Applying ARGET ATRP to the Growth of Polymer Brush Thin Films by Surface-initiated Polymerization

Introduction

In 2006, Matyjaszewski and co-workers first described controlled radical polymerization by Activators ReGenerated by Electron Transfer Atom Transfer Radical Polymerization (ARGET ATRP),^1-3 a development of the very widely used ATRP technology. ARGET ATRP offers two principle advantages over conventional ATRP: improved tolerance of oxygen and significantly reduced heavy metal catalyst concentrations.

Using initiator molecules with specific surface-binding groups (e.g., thiols, silanes, or catechols), monolayers containing a high density of initiator sites can be produced. When polymerization is initiated from a fraction of these sites (the exact fraction depending on the polymerization system and conditions), a layer of densely-packed chains is produced, known as polymer brushes. Experimentally, such polymerizations are conducted by immersing an initiator-functionalized surface into bulk monomer or monomer solution, with additional components such as catalysts and free (untethered) initiator added depending on the exact polymerization type.

From an applications perspective, these coatings can be considered as ultra-thin (typically 0-500 nm) polymer films with outstanding stability towards solvents due to covalent chain tethering. These films can be immersed in a good solvent for the polymer without fear of damage or dissolution. The coating will reversibly swell in a good solvent, an effect which can be utilized for sensing applications.⁵ This solvent stability has been particularly well-exploited in producing robust coatings of hydrophilic polymers, with applications such as non-biofouling/proteinresistant surfaces⁶ and antimicrobial coatings⁷ for proposed use in biosensors and medical devices. In addition, the “bottom-up” nature of the polymer growth allows thin films to be produced free of pin-hole defects commonly encountered with spin-coating, and makes them suitable for use as dielectric layers in electronic devices.⁸ Furthermore, patterning these polymer layers is simple, requiring only that the initiator layer is generated using any of the existing well-developed monolayer patterning techniques (e.g., micro-contact printing).⁹

In this article, we will discuss ARGET ATRP, particularly focusing on the application of this technique to surface-initiated polymerization (SIP). This case study provides an excellent example of how using ARGET ATRP in place of conventional ATRP can help to broaden the applicability and appeal of a polymerization process.

We will first introduce SIP and explore why controlled radical polymerization (CRP) in general, and ATRP in particular, has become such a popular choice of polymerization technology for SIP. After introducing ARGET ATRP, we will discuss how it enables SIP to be more widely applied outside of the synthetic chemistry laboratory.

Surface-initiated Polymerization (SIP) and Polymer Brushes

Chain-growth polymerizations (e.g., radical addition polymerization) frequently use initiators intimately mixed with the monomers either in solution or in bulk. However, if initiators can be tethered to a surface (using established chemistry from the field of self-assembled monolayers),⁴ then chains grown from these initiators will also be endtethered (or grafted) to the surface. Such a process is termed surfaceinitiated polymerization (SIP). The overall process for SIP is shown in Figure 1.

Controlled Radical Polymerization for SIP

Most chain-growth polymerization techniques which have been utilized to synthesize polymers in bulk or in solution have also been applied to growing polymer brushes. However, the vast majority of recent research in coatings has focused on CRP technology such as nitroxide-mediated polymerization (NMP), reversible addition/fragmentation chain transfer polymerization (RAFT) and most notable ATRP. For detailed coverage of the application of CRP to polymer brush growth, see reviews by Edmondson et al.¹⁰ and Barbey et al.¹¹

Using CRP allows the molecular weight of the grafted chains to be easily controlled, which in turn controls the thickness of the brush layer. Perhaps the simplest way of achieving this control is to vary the polymerization time — experimentally, this can be realized by simply removing samples from the reaction periodically as required. Very thin uniform films (just a few nanometers thick) can be produced in this way using slow polymerization methods and short growth times, although thicknesses of up to ~500 nm are possible. Gradients of thickness can also be produced by slowly feeding polymerization solution (monomer and catalyst in solvent) into or out of the vessel containing the substrate.¹² Thickness control can also be achieved by adding free (untethered) initiator and allowing the polymerization to run to high conversion, with the molecular weight being determined by the [monomer]/[initiator] ratio. However, this method produces free polymer which can make sample extraction and cleaning difficult.

A wide range of monomers have been utilized to prepare polymer brushes via CRP. Some polymer brushes can also be modified after growth, for example by coupling biomolecules, giving an even wider range of available functionality. Moreover, multi-block copolymers producing multi-layer films can also be made by sequential polymer growth with different monomers (Figure 2).

