Vapor Sensing Characteristics of Nanoelectromechanical
Chemical Sensors Functionalized Using Surface-Initiated
Polymerization
Heather C. McCaig
†
,
Ed Myers
‡
,
Nathan S. Lewis
†,*
, and
Michael L. Roukes
‡,*
†
Kavli Nanoscience Institute and Division of Chemistry and Chemical Engineering, California
Institute of Technology, Pasadena, California 91125, United States
‡
Kavli Nanoscience Institute and Condensed Matter Physics, California Institute of Technology,
Pasadena, California 91106, United States
Abstract
Surface-initiated polymerization has been used to grow thick, uniform poly(methyl methacrylate)
films on nanocantilever sensors. Cantilevers with these coatings yielded significantly greater
sensitivity relative to bare devices as well as relative to devices that had been coated with drop-cast
polymer films. The devices with surface-initiated polymer films also demonstrated high selectivity
toward polar analytes. Surface-initiated polymerization can therefore provide a straightforward,
reproducible method for large-scale functionalization of nanosensors.
Graphical Abstract
Keywords
Cantilever sensor; nanocantilever; nanomechanical resonator; chemical vapor sensor; surface-
initiated polymerization; ATRP
*
Corresponding Authors
. (N.S.L.) nslewis@caltech.edu. (M.L.R.) roukes@caltech.edu.
ASSOCIATED CONTENT
Supporting Information
Details of our experimental procedures, tabulated response data, sensor response characteristics, fitting procedures to deduce response
times, partition coefficients for the drop-cast and SI-ATRP-grown PMMA films, and further discussion of the results of sensor
responses to 5000 s exposures of vapors. This material is available free of charge via the Internet at
http://pubs.acs.org
.
The authors declare no competing financial interest.
HHS Public Access
Author manuscript
Nano Lett
. Author manuscript; available in PMC 2017 February 08.
Published in final edited form as:
Nano Lett
. 2014 July 09; 14(7): 3728–3732. doi:10.1021/nl500475b.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Resonant micro- and nanocantilever sensors,
1
,
2
modified with self-assembled monolayers
(SAMs) or polymer films, have been used to detect a variety of biological and chemical
species,
3
–
6
including chemical vapors.
7
–
11
Sorption of a chemical vapor onto the surface of
a cantilever changes factors such as the mass and stiffness of the cantilever, which in turn
induces shifts in the resonant frequency of the structure.
2
The analyte sensitivity increases as
the size of such resonant structures decreases
12
with nanocantilevers demonstrated to detect
mass changes down to the attogram (10
−18
g) scale in ambient conditions
10
and at and below
the zeptogram (10
−21
g) scale in vacuum.
13
–
16
The functionalization of nanocantilevers with polymer films increases the sorption of
chemical vapors onto the sensor, relative to the behavior of bare sensors. Functionalization
also imparts selectivity to the sensor based on the differences between chemical interactions
of various polymer/vapor pairs. In response to chemical vapors under ambient conditions,
the signal-to-noise ratio of the sensors increases as the film thickness increases. Previous
nanocantilever chemical vapor sensor studies have relied on thin, 2–10 nm drop-cast
polymer films,
10
,
11
which, while effective, limit the sensor’s dynamic range in terms of both
the minimum and maximum detectable vapor concentrations. Top-down coating techniques,
such as microcapillary-pipet-assisted drop-casting
17
and inkjet printing,
18
utilize solvent
evaporation to produce solid films. These methods result in films of nonuniform thickness,
resulting in a low yield of well-coated sensors, and a high degree of irreproducibility
between adjacent sensors. Surface-initiated polymerization (SIP) from a variety of
precursors has been widely used to grow polymers directly on the surfaces of devices.
19
The
resulting films are composed of polymer chains with one end tethered to a substrate. When
the interchain distance is small, steric repulsion leads to chain stretching, resulting in a
brushlike conformation. Functionalization of nanocantilevers with SIP-grown films provides
a method to deposit sorptive films. Surface-initiated atom-transfer radical polymerization
(SI-ATRP) is a particular polymer brushgrowth technique that is versatile and easily
implemented with a wide range of functional groups.
20
SI-ATRP has been used to grow
polymer brushes on microcantilevers that have been subsequently used to detect changes in
solvent quality,
21
,
22
pH,
22
and temperature,
22
,
23
as well as to detect the presence of glucose
in liquids
24
and to detect saturated toluene vapor in nitrogen.
