Article pubs.acs.org/JPCA
Reactive Molecular Dynamics Simulation on the Disintegration of Kapton, POSS Polyimide, Amorphous Silica, and Teflon during Atomic Oxygen Impact Using the Reaxff Reactive Force-Field Method A. Rahnamoun and A. C. T. van Duin* Department of Mechanical and Nuclear Engineering, Pennsylvania State University, 136 Research Building East, University Park, Pennsylvania 16802, United States ABSTRACT: Atomic oxygen (AO) is the most abundant element in the low Earth orbit (LEO). It is the result of the dissociation of molecular oxygen by ultraviolet radiation from the sun. In the LEO, it collides with the materials used on spacecraft surfaces and causes degradation of these materials. The degradation of the materials on the surface of spacecrafts at LEO has been a significant problem for a long time. Kapton polyimide, polyhedral oligomeric silsesquioxane (POSS), silica, and Teflon are the materials extensively used in spacecraft industry, and like many other materials used in spacecraft industry, AO collision degradation is an important issue in their applications on spacecrafts. To investigate the surface chemistry of these materials in exposure to space AO, a computational chemical evaluation of the Kapton polyimide, POSS, amorphous silica, and Teflon was performed in separate simulations under similar conditions. For performing these simulations, the ReaxFF reactive force-field program was used, which provides the computational speed required to perform molecular dynamics (MD) simulations on system sizes sufficiently large to describe the full chemistry of the reactions. Using these simulations, the effects of AO impact on different materials and the role of impact energies, the content of material, and temperature of material on the behavior of the materials are studied. The ReaxFF results indicate that Kapton is less resistant than Teflon toward AO damage. These results are in good agreement with experiment. These simulations indicate that the amorphous silica shows the highest stability among these materials before the start of the highly exothermic silicon oxidation. We have verified that adding silicon to the bulk of the Kapton structure enhances the stability of the Kapton against AO impact. Our canonical MD simulations demonstrate that an increase in the heat transfer in materials during AO impact can provide a considerable decrease in the disintegration of the material. This effect is especially relevant in silica AO collision. Considerable experimental efforts have been undertaken to minimize such AO-based degradations. As our simulations demonstrate, ReaxFF can provide a cost-effective screening tool for future material optimization.
1. INTRODUCTION
One of the most common materials used in the surface of spacecraft is Kapton. Kapton is a polyimide film that can remain stable in a wide range of temperatures, from 0 to 673 K. The main reasons for the interest in Kapton application on the spacecraft surface are the light weight, temperature stability, insulation characteristics, and UV stability of this material.6,7 For enhancing the stability and mechanical properties of Kapton, silicon oxide compound can be added to the material. Polyhedral oligomeric silsesquioxanes are molecules with the chemical composition (RSiO1.5)n that are used in the space applications. The most common composition is the composition of n = 8, which is shown in Figure 1. Kapton-POSS is a carbon-based material, and the reactions that lead to the formation of oxides lead to the removal of CO and CO2. Therefore, the AO collision can erode the surface of this material. The studies of these materials are mostly
The low earth orbital (LEO) environment subjects materials on the surface of a spacecraft to ∼4.5 eV collisions with ambient AO (atomic oxygen) and to ∼8 eV collisions with ambient N2 molecules.1 Such collisions cause the surface degradation of these materials. There have been many efforts to protect these materials by using nonvolatile metal oxides on the surface and in the bulk of the material. For enhancing the brittleness characteristics of such materials, compound-like fluoropolymers metal oxides are applied.2 Under the LEO space exposure condition, the materials used on the surface of the spacecrafts are affected by AO. Other parameters that can affect these materials such as ultraviolet (UV) radiation, micrometeoroids, debris, and thermal cycling are important to be considered for LEO exposure conditions. Simultaneous exposure of such materials to AO and UV radiation can dramatically increase the amount of the degradation.3,4 The impact of UV radiation is especially experienced on Teflon materials used on the surface of spacecraft.5 © 2014 American Chemical Society
Received: December 10, 2013 Revised: March 25, 2014 Published: March 28, 2014 2780
dx.doi.org/10.1021/jp4121029 | J. Phys. Chem. A 2014, 118, 2780−2787
The Journal of Physical Chemistry A
Article
Figure 2. Preparation of Kapton structure. In this preparation, 30 Kapton monomers are put together in an amorphous configuration. There are 1230 total atoms in the structure. The gray atoms represent C, the red atoms represent O, the blue atoms represent N, and the white atoms represent H.
