Controlled synthesis and characterization of poly[methyl(3,3,3-trifluoropropyl)siloxane] with selective end groups

Abstract

Anionic ring-opening polymerization of 1,3,5-tris(trifluoropropylmethyl)cyclotrisiloxane in the bulk was studied using dilithium diphenylsilanediolate as an initiator (I); and N,N-dimethylformamide (DMF), bis(2-methoxyethyl)ether (Diglyme), and 1,2-dimethoxyethane (DME) as promoters (P). A detailed study on the kinetics of polymerization with different molar ratios of promoter to initiator ([P]/[I]) that were equal to 2.0, 4.0, and 6.0 revealed that the yield of linear polymers was the highest when [P]/[I] = 2.0 for all the promoters, among which DME was the most efficient promoter for suppressing the backbiting reactions. The reaction promoted by DME had a very broad “termination window” with the highest yield of linear polymer and very narrow molar mass distribution. The results of the matrix-assisted laser desorption/ionization time of flight mass spectrometry indicated that the intermolecular redistribution occurred during the process of polymerization. PMTFPS with end groups such as vinyl, hydroxyl, hydrogen and chloromethyl were prepared and characterized by ¹H NMR, ²⁹Si NMR and FT-IR. Polymers having vinyl end groups displayed higher thermo stability than those having hydroxyl end groups under nitrogen.

---

Introduction

Poly[methyl(3,3,3-trifluoropropyl)siloxane] (PMTFPS) has a wide range of commercial applications due to superior properties, such as high thermal stability,^1–6 low temperature flexibility,^7–9 radiation resistance,¹⁰ low surface energy,¹¹ and outstanding fuel and chemical resistance because it is capable of introducing fluorine atoms into the polymer structure.^12–15 Traditionally, PMTFPS is synthesized by anionic ring-opening polymerization of 1,3,5-tris(trifluoropropylmethyl)cyclotrisiloxane (D₃^F) in the bulk initiated by sodium hydroxide (NaOH), potassium hydroxide (KOH) or tetramethylammonium hydroxide (TMAOH), which is an equilibrium reaction that has a significant tendency to form the cyclic compounds, D_m^F, (m = 4, 5, 6,…).^12,15–20 Therefore, the reaction residue contains a considerable amount of cyclic products and the yield of the resulting linear polymer is only about 18.3% (Scheme 1),^15–17,21 and the molar mass distribution is broad.

Scheme 1 Equilibrium between linear polymer and cyclic products in bulk or solution polymerization of D₃^F.

To overcome the shortcomings of the equilibrium polymerization, many efforts have been put to study the non-equilibrium polymerization.^{16,18,22–30} Yuzhelevskii and coworkers^22–25 studied the effect of promoters like aprotic and polar solvents including dimethylsulfoxide (DMSO), N,N-dimethylformamide (DMF), and organic phosphates on D₃^F polymerization and found that the polymerization rate could be increased by at least two orders of magnitude by promoting the dissociation of ion pairs between SiO⁻ and Li⁺. Veith et al.¹⁶ investigated the effect of the DMSO/butyllithium (BuLi), promoter–initiator (P–I) combination on the polymerization of D₃^F at 40 °C indicating that the initiator containing lithium ion instead of sodium ion could remarkably reduce the rate of backbiting reaction. Besides, according to their kinetic model, they predicted that the highest yield of polymer was 90%. Barrère et al.¹⁷ polymerized D₃^F by the miniemulsion technique using didodecyldimethylammonium bromide as an emulsifier and NaOH as an initiator. PMTFPS with hydroxyl as end group was prepared and they reported that the maximum yield of the polymerization was almost quantitative and analyzed the by-products formed due to intermolecular redistribution by matrix-assisted laser desorption/ionization time of flight (MALDI-TOF) mass spectrometry. The limitation of miniemulsion technique is that the number-average molar mass and molar mass distribution is difficult to be controlled and the polymerization cannot attain polymer with other end groups such as vinyl, hydrogen or chloromethyl. Therefore, it is important to explore a polymerization method to synthesize PMTFPS with narrow molar mass distribution, selective end groups and high yield.

