PMDA/ODA Polyimide

Fig. 24. Numbering of the earbon and oxygen atoms in the repeat unit of PMDA/ODA polyimide. Reprodueed by permission of John Wiley and Sons from Ref. [33].

Adhesion of copper films to PMDA/ODA polyimide was determined by peel tests conducted on samples that were prepared by vapor-depositing a thin layer of copper onto the polyimide and then building the thickness of the metal layer to about 18 p,m by electrodeposition of copper. Results of the adhesion measurements correlated well with substrate pretreatment. When the substrate... 

Negative TOF-SIMS speetra of PMDA/ODA polyimide before and after plasma treatment are shown in Fig. 53. The speetra generally show inereas-ing fragmentation as a function of plasma treatment time. This tendency was especially evident for the peak at m/z = 215 (PMDA + H ). 

The first patent of Edwards and Robinson147 claims the condensations of pyromel-litic acid and aliphatic diamine salt to prepare polyimide. Recently, that approach has been revisited, and biphenyl tetracarboxylic and pyromellitic acids give a salt monomer by reaction with 1 mol of an aliphatic diamine (octamethylene diamine and dodecamethylene diamine). The salts were polymerized under 250 MPa at 250°C for 5 h in closed reaction vessels (Fig. 5.32) giving crystalline polymers.148 By reaction of pyromellitic tetraacid with oxydianiline, it has been possible to isolate a monomeric salt. It was polymerized under 30 MPa giving a PMDA-ODA polyimide with water elimination. 

Fig.1 Dynamic mechanical spectra of PMDA/ODA polyimide and poly(phenylquinoxaline),...

Likewise, the mechanical properties of the copolymers were nearly identical or even somewhat enhanced towards the polyimide homopolymer in terms of the modulus and tensile strength values [44,47]. For most of the block copolymers, the elongations to break were substantially higher than that of PMDA/ODA polyimide (Table 4). The shape of the polyimide stress-strain curve is similar to that of a work-hardened metal with no distinguishable yield point... 

It is important to point out that the improvement in adhesion did not result from an increase in the solubihty of the imidized polymer containing the flexible coblock after the Tj cure cycle. In fact, the block copolymers demonstrated less than 2 % swelling (72 h) in the casting solvent, whereas PMDA/ODA polyimide homopolymer swells approximately 20-30 % (72 h). Clearly these data suggest that the improved auto-adhesion results from melt flow at 400 °C [44]. 

Table 14.4. Mechanical Properties of Perfluorinated (10FEDA/4FMPD), Partially Fluorinated (6FDA/TFDB), and Unfluorinated (PMDA/ODA) Polyimide Films ...

Abstract—The adhesion of pyromellitic dianhydride-oxydianiline (PMDA-ODA) polyimide to fluorine-contaminated silicon dioxide (F-SiO,) with y-aminopropyllriethoxysilane (APS) adhesion promoter has been studied as a function of the peel ambient humidity. The peel strength was not affected by the change in peel ambient relative humidity (RH) from 11-17% to 35-60% when APS was used at the interface. Without APS, the adhesion degraded significantly with this change in RH. It was found that although the dip application of APS caused the removal of about 80% of the initial atomic percentage of fluorine on the surface, it could not be totally removed even after several days in water at elevated temperature. 

It is evident from Table 5 that an order of magnitude change in the peel rate when peeling PMDA-ODA polyimide does not make a significant difference in the measured peel strength. Actually, the difference between consecutive table entries may be considered to be within the experimental error. 

The substrates used in this study were (0001) sapphire (A1203), (001) magnesia (MgO), and amorphous fused silica (Si02). All substrates were obtained with surface finish to 0.025 / m, and were cleaned with isopropylalcohol (IPA) prior to PA A or APS application. The surfaces and interfaces after peel test were characterized using X-ray photoelectron spectroscopy (XPS). The PMDA-ODA PAA was cast from NMP solution. Figure l shows the structure of the PAA and the thermally imidized PMDA-ODA polyimide. 

Figure 2 Protomer structures and carbon K-edge Nexafs spectra of a) poly(methyl phenylene oxide) or PMPO b) poly(vinyl methyl ketone) or PVMK c) PMDA-MBCA polyimide and d) PMDA-ODA polyimide.

Figure 2. External reflectance IR spectra (a) PMDA-ODA polyimide, (b) potassium polyamate and (c) polyamic acid. The starting polyimide is 870 A thick and the whole layer is modified. The IR incidence angle is 37° from the sample surface.

Figure 5. Relationship between PMDA-ODA polyimide thickness and imide carbonyl 1R (1740 cm-1) absorbance.

Figure 6. External reflectance 1R spectra of the PMDA-ODA polyimides modified with 1 M KOH aqueous solution at 22 °C for (a) 0 min, (b) 1 min and (c) 10 min. The angle of 1R incidence was 37 0 from the sample surface.

Sample Preparation. PMDA-ODA polyimide films were prepared by spin coating polyamic acid onto silicon wafers and curing to 360°C for 1 hour. The cured film thickness was approximately 10 pm. These samples were treated downstream of an 85% CFi -15% 02 microwave plasma for 30 minutes to produce a fluorinated layer, approximately 600 thick (16). Fluorinated layers less than 600 thick were obtained by additional treatment downstream from an 02 rich (85%) microwave plasma for various times. PTFE films (25 pm thick) were used as received from E. I. duPont de Nemours Inc. 

If the issue of n-complex formation with PMDA-ODA polyimide seems confused, the results of model compound studies have, in places, exacerbated the problem. Several research groups have turned to reactions of formally zerovalent chromium to model the chemistry of PMDA-ODA/Cr (58e, 60. 62) The studies assume that the chemistry of atomic Cr and of certain chromium carbonyl complexes is coextensive. Since atomic chromium has not been available as a benchtop reagent, popular sources of formally zerovalent chromium, disguised as Cr(CO)6 and (CH3CN)3Cr(CO)3, have been used. These compounds readily form (arene)Cr(CO)3 complexes (71). 

There are additional reasons why the (arene)Cr(CO)3 compounds can be misleading as models for PMDA-ODA polyimide/Cr interactions. Formation of (arene)Cr(CO)3 compounds of model PMDA or ODA systems suggests that it-complexes can be formed at the polyimide surface in the early stages of metal deposition. But the (arene)chromium(tricarbonyl) complexes can give little additional chemical or physical insight into surface phenomena because the properties of... 

---