Enzymatically Degassed Surface-Initiated Atom Transfer Radical Polymerization with Real-Time Monitoring

Article abstract of DOI:10.1021/jacs.8b12072

Polymer brush coatings are frequently prepared by radical polymerization, a notoriously oxygen sensitive process. Glucose oxidase (GOx) can inexpensively enable radical polymerization in solution by enzymatically consuming oxygen as it oxidizes glucose. Here, we report the growth of polymeric brushes using GOx-assisted atom transfer radical polymerization (ATRP) from a surface while open to air. Specifically, we grew a set of biomedically relevant polymer brushes, including poly(oligo(ethylene glycol) methacrylate) (POEGMA), poly(2-dimethylaminoethyl methacrylate) (PDMAEMA), poly(sulfobetaine methacrylate) (PSBMA), and poly(2-(methylsulfinyl)ethyl acrylate (PMSEA). For each of these polymers, we monitored GOx-assisted and GOx-free ATRP reaction kinetics in real time using quartz crystal microbalance (QCM) and verified findings with localized surface plasmon resonance (LSPR). We modeled brush growth kinetics considering bimolecular termination. This model fit our data well (r² > 0.987 for all samples) and shows the addition of GOx increased effective kinetic chain lengths, propagation rates, and reproducibility. We tested the antifouling properties of the polymer brush coatings against human blood plasma and were surprised to find that coatings prepared with GOx repelled more plasma proteins in all cases than their GOx-free counterparts.

Full text of DOI:10.1021/jacs.8b12072

Subscriber access provided by UNIV OF LOUISIANA
Article
Enzymatically-Degassed Surface Initiated ATRP with Real-Time Monitoring
Luis Navarro, Alan Enrique Enciso, Krzysztof Matyjaszewski, and Stefan Zauscher
J. Am. Chem. Soc., Just Accepted Manuscript • Publication Date (Web): 23 Jan 2019
Downloaded from http://pubs.acs.org on January 23, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 37
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Enzymatically-Degassed Surface-Initiated ATRP
with Real-Time Monitoring
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Luis A. Navarro1, Alan E. Enciso2, Krzysztof Matyjaszewski2, Stefan Zauscher1,*
1Department of Mechanical Engineering and Materials Science, Duke University, 101
Science Drive, Durham, North Carolina, 27708, United States
2Department of Chemistry, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, PA,
15213, United States
KEYWORDS: Glucose oxidase, non-fouling, antifouling, QCM, polymer brush, SIP
Abstract
Polymer brush coatings are frequently prepared by radical polymerization, a
notoriously oxygen sensitive process. Glucose oxidase (GOx) can inexpensively enable
radical polymerization in solution by enzymatically consuming oxygen as it oxidizes
glucose. Here, we report the growth of polymeric brushes using GOx-assisted atom
transfer radical polymerization (ATRP) from a surface while open to air. Specifically, we
grew a set of biomedically-relevant polymer brushes, including poly(oligo(ethylene glycol)
ACS Paragon Plus Environment
1

Journal of the American Chemical Society
Page 2 of 37
1
2
3
4
5
6
7
8
9
methacrylate) (POEGMA), poly(2-dimethylaminoethyl methacrylate) (PDMAEMA),
poly(sulfobetaine methacrylate) (PSBMA), and poly(2-(methylsulfinyl)ethyl acrylate
(PMSEA). For each of these polymers, we monitored GOx-assisted and GOx-free ATRP
reaction kinetics in real time using quartz crystal microbalance (QCM) and verified findings
with localized surface plasmon resonance (LSPR). We modeled brush growth kinetics
considering bimolecular termination. This model fit our data well (r2 > 0.987 for all
samples) and shows the addition of GOx increased effective kinetic chain lengths,
propagation rates, and reproducibility. We tested the antifouling properties of the
polymer brush coatings against human blood plasma and were surprised to find that
coatings prepared with GOx repelled more plasma proteins in all cases than their GOx-
free counterparts.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
120
Glucose
Oxidase
100
O2
O2
Polymer
Brush
80
60
40
20
0
CO2
O2
O2
Open
to Air
ATRP
CO2
Br
0
1
2
Time (hr)
(****TOC image to be displayed in the abstract of the main paper****)
Introduction
Polymer brushes are used as surface coatings in many applications to endow an
interface with a variety of useful properties, such as biocompatibility,1 lubrication,2
ACS Paragon Plus Environment
2

Page 3 of 37
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
colloidal stabilization,3 and protein-resistance.4 Fouling resistance is especially important
for both biomedical devices4-5 and marine applications.6 While a variety of methods are
employed to prepare polymer brushes, controlled radical polymerization techniques are
especially attractive because of their high functional group tolerance and compatibility
with organic and aqueous media.7-8 In particular, atom transfer radical polymerization
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(ATRP) is extremely versatile for making a variety of well-defined polymeric architectures.9-
12
However, conventional ATRP is sensitive to oxygen, leading to the development of
a new ATRP technique in which activators are regenerated by electron transfer (ARGET).
In ARGET ATRP, the activating Cu(I) species is formed in situ with reducing agents, such
as tin(II) octoate or L-ascorbic acid (LAscA), which reduce oxygen-insensitive Cu(II) catalyst
complexes to the Cu(I) activators.13-14 In ARGET ATRP, an excess of reducing agent is used
to reduce all initially added Cu(II) as well as any Cu(II) that accumulates following radical
termination. As Cu(I) species scavenge oxygen, ARGET ATRP can be used in non-degassed
solutions and still efficiently produce polymeric brushes.15 However, polymerization
kinetics can vary considerably in non-degassed mixtures,16 leading to highly variable
brush thickness. Thus, the search for new methods to perform surface-initiated ATRP (SI-
ATRP) without degassing continues.15, 17-18
Glucose oxidase (GOx) is an enzyme that oxidizes glucose, consuming oxygen in
the process (see Scheme 1).19-20 Recently, several groups have reported that the reaction
ACS Paragon Plus Environment
3

