查看文章

The continuing evolution of microprocessors requires the timely development of low-dielectric-constant (low-k) materials to replace dense silica, which with a k value of about 4 has been the dielectric insulator of choice since the inception of the semiconductor industry.[1–4] One leading approach to lower the k value to the ultralow-k range (ie k< 2.2) is to introduce porosity into the silica by taking advantage of the low k value of air (k% 1).[3, 4] Amorphous porous silicas such as sol–gel silica [5, 6] and organic-templated mesoporous silica [7–10] have been suggested as possible low-k materials. High thermal and chemical stability of porous silicas and their similarity to dense silica in terms of chemical composition are some of their attractive features over polymeric low-k materials.[3, 4] Porous silicas can also offer generational extendibility within the ultralow-k range.[2] However, the introduction of porosity into amorphous silicas thus far has been generally accompanied by a dramatic reduction in mechanical strength (commonly manifested in lower values of the elastic modulus, E),[3, 4] which makes their survivability during chemical mechanical processing (CMP) and packaging questionable. Germane to ultralow-k applications, puresilica-zeolites (PSZs) offer several potential advantages over amorphous silicas, including crystalline structure, intrinsically uniform and small pore size, and hydrophobicity. Here we hypothesize that PSZs exhibit a higher E value for any given k value than their amorphous counterparts because of their crystalline nature and provide support for this postulate below.

We have been investigating PSZs as possible low-k …