Synthetic resins or natural rubbers -- part of the class 520 ser – Synthetic resins – Polymers from only ethylenic monomers or processes of...
Reexamination Certificate
2001-02-26
2004-11-30
Choi, Ling-Siu (Department: 1713)
Synthetic resins or natural rubbers -- part of the class 520 ser
Synthetic resins
Polymers from only ethylenic monomers or processes of...
C526S348100, C526S346000, C528S244000, C528S401000
Reexamination Certificate
active
06825303
ABSTRACT:
BACKGROUND
This invention relates to precursors and processes for the preparation of thin films which have a low dielectric constant (“∈”), which are stable at high temperatures, and which are used in the manufacture of semiconductor integrated circuits (“ICs”). In particular, this invention relates to new precursors and processes for obtaining a low ∈ (approximately 2.2) fluorinated poly (para-xlylenes) and related films that are compatible with a barrier metal, such as Ta. Ta is currently the most critically useful barrier material for fabrication of future ICs that use copper as a conductor (Ryu et al,
Solid State Technology
, April 1999, pp 53-56).
All commercially available poly(para-xylylenes), PPX (III), such as PPX-N (X═Z═H, n=0), PPX-C (X═H, Z═Cl, n=1), PPX-D (X═H, Z═Cl, n=2) and PPX-Nf (X═H, Z═F, n=4); Desu, et al. in “Low K Material IV,”
Proceeding of MRS. Symposium
, Vol. 511(1998), MRS, pp 39-47)) are prepared by the Gorham method (Gorham et al. U.S. Pat. No. 3,342,754). The Gorham method employs a dimer precursor (I) that cracks under high temperatures ranging from 600 to 680° C. to generate the reactive, gaseous intermediate (II) under vacuum. When contacted with cold solid surfaces, the diradical (II) polymerizes according to reaction (2):
In equation (1), N is an integer denoting the number of molecules (I) and (II), and is at least 10, preferably at least 20, and more preferably at least 50. In equation (2), N is the average repeating unit of polymer PPX (III). It is an integer of at least 10, preferably at least 20, and more preferably at least 50. In both equations (1) and (2), n is zero, or an integer of 1 to 4.
The above commercially available PPX has dielectric constants (∈'s) ranging from about 2.5 to about 3.2, and is thermally unstable at temperatures of higher than 300 to 350° C. Thus, it is not very useful for future ICs that require low ∈ and good thermal stability.
PPX-F (III, X═F, Z═H, n=0, and N as defined above) has an ∈ value of 2.23 and is thermally stable up to 450° C. over 2.5 hours in a vacuum. Therefore, during the past few years, various attempts have been made to make PPX-F from the dimer (I, Z═H, n=0, X═F, N is as defined above) (Wary et al.,
Proc
. 2
nd intl. DUMIC
, 1996 pp. 207-213; Wary et al.,
Semiconductor Int'l
, 19(6), 1996, pp. 211-216) using commercially available equipment. However, these efforts were abandoned due to the high cost of the dimer and to the incompatibility of Ta with PPX-F films prepared by these methods (Lu et al.,
J.Mater.Res
. Vol,14(1), 1999, pp. 246-250; Plano et al.,
MRS Symp. Proc., Vol
. 476 (1998), pp. 213-218 and references herein).
Instead of using the Gorham method, Moore et al. (U.S. Pat. No. 5,268,202) employed a dibromo monomer (IV, Ar═C
6
H
4
, Y═Br) and metallic “catalysts” inside a reactor to generate reactive free radical (V) according to the reaction (3):
N in reaction (3) is the same as N in equation (1).
Further, to lower the cost of starting materials, Moore used a large proportion (>85 to 95 molar %) of a more readily available CF
3
—C
6
H4—CF
3
as the co-monomer to make PPX-F.
With regard to the use of monomer (IV, Ar═C
6
H
4
, Y═Br) in the above reaction (3), several points should be noted.
First of all, in U.S. Pat. No. 3,268,599, Chow discloses an example of the chemistry utilized to generate diradical (V). However, Chow only taught the method to prepare dimer (I, Z═H, X═F, n=0) by trapping (V) in a solvent. The equipment and processing methods Chow employed were not suitable for making the thin films that are useful for IC fabrication using Cu Dual Damascene process.