Journal of the American Chemical Society
Page 4 of 37
1
2
3
4
5
6
between GOx and glucose can be exploited to enzymatically degas a radical
polymerization medium, removing any oxygen that can deactivate growing radicals.17, 21-
7
8
9
23
GOx is attractive because it is an inexpensive enzyme that maintains activity in the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
presence of organic solvents,24-25 and thus enables RAFT polymerizations in 70%
dioxane,21 20% methanol,21 and 94 proof gin.22 Unfortunately, GOx produces hydrogen
peroxide as a side product that can interfere with polymerization by initiating new chains.
This interference is exacerbated in ATRP by a Fenton-like reaction with the copper
activators. To address this, Enciso et al. reported that the addition of sodium pyruvate
consumed the peroxide and allowed GOx-assisted ATRP reactions to achieve high,
predictable molecular weights.17 While these studies demonstrated GOx-assisted
polymerizations in solution phase, enzymatic degassing for surface-initiated
polymerizations (SIP) has yet to be explored. Therefore, we investigated the potential of
GOx to enable SIP reactions that are typically air sensitive.
SIP kinetics are most frequently studied by ellipsometry or AFM.7 As it is impractical
to continuously follow the progress of air-free polymerizations with these techniques,
kinetic studies are normally limited to a few periodic measurements of brush height. The
oxygen compatibility of GOx-assisted ATRP make it uniquely suited for techniques that
can continuously monitor SIP kinetics. One such technique is quartz crystal microbalance
(QCM) monitoring, which has been used to follow SIP kinetics in real-time.26-29 In a QCM
experiment, a thin quartz crystal’s vibrational modes are excited, and its resonant
ACS Paragon Plus Environment
4

Page 5 of 37
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
frequencies are measured. The addition of mass to the crystal surface lowers its resonant
frequency as described by the Sauerbrey30 or Parlak equations.31 When a polymer brush
is grown from the crystal surface, the relationship between frequency and mass can be
exploited to measure brush growth in real time.32-37
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Although QCM offers a remarkable degree of time resolution for SIP kinetics,
previous studies have only applied living polymerization models38-39 even when data
clearly showed severe nonlinearity.40 Furthermore, several studies required a special setup
to feed degassed polymerization mixture into a QCM chamber,38, 40 but still suffered from
variability attributed to trace oxygen.40 As GOx-assisted ATRP does not require air-free
reaction conditions, it dramatically simplifies the continuous monitoring of SIP kinetics
with techniques such as QCM and localized surface plasmon resonance (LSPR). To our
knowledge, LSPR has not yet been used to monitor SIP.
In this paper, we use QCM and LSPR to monitor GOx-assisted ARGET ATRP of
polymer brushes that resist protein adsorption. Specifically, we compare GOx-assisted and
GOx-free polymerizations by measuring brush growth over time for four different
polymers: poly(oligo(ethylene glycol) methyl ether methacrylate) (POEGMA), poly(2-
dimethylaminoethyl methacrylate) (PDMAEMA), poly(sulfobetaine methacrylate)
(PSBMA), and poly(2-(methylsulfinyl)ethyl acrylate (PMSEA). We exploit the high temporal
resolution of QCM and LSPR to study the kinetics of these reactions, and we use models
accounting for bimolecular termination to fit data more accurately than in previous
ACS Paragon Plus Environment
5

Journal of the American Chemical Society
Page 6 of 37
1
2
3
4
5
6
7
8
9
studies. By examining how the polymer brushes dissipate vibrational energy, we draw
inferences about adsorbed GOx during polymerization. Additionally, we test the
antifouling ability of the brush coatings, finding the addition of GOx to improve protein
resistance in all cases. We provide additional characterization by ellipsometry, contact
angle goniometry, FTIR, and XPS to complement our findings. We show that the addition
of GOx makes ARGET ATRP a more robust technique for preparing a variety of polymer
brushes in the presence of air.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
6

Page 7 of 37
Journal of the American Chemical Society
1
2
3
4
Experimental
5
6
7
8
9
Materials. This listing only shows materials used in procedures highlighted in this
document. For a comprehensive list, see Supporting Information. Acetone (99.5%,
Sigma-Aldrich), alumina (basic, 150 mesh, Sigma-Aldrich), L-ascorbic acid (99%, Sigma-
Aldrich), blood plasma (human, lyophilized, Sigma-Aldrich), bovine serum albumin
(Sigma-Aldrich), copper (II) bromide (CuBr2, anhydrous, 99%, Acros), ethanol (EtOH,
absolute, KOPTEC), α-D-glucose (96%, Sigma-Aldrich), 1,1,4,7,10,10-hexamethyl-
triethylenetetramine (HMTETA, 97%, Sigma-Aldrich), hydrogen peroxide (35%, BDH
Chemicals, for piranha), glucose oxidase (GOx, from Aspergillus niger, Sigma-Aldrich),
isopropyl alcohol (IPA, 99.5%, BDH Chemicals), 2-(N-3-sulfopropyl-N,N-dimethyl
ammonium)ethyl methacrylate (SBMA, DMAPS, 97%, Sigma-Aldrich), methanol (MeOH,
99.8%, EMD Millipore Corporation), potassium chloride (99.995 %, Alfa Aesar), potassium
phosphate monobasic (99.8 %, J.T. Baker), sodium bromide (NaBr, 99%, Sigma-Alrdich),
sodium chloride (99%, Macron Chemicals), sodium dodecyl sulfate (SDS, 99%, Sigma-
Aldrich), sodium phosphate dibasic (99%, Sigma-Aldrich), sodium pyruvate (99%, Sigma-
Aldrich), sulfuric acid (93-98%, BDH Chemicals), and tris(2-pyridylmethyl)amine (TPMA,
98%, KOEI Chemical Co.) were purchased and used as received.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Oligo(ethylene glycol) methyl ether methacrylate (OEGMA, Mn = 300 Da, Sigma-
Aldrich) and 2-(dimethylamino)ethyl methacrylate (DMAEMA, 98%, Sigma-Aldrich) were
passed through basic alumina to remove inhibitor and used within three hours prior to
ACS Paragon Plus Environment
7