Moreover, earlier, the researchers focused on the thermal stability of polydimethylsiloxane.^2–5,31,32 For example, Thomas and Kendrick^31,32 proposed the random depolymerization mechanism based on uncatalyzed siloxane bond cleavage. Budden⁵ put forward the unzipping depolymerization and bond exchange mechanism during thermal degradation. Their studies suggested that cleavage of backbone by moisture was the principal mode at lower temperature in the range of 120 to 275 °C. However, the thermal degradation and thermo-oxidative degradation mechanisms of PMTFPS have rarely been studied. Thomas⁴ investigated the thermal stability of end hydroxyl fluorosilicone vulcanisates. They found that fluorosilicones was more susceptible to oxidative attack than methyl silicones because of their high content of methylene groups.

Herein, we investigated the mechanism of non-equilibrium polymerization of D₃^F using dilithium diphenylsilanediolate (Ph₂Si(OLi)₂) as an initiator (I); N,N-dimethylformamide (DMF), bis(2-methoxyethyl)ether (Diglyme), and 1,2-dimethoxyethane (DME) as promoters (P), and achieved the polymerization with high yield, selective end groups, well controlled molar mass and molar mass distribution, and broad “termination window”. PMTFPS with end groups such as α,ω-vinyl, hydroxyl, hydrogen and chloromethyl were attained nearly quantitatively by using chlorodimethylvinylsilane, acetic acid, dimethylchlorosilane and chloromethyldimethylchlorosilane as termination reagent, respectively (Scheme 2). The α,ω-vinyl and hydroxyl PMTFPS are very important addition type and condensation type fluorosilicone rubber, respectively; α,ω-hydrogen PMTFPS is always used as chain extender; and α,ω-chloromethyl PMTFPS can derive out other functional end groups. Besides, –Ph₂SiO– units from Ph₂Si(OLi)₂ were used as markers in MALDI-TOF analysis. Subsequently, the microstructure of the polymer was determined. Furthermore, we investigated the thermal stability of PMTFPS with hydroxyl and vinyl end groups by TGA analysis under nitrogen and air, respectively.

Scheme 2 Synthetic routes of PMTFPS with different end groups.

Results and discussion

Theory

The backbiting reactions and intermolecular redistribution (Scheme 3) are both observed during the course of polymerization initiated by Ph₂Si(OLi)₂ using different promoters in bulk. The backbiting reaction can lead to the decrease of the molar mass; however, the redistribution leads to the increase of the molar mass and the molar mass distribution of the polymer.^16,17

Scheme 3 Possible reactions occurring during polymerization of D₃^F in bulk.

The kinetics of propagation and side reactions can be expressed as eqn (1) and (2), respectively:

The rates of propagation and side reactions are dependent on [SiO⁻]. Based on the assumption that k₁ ≫ k₂, k₃, and k₄, decreasing [SiO⁻] significantly suppresses the side reactions; and the rate of propagation slows down simultaneously. Furthermore, [SiO⁻] shows square root dependence on [I] and [P] (eqn (3)).¹⁶

[SiO⁻] = C[I]^1/2[P]^1/2 (3)

C is a variable that depends on types of promoters and reaction temperature, indicating that when [I] is kept constant, lowing [P] or decreasing the values of C by varying the promoters can decrease [SiO⁻], and therefore, the side reactions can be suppressed.

Kinetics of polymerization

To monitor the progress of the polymerization, the conversion of the monomer (D₃^F) was calculated from the ²⁹Si NMR spectra of the reaction products based on the ratios of peak areas of silicon atoms in linear polymer (δ −22.3 ppm) to the total peaks areas of silicon atoms in D₃^F (δ −9.6 ppm), D₄^F (δ −19.5 ppm), and linear polymer (δ −22.3 ppm): conversion of monomer = area (polymer)/[area (D₃^F) + area (D₄^F) + area (polymer)].^18,26 We could not observe any peaks for the silicon atoms of D₅^F and D₆^F in the ²⁹Si NMR spectra indicating that the amounts of D₅^F and D₆^F in the reaction products were very little. Therefore, they were not included in the calculation. Theoretically, when the value of [P]/[I] is less than 2.0, only some of the initiators will be associated with the promoters. Therefore, the rate of the polymerization will be slow and the molar mass of polymers will be large. Therefore, to achieve the reasonable reaction time and well controlled molar mass of the resulting polymers, we only chose the [P]/[I] = 2.0, 4.0, and 6.0.

In our study, DMF, Diglyme, and DME were used as promoters in the polymerization of D₃^F. The consumption of D₃^F was very rapid after promoters were introduced into the reactor due to the fact that the ion–dipole interactions between Li⁺ and promoters made –SiO⁻Li⁺ a loose ion pair. Thus, SiO⁻ became to be a more active species when performing propagation and other side reactions.¹⁶ The kinetics of D₃^F polymerization is shown in Fig. 1. The propagation rate increases rapidly with the increase in the value of [P]/[I]. Almost all the monomers are consumed in 2, 5, and 15 min when DMF is used as the promoter and [P]/[I] is equal to 2.0, 4.0, and 6.0, respectively (Fig. 1A).

