Thermal and Surface Properties of Phenolic
Nanocomposites Containing Octaphenol Polyhedral
Oligomeric Silsesquioxane
Han-Ching Lin, Shiao-Wei Kuo,* Chih-Feng Huang, Feng-Chih Chang Institute of Applied Chemistry, National Chiao Tung University, Hsin Chu, Taiwan Fax: 886 3 5131512; E-mail: [email protected]
Received: December 15, 2005; Revised: January 27, 2006; Accepted: January 27, 2006; DOI: 10.1002/marc.200500852 Keywords: nanocomposites; phenolic; POSS; resins; surface property
Summary:We have synthesized a new polyhedral oligome-ric silsesquioxane (POSS) containing eight phenol functional groups and copolymerized it with phenol and formaldehyde to form novolac-type phenolic/POSS nanocomposites exhib-iting high thermal stabilities and low surface energies. Our DSC results indicate that the glass transition temperature of these nanocomposites increased initially upon increasing their POSS content, but then decreased at POSS content above 10 wt.-%, presumably because of the formation of relatively low molecular weight species and POSS
aggrega-tion as evidenced from MALDI-TOF mass analyses. Our TGA analyses indicated that the 5-wt.-%-mass-loss temper-atures (Td) increased significantly upon increasing the POSS
content because the incorporation of the POSS led to the formation of an inorganic protection layer on the nanocom-posite’s surface. XPS and contact angle data provided posi-tive evidence to back up this hypothesis. In addition, contact angle measurements indicated a significant enhancement in surface hydrophobicity after increasing the POSS content.
Introduction
Polymers reinforced with well-defined nanosized inorganic clusters (i.e., polymeric nanocomposites) have attracted a tremendous degree of interest because of their potential applications. Among these systems, polyhedral oligomeric silsesquioxanes (POSSs) compounds, which possess uni-que cage-like structures and nanoscale dimensions, are of particular interest for use as hybrid materials. POSS com-pounds embody inorganic/organic hybrid architectures, i.e., they contain an inner inorganic framework composed of silicone and oxygen (SiO1.5)xand present organic
substi-tuents. Because POSS moieties can be readily incorporated into polymer matrices through copolymerization, many types of polymer/POSS nanocomposites have been synthe-sized.[1 – 7]
Phenolic resins are currently irreplaceable materials be-cause they exhibit excellent ablative properties, structural integrity, thermal stability, and solvent resistance. In gene-ral, phenolic resins have been practically neglected as materials for use in nanocomposites because they possess three-dimensional structures, even when the resins are not crosslinked, but some effort has been directed toward over-coming this drawback. For example, phenolic-based nano-composites have been prepared through the sol-gel processing[8] and through intercalative polymerization[9]
in the presence of montmorillonite modified with different surfactants. In contrast, very few studies have described the effects that POSSs have on enhancing the properties of phenolic resins,[10]especially their surface properties.
In this study, we developed a simple two-step synthesis of octaphenol-POSS through hydrosilylation of 4-acetoxys-tyrene with Q8M8
H
and subsequent hydrolysis of the acetoxy units. This octaphenol-POSS, which possesses eight poly-merizable phenol groups, is quite miscible with phenol in solution because of the existence of favorable intermolec-ular interactions. The eight phenol groups on the periphery of this POSS core copolymerize with phenol and form-aldehyde units to form high-molecular-weight novolac-type phenolic/POSS nanocomposites, as illustrated in Scheme 1.
Experimental Part
Syntheses of Phenolic/OP-POSS Nanocomposites
Octa(acetoxystyryl)octasilsesquioxane (AS-POSS) was ob-tained through the reaction of Q8M8Hand 4-acetoxystyrene, as
described previously.[11] AS-POSS was dissolved in THF, NaOH(aq) was added, and the hydrolysis was performed at
room temperature under nitrogen; it was complete within 2 d. Ethyl ether and deionized water (1:1) were added to the
solution, and then aqueous hydrochloric acid (10 wt.-%) was added slowly with stirring until the pH reached 8. Residual ethyl ether and water were evaporated under vacuum to provide octaphenol-POSS (OP-POSS, Mn¼ 1 560 g mol1, PDI¼
1.12 by GPC), which was soluble in THF and benzene. The structure of the OP-POSS was confirmed from its1H NMR and FT-IR spectra.