ATRP has proved to be the most popular CRP technology for the formation of polymer brushes, with hundreds of published examples. The beneficial features of ATRP for solution polymerization (e.g., tolerance to a wide range of functional groups, ability to polymerize at room temperature in environmentally benign solvents with good control) are also applicable to SIP brush growth. A wide range of acrylates, methacrylates, and styrenic monomers have been used to make brush coatings with surface-initiated ATRP. Many diverse classes of polymers have been grown this way, including hydrophobic, hydrophilic, charged (cationic and anionic), stimuli-responsive (pH and temperature), biocompatible, cell-adhesive, antimicrobial, and chemically reactive (for post-growth functionalization or crosslinking).¹¹

In addition to the wide range of accessible polymer functionality, the popularity of ATRP for SIP can also be attributed to the ease of synthesis of suitable alkyl halide initiators. An acyl bromide initiator, 2-bromoisobutyryl bromide (BIBB) (Figure 3) is commercially available (252271) and can be used to synthesize a range of surface-reactive initiators such as silanes (for silica and some metal oxides) and catechols (for some metal oxides). A disulfide-functional initiator is also commercially available (733350 in Figure 3) for use with noble metals (e.g., gold) and select oxide-free transition metal surfaces (e.g., copper).

Surfaces bearing organic hydroxyl groups (C-OH) can be more directly modified to incorporate initiating sites by directly reacting with BIBB, an approach which has been used to functionalize surfaces such as cellulose¹³ and oxidized/activated polymer surfaces (shown in Figure 4). Our group has also been working on more general (non-substratespecific) initiators, based on polyelectrolytes¹⁴ and dopamine polycondensation¹⁵ to further broaden the application of SIP.

ARGET ATRP for Robust Brush Growth

ATRP has been very successfully used for surface-initiated polymerization, but suffers one major drawback — the oxygen sensitivity of the Cu(I) complexes typically used for polymerization. This sensitivity necessitates careful deoxygenation of solvents and polymerization vessels. Matyjaszewski and co-workers have recently developed a version of ATRP, termed ARGET ATRP which has greatly reduced sensitivity to oxygen by the introduction of an excess of reducing agent in the reaction.^1-3

The effect of oxygen in a conventional ATRP polymerization is to oxidize the Cu(I) to Cu(II). Since the polymerization rate is proportional to the ratio [Cu(I)] / [Cu(II)], a small amount of oxygen leads to a large reduction in polymerization rate. ARGET ATRP overcomes this problem by having a reservoir of reducing agent in the reaction mixture. Cu(II) generated by oxidation, or other processes such as termination or disproportionation, is simply reduced back to the active Cu(I) (Figure 5).

ARGET ATRP offers further advantages over conventional ATRP:

• Copper catalyst can be added initially as Cu(II) and generate active Cu(I) in situ via reduction at the start of the polymerization. Therefore, only air-stable Cu(II) salts need to be stored and handled in the laboratory, rather than air-sensitive Cu(I) salts.
  
• The total amount of copper catalyst can be greatly reduced, even down to ppm levels,1 since the polymerization rate theoretically only depends on the ratio [Cu(I)] / [Cu(II)], not the total amount of copper. Conventional ATRP must employ much greater concentrations of copper than ARGET ATRP to minimize the effects of oxidation and termination.

The controlled synthesis of a range of polymers in solution (i.e., not tethered) using ARGET ATRP have been reported, for example hydrophobic polymers such as polystyrene (PS),2 poly(methyl methacrylate) (PMMA),^1,3 and poly(n-butyl acrylate) (PnBA),^1,3 as well as hydrophilic poly(2-hydroxyethyl methacrylate) (PHEMA).¹⁶ The high level of control possible with ARGET ATRP has been demonstrated by the synthesis of block copolymers.^1,3 In some cases ARGET ATRP allows the synthesis of polymers that are not accessible by conventional ATRP, such as high molecular weight acrylonitrile homopolymers¹⁷ and copolymers.¹⁸

Interestingly, some reported ARGET ATRP polymerizations do not require the addition of a reducing agent, for example tertiary amine groups present on the ligand¹⁹ or monomer (2-dimethylaminoethyl methacrylate, DMAEMA)²⁰ can act as ′intrinsic′ reducing agents.