Fig. 1 Kinetics of D₃^F polymerization at 40 °C in bulk initiated by Ph₂Si(OLi)₂ with (A) DMF, (B) Diglyme, and (C) DME as promoters.

When Diglyme is used, the propagation rates slow down insignificantly (Fig. 1B); however, the propagation rates slow down significantly when DME is used as the promoter (Fig. 1C), revealing the following order of the effect of promoting polymerization, DMF > Diglyme > DME. This is because the electron donating ability of nitrogen atoms is stronger than that of oxygen atoms resulting in greater ion–dipole interactions between Li⁺ and nitrogen atoms. The related propagation constants, k* calculated based on the equation: k* = ln([D₃^F]₀/[D₃^F]_t)/t,²⁶ are listed in Table 1. k* increases as the value of [P]/[I] grows for the same promoter, and for the same value of [P]/[I], a higher k* indicate a stronger promoter.

Table 1 The propagation constants, k* and linear polymer yields at different conditions

Promoter	[P]/[I]	k* (cm^−1)	Polymerization time (h)	Linear polymer yields^a (%)	D4^F^b (%)
a The yields were measured by ^29Si NMR after polymerizing for 5 or 10 h.b The yields were measured by ^29Si NMR after polymerizing for 5 or 10 h.
DMF	2.0	0.170	5	94.4	5.6
DMF	4.0	0.850	5	91.3	8.7
DMF	6.0	1.210	5	89.6	10.4
Diglyme	2.0	0.100	5	95.4	4.6
Diglyme	4.0	0.380	5	93.9	6.1
Diglyme	6.0	0.740	5	90.6	9.4
DME	2.0	0.033	10	Cyclic products were not detected
DME	4.0	0.041	5	Cyclic products were not detected
DME	6.0	0.052	5	99.0	1.0

---

However, the rapid rate of propagation is not always advantageous. Table 1 shows that for the weakest promoter, DME, almost no by-product, D₄^F is obtained in the reaction products; however, for DMF and Diglyme, little amount of D₄^F is obtained in the reaction products and the yields of D₄^F increase as the values of [P]/[I] increase. Therefore, the results indicate that when either stronger promoters or higher concentration of promoters are used in the reactions, –SiO⁻ not only attacks cyclic trimer silicon, but also Si–O bond in the same polymer chain resulting in the backbiting reactions. Therefore, although lowering the values of [P]/[I] slows down the propagation rate, simultaneously the backbiting reactions are suppressed. We further investigated the evolution of the molar mass during the process of polymerization with each of the three promoters when [P]/[I] = 2.0. For reactions facilitated by DMF and Diglyme, and terminated by chlorodimethylvinylsilane (Fig. 2A and B), although the monomer conversion reaches almost 100% in 30–40 min, subsequently, the number average molar mass (M_n) keeps on increasing, which is possibly because the intermolecular redistribution is dominant in 30–60 min. The M_n starts decreasing after 60 min because of the backbiting reactions.

Fig. 2 Kinetics of D₃^F polymerization at 40 °C in bulk initiated by Ph₂Si(OLi)₂ with (A) DMF, (B) Diglyme, and (C) DME as promoters; [P]/[I] = 2.0 in all the cases.

The ²⁹Si NMR spectra and GPC chromatograms of the above mentioned reactions after 5 h (Fig. 3A(1 and 2) and B(1 and 2)) clearly show the appearance of D₄^F in the products due to the backbiting reactions.¹⁸ However, when DME was used as the promoter and [P]/[I] = 2.0, the monomer is consumed in about 2 h, and later, the M_n does not show significant change for almost 8 h (Fig. 2C). D₄^F is not observed in the ²⁹Si NMR spectrum and GPC chromatogram of the polymerization after 10 h (Fig. 3A(3) and B(3)), indicating the absence of backbiting reactions. Furthermore, the resulting polymer has narrower molar mass distribution compared to those from reactions in which DMF and Diglyme were used as promoters (Fig. 2). Therefore, when DME was used as the promoter, the reaction had a broad “termination windows” with high polymer yields and a narrow molar mass distribution. This is mainly because DME is a weak promoter, and backbiting side reactions and intermolecular redistributions can be considerably suppressed.

