Novel multi end-functionalised polymers. Additives to modify polymer properties at surfaces and interfaces

Abstract

Fréchet-type poly(arylether) dendrons with up to eight trifluoromethyl peripheral groups have been used as initiators in the copper mediated living radical polymerisation of d₈-styrene, the surface adsorption behaviour and resulting properties of these polymers in blends have been investigated by ion beam analysis and contact angle measurements.

---

An ongoing challenge in polymer science is the preparation of materials with specific surface properties which differ from that of the bulk, for example, hydrophobicity, wettability, adhesion or biocompatibility, whilst retaining the advantageous mechanical properties of the bulk polymer. In the overwhelming majority of cases when the surface properties of a polymer are modified, this is achieved via a costly post-processing procedure. This may involve the application of a surface coating layer or chemical modification of the surface. Fluorinated polymer surfaces are particularly appealing in terms of their liquid repellence, chemical inertness and low coefficient of friction. Imparting these attributes to a surface can result in the polymers finding use in applications such as anti-fouling finishes, biomedical devices, release coatings and filter media.¹ Current methodologies for fluorinating polymer surfaces include plasma treatments,^1–4 wet chemical modification^5–7 and the application of polymeric surface coatings.^8,9 All or most of these approaches tend to suffer from (at least one of) being expensive, restricted to batch processing, safety hazards or the generation of solvent waste.

In this work, we describe an alternative strategy in which low surface energies may be achieved without any post-processing procedure by the use of a fluorinated polymer additive which may be added during processing. Previous attempts to use a surface segregating polymer additive to modify surface properties have had varying degrees of success. Previous work in our group, using polystyrenes end-functionalised with a single C₆F₁₃ group failed to achieve any measurable change in contact angle.¹⁰ A similar study¹¹ using low molecular weight polystyrene additives (circa 7000 g mol⁻¹) and a single C₆F₁₃ end group in a blend with high molecular weight unmodified polystyrene showed evidence of surface segregation and modest increases in contact angle with water to 93–94°. The additive in the current work consists of a low generation benzyl ether dendron (G0–G2) decorated with multiple trifluoromethyl groups at the periphery of the dendron and attached to a linear polymer chain at the focus of the dendron. The presence of multiple trifluoromethyl groups renders the additive surface active and the linear polymer tail compatibilises the additive with the bulk polymer which should provide a stable and durable surface layer.

The strategic plan for the synthesis of the dendrons involved the successive Williamson ether coupling of a trifluoromethyl decorated benzyl bromide with 3,5-dihdroxylbenzyl alcohol and bromination of the core benzylic alcohol functionality to generate an ‘active’ dendron which can be used for generation growth (Scheme 1), analogous to the methodology developed by Fréchet et al.^12,13 The brominated dendrons can then be used as initiators for the copper mediated living radical polymerisation of styrene to generate trifluoromethyl end-functionalised polymers of desired molecular weights and narrow polydispersities. A summary of deuterated polystyrene (DPS) samples prepared from initiators 1, 3 and 5 is given in Table 1.

Scheme 1 Synthesis of ditrifluoromethyl(benzene)-terminated dendrons. Reagents and conditions: (i) 3,5-dihydroxybenzyl alcohol, K₂CO₃, 18-crown-6, acetone, reflux 24 h; (ii) CBr₄, PPh₃, THF, ambient temperature, 6–18 h. Products were purified by recrystallisation.

Table 1 Deuterated polystyrene samples prepared by copper mediated living radical polymerisation^a

Initiator	DPS sample	Target Mn/g mol^−1	M n/g mol^−1	M w/Mn
a Polymerisations carried out in bulk under an inert atmosphere using a Cu(I)Br–2,2′-dipyridyl catalytic system. Ratio of components [bromo-initiator] ∶ [Cu(I)Br] ∶ [2,2′-dipyridyl] = 1 ∶ 1 ∶ 2. Reactions were left for 18–24 h at 110 °C. Purification was achieved by dissolving polymer in toluene, passing over an alumina column and precipitating into methanol.
1	1DPS	10 000	8800	1.33
3	3DPS	11 700	10 500	1.26
5	5DPS (10 k)	10 000	9200	1.26
5	5DPS (15 k)	15 000	15 600	1.14

---

Nuclear reaction analysis† was used to investigate the surface adsorption behaviour of these additives. This is a technique^14,15 which allows the absolute determination of concentration of the deuterium-labelled additive with respect to depth of the film. By judicious choice of sample orientation and beam energy, depth resolution better than 10 nm is routinely possible and therefore appropriate for the characterisation of surface excess layers. Polymer blends were prepared by co-dissolving 15 wt% of the trifluoromethyl functionalised d8-polystyrene additive and hydrogeneous polystyrene (M_n = 250 000 gmol⁻¹, M_w/M_n = 1.05) in toluene. Thin films were spin-coated onto silicon substrates from the toluene solutions and films of approximately 150 nm total thickness were obtained. Typical experimental data and model fit are shown Fig. 1.