25
These microcantilever-based
measurements of changes in gaseous environments were performed with a readout based on
the static deflection of the cantilever device of interest.
We describe herein the use of surface-initiated polymerization to grow thick, sorptive films
on nanocantilever chemical vapor sensors. Specifically, poly(methyl methacrylate) (PMMA)
has been grown directly from the surface of nanocantilevers via SI-ATRP, using a synthetic
method that confines the formation of the polymer to the cantilever surface. The SI-ATRP
PMMA-coated cantilevers were then exposed to a series of seven organic vapors, along with
both bare cantilevers and cantilevers functionalized with a drop-cast PMMA film. In contrast
to using a readout based on the static deflection of the device, dynamic detection based on
the resonance frequency shift of the cantilever was utilized as the sensing signal. The SI-
ATRP PMMA-coated cantilever response to polar analytes was enhanced relative to bare and
drop-cast PMMA-coated cantilevers, while all sensors exhibited mutually similar
magnitudes of responses to nonpolar vapors. The thick polymer films grown by SI-ATRP on
resonant nanocantilever sensors have enabled new studies in which the sensor responses are
McCaig et al.
Page 2
Nano Lett
. Author manuscript; available in PMC 2017 February 08.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
dominated by analyte absorption into polymers. Notably, these films are readily adaptable to
wafer-scale processing.
The properties of surface-functionalized sensors using SI-ATRP were explored by use of
piezoresistive, gold-coated, silicon nitride nanocantilevers
10
,
26
,
27
with a typical fundamental
resonance frequency of 10–12 MHz, quality factors (
Q
’s) of 100–200 in ambient conditions,
and a capture area of 1.5 μm
2
. The cantilever resonance was actuated thermoelastically using
integrated Joule heating elements in conjunction with an AC drive current.
28
The
nanocantilever sensors were controlled with custom, LabView-controlled, electronics
27
that
continuously tracked the resonance frequency of each sensor through the use of parallel and
independent phase-locked loops (PLLs). For surface polymerization, after a thorough
cleaning by a UV/ozone plasma the polymerization initiator bis(2-[2
′
-
bromoisobutyryloxy]ethyl)disulfide (BiBOEDS) (ATRP Solutions) was tethered to a gold
overlayer on each cantilever by self-assembly, involving immersion of the substrate in a 5
mM solution of BiBOEDS in C
2
H
5
OH for 24–36 h. The PMMA polymer brush was then
grown using a room-temperature, water-accelerated reaction
29
that was allowed to proceed
for between 30 min and 30 h. Additional details on the synthetic procedures are provided in
the Supporting Information.
Figure 1 presents ellipsometric measurements of PMMA films grown on flat, gold-coated
substrates. These films displayed an initial linear relationship between the reaction time and
the film thickness with the relationship deviating from linearity at long times due to chain
termination. For reaction times less than 20 h, the standard deviation of the film thickness
for a given reaction time was less than 3.5% of the average film thickness. A maximum film
thickness of ~90 nm was reached after 20 h of film growth. For reaction times of >20 h,
larger scatter was observed in the final film thickness, likely due to a higher rate of polymer
chain termination relative that observed at shorter reaction times. As shown in Figure 2,
scanning electron micrograph (SEM) images of a nanocantilever coated with a 90 nm thick
SI-ATRP PMMA film indicated that the resulting films were smooth with a uniform
thickness across the nanocantilever, which is in contrast to the morphology characteristic of
sensors coated with drop-cast PMMA films.
Nanocantilevers were exposed to analyte vapors using an automated vapor delivery system
controlled by LabView-based software.
30
The analytes (hexane, toluene, heptane, ethyl
acetate, chloroform, tetrahydrofuran, and isopropanol) were delivered at concentrations of
P/P
° = 0.0050−0.080 (where
P
is the partial pressure and
P
° is the saturated vapor pressure
of the analyte at room temperature). Each exposure consisted of 70 s of pure carrier gas, 400
s of analyte vapor exposure, and 630 s of carrier gas to purge the system. For single
concentration experiments, a given run consisted of five exposures to each analyte at
P/P
° =
0.020. To ascertain the linearity of the functionalized nanocantilever response with respect to
analyte concentration, five exposures per concentration, per analyte, were delivered in the
order
P/P
° = 0.030, 0.010, 0.048, 0.0050, 0.080, and 0.020 to minimize potential hysteresis
in the measured linearity profiles. SI-ATRP PMMA-coated cantilevers were also exposed to
polar vapors for longer times, that is, up to as much as 5000 s, to determine both the
equilibrium response and the response time of the sensors. Additional details of the
measurement protocols are provided in the Supporting Information.
