Resistance of POSS Polyimide Blends to...
0 downloads
55 Views
2MB Size
Research Article www.acsami.org
Resistance of POSS Polyimide Blends to Hyperthermal Atomic Oxygen Attack Min Qian,† Vanessa J. Murray, Wei Wei, Brooks C. Marshall, and Timothy K. Minton* Department of Chemistry and Biochemistry, Montana State University, 103 Chemistry and Biochemistry Building, Bozeman, Montana 59717, United States S Supporting Information *
ABSTRACT: Copolymers of polyhedral oligomeric silsesquioxane (POSS) and polyimide (PI) have shown remarkable resistance to atomic oxygen (AO) attack and have been proposed as replacements for Kapton on the external surfaces of spacecraft in the harsh oxidizing environment of low Earth orbit (LEO). POSS PI blends would be an economical alternative to the copolymers if they also resisted AO attack. Thus, blends of trisilanolphenyl (TSP) POSS and PI with different weight percentages of the Si7O9 POSS cage were cast into films and exposed to a hyperthermal AO beam, and they were characterized in terms of their recession, mass loss, surface morphology, and surface chemistry. In order to compare the AO resistance of the blends with POSS PI copolymers, samples of previously studied copolymers were also investigated in parallel with the blends. For all POSS PI materials, the AO resistance increased with increasing AO fluence and with increasing POSS cage loading. At similar POSS cage loadings and exposure conditions, the TSP POSS PI blends showed comparable erosion yields to the POSS PI copolymers, with specific samples of blends and copolymers achieving erosion yields as low as 0.066 × 10−24 cm3 atom−1 with an AO fluence of 5.93 × 1020 O atoms cm−2. SEM and XPS analyses indicated that passivating SiOx layers were formed on the surfaces of all POSS-containing polymers during AO exposure. Thus, a TSP POSS PI blend is proposed as a low-cost variant of a POSS polyimide for use in extreme oxidizing environments, such as LEO. KEYWORDS: atomic oxygen, polyimide, POSS, polyhedral oligomeric silsesquioxane, low Earth orbit, hyperthermal
1. INTRODUCTION Organic polymers are commonly used on the external surfaces of spacecraft in low Earth orbit (LEO), but the harsh oxidizing environment leads to rapid degradation of these materials through oxidation and etching if they are unprotected.1−16 Common approaches to make polymers more durable in LEO involve inorganic constituents, either as coatings,17−27 particle additives or fillers,28−32 or components of the polymer backbone.33−43 Spacecraft polymer coating materials include metals, metal oxides, and semiconductor oxides.17−27 Potential metal coatings, such as silver and copper, have been tested in the LEO environment, but they oxidize readily and would need to be protected.19,21,44 For some specialty applications, the noble metals, gold and platinum, should be durable in an oxidizing environment.44 The metal oxide coatings, Al2O3, TiO2, and SnO2, have shown remarkable resistance to atomic oxygen (AO) attack.24−27 Still, the most commonly used coating material is the semiconductor oxide, SiO2, which is almost as resistant to AO attack as Al2O3 and, like Al2O3, does not alter the thermo-optical properties of the material in a deleterious way.11,17,18,24,25,45,46 However, all inorganic coatings are subject to cracking through handling and thermal cycling and to damage from micrometeoroid and debris impacts, leaving the underlying material exposed to AO attack.11,12,18,19,47,48 © 2016 American Chemical Society
Particle additives to improve polymer durability include nanoclay, silica, and graphite.28−32 Adding particles of these materials is attractive because large-scale films of the composite materials can be produced simply and with relatively low cost. Some such polymer nanocomposites exhibited better AO resistance than control samples of pure polymer. For example, an epoxy/silica blend (with 5 wt % of 100 nm dia silica particles) showed an erosion yield of only 0.18 × 10−24 cm3 atom−1 after exposure to an AO fluence of 1.97 × 1021 O atoms cm−2.32 This erosion yield compares favorably to the commonly used Kapton H and HN polyimide films, whose erosion yields are 3.00 × 10−24 and 2.81 × 10−24 cm3 atom−1, respectively.10 On the other hand, polymer nanoclay composites did not have such high AO resistance, with reported erosion yields in an AO plasma of 0.78 × 10−24 and 2.91 × 10−24 cm3 atom−1 for nylon/7.5 wt % nanoclay (Nanocor I.30, with size of 8−10 μm) and epoxy/6.0 wt % nanoclay (Nanocor I.30, with size of 8− 10 μm) composites, respectively.30,31 As these inorganic additives do not dissolve in polymers, they usually aggregate to form inorganic protective layers as organic material erodes away.32 However, these protective layers are not continuous, so Received: August 23, 2016 Accepted: November 7, 2016 Published: November 7, 2016 33982
