| PATENT NUMBER | This data is not available for free |
| PATENT GRANT DATE | July 15, 2003 |
| PATENT TITLE |
Method for determination of cure and oxidation of spin-on dielectric polymers |
| PATENT ABSTRACT | This invention is a method comprising preparing a sample by coating a thin film of a precursor material, which is free of fluorescent probe molecules onto a substrate and subjecting the precursor material to conditions to attempt to cause cure of the precursor to an organic, aromatic, polymer having a dielectric constant of less than 3.0, exposing the sample to radiation having a wavelength in the range of 200 to 500 nm, detecting a resulting emission of radiation, and comparing the emission to the emission for a known cured, non-oxidized standard for the polymer |
| PATENT INVENTORS | This data is not available for free |
| PATENT ASSIGNEE | This data is not available for free |
| PATENT FILE DATE | December 28, 2000 |
| PATENT REFERENCES CITED |
C. Burgess et al., "Spectrophotometry, Luminescence and Colour; Science & Compliance" 1995, Elvsevier, 1st ed. vol. 6, pp. 178-186, 192-193 and 259-262.* Sun, et al., "Intrinsic Fluorescence Cure Sensor for Reaction Monitoring in Polyurethane," Polymer Preprints, vol. 35, No. 1, pp. 435-436, (Mar. 1994). Grunden, et al., "Cure Monitoring of Styrene Containing Polymers Using UV-Reflection and Fluorescence Spectroscopies," Polymer Preprints, vol. 37, No. 1, pp. 477-478 (Mar. 1996). N.H. Hendricks and K.S.Y Lau, Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chm.) 1996, 37 (1), pp. 150-151. Material Res. Soc, Symp. Proc. (1997), vol. 476 (Low Dielectric Constant Materials III), pp. 121-128. Song, et al., "Cure Monitoring of UV Coatings Using a Fiber-Optic Fluorescence Technique," Polymeric Materials Science and Engineering, vol. 72, pp. 591-592 (1995). "New Developments in Fluorescene Probe Technology for Cure Monitoring," Radtech '98 North American UV/EB Conf., pp. 348-355. Sung, et al., "UV and Emission Techniques for characterization of polymers and ocmposites," Polymeric Materials Science and Engineering, vol. 71, pp. 69-70 (1994). Radtech Report, In Situ Cure Monitoring, vol. 12, pp. 27-32, 1998. "Low-Dielectric Constant Materials-Synthesis and Applications in Microelectronics," Materials Research Society Symposium Proceeding, vol. 381, pp. 165-173 (1995). "Monitoring of Curing of Polyurethane Polymers with Fluorescence Method," Macromolecules, vol. 33, pp. 438-443 (2000). Aronow, et al., "In-Situ Cure Monitoring of Polyurethane/urea Cure with Fiberoptic Fluorescence Spectroscopy," Polymer Preprints, American Chemical Society, vol. 39, No. 2, pp. 642-643, (Aug. 1998). Jager, et al., "Organic Donor-Acceptor Salts: A New Type of Probe for Monitoring Photopolymerization Processes," Macromolecules, vol. 29, pp. 7351-7355 (1996). Conference Proceedings, "Monitoring Conventional and Microwave Curing of Polymers and Composites with the Fluorescence Optrode Cure Sensor (FOCS)," pp. 1668-1671 (1991). Sun, et al., "Cure Characterization in Polyeurethane and Model Urethane Reactions by an Intrinsic Fluorescence Technique," Macromolecules, vol. 29, pp. 3198-3202 (1996). Sung, et al., "Fluorescence characterization of cure and water uptake in polymers and composites," Materials Science and Engineering, A162, pp. 241-247(1993). Application of Fluorescence Spectroscopy in Polymer Science: Quantaitiative Analysis of Cure, Diffusion, Phase Behavior, Sorption and Relaxation Behavior, p. 65. Wang, et al. "Cure Monitoring of Bulk MDI-Based Polyure-Polyurethanes by In-Situ Emission Fluorescence," Polymer Preprints, vol. 37, No. 1, pp. 467-468 (Mar. 1996). Woerdeman, et al. "Cure Monitoring in RTM Using Fluorescence," Plastics Engineering, Plastic Ayalysis, pp. 25-27 (1995). Xu, et al., "Curve Characterization of Bisphenol A Dicyanate Ester by Fluorescence, UV and FTIR spectroscopy," Polymer Preprints, vol. 36, No. 2, pp. 356-357 (Aug. 1995). Song, et al., "Monitoring Degree of Cure and Coating Thinckness of Photocurable Resins Using Fluorescence Probe Techniques," American Chemical Society, Chapter 28, pp. 473-487 (1995). |
| PATENT CLAIMS |