Journal of the American Chemical Society
Page 8 of 37
1
2
3
4
5
6
7
8
9
any polymerization. PBS-Br was prepared as PBS (phosphate buffered saline), except NaCl
was replaced with NaBr to form copper to bromide salts rather than chloride. The GOx-
POEGMA hybrid was produced by coupling an ATRP initiator to GOx by NHS ester
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
chemistry followed by ATRP. The hybrid’s structure was determined by NMR, the POEGMA
molar mass was measured by GPC, and its enzymatic activity was determined by colorimetric
assay. BIBOED (disulfide ATRP initiator), MSEA (DMSO-mimicking acrylate), GOx-POEGMA
hybrid, and PBS-Br were prepared with full synthetic procedures and spectra available in
Supporting Information. Water used was Milli-Q grade and obtained using a NANOpure
Diamond water purification system (Barnstead, 18.2 MΩcm) and filtered through a 0.22
µm Millipak 40 Gamma Gold (Millipore) filter.
Instrumentation. For details on chemical characterization, ellipsometry, contact angle
goniometry, and x-ray photoelectron spectroscopy, see Supporting Information.
Quartz Crystal Microbalance (QCM). QCM data was measured on a QSense E4 (Biolin
Scientific) with a temperature controller with a steady flow rate maintained by an IPC
multichannel dispenser (model ISM935C, ISMATEC). QCM samples were prepared on
5 MHz gold-coated quartz sensor (QuartzPro). QCM data was acquired using QSoft401
(v2.5.33.748), exported with QTools (v3.1.25.604), and analyzed using a MATLAB (R2018a,
v9.4.0.813654) script.
Localized Surface Plasmon Resonance (LSPR). LSPR data was obtained using an Open
SPR instrument (model REV3.0, Nicoya Lifesciences) using OpenSPR software
ACS Paragon Plus Environment
8

Page 9 of 37
Journal of the American Chemical Society
1
2
3
4
5
6
(v3.10.6655.20567). LSPR data was analyzed using the MATLAB Curve Fitting Toolbox
(v3.5.7, MATLAB R2018a, v9.4.0.813654).
7
8
9
Highlighted Procedures
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Preparation of ATRP-initiator Surfaces. Gold-coated QCM sensors (5 MHz) were
rinsed and sonicated (with 1% SDS, H2O, acetone, and EtOH), cleaned with piranha
solution, cleaned by oxygen plasma, and immersed into a 0.1% m/v solution of BIBOED
initiator in ethanol at room temperature for at least 20 hours. LSPR sensors (50 nm gold
nanoparticles, Nicoya Lifesciences, 9 nm EM decay length) were similarly prepared, but
omitting sonication. Detailed procedures are available in the Supporting Information.
Sensors were checked by ellipsometry and contact angle to ensure consistent and clean
surfaces.
Preparation of polymerization solutions. To offer controls with matched
concentrations, polymerization mixtures were prepared as combinations of a 2x glucose
mixture and 2x monomer mixture (see details below). Polymerization mixtures lacking a
given reagent (i.e., LAscA) had the relevant stock solution(s) replaced with an equivalent
volume of solvent. All solutions were prepared volumetrically using stock solutions of
reagents in PBS-Br unless otherwise noted. Solutions were not degassed and were left
completely open to atmosphere throughout the polymerization, with the exception of
DMAEMA, which was lightly covered with foil to prevent inhalation of the toxic monomer.
ACS Paragon Plus Environment
9

Journal of the American Chemical Society
Page 10 of 37
1
2
3
4
5
6
7
8
9
2x Glucose Mixture. A 30% glucose stock was prepared using α-D-glucose, which
tautomerizes to β-D-glucose in aqueous solution. A 2x solution (50 mL) was prepared by
mixing 26 mL of PBS-Br, 12 mL of 30% glucose (final 2x conc. 72 mg/mL, 400 mM, 2 eq),
11 mL of 10% sodium pyruvate (final 2x conc. 22 mg/mL, 200 mM, 1 eq), and 1 mL of
5.0 kU/mL GOx (final 2x conc. 100 U/mL, 37 mg/mL). GOx-free samples had the GOx
solution replaced with an equivalent volume of PBS-Br. For the GOx-POEGMA sample, the
GOx in the mixture was replaced with an equivalent volume of GOx-POEGMA (final 2x
conc. 162 mg/mL), and it was compared alongside 155 kU/g GOx used as a precursor for
the hybrid.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2x POEGMA Mixture. A 2x mixture (50 mL) was prepared by mixing 25 mL of PBS-Br,
17.6 mL of OEGMA (Mn = 300 Da, final 2x conc. 0.369 g/mL, 1.232 M, 1166 eq), 1.18 mL
of 10 mg/mL CuBr2 (final 2x conc. 0.236 mg/mL, 1.06 mM, 1 eq), 19.8 mg HMTETA (final
2x conc. 0.396 mg/mL, 1.72 mM, 1.6 eq), and L-ascorbic acid (added last, approx. 5 min
prior to polymerization). For 20xAA conditions, 6.2 mL of 30 mg/mL L-ascorbic acid was
used (final 2x conc. 3.76 mg/mL, 21.3 mM, 20 eq). For 0.2xAA conditions, 6.2 mL of
0.3 mg/mL L-ascorbic acid was used (final 2x conc. 37.6 µg/mL, 0.213 mM, 0.2 eq). L-
ascorbic acid was added at the last possible moment (approx. 5 min prior to
polymerization), causing the 20xAA mixtures to turn from medium intensity navy blue to
colorless, while no such color change was observed in 0.2xAA mixtures.
ACS Paragon Plus Environment
10

