Highly-reactive haloester surface initiators for ARGET ATRP readily prepared by radio frequency glow discharge plasma

Abstract

New surface initiators for ARGET ATRP (activators regenerated by electron transfer atomic transfer radical polymerization) have been prepared by the plasma deposition of haloester monomers. Specifically, methyl 3-bromopropionate (M3BP), methyl 2-chloropropionate, and ethyl 2-fluoropropionate (E2FP) were plasma deposited onto glass discs using RF glow discharge plasma. This technique creates surface coatings that are resistant to delamination and rich in halogen species making them good candidates for surface initiators for ARGET ATRP. Of all the plasma polymerized surface coatings, M3BP showed the highest halogen content and was able to grow 2-hydroxyethyl methacrylate (HEMA) polymer brushes on its surface via ARGET ATRP in as little as 15 min as confirmed by XPS. Surprisingly, E2FP, a fluoroester, was also able to grow HEMA polymer brushes despite fluorine being a poor leaving group for ARGET ATRP. The versatility of RF glow discharge plasma offers a clear advantage over other techniques previously used to immobilize ARGET ATRP surface initiators.

I. INTRODUCTION

The ability to modify surfaces with thin polymeric films allows one to tailor properties such as wettability,¹ biocompatibility,² biocidal activity,³ adhesion,⁴ and adsorption.⁵ Surface initiated atom transfer radical polymerization (SI ATRP) is one such technique that has become a powerful tool for the preparation of functional surfaces and interfaces to match the desired needs of a specific biomedical application.⁶ Using this technique, researchers have been able to graft various polymers, such as polystyrene,⁷ poly(methyl methacrylate) (PMMA),⁸ poly(oligoethyleneglycol methacrylate),⁹ poly(hydroxyethyl methacrylate) (PHEMA),¹⁰ poly(N-isopropyl acrylamide),¹¹ poly(carboxybetaine methacrylate),^12,13 poly(sulfoxybetaine methacrylate),^12,14 and many other polymers onto various surfaces.¹⁵

One major drawback of ATRP is that it requires inert atmospheric conditions as a small amount of oxygen causes Cu(I) to oxidize to Cu(II), which results in a significant reduction in polymerization rate.¹⁶ In order to overcome this, activators regenerated by electron transfer atomic transfer radical polymerization (ARGET ATRP) use a reservoir of reducing agent such as ascorbic acid¹⁷ in the reaction mixture, which reduces Cu(II) to Cu(I) in situ. The benefits of ARGET ATRP over conventional ATRP make it a particularly attractive system for surface initiated polymerization of polymer brushes, such as PMMA (Ref. 18) and PHEMA (Ref. 19) on a variety of substrates.

The composition and density of surface initiator are important parameters that determine the grafting density and thickness of the grafted polymer.²⁰ In order to graft a polymer from a surface via SI ATRP or SI-ARGET ATRP, a halogen is required to initiate the reaction. Bromoester based initiators such as bromoisobutyryl bromide have often been used to modify the surface to act as initiators for the SI ATRP reaction.^7,12,21,22 Some researchers have also used alpha methyl bromopropionate as an effective initiator to graft polymer chains.^23,24

RF glow discharge plasma polymerization has historically been used to create uniform thin polymer films that are strongly bound to surfaces.^25–27 The Badyal group have used pulsed plasma polymerization of 4-vinylbenzyl chloride^28,29 and 2-bromoethyl acrylate²⁸ to prepare initiator nanofilms for use in ATRP. Several other research groups have immobilized ATRP initiators by plasma polymerizing bromoesters such as allyl-2-bromo-2-methylpropionate³⁰ and ethyl α-bromoisobutyrate^31,32 onto substrates. Additionally, our lab has previously plasma polymerized 2-chloroethyl methacrylate (CEMA) as a surface initiator in order to synthesize and pattern surfaces with poly[(oligoethylene glycol) methyl methacrylate] (POEGMA) polymer brushes grown by SI ATRP.³³

In this paper, we explore ARGET ATRP surface initiators that were prepared using a robust one step, solvent-free synthesis method using RF glow discharge plasma polymerization of haloesters, specifically, methyl-3-bromopropionate (M3BP), methyl-2-chloropropionate (M2CP), and ethyl-2-fluoropropionate (E2FP). Once prepared, we used the plasma polymerized haloester surface as a surface initiator for SI-ARGET ATRP synthesis of PHEMA as proof of concept (Scheme 1). To the best of our knowledge, this is the first time that chloro- and fluoroesters have been plasma polymerized on surfaces and used as surface initiators for ARGET ATRP synthesis of polymer brushes.