What is claimed is: 1. A method for simultaneous monitoring of the degree of cure and oxidation of an organic, spin-on dielectric material comprising the steps of preparing the sample by coating a thin film of a precursor material, which is free of fluorescent probe molecules onto a substrate and subjecting the precursor material to conditions to attempt to cause cure of the precursor to an organic, aromatic, polymer having a dielectric constant of less than 3.0, exposing the sample to radiation having a wavelength in the range of 200 to 500 nm, detecting a resulting emission of radiation, and comparing the emission to an emission for at least one standard for the polymer, which standard has a known degree of cure and a known degree of oxidation, thereby obtaining an estimated degree of cure and estimated degree of oxidation for the sample. 2. The method of claim 1 wherein the polymer is a polyarylene. 3. The method of claim 1 wherein the polymer is the reaction product of a cyclopentadienone functional compound and an aromatic acetylene functional compound. 4. The method of claim 1 wherein the detecting and comparing steps comprise measuring emission intensity at emission wavelengths corresponding to a cure responsive wavelength, an oxidation responsive wavelength, and a wavelength not responsive to either cure or oxidation; normalizing by dividing each of the emission intensities at the cure responsive wavelength and the oxidation responsive wavelength by the emission intensity at the non-responsive wavelength to obtain a cure intensity ratio and an oxidation intensity ratio; and comparing the cure intensity ratio and the oxidation intensity ratio to a cure intensity ratio and an oxidation intensity ratio for at least one standard having a known degree of cure and oxidation. 5. The method of claim 1 further comprising adjusting cure conditions based on the estimated degree of cure and the estimated degree of oxidation. 6. The method of claim 5 wherein the cure condition, which is adjusted, is selected from temperature, time, ambient environment, or a combination thereof. 7. The method of claim 1 wherein the comparing step comprises comparing either the peak excitation wavelength or the peak emission wavelength of the sample to a peak excitation wavelength or peak emission wavelength of at least one standard having a known degree of cure and oxidation. 8. The method of claim 4 wherein the polymer is the reaction product of a cyclopentadienone functional compound and an acetylene functional compound. 9. The method of claim 8 wherein the cure responsive wavelength is in the range of about 380 to 440 nm, the oxidation responsive wavelength is in the range of 500 to 650 nm, and the non-responsive wavelength is preferably in the range of 460 to 500 nm. 10. The method of claim 8 wherein the cure responsive wavelength is in the range of 390 to 400 nm or 420 to 430 mn, the oxidation responsive wavelength is in the range of 520 to 550 nm, and the non-responsive wavelength is preferably in the range of 470 to 480 nm. 11. The method of claim 1 wherein the substrate comprises a silicon wafer. 12. The method of claim 1 wherein the polymer is a benzocyclobutene based resin. 13. The method of claim 1 wherein the comparing step comprises comparing, on a plot of emission intensity versus either excitation wavelength or emission wavelength, a width of peak emission intensity at half the height of the peak to a width of peak emission intensity at half the height of the peak for a standard having a known degree of cure and oxidation. -------------------------------------------------------------------------------- |
| PATENT DESCRIPTION |
FIELD OF THE INVENTION This invention relates to a method for determining cure and/or detecting oxidation of spin-on dielectric polymers. BACKGROUND OF THE INVENTION The microelectronics fabrication industry is moving toward smaller geometries in its devices to enable lower power and faster speeds. As the conductor lines become finer and more closely packed, the requirements of the dielectrics between such conductors become more stringent. One class of materials being examined as a replacement for the standard dielectric material, silicon dioxide, is spin-on dielectric (SOD) polymers. Unlike the traditional silicon dioxide dielectric layers, these dielectric layers are formed by applying a solution containing the oligomeric precursor to the dielectric polymer, spinning to evenly coat and to remove solvent, followed by curing of the polymer. Curing typically occurs by heating the coated substrate to initiate additional polymerization reaction and/or cross-linking. Achieving an adequate degree of cure is essential to minimizing change in mechanical or other properties later on during fabrication or even during use of a device having a SOD polymer. An undesirable oxidation reaction may also occur if cure conditions (such as exposure to oxygen or other oxidants) are not adequately controlled. Detection of such oxidation reactions may also be important to assuring quality