A novel strategy for enhancing propagation rate of polystyrene grown from silica nanoparticles or carbon nanotubes

Abstract

Addition of PS into St monomer can accelerate the SI-ATRP and the PS chains were grown to high molecular weight in a short time from silica nanoparticles and carbon nanotubes.

---

Atom transfer radical polymerization (ATRP) provides a robust tool for grafting polymers onto solid substrate surfaces.^1–4 This strategy includes modification of the solid surface with ATRP initiators and surface-initiated ATRP (SI-ATRP). Using this strategy, a variety of polymers have been grafted onto various materials of different geometries and dimensions including flat substrates,¹nanoparticles (NPs),²carbon nanotubes (CNTs)³ and porous materials.⁴

However, for most monomers, SI-ATRP from the surface of silica NPs was ill-controlled,⁵ so addition of a sacrificial initiator is necessary to ensure a sufficient concentration of deactivating species for enabling the polymerization to be controlled.⁶ Compared to the homogeneous polymerization, the SI-ATRP rate was much slower, especially when CuBr₂ was added or a mixed halide CuCl/CuBr₂/ligand was used, which is probably due to geometric constrain resulting from high density of the chains.⁷ When the SI-ATRP was carried out in polar solvent or in aqueous solution, the polymerization rate increased significantly.⁸ But for most hydrophobic monomers, SI-ATRP was usually carried out in bulk or in low-to-middle polar organic solvent, the polymerization rate was fairly slow especially for styrene (St). For example, when polystyrene (PS) was grafted onto the silica NPs via SI-ATRP in bulk without use of Cu(II), the conversion of St was about 25%, and the number-average molecular weight (M_n) of the resultant PS was less than 16000 g mol⁻¹ after 20 h of polymerization.⁹ When CuBr₂ (10 mol% relative to CuBr) was used, the conversion was 13%, and M_n was 18200 g mol⁻¹ after 31.2 h of polymerization.¹⁰ Gao et al. grafted PS onto CNTs via SI-ATRP in diphenyl ether, the M_n of grafted PS was only 11000 after 50 h of polymerization.¹¹ So, it is interesting to find a new strategy for accelerating polymerization, at the same time keeping better control of SI-ATRP. In this article, we report a facile and feasible method for accelerating the SI-ATRP and producing the high molecular weight (M_n>90000) PS in a short time (2 h) only by addition of small amount of thermal, self-polymerized PS into the reaction system.

To prepare PS-grafted silica NPs via SI-ATRP, the first step is synthesis of ATRP initiator-functionalized silica. The synthetic approach is outlined in Scheme S1 of the ESI.† Bare silica NPs with average diameter (D) of 50 nm were prepared by hydrolysis polycondensation of tetraethoxysilane.^2e By treating NPs with commercially available coupling silane agent, 3-aminopropyltriethoxysilane, yielding primary amino group bound onto silica NPs (NH₂-SiO₂), and then it was reacted with glycidol 2-bromoisobutyrate to yield ATRP initiator-functionalized silica NPs (Br-SiO₂). The successful synthesis of Br-SiO₂ was verified by the characteristic carbonyl absorption band at 1730 cm⁻¹ and a C–H stretching band at 2986 cm⁻¹ in the FT-IR spectrum (Fig. S1†). Thermogravimetric analysis (TGA) demonstrates that ∼5.2% weight difference between NH₂-SiO₂ and Br-SiO₂ was observed based on the weight retentions at 700 °C in Fig. 1a. Based on these data, we can calculate the amount of ATRP initiator on silica NPs, which was approximately 320 μmol g⁻¹.

Fig. 1 A: TGA of SiO₂ (1), SiO₂–NH₂ (2), SiO₂–Br (3) and PS–grafted silica NPs obtained from SI–ATRP of St containing 0.5% of PS at different polymerization times; B: TGA of PS–grafted silica NPs obtained from SI–ATRP of St containing 1.0% PS at different polymerization times.

Initially, SI-ATRP of freshly purified St was carried out in anisole (1 : 1 to St, w/w), only a small amount of PS was grafted on silica NPs after 24 h, the grafted PS is still low even after 48 h of polymerization. However, when St containing self-polymerized PS, which resulted from storage in a refrigerator for two months, was used in SI-ATRP occasionally, we unexpectedly observed a viscosity increase of the reaction solution. From above fact, we postulated that the PS in St might have great effect on the SI-ATRP. In order to understand this phenomenon, further study is necessary.

The St containing PS was obtained by setting freshly distilled St at room temperature until thermal, self-initiated polymerization of St, and the content of PS was estimated by weight method. The St (5.03g) was dropped into 50 mL of methanol, and then the collected PS was dried in vacuum at 40 °C until constant weight. St containing 1.8% of PS with M_n(GPC) = 483200 and M_w/M_n = 2.10 was obtained. This St was equally divided into two parts, and each part was diluted with freshly purified St to obtain St containing 0.5% and 1.0% (w/w) of PS, respectively, which were respectively designated as St/PS0.5 and St/PS1.0.