Fig. 1 DPS concentration versus depth profile for 5DPS (15 k) blend obtained via nuclear reaction analysis. The peak close to the surface (x = 0) indicates the excess of this material found at the surface after annealing. The solid curve indicated the best fit obtained using a two-slab model concentration profile.

A two-layer model was used to fit the experimental data comprising a thin surface layer of variable concentration, ϕ_surface, to describe the adsorbed layer of functionalised polymer, and a thicker sub-layer to describe the functional polymer concentration in the bulk of the film, ϕ_bulk. The two-layer model was convolved with the instrumental resolution (∼6 nm) and fitted to the experimental data by varying ϕ_surface.

The surface excess, z*, of adsorbed functionalised polymer is defined as the integral of volume fraction over the surface region in excess of the bulk concentration. For the two layer model, z* is defined by z* = h(ϕ_surface − ϕ_bulk), where h is the thickness of the surface layer. Surface excess data were found to be insensitive to values of h that were comparable to or less than the instrumental resolution. Given that the radius of gyration of the largest DPS chain was ∼3.5 nm, z* values derived in this way are reliable. For comparison between adsorbates of differing molecular weight, it is convenient to normalise the surface excess with respect to the single polymer chain volume, yielding an effective area per adsorbed chain. Table 2 shows that surface activity increases with both annealing of the film (130 °C for one hour) and the number of CF₃ groups per polymer chain. No surface excess was found for chains end-functionalised with two CF₃ groups, however we note that extremely small surface excesses may be impossible to detect within the finite instrumental resolution of NRA.

Table 2 Surface excess data obtained from nuclear reaction analysis of blended films^a

Sample in blend	Volume/nm^3	ϕ surface ^b	z*/nm^c	nm^2/chain
a The thickness of the excess layer of the model and the bulk volume fraction of DPS were fixed at 4 nm and 0.15 respectively. The volume of each DPS chain was estimated from the weight-average molecular weight assuming a density of 1.12 g cm^−3. b ϕ surface concentration of DPS at surface. c The surface excess, z*, of adsorbed functionalised polymer is defined as the integral of volume fraction over the surface region in excess of the bulk concentration.
1DPS (unannealed)	17.35	0.15	0.00	—
1DPS (annealed)	17.35	0.15	0.00	—
3DPS (unannealed)	19.61	0.28	0.52	37.98
3DPS (annealed)	19.61	0.55	1.61	12.16
5DPS (10 k) (unannealed)	17.19	0.40	0.99	17.30
5DPS (10 k) (annealed)	17.19	0.87	2.89	5.94
5DPS (15 k) (unannealed)	26.37	0.56	1.64	16.11
5DPS (15 k) (annealed)	26.37	1.02	3.49	7.55

---

Since the films are glassy, the formation of a surface excess in the unannealed blended films indicates some surface activity of the functionalised DPS in the toluene solution. This spontaneous adsorption must occur within the timescale of the evaporation of the solvent during the spin-coating process, which is of the order of several seconds. The z* values appear to increase by a factor of approximately two with increasing the number of CF₃ from four to eight per polymer. We can see an interesting dependence of surface excess on molecular weight of DPS chain for the two polymers synthesized from initiator 5. The surface excess of the larger polymer chain is consistently slightly larger than for the smaller chain whether the sample is annealed or not. However, the normalized surface excess indicates a greater number of adsorbed chains per unit area in the case of the lower molecular weight polymer. This indicates that the maximum number of CF₃ groups adsorbed per unit area (and therefore maximum hydrophobicity) is likely to be achieved with relatively low molecular weight functionalised polymers.

Clearly the NRA data demonstrate that the fluoralkyl functionalised additives are surface active and are adsorbed to the polymer/air interface. To determine the effect of the additives on the surface properties of the films contact angle measurements with both water and dodecane were carried out. The bulk unfunctionalised polystyrene used in the blends gave a contact angle of 90° with water whereas dodecane completely wets the surface of the unfunctionalised polystyrene film making the angle unmeasurable. It is found that the 1DPS additive does not affect contact angle either in the annealed and unannealed films which is consistent with the absence of any measurable surface excess from the nuclear reaction analysis. Modest changes in contact angle with dodecane were observed with films containing 3DPS and 5DPS (15 k) with angles around 4° measured. The additive which showed the largest effect on surface properties was 5DPS (10 k) with angles of approximately 92° (water) and 10° (dodecane) observed in both annealead and unannealed films. It is to be noted that the angle with water for both 5DPS containing blends were initially 96° but dropped to 92° after one minute.