control. Fluorescence has been taught to measure cure, potentially in-situ, of various polymers with the addition of a fluorescent probe (see, e.g., U.S. Pat. No. 5,100,802 and WO86/07456). However, the addition of a probe molecule would be undesirable due to the need for purity and due to high processing temperatures in the microelectronics fabrication process. Intrinsic fluorescence has also been taught as a method for monitoring cure in polyurethanes, (see, e.g., Sun, et al., "Intrinsic Fluorescence Cure Sensor for Reaction Monitoring in Polyurethane," Polymer Preprints, Vol. 35, No. 1, page 435, March 1994, and in polyester/styrene polymers (see, e.g., Grunden, et al., "Cure Monitoring of Styrene Containing Polymer Using UV-Reflection and Fluorescence Spectroscopies," Polymer Preprints, Vol. 37, No. 1, March, 1996). However, it was unknown whether SOD polymers possessed this characteristic. Various analysis methods have been examined to determine cure and oxidation of SOD polymers. FT-Raman analysis can be used to monitor cure of extent of cure, but this method is destructive to the sample. Refractive index may also provide some indication of cure, but it is not very sensitive to oxidation. FT-IR analysis can be used to detect oxidation but is ineffective at detecting cure of important SOD polymers. Therefore, a need remains for an efficient and cost-effective manner for detecting extent of cure and/or oxidation for spin-on dielectrics. SUMMARY OF THE INVENTION The Applicants have discovered a method that not only enables monitoring on-line of SOD for extent of cure but would also allow for simultaneous examination for oxidation. Thus, this invention is a method comprising preparing a sample by coating a thin film of a precursor material, which is free of fluorescent probe molecules onto a substrate and subjecting the precursor material to conditions to attempt to cause cure of the precursor to an organic, aromatic, polymer having a dielectric constant of less than 3.0, exposing the sample to radiation having a wavelength in the range of 200 to 500 nm, detecting a resulting emission of radiation, and comparing the emission to the emission for a known cured, non-oxidized standard for the polymer. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a graph of the fluorescence emission spectra (emission intensity versus emission wavelength) of the polymeric reaction product of a cyclopentadienone functional compound and an aromatic acetylene functional compound cured in a nitrogen atmosphere. FIG. 2 is a graph of the normalized ratios of emission intensity at two emission wavelengths versus time of cure for the polymeric reaction product of a cyclopentadienone functional compound and an aromatic acetylene functional compound cured in a nitrogen atmosphere. FIG. 3. is a graph of the fluorescence emission spectra of the polymeric reaction product of a cyclopentadienone functional compound and an aromatic acetylene functional compound cured in air. FIGS. 4-6 are graphs of the normalized emission intensity at three wavelengths for cures occurring in air and nitrogen versus time for the polymeric reaction product of a cyclopentadienone functional compound and an aromatic acetylene functional compound. FIGS. 7 and 8 are FT-IR spectra for samples of the polymeric reaction product of a cyclopentadienone functional compound and an aromatic acetylene functional compound cured in air and nitrogen atmospheres. FIG. 9 is a graph normalized by forcing peak intensity to a maximum value showing peak emission versus excitation wavelength for CYCLOTENE.TM. BCB based resin at varying degrees of cure in nitrogen atmosphere. FIG. 10 is a graph normalized by forcing peak intensity to a maximum value showing emission spectra for CYCLOTENE BCB based resin at varying degrees of cure in a nitrogen atmosphere. FIG. 11 is a graph normalized by forcing peak intensity to a maximum value showing emission spectra for CYCLOTENE BCB based resin at varying degrees of cure in air. FIG. 12 is a graph normalized by forcing peak intensity to a maximum value showing peak emission versus excitation wavelength for CYCLOTENE BCB based resin at varying degrees of cure in air. FIG. 13 is a graph of emission intensity versus excitation wavelength for the polymeric reaction product of a cyclopentadienone functional compound and an aromatic acetylene functional compound cured in a nitrogen atmosphere. FIG. 14 is a graph of the normalized ratios of emission intensity at two emission wavelengths versus time of cure for the polymeric reaction product of a cyclopentadienone functional compound