Page 11 of 37
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
2x PDMAEMA Mixture. A 2x mixture (30 mL) was prepared by mixing 15.7 mL of PBS-
Br, 6.0 mL of DMAEMA (final 2x conc. 200 mg/mL, 1.18 M, 592 eq), 1.34 mL of 10 mg/mL
CuBr2 in MeOH (final 2x conc. 0.446 mg/mL, 2 mM, 1 eq), 34.8 mg TPMA (final 2x conc.
1.16 mg/mL, 4 mM, 2 eq), and L-ascorbic acid (added last, approx. 5 min before
polymerization). For 20xAA conditions, 7.0 mL of 30 mg/mL L-ascorbic acid in MeOH was
used (final 2x conc. 7.0 mg/mL, 40 mM, 20 eq). For 0.2xAA conditions, 7.0 mL of
0.3 mg/mL L-ascorbic acid in MeOH was used (final 2x conc. 70 µg/mL, 0.4 mM, 0.2 eq).
CuBr2 and TPMA were mixed for at least 10 minutes (yielding a medium intensity Kelly
green soln.). The color of the copper complex did not change color upon mixing with
DMAEMA, but turned cyan upon mixing with 2x glucose mixture (containing PBS-Br).
2x PSBMA Mixture. This 2x mixture was prepared as with the 2x POEGMA mixture
shown above with the following change. The OEGMA was replaced with 40% SBMA stock
in PBS-Br for a final 2x SBMA concentration of 0.615 M (583 eq, for 0.5x monomer conc.)
or 1.23 M (1166 eq, for 1x monomer conc.).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2x PMSEA Mixture. This 2x mixture of PMSEA was prepared like the 2x PDMAEMA
mixture shown above with two changes. First, the MSEA mixture replaced all MeOH with
PBS-Br. Second, DMAEMA was replaced with 42% MSEA stock in PBS-Br for a final 2x
MSEA concentration of 924 mM (463 eq, for 0.5x monomer conc.) or 1.85 M (925 eq, for
1x monomer conc.).
ACS Paragon Plus Environment
11

Journal of the American Chemical Society
Page 12 of 37
1
2
3
4
5
6
7
8
9
General QCM Experimental Design. Initiator-coated QCM sensors were rinsed with
ethanol, blown dry with nitrogen, and loaded into a QCM chamber. The flow cells were
assembled following manufacturer instructions, using PTFE tape to secure tubes, and
tapping each chamber ten times on each side with a metal spatula to relieve stress on the
crystal. A peristaltic pump was used to flow solution through the QCM cells at steady rate
(0.1 mL/min), while the temperature of the cells was maintained at 25 oC. The frequencies
of each overtone (n = 1, 3, 5, 7, 9, 11, and 13) of each sensor were found in air, then the
chambers were filled with PBS-Br, and the frequencies of each overtone were found again.
Overtone frequencies in PBS-Br were used for data analysis. Before starting the
experiment, the flow of the peristaltic pump was repeatedly reversed (5 second delay) to
check for the presence of bubbles, which typically show a reversible frequency spike when
the change in flow direction sends a pressure wave through the liquid. For detailed
procedures on preparing polymerization mixtures, see above. The order of solutions for
the polymerizations were as follows: (1) PBS-Br, (2) 1x glucose mixture (contains glucose,
sodium pyruvate, and GOx if applicable), (3) 1x glucose mixture with monomer, (4) full
polymerization mixture (contains glucose, sodium pyruvate, GOx if applicable, monomer,
CuBr2, ligand, and L-ascorbic acid), (1) PBS-Br, (5) 10% plasma (8 mg/mL), (1) PBS-Br, and
(6) H2O (see Figure 1). For solutions where L-ascorbic acid or GOx were used, these
reagents were added roughly 5-10 min prior to flowing a reaction solution containing
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
12

Page 13 of 37
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
them. Plasma solutions were reconstituted from a lyophilized powder in water within
20 min of usage. For details on data analysis, see Supporting Information.
LSPR Experimental Design. LSPR was conducted with a similar experimental design to
QCM experiments. For details, see Supporting Information.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Results and Discussion
a
Scheme 1. Schematic GOx deoxygenation and surface-initiated polymerization of
various monomers from initiator-modified surfaces.
i) Deoxygenation
ii) ATRP
Monomer:
Side Chain:
Monomer:
R
O
O
R
=
OEGMA
DMAEMAb
SBMA
O
O
3.5
O R
O
O
O
N
N
Br
Br
n
O
O
SO3
CuBr2, Ligand
AscA, PBS-Br
S
S
O
S
MSEAc
Au
Au
a
First, GOx consumes oxygen to oxidize glucose, producing an undesired hydrogen
peroxide side product. Second, pyruvate in solution consumes hydrogen peroxide without
the aid of GOx to yield CO2, H2O, and acetate as benign side products.
b DMAEMA used 50% PBS-Br, 40% MeOH, and 10% monomer.
c
MSEA is an acrylate monomer with a slightly different backbone than the other
methacrylate.
ACS Paragon Plus Environment
13

Journal of the American Chemical Society
Page 14 of 37
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. General overview of polymerization experiments. (a) Mass density on a QCM
sensor during a typical experiment for growing and testing a POEGMA brush grown by
GOx-assisted ARGET ATRP. Mass density was calculated by the Sauerbrey equation, in
which increases in fluid density and viscosity produce artificially inflated values when
changing solutions. The table shows reagents in solutions flowed over the surface of an
initiator-coated QCM sensor starting at the annotated points on the graph. For reference,
a 1.06 Da/Å2 corresponds to a 1 Hz drop in frequency. (b) Apparent mass density on an
initiator-coated QCM sensor upon adding glucose, pyruvate, and glucose oxidase (GOx),
followed by a buffer rinse (PBS-Br). Curves are shown for a glucose mixture lacking GOx
(black), containing GOx (red), or containing a GOx-POEGMA hybrid (blue dash). Mass
ACS Paragon Plus Environment
14

