This document discusses Fourier transform infrared spectroscopy (FTIR) and its application in analyzing polydimethylsiloxane (PDMS). It begins with an introduction to FTIR, describing how it uses Fourier transforms to measure infrared absorption spectra. It then covers the basic concepts of FTIR including Michelson interferometers, Fourier transforms, and advantages over dispersive spectrometers. Applications discussed include chemical reaction analysis and surface functional group analysis. It provides an example using FTIR to analyze the reaction mechanism of photo-defined PDMS.

1. Fourier Transform Infrared Spectroscopy
Introduction and its application in PDMS
Yang He(10400425) Advisor: Prof. S. Sukhishvili
1. Introduction
The analysis of any exchange of diverse signals into a group of its relative frequency elements are
called “Fourier Spectroscopy” [1]. The mathematic method, i.e. Fourier Transforms, can be
applied to this spectroscopy. In fact, several powerful spectroscopies including nuclear magnetic
resonance (NMR), Fourier transform infrared (FTIR) and electron spin resonance (ESR) etc. use
the same mathematical method, which gives us a synchronous analysis of numerous frequency
constituents in a single Fourier transform operation.
FTIR spectrometers are rising in popularity due to the precision to the nano-gram level and short
time necessity. In the conventional continuous wave spectrometer, an electromagnetic radiation
with changing frequency is used. After scanning across a line, the peaks in the spectrum caused
by the characteristic absorption can be used to pinpoint information of the sample.While a single
pulse of radiation comprising a certain range of frequencies is utilized in FTIR. After completing
the Fourier transform on the signal obtained by exposure, the responded frequency can be
derived. In this way, the same type of spectrum can be measured in a much more convenient
manner.
At the beginning of 1960 study in interferometer spectroscopies boomed. Cooley and Tukey
created the fast Fourier Transform algorithm which permitted the Fourier transforms to be
calculated ably on a computer [2]. Then, Connes et al. invited the first near-frared planetary
spectra [3]. Three years after, they produced spectra in great resolution [4]. Later, Jacquinot point
out that a proper apparatus of grating spectrometer can largely enhance the speed of operation
[5]. The difference of grating spectrometer and FTIR spectrometer was compared by Gibbie and
he highlighted the tremendous application of FTIR in the future [6]. Now commercial FTIR
spectroscopies were widely available in research laboratory.
2. Basic concepts
As we know, the entire internal energy of amolecule consists ofthe sumof rotational, vibrational,
and electronic energy. The infrared radiaton has a strong connection with the internal energy of
a molecule which can absorb energy of infrared light, so molecular vibrational energy can bound
to a upper level especially when the frequency of infrared radiation is just coupled with the
molecule’s characteristic vibrational frequency. The spectrum of infrared can leave exhaustive
data of the chemical composition information of the sample. The most fundamental knowledgr
of FTIR technique is the Michelosn interferometer and Fourier transform.
2.1 The Michelosn interferometer

2. Michelson interferometer has concise construction, higher resolving power and easier operation
[1, 7]. There are four components constituting the Michlson interferometer as shown on figure
1a, viz., a. a light source which can emit the infrared light; b. a beamsplitter composed of CsI or
KBr; c. two front surface coated mirrors, one
moving and the other one fixed; d. a detector.
a b
FIGURE 1a the schematic diagram of a Michelson interferometer [8]
1b the schematic diagram of interferogram and spectrum [9]
The whole optical path is illustrated as follow. The incident light of the beamsplitter is released
from the light source. Half of the light is reflected and another half is transmitted. The reflected
light is directed to the fixed mirror while the transmitted one goes to the moving mirror. Both
lights are reflected back towards the beamsplitter and both are independently separated into
two parts: one is sent to the detector; the other one is sent to the light source and is lost.