Figure 1. (SiO1.5)n, n = 8 cage used in Kapton-POSS structure. The yellow atoms represent Si and the red atoms represent O.
experimental, and there are fewer theoretical studies than experimental studies. Teflon is another important material used on spacecrafts. The main usage of Teflon is protection from solar heating. However, the drawbacks of using Teflon in such applications can be erosion and cracking.2 In this research, atomistic scale simulations of surface degradation of Kapton, Kapton-POSS, amorphous silica, and Teflon, which are among the most common materials used on the surface of the spacecrafts, under collision from AO are evaluated. For performing these simulations, the ReaxFF reactive forcefield program was used. This classical force field can provide the computational requirements to perform MD simulations on systems that are sufficiently large. Therefore, we will be able to describe the full chemistry of the reactions using this program. ReaxFF is a general bond-order-dependent force field that provides reasonable descriptions of bond breaking and bond formation.8−15 The main difference between traditional nonreactive force fields and ReaxFF is that in ReaxFF the connectivity is determined by bond orders calculated from interatomic distances that are updated every MD simulation step. This allows for bonds to break and form during the simulation. To account for nonbonded interactions such as van der Waals and Coulomb interactions, for a system with changing connectivity, these interactions are calculated between every pair of atoms, irrespective of connectivity, and any excessive close-range nonbonded interactions are avoided by inclusion of a shielding term. In addition, ReaxFF accounts for polarization effects by using a geometry-dependent charge calculation scheme.
Figure 3. Preparation of the Kapton-POSS structure. In this preparation, 20 Kapton-POSS monomers are put together in an amorphous configuration. There are 2120 total atoms in the structure. The gray atoms represent C, the red atoms represent O, the blue atoms represent N, the white atoms represent H, and the yellow atoms represent Si.
Figure 4. Preparation of the Teflon structure. In this preparation, five Teflon monomers are put together in an amorphous configuration. There are 2200 total atoms in the structure. The gray atoms represent C and the green atoms represent F.
put in a box with enough capacity, and after that, by compressing the system, the final structure that is one molecule is produced. It should be mentioned that these methods have been used to obtain these amorphous structures. The main target in preparing these structures was to construct models that have densities close to the densities of these materials under the stable conditions. The sizes and lengths of the monomers used to construct these models have been selected arbitrarily. Therefore, these structure are not unique. 2.2. Simulation Steps. After constructing the structures, the simulation is performed. Each simulation involves four steps. These steps are shown in Figure 6. After preparing the structures, using geometry optimization followed by NVT equilibration at 300 K, the materials are ready for performing the AO impact simulations. The time step of the simulations is 0.1 fs, and AOs are added to the simulation every 200 fs and the positions of these AOs are ∼65 Å over the surface of the materials. The positions of the AO in the plane parallel to the surface of the material (x and y coordinates) are picked
2. METHODS 2.1. Slabs Preparation. The preparation of the structures of these four materials for the AO collision simulations is depicted in Figures 2−5. In the preparation of these slabs, the densities of 1.3, 2, 2.62, and 2.2 g/cm2 were obtained for Kapton, Kapton-POSS, silica, and Teflon slabs, respectively. For preparing Kapton and Kapton-POSS and Teflon structures, the monomers shown in Figures 2−4 are used for the polymerization. In the Kapton, Kapton-POSS, and Teflon structures, 30, 20, and 5 monomers are used, respectively. For preparing the amorphous silica structure, 71 Si8O12 cages are 2781
dx.doi.org/10.1021/jp4121029 | J. Phys. Chem. A 2014, 118, 2780−2787
The Journal of Physical Chemistry A
Article
Figure 5. Preparation of the Silica structure with 2600 atoms. Putting 130 Si8O12 cages at low density in the big enough box and slowly compressing the system. The final structure is one big molecule. The red atoms represent O and the yellow atoms represent Si.