DOI: 10.1021/acsami.6b10612 ACS Appl. Mater. Interfaces 2016, 8, 33982−33992
Research Article
ACS Applied Materials & Interfaces
yielded a very low erosion yield of 0.026 × 10−24 cm3 atom−1 for 3.5 wt % Si8O11 MC POSS PI. Higher AO fluence exposures of POSS PIs typically lead to lower erosion yields, as a result of the growing and increasingly impenetrable SiO2 layer on the surface.37,40 Given the excellent AO resistance of the POSS PI copolymers and the fact that the POSS content did not adversely affect the thermal or mechanical properties of the material43,55,56 up to a POSS cage loading of 7 wt %, POSS PI seems to be ideal as a drop-in replacement for Kapton on the external surfaces of spacecraft in LEO. However, the POSS monomers that are functionalized for copolymerization are expensive to synthesize, and the copolymerization process is not conducive to low-cost and large-scale fabrication of POSS PI films. Thus far, a 45 in. diameter sheet of the 7 wt % Si8O12 SC POSS PI is the largest POSS PI film that has ever been produced.57 A potential approach to an AO-resistant polymer that combines the ease of additives with the durability of POSS would be to blend POSS monomers with a polymer. Such POSS blends are commercially available for various purposes,58−61 but they have not been investigated in detail for their durability under AO bombardment. The use of POSS PI blends has been explored in the context of reducing AO attack in the region of a hypervelocity impact, and it was found that the addition of POSS to the PI did significantly reduce the etching of the material in the vicinity of the impact where the polymer had been strained.12,52 These studies used blends of trisilanolphenyl (TSP) POSS monomers in a PMDA-ODA polyimide. TSP POSS, which is readily available from a commercial source, is soluble in polar solvents, presumably allowing for POSS PI blends with fully dispersed POSS monomers.62 If, as was shown for the MC and SC POSS PI copolymers, the AO resistance only depends on the POSS content then blends of TSP POSS PI with the appropriate wt % POSS cage should perform as well as the more costly MC and SC POSS PI copolymers. Therefore, we prepared a number of TSP POSS PI blends with specific weight percentages of POSS cage, and we exposed these polymers to beams of hyperthermal AO side-byside with POSS PI copolymers having nearly identical weight percentages of POSS cage. The effects of AO on the exposed materials were interpreted in terms of mass loss, etch depth, surface chemistry, and surface morphology.
O atoms can penetrate between the gaps and continue the etching of the organic component.31,32,49,50 Another potential problem with particle additives is that they might lead to undesirable mechanical properties of the material.28 Copolymerization with an appropriate inorganic monomer has the advantage that the inorganic component is uniformly dispersed and can form an essentially continuous passivating oxide layer when the hybrid copolymer is subjected to AO bombardment.34−43 Because the polyimide (PI), Kapton, is often used on the external surfaces of spacecraft, the majority of the hybrid copolymers that have been studied involve PI as the main component. Copolymers of the PI, oxidiphthalic dianhydride oxidianiline (ODPA-ODA), with phenylphosphine oxide groups have been shown to form a passivating phosphate layer on the surface upon exposure to AO.34−36,38,39 For example, one such phosphorus-containing PI was reported to have an erosion yield of 0.48 × 10−24 cm3 atom−1.36 Another study which used the copolymerization of phenylphosphine oxide and a polyimide in a composite reported an erosion yield as low as 0.68 × 10−24 cm3 atom−1 for an AO fluence of 2.27 × 1020 O atoms cm−2.51 A hyperbranched polysiloxane PI copolymer functioned in a similar way as the phosphorus PI copolymers, forming a passivating oxide layer (SiO2) and exhibiting a fairly low erosion yield of 0.24 × 10−24 cm3 atom−1 in a test with a reported AO fluence of 3.87 × 1020 O atoms cm−2.41 The thermal and optical properties of the hyperbranched polysiloxane PI copolymers were comparable to control PIs, while the mechanical properties degraded slightly.41 The major drawbacks of the siloxane block PI copolymers are complicated synthesis and high cost. Another copolymer involving silicon that has received a great deal of study is made from polyhedral oligomeric silsesquioxane (POSS) and the Kapton-like PI, pyromellitic dianhydride oxidianiline (PMDA-ODA), which also forms a passivating SiO2 layer when exposed to AO.12,37,40,42,43,52,53 POSS PI copolymers have shown excellent stability in a laboratory test environment with hyperthermal AO and in exposures on the International Space Station as part of the Materials Interactions