1H NMR (500 MHz, CDCl
3, ppm): d¼ 6.75–7.13 (4H,
aromatic CH), 4.93 (1H, C6H4OH), 2.91 (2H, CCH2Ar), 1.25
(2H, SiCH2C), 0.54 ppm (6H, Si(CH3)2).
FT-IR (KBr, cm1): 3 525 (free OH), 3 350 (hydrogen bonded OH), 3 037 (benzene ring CH stretching), 2 957 (CH2
stretching), 1 603 (in-plane aromatic C–C stretching), 1 065 (Si–O–Si stretching).
Aqueous formaldehyde and a desired amount of OP-POSS were added into a phenol solution, stirred for 5 min, and then the mixture was added to the flask; Table 1 summarizes the compositions of the phenolic/OP-POSS nanocomposites. Sul-furic acid was added via syringe to the flask and the mixture was then heated at 100 8C under nitrogen; the reaction was completed within 22 h. The solution was washed three times with hot (90 8C) water to remove any unreacted monomer and then it was extracted to remove any residual water. The product was dried in a vacuum oven at 180 8C for 24 h.
Characterization
1H NMR spectra were recorded on a Varian Unity Inova 500 FT
NMR spectrometer operated at 500 MHz; deuterated chloro-form was used as the solvent. Thermal analyses were per-formed using a DuPont DSC-9000 differential scanning calorimeter operated at a scan rate of 20 8C min1within a
temperature range from50 to 150 8C. Thermal stabilities of the cured samples were investigated using a Du Pont 2050 TGA instrument operated at a rate of 10 8C min1from 30 to 800 8C under a nitrogen flow. All mass spectra were recorded using a Bruker Biflex 3 time-of-flight mass spectrometer equipped with a 337-nm nitrogen laser, a 1.25-m flight tube, and a sample target having the capacity to load 384 samples simultaneously; the accelerating potential was set at 19 kV. For contact angle measurements, deionized water and diiodomethane (99%, Aldrich) were chosen as testing liquids because significant amounts of data are available for these liquids. The advancing contact angle measurements of a polymer sample were deter-mined at 25 8C after injecting a liquid drop (5 mL) onto the surface and then using a Kru¨ss GH-100 goniometer interfaced to image-capture software to perform the measurement. We use the two-liquid geometric method to determine the surface energy.[12] For X-ray photoelectron spectroscopy (XPS),
samples were spin-coated onto silicon wafers at 1 500 rpm for 45 s; XPS was performed using a VG Microlab 310F spectrometer equipped with an Al Ka X-ray source
(1486.6 eV).
Results and Discussion
Figure 1 displays the conventional second-run DSC and TGA thermograms of phenolic/OP-POSS nanocomposites that we prepared at various weight ratios. Each of these hybrids possesses essentially a single value of Tg,
suggest-ing that these hybrids exhibit a ssuggest-ingle phase. The glass transition temperatures of the nanocomposites were signifi-cantly enhanced after incorporation of the POSS units, but suggested that some incompletely reacted functional groups remained on the POSS and phenolic units. We believe that the enhanced glass transition
Table 1. Formulations and thermal properties of phenolic/OP-POSS nanocomposites.
OP-POSS OP-POSS Phenol Formaldehyde Tg Td Char yield at 800 8C
wt.-% g g g 8C 8C wt.-%
0 0 47.44 12.58 82 252 39.8
2 1.23 46.53 12.58 104 311 53.3
5 3.71 45.59 12.58 106 331 53.6
10 6.18 44.65 12.58 96 375 48.2
Figure 1. Thermal analyses of phenolic/OP-POSS nanocompo-sites containing different OP-POSS contents: (a) DSC and (b) TGA.
temperatures resulted from the restricted motion of the polymer chains that was caused by the even distribution of POSS units on the segmental level.[13]In addition, because phenolic resin contains a high density of hydroxyl groups, strong self-associated hydrogen bonding with the siloxane groups of OP-POSS serves as a physical crosslink that increases the values of Tgof the nanocomposites; we have
discussed this phenomenon in depth previously.[14] The values of Tg of these nanocomposites increased upon
increasing the POSS content up to 5 wt.-%, but then they decreased unexpectedly at loadings above 10 wt.-%.
To understand this change in behavior of the glass tran-sition temperature of these nanocomposites, we obtained MALDI-TOF mass spectra analyses to monitor the molec-ular weight distribution. The molar mass distributions of silsesquioxanes can be determined conveniently through the use of UV-MALDI-TOF mass spectrometry.[15] Figure 2 displays the normalized MALDI-TOF mass spectra of the phenolic/OP-POSS nanocomposites that we prepared using various OP-POSS concentrations.