Figure 6. Steps used in the AO collision simulations.
randomly for each AO added to the system. These AOs are directed toward the material with 4.5 eV energy and impact the surfaces of the materials perpendicularly.
3. RESULTS 3.1. AO Impact. After minimization and equilibration of the structures, the surfaces of these three structures are collided with high-energy AOs. These AOs are directed toward the slabs with 4.5eV energy. Every 200 fs time interval, one AO is added to the system. The MD simulations of these reactions are NVE
Figure 8. Temperature evolution of Kapton, Kapton-POSS, Teflon, and amorphous silica under AO impact (NVE simulations).
simulations, indicating conservation of energy, and some key snapshots of the changes in Kapton, POSS, and silica during the simulation are shown in Figure 7. The temperature evolution and the total remaining mass of the slabs under the AO attack are shown in Figures 8 and 9, respectively. For better understanding and because the initial weights of the slabs are different, the total mass is divided by the initial mass to obtain the normalized slab mass in Figure 10. With continuing the AO impact, the materials start to disintegrate at different rates, and in some stage the materials are almost disintegrated. This is because of the limitation of the
Figure 7. Key snapshots of AO collisions on silica, Kapton, POSS, and Teflon materials. 2782
dx.doi.org/10.1021/jp4121029 | J. Phys. Chem. A 2014, 118, 2780−2787
The Journal of Physical Chemistry A
Article
Figure 9. Total mass of the Kapton, Kapton-POSS, Teflon, and amorphous silica slabs under AO impact (NVE simulations).
The chemical compositions of the materials released from the surface of Kapton, POSS, Teflon, and silica materials in the first stages of AO impacts are shown in Table 1. These simulations show that in the first stages, water and O2 are the first small molecules separating from the Kapton surface, and with the progress of AO collisions, organic compound containing carbon starts separating from the surface, while for the POSS, we can see such carbon-containing organic materials coming off the surface even in the first stages of the simulation. The simulation of AO collisions on the silica surface shows the more stable performance of this material in the first stages, and all of the AOs reaching the surface attach to the surface. By keeping AO collisions, first O2 and later silicon-oxidecontaining fragments separate from the surface of the silica structure. For showing the first silicon containing molecules separating from the surface, the results of the simulation after impacting 250 atomic oxygen atoms on the silica surface are shown in Table 1. The results show that also Teflon shows very good stability against AO impact. The difference between the performance of Teflon compared with other materials under study is that most of the atomic oxygen atoms impacting the surface of Teflon do not react with the Teflon, while for other three cases, most of the atomic oxygen atoms react with the surfaces. The data in Table 1 show that after adding 51 AO, several different molecules start to separate from the surface of Kapton and Kapton-POSS, while silica and Teflon still stay stable and just a few small molecules separate from their surfaces. For comparing the simulations results with experimental results, the AO erosion yield observations of Materials International Space Station Experiment 2 (MISSE 2) have
Figure 10. Normalized slab mass changes of the Kapton, KaptonPOSS, Teflon, and amorphous silica slabs obtained by dividing the remaining slabs mass by the initial mass of the slabs.
