Surface Modification of the Polyimide Films by Electrical Discharges in Water

Full Paper

798

Surface Modification of the Polyimide Filmsby Electrical Discharges in Water

Camelia Miron,* Ion Sava, Ionut Jepu, Petre Osiceanu,Cristian Petrica Lungu, Liviu Sacarescu, Valeria Harabagiu

Pulsed electrical discharges in distilled water were used for the surface modifications of thearomatic polyimides. Optical emission spectroscopy has shown the formation of excitedspecies in the plasma due to water dissociation and ionization. The reactive species have
induced surface modifications of the polymer films whichwere investigated by FTIR, AFM, XPS, and static contactangle measurements. The surface hydrophobicity of thepolyimide films increased with the plasma treatment timedue to the segregation of the CF3 groups at the polymersurface and the imidization of the polyimides during theplasma discharge.
C. Miron, I. Sava, L. Sacarescu, V. HarabagiuInstitute of Macromolecular Chemistry ‘‘Petru Poni’’, AleeaGrigore Ghica, Voda No 41A, Iasi 700487, RomaniaE-mail: [email protected]. Jepu, C. P. LunguNational Institute for Laser, Plasma and Radiation Physics INFLPR,Atomistilor Street No 409, Magurele Ilfov 077125, RomaniaP. OsiceanuInstitute of Physical Chemistry ‘‘IlieMurgulescu’’ of the RomanianAcademy, Splaiul Independentei 202, Bucharest 060021, Romania

Plasma Process. Polym. 2013, 10, 798–807� 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim wileyonlinelibrary.com

1. Introduction

Polyimides (PIs) are both scientifically and commercially

important due to their excellent electrical, thermal, and

mechanical properties.[1–3] PIs are prepared from inexpen-

sive monomers by several synthetic routes and may serve

as an importantmaterial formicroelectronic devices and as

coating materials on different substrates such as plastics,

glass, metals, silicon.[4–5] The good thermal properties and

low moisture absorption make them suitable for the

microelectronic applications. These requirements are

successfully met by the aromatic PIs. The aromatic PIs

usually encounter processing difficulties due to infusibility

and poor solubility in common organic solvents caused by

strong inter-chain forces or inherent macromolecular

stiffness.[5,6] One of the successful approaches used to

overcome these drawbacks is the introduction of flexible

linkages into the backbone of fully aromatic polymers.[6]

The synthesis of PIs containing hexafluoroisopropylidene

(6F) groups is a promisingway to process easily compounds

having high thermal stability.[6,7]

The wettability of the PI surface is important for the

application in themicroelectronic industry. This property is

governed by the surface energy and the geometric

structure. Several surface treatment techniques are used

to modify the surface wettability, roughness, and cross-

linkingdensity, suchaselectronbeams,plasma,photons,X-

rays, etc. Electrical discharges in vacuum or at atmospheric

pressurewereusedasaneffective tool foryieldingmostlyPI

withmodified surfacewettability.[8–10] The reactive species

formed in the plasma interact with the polymer to attach

functional groups to the surface. The O2 plasma treatment

of the fluorinated PI induced a significant decrease of the

DOI: 10.1002/ppap.201300015

Surface Modification of the Polyimide Films

C—C/C—Hbonding, while theN2 plasma treated fluorinat-

edPIsdeterminedthedefluorinationof thePI surfaceandan

increase of the surface water absorption.[8] The surface

treated with the N2/He/SF6/O2 plasma increased the

surface wettability, while the N2/He/SF6 plasma increased

thecontactangleof thesample surfacedue to the formation

of the C—Fx bonding on the surface of the PI film.[9] The

surface free energy of the PIs depends on the plasma

parameters used in the experiments. However, these

techniques used for the surface functionalization of the

PI film have some drawbacks due to the excess of energy

delivered to the film surface. When the polymer is exposed

to the plasma, the energy delivered by the UV radiation or

particle bombardment may lead to the random chain

scission, molar mass degradation, or crosslinking in

polymers.[11] Therefore, the development of novel techni-

ques is necessary for the generation of selective, mono-sort

functional groupscontainingpolymersurfacenecessary for

further industrial applications.

