Poly(imide-siloxane)s based on hyperbranched polyimides

Petr Sysel; Anna Patrova; Marek Lanc; Karel Friess

doi:10.1515/epoly-2017-0103

Article Publicly Available

Poly(imide-siloxane)s based on hyperbranched polyimides

Petr Sysel , Anna Patrova , Marek Lanc and Karel Friess

Published/Copyright: July 25, 2017

Published by

Become an author with De Gruyter Brill

Submit Manuscript Author Information

From the journal e-Polymers Volume 18 Issue 2

Abstract

Poly(imide-siloxane)s differing in their composition were prepared and characterized. The starting polymer (control) was the hyperbranched polyimide based on 4,4′-oxydiphthalic anhydride and 4,4′,4″-triaminotriphenylmethane. In the poly(imide-siloxane)s, 10 or 40–50 mol% of the 4,4′,4″-triaminotriphenylmethane was theoretically substituted for by amine-terminated siloxane dimer or oligomers. 1,3-Bis(3-aminopropyl)-1,1,3,3-tetramethyldisiloxane, bis(3-aminopropyl)-terminated poly(dimethylsiloxane) with a theoretical number-average molar mass of 1000 g·mol⁻¹ and bis(3-aminopropyl)-terminated poly(dimethylsiloxane) with a theoretical number-average molar mass of 2500 g·mol⁻¹ were used for this purpose. The thermo-mechanical properties and gas separation characteristics for hydrogen, carbon dioxide and methane of these polymeric products were shown to be dependent on their composition.

Keywords: gas transport properties; hyperbranched polyimide; poly(imide-siloxane); siloxane moiety; thermo-mechanical properties

1 Introduction

The dendritic topology (dendrimers, hyperbranched polymers) has been demonstrated as a new macromolecular architecture group. Hyperbranched polymers do not have a perfectly branched architecture, but they possess some similar properties to dendrimers, and they can be employed to replace them for certain purposes. Their highly branched structure with a specific polymer chain hierarchy (arrangement) and large number of functional end-groups are the basic structural features that distinguish them from linear polymers (1), (2). Increasing attention has been paid to hyperbranched polyimides (HBPIs) (3), (4) due to the potential combination of the known advantages of linear (5) or crosslinked (6) polyimides with those of hyperbranched polymers (2), (7).

Polyimides show good mechanical, dielectric and chemical stability at temperatures up to 250°C. These rigid polymers with high glass transition temperatures are mostly used in the aircraft industry, electronics (5) and as polymeric separation membranes (8). Linear polyimides are commonly synthesized by a two-step procedure via a polyimide precursor. The precursor (polyamic acid) is prepared from an aromatic dianhydride and an aromatic diamine in solution at room temperature. The precursor is transformed into a polyimide using thermal or chemical imidization (5). HBPIs are often prepared from a bifunctional anhydride and a trifunctional amine under specific reaction conditions to avoid gel (3-dimensional structure) formation. The molar ratio of the anhydride and amine components determines the type and/or ratio of the end-groups (2), (3), (4).

We have previously used 4,4′-oxydiphthalic anhydride and 4,4′,4″-triaminotriphenylmethane as monomers for this purpose (4), (9). These compounds, used in a 1:1 molar ratio, provided an amine end-capped HBPI. The self-standing, homogeneous membranes prepared from this polymer showed a 2–4-fold increase in gas permeability compared with a membrane prepared from the linear polyimide based on 4,4′-oxydiphthalic anhydride and 4,4′-diaminodiphenylmethane at comparable separating ability (selectivity). Nevertheless, the brittleness of these materials – due to the lack of chain entanglements – renders them unsuitable for a larger scale use. A reduction of the degree of branching from forming a “hyperbranched-linear hybrid structure” (2) can contribute to better mechanical stability.