number of the atoms of the initial slab used in the simulations. Because of this limitation in the size of the system, the trend of the mass loss in the first stages of the simulations (the values of first 200 000 iterations) can best indicate the performance of these materials under AO collisions. The AO impact on the silica results in more exothermic reactions; however, as shown in Figure 9, the silica is more stable than Kapton, POSS, and Teflon. On the basis of the results in Figures 8 and 9, Kapton-POSS shows much more stability compared with Kapton under AO impacts. It is clear that changing the Kapton structure to Kapton-POSS structure can improve the stability of this material against AO impact. 2783
dx.doi.org/10.1021/jp4121029 | J. Phys. Chem. A 2014, 118, 2780−2787
The Journal of Physical Chemistry A
Article
Table 1. Chemical Compositions of the Materials Separating from the Surface of Kapton, POSS, and Silica Materials in the First Stages of AO Impacts material
number of AO added
number of AO reacted
Kapton Kapton-POSS Silica Teflon Kapton Kapton-POSS Silica Teflon Kapton Kapton-POSS Silica Teflon Silica
21 21 21 21 36 36 36 36 51 51 51 51 250
11 14 12 1 27 28 24 3 42 43 40 8 214
molecules separated from the surface 1H2O 1C6H5O2
1HO, 1H, 2O2, 1C2H2O2, 1H2O 1CO2, 1H2O, 1HO, 1C3H2O, 1CO, 1C6H5O2, 1O2
1C6H6O3, 1C6H5O, 1CO2, 1C2HO, 1C4H2O3N, 3O2, 1CHO, 1H, 1C2HO2,1 HO 2H2O, 1CO2, 3HO, 1C5H4O2, 2CO, 1C3H2O, 1C2HO, 1C4H4O2, 1O2 1O2 4COF2 16 O2Si, 38O2,6O3Si,2O3Si2,2O4Si2,1O5Si2
(LEO) environment on the exterior of the International Space Station (ISS) for nearly 4 years. On the basis of these experimental results, the erosion yield for Kapton is 4.28 × 10−24 g/oxygen atom, and this value for Teflon is 3.05 × 10−25 g/oxygen. The erosion yield results of the simulations for Kapton and Teflon are shown in Table 2. If we consider the erosion yield of Kapton and Teflon in the first stages of the simulation where separation of small molecules start from the surfaces, the erosion yield values are close to the experiment results. Because there are some unclear origins in the interactions of AO and the surface of the materials during the AO impacts and also because there are some differences between experimental
Table 2. Erosion Yield Results of Kapton and Teflon AO Collision Simulations material
number of AO added
Kapton Teflon Kapton Teflon Kapton Teflon
21 21 36 36 51 51
erosion yield (g/oxygen atom) 1.42 0 7.29 0 2.00 5.86
× 10−24 × 10−24 × 10−23 × 10−25
been used.16 In these experiments, different materials that have usage in spacecrafts have been exposed to the low Earth orbit
Figure 11. Temperature evolution of the Kapton, POSS, and Silica during AO impact in NVE and NVT simulations. 2784
dx.doi.org/10.1021/jp4121029 | J. Phys. Chem. A 2014, 118, 2780−2787
The Journal of Physical Chemistry A
Article
Figure 12. Total mass of the lightweight molecules coming off the surface of the Kapton-POSS during AO impact.
Figure 13. Two different monomers used for building two Kapton-POSS structures with different silicon contents. Using the left monomer can provide higher silicon content inside the material, provided that both final structures will have the same total number of atoms.
Table 3. Numbers of Different Atoms in Two Different Kapton-POSS Structures atoms
lower silicon content structure
higher silicon content structure
C H O Si N total
880 560 440 160 80 2120
660 540 510 240 60 2010
results of such interactions, further investigations of these interactions are needed. At high impact energies, the issue of electron excitation might be an important feature to be considered too. Therefore, physical chemistry investigations of such phenomenon can be very helpful in achieving better understanding of these interactions. 3.2. Heat Transfer Effects. Considering that there is some heat transfer between the materials on the surface of the
Figure 14. Total remaining mass of the two Kapton-POSS structures with different Si contents during AO impact.