Among other plasma processes, electrical discharge in

liquids has become one of the most interesting techniques

usedfor the functionalizationof thesurfacepolymers.[11–13]

Plasma formed in liquids allows the formation of reactive

species which effectively may react with the polymer

surface, leading to the attachment of chemical groups to

it.[14–16] The density ofmolecules in liquids is several orders

of magnitude higher than that under the low-pressure

plasma, contributing to a faster and more efficient

functionalization process. Hydrogen, oxygen, and hydroxyl

radicals are the main excited species formed in the liquid

plasmas.Whenwater is subjected toahighelectricfield, the

dissociation of the water molecules occurs before the

plasma breakdown and themechanism is similar to that of

the water electrolysis, leading to the formation of Hþ and

OH�.[17] The local current density increases with the

increase of the applied voltage at the electrodes and

gaseous bubbles are formed due to the Joule heating. At

sufficiently high values of the applied voltage, highly

conductive channels called streamers develop between the

electrodes and the plasma is spontaneously formed in the

interelectrode gap.[18] The electrons in these channels

efficiently excite and ionize water molecules, producing a

high amount of excited species. The energy stored in these

ionized channels is consumed by radiation, shock waves,

and thermal conduction to the surrounding water. Some

authors stated that the water molecules break up near the

plasma/liquid interface forming hydrogen and hydroxyl

radicalsdue to thehighly energizedH2Oþgas that escape from

the plasma which impinge onto the water surface with

kinetic energies higher than 100 eV.[19] UV radiation

generated in the discharge has an additional effect in the

formation of reactive species in the liquid.[20,21] The UV

radiation is absorbed by the water layer surrounding the

plasma, determining a strong hydroxyl radical emission

Plasma Process. Polym. 2013, 10, 798–807� 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

around the plasma channels.[20] The hydroxyl radical is a

very powerful oxidant that has the potential to oxidize

organic compounds.[22,23] The water conductivity mea-

sured in the discharge chamber was also seen to increase

after the plasma breakdown, indicating the presence of

chemically active species in the liquid volume.[17] These

speciesmay interactwith thepolymerfilmimmersed in the

water at a certain distance from the plasma core,

determining the attachment of functional groups on the

polymer surface. Using the electrical discharges in water

techniques some authors found that the content of the

functional groups containing oxygen (such as hydroxyl,

carbonyl, or carboxyl)was foundtoexceedbya factorof two

that of the oxygen glow discharge plasmas.[12] Almost 30%

of all oxygen functional groups attached on the polymer

surface were OH-groups, compared to less than 10% OH

produces in gaseous oxygen plasma.[13] Thus, the discharge

in liquids technique seems tobe suitable for anefficient and

selective surface functionalization of a wide range of

polymers.

The purpose of this study was to investigate the surface

modifications of aromatic PIs induced by the pulsed

electrical discharges initiated in distilled water. Electrical

and optical diagnostic techniques were employed for the

investigation of the plasma characteristics. Therefore,

current–voltage characteristics were acquired in order to

establish the discharge type initiated between tungsten

electrodes. Optical emission spectroscopy was used for the

identificationof reactive species formed inplasmaandtheir

evolution with the discharge time. The static contact angle

(u) of water (doubly distilled) was measured with a

goniometer to investigate the surface wettability. The

modifications produced at the polymer surface by the

plasma were mainly examined by the X-ray photoelectron

spectroscopy (XPS), Fourier transform infrared spectrosco-

py (FTIR), andatomic forcemicroscopy (AFM) techniques. To

our knowledge, this is the first work aiming to investigate

the structural modifications of the aromatic PIs induced by

the water plasma.

2. Experimental Section

2.1. Water Plasma Reactor

The pulsed electrical discharges were generated in a cylindrical

Pyrex vessel between tungsten electrodes immersed in 250ml

distilled water. Ceramic tubes were used as insulator material

(Nilaco) and the protrusion of the electrodes from the insulation

was 1mm,with a flat end. The electrodes of 1.6mmdiameter (99%

purity,Nilaco)wereplaced in the center of thevessel in a rod-to-rod

configuration. The interelectrode gap was set to 0.1mm. The

experimental configuration is shown schematically in the Figure 1.

In all the performed experiments the applied pulsed voltage was

provided by a unipolar power supply having a 30-ms rise time—

5kV maximum voltage pulse. The voltage pulse width was set to

799www.plasma-polymers.org

Figure 1. Experimental configuration of the pulsed electricaldischarges in distilled water.