SI-ATRPs of St/PS0.5 and St/PS1.0 in the presence of Br-SiO₂ were carried out in anisole [1 : 1 (w/w) relative to St] at 90 °C using CuBr and N,N,N′,N′,N′′-pentamethyldiethylenetriamine (PMDETA) as the catalytic system. The ratio of [St]₀ : [CuBr]₀ : [PMDETA]₀ was 310 : 1 : 1. Different from the previous reports,⁶ addition of free sacrificial initiator was not needed because the SI-ATRP of St in the presence of Br-SiO₂ with less than 75 nm diameter was well controlled.⁹ The reaction was stopped by plunging the tube into ice water after predetermined time. The resultant PS was diluted with THF, and the polymer solution was placed into excess of methanol. The precipitate was collected, and dried in vacuum at 40 °C until constant weight was reached. The monomer conversion was calculated based on the weight difference between the dried PS and the Br-SiO₂ added, the PS present in the converted monomer, and the results are listed in Table S1.†

To wash out the absorbed PS from the PS-grafted silica NPs (PS-SiO₂), three cycles of dispersion in THF–centrifugation were adopted. The amount of grafted PS was determined by TGA. Fig. 1 shows the TGA curves of PS-SiO₂ obtained from SI-ATRP of St/PS0.5 and St/PS1.0 at different polymerization time. For the polymerization of St/PS0.5 for 4 h, the weight loss was 65.9% at 700 °C, while for polymerization of the St/PS1.0 for 2 h, the weight loss was as high as 81.7%. The results agreed well with the monomer conversion, indicating that almost all converted monomer was grafted onto the silica NPs, and the PS in St can enhance the SI-ATRP rate of St (relative to the polymerization without PS), while the rate increase became bigger with the content increase of PS in the St.

To get information about the molecular weights and polydispersities of the grafted polymer, the PS-SiO₂ NPs were treated with hydrofluoric acid in THF. The cleaved PS was hydrolyzed under alkaline conditions, and then the PS isolated by precipitation was analyzed by GPC, and the results are shown in Fig. S2.† We can clearly see that the elution peaks continuously shift to higher molecular weight with increase of the reaction time in both cases. For SI-ATRP of St/PS05 for 4 h, the monomer conversion was 18.7%, M_n and M_w/M_n of the cleaved PS are 42700 g mol⁻¹ and 1.10. While for SI-ATRP of St/PS1.0 for 2 h, the conversion was as high as 35.3%, M_n and M_w/M_n were 96900 g mol⁻¹ and 1.10, respectively. In both cases, the conversions and M_ns are much higher than the data reported in literature.^9–11

We examined whether the polymerization is of “living” character for the SI-ATRP without sacrificial initiator. The plots for the SI-ATRP of St/PS0.5 and St/PS1.0 in the presence of Br-SiO₂ are shown in Fig. 2a and 2b, and both exhibited linear first-order kinetics. Fig. 3a and 3b show the relationship of M_n and M_w/M_n with conversion of St for SI-ATRP of St/PS0.5 and St/PS1.0, respectively. We can clearly observe the linear increase of M_n with monomer conversions in both cases. Moreover, the polydispersity values remained narrow (≤1.20) throughout the polymerization. These results suggest that the PS brushes were grown from the Br-SiO₂ NPs in a controlled fashion, which is similar to the previous result.⁹ But the both straight lines did not pass through the original point, which may be attributed to low initiation efficiency of Br-SiO₂ NPs.

Fig. 2 Kinetic plot of SI-ATRP of St/PS0.5 (a) and St/PS1.0 (b) in the presence of Br-SiO₂ NPs in anisole at 90 °C.

Fig. 3 Relationship of the molecular weight and molecular weight distribution with conversion for SI-ATRP of St/PS0.5 (a) and St/PS1.0 (b) at 90 °C in the presence of Br-SiO₂.

The growth of PS brushes with the monomer conversion can be directly observed from TEM images of PS-SiO₂ NPs obtained from different conversions of St, generally, the average interparticle distances increased with chain length increase of the grafted polymer.^5a This phenomenon was also observed in the TEM image shown in Fig. S3,† we can see that the distance between the NPs increased obviously when M_n of the grafted PS increased from 29200 formed at 7.3% conversion to 96900 g mol⁻¹ at conversion of 35.3%.