Scheme 1.

Schematic representation of the process to grow PHEMA polymer brushes via SI-ARGET ATRP from plasma polymerized haloester surfaces. We first create a surface rich in Br, Cl, or F by plasma depositing haloesters onto glass coverslips. This is followed by placing the plasma polymerized haloester coverslips into an ARGET ATRP solution in excess of HEMA to grow PHEMA brushes.

II. EXPERIMENT

A. PART I: Plasma deposition and characterization of haloesters: M3BP, M2CP, and E2FP

1. Materials

8 mm glass coverslips were purchased from Knitell Glass (Cat. No. G401-08). M3BP was purchased from Sigma-Aldrich (Cat. No. 3395-91-3), M2CP was purchased from Alfa Aesar (Cat. No. A18541), E2FP was purchased from TCI America (Cat. No. 349-43-9), and CEMA was purchased from Pfaltz & Bauer (Cat. No. 1888-94-4). All organic solvents were purchased from Fisher Scientific.

2. Plasma polymerization of haloesters

Prior to plasma deposition, 8 mm glass coverslips were placed in a sample holder within a glass container and ultrasonicated in a water bath for two 10 min cycles each in methylene chloride, followed by acetone and then methanol. The glass samples were then air-dried in a fume hood.

Plasma polymerization of haloesters was carried out in a custom-built plasma reactor setup based on the design reported in Lopez et al.³⁴ Cleaned glass coverslips were loaded in the plasma reactor and positioned in between the powered and grounded electrodes (100 mm apart), and a mechanical pump was used evacuate the reactor to the desired base pressure in the low 10⁻³ Torr range. A liquid nitrogen cooled cold trap between the vacuum pump and the reactor was used to condense organic materials. The powered electrode was connected to a 13.56 MHz radio frequency power source and an automatic impedance matching network. The coverslips were first etched in argon gas for 10 min at 40 W while maintaining a pressure of 250 mT and then plasma coated with a methane deposition at 80 W for 5 min at a pressure of 140 mT. This plasma-deposited methane layer generated at high plasma power has been found to allow plasma coatings to strongly adhere to the substrate, thus preventing delamination. Following this, the coverslips were coated with the haloester of choice. In brief, a flask containing approximately 20 ml of M3BP, M2CP, or E2FP was connected to a caliper valve and was used to control the monomer delivery rate into the plasma reactor. The coverslips were coated for 1 min at 80 W in order to form an adhesion promoting layer, after which the power was lowered to 10 W for 10 min, while maintaining a pressure of 150 mT for the monomer throughout the plasma deposition process. Finally, samples were quenched under M3BP, M2CP, or E2FP vapor for 5 min before retrieving them.

CEMA plasma-deposited glass coverslips were used as a positive control for all experiments and were prepared by a modified protocol as described by Hucknall et al.³³ Following the argon etching process and methane plasma deposition, glass coverslips were coated with CEMA for 1 min at 80 W and for 10 min at 10 W, followed by quenching for 5 min. A pressure of 150 mT was also maintained for CEMA throughout the process.

3. Surface characterization

X-ray photoelectron spectroscopy (XPS) was performed on all plasma polymerized glass disc samples on a Surface Science Instruments S-Probe equipped with a monochromatic Al Kα source and a low energy electron flood gun for charge neutralization. X-ray analysis for these acquisitions was in a circle approximately 800 μm across. Pressure in the analytical chamber during spectral acquisition was less than 5 × 10⁻⁹ Torr. Low-resolution survey scans were obtained from 0 to 1100 eV with a 1 eV step size and a pass energy of 150 eV. High-resolution C 1s scans were obtained from 270 to 290 eV with a 0.065 eV step size and a pass energy of 50 eV. The spectra were analyzed offline with Service Physics hawk version 7 data analysis software to calculate the elemental compositions from peak areas and to peak fit high-resolution spectra. All binding energies were calibrated with reference to the aliphatic carbon at C 1s = 285.0 eV. A Shirley background was used for all spectra.

