Wafer testing

Film thickness metrology has been integrated into CMP tools, which reduces or eliminates lost productivity due to test wafer queuing delays. 

It should be noted that approximately 1% of the APS used in the last experiment appears to be APS monomers before dilution. Could this low monomer concentration be responsible for adhesion It should be noted that the residual monomers present in this highly-oligomerized solution correspond to a 0.001 vol % solution if one were to remove the oligomerized APS. As a result of this observation, the dependence of the adhesion of thin films to native-oxide silicon wafers as a function of the concentration of the APS under conditions of T H stress was investigated. Adhesion studies were performed using APS solutions with concentrations that varied from the industry standard of 0.1 vol % down to 0.00001 vol %. The test wafers were prepared and exposed to T(200) and T(500) conditions as discussed above. Adhesion was measured by 90° peel test, as discussed above. The results of this study are presented in Fig. 13. The Lx-axis is APS concentration which decreases from left to right. The y-axis is the adhesion in the units of g mm"The three curves are the results at T(0), T(200) andT(500). 

FIGURE 7.4 Material removal rate on 8" tungsten blanket test wafers as a function of hydrogen peroxide concentration (from Ref. 12). 

FIGURE 7.20 Removal rate versus downforce on 8" Cu blanket test wafers polished at 75/65 rpm table/carrier speed and 200ml/min slurry flow rate on Strasbaugh u-Hance polisher. Diamond data points indicate the removal rate values with the organic particles, and square data points denote removal rate values for silica particles, both polished under identical formulation and abrasive concentration (from Ref. 110). 

TABLE 7.3 Removal Rate of 8" Blanket Test Wafers of Different Snbstrates at Similar Process Conditions with the Organic Abrasive Slurry [109]. 

FIGLfRE 7.22 Representative surface quality image of Cu blanket test wafer polished with a slurry consisting of organic abrasive particles. Surface roughness RMS value of... 

FIGURE 9.4 A cross section of a typical testing wafer in which tungsten lines are used as an interconnect (from Ref. 13). 

REPRESENTATIVE TESTING WAFERS FOR STI PROCESS AND CONSUMABLE EVALUATIONS... 

Next, test wafers were fabricated with copper deposited on titanium. The copper was subsequently etched from half of the wafer, leaving titanium exposed on the etched half. The wafers were polished for 30 seconds to determine if the copper polished from one half of the wafer interacted with the titanium on the other half. Copper and titanium polish rates, measured on the same wafer, were found to be 520 nm/min and 252 nm/min, respectively, giving a selectivity of 2.1. The reduction in selectivity suggests that copper polished from the wafer interacted with titanium to increase the polish rate of the titanium. In addition, the titanium appears to have caused a decrease in the polish rate of copper. 

The test wafers had the following process flow. After the deposition of 200 nm of oxide the metal stack with a total thickness of 2200 nm was sputtered. The wafers... 

There were two types of wafers used in this experiment blanket test wafers and fully integrated production wafers. The blanket test wafers used in this experiment were P(IOO) silicon deposited with 7500A of TEOS oxide followed by a sputtered TiN adhesion layer and a deposited SOOOA W film. The tungsten film used in this experiment was deposited in an Applied Materials CENTURA reactor, using WFe reduced with SiHt and H2. After W deposition, W CMP was performed using different hardware and process parameters. 

The RPL values obtained by puddle development for PC-129 and HPR-204 are both larger than those obtained by dip development, thus indicating that the dip method is better with the developers employed. In addition, resist image clearing uniformity across the 3" test wafers was better for the dip developed wafers than for those puddle developed. Furthermore, the dip development method yielded statistically better wafer to wafer development reproducibility for PC-129. This is not to say that puddle development is not a useable entity but that dip is preferred. In fact, Leonard and coworkers (6) have developed a puddle development method with production capability. Similarly, the RPL data for KTI-II favors dip development over that of spin/spray, therefore, dip development is the favored development method overall for these three example resists. 

The patterned and subsequently pyrolyzed polyimide films on silicon wafers adhered well to the substrate without cracking or peeling (Figure 2). The measured conductivity of the CPI was 10 (ohm-cm). This value is about ten times lower than that reported in reference (2) and is likely due to the relatively short test wafer heating period of 30 minutes. It is suspected that if higher temperatures or extended times during heat treatment are used, conductivity should increase by a factor of 10. 

The easiest way to measure stress in thin films after deposition is to analyze the change in the radius of curvature of the wafers before and after film deposition on one side, as first proposed by Stoney [7]. However, this technique usually requires the use of test wafers. After complete processing of the wafers, the stress can be obtained by measuring the deflection of membranes or indicator structures [8], To measure compressive stress, the buckling technique on double-side supported bridges [9] and harp-like structures [10] can be applied. 

Figure (5) presents sample experimental current versus potential data for the test wafer with a 2cm circle exposed to plating. Similar data was also obtained for the test wafer with a 3-4cm ring exposed to plating. Limiting current density values gleaned from... 

Two-dimensional, axisymmetric and three-dimensional simulations are conducted that mimic the experiments by forcing the wafer Cu concentration to zero across the same exposed areas. The one-dimensional model is independent of radial variations, and so it is not considered here. Table (3) compares the model predictions of k with experimental values for the 2cm circle test wafer. The two-dimensional model shows poor agreement with the data for the no rotation case. The impinging jet flows near the center of the wafer enhance the mass transfer in this region, and the two-dimensional model is incapable of capturing these effects. However, as wafer rotation effects dominate, two-... 

Figure 5. Experimental limiting current density data for test wafer with a 2cm circle exposed to plating.

The gap filling capability of the bath was also tested throughout the experiment. Test wafer used for this study is 0.3 um trenches with aspect ratio of S.S. Fig. 6 contains three SEM pictures of the sample deposited in the fresh bath, 5-turnover, and 10 turnover with the same process recipe. All three samples have seamless filling of copper film. These results indicate that the gap filling capability of the bath remains good as bath ages. 

The experimental apparatus Includes 1) a rotating disk system (RDE) to mount a test wafer, control the rotating speed, and adjust the distance between the cathode and anode ... 

Figure 6.11 Pattern type and pattern density within a die of an SKW7-2 test wafer (Fan, 2012) (a) layout of a die on SKW7-2 wafer (MIT standard oxide CMP characteization layout). A P preceding a number indicates a pitch structure with 50% density, with the number following in microns. All other numbers are localized densities, with the number indicating the density. Density stmctures have a fixed 100 pm pitch, (b) Topography of the 70% STEP array in a die.

Design of the experiment. Test wafers or consumables are designed to assess the polishing performance and enable the study of target factors in a CMP process. Process parameters are selected to explore the target polishing requirements. 

---