Page 15 of 37
Journal of the American Chemical Society
1
2
3
4
5
6
density was calculated using the Parlak equation,31 which corrects for fluid viscosity
effects.
7
8
9
Polymerization Experimental Design
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For this study, we chose to use ATRP for its compatibility with a broad range of
monomers. Specifically, we chose the oxygen-tolerant ARGET ATRP because its oxygen
tolerance can be tuned by varying the amount of reducing agent per copper. L-ascorbic
acid (LAscA) was chosen as a reducing agent because it reduces Cu(II) species to Cu(I)
with a semiquinone-like byproduct that rapidly consumes oxygen.41-42 We compared
reactions with 20 equivalents of LAscA (20-40 mM, 20xAA) which is sufficient to react with
both copper and dissolved oxygen, and 0.2 equivalents (0.2-0.4 mM, 0.2xAA) which can
only reduce a fraction of the copper. As all mixtures were non-degassed and completely
open to air, reactions with 0.2 equiv of LAscA would require some external aid to remove
dissolved oxygen (0.26 mM at STP) that hampers the polymerization by both oxidizing
copper and by terminating propagating chains. In contrast, 10-20 mM of LAscA is a
proven method of growing polymer brushes by ARGET ATRP in non-degassed
polymerization mixtures,15-16 although it may yield variable results.16
We first tested the ability of GOx to enable SIP. Briefly, we started with an initiator-
coated gold surface in a QCM or LSPR flow cell and subsequently flow, in order: a GOx
mixture, a GOx mixture containing monomer, a full GOx-containing polymerization
mixture, buffer, 10% blood plasma, and buffer (see Figure 1a). This procedure distinctly
ACS Paragon Plus Environment
15

Journal of the American Chemical Society
Page 16 of 37
1
2
3
4
5
6
7
8
9
shows three major events: fouling of the clean sensor by GOx, polymerization without
degassing, and the fouling of the completed polymer brush with plasma protein. The
addition of protein and polymer mass can be tracked with these techniques, and the
instantaneous slopes in Figure 1a reveal the evolving polymerization rate. In addition to
QCM and LSPR, polymer brushes were characterized by contact angle, ellipsometry, FTIR-
ATR, and XPS (for POEGMA) to verify brush growth and identity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
GOx reliably fouled clean sensors, as illustrated in Figure 1b (12.9 ± 4.8 Da/Å2 by Parlak,
14.1 ± 4.6 Da/Å2 by Sauerbrey, mean ± SD across 8 QCM sensors). Surprisingly, GOx
fouling did not inhibit the polymerization of any of the polymers we tested (Figure 2).
This is consistent with findings by Divandari et al., who showed PNIPAM brushes would
still grow by ATRP when fouled by hemoglobin.43
ACS Paragon Plus Environment
16

Page 17 of 37
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Sauerbrey-calculated mass densities of various polymer brushes grown from
QCM crystals during ARGET ATRP with 20xAA per Cu (black) or 0.2xAA per Cu (red) in the
absence of GOx (solid) or presence of GOx (dash). GOx-POEGMA-assisted ATRP with
0.2xAA (blue dash ▽) is also shown. At time t = 0, the sensors were equilibrated with
glucose, pyruvate, GOx, and monomer solution (point ③ in Figure 1a). Annotated time
points denote: (Cu) the addition of copper catalyst and reducing agent and (2x) the
doubling of monomer concentration in a polymerization mixture. Plots show SIP for (a)
POEGMA, (b) PDMAEMA, and (c) PSBMA.
ACS Paragon Plus Environment
17

Journal of the American Chemical Society
Page 18 of 37
1
2
3
4
5
6
7
8
9
Table 1. Compilation of QCM and spectroscopic ellipsometry measurements of polymer
brushes grown from QCM crystals. The rate constants shown reflect parameters from
fitting QCM data to a bimolecular termination model (mass = kp[M]/2kt ln(1+ktI0t)) based
on Sauerbrey-calculated mass densities from the 9th overtone.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
[M]
Plasma
(mM Polyme Foulin kp[M]/2 kp [M] I0
LAscA
/Cu
Ellipsometr
yb (nm)
Polymer
GOx
)
ra
g
kt
(Da/Å2/h
r)
(Da/Å2) (Da/Å2 (Da/Å2)
)
Bare
Bare
-
-
No
Yes
No
No
Yes
No
Yes
Yes
Yes
Hybri
d
-
-
-
-
95.2
67.8
1.4c
-
-
-
-
3.8 ± 0.2
3.2 ± 0.1
6.0 ± 0.6 c
3.6 ± 0.1
8.5 ± 0.8 c
1.2 ± 0.2 c
5.9 ± 0.1 c
5.6 ± 0.4
4.4 ± 0.3d
POEGMA
POEGMA
POEGMA
POEGMA
POEGMA
POEGMA
POEGMA
20
20
20
0.2
0.2
0.2
0.2
620
620
620
620
620
620
620
120.2
65.9
28.8
53.2
26.7
5.9
13.9
15.7
19.6
375.4
33.9
637.9
11.3
336.1
183.4
175.1
19.9
139.4 -1.3 c
24.9
73.0
80.2
82.1
6.6 c
1.0 c
14.2
-
POEGMA
0.2
620
101.1
2.7
18.7
218.5
8.6 ± 0.2
PDMAEMA
PDMAEMA
20
0.2
No
Yes
590
590
310
620
310
620
460
920
460
460
920
460
32.3
34.6
32.5
-
7.5
16.6
5.0
8.2
6.4
16.4
59.0
9.3
7.4
6.0
11.7
47.5
10.2
10.7
7.6
18.0
72.9
25.4
10.7
81.1
4.5
2.3 ± 0.3
4.5 ± 0.1
1.3 ± 0.1d
3.0 ± 0.1
2.4 ± 0.1 d
3.1 ± 0.1
PSBMAe
20
No
39.5
45.8
64.5
50.0
PSBMAe
0.2
Yes
1.6 ± 0.1 d
3.3 ± 0.1
PMSEAe
PMSEAe
20
No
66.3
65.9
1.2 ± 0.1 d
3.5 ± 0.1
0.2
Yes
42.3
55.8
300.5
1.2
10.2
aThis quantity represents the total mass on the QCM crystal following polymerization,
including any GOx which may be on the crystal.
ACS Paragon Plus Environment
18