and an aromatic acetylene functional compound cured in a nitrogen atmosphere. FIG. 15 is a graph of emission intensity versus excitation wavelength for the polymeric reaction product of a cyclopentadienone functional compound and an aromatic acetylene functional compound cured in air. DETAILED DESCRIPTION OF THE INVENTION Generally, fluorescence behavior can be tested in two manners. First, in an excitation scan method, the wavelength of radiation applied to a sample is varied over a range of wavelengths and the sample is observed for emissions of radiation at one or more set wavelengths. This method maps the excited state for a fluorescent molecule, i.e., the excitation spectrum. However, preferably, an emission spectrum is obtained. In this second method, the wavelength of the radiation applied to a sample is held at a fixed value and the wavelengths of emission are collected over a range of wavelengths. This method maps the energy level for the ground state. After obtaining the excitation or emission spectrum for the sample, that spectrum is compared to a spectrum for a sample having a known degree of cure and a known degree of oxidation. This comparison may reveal cure and/or oxidation in at least one of three manners--change in band wavelength position, change in band width, or change in band or peak intensity. For the first two approaches, there is a shift in the wavelength where peak intensity occurs and/or a broadening of the peak or band. Alternatively, or in addition, the intensity of the emission at a given wavelength may have changed. A potential drawback of the latter method is that due to the inherent variability in determining absolute fluorescence intensities, the comparison of the test sample to the standard will need to be comparison of a normalized value for the test sample to a normalized value for the sample. Thus, according to one preferred embodiment, a sample is exposed to radiation having an effective excitation wavelength, and the emission intensity is measured for at least three predetermined wavelengths. The first predetermined wavelength, referred to herein as the cure responsive wavelength, corresponds to a wavelength at which emission intensity is known to vary with cure. The second predetermined wavelength, referred to as the oxidation responsive wavelength, corresponds to a wavelength at which emission intensity is known to vary with oxidation. The third predetermined wavelength, referred to as the non-responsive wavelength, corresponds to a wavelength, which remains relatively unchanged with cure and oxidation reactions. The raw values for intensity at each of the cure responsive and oxidation responsive wavelengths are normalized by dividing the raw values of intensity for each by the value for intensity at the non-responsive wavelength to yield a cure intensity ratio and an oxidation intensity ratio. These ratios can then be compared to the cure intensity ratio and oxidation intensity ratio for at least one standard having a known degree of cure and oxidation. Thus, the degree of cure and oxidation can be estimated based on the difference in these ratios between the test sample and the standard. Use of more than one standard having different known degrees of cure and oxidation will provide more precise information about the degree of cure and oxidation. According to a second embodiment, the first preferred embodiment is used except, rather than plotting emission wavelength versus intensity, the emission (detection) wavelength is held constant and the excitation wavelength is varied. Intensity of emission is then plotted against excitation wavelength and a similar normalization procedure as set forth in the first embodiment is used. According to a third preferred embodiment, the detection wavelength is held constant but the excitation wavelength is varied and the intensity of emission at the detection wavelength for the various excitation wavelengths is determined. The excitation wavelength, which causes the maximum (i.e., peak) emission intensity, is determined and is compared to the excitation wavelength, which causes the peak emission intensity for at least one standard having a known degree of cure and oxidation. The difference in peak wavelength will indicate how close the sample is to the degree of cure and oxidation in the standard. This method avoids the necessity of normalization, but suffers from the fact that the effect of cure reaction on peak shift could be negated or enhanced by the effect of the oxidation reaction on peak shift, thereby yielding confusing results. In fact, unless there is