Ion-Enhanced Dry Etching of Magnetic Multilayers: Post-Etch Cleaning and Effects of UV Illumination
- PDF / 1,589,867 Bytes
- 6 Pages / 417.6 x 639 pts Page_size
- 51 Downloads / 197 Views
structures.[2-9] Due to the fact that the component materials used in these magnetic multilayers (e.g. NiFe-based Giant Magneto Resistive (GMR) materials) form relatively involatile reaction products that cannot be removed readily from the surface in conventional dry etching techniques such as ion milling and reactive ion etching, high resolution pattern transfer has only been realized in high density plasmas. Furthermore, a problem often encountered with ion milling techniques is redeposition onto the sidewall of the features being patterned, leading to degradation of the magnetic performance of the device.[ 10-13] It has been reported that C12/Ar plasma chemistry can produce practical etch rate for magnetic multilayer structures under high ion density conditions,[7,8] but there is one concern with C12/Ar plasma processes. The residual chlorine or chlorinated reaction residue remaining on the feature sidewalls causes corrosion and the resultant considerable degradation in structural and magnetic properties. Recently, some research groups have reported that UV illumination during high density C12/Ar etching of Cu provides a significant enhancement by transforming involatile CuClx reaction products into a slightly more volatile reaction products.[1, 14,15] In this work we examined four different post-etch cleaning methods and magnetic and structural stability over an extended period. The effect of UV illumination on NiFe, NiFeCo, TaN and CrSi was also investigated.
123 Mat. Res. Soc. Symp. Proc. Vol. 585 © 2000 Materials Research Society
EXEPERIMENTAL Full MRAM structures with total thickness -2000 A (80 A NiFeCo/15 A CoFe/35 A Cu/15 A CoFe/8o A NiFeCo/200 A TaI750 A TaN/800 A CrSi) and with additional 2000A thick SiO2 masks, and single layers of Nio.sFeo.2 , Nio.sFeo.13CoO.07 , TaN and CrSi deposited on Si (100) substrates by dc magnetron sputtering from composite targets were used for this experiment. Typical thicknesses of the single layer structures were 3000-5000A and they were masked with Apiezon wax for etch rate measurements. High density plasma etching was performed in a Plasma-Therm 790 reactor. The ICP source power was varied from 100-1000W, while the rf chuck power was varied from 100200W. Electronic grade C12 and Ar were injected into the ICP source at flow rates of 10 and 5 standard cubic centimeters per minute (sccm), respectively. The process pressure was fixed at 2 mTorr. An unfiltered 400W Hg arc lamp was placed on top of the quartz window directly on top of the ICP source, a distance of -20 cm from the samples, and provided the illumination to the single layer structure samples. The magnetic properties of MRAM structures before and after etching were examined using SQUID magnetometry (Quantum Design MPMS-5S) at 4.2K. We measured the patterned area on the dry etching samples by SEM, and normalized the data to the number of spins in the control (unpatterned) and etched (patterned) samples. Each etch rate and magnetization data were examined multiple times under the same conditions to secure repro
Data Loading...