Furthermore, in order to test the robustness of the plasma polymerized surfaces, samples were placed in a glass disc holder and soaked in methanol for 2 h. The samples were then air-dried in a chemical hood and analyzed via XPS.

Additionally, plasma polymerized M3BP surface coatings were analyzed using time-of-flight secondary ion mass spectroscopy (ToF-SIMS) (IONTOF TOF.SIMS 5 spectrometer) using a 25 keV Bi₃⁺ cluster ion source in the pulsed mode and were compared to M3BP that was drop coated onto silicon wafer and air-dried. Plasma polymerized M3BP surface coatings were also soaked in methanol for 2 h, and the surface coatings post methanol soak as well as the extract from the methanol soak were analyzed using ToF-SIMS. Spectra were acquired for both positive and negative secondary ions over a mass range of m/z = 0–850. The ion source was operated with at a current of 0.17 pA. Secondary ions of a given polarity were extracted and detected using a reflectron time-of-flight mass analyzer. Spectra were acquired using an analysis area of 100 × 100 μm². The primary ion dose for each spectrum was 8.7 × 10¹¹ ion/cm². Positive ion spectra were calibrated using the CH₃⁺, C₂H₃⁺, and C₃H₅⁺ peaks. The negative ion spectra were calibrated using the CH⁻, OH⁻, C₂H⁻, C₄H⁻, and Br₂⁻ peaks. Calibration errors were kept below 25 ppm. Mass resolution (m/Δm) for a typical spectrum was between 5000 and 6000 for m/z = 27 (pos) and between 4500 and 6000 for m/z = 25 (neg).

B. PART II: SI-ARGET ATRP reaction of HEMA using plasma-deposited M3BP, M2CP, and E2FP as the surface initiator

1. Materials

HEMA (ophthalmic grade) was purchased from Polysciences, Inc. (Cat. No. 04675). l-Ascorbic acid (Cat. No. A5960), tris(2-pyridylmethyl)amine (Cat. No. 723134), and copper (II) bromide (Cat. No. 221775) were all purchased from Sigma-Aldrich.

2. Synthesis of HEMA polymer brushes on plasma coated haloester surfaces

To grow HEMA brushes on M3BP, M2CP, E2FP, and CEMA plasma-deposited surfaces, we modified a protocol provided by Paterson et al.¹⁹ A solution prepared from HEMA (2 ml, 16.4 mmol), methanol (1.5 ml), stock solutions of CuBr₂/TPMA (44 μl, 0.84 μmol CuBr₂, 4.2 μmol TPMA), and l-ascorbic acid (460 μl of 25 mg/ml of ascorbic acid in methanol) in a round bottom Schlenk flask was degassed by five freeze-pump-thaw cycles, and the flask was backfilled with nitrogen. The solution was then pipetted into a 50 ml falcon tube containing a sample holder with the plasma polymerized substrates, capped, and then allowed to polymerize at ambient room temperature. After the desired polymerization time (5 min, 15 min, 30 min, 1 h, and 2 h), the glass disc holder containing the samples was removed and subsequently rinsed five times with 50% v/v methanol in water and dried in air in a chemical hood. Clean glass discs were used as negative controls.

3. Surface characterization

The polymer brush surface synthesized by SI-ARGET ATRP was analyzed using XPS as well as ToF-SIMS as described inSec. II A.

Additionally, the thickness of the polymer brushes synthesized by SI-ARGET ATRP was determined by profilometry on a Bruker DektakXT Stylus Profiling system with a 2 μm stylus. Scratches were created on the surfaces using a razor blade, and the profile across the scratch was used to determine the polymer brush thickness. All samples were dry at the time of thickness measurement.