more than one peak that can be analyzed, this method is less accurate in providing simultaneous cure and oxidation information. There is an analogous fourth method, which bases the determination of degree of cure and oxidation on peak shift but holds excitation wavelength constant and plots intensity of emission versus emission wavelength. This method suffers from the same drawback that, unless there is more than one peak, the separate effects of cure and oxidation may be difficult to determine. According to a fifth embodiment, a sample is exposed to radiation having an effective excitation wavelength, the emission spectra is obtained. Band width of a select peak or band is measured at half the band height and compared to that for a standard having a known degree of cure and oxidation. The difference in band width will indicate how close the sample is to the known degree of cure and oxidation. There is an analogous method, which base varies the excitation wavelength and measures emission at a set wavelength. The plot obtained is a plot of emission intensity versus excitation wavelength. A similar band width measurement can be made and compared to band width for a standard. Which methods (excitation scan vs. emission scan and peak shift, width or intensity) are more appropriate may depend upon the characteristics of a specific SOD. Applicants believe the method of this invention would be useful in simultaneously detecting cure and oxidation of various aromatic SOD polymers. Preferably, the SOD polymer is an organic polymer having no, or substantially no, Si atoms in the backbone. Polyarylenes are especially preferred. Examples of polyarylenes include the poly(arylene ethers) (i.e., PAE resins--Air Products) that are described in EP 0 755 957 B1, Jun. 5, 1999 and/or the FLARE resins made by Allied Signal Corp. (see N. H. Hedricks and K. S. Y Liu, Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chm.) 1996, 37(1), pages 150-1; also J. S. Drage, et al., Material Res. Soc, Symp. Proc. (1997), Volume 476 (Low Dielectric Constant Materials III), pages 121-128 and those described in U.S. Pat. Nos. 5,115,082; 5,155,175; 5,179,188 and 5,874,516 and in PCT WO91/09081; WO97/01593 and EP 0755957-81. Alternatively, the polyarylene may be as disclosed in WO97/10193. Preferably, however, the polyarylene is the reaction product of a cyclopentadienone functional compound and an aromatic acetylene functional compound, as disclosed in those disclosed in U.S. Pat. No. 5,965,679, incorporated herein by reference. Fluorescence analysis for simultaneous detection of cure and oxidation for these latter preferred polyarylenes works particularly well as these polymers display peaks at distinct wavelengths for oxidation and cure, respectively. The precursors (i.e., curable oligomer or polymer) are preferably of the general formula: [A].sub.w [B].sub.z [EG].sub.v wherein A has the structure: ##STR1## and B has the structure: ##STR2## wherein EG are end groups having one or more of the structures: ##STR3## wherein R.sup.1 and R.sup.2 are independently H or an unsubstituted or inertly-substituted aromatic moiety and Ar.sup.1, Ar.sup.2 and Ar.sup.3 are independently an unsubstituted aromatic moiety or inertly-substituted aromatic moiety, M is a bond, and y is an integer of three or more, p is the number of unreacted acetylene groups in the given mer unit, r is one less than the number of reacted acetylene groups in the given mer unit and p+r=y-1, z is an integer from 0 to about 1000; w is an integer from 0 to about 1000 and v is an integer of two or more. Such oligomers and polymers can be prepared by reacting a biscyclopentadienone, an aromatic acetylene containing three or more acetylene moieties and, optionally, a polyfunctional compound containing two aromatic acetylene moieties. Such a reaction may be represented by the reaction of compounds of the formulas (a) a biscyclopentadienone of the formula: ##STR4## (b) a polyfunctional acetylene of the formula: ##STR5## (c) and, optionally, a diacetylene of the formula: ##STR6## wherein R.sup.1, R.sup.2, Ar.sup.1, Ar.sup.2, Ar.sup.3 and y are as previously defined. The definition of aromatic moiety includes phenyl, polyaromatic and fused aromatic moieties. Inertly-substituted means the substituent groups are essentially inert to the cyclopentadienone and acetylene polymerization reactions and do not readily react under the conditions of use of the cured polymer in microelectronic devices with environmental species, such as water. Such substituent groups include, for example, F, Cl, Br, --CF.sub.3, --OCH.sub.3, --OCF.sub.3, --O--Ph