The benefits of ARGET ATRP over conventional ATRP make it a particularly attractive system for the growth of polymer brushes by SIP. The less stringent experimental conditions required make ARGET ATRP ideal for use by scientists and engineers with all levels of synthetic expertise, greatly broadening the application of polymer brush growth. Furthermore, low catalyst concentrations, ambient temperature polymerization conditions (for many monomers) and relatively benign solvents and reducing agents (e.g., methanol/water mixes and ascorbic acid, respectively) make this technique more applicable for use outside the chemistry laboratory, and extendable into industrial settings.

Sample handling with ARGET ATRP is also simplified since it is sufficient for the polymerization to be conducted in a sealed jar under air, because the small amount of oxygen present will be consumed by the excess reducing agent. However, to prepare polymer brush surfaces by conventional ATRP the system must be enclosed in rigorously deoxygenated flasks or entirely conducted in a glove box.

The tolerance of the polymerization to low levels of oxygen also allows samples to be removed from a polymerization vessel during the reaction without terminating the polymerization. Samples can even be analyzed (e.g., ellipsometric thickness) and returned to the polymerization reaction if a thicker layer is required. Conventional ATRP would require an extensive process for re-initiation including transfer to a new deoxygenated reaction flask with fresh polymerization solution.

Research utilizing ARGET ATRP for surface initiated polymerization is continuing to grow. Matyjaszewski and coworkers first demonstrated the technique by synthesizing PnBA brushes and PnBA-block-PS copolymer brushes.²¹ More recently, chemically functional polymer brushes such as epoxy-functional poly(glycidyl methacrylate)¹³ and hydroxyl-functional PHEMA¹⁵ have also been prepared by ARGET ATRP. In addition, PMMA brushes via ARGET ATRP have been grown from a variety of substrate materials including silicon wafers,¹⁵ high-surface-area porous silica,²² and imogolite nanotubes (an aluminosilicate clay).²³

Our group has had significant success in simply adapting existing ATRP protocols into ARGET ATRP protocols, especially those run at ambient temperature using polar solvents (e.g., methanol and water), by simply applying the following ′algorithm′:

• Monomer and solvents quantities are kept identical
  
• Replace Cu(I) salts with CuBr2 and greatly reduce the overall copper concentration. (Typical [monomer]:[Cu] = 5000:1)
  
• Keep the same ligand species, but increase the amount of ligand. (Typical [ligand]:[Cu] = 10:1)
  
• Add reducing agent (typically ascorbic acid or sodium ascorbate), in large excess over copper. (Typical [reducing agent]:[Cu] = 10:1)

To conduct ARGET ATRP in the most ideal manner (achieving the best control and the lowest polydispersity), further optimizations need to be made. For example, a stronger chelating ligand (e.g., Me₆TREN or TPMA)¹ may need to be used and the concentration and nature of the reducing agent should be optimized. However, in surface-initiated polymerization a simple recipe produced by the algorithm above is often sufficient to grow a brush layer, where polydispersity and good re-initiation efficiency are not of critical importance. A typical polymerization procedure for PMMA brushes, developed using this algorithm, is given at the end of this article. It should be noted that if an ARGET ATRP brush growth protocol produces brush layers which are too thin, as measured by ellipsometry or other techniques, this could occur for two quite different reasons:

• The polymerization is too fast, causing high levels of termination. In this case, the polymerization can be slowed by using less polar solvents or less active reducing agents.
• The polymerization is too slow. In this case, the polymerization can be accelerated by using more polar solvents (e.g., adding water to the solvent mix), more active reducing agents, or adopting more rigorous oxygen exclusion.

These two scenarios can be differentiated by growing samples for a range of times, making obvious the difference between early termination (constant thickness with increasing time) and slow propagation (slowly increasing thickness).

Conclusion

In this short article, we have reviewed the fundamental principles of ARGET ATRP and discussed the advantages of this technique over conventional ATRP. We have focused on the application of ARGET ATRP for surface-initiated polymerization to prepare polymer brushes. Using ARGET ATRP in this application reduces the experimental complexity of brush growth, in particular by requiring less rigorous inert atmosphere conditions. The synthesis of a range of useful polymers (both surfacetethered and free) by ARGET ATRP has been demonstrated in literature, and this useful technique is likely to grow in importance for controlled polymer synthesis both inside and beyond the chemistry laboratory.