The range of polyimide use can be broadened by combining them with another polymer. Considerable attention has been focused on the preparation of linear poly(imide-siloxane)s. By incorporating rubbery, hydrophobic siloxane segments – most frequently poly(dimethylsiloxane)s – into polyimide chains, products with higher solubility, impact strength and gas permeability and lower modulus, moisture absorption and relative permittivity have been obtained. Simultaneously, it is important that the long-term thermal stability in air (up to 180°C) of poly(dimethylsiloxane)s is only slightly lower than that of polyimides. The final properties depend on both the type of copolymer and/or its weight composition as well as the molar mass of the components used (10), (11), (12), (13). Researchers in one study (14) prepared a hyperbranched siloxane precursor. After transforming it to a bifunctional component, they incorporated it into a polyimide backbone. The resistance of these materials to the simulated atomic oxygen environment was higher than that of a pure polyimide. To our best knowledge, poly(imide-siloxane)s derived from HBPI have not yet been prepared and characterized. In the work (15) the properties of the poly(urethane-siloxane)s based on a hyperbranched polyester were controlled by the ratio of components.

Therefore, poly(imide-siloxane)s prepared by substituting a defined fraction of trifunctional 4,4′,4″-triaminotriphenylmethane by amine-terminated siloxane dimer or oligomers (with a number-average molar mass of 1000 or 2500 g·mol⁻¹) were synthesized, and their structure, thermo-mechanical properties and gas transport characteristics were studied.

2 Materials and methods

2.1 Chemicals (materials)

4,4′-Oxydiphthalic anhydride (ODPA) (Chriskev, USA) was heated at 160°C for 5 h under vacuum before use. 4,4′,4″-Triaminotriphenylmethane (TTM) (Dayang Chemicals, China), 1,3-bis(3-aminopropyl)-1,1,3,3-tetramethyldisiloxane (N2Si), bis(3-aminopropyl)-terminated poly(dimethylsiloxane) with a theoretical number-average molar mass of 1000 g·mol⁻¹ (N2Si10) and bis(3-aminopropyl)-terminated poly(dimethylsiloxane) with a theoretical number-average molar mass of 2500 g·mol⁻¹ (N2Si25) (all Hüls Petrarch, Germany) were used as received. 1-Methyl-2-pyrrolidone (NMP) was distilled under vacuum over phosphorous pentoxide, and tetrahydrofuran (THF) (both Aldrich, Czech Republic) was distilled at atmospheric pressure over sodium.

2.2 Synthesis of hyperbranched polyamic acid

The hyperbranched polyamic acid (HBPAA) was synthesized in a three-necked flask equipped with a magnetic stirrer, nitrogen inlet/outlet and dropping funnel. At room temperature, a 4,4′-oxydiphthalic anhydride (ODPA) solution in NMP [2.069 g (0.00667 mol) of ODPA in 50 ml of NMP] was added dropwise to the 4,4′,4″-triaminotriphenylmethane (TTM) solution in NMP [1.931 g (0.00667 mol) of TTM in 46 ml of NMP]. This reaction mixture was then stirred at room temperature for 24 h.

2.3 Synthesis of poly(amic-siloxane) acids

The preparation of the poly(amic-siloxane) acid (PASA) based on ODPA, TTM and N2Si in a molar ratio 0.5:0.45:0.05 is as follows:

PASA was synthesized in a three-necked flask equipped with a magnetic stirrer, nitrogen inlet/outlet and dropping funnel. At room temperature, an ODPA solution in NMP [2.069 g (0.00667 mol) of ODPA in 50 ml of NMP] was added dropwise to the solution consisting of 1.731 g (0.00599 mol) of TTM, 0.166 g (0.00068 mol) of N2Si and 46 ml of NMP. This reaction mixture was then stirred at room temperature for 24 h.
If bis(3-aminopropyl)-terminated poly(dimethylsiloxane) with a theoretical number-average molar mass of 1000 g·mol⁻¹ (N2Si10) or bis(3-aminopropyl)-terminated poly(dimethylsiloxane) with a theoretical number-average molar mass of 2500 g·mol⁻¹ (N2Si25) were used as comonomers, a mixed solvent of NMP/THF [40/60 (vol%/vol%)] was used instead of NMP.