Fig. 3 ²⁹Si NMR spectra (A) and GPC trace (B) of resulting polymers from reactions (1) DMF as the promoter, [P]/[I] = 2.0, polymerization for 5 h, (2) Diglyme as the promoter, [P]/[I] = 2.0, polymerization for 5 h, and (3) DME as the promoter, [P]/[I] = 2.0, polymerization for 10 h.

Intermolecular redistributions analyzed by MALDI-TOF-MS

To investigate the intermolecular redistributions during the polymerization, the resulting polymer of the reaction in which DME was used as promoter and [P]/[I] = 2.0 was analyzed by MALDI-TOF-MS. The reaction was quenched by addition of acetic acid before the crude product was characterized by MALDI-TOF. The mass spectrum of the reaction product (Fig. 4) shows four groups of peaks: A, B, C, and D. The m/z values of peaks A1, A2, A3, and A4 correspond to molecules containing one –Ph₂SiO– unit (P) and 18, 19, 20, 21 –SiOCH₃CH₂CH₂CF₃ units (F) with a sodium ion, respectively (H–PF₁₈–OH + Na⁺: [M + Na]⁺/z = 3047.2; H–PF₁₉–OH + Na⁺: [M + Na]⁺/z = 3230.1; H–PF₂₀–OH + Na⁺: [M + Na]⁺/z = 3358.9; H–PF₂₁–OH + Na⁺: [M + Na]⁺/z = 3515.2). The A series only has one P unit, which indicates the absence of intermolecular redistribution in these polymers. The m/z values of peaks B1, B2, B3, and B4 correspond to H–P₂F₁₇–OH + Na⁺ ([M + Na]⁺/z = 3088.3), H–P₂F₁₈–OH + Na⁺ ([M + Na]⁺/z = 3245.1), H–P₂F₁₉–OH + Na⁺ ([M + Na]⁺/z = 3401.1), and H–P₂F₂₀–OH + Na⁺ ([M + Na]⁺/z = 3557.3), respectively. There are two P units presented in B series; therefore, this group of molecules must have undergone intermolecular redistribution at least once (Scheme 4(1)). The m/z values of peaks C1, C2, C3, and C4 correspond to H–P₃F₁₆–OH + Na⁺ ([M + Na]⁺/z = 3131.0), H–P₃F₁₇–OH + Na⁺ ([M + Na]⁺/z = 3286.7), H–P₃F₁₈–OH + Na⁺ ([M + Na]⁺/z = 3443.0), and H–P₃F₁₉–OH + Na⁺ ([M + Na]⁺/z = 3599.1), respectively. The three P units indicate that the intermolecular redistribution occurs in this C series at least twice (Scheme 4(2)). The m/z values of peaks D1, D2, D3, and D4 correspond to H–F₂₀–OH + Na⁺ ([M + Na]⁺/z = 3161.2), H–F₂₁–OH + Na⁺ ([M + Na]⁺/z = 3316.8), H–F₂₂–OH + Na⁺ ([M + Na]⁺/z = 3473.1), and H–F₂₃–OH + Na⁺ ([M + Na]⁺/z = 3628.9), respectively. The D series also displays products of intermolecular redistribution (Scheme 4(1) and (2)). The MALDI-TOF analysis suggests that although DME was a weak promoter, the intermolecular redistribution cannot be absolutely suppressed during the process of polymerization.

Fig. 4 MALDI-TOF mass spectrum of resulting polymer obtained by polymerization initiated by Ph₂Si(OLi)₂ and using DME as the promoter, [P]/[I] = 2.0, polymerization for 2 h, terminated by equivalent amount of acetic acid.

Scheme 4 Possible intermolecular redistributions occurred during the process of polymerization.

Chracterization of PMTFPS

¹H NMR analysis. The ¹H NMR spectra of PMTFPS with end groups such as vinyl (P1), hydroxyl (P2), hydrogen (P3) and chloromethyl (P4) are shown in Fig. 5A and B. Peak 1 (Fig. 5A) (δ 0.30 ppm) corresponds to –CH₃ attached to silicon, peak 2 (from δ 0.87 to δ 0.91 ppm) corresponds to α –CH₂– of trifluoropropyl and peak 3 (from δ 2.14 to δ 2.30 ppm) corresponds to β –CH₂– of trifluoropropyl. Both of the methylene groups appear to be multiple peaks duo to the coupling action between hydrogen and fluorine. Peak 4 (from δ 7.21 to δ 7.75 ppm) (Fig. 5B) is assigned to the phenyl group. Peak 5 (from δ 5.60 to δ 6.76 ppm) is assigned to end group, vinyl. Peak 6 (δ 5.67 ppm) is assigned to end group, hydroxyl. Peak 7 (δ 4.78 ppm) is assigned to end group, hydrogen, and peak 8 (δ 2.91 ppm) is assigned to end group, chloromethyl.