Therefore the detector receives two beams. As the two beams originate from the same source,
their phase difference is only caused by the difference of the light path which can be changed by
putting the moving mirror in diverse sites. If the light source only emits light with unvarying
frequency, we can get a sinusoidal signal when spanning the moving mirror over a range. The
highest amplitude is in accord with the constructive interference, while the lowest matches the
destructive interference. This sinusoidal signal is known as interferogram namely the intensity
(signal in detector) against optical path difference. The signal obtained by the detector is an
accumulation of all the sinusoidal signals. The interferogram has an eruption-shape peak at the
center and decrease abruptly around the center. As illustrated on figure 1.b, the signal can be
adapted to the frequency spectrum through the inverse Fourier transform. Note that it is
necessitous to record a pertinent background spectrum for every sample. If we do not put the
sample inthe optical path, we can gainthe background spectrum demonstrating the instrument’s
energy profile. Furthermore it can be influenced by a. the innate attribute of the light source, b.

3. the energy of air (especially H2O and CO2) absorption, c. the performance of `beamsplitter and d.
the sensitivity of the detector.
2.2 Fourier transforms and mathematical expressions
It is said that the interferogram is created by the interferometer. The signals of interferogram
records the infrared intensity distribution I(x) along the path difference, namely, the travelling
distance x.The reverse Fourier transform which is a bond to connect the frequency spectrum and
the interferogram. This operation can be written as:
∫ 𝐼( 𝑥) 𝑒+𝑖2𝜋𝜐𝑥
𝑑𝑥 = 𝐹−1
(𝐼( 𝑥)) = 𝑆(𝜈)
+∞
−∞
(1)
Now we can express the wave equation of the incident light on the beamsplitter.
𝐸( 𝑥, 𝜐) 𝑑𝜐 = 𝐸0(𝜐)𝑒 𝑖(𝜔𝑡−2𝜋𝜐𝑥)
𝑑𝜐 (2)
Where E0 is the maximum amplitude, x is the optical path, ν is the frequency of the spectrum and
t is time. After the light proceeds through the beamsplitter, two autonomous beams are created.
Before them reorganizing into one beam, they travel distance x1, x2 respectively. Hence the wave
function ER of the recombined beam is
𝐸 𝑅 ( 𝑥1, 𝑥2, 𝜐) 𝑑𝜐 = 𝑇𝑅𝐸0( 𝜐)(𝑒 𝑖( 𝜔𝑡−2𝜋𝜐 𝑥1)
+ 𝑒 𝑖( 𝜔𝑡−2𝜋𝜐 𝑥2)
)𝑑𝜐 (3)
Where T is the transmission coefficient, R is the reflection coefficient.
According to the definition, intensity is proportional to the square of the conjugate amplitude.
Now we attain the intensity as
𝐼( 𝑥1, 𝑥2, 𝜐) 𝑑𝜐 = 𝐸 𝑅 ( 𝑥1, 𝑥2, 𝜐) 𝐸 𝑅
∗ ( 𝑥1, 𝑥2, 𝜐) 𝑑𝜐 (4)
𝐼( 𝑥1, 𝑥2, 𝜐) 𝑑𝜐 = 2𝐸0
2( 𝜐)| 𝑅𝑇|2(1 + cos2𝜋( 𝑥1− 𝑥2) 𝜐) 𝑑𝜐 (5)
The term (x1-x2) represents the path difference. Because this equation obeys the theorem of
superposition of waves, we can derive the total path difference via integrating the equation (5):
𝐼 𝑅 (𝑥) = 2| 𝑇𝑅|2
∫ 𝐸0
2( 𝜐) 𝑑𝜐 +
∞
0
2| 𝑇𝑅|2
∫ 𝐸0
2( 𝜐)cos(2𝜋𝑥𝜐) 𝑑𝜐
∞
0
(6)
IR(x) signifies the interferogram which is the interference with the path difference x. Carrying out
the cosine reverse Fourier transform as equation (1), the equation (6) can be altered into:
𝐸0
2( 𝜐) = (
1
𝜋| 𝑇𝑅|2) ∫ (𝐼 𝑅(x) −
𝐼 𝑅 (0)
2
)cos(2𝜋𝑥𝜐) 𝑑𝜐
∞
0
(7)
Where IR(0) corresponds with the coherent interference. Hence, (𝐼 𝑅(x) −
𝐼 𝑅 (0)
2
) implies the
wavering of the intensity around the mean value IR(0)/2. It is also seeming that
𝑆(𝜐) ∝ 𝐸0
2( 𝜐) (8)

4. We can finally get the spectrum S(ν). Besides that, various advanced mathematical steps, such as
apodization, phase correction and Fourier self-deconvolution, are developed to make the
spectrum computation more practical too.
