness
4. Pre-exposure reflectivity
5. Post-exposure reflectivity
6. Photoresist thickness versus time (develop process)
7. Dynamic photoresist develop rate
8. Develop "end-point"time

The INS 800 from SC Technology is the only process monitoring tool currently capable of sequentially performing all of the measurements described above, in-line and in real-time.

Case 1 : Constant exposure energy process

If the process is to use a constant exposure energy then the primary property to control and maintain a constant post-exposure [PAC} is the pre-exposure reflectivity. This can be accomplished by controlling substrate layer thickness and/or photoresist film thickness.

The dominant control of reflectivity is the layer thickness. Figure 3 illustrates the relationship of reflectivity with respect to layer thickness. To develop a robust process, the substrate layer(s) and the photoresist layer thickness must be co-optimized to produce a pre-exposure reflectivity versus layer thickness curve. This practice will allow the maximum latitude in the layer thickness while concurrently minimizing the effect of thickness change upon pre-exposure reflectivity.