The Impact of Temperature and Concentration on SC2 Cost and Performance in a production Environment
- PDF / 1,639,149 Bytes
- 8 Pages / 414.72 x 648 pts Page_size
- 69 Downloads / 182 Views
Under these
conditions, the clean is extremely effective at removing metallic contamination from wafer surfaces. The low pH of the clean leads to good metal oxide solubility, while the chloride ion acts as an effective complexing agent. In recent years, industry interest has developed in moving towards a more dilute SC2 formulation. Numerous studies [2-8] have demonstrated that solutions with very low HCI concentration can successfully control metallic levels on wafer surfaces. Further, the suggestion is made that extremely low (< I02 M) acid concentrations provide a particle advantage relative to concentrated mixtures [4]. The need for hydrogen peroxide in the SC2 mixture has been questioned, with some researchers [2-5] advocating its complete removal from the clean. The primary role of peroxide in SC2 is the oxidation of noble metals. Argument is frequently made that with high purity chemicals, noble metals are no longer an issue. If metals requiring a high oxidation potential for removal are not present, peroxide offers no benefit; Norga and Kimerling [9], in fact, point out that low pH solutions are actually less effective at removing Fe when the oxidation potential is high. While Au and Ag are no longer considered common cleanroom contaminants, Cu remains a concern. Christenson and Smith [6] have demonstrated difficulty in removing Cu residue without a peroxide presence; with no peroxide in their study, only 15-25% of Cu was removed in an HCI clean. When a very small volume of peroxide was added to the clean ( - 2) to induce a wafer/particle repulsion. At the same time, ionic strength must be maintained at a relatively low level; the strength and the range of any existing electrostatic interactions will be significantly reduced in a high ionic strength solution. Hurd, et al [4] have demonstrated that if HCI concentration is kept below .01 M, particle deposition levels can be significantly reduced. The recipe being evaluated in this study has an HCI concentration of approximately 0.55 M. At this concentration level, the particle protection afforded by a significant potential energy barrier should not exist. Despite this fact, testing has indicated that particle addition onto bare Si surfaces is significantly reduced with the 1:1:20 recipe. 3 shows particle results 140 _Figure from one test series. This data 120 represents information from two ,100 Bare Si 200 A Oxide separate runs with each SC2 formulation; all data in Figure 3 was s80 S60 collected in the same wet station. mThe
C4
•:•addition S20
tendency for reduced particle
m•
-20 1:1:6
1:1:20
1:1:6
1:1:20
Figure 3: Particle performance of 1:1:6 85 C SC2 and 1:1:20 60 C SC2 on bare wafers and thermal oxide wafers
onto bare Si with the 1:1:20 recipe was noted each time this test was performed. Results on oxide, however, have been inconsistent. In one test series, particle density on oxide was actually observed to be higher on wafers which had received
the 1:1:20 clean (see Figure 4).
522
The bare silicon and oxide data in Figure 4
100
S80 S606
Data Loading...