Page 19 of 37
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
bEllipsometry measurements are taken after fouling with plasma unless otherwise noted,
including any dry thickness which may result from polymer, GOx, and plasma. Values are
shown as mean ± standard deviation resulting from three independent measurements on
the same crystal.
9
cThese quantities correspond to samples fouled with BSA, not plasma.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
dThese ellipsometry measurements were taken on a polymer sample that was not fouled
with BSA nor plasma.
ePSBMA and PMSEA samples had multiple monomer concentration mixtures flow over
the growing brush surface. The different monomer concentrations are listed sequentially
(top to bottom) in the order they were used.
GOx Assistance in ATRP of POEGMA
We first tested polymerizations with POEGMA because it is well controlled in ATRP44
and it is an antifouling polymer (Figure 2a). ATRP with 20xAA was mildly accelerated by
GOx even though GOx was not needed to fully remove oxygen. With only 0.2 equiv of
LAscA, GOx-assistance was necessary for the reaction to proceed at a reasonable rate.
Based on the higher observed rates of brush growth, we concluded that GOx provides
beneficial degassing of the reaction mixture and that GOx fouling does not significantly
impair polymerization. This holds true regardless of the amount of reducing agent added.
Furthermore, both the shapes of the growth curves and the values of fit rate constants were more
consistent across duplicate experiments for polymerizations using GOx with 0.2xAA than for
GOx-free with 20xAA. From this, we conclude that GOx activity leads to more reproducible open-
to-air polymerizations.
We also investigated the effect of using a GOx-POEGMA biohybrid instead of GOx to
perform this polymerization. We expected this modification to reduce GOx fouling to
ACS Paragon Plus Environment
19

Journal of the American Chemical Society
Page 20 of 37
1
2
3
4
5
6
7
8
9
allow us to observe the degassed polymerization without surface-active GOx. However, it
instead caused more mass to adsorb commensurate with its higher molar mass (Figure
1b). Furthermore, using a colorimetric assay, we found the POEGMA modification nearly
extinguished the hybrid’s enzymatic activity (0.2% remaining activity). Despite its greatly
reduced activity, the modified enzyme still produced SIP kinetics similar to that of native
GOx. This implies only a small amount of GOx activity is actually necessary to assist SI-
ATRP. Furthermore, a small mass loss (~4 Da/Å2) is observed in the first two minutes of
the reaction, a feature observed to a lesser degree with the lighter native GOx (Figure 2a).
This implies that GOx-assisted SIP of POEGMA starts with some desorption of GOx. As
adsorbed GOx is still observed immediately before the addition of catalyst (see Figure
S1), we conclude that GOx desorption was induced by the formation of a POEGMA brush
beneath it.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
GOx-Assisted ATRP of Other Polymer Brushes
Following successful GOx-assisted ATRP of OEGMA, we tested this technique with other
monomers. We preserved the (meth)acrylate functionality and focused on two ATRP
conditions using 20xAA (GOx-free) and GOx-assisted 0.2xAA. Because POEGMA growth
caused GOx desorption, we wanted to challenge the reaction with a polymer brush that
could prevent GOx desorption. Thus, we tested PDMAEMA, a polycation known for fouling
BSA.45 GOx-assisted 0.2xAA ATRP produced a thicker PDMAEMA brush than unassisted
20xAA ATRP (Figure 2b). Furthermore, GOx assistance also leads to a more even
ACS Paragon Plus Environment
20

Page 21 of 37
Journal of the American Chemical Society
1
2
3
4
5
6
polymerization rate of DMAEMA throughout brush growth, showing that GOx is beneficial
even with fouling polymers.
7
8
9
To further explore the variety of polymers amenable with GOx, we similarly produced
PSBMA (poly(sulfobetaine methacrylate)), an antifouling polyzwitterion (Figure 2c).
Similar to the case of DMAEMA, GOx led to more even growth rates of SBMA throughout
the reaction. We also tested poly(2-(methylsulfinyl)ethyl acrylate) (PMSEA), a DMSO-
mimicking acrylic polymer with promise in biomedical applications.46 However, this
polymerization suffered from monomer-specific challenges even for GOx-free ATRP.
Polymerization Kinetics
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
QCM and LSPR offer a unique way to continuously follow the progress of surface-
initiated polymerizations. These real-time monitoring techniques excel at providing much
more detailed information on polymerization by revealing subtle changes in
polymerization rate caused by phenomena such as termination. While this is not the first
study to show SIP in real time, it is the first study to take full advantage of the high
temporal resolution offered by these methods to examine non-linear growth kinetics
during polymerization. Furthermore, their high resolution make it possible to accurately
distinguish polymerization from simultaneously occurring GOx desorption. Our reactions
conditions are appropriate for testing the limits of QCM and LSPR for monitoring SIP.
QCM data show that all brush growth curves were concave down (i.e., decreasing slope)
regardless of the presence of GOx. Since fresh reagent mixture was flowed over the surface
throughout the reaction, this observation indicates a gradual slowdown of polymerization caused
ACS Paragon Plus Environment
21