Fig. 5 (A) ¹H NMR spectra of PMTFPS with vinyl (P1), hydroxyl (P2), hydrogen (P3) and chloromethyl (P4) end groups. (B) The zoomed view of (A) from δ 8.0 to δ 2.5 ppm.

²⁹Si NMR analysis

Polymers P1, P2, P3 and P4 are characterized by ²⁹Si NMR also. As shown in Fig. 6, peak a (δ −22.3 ppm) corresponds to silicon of repeat unit; peak b (δ −46.1 ppm) corresponds to silicon in the main chain that is from initiators. Both of peaks appear for all the polymers, which mean the initiators were successful in initiating the polymerization. Peak c (δ 1.8 ppm) is assigned to silicon of –OSi(CH₃)₂CHCH₂; peak d (δ −14.5 ppm) is assigned to silicon of –OSiCH₃CH₂CH₂CF₃OH; peak e (δ −4.7 ppm) is assigned to silicon of –OSi(CH₃)₂H; peak f (δ 3.7 ppm) is assigned to silicon of –OSi(CH₃)₂CH₂Cl. Both of the ¹H NMR and ²⁹Si NMR data clearly demonstrates that the PMTFPS with different end groups were synthesized successfully and the living ends of polymers were terminated completely.

Fig. 6 ²⁹Si NMR spectra of PMTFPS with vinyl (P1), hydroxyl (P2), hydrogen (P3) and chloromethyl (P4) end groups.

FT-IR analysis

These polymers are further characterized by FT-IR. As displayed in Fig. 7. Peaks at 1068.6 and 1017.6 cm⁻¹ corresponds to the extension vibration of Si–O–Si in the backbone; peaks at 1210.9, 1127.7, and 898.7 cm⁻¹ are attributed to extension vibration of SiCH₂CH₂CF₃ present in the side chains; peaks at 1369.5 and 1446.4 cm⁻¹ are attributed to flexural vibration of SiCH₃; peaks at 2966.6 and 2909.9 cm⁻¹ corresponded to the extension vibration of –CH₂– and –CH₃, respectively. The peaks observed at 3057.2 and 1594.1 cm⁻¹ are ascribed to the end Si–CHCH₂ groups and phenyl groups. Peaks at 3699.4 cm⁻¹ is ascribed to the end Si–OH groups, peaks at 2130.8 cm⁻¹ is ascribed to the end Si–H groups. The absorption bands of C–Cl are from 800 to 600 cm⁻¹, which are covered by those of other groups. The amount of initiators and termination reagents added was so little that the characteristic absorption bands of end groups exhibited in the ¹H NMR, ²⁹Si NMR and FT-IR spectra are very weak.

Fig. 7 FT-IR spectra of PMTFPS with vinyl (P1), hydroxyl (P2), hydrogen (P3) and chloromethyl (P4) end groups.

Thermal stability analysis

α,ω-Vinyl and hydroxyl PMTFPS are very important addition type and condensation type fluorosilicone rubber, respectively; and they are widely used as heat resistance materials. Therefore, the thermal stability of the polymers with vinyl and hydroxyl as end groups under N₂ and air were evaluated by TGA studies (Fig. 8A and B). Fig. 8A displays that the temperature at which there is 5% weight loss (Td₅) is 417 °C for divinyl-terminated PMTFPS (P1) and Td₅ is around 327 °C for dihydroxyl-terminated PMTFPS (P2) when the TGA analysis was performed under N₂. Thomas and Kendrick studied the thermal degradation of PDMS and proposed the formation of the cyclic dimethylsiloxane due to the rearrangement of the siloxane bond during the depolymerization.^31,32 Probably, similar process happened during the thermal analysis of PMTFPS. For P2, the hydroxyl groups at the end attack the Si–O bond readily leading to the formation of D_m^F; thus, resulting in the weight loss (Scheme 5(1)); however, for P1, a higher temperature is needed for random intermolecular and intramolecular degradations, among which only intramolecular degradations produce D_m^F (Scheme 5(2) and (3)). The weights of the residues from P1 and P2 are nearly zero, which indicates that all the samples are degraded to the volatile D_m^F with small repeat units.