C O

C C 3

CF3

CF

O

O

C C

O

O

O CH 2 NH 2 H 2 N CH 3 3 CH

Dianhydride Diamine

O

Room Temp. 4h

CF 3 N

H

N

O

n

C

CF 3 O

C

C

O

HO C

C

H

OH

CH 2

CH 3 CH 3

Thermal Treatment

-2H2O

3

2 3

3 3

N

n

CH

CH

CH

O

O CF

C

CF

O

O

N

Scheme 1. Synthesis of the PI 6FDA-MMDA.

C. Miron et al.

800

50ms and the repetition frequency was 7kHz. The details of

the pulsed power construction are described elsewhere.[16]

The current and the voltage waveforms were measured using a

current probe (Tektronix P-6021) and a voltage probe (Tektronix

P-6015) respectively, connected to an oscilloscope (Tektronix

TDS 310).

2.2. Optical Emission Spectroscopy

The light emitted by the reactive species formed in the water

plasma was collected with an optical fiber immersed in water at

5mm above the gap between the electrodes using a PTFE tube

ended with a fused silica window of 2mm width and 15mm

diameter. The optical fiber was connected to a spectrograph

(Spectral Products SM 242) which has an optical resolution of

0.4 nm from 200 to 1 050nm. The experimental conditions were

kept constant during the detection of the spectra, except for the

discharge time, which was set for 2, 4, and 6min. The electrode

erosionand the slightly changeof the interelectrode gapdetermine

the generation of the instabilities inwater plasmawhichwere not

reproducible for the same experimental conditions. The intensity

fluctuations in the acquired emission spectra averaged over 30

samples was of about 12%.

2.3. Materials

Aromatic PIs have been synthesized by solution polycondensation

of certain aromatic dianhydrides (hexafluoroisopropylidendiph-

thalic dianhydride-6FDA) with aromatic diamines (4,40-diamino-

3,30-dimethyl diphenylmethane-MMDA) by a method previously

reported.[6] The PI films were obtained by casting onto glass

plate the solution of the PI in DMAc followed by thermal

treatment in a vacuum oven to remove the solvent (in the case

of PI solution) (Scheme 1). The detailed procedure is described

elsewhere.[6]

The PI film of 40mm thickness and 15�15mm2 size was

mounted on a metal support and immersed in distilled water

(Figure 1). The distance between the interelectrode gap and the

polymer surface was set at 20mm. The film was subjected to the


plasma treatment for 2, 4, and 6min. The samples were

investigated by means of FTIR, AFM, XPS, and static contact angle

measurements to observe the surfacemodifications inducedby the

plasma treatment.

2.4. Fourier Transform Infrared Spectroscopy (FTIR)

Infrared spectra for samples were recorded using the spectrometer

Bruker Vertex 70 FT-IR in reflectance mode (ATR mode ZnSe crystal:

550–8000 cm�1). The spectra of the samples were recorded in the

range of 600–4000 cm�1 in ambient air at room temperature,with a

resolution of 2 cm�1 and accumulation of 32 scans. FTIR was

employed to investigate the functional groups present on the

PI film surface before and after 2, 4, and 6min of water plasma

treatment.

2.5. Atomic Force Microscopy (AFM)

The water plasma PI treated films surfaces were scanned by AFM

(Ntegra Spectra) to observe their surface morphology. A silicone

cantilever NSG10was used and an area of 10�10mm2 squarewas

scanned.The rootmeansquare (rms)of the surface roughness (Rrms)

was calculated from the roughness profile determinedbyAFM. The

PI film surfacemorphologywas analyzed before and after 2, 4, and

6min of exposure to water plasma.

2.6. X-ray Photoelectron Spectroscopy (XPS)

SurfaceanalysisperformedbyXPSwascarriedoutonPHIQuantera

equipment with a base pressure in the analysis chamber of

10�9 Torr. The X-ray source wasmonochromatized Al Ka radiation

DOI: 10.1002/ppap.201300015


(1 486.6 eV) and the overall energy resolution is estimated at

0.65 eV by the fullwidth at half-maximum (FWHM) of theAu4f7/2

photoelectron line (84 eV). Although the charging effect was

minimized by using a dual beam (electrons and Ar ions) as

neutralizer, the spectra were calibrated using the C1s line

(BE¼ 284.8 eV) of the adsorbed hydrocarbon on the sample

surface (C—C or (CH)n bondings). As this spectrum was recorded

at the start and the end of each experiment the energy calibration

during experiments was quite reliable. We have to emphasize

that the errors in our quantitative data are found in the range of

�10%, while the accuracy for BEs assignments is �0.2 eV.