2.3 advantages of FTIR
FTIR has a number of commonly documented chief advantages. Those key merits will be
concisely discussed below to show why FTIR spectroscopies can distinguish themselves from
other mediocre dispersive-type spectrometers.
2.3.1 Multiplex (Fellgett) Advantage
Generally, a grating or a prism is manipulated to unfold light into individual frequency in the
traditional spectrometer. The monochromator only permits one unit resolution section of the
spectrum to be measured at a time. There is aslitplaced in front of the detector, thus frequencies
belonging to other regions are impassable. On the other hand, the FTIR senses every minor
resolution element all the time. Suppose N means the number of the spectral components in the
broad band and the detecting time is T, the time of conventional spectrometer scanning through
individual resolution elements is T/N, whereas it is T in FTIR. From this aspect we also can
compare the signal-to-noise ratio of conventional spectroscopy (equation(9)) and FTIR
(equation(10)).
(
𝑆
𝑁
) 𝐹 ∝ √ 𝑇 (9)
(
𝑆
𝑁
)
𝐶
∝ √
𝑇
𝑀
(10)
It is clear that the ratio of FTIR is significantly higher than the ratio of conventional spectroscopy.
2.3.2 Throughput (Jacquinot) Advantage
The infrared beam energy per unit time is very vital to achieve spectrum with high quality. In
order to observe the signals in conventional spectroscopy, the view time of each frequency
interval is prolonged. However, that inevitably lessens the efficacy. In contrast, the
interferometer throughput is determined only by the diameter of the collimated beam emerging
from the source.
2.3.3 Laser reference (Connes) Advantage and resolution
The dispersive-style spectroscopies are subject to the electromechanical adjustability of
movement of gratings and slits and tuning of external instruction. Oppositely, a detectible laser
source is used in FTIR spectrometer. Via the monochromatic light of laser, the moving mirror’s
position and the sample’s position are logged. Because the wavelength of the laser is static and
accurately known, the interferometer information (interferogram) can be viewed in the
magnitude of wavelength. So the Fourier transform of the interferogram belongs to
the wavenumber domain. The spectral resolution is in wavenumbers per cm-1.
2.4 limitations

5. There are two former foremost disadvantages of FTIR spectroscopies which are now fully
overcame. Originally, FTIR spectroscopies do not directly measure the spectra, they measure the
interferograms which are problematic to interpret without performing a Fourier transform to
generate a spectrum. However since the histrionic growth in computing competence, calculating
a Fourier transform is no long a matter. Secondly, because all regions of the spectrum are
observed simultaneously, if noise occurs in one part of the radiation from the infrared source, it
will be multiplied throughout the spectrum. In contrast, the noise would be viewed only in the
domain of the spectrum in which it appears in a dispersive spectrometer.
FTIR instruments have a single light beam, whereas dispersive spectrometers usually have a
double beam. If there is no change in atmospheric conditions during the experiment, this dose
not render a problem. Nevertheless, very sensitive experiment needs an extensive measure
period, variation in infrared absorbing gas concentration can severely affect the results.
Therefore it is necessary to eliminate the CO2 and H2O and use an infrared-transparent gas such
as N2.