Fig. 8 TGA of P1 (divinyl-terminated PMTFPS, M_n = 1.2 × 10⁴ g mol⁻¹) and P2 (dihydroxyl-terminated PMTFPS, M_n = 1.4 × 10⁴ g mol⁻¹). (A) P1 and P2 were under nitrogen atmosphere with heating rate 10 K min⁻¹; (B) P1 and P2 were under air atmosphere with heating rate 10 K min⁻¹.

Scheme 5 Possible depolymerization and oxidation at high temperature for PMTFPS with vinyl or hydroxyl end group.

When the thermal analysis was performed under air, the Td₅ are 312 °C for P1 and 306 °C for P2, which are close to each other. This is because not only the end depolymerization and random degradation (Scheme 5(1) and (2)) occur, but also in the presence of O₂, the free radicals are formed and the side chains of the polymer start degrading, CO₂, H₂O, HCHO and a few of CF₃CHCH₂ are released (Scheme 5(4)). The free radicals from the oxidative degradation of the side chain cross-links with each other. The formation of volatile D_m^F led to the lowering of the weight percentages of the residues of P1 and P2 compared to their theoretical value (38%), if it was assumed that only the reactions represented in Scheme 5(4) happened and SiO₂ was the sole product. Furthermore, the weight percentage of the residue of P1 was higher than that of P2. This is probably because vinyl groups at the end of P1 undergo the free radical reaction more easily and form the cross-linking structures on the surface of the polymer (Scheme 5(5)); thus, preventing the random depolymerization to some extent.

Conclusions

The kinetics studies of polymerization of D₃^F using three promoters with different ability to facilitate the reactions at various ratios of [P] to [I] showed that the backbiting reactions and intermolecular redistribution could be suppressed by using weak promoters or by decreasing the concentration of promoters. Therefore, a broad “termination window” with high yield of linear polymer and narrow molar mass distribution could be achieved by manipulating the types of the polymer and values of [P]/[I]. PMTFPS with end groups such as α,ω-vinyl, hydroxyl, hydrogen and chloromethyl were successfully synthesized with quantitative transformation. Moreover, the results indicated that MALDI-TOF was extremely effective in investigating the intermolecular redistribution during the polymerization. PMTFPS with vinyl end groups showed higher thermo stability than those with hydroxyl end groups under nitrogen. This study helped to better understand the mechanism and process of D₃^F anionic polymerization and provided a simple and efficient method to prepare PMTFPS with different end groups exhibiting unique properties.

Experimental

Materials

D₃^F was purchased from Weihai Newera Chemical Co., Ltd., dried using calcium hydride (CaH₂), and distilled under reduced pressure. The purity of D₃^F (99.5%) was assessed by ²⁹Si NMR and ¹H NMR. n-Butyllithium (2.5 mol L⁻¹ solution in n-hexane), dimethylchlorosilane ((CH₃)₂SiHCl; purity: 97%) and chloromethyldimethylchlorosilane (ClCH₂(CH₃)₂SiCl; purity: 97%) were purchased from Beijing InnoChem Science & Technology Co., Ltd. DMF, bis(2-methoxyethyl)ether (Diglyme), and 1,2-dimethoxyethane (DME) were bought from J & K Chemical Co., Ltd. (purity: 99%). Chlorodimethylvinylsilane ((CH₃)₂ViSiCl; purity: 97%) was purchased from Aldrich. Ph₂Si(OLi)₂ was synthesized in our laboratory whose purity was 97% estimated by solid state ²⁹Si NMR, ⁷Li NMR, and titration.

Measurements

Gel permeation chromatography (GPC) was performed on a system composed of a Waters 515 HPLC Pump, a set of Waters μ-styragel columns (pore size: 10³, 10⁴, 10⁵ Å), and a Waters 2410 refractometric detector. Tetrahydrofuran (THF) was used as the eluent and the flow rate was set at 1 mL min⁻¹ (30 °C). Samples were prepared in THF at 5–10 mg mL⁻¹ and rapidly injected after preparation to avoid depolymerization. Molar mass and molar mass distribution of PMTFPS was determined from a calibration curve based on narrowly dispersed polystyrene standards. The data were collected and analyzed with Millennium 3.05.