In order to check that the propagation rate increase of SI-ATRP was caused by PS chains or by monomer St, the PS with M_n(GPC) = 200300 and M_w/M_n = 2.23 was prepared by conventional free radical polymerization, and then was used in SI-ATRP of the freshly distilled St, almost the same result was obtained. For example, when the SI-ATRP of freshly distilled St containing 0.5% of PS was carried out 90 °C for 4 h, the monomer conversion was 19.1%, which was no difference from the SI-ATRP of St containing 0.5% of thermal, self-initiated polymerized PS (18.7% in Table S1†) in the range of experiment error. This result indicated that the rate increase of SI-ATRP was resulted from PS exclusively.

On the other hand, homogeneous ATRPs of freshly distilled St and St/PS0.5 with the same recipe and the same reaction conditions was carried out for 4 h, respectively, the conversions were 41.6% and 42.0%, respectively. The results suggested that the PS chains presented in the polymerization system have no effect on the homogeneous ATRP. Thus the freshly distilled St and the storage St do not affect the SI-ATRP.

The slow growth of PS from the silica NPs was attributed to the faster irreversible termination of the radicals in the local surface area of SiO₂ particles than that formed in solution due to high density of surface initiating sites and high initiation rate at initial polymerization, leading to significant decrease of the bromine groups on the surface of silica particles and an increase in the Cu(II) concentration in the solution, in turn, to a shift in the activation–deactivation equilibrium, thereby, the formation of dormant species is favored.^7,10 This explanation was supported by the fact that the initiation efficiency (f_in) of SI-ATRP was lower than that of ATRP in the solution. The following phenomenon we observed might also support this explanation. When the freshly purified St was used in SI-ATRP, we saw change of the color from light green to turquoise blue, indicating the formation of more Cu(II), and the viscosity of polymerization solution had no obvious change during the polymerization, and amount of the grafted PS was low after 24 h of polymerization. This may be resulted from that irreversible termination of the neighbor radicals tethered on the surface was much faster than the initiation. But for the SI-ATRP of St/PS, the color change was not so obvious, and the viscosity increased rapidly with increase of polymerization time. The possible reason was that the PS chains were absorbed on the silica NPs, and they affected the polymerization in two ways: 1 these initiating sites embedded in the adsorbed PS couldn't be activated because the catalyst was difficult to reach them; 2 some of the unembedded initiating sites might be activated, and initiated the polymerization to form chain radicals, their lifetime increased because the adsorbed PS chains hinder their irreversible termination. Both cases would reduce the formation of Cu(II), increase the lifetime of free radicals, thereby, favor the propagation reaction to produce long PS chains in short time. This explanation was also supported by reduction of the f_ins shown in Table S1.† Because some of the initiating sites were embedded in the adsorbed PS, the f_in would be reduced. Both PS chains, the adsorbed PS and the PS grown from the unembedded initiating sites, would hinder the catalyst reaching the residual unembedded initiating sites, resulting in further reducing of the f_in. Moreover, we observed the decrease of f_in with increase of the adsorbed PS. As shown in Table S1,† when the content of PS in the original St increased from 0.5% to 1.0%, the f_in decreased from about 15% to 12%, which were much lower than that (25%) reported.⁵ This further confirms that the PS chains in the St were absorbed on the surface of Br-SiO₂, and blocked the attaching of the catalyst. In addition, the results in Table S1 shows that the f_ins are independent with conversion of St, which is different from increase of f_in with increase of conversion described in the previous reports.⁵ Probably, contribution of the absorbed PS chains on blocking initiating sites is bigger than that of the PS chains grown from silica NPs, and the exact reason needed to be further studied.

To test whether the strategy of accelerated polymerization using the SI-ATRP of St/PS can be applied to other substrate surface, multiwalled carbon nanotubes (MWCNTs) were selected. The attachment of ATRP initiators onto MWCNTs was similar to the previous report.¹¹ Amount of the anchored ATRP initiator was 270 μmol g⁻¹ determined by TGA. SI-ATRP of St from MWCNTs was carried out in anisole at 90 °C using CuBr and PMDETA as catalytic system. The ratios of [St]₀ : [initiator]₀ :  [CuBr]₀ :  [PMDETA]₀ was 740 : 1 : 1 : 1. After washing with THF three times for removal of the free PS, the MWCNT-g-PSs were analyzed by TGA to determine the amount of grafted PS. As shown in Fig. 4, for MWCNT-g-PS prepared from SI-ATRP of the purified St for 24 h, the weight loss was 66.1% at 800 °C, while for the MWCNT-g-PS obtained from SI-ATRP of St/PS0.5 for 6 h, the weight loss was as high as 88.9%. It must point out that unlike SI-ATRP from silica NPs, the influence of the PS in the monomer on the SI-ATRP of St from MWCNTs is not so big. This preliminary study suggests that PS in the monomer can accelerate the propagation rate of PS chains on the MWCNTs. Based on these results; we can optimistically predict that this method can be applied to modification of other substrate surfaces with different geometry and dimensions. But this prediction needs to be tested by further study.