III. RESULTS AND DISCUSSION

In most previous work, substrates modified with polymer brushes are limited to materials such as gold, silicon, and metal oxides that support the formation of self-assembled monolayers that present ATRP initiators.^35,36 However, surfaces of many technologically relevant materials, such as glass, plastics, and metals, do not support self-assembled monolayer formation. To overcome this limitation, we used glow discharge RF plasma for deposition of a surface initiator onto glass disc substrates to initiate SI-ARGET ATRP of HEMA. We chose glow discharge RF plasma because it is capable of functionalizing a surface with halogen precursors independent of substrate material.

XPS was used to assess the functionalization of glass coverslips with the surface initiators. The survey scans seen in Fig. 1, and summarized in Table I, provide information on the overall surface chemical composition before and after methanol wash. We observed that the bromine content in plasma polymerized M3BP coatings, 33 at. % Br, was significantly higher than the expected amount of bromine (14 at. % Br) based on the stoichiometry of the compound. Conversely, the chlorine and fluorine contents in plasma polymerized M2CP and E2FP coatings were much lower than expected—4 at. % Cl versus the expected 14 at. % Cl in M2CP and 3 at. % F versus the expected 14 at. % F in E2FP. For the plasma polymerized CEMA coatings, we found that the overall chemical composition matched the expected composition reasonably well.

Fig. 1.

Representative XPS survey scans and corresponding monomer chemical structures (inset) of plasma polymerized (a) M3BP, (b) M2CP, (c) E2FP, and (d) CEMA on glass coverslips. Major elemental peaks are labeled in each survey scan.

Table I.

Summary of surface elemental composition of plasma polymerized M3BP, M2CP, E2FP, and CEMA surface coatings as determined by survey scans obtained from XPS (n = 3).

Plasma coating		Survey					C 1s
		C	O	Br	Cl	F	Peak 1 CH	Peak 2 CX +/ COR X = Br/Cl	Peak 3 CF	Peak 4 COOR
Before methanol wash
M3BP	Observed	59.8 ± 0.6	7.0 ± 0.7	33.2 ± 0.1	—	—	38.6 ± 2.5	56.3 ± 2.6	—	5.1 ± 0.2
	Expected	57.1	28.6	14.3	—	—	25.0	50.0	—	25.0
M2CP	Observed	79.2 ± 1.4	16.6 ± 1.1	—	4.1 ± 0.2	—	72.2 ± 1.3	18.6 ± 1.0	—	9.3 ± 0.7
	Expected	57.1	28.6	—	14.3	—	25.0	50.0	—	25.0
E2FP	Observed	80.5 ± 1.6	16.7 ± 1.7	—	—	2.8 ± 0.2	71.3 ± 2.1	10.8 ± 0.5	9.2 ± 1.3	8.7 ± 1.1
	Expected	62.5	25	—	—	12.5	40.0	20.0	20.0	20.0
CEMA	Observed	78.0 ± 0.6	13.5 ± 0.1	—	8.5 ± 0.6	—	70.4 ± 1.8	22.0 ± 1.7	—	7.6 ± 0.1
	Expected	66.6	22.2	—	11.1	—	50	33.3	—	16.7
After methanol wash
M3BP	Observed	69.8 ± 0.5	19.5 ± 1.4	10.8 ± 1.0	—	—	61.5 ± 2.2	27.6 ± 1.9	—	10.9 ± 0.3
	Expected	57.1	28.6	14.3	—	—	25.0	50.0	—	25.0
M2CP	Observed	74.1 ± 0.4	18.6 ± 0.4	—	7.3 ± 0.6	—	64.5 ± 1.6	24.7 ± 1.3	—	10.8 ± 0.3
	Expected	57.1	28.6	—	14.3	—	25.0	50.0	—	25.0
E2FP	Observed	82.3 ± 0.7	15.2 ± 0.7	—	—	2.5 ± 0.0	70.4 ± 9.8	14.9 ± 8.3	8.8 ± 1.5	5.9 ± 0.1
	Expected	62.5	25	—	—	12.5	40.0	20.0	20.0	20.0
CEMA	Observed	77.0 ± 0.1	13.0 ± 0.3	—	10.0 ± 0.3	—	68.3 ± 0.4	24.9 ± 0.3	—	6.8 ± 0.1
	Expected	66.6	22.2	—	11.1	—	50	33.3	—	16.7