A Bruker Autoflex III time-of-flight mass spectrometer equipped with a nitrogen laser (337 nm), a delayed extraction, and a reflector was used for the MALDI-TOF analysis. 4-Chloro-alpha-cyanocinnamic acid was used as matrix and it was dissolved in THF at 10 mg mL⁻¹. Polymer solution (10 μL, 2–5 mg mL⁻¹) in THF was mixed with the matrix solution (50 μL), and sodium iodide solution (10 μL, 5 mg mL⁻¹ in THF) was added to facilitate the ionization by cation attachment. A 1 μL portion of the final solution was deposited onto the sample target and allowed to dry in air at room temperature. Internal standards (peptides) were used to calibrate the mass scale. The data were collected and analyzed with FlexControl and FlexAnalysis, respectively.

The PMTFPS terminated by (CH₃)₂ViSiCl was dissolved in deuterated acetone and their ¹H NMR spectra were recorded on a Bruker AVANCE 400 spectrometer at ambient temperature. Chromium acetylacetonate was added to the polymer solution in deuterated acetone, as a relaxing reagent and the ²⁹Si NMR spectra were recorded on a Bruker DMX 300 spectrometer.

Solid state NMR experiments were carried out at room temperature on a Bruker AVANCE III 400 spectrometer operating at a static magnetic field of 7.05 T using a 4 mm magic angle spinning (MAS) probe. ²⁹Si and ⁷Li NMR spectra were recorded under MAS conditions (sample rotation frequency: ²⁹Si NMR was 5 kHz, and ⁷Li NMR was 12 kHz.) with cross polarization (CP) excitation, using π/2 pulse widths of 4.0 μs. The ²⁹Si and ⁷Li NMR chemical shifts were determined relative to external standards of TMS and LiCl (δ = 0 ppm), respectively. CP MAS ²⁹Si NMR: δ −34.57 [s, (C₆H₅)₂Si(OLi)₂]. CP MAS ⁷Li NMR: δ 2.62 [s, (C₆H₅)₂Si(OLi)₂].

Fourier transform infrared (FT-IR) of PMTFPS with different end groups spectra were collected on a Bruker TENSOR-27 IR spectrometer.

Thermogravimetric analyses (TGA) were performed on a SII EXTRA 6300 TG/DTA apparatus, using 6–10 mg of polymer, at a heating rate of 10 K min⁻¹ with nitrogen or air as the purge gas at 20.0 mL min⁻¹.

Synthesis of dilithium diphenylsilanediolate (Ph₂Si(OLi)₂)

The initiator, Ph₂Si(OLi)₂ was prepared in a three neck flask from a solution of dry diphenylsilanediol (10.80 g, 50 mmol) dissolved in 60 mL THF under dry nitrogen atmosphere. n-Butyllithium (21.00 mL, 52.5 mmol) was slowly introduced to the solution in 1 h at −78 °C so that it turned to a primrose yellow color. THF was removed with rotary evaporation; the resulting solid product was washed 3 times with n-hexane; and a 10.40 g faint yellow powder was obtained.³³

Polymerization

For all the reactions, D₃^F (9.36 g, 20 mmol) and Ph₂Si(OLi)₂ (0.1142 g, 0.5 mmol) were well mixed by sonication in a polymerization tube under nitrogen. Then, the mixture was heated to 40 °C in an isothermal oil bath and continuously stirred at 300 rpm. The process of polymerization started when the promoters were added to the reaction tubes. The amount of the promoters added to the reaction were determined according to the value of [P]/[I]. Aliquots were taken from the reaction mixture, quenched by adding slightly excess amount of chlorodimethylvinylsilane, acetic acid, dimethylchlorosilane or chloromethyldimethylchlorosilane (Scheme 2), and analyzed by ²⁹Si NMR in order to monitor the progress of polymerization. PMTFPS with different end groups were synthesized using Ph₂Si(OLi)₂ as initiator and DME as the promoter, [P]/[I] = 2.0. Samples were purified before TGA test as follows: 5 g of the synthetic polymer was dissolved in 10 mL of ethyl acetate and then precipitated through adding 10 mL of methyl alcohol. Repeated this manipulation several times until the lithium chloride was removed completely.

Acknowledgements

We are grateful to National Nature Sciences Foundation of China (no. 50803070) for financial support. Special acknowledgements are given to Prof. Xie Zemin for his enduring help during the work and Mrs Guiyun Xu for her assistance on TGA tests.