McCaig et al.
Page 3
Nano Lett
. Author manuscript; available in PMC 2017 February 08.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
For all vapor exposure experiments, the nanocantilevers were housed in a sealed brass
chamber with an internal volume of 100 mL. One to four sensors were tested in each
experimental run, and all sensors were “broken-in” prior to data collection by multiple
exposures to each analyte. The temperature of the device and chamber were not controlled
directly but were stable at 21 °C to within ±1 °C. Frequency data were corrected for any
baseline drift prior to extraction of the sensor responses. The baseline noise was computed
as the standard deviation of the drift-corrected baseline frequency over a period of 10 s prior
to the sensor response. The signal-to-noise ratio (SNR) was calculated as the average
response divided by three times the baseline noise.
Figure 3 shows data from analyte exposures of 400 s, indicating that cantilevers coated with
the SI-ATRP PMMA film produced larger responses to polar vapors relative both to devices
without coatings as well as compared to devices coated with drop-cast PMMA films.
However, no signal enhancement was observed for nonpolar vapors. Figure 4 presents the
dependence of the sensor response on the vapor concentration for 400 s exposures of vapor
to a cantilever-coated with a PMMA film grown by SI-ATRP. The sensor showed a nearly
linear response to toluene vapor, but the responses to ethyl acetate and to isopropanol
deviated from linearity at high analyte concentrations.
The enhanced sensitivity to polar analytes and the lack of sensitivity enhancement for
nonpolar analytes cannot be readily explained from differences in the respective partition
coefficients of analytes into PMMA films grown by SI-ATRP. The partition coefficient (
K
eq
)
for an analyte/polymer pair is defined as
(1)
where
C
f
is the concentration of the analyte in the polymer film and
C
v
is the concentration
of the analyte in the vapor phase.
31
Hence, the number of molecules absorbed into the
polymer film is not correlated with the magnitude of the response. The relative mass loading
of the polymer film (calculated as the product of the partition coefficient and the molecular
weight of the analyte) also does not correlate with response magnitude. (The Supporting
Information provides
K
eq
values for both bulk PMMA and SI-ATRP PMMA films for all
analytes employed in this study).
The enhanced sensitivity to polar vapors of nanocantilevers that had been coated with
PMMA grown by SI-ATRP also cannot be ascribed to vertical swelling of the polymer brush
in response to the presence of analyte vapors. The largest relative change in thickness of SI-
ATRP PMMA films was observed upon exposure to saturated chloroform vapor.
Progressively smaller responses were observed upon exposure to tetrahydrofuran, ethyl
acetate, isopropanol, toluene, heptane, and hexane vapors, respectively. The differences in
magnitude of the relative thickness changes do not correlate with the observed responses of
the functionalized cantilever sensors. Additionally, the ratio of the relative film swelling to
the
K
eq
for each vapor was an order of magnitude greater for chloroform and
McCaig et al.
Page 4
Nano Lett
. Author manuscript; available in PMC 2017 February 08.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
tetrahydrofuran, two good solvents for PMMA,
32
compared to that of the other analyte
vapors.
The magnitude of the response of SI-ATRP PMMA-coated nanocantilevers correlates with
the dipole moment of the analyte vapors (see Supporting Information). To test the validity of
this correlation, cantilevers coated with PMMA grown by SI-ATRP were exposed to carbon
tetrachloride, which is chemically similar to chloroform, but that has no dipole moment. As
shown in Table 1, a 400 s exposure to chloroform induced a relative frequency shift of 2.19
× 10
−4
, whereas a 400 s exposure to carbon tetrachloride caused a relative frequency shift of
only −4.11 × 10
−5
. This behavior is consistent with expectations in which analytes with
nonzero dipole moments interact more strongly with PMMA and induce changes in the
polymer film that yield increased sensor stiffness that in turn is manifested as large positive
shifts in the resonance frequency of the cantilever. The sensitivity of sorption-based vapor
sensors has been shown to correlate primarily with the fractional vapor pressure of the
analyte, as opposed to the absolute value of the analyte concentration in the gas phase.