Journal of the American Chemical Society
Page 22 of 37
1
2
3
4
by termination (Figure 2). Thus, we fit the kinetics of these reactions following a model that
5
6
7
accounts for bimolecular termination,
8
9
푋푛(푡) = 푘 [푀]ln 1 + 2푘푡2퐼0 푡
푝
(
)
,
2푘푡2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
where Xn is the number average degree of polymerization within the polymer brush
(proportional to areal mass density), kp is the propagation rate, [M] is the monomer
concentration, kt2 is the bimolecular termination rate constant, and I0 is the starting active
initiator area density. We compared this model to unimolecular termination by a constant
concentration of dissolved oxygen,
푋푛(푡) = 푘 [푀]퐼 (1 ― e ― 푘 푡
푝
0
′
푡1
,
)
푘′푡1
where kt1’ is the effective unimolecular termination rate constant times oxygen
concentration. For full derivations for both models, see Supporting Information:
Derivations. For representative fits, see Figure 3. For all fits, see Figures S19, S26, S30,
and S33.
LSPR, 0.2xAA+GOx
Raw Data
Fit
QCM, 20xAA
Raw Data
Fit
QCM, 0.2xAA+GOx
Raw Data
Fit
100
80
60
40
20
0
(a)
(b)
(c)
50
40
30
20
10
0
0.6
0.5
0.4
0.3
0.2
0.1
0
0
2000
4000
6000
0
2000
Elapsed Time (s)
4000
6000
0
2000
4000
6000
Elapsed Time (s)
Elapsed Time (s)
ACS Paragon Plus Environment
22

Page 23 of 37
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Figure 3. Kinetic fits of representative polymerizations of POEGMA with raw data (black)
and fits (red, dash). (a,b) show Sauerbrey-calculated mass densities from QCM data fit to
a bimolecular termination model (m = A ln(1+B t)). (a) ARGET ATRP of POEGMA on QCM
with 20xAA. (b) GOx-assisted ATRP of POEGMA on QCM with 0.2xAA. For fit values, see
Table 1. (c) LSPR data of POEGMA grown by GOx-assisted ARGET ATRP with 0.2xAA fit to
a living polymerization model (Δλ = A (1-exp(-B t)), yielding A = 0.7299 nm and B =
3.762*10-4 s-1. See Supporting Information for derivations.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Both termination models fit QCM data accurately (r2 > 0.97). At longer times, the
difference between the two models is that a brush undergoing bimolecular termination
grows at a continuously slowing rate (as a log), while a brush undergoing unimolecular
termination approaches a plateau (as an exponential). However, drift inherent to QCM
prevents one from distinguishing between these models at those later times. In QCM,
frequency may drift by a few Hz per hour due to thermal drift or changes in polymer
swelling. Therefore, once the polymerization has slowed significantly, it becomes difficult
to discriminate between the two termination mechanisms. For our analysis, we will focus
on bimolecular termination due to more consistent fit results and prior work.47
To test the validity of the bimolecular termination model for our SIP reactions, we
analyzed the parameters obtained from our fits. Specifically, we looked at the effective
kinetic chain lengths, kp [M]/kt2, which are independent of the fraction of dead chains.
First, we observe that duplicate polymerizations of OEGMA have very different kinetic
ACS Paragon Plus Environment
23

Journal of the American Chemical Society
Page 24 of 37
1
2
3
4
5
6
7
8
9
chain lengths in the absence of GOx. In contrast, hybrid-assisted and duplicate GOx-
assisted polymerizations show quantitatively consistent results, i.e., kp[M]/kt2 varies with
18% relative deviation. Furthermore, we grew PSBMA and PMSEA by sequentially flowing
polymerization mixtures of doubled monomer concentration over the same surfaces to
test that the fit-derived effective kinetic chain length is proportional to monomer
concentration. Doubling monomer concentration for PSBMA increased effective kinetic
chain length by 62% (20xAA) and 154% (0.2xAA, GOx-assisted). This result is within error
of the expected increase of 100% ± 50% (S.D. after error propagation). Kinetic results for
MSEA are more consistent with a non SI-ATRP growth mechanism. With the exception of
PMSEA, all the tested polymerizations yielded consistent results under a bimolecular
termination model. Despite this, further study using variable initiator surface density is
necessary to clearly distinguish between unimolecular and bimolecular termination.
To complement QCM results, we repeated GOx-assisted, 0.2xAA ARGET ATRP in an
LSPR (Figure 3c). Where QCM detects a growing polymer brush through acoustic waves,
LSPR gives complementary measurements by probing a polymer brush through its optical
properties. LSPR successfully detected POEGMA growth during GOx-assisted ATRP,
confirming QCM results (see Supporting Information for formulas and fits).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
24

Page 25 of 37
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Evolution of normalized dissipation versus frequency drop during the growth
of (a) POEGMA and (b) PDMAEMA brushes. Conditions are shown for 20xAA (black),
0.2xAA with GOx (red), and 0.2xAA with GOx-POEGMA (blue, O). For reference, a guideline
(black dot) is shown for the slope that corresponds to pure fluid viscosity. We note that
frequency drop (abscissa) is directly proportional to polymer brush mass density.
Brush Properties
Both QCM and LSPR give additional information about polymer brush properties
beyond kinetics and fouling. For example, by fitting LSPR data, we determined a swollen
POEGMA brush to be 5.1% v/v POEGMA, which is consistent with the 4.0% volume fraction
measured by liquid ellipsometry. Furthermore, by analyzing dissipation versus frequency
in QCM data, it is possible to identify changes in energy dissipation per unit mass of
polymer brush during polymerization (Figure 4). In this plot, the growth of a thin,
homogeneous polymer layer would produce a straight line of reproducible slope that is
characteristic of the layer’s dissipation per unit mass. GOx-free POEGMA growth follows this
ACS Paragon Plus Environment
25