Notes and references

1. Y. Liu, H. Liu, R. Zhang, C. Zhou and S. Feng, Polym. Eng. Sci., 2013, 53, 52–58.
2. S. Luciano and H. William, Ind. Eng. Chem., 1958, 50, 1583–1584.
3. J. M. Nielsen, J. Polym. Sci., Polym. Symp., 1973, 40, 189–197.
4. D. K. Thomas, Polymer, 1972, 13, 479–483.
5. G. Budden, J. Coated Fabr., 1998, 27, 294–308.
6. S. Z. Wu and J. E. Mark, Polym. Rev., 2007, 47, 463–485.
7. K. E. Polmanteer, Rubber Chem. Technol., 1988, 61, 470–502.
8. T. Hu, L. Dai, X. Gao, J. Xie, W. Xie, Z. Xie and Z. Zhang, J. Polym. Sci., Part A: Polym. Chem., 2014, 52, 1408–1421.
9. T. M. Madkour and J. Mark, Macromolecules, 1995, 28, 6865–6870.
10. Y. Liu, C. Zhou and S. Feng, Mater. Lett., 2012, 78, 110–112.
11. H. Inoue, A. Matsumoto, K. Matsukawa, A. Ueda and S. Nagai, J. Appl. Polym. Sci., 1990, 40, 1917–1938.
12. B. Li, S. Chen and J. Zhang, Polym. Chem., 2012, 3, 2366–2376.
13. H. Kobayashi and W. Nishiumi, Macromol. Chem. Phys., 1993, 194, 1403–1410.
14. C.-M. Kuo, J. C. Saam and R. B. Taylor, Polymer Int., 1994, 33, 187–195.
15. L. M. Yi, X. L. Zhan, F. Q. Chen, F. Du and L. B. Huang, J. Polym. Sci., Part A: Polym. Chem., 2005, 43, 4431–4438.
16. C. A. Veith and R. E. Cohen, J. Polym. Sci., Part A: Polym. Chem., 1989, 27, 1241–1258.
17. M. Barrère, C. Maitre, M. A. Dourges and P. Hèmery, Macromolecules, 2001, 34, 7276–7280.
18. W. Peng and Z. Xie, Macromol. Chem. Phys., 2000, 201, 1292–1294.
19. Z. Shi and X. Wang, Polym. Adv. Technol., 2009, 20, 1017–1023.
20. J. E. Mark, Acc. Chem. Res., 2004, 37, 946–953.
21. A. K. Sutar, T. Maharana, S. Dutta, C.-T. Chena and C.-C. Lin, Chem. Soc. Rev., 2010, 39, 1724–1746.
22. Y. A. Yuzhelevskii, V. V. Pchelintsev and Y. G. Kagan, J. Polym. Sci., 1973, 15, 1795–1799.
23. Y. A. Yuzhelevskii, Y. G. Kagan, N. P. Timofeyeva, T. D. Doletskaya and A. L. Klebanskii, J. Polym. Sci., 1971, 13, 208–216.
24. Y. A. Yuzhelevskii, Y. G. Kagan and N. N. Fedoseyeva, J. Polym. Sci., 1970, 12, 1800–1809.
25. Y. A. Yuzhelevskii, Y. G. Kagan, E. V. Kogan, A. L. Klebanskii and N. N. Nikiforova, J. Polym. Sci., 1969, 11, 1745–1750.
26. Y. Zhang, Z. Zhang, Q. Wang and Z. Xie, J. Appl. Polym. Sci., 2006, 102, 3510–3516.
27. H. W. Ahn and S. J. Clarson, Silicon, 2011, 3, 157–161.
28. X. Gao, Q. Wang, H. Sun, Y. Tan, Z. Zhang and Z. Xie, Phosphorus, Sulfur, and Silicon and the Related Elements,  DOI:10.1080/10426507.2013.860534.
29. T. Suzuki, Polymer, 1989, 30, 333–337.
30. M. Cypryk, B. Delczyk, A. Juhari and K. Koynov, J. Polym. Sci., Part A: Polym. Chem., 2009, 47, 1204–1216.
31. T. H. Thomas and T. C. Kendrick, J. Polym. Sci., Part A-2, 1970, 8, 1823–1830.
32. T. H. Thomas and T. C. Kendrick, J. Polym. Sci., Part A-2, 1969, 7, 537–549.
33. P. B. Kevin, C.-M. Kuo, L. M. Robert and C. S. John, Macromolecules, 1995, 28, 790–792.

---

Footnote

† Electronic supplementary information (ESI) available: CP MAS ²⁹Si-NMR spectra, CP MAS ⁷Li-NMR. See DOI: 10.1039/c4ra12268b

---