33
For
a given concentration (mol/volume) of vapor, analytes with higher vapor pressures (such as
those used in this study) experience a lower thermodynamic driving force to absorb into the
polymer film than analytes with low vapor pressures (e.g., organophosphate nerve agents
and explosives). Therefore, nanocantilevers coated with an appropriate polymer film are
expected to be more sensitive to low vapor pressure analytes compared to higher vapor
pressure analytes.
These positive shifts in nanocantilever resonance frequency can be represented by the
relation
(2)
In eq 2, Δ
f
is the change in frequency,
f
O
is the fundamental resonance frequency,
k
is the
initial stiffness,
δ
k
is the change in stiffness,
m
eff
is the initial effective mass, and
δ
m
eff
is
the change in effective mass.
7
The simple sorption of vapor molecules onto a nanocantilever
will result in an increase in mass. If the sorption-induced mass increase is the sole effect, the
cantilever should therefore experience a decrease in its resonance frequency. For a positive
frequency shift to be observed in response to sorption of an analyte vapor, a concomitant
increase in sorption-induced sensor stiffness must occur, and this effect must dominate the
effects of mass loading. This phenomenon has been observed in microcantilevers used for
gas sensing,
7
as well as for nanocantilevers used to detect chemical vapors
34
and biological
species.
35
–
37
Consistently, the resonance frequency of a microcantilever either increased or
decreased after evaporation of a gold film onto the device, depending on whether the gold
was deposited at the clamped end or at the free end of the cantilever, respectively.
38
For
vapor absorption into glassy PMMA films grown by SI-ATRP, the effects of small molecules
interpenetrating the polymer chains could account for the observed increase in sensor
stiffness.
2
McCaig et al.
Page 5
Nano Lett
. Author manuscript; available in PMC 2017 February 08.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Figure 4 shows the response of nanocantilevers coated with PMMA films grown by SI-
ATRP over a range of partial pressures of toluene, isopropanol, and ethyl acetate. Such
sensors showed a linear response for small toluene partial pressures but showed nonlinear
responses during exposure to the same range of partial pressures of ethyl acetate or
isopropanol. The shapes of the response data can be explained by the relative diffusion rates
of the analytes partitioning into the 90 nm thick PMMA films grown by SI-ATRP. The
nanocantilever sensors were operated at ~30 °C, whereas the glass transition temperature
(
T
g
) of the bulk PMMA (Scientific Polymer Products, Inc.; molecular weight = 35 000) used
for the drop-cast films is 105 °C.
39
At temperatures well below
T
g
, the individual chains of a
polymer are locked into a small set of configurations, rendering the polymer “glassy” and
decreasing the diffusion rate of vapor molecules into the film relative to the diffusion rate
above the same polymer’s
T
g
. Glassy polymers such as PMMA are known to exhibit
diffusion of analytes that does not follow Fick’s Law. Instead, diffusion involves delayed
relaxation of the polymer chains, which can greatly increase the time required for the
absorbed analyte to reach its steady-state concentration.
40
–
43
Specifically, the profiles of the
sensor responses were similar to the behavior observed in dual-mode sorption in which the
following two populations of sorbed analyte molecules are present: those dissolved within
matrix of the polymer chains (described by Henry’s Law) and those residing in holes of free
volume in polymer film (described by a Langmuir expression).
44
,
45
The nanocantilever
responses did not reach steady state during 400 s exposures to ethyl acetate at any
concentration explored. Similar behavior was observed for exposure to isopropanol vapor at
concentrations above
P/P
° = 0.02. The sensors only reached a steady-state response to ethyl
acetate after ~5000 s of exposure.
We therefore have described a method for enhancing the absorption of vapor onto
nanocantilevers sensors by deposition of thick, uniform polymer films via the SI-ATRP
process. The approach circumvents the limitations of top-down functionalization schemes,
such as standard drop-coating techniques, and yields sensors with both improved sensitivity
and enhanced saturation limits. The method also enables facile tailoring of the physical and
chemical properties of the polymer films for specific sensing applications. Advanced
chemical functionalization techniques, such as the surface-initiated polymerization presented
here, will accelerate the adoption of miniaturized, nanocantilever-based vapor detection
platforms for a wide spectrum of applications.