In addition, another common spectrometry is Raman spectrometry measuring the relative
frequencies at which a sample scatters radiation. The major advantage of Raman spectroscopy
is that it requires little to no sample preparation while the FTIR method has constraints on sample
thickness, regularity and dilution to avoid saturation.
3. Application
FTIR spectrometers are nearly ubiquitously used in industry and research. Here I will succinctly
introduce the application in several facets, and then I will focus on research of photo-defined
PDMS and its surface modification using the FTIR spectrometers as the tool.
FIGURE 3. PDMS curing reaction [13]
TABLE 1. expected and measured peaks of photo-PDMS [13]
FTIR can be set as GC-IR (gas chromatography-infrared spectrometry) which can filter the
information of gas mix [10]. Rather than recording the spectrum of light transmitted through the
sample, FTIR spectrometer can be used to acquire spectrum of light emitted by the sample [11].
Such emission could be motivated by various processes like luminescence and Raman scattering.

6. In addition to emissionspectra,FTIR can help to get Photocurrent spectra [12].This mode exploits
a typical, absorption FTIR spectrometer.
3.1 Chemical reaction analysis
For qualitative analysis, one of the excellent traits of an infrared spectrum is that the absorption
or the absence of absorption in exactwavelengths can be associatedwith specificmolecular bond
information. Thus, it is viable to analysis a very tiny change of functional group at nano-gram
scale. When one would like to find out the reaction mechanism of the photoresist during
exposure and post-exposure bake, FTIR becomes especially worthwhile.
FIGURE 4. Proposed reaction mechanism of photo-PDMS [16]
FIGURE 5.photoPDMS FTIR spectra showing the reaction between benzophenone,PDMS monomer and crosslinker
taken before (red) and after (blue) UV exposure. [13]
PDMS (polydimethylsiloxane) is one of the most prevalent silicone elastomers used in the
fabrication of lab-on-a-chip microfluidics device. A number of research groups try to develop a
recipe (photo-initiator) to make PDMS photo-definable to avoid the high agent price and the
need for clean room [13-16]. The base prepolymer is composed of roughly sixty repeating units
of -OSi(CH3)2- and is ended with vinyl –CH=CH2 =. The curing agent has a short chain with silicon
hydride –OSiHCH3- units. During curing, PDMS monomer vinyl groups react with the silicon
hydride in the curing agent forming Si-CH2-CH2-Si links (Fig 3). In term of the photo-initiator

7. ( benzophenone), it will undergo a different reaction engendering a benzophenone radical.. All
assumptions of reaction mechanism can be verified by the comparison of the FTIR spectra. Table
1 shows the predictable and measured characteristic frequencies of the functional groups in the
photo-PDMS and Figure 4 shows a possible reaction means.
In order to confirm the proposed reaction mechanism, three FTIR spectras are measured [13].
The first sample contains base monomer and benzophenone. The second mixture contains
crosslinkers and benzophenone. The last one has three-component: base, crosslinker and
benzophenone. As the figure 5 illustrated, the vinyl C=C peak at 1600 cm-1 that relates to the
CH=CH2 reduced in intensity. Another vinyl functionality at 960 cm-1 stayed invariant, indicating
the monomer reaction equation in figure 4. The FTIR also exhibited a fall in the C=O carbonyl
group peak at 1664 cm-1, validating the formation of benzophenone radical as displayed in the
figure 4a. Besides, the Si-H silicon hydride group peaks at 2169 cm-1 also fell, verifying that
hydrosilanes reduce carbonyl groups by serving as hydrogen donors. This implies the crosslinker
reaction manner. Thus the FTIR spectra confirms that, the benzophenone precludes the
conventional polymerization of the base and the crosslinker through hyrosilation of olefins and
instead forms a weakly crosslinked region.
