Supporting Information

for

Advanced Materials, adma.200701419

© Wiley-VCH 200769451 Weinheim, Germany

Supplementary Information

Polymer-Silicon Flexible Structures for Fast Chemical Vapor Detection

Surface morphology of sensitive coatings:

Figure 1 shows the surface morphology of various sensitive coatings studied here. It is

important to note that all the plasma polymers exhibit a typical granular morphology

markedly different from the spin coated counterpart.

Swelling of pp PMAN under humidity:

500 nm

Z range10 nm

1 µmZ range10 nm

200 nm 200 nm

Figure 1: AFM image showing the surface morphology of (a) Spincoated PMAN film (b) pp PMAN (c) pp PAN and (d) pp PTSA.

1 µmZ range50 nm

Z range10 nm

Z range10 nm

(a) (b)

(c) (d)

Temperature dependence of the response of humidity sensor:

The response of pp PMAN coated cantilever for humidity was tested at different

temperatures. Figure 2 (a) shows the response of the cantilever position at three different

temperatures for 7% RH. The response of the cantilever exhibited a significant

60° C

40° C

20° C

(a)

50µm

Figure 3: (a) Position of the cantilever at various temperatures at 6% RH (b)

schematic representation of the reference cantilever working in conjunction

with the sensor cantilever to compensate the temperature effects.

Temperaturecompensateddeflection

Reference

Sensor

∆d

(b)

Figure 2: Thickness of the pp PMAN coated film at different relative humidities

measured by AFM by scanning along the edge film and substrate

0 10 20 30 40 50 60 70

280

285

290

295

300

305

310

315

320

325

Thic

knes

s (n

m)

Relative Humidity (%)

dependence on the temperature. Establishing a precise thermal coefficient of sensitivity

or maintaining the sensor at a constant temperature would ensure demonstrated

unprecedented sensitivity of the cantilever.

Elastic modulus and adhesion of plasma polymers:

Delamination and rupture of the sensitive coatings is a significant problem encountered in

microcantilever based sensors. AFM imaging was performed on various locations of the

cantilever after a series of extreme (300 µm) deflections to probe the possible

50 µm

Figure 4: AFM surface force measurements (a) Average of 35 AFM force-distancecurves obtained from pp PMAN. (c) 16×16 resolution AFM micromechanicalmapping height image and (d) corresponding young’s modulus mapping.

(a) (b)

(c)

-40

-30

-20

-10

0

10

20

30

0 20 40 60 80

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

Forc

e (n

N)

Piezo Movement (nm)

Def

lect

ion

(nm

)

Extension Retraction

catastrophes in the sensitive coatings. Figure 5 shows the representative AFM image of

the coating on the cantilever and it can be observed that the coating retains its surface

morphology with no observable signs of rupture.

FTIR and XPS studies:

Figure 4 shows the FTIR spectra of the pp PMAN and pp PAN. It can be observed that

both the plasma polymers show the characteristic C≡N (cyano group) stretch mode. The

other important aspect of the spectra is the presence of a strong C=N stretch mode which

indicates the dissociation of a C≡N, resulting in high degree of crosslinking.

Figure 5: Representative image of numerous AFM scans performed on thesensitive coating of the tip after a series of extreme deflections showing no signsof delamination or rupture of the sensitive coating.

1.0µm

Table 1: Atomic compositions of the plasma polymer films obtained from XPS.

Plasma

Polymer

C 1s

(%)

O 1s

(%)

N 1s

(%)

S 2p

(%)

Si 2p

(%)

pp PMAN 71.0 17.9 9.4 0.3 1.5

pp PAN 72.2 14.6 13.0 0.2 -

1650 cm-1

C=Nstretch

2200 cm-1

C≡Nstretch

pp PANpp PMAN

Figure 5: FTIR spectra of (a) pp PMAN and (b) pp PAN

1650 cm-1

C=Nstretch

2200 cm-1

C≡Nstretch

(a) (b)

1000 1500 2000 2500 3000 3500 4000Tr

ansm

ittan

ce (a

.u.)

Wavenumber (cm-1)

1000 1500 2000 2500 3000 3500 4000

Tran

smitt

ance

(a.u

.)

Wave number (cm-1)

Trace detection of Explosive vapors:

Pp PAN coated cantilever exhibited a 60µm deflection under exposure to Naphthalene at

40°C which exhibits a vapor pressure of 0.25 Torr.1 Considering the smallest detectable

deflection to be 0.2 nm the detection limit (thermal noise limited) is extrapolated (by

linear approximation of the sensitivity) to be 1 ppbv.

On the other hand, microcantilevers coated with pp PBN exhibited ~ 300 µm deflection

50 µm

Figure 6: Response of pp benzonitrile coated cantilever to Hydrazine vaporsclose to saturation vapor pressure.

Figure 5: X-ray photoelectron spectra of (a) pp-MAN and (b) pp-AN

(a) (b)

0 200 400 600 800 1000

0

5000

10000

15000

20000

Cou

nts

Binding Energy (eV)0 200 400 600 800 1000

-20000

2000400060008000

1000012000140001600018000

Cou

nts

Binding energy (eV)

(shown in Figure 4) for saturated vapor at room temperature with an estimated detection

limit of 12 ppbv. From these preliminary successful attempts, one can easily envision that

the proper choice of the polymer coatings makes the demonstrated technology to be

highly prospective for the realization of extremely sensitive artificial nose capable of

detecting multitude of odors simultaneously.