2.4 Preparation of hyperbranched polyimide (HBPI) and poly(imide-siloxane) (PIS)

The solution of HBPAA or PASA was cast on a Teflon substrate and heated in an oven at 30°C for 6 h, 60°C for 6 h, 100°C 1 h, 150°C 1 h, 200°C for 2 h and 230°C for 1 h. Self-standing films with a thickness of approximately 50 μm were obtained.

2.5 Instrumental techniques

IR spectra were taken on a Nicolet 6700 spectrometer in reflective mode. Thermogravimetric measurements (TGA) were performed in air from 25 to 800°C with a temperature gradient of 10°C·min⁻¹ using a TGA Q500 instrument. Dynamic mechanical analysis (DMA) was recorded using a DMA DX04T (RMI, Bohdanec, Czech Republic) operating at 1 Hz from 25 to 400°C with a temperature gradient of 3°C·min⁻¹. Mechanical properties were evaluated by using an Instron 3365. The intrinsic viscosities of the polyimide precursors were measured by using an Ubbelohde capillary viscometer in NMP or NMP/THF at 25°C. The permeation measurements were conducted using a self-developed manometric integral apparatus (16).

3 Results and discussion

The main aim of this work was the preparation and characterization of poly(imide-siloxane)s (PIS) derived from a hyperbranched polyimide (HBPI). This starting material (control) is represented by the amine end-capped HBPI based on a bifunctional ODPA and a trifunctional TTM [HBPI(ODPA-TTM)]. It was prepared using a two-step synthesis via a polyimide precursor, hyperbranched polyamic acid [HBPAA(ODPA-TTM)] (Figure 1).

Figure 1:

Two-step synthesis of of HBPI based on ODPA and TTM.

To synthesize the amine-terminated HBPAA(ODPA-TTM), an ODPA:TTM molar ratio of 1:1 (anhydride groups:amino groups 2:3) was used. When trifunctional and bifunctional monomers are employed, a three-dimensional structure (gel) is created during the course of polymerization (2). The gel formation was suppressed by the slow dropwise addition of a dilute solution of one monomer to a solution of the other. The highest concentration of solids without the gel formation was found to be 4 wt% [4, 9]. The intrinsic viscosity of the HBPAA(ODPA-TTM) was 30.7 cm³·g⁻¹ (in NMP, at 25°C).

For the syntheses of polyimide precursors containing siloxane moieties (PASA), N2Si [PASA(ODPA-TTM-N2Si)], N2Si10 [PASA(ODPA-TTM-N2Si10)] and N2Si25 [PASA(ODPA-TTM-N2Si25)] were used in addition to ODPA and TTM monomers. After their heat treatment (see Materials and methods, 2.3) the corresponding poly(imide-siloxane)s (PIS) were obtained. The chemical structure of such a PIS is shown in Figure 2. The designation of the materials together with the molar ratios of the starting compounds used for their syntheses are given in Table 1.

Figure 2:

Chemical structure of PIS.

Table 1:

Theoretical monomer molar ratios in the materials.

Material	Monomers (mol%)
Material	ODPA	TTM	N2Si	N2Si10	N2Si25
HBPI	50	50	–	–	–
PIS1	50	45	5	–	–
PIS2	50	25	25	–	–
PIS3	50	45	–	5	–
PIS4	50	30	–	20	–
PIS5	50	45	–	–	5
PIS6	50	30	–	–	20

For the synthesis of a polyimide precursor, it is important to find a medium dissolving both the monomers and the product (polyamic acid). In the cases of HBPAA(ODPA-TTM) and PASA(ODPA-TTM-N2Si), NMP meets these requirements. On the other hand, N2Si10 and N2Si25 are not soluble in NMP. A mixed solvent of NMP/THF in a 2:3 volume ratio was suitable for the preparation of 4 wt% solutions of N2Si10 and N2Si25. PASA(ODPA-TTM-N2Si10) and PASA(ODPA-TTM-N2Si25) containing theoretically not more than 20 mol% of siloxane oligomers were soluble in this mixed solvent. Although the preparation of the materials containing theoretically 5 or 25 mol% of the siloxane moieties was the aim of this work, the level was decreased to 20 mol% for PASA(ODPA-TTM-N2Si10) (providing PIS4) and PASA(ODPA-TTM-N2Si25) (providing PIS6) (see Table 1). The intrinsic viscosities of PASA(ODPA-TTM-N2Si10) containing 5 mol% of N2Si10, PASA(ODPA-TTM-N2Si10) containing 20 mol% of N2Si10, PASA(ODPA-TTM-N2Si25) containing 5 mol% of N2Si25 and PASA(ODPA-TTM-N2Si25) containing 20 mol% of N2Si25 were 30.6, 26.5, 30.5 and 35.2 cm³·g⁻¹ (in NMP/THF=4:6, 25°C), respectively.