2785
dx.doi.org/10.1021/jp4121029 | J. Phys. Chem. A 2014, 118, 2780−2787
The Journal of Physical Chemistry A
Article
characteristics of the Kapton-POSS is the amount of silicon in the bulk of the material. For evaluating the effect of this silicon content, the simulations on Kapton-POSS with two different silicon contents are performed. The monomers that have been used for making two Kapton-POSS structures with different silicon contents are depicted in Figure 13. The monomer on the left is used for building the higher silicon content structure, and the monomer on the right is used for building the lower silicon content structure. The numbers of the atoms in these two different structures are shown in Table 3. In the simulations, the AOs are directed toward the slabs with 4.5 eV energy. Every 200 fs time interval, one AO is added to the system, and the time step of the simulations is 0.1 fs. The total remaining mass of the slabs and the normalized slab mass under the AO attack are shown in Figures 14 and 15, respectively. On the basis of these results, at the end of the simulation, the slab with lower Si content loses 3% of the initial mass, and the slab with higher Si content loses %1 of the initial mass. The temperature evolution of these slabs during AO impact is depicted in Figure 16. These results indicates that there would be slight increase in the temperature of the Kapton-POSS with higher Si content. Considering the mass and temperature limitations for the surface of the spacecraft materials, computational chemical methods as used here can provide a useful tool for evaluating optimal material composition under LEO conditions.
Figure 15. Normalized slab mass changes of the Kapton-POSS with different Si contents obtained by dividing the remaining slabs mass by the initial mass of the slabs.
4. CONCLUSIONS Our ReaxFF simulations of Kapton, Kapton-POSS (polyhedral oligomeric silsesquioxane), amorphous silica, and Teflon subject to the high-energy AO collision indicate that the amorphous silica shows the highest stability among these materials. However, once the disintegration of the silica starts under AO collision, it continues with a very high rate due to the high exothermicity of silicon oxidation. During the Kapton AO impact simulation, the initial gasphase species are dominated by H2O and O2, while lower quantities of H and OH radicals are observed, together with a wide range of larger Kapton degradation products. In the simulation in which POSS material is subject to AO collisions, initial gas species also contain CO and CO2. In contrast, for the silica AO impact, the initial gas-phase products are O2 and light silicon oxides, and later heavier silicon oxides appear as the gas phase. Similar to silica, Teflon shows very good stability, while first 50 atomic oxygen atoms are impacted on it is surface. With continuing the AO collision, carbonyl fluorides are observed as the first molecules separating from the surface of the Teflon. Comparing the stabilities of Kapton and Kapton-POSS stabilities under AO collision shows that adding silicon to the bulk of the Kapton structure enhances the stability of the structure against AO impact. The NVT simulation results for controlling the temperature evolution of the Kapton-POSS material show that heat transfer in the material during AO impact can provide a considerable decrease in the disintegration of the material. This effect is especially observed in silica AO collision, and by lowering the temperature of the material using thermostats, the stability of silica is increased considerably. Comparing the simulations results to experimental results of AO erosion yield of Materials International Space Station Experiment 2 (MISSE 2) shows that the erosion yields of Kapton and Teflon in first stages of the simulation are close to
Figure 16. Temperature evolution of the Kapton-POSS with different Si contents during AO impact in NVE simulations.
spacecraft and other parts, we repeated the NVE simulations described in the previous section using a canonical NVT/ Berendsen thermostat ensemble with three different thermostats: strong (Tdamp = 100 fs), medium (Tdamp = 500 fs), and weak (Tdamp = 5000 fs). The results of these simulations and the effect of using thermostats in the simulation compared with the NVE simulation for Kapton, POSS, and Silica are depicted in Figures 11 and 12. If we compare the simulations with no thermostat and weak thermostat, it is clear that using thermostats all thermostats show the same performance for such systems, and these thermostats keep the slab temperature around a constant value, and preventing the temperature increase in each of the slabs during AO impacts will prevent mass losses in the slabs. This result indicates the importance of the material temperature in regards to the stability of them. It should be noted that the heat transfer of these materials on the surface of the spacecraft to the body of the spacecraft is not desired. All of these requirements highlight the need for materials, which can show the desired qualities of current materials used, but they should be less exothermic when reacting with the high-energy AO. Effect of Different Silica Ratios in Kapton-POSS. An important parameter that can play the main role in the 2786
dx.doi.org/10.1021/jp4121029 | J. Phys. Chem. A 2014, 118, 2780−2787
The Journal of Physical Chemistry A
Article
(15) van Duin, A. C.; et al. Atomistic-Scale Simulations of the Initial Chemical Events in the Thermal Initiation of Triacetonetriperoxide. J. Am. Chem. Soc. 2005, 127, 11053−11062. (16) de Groh, K. K.; et al. MISSE 2 PEACE Polymers Atomic Oxygen Erosion Experiment on the International Space Station. High Perform. Polym. 2008, 20, 388−409.