Fig. 4 TGA of pristine MWNTs (a), MWNT-OH (b), MWNT-Br (c), MWNT-g-PS prepared by SI-ATRP of the purified St (d) and MWNT-g-PS obtained from SI-ATRP of St/PS0.5 (e).

In summary, by addition of around 1% PS into St, the rate of SI-ATRP can be significantly increased, and PS chains anchored the silica NPs and MWCNTs can be grown to high molecular weight in short time, and the polymerization rate increased with increase of the PS content in monomer St. A possible reason is absorption of the added PS on the surface of substrates, leading to a reduction in irreversible termination.

This work is supported by the Dr Foundation of Hefei University of Technology under contract No. 2010HGBZ0297 and the National Natural Science Foundation of China under contract No. 50673086 and 50633010.

References

1. (a) Y. Zheng, M. L. Bruening and G. L. Baker, Macromolecules, 2007, 40, 8212–8219; (b) W. Huang, G. L. Baker and M. L. Bruening, Angew. Chem., Int. Ed., 2001, 40, 1510–1512; (c) J. Y. Wang, W. Chen, A. H. Liu, G. Lu, G. Zhang, J. H. Zhang and B. Yang, J. Am. Chem. Soc., 2002, 124, 13358–13359; (d) A. M. Granville and W. J. Brittain, Macromol. Rapid Commun., 2004, 25, 1298–1302.
2. (a) P. Zhou, G. Q. Chen, C.-Z. Li, F. S. Du, Z. C. Li and F. M. Li, Chem. Commun., 2000, 797; (b) T. Werne and T. E. Patten, J. Am. Chem. Soc., 1999, 121, 7409–7410; (c) D. Li, G. L. Jones, J. R. Dunlap, F. Hua and B. Zhao, Langmuir, 2006, 22, 3344–3351; (d) T. K. Mandal, M. S. Fleming and D. R. Walt, Nano Lett., 2002, 2, 3; (e) C. H. Liu and C. Y. Pan, Polymer, 2007, 48, 3679–3685; (f) J. T. Park, J. H. Koh, J. K. Koh and J. H. Kim, Appl. Surf. Sci., 2009, 255, 3739–3744.
3. (a) H. Kong, C. Gao and D. Yan, J. Am. Chem. Soc., 2004, 126, 412–413; (b) C. Y. Hong, Y. Z. You, D. Wu, Y. Liu and C. Y. Pan, Macromolecules, 2005, 38, 2606–2611; (c) L. P. Yang, P. Zou and C. Y. Pan, J. Mater. Chem., 2009, 19, 1843–1849.
4. (a) Q. Fu, G. V. R. Rao, L. K. Ista, Y. Wu, B. P. Andrzejewski, L. A. Sklar, T. L. Ward and G. P. Lopez, Adv. Mater., 2003, 15, 1262–1266; (b) C. Y. Hong, X. Li and C. Y. Pan, Eur. Polym. J., 2007, 43, 4114–4122.
5. (a) Y. P. Wang, X. W. Pei, X. Y. He and K. Yuan, Eur. Polym. J., 2005, 41, 1326–1332; (b) M. Save, G. Granvorka, J. Bernard, B. Charleux, C. Boissière, D. Grosso and C. Sanchez, Macromol. Rapid Commun., 2006, 27, 393–398.
6. (a) A. El Harrak, G. Carrot, J. Oberdisse, J. Jestin and F. Boue, Polymer, 2005, 46, 1095–1104; (b) P. Pasetto, H. Blas, F. Audouin, C. Boissière, C. Sanchez, M. Save and B. Charleux, Macromolecules, 2009, 42, 5983–5995.
7. S. Edmondson, V. L. Osborne and W. T. S. Huck, Chem. Soc. Rev., 2004, 33, 14–22.
8. (a) R. Barbey, L. Lavanant, D. Paripovic, N. Schüwer, C. Sugnaux, S. Tugulu and H.-A. Klok, Chem. Rev., 2009, 109, 5437–5527; (b) X.-S. Wang, S. F. Lascelles, R. A. Jackson and S. P. Armes, Chem. Commun., 1999, 1817–1818.
9. T. Werne and T. E. Patten, J. Am. Chem. Soc., 2001, 123, 7497–7505.
10. J. Pyun, S. Jia, T. Kowalewski, G. D. Patterson and K. Matyjaszewski, Macromolecules, 2003, 36, 5094–5104.
11. H. Kong, C. Gao and D. Yan, Macromolecules, 2004, 37, 4022–4030.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental section, Table S1 and Fig. S1–S3. See DOI: 10.1039/c0py00338g

---