High-resolution C 1s spectra (Fig. 2) provide insight into the chemical bonding environment of carbon. In Table I, we compare the observed contributions of each chemical state of carbon to the expected contributions based on the stoichiometric ratios present in each chemical structure. As seen in Fig. 2, the shape and features of the high-resolution C 1s peak of the M3BP, M2CP, and CEMA plasma polymerized coatings were fitted with three width-constrained peaks corresponding to three different bonding environments: CH_y at 285 eV, C–X (where X is either Br or Cl) and C–OR both at ∼286.5 eV, and COOR at 289 eV.³⁷ Similarly, four width-constrained peaks were used to fit the high-resolution C 1s peak of E2FP plasma polymerized coatings: CH_y at 285.0 eV, C–OR at 286.5 eV, C–F at 287.5 eV, and COOR at 289 eV.³⁷ It should be noted that all XPS spectra were shifted ∼9.3 eV so that the hydrocarbon peak could align at 285.0 eV (charging binding energy correction).

Fig. 2.

Representative high-resolution spectra of the C 1s peak (black envelope) and the corresponding monomer chemical structure (inset) of plasma polymerized (a) M3BP, (b) M2CP, (c) E2FP, and (d) CEMA on glass coverslips. Fitted peaks are labeled to correspond to different bonding environments within each chemical structure.

In the case of plasma polymerized M2CP and E2FP coatings, the observed contribution of the corresponding halogen is less than expected. On the other hand, the amount of Br seen on plasma polymerized M3BP coatings is significantly higher (∼230%) than what is stoichiometrically expected. This confounding result is the outcome of a complex reaction and can be attributed to radical chemical reactivity, concentration of species in the gas phase, transport to the surface from the gas phase and the probabilities for surface reaction. However, it is outside the scope of this paper to develop a complete mass transport-reaction model to explain the chemical observations. Furthermore, from the high-resolution C 1s spectra of plasma polymerized M3BP, we observe nearly 56% of the total signal from C–Br and C–OR, which suggests that the surface of plasma polymerized M3BP coatings is rich in CBr_x species.

XPS was also used to evaluate whether the plasma polymerized coatings delaminate or change after soaking in methanol (Table I). Importantly, after the methanol wash, no elemental peaks associated with glass discs (specifically silicon) were observed. This indicates that the plasma polymerized M3BP, M2CP, E2FP, and CEMA coatings are stable and do not delaminate. Furthermore, the chlorine content on plasma polymerized M2CP and CEMA coatings increased slightly from 4.1% ± 0.2% to 7.3% ± 0.6% and 8.5% ± 0.6% to 10.0% ± 0.3%, respectively, while the fluorine content on plasma polymerized E2FP coatings did not change significantly. These increases in the Cl signal suggest that there may be a gradient in Cl concentration through the depth of the M2CP and CEMA plasma coatings and that an overlayer of loosely bound, deposited material was removed by the solvent revealing an underlayer of the polymer network that has a higher concentration of Cl.

Interestingly, for plasma polymerized M3BP coatings, the bromine content significantly reduced from 33.2 ± 0.1% to 10.8 ± 1.0% after the methanol wash. This significant reduction in Br after the methanol wash warranted further investigation. ToF-SIMS was used to investigate the surface of plasma polymerized M3BP samples before soaking and after soaking in methanol. In addition, we also analyzed the methanol extract and compared these to M3BP that was drop coated onto silicon wafers and air-dried. From the ToF-SIMS data, we observed that the plasma polymerized M3BP samples contained Br_x (x = 1–4) species whereas the drop-coated M3BP samples only contained Br and a small Br₂ signal. It should be noted that the exact structure of the bromine on the surface could not be determined from the Tof-SIMS data; however, the presence of higher order Br_x species suggests that polybromine compounds were present. Since the XPS composition data for the M3BP sample showed a significant amount of carbon and bromine with much lower amounts of oxygen, we also looked for the presence of Br_xC_y peaks. Polybromide species though rare have been studied theoretically^38,39 and demonstrated experimentally.⁴⁰ It was noted that a series of peaks of the general formula of BrC_y⁻ (y = 1–12) were seen in the M3BP spectra along with Br₂C⁻ and Br₂C₃⁻. All of these were minor peaks in the spectra, suggesting that they do not easily ionize, they are not stable ions, or they are present in low levels at the surface. Since bromine has a unique isotopic pattern, these peaks were confirmed by matching both their exact mass and isotopic pattern. No other Br_xC_y⁻ peaks were found. In comparison, only BrC⁻, BrC₂⁻, and BrC₄⁻ were clearly seen in the drop-coated M3BP sample (see Table S1 in the supplementary material⁴⁴). The presence of Br_xC_y peaks with y > 4 could be indicative of a highly crosslinked structure within the plasma-deposited film. To further investigate this, we looked at the relative intensity of hydrocarbon peaks between the plasma polymerized and drop coated M3BP samples. It was noted that the relative intensity of C_xH_y⁺ peaks with y < x was higher on the plasma polymerized samples while the relative intensity of C_xH_y⁺ peaks with y > x was higher on the controls (see Table S1 in the supplementary material⁴⁴). This further supports the hypothesis that the plasma polymerized sample is heavily crosslinked.