the experiment results, and Teflon shows lower erosion yield compared with Kapton. Changing the silicon content and also using other enhancement methods for Kapton material like the effect of different ions additives in the bulk of Kapton can enhance the stability of the Kapton structure in the LEO environment, but the effect of these changes on the material characteristics like brittleness should be evaluated, and the best material composition should be figured out considering all limitations. As these simulations demonstrate, ReaxFF can provide a cost-effective screening tool for such material optimization.
■
AUTHOR INFORMATION
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was sponsored by the U.S. Air Force Office of Scientific Research (AFOSR), Grant No. FA9550-11-1-0158. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of the AFOSR or the U.S. Government.
■
REFERENCES
(1) Zhang, J.; Minton, T. K. Production of Volatile CO and CO2 from Oxidized Polyethylene and Graphite Surfaces by Hyperthermal Atom−Surface Collisions. High Perform. Polym. 2001, 13, 467−481. (2) Banks, B. A. The Use of Fluoropolymers in Space Applications; John Wiley & Sons: New York, 1997. (3) Bowles, D. E.; Tenney, D. R. Composite Tubes for the Space Station Truss Structure. Natl. SAMPE Tech. Conf. 1986, 18, 414−428. (4) Visentine, J. T. Atomic Oxygen Effects Measurements for Shuttle Missions STS-8 and 41-G; NASA Technical Memorandum 100459; National Aeronautics and Space Administration (NASA): Houston, TX; 1988, Vol. 1. (5) Koontz, S.; et al. Vacuum Ultraviolet Radiation/Atomic Oxygen Synergism in Materials Reactivity. J. Spacecr. Rockets 1990, 27, 346− 348. (6) Bedingfield, K. L., Leach, R. D.; Alexander, M. B. Spacecraft System Failures and Anomalies Attributed to the Natural Space Environment; NASA Reference Publication 1390; National Aeronautics and Space Administration: Huntsville, AL, 1996. (7) Tribble, A. C. The Space Environment: Implementation for Spacecraft Design; Princeton University Press: Princeton, NJ, 2003 (8) Chenoweth, K.; van Duin, A. C. T.; Goddard, W. A. A ReaxFF Reactive Force Field for Molecular Dynamics Simulations of Hydrocarbon Oxidation. J. Phys. Chem. A 2008, 112, 1040−1053. (9) Nielson, K. D.; et al. Development of the ReaxFF Reactive Force Field for Describing Transition Metal Catalyzed Reactions, with Application to the Initial Stages of the Catalytic Formation of Carbon Nanotubes. J. Phys. Chem. A 2005, 109 (3), 493−499. (10) Strachan, A.; et al. Shock Waves in High-Energy Materials: The Initial Chemical Events in Nitramine RDX. Phys. Rev. Lett. 2003, 91 (9), 098301. (11) Buehler, M. J.; van Duin, A. C. T.; Goddard, W. A. Multiparadigm Modeling of Dynamical Crack Propagation in Silicon Using a Reactive Force Field. Phys. Rev. Lett. 2006, 96 (9), 095505. (12) Ojwang, J. G. O.; et al. Modeling the Sorption Dynamics of NaH Using a Reactive Force Field. J. Chem. Phys. 2008, 128 (16), 164714. (13) Strachan, A.; et al. Thermal Decomposition of RDX from Reactive Molecular Dynamics. J. Chem. Phys. 2005, 122 (5), 054502. (14) van Duin, A. C. T.; et al. ReaxFF: A Reactive Force Field for Hydrocarbons. J. Phys. Chem. A 2001, 105 (41), 9396−9409. 2787
dx.doi.org/10.1021/jp4121029 | J. Phys. Chem. A 2014, 118, 2780−2787