Space Station Experiments (MISSE-1, MISSE-5, and MISSE-6).40 Two types of POSS copolymers were studied. In the first such material to be prepared, a cyclopentyl POSS monomer was incorporated into the main polymer backbone, referred to as main-chain (MC) POSS PI.40,54 Later, a second and somewhat less expensive copolymer was prepared by grafting a N-[(heptaisobutylPOSS)propyl]-3,5-diaminobenzamide POSS monomer as a pendant group on the main polymer backbone; this copolymer was referred to as side-chain (SC) POSS PI.40 It was found that the erosion yield of the POSS PI materials depended essentially on the wt % of the silicon−oxygen cage and not on how the POSS monomer was bonded to the polymer backbone.40 With a hyperthermal AO fluence of 2.70 × 1020 O atoms cm−2, the erosion yields of 7.0 wt % Si8O12 SC POSS PI and 7.0 wt % Si8O11 MC POSS PI were 0.15 and 0.13 × 10−24 cm3 atom−1, respectively.40 In a space-flight AO test of 1.80 × 1020 O atoms cm−2 on MISSE-5, the erosion yield of 7.0 wt % Si8O11 MC POSS PI was 0.14 × 10−24 cm3 atom−1.40 The erosion results from space-flight and hyperthermal AO beam tests at similar AO fluences were comparable, indicating that the hyperthermal AO beam exposure faithfully replicates the AO effects observed in LEO experiments. In a space-flight AO test of 1.97 × 1021 O atoms cm−2 on MISSE-6, the erosion yields of 7.0 wt % Si8O12 SC and Si8O11 MC POSS PIs were 0.07 and 0.05 × 10−24 cm3 atom−1, respectively.40 A much longer exposure on MISSE-1 to an AO fluence of 8 × 1021 O atoms cm−2
2. EXPERIMENTAL DETAILS 2.1. Preparation of TSP POSS Poly(amic acid) Solutions. A poly(amic acid) (HD MicroSystems PI-5878G) that is the precursor to the PMDA-ODA polyimide (with the same chemical repeat unit as Kapton) was used to produce PI films.40 This poly(amic acid) solution has a solid content of 16.40%, with N-methyl-2-pyrrolidone (NMP) as the solvent.63 Different weights (0, 0.13, 0.27, 0.43, 0.62, and 0.82 g) of TSP POSS (Hybrid Plastics SO1458) powder were dissolved in 15 g of poly(amic acid) solution to make homogeneous TSP POSS poly(amic acid) solutions with POSS monomer loadings of 0, 5, 10, 15, 20, and 25 wt %, which correspond to Si7O9 cage loadings of 0, 1.8, 3.7, 5.5, 7.3, and 9.1 wt %, respectively, as shown in Figure S1. These solutions were clear, indicating complete dissolution of the TSP POSS. 2.2. Solution Casting of TSP POSS PI Blend Free-Standing Films. The poly(amic acid) or TSP POSS poly(amic acid) solutions were cast onto 1-in. square plates of glass, which were placed in aluminum molds with 1-in. square pockets, as shown in Figure S2. Doctor blading was used to remove excess solution by wiping with a sharp cutting edge. During the casting and doctor blading process, attention was paid to avoid any air bubbles in the solution. The molds with coated glass plates were put into a glass container with a small air gap for 2 days in order to allow the NMP solvent to evaporate slowly and 33983
DOI: 10.1021/acsami.6b10612 ACS Appl. Mater. Interfaces 2016, 8, 33982−33992
Research Article
ACS Applied Materials & Interfaces avoid bubble formation. The depths of the pockets in the aluminum molds were designed to produce 25−50 μm thick TSP POSS PI freestanding films after curing. 2.3. Spin-Coated TSP POSS PI Blend Films. The poly(amic acid) or TSP POSS poly(amic acid) solutions were further diluted by NMP solvent, with a mass ratio of 1:0.5, to make the solutions less viscous for spin coating onto 5 MHz, 0.5 in. diameter, sensor crystals (Inficon Maxtek SC-150) for a quartz crystal microbalance (QCM) (Inficon Maxtek RCQM and associated front-load crystal holder). An adhesion promoter (HD MicroSystems VM-651) was used to treat all surfaces before spin coating the poly(amic acid). Dilution of the poly(amic acid) and spin-coating conditions were adjusted by trial and error to yield 1−2 μm thick PI films after cure. The typical spincoating procedure involves the following steps. (1) Dilute 0.1 mL of adhesion promoter with 120 mL of deionized water. (2) Apply promoter solution to the static QCM disc and let sit for 60 s. (3) Spin at 3000 rpm for 30 s. (4) Bake the promoter-coated disc in an oven at 60 °C for 1.5 h, remove from heat, and allow to cool to room temperature. (5) Apply a few drops of poly(amic acid) or TSP POSS poly(amic acid) solution to completely cover the static disc. (6) Spin at 500 rpm for 10 s. (7) Spin at 3000 rpm for 30 s. The spin-coating program for steps (6) and (7) is shown in Figure S3. Then the spincoated QCM discs were covered with a watch glass, with a small air gap for 2 days to allow the solvent to evaporate slowly. 2.4. Curing Process. Figure S4a shows a schematic diagram of the curing process. Pure poly(amic acid) and POSS poly(amic acid) films were cured in a temperature programmable vacuum oven (