It was noted that the Br₄⁻ intensity was very small when detected. As expected, we also observed that the drop-coated M3BP samples preserved the structure of the monomer better than plasma polymerized M3BP samples. Furthermore, the M3BP samples soaked in methanol showed a significant reduction in all Br_x (x = 1–4) species and bromine signals [Br_x (x = 1–4)] were detected in their methanol wash. These data show that some bromine is lost from the surface after washing with methanol as was seen in the XPS data. It is likely that the majority of the bromine species lost were polybromine compounds as the relative intensity decrease of Br_x (x = 2–4) was greater than that of Br. These ToF-SIMS results can be found summarized in the supplementary material (Fig. S1).⁴⁴ Thus, the results from XPS and ToF-SIMS suggest that the plasma polymerized M3BP coatings are comprised of highly crosslinked CBr_x species as well as small branched polybromine species that have not been tied into the polymer network and are readily washed away during the methanol wash.

After characterization of plasma polymerized haloester surface coatings, their function as surface initiators for ARGET ATRP was assessed. To this end, as a proof of concept, we attempted to grow HEMA polymer brushes from the surface of these coatings and then characterized the surfaces using XPS, profilometry, and ToF-SIMS. Surfaces analyzed included plasma polymerized M3BP, M2CP, E2FP, and CEMA coatings soaked in ARGET ATRP solution for 5 min, 15 min, 30 min, 1 h, or 2 h.

A summary of the surface chemical composition after the ARGET ATRP reaction can be found in Table II. In general, as the reaction time increases we observe a reduction in the amount of halogen on the surface of all plasma polymerized coatings. This result would suggest that the HEMA polymer brush is increasing in grafting density and/or polymer chain length which is why we notice a reduction in the concentration of the halogen on the surface as the ATRP reaction progresses. In all cases, by the 2 h time point, the plasma polymerized coated surfaces only contain carbon and oxygen as seen in the survey scans (Table II). Moreover, the high-resolution C 1s peaks of all plasma polymerized haloester coatings resemble that of HEMA polymer brushes and match the C/O ratio expected for PHEMA (2/1) (data not shown). Particularly, plasma polymerized M3BP coatings are capable of growing HEMA polymer brushes that are at least 10 nm thick as rapidly as 15 min, whereas plasma polymerized M2CP, E2FP, and CEMA coatings begin to grow HEMA polymer brushes that are at least 10 nm thick by 30 min. To the best of our knowledge, this is the fastest reported time for ARGET ATRP reaction of HEMA reported in the literature. However, further experiments would be required to specifically measure kinetics and the reaction rates and compare them with other published studies.

Table II.

Surface elemental compositions by XPS (n = 3) of plasma polymerized M3BP, M2CP, E2FP, and CEMA surface coatings after being in ARGET ATRP solution for 5, 15, 30, 60, and 120 min.