Secondly, according to Moore, without using the “catalysts,” the above reaction (3) would need a cracking temperature ranging from 660-680° C. However, metallic “catalysts,” such as Zn or Cu, would readily react with organic bromine at temperatures ranging from 300 to 450° C. which are the pyrolyzer temperatures employed by Moore. Formation of metallic halides on surfaces of these “catalysts” would quickly deactivate these “catalysts” and inhibit further debromination as shown in the reaction (3). In addition, the presence of Zn and Cu halides inside a pyrolyzer would cause contamination for the deposition system and dielectric films on wafer.
Thirdly, all commercial equipment used for cracking dimers consists of a Stainless Steel pyrolyzer. This pyrolyzer is not compatible with monomers that contain halogens, such as bromine or iodine (IV, Ar═C
6
H
4
, Y═Br or I), because the iron inside the pyrolyzer's surface would react with these halogens at temperatures higher than about 450 to 500° C.
Lastly, all fluorinated PPX-F films prepared from either monomers or dimers failed to pass a Ta-compatibility test so far. Specifically, when PPX-F was coated directly onto Ta or Ta was coated over PPX-F, film integrity of the Ta layer would fail after annealing these bilayer structures at temperatures ranging from 300 to 350° C. in vacuum or under inert atmosphere for 30 minutes.
In fact, so far there is no disclosure for methods and suitable equipment to make any low ∈ of lower or equal to 2.5 solid (not porous) dielectric films that can pass the Ta-compatibility test. The Ta-compatibility test is considered to be the most critical criterion for acceptance of any dielectric in future ICs that use Cu Dual Damascene process. This is because electroplated copper needs to be annealed at high temperatures ranging from 300 to 350° C., for 30 minutes to one hour, in order to achieve the desirable conductivity. A thermally stable conductive barrier such as Ta is sandwiched in between Cu and the dielectric to prevent diffusion of Cu into the dielectric during the annealing.
In addition, according to Wary et al,
DUMIC Proceeding
, Feb. 8-9, 1999, pp. 272-281, other fluorinated PPX-F, including perfluoro PPX, PF-PPX, (III, X═F, Z═F, n=4, and N is at least 20, and preferably at least 50), in principle, should have a lower ∈ than that of PPX-F (III, X═F, Z═H, n=0, ∈=2.23). PF-PPX is thermally more stable and has higher rigidity than PPX-F. However, starting dimers useful for making these polymers are very difficult to make. These dimers are only available in small quantities(few grams) and are very expensive. Thus, they are not economically feasible for manufacturing ICs.
Beach et al., in an International Patent Application (Number WO97/15541), claimed the preparation of PF-PPX. Beach et al. stated that PF-PPX can be prepared from its corresponding dimer, (—CF
2
—C
6
F
4
—CF
2
—)
2
and that “[t]he methods of producing the above described, (F) ring-substituted AF-4 dimer are not critical to the instant invention, as such methods could be implemented by those skilled in the art.” This International Patent Application does not teach or demonstrate the methods to prepare the dimer(—CF
2
—C
6
F
4
—CF
2
—)
2
, nor does the Application teach the making of PF-PPX.
In fact, for many years several attempts both at Los Alamos Lab, NM (Jorgensen et al.,
JOWOG
Meeting-Org. Coating, Oct. 6-8, 1997) and at Marshallton Research Laboratory, King, N.C., have failed to make the desirable dimer(—CF
2
—C
6
F
4
—CF
2
—)
2
and monomer, YCF
2
—C
6
F
4
—CF
2
Y (Y═Br or I). The difficulty in making the dimer is even greater because it will take more reaction steps to produce the dimer.
Lee et al. in U.S. Pat. No. 6,020,458 disclosed fluorinated chemical precursors and methods of manufacture of polymer thin films with low dielectric constants and integrated circuits comprised of sp
2
C—F bonds and some hyperconjugated sp
3
C—F bonds. The reference discloses precursors for creating fluorinated silanes and siloxanes and fluorinated hydrocarbon polymers.
Thus, the prior art contains no teaching for the methods and suitable equipment to make any low ∈≦2.5 solid (not porous) dielectric films t
Choi Ling-Siu
Dielectric Systems Inc.
Kolisch & Hartwell, P.C.
LandOfFree
Integration of low &egr; thin films and Ta into Cu dual... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Integration of low &egr; thin films and Ta into Cu dual..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Integration of low &egr; thin films and Ta into Cu dual... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3313457