Silicon dioxide, plasma etching

A representative sample of terpolymers was exposed to a variety of etchants for polysilicon and silicon dioxide, and the results are given in Table V. The ratio of the etch rate of the substrate to the etch rate of the resist must be at least 2 1 for the resist to be a viable etch mask. Inspection of Table V, shows that the materials examined are unacceptable for only the QFj — CF3CI (4 1) plasma. The etch rates are comparable to those for PMMA the a-keto-oxime exhibits essentially no effect on that rate and the nitrile affords a slight decrease in the plasma etch rate. The etch rates of some commercially available materials are shown for comparison. 

Figure 9. Dependence of silicon and silicon dioxide etch rates on the percentage of Hz in CF4-H2 plasmas, (seem is standard cubic centimeters per minute.) (Reproduced with permission from reference 118. Copyright 1979 The Electrochemical Society, Inc.)...

These authors found that it was possible to deposit amorphous films whose Ta concentration ranged from 10 to 80 mol % by changing the reactive gas mix. Another feature of the films was that under certain conditions they contained substantial quantities of chlorine and hydrogen. Also, they did not adhere to either silicon or silicon dioxide after annealing (argon atmosphere for 1 hour at 900°C). When the substrates were dry etched in an HCI plasma for 2 minutes, they adhered to the substrate even after annealing. Since this etch removed about 50 A, it appears that the native oxide on the silicon and/or some other surface impurities on both the silicon and silicon dioxide were causing the lack of adhesion. 

Contact cuts were plasma etched through the polyimide (using a photoresist mask) and buffered HF was used to etch the silicon dioxide underneath. After aluminum was deposited and patterned, both test and control wafers were sintered in forming gas at 400°C for 10 minutes. A CPI strip was fabricated beside each FET with four aluminum contacts to allow CPI 4 point conductivity measurements. 

A possible mechanism of formation of a sidewall passivation film in the case of silicon etching in a HCI/O2/BCI3 plasma is shown in Fig. 18 [77]. SiCf Hv-type byproducts are sputtered away by ion bombardment from the bottom of the trench. A portion of the sputtered flux strikes and sticks on the sidewalls on the trench. Oxygenation of the byproducts on the sidewalls results in a silicon dioxide type of film that resists etching. The sidewalls do not receive any appreciable ion bombardment and hence, depending on conditions, a rather thick inhibitor film may be formed. 

In plasma etchers, specific radicals are selected from the mix of the species generated within the chamber to effect the etching action. For the specific case of species generated from CF4 gas within a plasma chamber, for example, the fluorine radical (F) is selected by means of an appropriately configured perforated aluminum shield or other contraption that blocks the other species from reaching the wafer. In this way, etching of the wafer proceeds only by the reaction of the fluorine radical. Substrates such as silicon, silicon dioxide, and silicon nitride are readily etched by this technique. ... 

In one approach" " thin-film aluminum is vapor deposited or sputtered onto the silicon wafer and photoetched to form a mask to etch the vias. The vias are plasma etched using fluorine or a gaseous fluoride compound such as carbon tetrafluoride but etching is stopped short of going completely through the silicon. The aluminum mask is then removed and silicon dioxide is deposited over the entire wafer to isolate the silicon from subsequent deposition of copper. A tie layer of Ti/W is generally deposited prior to copper deposition. The vias are then filled with either a conductive... 

CF plasma etching of silicon and silicon dioxide is widely used in the electronics industry (47). When CF is RF excited at pressures above... 

CF4 plasma etching tends to etch silicon dioxide faster than silicon and is commonly used to preferentially etch SiOj (47,49,50). When oxygen is added to CF4, the etch rate of Si becomes equal to SiOj and is increased over that of pure CF4. Figure 18 shows a survey spectrum of... 

The application of photoelectron spectroscopy to semiconductors euid semiconductor device structiires has been demonstrated through its application to the silicon dioxide-silicon interface, ni-V compound semiconductor metal jimctions, and plasma etching residues. Through the use of profiling methods, chemical depth profiles are obtained and are extremely useful to device structural studies. Many methods such as in-situ film growth, film deposition, air-lock mounted pretreatment chambers, etc., have been employed to study semiconductor surfaces and device structures. 

Figure 9. Formation of micropillars from a solid precursor (schematic). (1) A175 nm thick thermal silicon dioxide film is grown on a silicon wafer. (2) A photoresist is applied by spin coating. (3) The oxide film Is selectively exposed, and then (4) selectively plasma etched with CHFs. (5) The photoresist is stripped, and (6) the silicon water is plasma etched with CI2/BCI3. (7) The residual oxide film is removed with a HF etch, and the silicon miaopillars as-formed are ready for use. Courtesy of Dr. A. Perez, Cornell University, Ithaca, NY.

Kushner, M. J. A kinetic study of plasma-etching process. I. A model for the etching of silicon and silicon dioxide in carbon fluoride (CnFm)/hydrogen and carbon fluoride (CnFm)/oxygen plasmas. J. Appl. Phys. 53(4), 2923-2938,1982. 

---