The abundances of low-molecular-weight species (m/z < 1 500) increased upon increasing the POSS content;
this finding is consistent with an increasing number of POSS-POSS interactions.[16]In addition, the greater steric hindrance about the aggregated POSS clusters tends to decrease the reactivity of the phenolic groups; this pheno-menon results in the production of a relatively larger fractions of low-molecular-weight components and the concomitant decrease in values of Tg. As a result, two
competitive factors are involved in determining the final glass transition temperature of the phenolic/OP-POSS nanocomposites: the hindering effects that the POSS cages have on the motion of the polymer chains tends to increase the value of Tg, while the inclusion of the bulky POSS
groups tends to increase the free volume of the system and retard the reactivity of the monomers to produce relatively lower MW nanocomposites, i.e., those having lower values of Tg.[17]
We also applied TGA to evaluate the thermal stability of the phenolic/OP-POSS nanocomposites [Figure 1(b)]. Typ-ical of the degradation of phenolic resins, our neat phenolic resin and all of the POSS-containing nanocomposites degraded thermally over three steps. Table 1 summarizes the results of the DSC and TGA analyses, including the 5-wt.-%-mass-loss temperatures (Td) and the char yields at
800 8C. The char yield increased upon increasing the POSS content, except that it decreased for the sample incorporat-ing 10 wt.-% of POSS, which provides further evidence that lower-molecular-weight species were present in the phenolic/OP-POSS 10 wt.-% sample. The value of Td
increased significantly upon increasing the POSS content; e.g., the phenolic/OP-POSS 10 wt.-% sample exhibited a value of Tdthat was 123 8C higher than that of the pure
phenolic resin. This phenomenon can be explained in terms of the nanoreinforcement effect of incorporating POSS moieties into polymeric matrixes. The nanoscale dispersion of POSS moieties within the matrix and their covalent and hydrogen bonds to the phenolic resin are responsible for enhancing the initial decomposition temperature. A recent study demonstrated that physical crosslinks formed by POSS units can significantly retard thermal motion; at the same time, they can act as flow aids at elevated
Figure 2. MALDI-TOF mass spectra of phenolic and phenolic/ OP-POSS nanocomposites.
Table 2. Advancing contact angles, surface free energies, and XPS analyses of phenolic/OP-POSS nanocomposites of various compositions.
OP-POSS Testing liquid Surface free energy XPS analysis
wt.-% Water Diiodomethane gsd gsp gs C O Si
mJ m2 mJ m2 mJ m2 % % %
0 79.4 42.2 34.5 4.8 39.3 84.9 15.1 0.0
2 91.1 71.0 19.1 4.6 23.7 77.3 11.8 10.9
10 97.0 73.8 18.5 2.8 21.3 67.9 12.1 20.0
After treatment at 210 8C for 24 h
0 –a) – – – –
2 77.5 65.0 20.0 10.7 30.7
10 74.7 66.1 18.8 13.0 31.8
a)
temperatures.[18]In addition, the increased values of Tdfor
the nanocomposites may also result from increased chain spacing, which gives rise to lower thermal conductivity.[19] It was reported recently that the POSS units in related nanostructures prefer to be oriented toward the air-side, an arrangement that screens out the polar groups (e.g., urethane and carboxyl units).[20,21]We anticipated that the POSS moieties in our phenolic/OP-POSS nanocomposites were also oriented toward the air-side to form an inorganic protection layer on the surface of each nanocomposite. To provide evidence in support of this hypothesis, we performed contact angle measurements and XPS analyses to investigate the surface behavior of the nanocomposites. Table 2 lists the surface advancing contact angles and surface energies, measured using two testing liquids (water and diiodomethane), of the pure phenolic resin and of the phenolic/OP-POSS nanocomposites. We used the contact angle data to calculate the polymer surface energy (gs), and
its polar (gsp) and dispersive (gsd) components, according to
the two-liquid geometric method. Strikingly, the phenolic/ OP-POSS 10 wt.-% sample had a surface energy (21.3 mJ m2) lower than that of poly(tetrafluoroethylene) (PTFE) (22 mJ m2, as measured using the same method).[22] Noticeably, the polar component of the surface energy (gsp)
was very sensitive to the presence of even a low distribution of POSS moieties; it decreased upon increasing the POSS content.