3.2 Surface functional group analysis
The surface of PDMS is inherently hydrophobic. A number of attempts have been made to modify
the surface of PDMS micro-channels in order to enhance hydrophilicity [17, 18]. Conversely, there
are several researches to make PDMS with super-hydrophilic surface [19-21].
FIGURE 6. The schematic of optical path in ATR crystal.
FIGURE 7. FTIR spectra of PDMS samples untreated and treated with UV and plasma. (A) Range of 3100~2750c m-
1, (B) range of 1300–700 cm-1. [18]
FTIR measurements can be used to characterize the surface of the modified PDMS and elucidate
observed differences in hydrophilicy [18]. FTIR spectra reveals additional absorption peaks at
1069 and 1084 cm-1 related with Si-O-Si bonds, and changes in silanol (Si-OH) and methyl (-CH3)
groups, specifying modifications in the structure of polar groups on the surface of PDMS thin
layer by ultraviolet radiation exposure. Extra noticeable surface change can be obtained after
increasing UV exposure time (Figure 7 (B)). Then that samples exposed to UV 120 minutes and
subsequent 60 second treatment of O2 plasma. This process revealed relatively strong peaks at
Si-OH–, Si-O-Si, and CH3 as shown on Figure 7 (A). As the UV exposure time is prolonged more
molecules of methyl group are trapped in the PDMS layer, forming SiOx structure. The change of

8. CH3-Si-O- to –O-Si-O- enriches the polar similarity with water molecules, and enhances the
hydrophilicity of PDMS film, which explains the high hydrophilic property observed on UV and
plasma treated samples. Therefore, the existence of Si-O-Si, and Si-OH bonds improves the
polarity of surface leading the hydrophilicity of PDMS films. A similar experiment uses plasma
and chemical treatment to get stable hydrophilic surface [17]. They use Attenuated total
reflection (ATR)-FTIRto measure the spectrum. ATR is a sampling technique used in combination
with FTIR which facilitates samples to be examined directly in solid or liquid state without further
preparation [9]. When total internal reflection occurs, there exists an evanescent wave covers
beyond the surface of the crystal, thus when a sample is carried in contact with the totally
reflecting surface of the ATR crystal, the evanescent wave will interact with sample. It will be
gradually reduced in spectral regions where the sample absorbs energy from infrared beams. The
infrared light experiences several reflections inside the crystal to increase the interaction with
the sample (figure 6). Henceforth a high quality interferogram can be obtained.
UV/ozone was employed to modify of PDMS surface in another case. Figure 9 shows typical
spectra from native PDMS and treated PDMS. It is comprehensible from Fig. 9 that the native
patterned PDMS and UV/ozone treated PDMS have dissimilar attributes. The irregular extension
of function group ≡Si−O−Si≡ indicated the peaks among 1055–1090cm−1. In addition, there is
a decline of intensity in the UV/ozone treated samples which can be attributed to chain scission.
Also, there is a rise of intensity in area of 825–865 cm−1, 875–920 cm−1 and 3050–3700cm−1. The
first two peaks are consistent with ≡Si−O stretching in ≡Si−OH and the last one matches the –
OH stretching in ≡X−OH (X=C or Si). The experiment also showed low absorption at the
implemented frequency of infrared source. There is no evidence to show (−OH)associatedpeaks.
So the information of FTIR spectra can gives an explanation of the experimental observation.
FIGURE 8. ATR-FTIR spectra of native patterned and UV/ozone treated patterned PDMS. The spectra have been
baseline corrected [19]
FIGURE 9. ATR-FTIR spectra of(a) HEMA grafted PDMS with the CO3-pulsed laser and (b) untreated PDMS. [20]
Another comparable modification was carried out using CO2-pulsed laser [20]. The ATR-FTIR
measurement shows that this method can initiate peroxide groups onto the PDMS surface
without photosensitizer at ambient condition. Figure 9 shows that the characteristic absorption
bands of HEMS emerging at 1712 cm-1 and 3335 cm-1 correspond to the HEMA ester and hydroxyl
groups, respectively. Comparison of this spectrum with the original surface supplies solid proof
for the presence of grated poly(HEMA) onto the PDMS surface. M. T. Khorasani et al also

9. manipulate laser to bring on the surface modification [21]. ATR-FTIR works as the same way to
indicate the change of the surface after laser exposure.