Plasma coating	Reaction time	Survey scan
		C	O	Br	Cl	F
M3BP	0 min	69.8 ± 0.5	19.5 ± 1.4	10.8 ± 1.0	—	—
	5 min	68.6 ± 1.1	27.8 ± 1.7	3.6 ± 2.1	—	—
	15 min	68.7 ± 0.9	31.3 ± 0.9	—	—	—
	30 min	68.9 ± 0.3	31.1 ± 0.3	—	—	—
	1 h	68.5 ± 0.8	31.5 ± 0.8	—	—	—
	2 h	68.3 ± 0.2	31.7 ± 0.2	—	—	—
M2CP	0 min	74.1 ± 0.4	18.6 ± 0.4	—	7.3 ± 0.6	—
	5 min	72.1 ± 0.2	22.3 ± 0.1	—	5.5 ± 0.1	—
	15 min	71.1 ± 0.8	27.1 ± 0.6	—	1.8 ± 0.2	—
	30 min	69.0 ± 0.6	31.0 ± 0.6	—	—	—
	1 h	68.8 ± 0.8	30.7 ± 0.8	—	0.5 ± 0.1	—
	2 h	68.3 ± 0.3	31.7 ± 0.3	—	—	—
E2FP	0 min	82.3 ± 0.7	15.2 ± 0.7	—	—	2.5 ± 0.0
	5 min	79.5 ± 5.5	18.7 ± 5.9	—	—	1.8 ± 0.4
	15 min	72.1 ± 0.3	27.0 ± 0.4	—	—	0.9 ± 0.2
	30 min	69.6 ± 0.5	30.4 ± 0.5	—	—	—
	1 h	69.3 ± 0.4	30.7 ± 0.4	—	—	—
	2 h	70.2 ± 1.6	29.8 ± 1.6	—	—	—
CEMA	0 min	77.0 ± 0.1	13.0 ± 0.3	—	10.0 ± 0.3	—
	5 min	71.9 ± 2.0	24.6 ± 2.5	—	3.5 ± 0.5	—
	15 min	75.5 ± 0.5	19.7 ± 0.5	—	1.8 ± 0.4	—
	30 min	70.6 ± 0.2	29.4 ± 0.2	—	—	—
	1 h	68.7 ± 0.4	31.1 ± 0.2	—	—	—
	2 h	69.5 ± 1.5	30.5 ± 1.5	—	—	—

As a negative control, we used clean glass coverslips and attempted to grow HEMA polymer brushes via SI-ARGET ATRP. XPS data showed that no polymer brush resulted from the SI-ARGET ATRP reaction even after being in solution for 5 h (data not shown). This is not surprising because a surface initiator that contains a halogen is necessary for the ARGET ATRP reaction to initiate and propagate.

Using profilometry, we were able to determine the thickness of the HEMA polymer brushes at different points during the ARGET ATRP reaction (Fig. 3). As expected, the thickness of the HEMA polymer brush increases over time for all plasma polymerized haloester coatings. In all cases, except for E2FP, at the 2 h time point we were able to grow films that were nearly 2 μm thick. Furthermore, we observed that the thickness of the HEMA polymer brushes on plasma polymerized E2FP surface initiators increases at a slower rate compared to M3BP and M2CP. It should also be noted that, as reaction time increases the surface of the HEMA polymer brushes grown on plasma polymerized surfaces changes from smooth to rough (data not shown). This resulted in the large standard deviation seen at longer reaction times.

Fig. 3.

HEMA polymer brush coating thickness grown via SI-ARGET ATRP on plasma polymerized M3BP, M2CP, E2FP, and CEMA surface coatings. All samples were dry during thickness measurements (n = 4).

ToF-SIMS was used to determine the presence of PHEMA on the plasma polymerized haloester surfaces following the ATRP reaction for 2 h. PHEMA produces a number of characteristic secondary ion fragments including the HEMA-OH monomer fragment at 113.06 m/z and the methacrylate fragment C₄H₅O⁺ at 69.04 m/z.⁴¹ A peak list for PHEMA was generated and can be seen in Table S2 in the supplementary material.⁴⁴ Positive ion spectra were then collected for the ATRP reaction surfaces with CuBr₂ (Fig. 4). The characteristic PHEMA positive ions C₂H₃O⁺, C₂H₅O⁺, C₃H₃O⁺, C₄H₅O⁺, C₈H₁₁⁺, C₆H₉O₂⁺, and C₉H₁₃⁺ were detected for all plasma polymerized coatings after 2 h in the ATRP reaction (Fig. 4).