All final materials were insoluble in NMP, N,N-dimethylformamide, toluene, chloroform, tetrahydrofuran and methanol. The IR spectra of the materials HBPI, PIS2, PIS3 and PIS5 are collected in Figure 3. The absorption bands at approximately 1780 and 1720 cm⁻¹ (symmetric and asymmetric stretching of the ring carbonyl groups) and the band at 1360 cm⁻¹ (stretching of the ring C-N bond) are distinct in the spectra and characterize the formation of imide structures. The absence of the band at 1670 cm⁻¹ (amide group of polyamic acid) supports the conclusion that the thermal treatment (see Materials and Methods, 2.3) led to almost complete imidization. In the spectra of PIS2, PIS3 and PIS5, the broad band at 1010–1080 cm⁻¹ characterizes siloxane moieties (Si-O-Si) in these materials (12). Because the spectra were not standardized, they offer qualitative information only.

Figure 3:

IR spectra of the materials HBPI, PIS2, PIS3 and PIS5 (from top to bottom).

The thermal and mechanical properties of the products are collected in Table 2. To draw any conclusions from the data in Table 2, it is necessary to realize that the amount of N2Si10 in the materials PIS3 and PIS4 is slightly lower and the amount of N2Si25 in the materials PIS5 and PIS6 is lower than the theoretical assumption. The similar phenomenon was observed for the linear poly(imide-siloxane)s (17). It is probably a consequence of a different solubility of the PASA based on N2Si10 or N2Si25 in a mixed solvent NMP/THF. The amount of siloxane moieties in the products was re-calculated from the incombustible fractions on the presumption that it represents silicon dioxide (silica) (Table 2). Both the glass transition temperature (T_g) and the temperature corresponding to 10% weight loss decrease slightly with siloxane content. Nevertheless, these materials can still be considered to be highly thermally resistant. The T_g values given in Table 2 characterize the imide phase (T_g of the siloxane phase is approximately −120°C). The shift of these T_gs values in comparison with the T_g value of the pure HBPI is influenced by a miscibility/immiscibility of the imide and siloxane phases (the better miscibility, the deeper T_g decrease). It is supposed that the phase separation is more distinct with a length of siloxane oligomer (12). Then, the T_g drop is more pronounced for the materials containing the shorter siloxane oligomers with molar mass of 1000 g mol⁻¹. The siloxane dimer mainly contributes to making these products more linear in comparison with the hyperbranched control. As a consequence of this, the tensile strength increases and the modulus decreases. On the other hand, the presence of siloxane oligomers renders the materials not only more linear but also more elastic (rubbery), and therefore, both tensile strength and modulus decrease and elongation at break increases slightly.

Table 2:

Thermal and mechanical properties of the products.

Material	T_g^a (°C)	T₁₀^b (°C)	Ash^c (%)	σ^d (MPa)	E^e (MPa)	ε^f (%)
HBPI	369	492	0 (0)	78	4500	2.4
PIS1	361	500	2 (2)	128	4800	3.4
PIS2	–	456	10 (10)	157	3100	–
PIS3	354	430	9 (11)	50	2100	4.4
PIS4	323	392	19 (24)	33	750	6.5
PIS5	361	433	10 (34)	41	1600	4.9
PIS6	331	409	27 (53)	20	250	13.7

^aGlass transition temperature (by using DMA).
^bTemperature corresponding to 10 % weight loss (by using TGA in air).
^cIncombustible fraction (by TGA at 800°C in air); theoretical amount is given in parentheses.
^dTensile strength (by Instron).
^eTensile modulus (by Instron).
^fElongation at break (by Instron).