3.3 Comparison with other spectroscopies
X-ray photoelectron spectroscopy (XPS) and electron energy loss spectroscopy (EELS) canalso
give the information of bonds.
The chemical shifts of XPS are the change of binding energy. That change at valence shell
prevent astrong interaction between the inner shellelectrons and the nuclear positivecharge.
So the kinetic energy of photoelectrons can leave the ‘footprint’ of the state of certain atoms.
Moreover, Multiplet splitting and shake-up satellites phenomena also add several clues to
the spectrum which can be helpful to deduce the information of bonds. However charging
effects can be a big challenge when measuring insulating samples such as PDMS. The
preparation of XPS is much more complicated. It needs an overnight vacuum pump of the
chamber. The qualitative analysis takes about 10 minutes and quantitative analysis spend
more than one hour. In our case, the amount of reacted functional groups is extremely small.
So XPS is not a good candidate instrument to give the spectra. In fact a new technique based
on XPS is developed, that is Chemical derivatization X-ray photoelectron spectroscopy. This
method is widely used to show the presence or absence of carbon or silicon based functional
groups.
EELSalso requires a high vacuum system. Energy dispersivex-ray spectroscopy (EDS) alsohas
an ability to show the atoms in different chemical states, but ELLD has more advantages to
detect different type of states of the identical elements than EDS. But the sample needs to
be a very thin filmto allow electrons to transmit. The electron beams can damage the sample.
Therefore it is not suitable for PDMS sample.
As for surface analysis, Auger electron spectroscopy (AES) has high lateral resolution and can
distinguish the compositional variations. Since a very thin PDMS sample can be made, AES
can provide the in-depth compositional evaluation. It seems that the mechanism of PDMS
surface can be clearly illustrated by AES. However, the source of AES device is an electron
gun. The destruction of sample, especially the sample is biology or organic materials, caused
by electron beams during operation will prevent its usage. Thus, it is not suitable to use AES
to characterize PDMS surface. Low energy ion scattering (LEIS) cannot be applied to detect
the surface of PDMS due to the same reason.
Secondary ion mass spectroscopy (SIMS) has the highest elemental detection sensitivity. This
technique physically sputter the atoms of the surface and collect the secondary ions in
detector. But the study about PDMS above is to find the certain functional group. Sputtering
means the impossibility to find them.

10. Reference
[1] Jaggi,Neena, and D R Vij. 2006. “Fourier Transform Infrared Spectroscopy.” Handbook of Applied Solid State Spectroscopy, 411–
50.
[2] Cooley, James W., and John W. Tukey. "An algorithm for the machine calculation of complex Fourier series." Mathematics of
computation 19, no. 90(1965): 297-301.
[3] Connes, Janine,and Pierre Connes. "Near-infrared planetary spectra by Fourierspectroscopy. I. Instruments and results." JOSA 56,
no. 7 (1966): 896-910.
[4] Connes, Janine, Hervé Delouis, Pierre Connes, Guy Guelachvili, Jean-Pierre Maillard, and Guy Michel. "Spectroscopie de Fourier
avec transformation d'un million de points." Nouvelle Revued'OptiqueAppliquee 1 (1970):3-22.
[5] Jacquinot, Pierre. "Interferometryand grating spectroscopy: An introductory survey." Applied optics 8,no. 3 (1969): 497-499.
[6] Gebbie,H. Alistair. "Fouriertransform versus grating spectroscopy." Appliedoptics 8, no. 3 (1969):501-504.