Given the surface activity that is apparent from nuclear reaction analysis, the contact angles obtained are smaller then expected. Although these changes in contact angle are small, previous work using polystyrenes end-functionalised with a single C₆F₁₃ group either achieved similarly modest changes¹¹ or failed to achieve any measurable change at all in contact angle.¹⁰ In the present study, even in the case of the most surface-active material, the effective area per adsorbed chain was approximately 6 nm². The effective area per CF₃ group can be estimated from the cross-sectional area perfluorocarbon chain to be approximately 0.28 nm². In this case the surface coverage is at most 37% CF₃, possibly less if some proportion of the CF₃ groups cannot reach the polymer surface. Finally, we should note that the CF₃ group has a small dipole due to the tetrahedral arrangement of the electronegative F atoms. Although adsorption of the CF₃-functionalised polymer must reduce the overall surface energy, it is possible that this process leads to a slight increase in the polarity of the surface. This may explain the small change in contact angle with respect to water (a polar solvent) when compared to the more pronounced changes in contact angle for dodecane. The increasing dodecane contact angle with increasing adsorption of the additive is consistent with a reduction in surface energy. The reason for the unstable contact angle measurements with respect to water for the strongly adsorbing systems is not yet clear. This is an intriguing situation, and suggests that even though the polymer films are glassy, some local reorganisation at the film surface may be possible, allowing a reduction in the number of unfavourable contacts between CF₃ groups and water.

Without doubt these preliminary results demonstrate that the use of polymer additives, end-capped with a dendron carrying multiple trifluoromethyl functionalities does result in a surface layer of a few nanometres depth, which is rich in fluorine. Nuclear reaction analysis data show a significant surface excess of the additives with 8 CF₃ groups although the effect of this fluorine rich surface layer with respect to contact angles is a little disappointing. This is very much a work in progress and we are currently preparing analogous polymer additives end-functionalised with dendrons decorated with C₈F₁₇ groups. It is expected that these materials will show even stronger surface adsorption behaviour, provide better surface coverage and result in significantly improved liquid repellency.

Acknowledgements

The authors are grateful to the European Union, Nanotech UIC and One Northeast for financial support. We also wish to thank Dr A. M. Kenwright for NMR analysis.

Notes and references

1. L. J. Ward, J. P. S. Badyal, A. J. Goodwin and P. J. Merlin, Polymer, 2005, 46, 3986.
2. T. S. Cheng, H. T. Lin and M. J. Chuang, Mater. Lett., 2004, 58, 650.
3. E. Selli, G. Mazzone, C. Oliva, F. Martini, C. Riccardi, R. Barni, B. Macandalli and M. R. Massafra, J. Mater. Chem., 2001, 11, 1985.
4. S. Sigurdsson and R. Shishoo, J. Appl. Polym. Sci., 1997, 66, 1591.
5. T. Kawase and H. Sawada, J. Adhes. Sci. Technol., 2002, 16, 1121.
6. T. Kawase, M. Yamane, T. Fuji and M. Minigawa, J. Adhes. Sci. Technol., 1997, 11, 1381.
7. S. Biltresse, D. Deschamps, T. Boxus and J. Marchand-Brynaert, J. Polym. Sci., Part A: Polym. Chem., 2000, 38, 3510.
8. D. L. Schmidt, C. E. Coburn, B. M. DeKoren, G. E. Potter, G. F. Meyers and D. A. Fischer, Nature, 1994, 368, 39.
9. S. Saïdi, F. Guittard, C. Guimon and S. Géribaldi, Macromol. Chem. Phys., 2005, 206, 1098.
10. L. R. Hutchings, R. W. Richards, R. L. Thompson and D. G. Bucknell, Eur. Phys. J. E, 2002, 8, 121.
11. T. F. Schaub, G. J. Kellogg, A. M. Mayes, R. Kulasekere, J. F. Ankner and H. Kaiser, Macromolecules, 1996, 29, 3982.
12. C. J. Hawker and J. M. J. Fréchet, J. Am. Chem. Soc., 1990, 112, 7638.
13. C. J. Hawker and J. M. J. Fréchet, J. Chem. Soc., Chem. Commun., 1990, 15, 1010.
14. R. S. Payne, A. S. Clough, P. Murphy and P. J. Mills, Nucl. Instrum. Methods Phys. Res., Sect. B, 1989, 42, 130.
15. U. K. Chaturvedi, U. Steiner, O. Zak, G. Krausch, G. Schatz and J. Klein, Appl. Phys. Lett., 1990, 56, 1228.

---

Footnote

† Nuclear reaction analysis experiments were carried out using an NEC 5SDH pelletron ion beam accelerator. A beam of ³He⁺ ions was accelerated to 0.7 MeV and directed onto the sample surface at 83° to the sample normal (“grazing incidence”). Backscattered protons arising from the d(³He,p)α nuclear reaction were detected at 170° to the incident beam. The beam diameter was 2 mm, and measurements were taken over a total charge of 5 µC ³He⁺. Proton spectrum data were converted to volume fraction of deuterated polymer versus depth according to the thick target approximation.

---