The gas transport properties of these materials (in the form of self-standing films with a thickness of approximately 50 μm) are also in good agreement with their chemical compositions. The permeability coefficients of hydrogen, carbon dioxide and methane in the membranes prepared from the pure HBPI and the PISs are compared in Table 3. Nevertheless, the permeability coefficients of methane are very low in some cases. Therefore, the repeatability of such values is very poor and they were left out.

Table 3:

Gas transport properties of HBPI and PISs.

Material	Permeability coefficient (10¹⁷ mol m⁻¹·s⁻¹·Pa⁻¹)			Selectivity
Material	H₂	CO₂	CH₄	H₂/CO₂	CO₂/CH₄
HBPI	27	7	0.3	3.9	23.3
PIS1	22	5	–	4.4	–
PIS2	23	7	–	3.3	–
PIS3	27	7	–	3.9	–
PIS4	559	997	405	0.6	2.5
PIS5	32	6	0.3	5.3	20
PIS6	636	2914	768	0.2	3.8

From the data given in Table 3, it is clear that the membranes based on HBPI and PIS1, PIS2, PIS3 and PIS5 show a common order of the gas permeability coefficients in the flat, homogeneous (non-porous) polymeric membranes [hydrogen>carbon dioxide (>methane)]. The carbon dioxide/methane selectivities of approximately 20 are rather lower in comparison with those presented in the literature (17), (18). The size of gas molecules and their ability to diffuse through the membrane are decisive in this case. However, this order is not valid for the membranes made of PIS4 and PIS6. A higher content of longer siloxane moieties increases the contribution of the solubility of carbon dioxide in the membranes, and their behaviour for the hydrogen/carbon dioxide correspond to so called reverse-selective polymeric membranes (18).

4 Conclusion

Poly(imide-siloxane)s based on the hyperbranched polyimide based on 4,4′-oxydiphthalic anhydride and 4,4′,4″-triaminotriphenylmethane were prepared. The use of both an amine-terminated dimethylsiloxane dimer and amine-terminated dimethylsiloxane oligomers differing in their molar mass as the source of siloxane moieties enabled us to vary the properties of the final materials. All of these products form self-standing films with a thickness of approximately 50 μm, and due to their lower brittleness compared to the pure HBPI, they are potentially useful as polymeric membranes. Nevertheless, the optimal reaction conditions for the preparation of materials based on the amine-terminated siloxane oligomers will be studied. It will enable us to synthesize materials with more accurate chemical compositions and to explore the composition-property relationships in detail.

Acknowledgements

The support of the Grant Agency of the Czech Republic (No. 15-06479S) and from the specific university research fund is acknowledged.

References

1. Tomalia DA. Birth of new macromolecular architecture dendrimers as quantized building blocks for nanoscale synthetic polymer chemistry. Prog Polym Sci. 2005;30(3–4):294–324.10.1016/j.progpolymsci.2005.01.007Search in Google Scholar

2. Voit BI, Lederer A. Hyperbranched and highly branched polymer architecture-synthetic strategies and major characterization aspects. Chem Rev. 2009;109(11):5924–73.10.1021/cr900068qSearch in Google Scholar PubMed

3. Jikei M, Kakimoto M. Dendritic aromatic polyamides and polyimides. J Polym Sci. Part A. 2004;42(6):1293–309.10.1002/pola.20018Search in Google Scholar

4. Abadie MJM, editor. High performance polymers-polyimides based (from chemistry to applications). Rijeka: InTech; 2012. 37–62 p.10.5772/2834Search in Google Scholar

5. Liaw D-J, Wang K-L, Huang Y-C, Lee KR, Lai JY, Ha CS. Advanced polyimide materials. Syntheses, physical properties and applications. Prog Polym Sci. 2012;37(7):907–74.10.1016/j.progpolymsci.2012.02.005Search in Google Scholar