Supplementary Material
Refer to Web version on PubMed Central for supplementary material.
Acknowledgments
We gratefully acknowledge assistance from Derrick Chi in fabrication of the NEMS devices, from Xinchang Zhang
for development of the NEMS control electronics, the Caltech Geology and Planetary Science Analytical Facility
and the Kavli Nanoscience Institute for SEM imaging, and from Bruce S. Brunschwig and the Molecular Materials
Research Center for the use of their ellipsometer. We acknowledge support for this work from DARPA/MTO-MGA
through Grant NBCH1050001, and from the Department of Homeland Security, Centers of Excellence, Agreement
2008-ST-061-ED0002.
McCaig et al.
Page 6
Nano Lett
. Author manuscript; available in PMC 2017 February 08.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
REFERENCES
1. Boisen A, Dohn S, Keller SS, Schmid S, Tenje M. Rep. Prog. Phys. 2011; 74(3):30.
2. Eom K, Park HS, Yoon DS, Kwon T. Phys. Rep. 2011; 503(4–5):115–163.
3. Lavrik NV, Sepaniak MJ, Datskos PG. Rev. Sci. Instrum. 2004; 75(7):2229–2253.
4. Singamaneni S, LeMieux MC, Lang HP, Gerber C, Lam Y, Zauscher S, Datskos PG, Lavrik NV,
Jiang H, Naik RR, Bunning TJ, Tsukruk VV. Adv. Mater. 2008; 20(4):653–680.
5. Zougagh M, Rios A. Analyst. 2009; 134(7):1274–1290. [PubMed: 19562189]
6. Hwang KS, Lee SM, Kim SK, Lee JH, Kim TS. Annu. Rev. Anal. Chem. 2009; 2:77–98.
7. Thundat T, Chen GY, Warmack RJ, Allison DP, Wachter EA. Anal. Chem. 1995; 67(3):519–521.
8. Maute M, Raible S, Prins FE, Kern DP, Ulmer H, Weimar U, Gopel W. Sens. Actuators, B. 1999;
58(1–3):505–511.
9. Pinnaduwage LA, Ji HF, Thundat T. IEEE Sens. J. 2005; 5(4):774–785.
10. Li M, Tang HX, Roukes ML. Nat. Nanotechnol. 2007; 2(2):114–120. [PubMed: 18654230]
11. Li M, Myers EB, Tang HX, Aldridge SJ, McCaig HC, Whiting JJ, Simonson RJ, Lewis NS,
Roukes ML. Nano Lett. 2010; 10(10):3899–3903. [PubMed: 20795729]
12. Ekinci KL, Yang YT, Roukes ML. J. Appl. Phys. 2004; 95(5):2682–2689.
13. Yang YT, Callegari C, Feng XL, Ekinci KL, Roukes ML. Nano Lett. 2006; 6(4):583–586.
[PubMed: 16608248]
14. Chiu HY, Hung P, Postma HWC, Bockrath M. Nano Lett. 2008; 8(12):4342–4346. [PubMed:
19053791]
15. Jensen K, Kim K, Zettl A. Nat. Nanotechnol. 2008; 3(9):533–537. [PubMed: 18772913]
16. Chaste J, Eichler A, Moser J, Ceballos G, Rurali R, Bachtold A. Nat. Nanotechnol. 2012; 7(5):
301–304. [PubMed: 22466856]
17. Wright YJ, Kar AK, Kim YW, Scholz C, George MA. Sens. Actuators, B. 2005; 107(1):242–251.
18. Bietsch A, Zhang JY, Hegner M, Lang HP, Gerber C. Nanotechnology. 2004; 15(8):873–880.
19. Rühe, J. Polymer Brushes. Berlin: Wiley-VCH Verlag GmbH & Co. KGaA; 2005. Polymer
Brushes: On the Way to Tailor-Made Surfaces; p. 1-31.
20. Matyjaszewski K, Xia JH. Chem. Rev. 2001; 101(9):2921–2990. [PubMed: 11749397]
21. Bumbu GG, Wolkenhauer M, Kircher G, Gutmann JS, Berger R. Langmuir. 2007; 23(4):2203–
2207. [PubMed: 17279715]
22. Abu-Lail NI, Kaholek M, LaMattina B, Clark RL, Zauscher S. Sens. Actuators, B. 2006; 114(1):
371–378.