and alkyl of from one to eight carbon atoms, cycloalkyl of from three to about eight carbon atoms. For example, the moieties, which can be unsubstituted or inertly-substituted aromatic moieties, include: ##STR7## ##STR8## wherein Z can be: --O--, --S--, alkylene, --CF.sub.2 --, --CH.sub.2 --, --O--CF.sub.2 --, perfluoroalkyl, perfluoroalkoxy, ##STR9## wherein each R.sup.3 is independently --H, --CH.sub.3, --CH.sub.2 CH.sub.3, --(CH.sub.2).sub.2 CH.sub.3 or Ph. Ph is phenyl. For these preferred SODs, i.e., the reaction product of cyclopentadienone functional compound and aromatic acetylene compounds--an emission spectrum is preferably attained after applying radiation having a wavelength in the preferred range of 300 to 450 nm, more preferably 300 to 400 nm, and most preferably 330 to 390 nm. The cure responsive emission wavelength is in the range of about 380 to 440 nm, preferably in the range of 390 to 400 nm and/or 420 to 430 nm. The oxidation responsive wavelength is in the range of 500 to 650 nm, preferably 520 to 550 nm. The non-responsive wavelength is preferably in the range of 460 to 500 nm, more preferably 470 to 480 nm. To ensure a substantial degree of cure, the cure intensity ratio at wavelengths in the range 390 to 400 nm is preferably less than about 0.5 or in the range 420 to 430 nm less than about 1.4. The oxidation intensity ratio is preferably less than about 0.5 for oxidation responsive wavelengths in the range of 520 to 550 nm. The methods outlined may also be applicable to BCB based polymers, which are the reaction product of monomers comprising (a) a cyclobutarene monomer having the formula: ##STR10## wherein B.sup.1 is a n-valent organic linking group, preferably comprising ethylenic unsaturation, Ar.sup.1 is a polyvalent aromatic or heteroaromatic group and the carbon atoms of the cyclobutane ring are bonded to adjacent carbon atoms on the same aromatic ring of Ar.sup.1 ; m is an integer of 1 or more; n is an integer of 1 or more; and R.sup.1 is a monovalent group. The preferred BCB based polymers are the reaction product of the monomer (a) and has the formula ##STR11## wherein each R.sup.3 is independently an alkyl group of 1-6 carbon atoms, trimethylsilyl, methoxy or chloro; preferably R.sup.3 is hydrogen; each R.sup.4 is independently a divalent, ethylenically unsaturated organic group, preferably an alkenyl of 1 to 6 carbons, most preferably --CH.sub.2.dbd.CH.sub.2 --; each R.sup.5 is independently hydrogen, an alkyl group of 1 to 6 carbon atoms, cycloalkyl, aralkyl or phenyl; preferably R.sup.5 is methyl; each R.sup.6 is independently hydrogen, alkyl of 1 to 6 carbon atoms, chloro or cyano, preferably hydrogen; n is an integer of 1 or more; and each q is an integer of 0 to 3, some of which are commercially available under the trade name CYCLOTENE from The Dow Chemical Company. BCB based polymers show an excitation wavelength peak shift to lower wavelengths with cure or an emission wavelength peak shift to higher wavelengths with cure. If oxidation is also occurring, the shift in peak emission wavelength to higher wavelengths is much larger. The width of the emission and excitation bands can also be used to detect oxidation since the width of these bands becomes much broader with oxidation. The fluorescence spectrum of SOD films can advantageously be obtained by using flexible fiber optic probes attached to the fluorimeter. The center of the fiber optic bundle carries the excitation light to the sample, while the emitted light is returned to the emission monochromator of the fluorimeter via an outer ring of bundled fibers. This arrangement is advantageous since (1) it provides a very small area of the sample to be probed for cure and oxidation, (2) it allows the remote measurement of cure and oxidation, and (3) it allows for easy incorporation of the fluoresence emission technique to be automated into a tool capable of mapping the cure and oxidation across a sample. Spatial resolution of cure and oxidation can provide valuable feedback on curing tools (hot plates, vertical tube furnaces), in order to determine their temperature uniformity, as well as the quality of the nitrogen purge environment, in addition to ensuring uniformity in physical properties to improve device yield and performance. The feedback could also be used to increase cure time and/or temperature if the emission indicates incomplete cure. |
| PATENT EXAMPLES | available on request |
| PATENT PHOTOCOPY | available on request |
|
Want more information ? Interested in the hidden information ? Click here and do your request. |