Fig. 4.

Representative ToF-SIMS spectra of HEMA polymer brushes grown on plasma polymerized (c) M3BP, (d) M2CP, (e) E2FP, and (f) CEMA surfaces via SI-ARGET ATRP. A glass disc without a surface initiator is used as a negative control (a), while a PHEMA hydrogel was used as a positive control (b). Positive polarity spectra were acquired in high mass resolution mode with Bi₃⁺ primary ion species over a 0–200 m/z mass range for HEMA. Major peaks associated with HEMA are highlighted in each spectrum, and the molecular formula and normalized intensities are displayed in Table S2 in the supplementary material (Ref. 44).

Since fluorine is a poor leaving group, fluorine-based compounds are not used as surface initiators for ARGET ATRP reaction. Lanzalaco et al. have reported that F-based ATRP systems are less efficient compared to Br- and Cl-based systems.⁴² For this reason, we were surprised to see that we were able to grow HEMA polymer brushes on the surface of plasma polymerized E2FP coatings as determined through profilometry and ToF-SIMS. This unexpected result led us to further investigate whether we are growing HEMA polymer brushes via SI-ARGET ATRP or through another reaction mechanism.

To this end, we placed plasma polymerized M3BP, M2CP, E2FP, and CEMA in the ARGET ATRP reaction mixture in the absence of CuBr₂ for 2 h. As confirmed by ToF-SIMS, no PHEMA peaks were seen on any of these plasma polymerized M3BP, M2CP, E2FP, and CEMA surface coatings (Fig. S2 in the supplementary material⁴⁴); the major peaks detected were mainly hydrocarbon chain fragments at 41.03 (C₃H₅⁺), 43.05 (C₃H₇⁺), and 55.05 m/z (C₄H₇⁺). Since CuBr₂ is crucial as a catalyst for ARGET ATRP, these results suggest that we are in fact growing HEMA polymer brushes via SI-ARGET ATRP.

To investigate the specificity of plasma polymerized E2FP coatings to initiate ARGET ATRP reaction, we also plasma polymerized hexafluoropropylene (C3F6) onto glass coverslips using a protocol previously described by Garrison et al.⁴³ and placed it in the ARGET ATRP reaction with or without CuBr₂ for 2 h. In both cases, ToF-SIMS verified that there were no HEMA polymer brushes on plasma polymerized C3F6 surface coatings with or without CuBr₂. The results from the C3F6 experiments suggest that plasma polymerized E2FP coatings are unique in its ability to grow HEMA polymer brushes compared to other plasma polymerized fluoro surfaces. We hypothesize that the fluorine-carbon bonds in plasma polymerized E2FP coatings are weaker which allows the fluorine to participate in the ARGET ATRP reaction. The specific reaction mechanism, however, is currently unknown and warrants further investigation. These ToF-SIMS data can also be found summarized in the supplementary material (Fig. S3).⁴⁴

IV. CONCLUSIONS

The ability to control surface chemistry is critical for many applications. Though conceptually this appears simple, it is often difficult to directly modify the surface chemistry of some materials. In this work, we describe a straightforward method for the deposition of haloester monomers using RF glow discharge plasma that enables further surface modifications. Due to the versatility of RF glow discharge plasmas for surface modification, this process is scalable and can be applied to any surface of interest. We have also demonstrated that these surface coatings have high halogen contents and do not delaminate making them good candidates as surface initiators for the ARGET ATRP reaction. As a demonstration of this process, we showed that all plasma polymerized haloester coatings can grow HEMA polymer brushes after 2 h in the ARGET ATRP HEMA solution. Specifically, plasma polymerized M3BP can grow HEMA polymer brushes within 15 min. Furthermore, even plasma polymerized E2FP (a fluoroester) was capable of growing HEMA polymer brushes. We believe that the methods described in this work have the potential to extend the use of HEMA coatings to a broad range of commercially interesting applications in the areas of biomedical devices, diagnostics, as well as cell and tissue engineering. Future studies will assess the efficacy of these haloester surface coatings to grow other nonfouling zwitterionic coatings such as carboxybetaine methacrylate and sulfoxybetaine methacrylate via ARGET ATRP for various biomedical applications.