23. Bradley C, Jalili N, Nett SK, Chu LQ, Forch R, Gutmann JS, Berger R. Macromol. Chem. Phys.
2009; 210(16):1339–1345.
24. Chen T, Chang DP, Liu T, Desikan R, Datar R, Thundat T, Berger R, Zauscher S. J. Mater. Chem.
2010; 20(17):3391–3395.
25. Bumbu GG, Kircher G, Wolkenhauer M, Berger R, Gutmann JS. Macromol. Chem. Phys. 2004;
205(13):1713–1720.
26. Yang YT, Ekinci KL, Huang XMH, Schiavone LM, Roukes ML, Zorman CA, Mehregany M.
Appl. Phys. Lett. 2001; 78(2):162–164.
27. Bargatin I, Myers EB, Arlett J, Gudlewski B, Roukes ML. Appl. Phys. Lett. 2005; 86(13):3.
28. Bargatin I, Kozinsky I, Roukes ML. Appl. Phys. Lett. 2007; 90(9):3.
29. Jones DM, Huck WTS. Adv. Mater. 2001; 13(16):1256–1259.
30. Severin EJ, Doleman BJ, Lewis NS. Anal. Chem. 2000; 72(4):658–668. [PubMed: 10701248]
31. Grate JW, Snow A, Ballantine DS, Wohltjen H, Abraham MH, McGill RA, Sasson P. Anal. Chem.
1988; 60(9):869–875.
32. Severin EJ, Lewis NS. Anal. Chem. 2000; 72(9):2008–2015. [PubMed: 10815958]
33. Doleman BJ, Severin EJ, Lewis NS. Proc. Natl. Acad. Sci. U.S.A. 1998; 95(10):5442–5447.
[PubMed: 9576901]
McCaig et al.
Page 7
Nano Lett
. Author manuscript; available in PMC 2017 February 08.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
34. Bargatin I, Myers EB, Aldridge JS, Marcoux C, Brianceau P, Duraffourg L, Colinet E, Hentz S,
Andreucci P, Roukes ML. Nano Lett. 2012; 12(3):1269–1274. [PubMed: 22280452]
35. Gupta AK, Nair PR, Akin D, Ladisch MR, Broyles S, Alam MA, Bashir R. Proc. Natl. Acd. Sci.
U.S.A. 2006; 103(36):13362–13367.
36. Tamayo J, Ramos D, Mertens J, Calleja M. Appl. Phys. Lett. 2006; 89(22):3.
37. Ramos D, Tamayo J, Mertens J, Calleja M, Zaballos A. J. Appl. Phys. 2006; 100(10):3.
38. Lee D, Kim S, Jung N, Thundat T, Jeon S. J. Appl. Phys. 2009; 106(2):024310.
39. Poly(methyl methacrylate). Ontario, NY: Scientific Polymer Products, Inc; 2012 Feb. MSDS No.
037A
40. Berens AR, Hopfenberg HB. J. Membr. Sci. 1982; 10:283–303.
41. Sanopoulou M, Petropoulos JH. Macromolecules. 2001; 34(5):1400–1410.
42. Ferrari M-C, Piccinini E, Giacinti Baschetti M, Doghieri F, Sarti GC. Ind. Eng. Chem. Res. 2008;
47(4):1071–1080.
43. De Angelis MG, Sarti GC. Annu. Rev. Chem. Biomol. Eng. 2011; 2(1):97–120. [PubMed:
22432612]
44. Guo J, Barbari TA. Macromolecules. 2008; 41(20):7762–7764.
45. Subramanian S, Heydweiller JC, Stern SA. J. Polym. Sci., Part B: Polym. Phys. 1989; 27(6):1209–
1220.
46. CRC Handbook of Chemistry and Physics. 82nd. Boca Raton, FL: CRC Press; 2001.
McCaig et al.
Page 8
Nano Lett
. Author manuscript; available in PMC 2017 February 08.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript
Figure 1.
Dependence of the thickness of PMMA films grown by SI-ATRP PMMA on the reaction
time. The film reached a maximum film thickness of ~90 nm after ~20 h of film growth. For
data points representing multiple trials, the error bars indicate the standard deviation of the
thickness of the PMMA films grown for a given reaction time.
McCaig et al.
Page 9
Nano Lett
. Author manuscript; available in PMC 2017 February 08.
Author Manuscript
Author Manuscript
Author Manuscript
Author Manuscript