Table 2 also lists the XPS results, we observed that the atomic percentage of silicon increased dramatically upon increasing the POSS content. Thus, both the contact angle measurements and the XPS results indicate that the POSS moieties were distributed preferably on the surface of the nanocomposite to provide a barrier against the direct contact of the polar phenolic units with the air. To further confirm this structure, we measured the contact angles of samples of pure phenolic resin and phenolic/OP-POSS nanocomposites that we had heated at 210 8C for 24 h; the results are presented in Table 2. After thermal treatment, both testing liquids exhibited complete wetting on the sur-face of the pure phenolic resin, i.e., we could not measure its contact angles; this phenomenon indicates that the surface properties of the pure phenolic resin changed dramatically after such high-temperature treatment. In contrast, the contact angles of the phenolic/OP-POSS nanocomposites decreased and resulted in higher surface energies, indicat-ing that the incorporation of the POSS moieties enhanced the surface thermal stabilities of these nanocomposites.
Conclusion
We synthesized a series of POSS-based hybrid phenolic resins and characterized their thermal properties
and surface free energies. Our DSC and TGA results indicate that the enhancement in the thermal properties was due to strong hydrogen bonding serving as a physical crosslink between the phenolic resin and the OP-POSS units. The results of TGA, XPS, and contact angle analyses all provided evidence that the incorporation of POSS led to the formation of a surface barrier that minimized direct contact of the polar phenolic units with the air. The presence of such a barrier not only enhanced the thermal stability of the bulk and surface of these POSS-containing composites but also led to the surface energy being maintained after treatment at high temperature.
[1] T. S. Haddad, J. D. Lichtenhan, Macromolecules 1996, 29, 7302.
[2] M. J. Abad, L. Barral, D. P. Fasce, R. J. J. Williams, Macromolecules 2003, 36, 3128.
[3] Y. Liu, F. Meng, S. Zheng, Macromol. Rapid Commun. 2005, 26, 926.
[4] H. Li, S. Zheng, Macromol. Rapid Commun. 2005, 26, 196. [5] C. M. Leu, Y. T. Chang, K. H. Wei, Macromolecules 2003,
36, 9122.
[6] Q. Chen, R. Xu, J. Zheng, D. Yu, Macromol. Rapid Commun. 2005, 26, 1878.
[7] B. X. Fu, W. Zhang, B. S. Hsiao, G. Johansson, B. B. Sauer, S. Phillips, R. Balnski, M. Rafailovich, J. Sokolov, Polym. Prepr. 2000, 41, 587.
[8] C. C. M. Ma, S. C. Sung, F. Y. Wang, L. Y. Chiang, L. Y. Wang, C. L. Chiang, J. Polym. Sci., Polym. Phys. 2001, 39, 2436.
[9] H. Y. Byun, M. H. Choi, J. Chung, Chem. Mater. 2001, 13, 4221.
[10] C. U. Pittman, Jr., G.-Z. Li, H. Ni, Macromol. Symp. 2003, 196, 301.
[11] S. W. Kuo, H. C. Lin, W. J. Huang, C. F. Huang, F. C. Chang, J. Polym. Sci., Polym. Phys. 2006, 44, 673.
[12] F. W. Fowkes, ‘‘Adhesion and Adsorption of Polymers, Polymer Science and Technology’’, L. H. Lee, Ed., Plenum Press, New York 1980, Vol. 12A, p. 43.
[13] M. Goldman, L. Shen, Phys. Rev. 1966, 114, 321.
[14] Y. J. Lee, S. W. Kuo, W. J. Huang, H. Y. Lee, F. C. Chang, J. Polym. Sci., Polym. Phys. 2004, 42, 1127.
[15] W. E. Wallace, C. M. Guttman, J. M. Antonucci, J. Am. Soc. Mass Spectrom. 1999, 10, 224.
[16] H. Xu, S. W. Kuo, C. S. Lee, F. C. Chang, Macromolecules 2002, 35, 8788.
[17] Y. Ni, S. Zheng, Chem. Mater. 2004, 16, 5141.
[18] S. H. Phillips, T. S. Haddad, S. J. Tomczak, Curr. Opin. Solid State Mater. Sci. 2004, 8, 21.
[19] H. Liu, S. Zheng, K. Nie, Macromolecules 2005, 38, 5088.
[20] S. Turri, M. Levi, Macromolecules 2005, 38, 5569. [21] S. Turri, M. Levi, Macromol. Rapid Commun. 2005, 26,
1233.