6. Vanherck K, Koeckelberghs G, Vankelecom FJ. Crosslinking polyimides for membrane applications. A review. Prog Polym Sci. 2013;38(6):874–96.10.1016/j.progpolymsci.2012.11.001Search in Google Scholar

7. Gao C, Yan D. Hyperbranched polymers: from synthesis to applications. Prog Polym Sci. 2004;29(3):183–275.10.1016/j.progpolymsci.2003.12.002Search in Google Scholar

8. Rybak A, Grzywna ZJ, Sysel P. Mixed matrix membranes composed of various polymer matrices and magnetic powder for air separation. Sep Purif Technol. 2013;118:424–31.10.1016/j.seppur.2013.07.026Search in Google Scholar

9. Sysel P, Minko E, Cechova R. Preparation and characterization of hyperbranched polyimides based on 4,4′,4″-triaminotriphenylmethane. E-Polymers. 2009;9(1):976–85.10.1515/epoly.2009.9.1.976Search in Google Scholar

10. Bershtein VA, Egorova LM, Sysel P. DSC study of main relaxations in the poly(imide-dimethylsiloxane) block copolymers. J Macromol Sci Phys. 1998;B37(6):747–63.10.1080/00222349808212414Search in Google Scholar

11. Sysel P, Hobzova R, Sindelar V, Brus J. Preparation and characterization of crosslinked polyimide-poly(dimethylsiloxane)s. Polymer. 2001;42(26):10079–85.10.1016/S0032-3861(01)00579-1Search in Google Scholar

12. McGrath JE, Dunson DL, Mecham SJ, Hedrick JL. Synthesis and characterization of segmented polyimide-polyorganosiloxane copolymers. In: Vol 140 of Progress in polyimide chemistry I. Advances in polymer science. Berlin: Springer; 1999. 61–105 p.10.1007/3-540-49815-X_3Search in Google Scholar

13. Yilgör E, Yilgör I. Silicone containing copolymers: synthesis, properties and applications. Prog Polym Sci. 2014;39(6):1165–9.10.1016/j.progpolymsci.2013.11.003Search in Google Scholar

14. Lei X, Qiao M, Tian L, Chen Y, Zhang Q. Evolution of surface chemistry and morphology of hyperbranched polysiloxane polyimides in simulated atomic oxygen environment. Corros Sci. 2015;98:560–72.10.1016/j.corsci.2015.05.060Search in Google Scholar

15. Pergal VM, Dzunuzovic JV, Poreba R, Ostojic S, Radulovic A, Spirkova M. Microstructure and properties of poly(urethane-siloxane)s based on hyperbranched polyester of the fourth pseudo generation. Prog Polym Coat. 2013;76(4):743–56.10.1016/j.porgcoat.2013.01.007Search in Google Scholar

16. Friess K, Sysel P, Minko E, Hauf M,Vopicka O, Hynek V, Pilnacek K, Sipek M. Comparison of transport properties of hyperbranched and linear polyimides. Desal Water Treat. 2010;14(1–3):165–9.10.5004/dwt.2010.1022Search in Google Scholar

17. Ghosh A, Sen SK, Dasgupta B, Banerjee S, Voit B. Synthesis, characterization and gas transport properties of new poly(imide-siloxane) copolymers from 4,4′-(4,4′-isopropylidenediphenoxy)bis(phthalic anhydride). J Membr Sci. 2010;364(1–2):211–18.10.1016/j.memsci.2010.08.015Search in Google Scholar

18. Lau CH, Li P, Li F, Chung TS, Paul DL. Reverse-selective polymeric membranes for gas separations. Prog Polym Sci. 2013;38(5):740–66.10.1016/j.progpolymsci.2012.09.006Search in Google Scholar

Received: 2017-5-25

Accepted: 2017-7-4

Published Online: 2017-7-25

Published in Print: 2018-2-23

Articles in the same Issue

https://doi.org/10.1515/epoly-2017-0103

Keywords for this article

gas transport properties; hyperbranched polyimide; poly(imide-siloxane); siloxane moiety; thermo-mechanical properties