Journal of the American Chemical Society
Page 26 of 37
1
2
3
4
5
6
7
8
9
type of linear trend (Figure 4a). In contrast, GOx and GOx-POEGMA assisted ATRP both show
large downward hooks in the dissipation vs. frequency plot at the start of polymerization. The
backward slope in this distinctive feature indicates mass loss at the start of polymerization,
followed by a downward slope that reveals initial stiffening of the adsorbed layer. The mass loss
is associated with GOx desorption from the surface. As polymerization progresses, GOx and GOx-
POEGMA produce curves with an upward concavity (i.e., increasing slope), which indicates a
gradual increase in dissipation/mass throughout the growth process. In the case of PDMAEMA,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
GOx leads to substantially stiffer brushes that dissipate 60% less energy per unit mass
than GOx-free PDMAEMA (Figure 4b). This suggests GOx intimately integrates into the
PDMAEMA, consistent with observations of BSA being entrapped in PDMAEMA.45
Figure 5. Fouling of polymer-coated QCM sensors by plasma. Times indicated by arrows
denote exposure to 10% human blood plasma and rinsing with PBS-Br buffer. The various
lines denote: (black, +, solid) initiator-coated gold without polymer, (black, +, dash)
initiator-coated gold previously coated with GOx, (red, ■) POEGMA brushes, (yellow, ○)
ACS Paragon Plus Environment
26

Page 27 of 37
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
PDMAEMA brushes, (black, ▽) PSBMA brushes, and (red, ○) PMSEA brushes. Line styles
denote surfaces prepared by ATRP with: (solid) 20xAA without GOx, (dash) 0.2xAA with
GOx, and (dot) 0.2xAA with GOx-POEGMA. Mass densities are calculated from the QCM
sensors’ 9th mode using the Sauerbrey equation.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Plasma Fouling
One could envision GOx fouling to produce patches where polymers do not grow,
decreasing the quality of polymer coatings made by GOx-assisted SIP. Thus, we tested the
antifouling capabilities of brushes prepared with and without GOx because patches could
serve as sites for fouling. We first tested fouling by exposing various POEGMA coatings
to 1% BSA solution and measuring any adsorbed mass by QCM (Table 1). The addition of
GOx or high amounts of reducing agent (20xAA) were both effective measures for
producing BSA-resistant polymer coatings. However, the coating prepared without GOx
and minimal reducing agent (GOx-free 0.2xAA) appreciably fouled BSA (6 Da/Å2).
Emboldened by these results, we probed our brush coatings using 10% human blood
plasma, which is a much stronger challenge for antifouling because it contains a myriad
of proteins (Figure 5). Plasma was used at a reduced concentration (10%) to avoid
potential complications with viscosity. As even 0.1% plasma will adsorb to saturation,48
this reduced concentration should minimally impact total adsorption.
ACS Paragon Plus Environment
27

Journal of the American Chemical Society
Page 28 of 37
1
2
3
4
5
6
7
8
9
As a first control, we found that 10% plasma extensively fouled initiator-coated
gold surfaces (95.2 Da/Å2) and GOx-coated gold surfaces to a lesser extent (67.8 Da/Å2).
For the purposes of our comparisons, we kept brush thickness consistent between GOx-
assisted and GOx-free samples of a given polymer. However, maintaining brush thickness
between different polymer types is impractical because of differences in swelling and
chain lengths. Furthermore, polymer coatings were kept thin to minimize “hearing loss”
of the quartz crystals. Hearing loss is a reduction of sensitivity to adsorption on top of a
thick, dissipative film that attenuates acoustic waves.49 Despite their low film thickness, all
polymer coatings repelled a significant amount of blood plasma, and GOx-assisted
brushes all exhibited less fouling than their GOx-free counterparts. Similar to the bare
gold surfaces, we suspect that GOx does not form vulnerable patches, but rather acts as
a blocking agent for plasma, like BSA in ELISA. PSBMA and POEGMA are both known
antifouling coatings, but our POEGMA samples appear far superior, likely due to thicker
brush layers as seen by ellipsometry (Table 1). However, we note that thicker brushes may
reduce apparent adsorption in QCM by both blocking adsorption and reducing sensitivity
through hearing loss.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Conclusions
We demonstrated that GOx-assisted SI-ATRP is an effective and robust route for
synthesizing polymer brushes. GOx rapidly consumes oxygen, which enables the open-
to-air preparation of polymer brush coatings. GOx-assistance also makes non-degassed
ACS Paragon Plus Environment
28

Page 29 of 37
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
ARGET ATRP far more reliable by reducing variability in SIP kinetics. Although GOx fouled
initiating surfaces, it surprisingly did not prevent SI-ATRP of a range of monomers. Quite
the opposite, the presence of GOx benefitted the polymerization of anti-fouling coatings
and improved their fouling resistance. QCM and LSPR were both effective and
complementary methods for monitoring polymer brush growth. However, short EM decay
length limits LSPR to follow brush growth kinetics to the early stages of polymerization.
QCM is effective until the rates of signal drift and brush growth become comparable in
magnitude. Despite the commonly observed nonlinear brush growth kinetics, previous
QCM studies only used living polymerization models. Herein, we exploited the high
resolution of QCM to probe the non-negligible contribution of termination to brush
growth. Furthermore, by examining plots of dissipation versus frequency, we made
qualitative inferences about the stiffness of the polymer brush layers during
polymerization. Lastly, antifouling coatings, most notably POEGMA, successfully reduced
fouling by blood plasma.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Moreover, the oxygen tolerance provided by GOx facilitates monitoring techniques
such as QCM and SPR to follow SIP kinetics in real time. In practice, these techniques can
also be used to prepare sensors with well-defined brush thicknesses by indicating when
to halt the polymerization to achieve a specified brush height. The inexpensive reagents
and oxygen tolerance inherent to GOx-based deoxygenation enables producing
antifouling brush coatings for larger-scale applications, like marine antibiofouling. While
ACS Paragon Plus Environment
29