Videos:

Video 1: Real time video showing the response of the pp-MAN coated cantilever andreference uncoated cantilever to desiccating nitrogen pulses in a humid ambient (Video 1in the supporting information).

Video 2: The first clip of the real time video shows the response of the pp PMAN coated

cantilever and reference uncoated cantilever to cycling pulses of nitrogen (3 Hz) by a

circular chopper and a the second clip shows the response of the cantilever to nitrogen

pulses attained by a linear chopper. (Video 2 in the supporting information).

Methods

Materials. Monomers used as precursors were purchased from Aldrich (purity 99%+), and

directly used for the plasma polymerization. The silicon microcantilevers (MicroMasch,

USA) were rectangular shaped with the following dimensions: L = 350µm; W= 20µm; T =

0.7 – 1.3µm, as verified by SEM. The spring constant of the cantilevers was determined to

be on the order of 0.01 N/m and the thermal vibrations at room temperature are calculated to

be 0.2 nm. A silicon wafer was simultaneously coated for independent characterization with

ellipsometry. The cantilevers were mounted on the corresponding wafer, and placed in the

PECVD reaction chamber so that only one side of the cantilever was coated. The PECVD

chamber was custom built, and details of the system are published elsewhere.2 Briefly,

Argon (50-200 cm3/min, 99.999%) used as the carrier gas for generating a plasma, flows

into a 10-cm diameter reactor at 0.02-0.5 Torr vacuum through a capacitively coupled radio

frequency (RF, 13.56 MHz) discharge of 20 to 45W power. The substrate was located about

1-3 cm further downstream from the precursor inlet. Precursor flow rates of 0.5 and 1.125

cm3/min were utilized for the coatings. AFM was carried out on a Dimension 3000 (Veeco)

according to the normal procedure adapted in our lab.3 Chemical compositions of the

plasma-polymerized films were identified through Fourier transform infrared (FTIR)

analysis. FTIR was performed on a Perkin-Elmer Spectrum 2000 FT-IR spectrometer in the

transmission mode. A range of 400 to 4000 cm-1 was scanned 128 times with 2 cm-1

resolution and averaged. Surface composition acquired by XPS was performed using a

Surface Science Instruments’ M-Probe Spectrometer equipped with a monochromatic AlKa

X-ray source (energy 1486.6 eV).

Sensor response. A hygrometer (EuTech Instruments, Digi-Sense®) with a resolution of ±

0.1%RH was used to track the RH of the enclosed chamber. The humidification of the

chamber was achieved by controlling the flow of dry nitrogen through a water-filled bubbler

into the chamber while the desiccation was achieved by dry nitrogen flow. Optical

microscopy was used to capture the deflection of the sensor at different humidity values.

For measuring the response of the sensor at different temperatures the cantilevers were

mounted on a Peltier heating element. An AFM (Nanoscope IIIa-Multimode) optical

photodiode detection system was utilized to monitor the response of the sensor for small

changes in humidity. The photodetector was calibrated through the standard method of

fitting curves performed on piranha cleaned <100> silicon. For this experiment, the AFM

scanner was enclosed in a chamber and the humidity in the chamber was altered as

mentioned earlier. For measuring the dynamic response of the sensors the humidity was

cyclically varied is the chamber and the deflection of the cantilever was recorded. The

response of the cantilever was recorded using optical microscope at a rate of 30 frames/sec.

For experiments involving the detection of naphthalene and hydrazine vapors, saturated

vapor of the desired chemical was introduced in to the test chamber along with a gentle

stream of nitrogen.

Mechanical properties and adhesion measurements. AFM force-volume measurements

were performed with a Multimode AFM equipped with a Picoforce module. The

measurements were performed using relatively stiff cantilevers, with a resonant frequency of

261 kHz (NSC 21B, MikroMasch). The tip radius was measured to be 74 nm by scanning

samples with 5 nm standard gold nanoparticles. The tip was scanned before and after the

force measurements were performed to ensure that the tip diameter was constant throughout

the measurements and free from contamination. The spring constant was measured to be

23.7 N/m using the standard tip-on-tip approach. The force measurements were performed

on the Si wafer, which was used as a substrate for the 283 nm thick PMAN coated

cantilever. A 16×16 curve force-volume was acquired at 1 Hz utilizing a 100nm ramp and 1

nm trigger; under these conditions the penetration was around 1nm. The Hertzian

approximation was used to derive the elastic modulus of the film.4

Finite Element Analysis. Theoretical deflections of the cantilever have been estimated

with FEA using structural mechanics module of the COMSOL Multiphysics 3.2. The

mechanical properties of the polymer coatings employed in the simulation include elastic

modulus of 1.6 GPa obtained from the AFM measurements and Poisson’s ratio of 0.3

(typical value for glassy polymers). Linear swelling coefficient (19.8×10-4) defined as the

ratio of change in thickness of the polymer to initial thickness for 1% change in RH was

obtained from the AFM measurements. Von Mises stress (a scalar function of the

components of stress tensor) has been recorded for various values of thickness and

humidity.

[1] L. Thompson, J.-G. Lee, P. Maksymovych, J. Ahner, J. T. Yates, Jr. J. Vac. Sci.

Technol. A 2003, 21, 491.[2] H. Jiang, J. T. Grant, K. Eyink, S. Tullis, J. Enlow, T.J. Bunning Polymer 2005, 46,

8178.

[3] V.V. Tsukruk, Rubber Chem. Techn. 1997, 70, 430.

[4] S. A.Chizhik, Z. Huang, V. V. Gorbunov, N. K. Myshkin, V. V. Tsukruk, Langmuir

1998, 14, 2606.