Appeal 2007-2035
Application 09/848,846
“blanket implant”) consisting of boron is applied. Because the polysilicon
completely covers the n wells 21, the halo implant is effectively only applied
to the p type material (Lowrey, col. 6, ll. 57-66; Fig. 6).
Although Lowrey does not actually show the halo implants in Figure
6, we presume that they would be implanted to the p type material in areas
which are not directly underneath the polysilicon layer 45 (i.e., in the areas
adjacent the n-wells 21). In this regard, we presume that the polysilicon
layer 45 functions as the “common mask” as claimed.
It is undisputed that implanting this boron halo implant directly affects
the device’s threshold voltage. Furthermore, Lowrey discusses in the
Background section that in DRAM applications, access devices generally
need a higher threshold than the periphery to optimize dynamic refresh
characteristics. Peripheral transistors are optimized at reduced threshold
values for maximum high speed performance. The conventional solution,
therefore, is to separately adjust the threshold of these two groups of
transistors using a photomasking level (Lowrey, col. 2, ll. 4-13).
When these two teachings are read together, we generally agree with
the Examiner that the skilled artisan would have readily adjusted the
threshold of two different devices (i.e., access and peripheral devices) by
selectively implanting a halo implant using a common mask (i.e., the
polysilicon layer noted above). But we fail to see how these teachings
reasonably suggest providing three or more of such devices with respective
different threshold values using the claimed technique.
At best, Lowrey teaches masking n-channel areas (i.e., areas over n-
wells 21) to selectively apply a halo implant to p-type material. While we
can see how such a technique would achieve two different threshold voltages
6
Page: Previous 1 2 3 4 5 6 7 8 9 Next
Last modified: September 9, 2013