[7] http://en.wikipedia.org/wiki/Fourier_transform_infrared_spectroscopy
[8] http://en.wikipedia.org/wiki/Michelson_interferometer
[9] Doyle, W.M.1992. “Principles andApplications of FourierTransforminfrared(FTIR)Process Analysis.” Process Control Quality.
[10] White, Robert.Chromatography/Fourier transforminfrared spectroscopyand its applications. Vol. 10. CRC Press, 1989.
[11] Gaft,Michael, Renata Reisfeld, and Gérard Panczer. Modern luminescencespectroscopy ofminerals and materials. SpringerScience
& Business Media, p. 263. 2005.
[12] Jef Poortmans, Vladimir Arkhipov. Thin film solar cells: fabrication, characterization and applications. John Wiley and Sons.
p. 189. ISBN 0-470-09126-6. 2006.
[13] Cheng,ChaoMin, Joel T. Dawson,QianpingGuo, PhilipR. Leduc, and Bin Li. 2011. “Maskless Fabrication ofSmall-Scale Structures
through Controlling Phase Interactions.” Applied Physics A: Materials Science and Processing 102 (1): 185–88. doi:10.1007/s00339-010-
6128-8.
[14] Jothimuthu,Preetha, Andrew Carroll, Ali AsgarS Bhagat, Gui Lin, James EMark,and Ian Papautsky. 2009.“Photodefinable PDMS
Thin Films for Microfabrication Applications.” Journal of Micromechanics and Microengineering 19 (4): 045024. doi:10.1088/0960-
1317/19/4/045024.
[15] Sawano,Satoshi,Keisuke Naka, Armin Werber, Hans Zappe, and Satoshi Konishi.2008. “Sealing Methodof Pdms as ElasticMaterial
for Mems.” Proceedings of the IEEE International Conference on Micro Electro Mechanical Systems (MEMS) 2: 419–22.
doi:10.1109/MEMSYS.2008.4443682.
[16] Jothimuthu,Preetha, a. a S Bhagat, and Ian Papautsky. 2008. “PhotoPDMS: PhotodefinablePDMS forRapid Prototyping.” Biennial
University/Government/Industry Microelectronics Symposium - Proceedings,183–86. doi:10.1109/UGIM.2008.54.
[17] Bodas, Dhananjay, and Chantal Khan-Malek. 2006. “Formation of More Stable Hydrophilic Surfaces of PDMS by Plasma and
Chemical Treatments.” Microelectronic Engineering83(4-9SPEC.ISS.): 1277–79. doi:10.1016/j.mee.2006.01.195.
[18] De Menezes Atayde, Cleuson, and Ioshiaki Doi. 2010. “Highly Stable Hydrophilic Surfaces of PDMS Thin Layer Obtained by UV
Radiation and Oxygen Plasma Treatments.” Physica Status Solidi (C) Current Topics in Solid State Physics 7 (2): 189–92.
doi:10.1002/pssc.200982419.
[19] Holgerson, P.,D. S. Sutherland, B. Kasemo, and D. Chakarov. 2005.“Patterningand Modification of PDMS Surface through Laser
MicromachiningofSilicon Masters and Molding.” AppliedPhysics A: Materials Scienceand Processing 81 (1):51–56.
doi:10.1007/s00339-004-3082-3.
[20] Khorasani, M. T., H. Mirzadeh, and P. G.Sammes. 1999.“Laser Surface Modification ofPolymers to Improve Biocompatibility:
HEMA Grafted PDMS, in VitroAssay - III.” Radiation Physics andChemistry 55(5-6): 685–89. doi:10.1016/S0969-
806X(99)00212-1.
[21] Khorasani, M.T., H. Mirzadeh,and P.G. Sammes.1996. “Laser Induced Surface Modification of Polydimethylsiloxane as a Super-
HydrophobicMaterial.” Radiation Physics andChemistry 47(6): 881–88. doi:10.1016/0969-806X(95)00166-U.