Figure 2. Current and voltage waveforms acquired in the waterplasma.

3. Results and Discussions

The plasma could be generated in distilled water at a

voltage of 5 kV (peak-to-peak voltage), current of 11A. The

repetition rate of the voltage pulses was set to 7 kHz. A

stable discharge between the electrodes could be obtained

at about 2 kV, and the discharge current was of the order of

70mA. The current and voltage waveforms were acquired

inwater discharges (Figure 2). The injected energyper pulse

in the discharge was calculated from the integration of the

measured input power (namely the product of the

measured gap voltage and the measured current) over

the pulse duration, and was in the range of 0.2–1.5 J.

Assuming that the distribution of the electric field was

quasiuniform between the electrodes, we have used the

relation E ¼ U=d to calculate the electric field in the

discharge regime, where U is the discharge voltage

(peak value) and d is the interelectrode gap. The electric

field required for the discharges in water was of about

200 kV � cm�1. The PI filmswere treated in the plasma for 2,

4, and 6min. The energy per pulse increased with the

discharge time from 0.24, 0.45, and 1.2 J for the 2, 4, and

6min of plasma discharge, respectively.

Figure 3. Optical emission spectra of pulsed electrical dischargesin distilled water.

3.1. Optical Emission Spectroscopy

In the electrical discharges in water the hydroxyl radicals,

oxygen and hydrogen excited atoms and molecules were

investigated by optical emission spectroscopy. The experi-

mental conditions were kept constant during the spectra

acquisition (discharge voltage U¼ 1 700V, frequency n¼ 7

kHz, pulse width 50ms). The polymer film was placed at

20mmin front of the interelectrodegap. Theexcited states/

of atomic hydrogen of the Balmer series Ha at l¼ 656.28

nm, Hb at l¼ 486.13nm, and atomic oxygen OI (3p5P! 3

s5S0 at l¼ 777.5 nm)were detected in the emission, as well

as the molecular bands of the hydroxyl radicals OH

(Emission System, A2Sþ!X2P at 306.4 nm (Figure 3).

The weak emission band with the band head at l¼ 400.9

nm might be attributed to the O2 Chamberlains airglow

system (C3Du! a1Dg).[24,25] The Oþ

2 second negative

system(A2Pu!X2Pgatl¼ 240nm)couldbedistinguished


in the spectra. The O3 Ultra-Violet System (1B2! 1A1) at

l¼ 310nm may overlap the molecular bands of the

hydroxyl radicals. In the emission spectra could not be

identified reactive species from the PI structure, such as

carbon or nitrogen, indicating that the PI structure was not

drastically affected during the plasma discharge. The color

of the polymer was not changed and no significant gloss

reduction was observed after 2, 4, or 6min of plasma

treatment.

Figure 4 shows that the hydrogen emission intensity

ratio ðIHa=IHb

Þ increased with the water plasma treatment

time. Under the assumption of a local thermal equilibrium

(LTE),[17,26] the mean electron temperatures calculated

during 6min of discharge was in the range of 0.5� 0.1

and 2.5� 0.3 eV. Therefore, the generation of the reactive


Figure 4. The hydrogen emission intensity ratio’s dependence ondischarge time during 6min of electrical discharge. Figure 5. ATR-FTIR spectra for the untreated and water plasma

treated PI films.

C. Miron et al.

802

species in the plasma increased with the increase of the

electron temperature. The electron temperature dependen-

cy on time played a major role in the polymer surface

modifications after plasma exposure, as will be discussed

later on in this paper.

3.2. Fourier Transform Infrared Spectroscopy (FTIR)

A detailed study of ATR-FTIR spectra of pristine polymer

film along with different water plasma treated polymer

samples were performed within 6min of PI surface

treatment (Figure 5). The spectra were assigned as I for

the untreated plasma samples, II, III, and IV for the 2, 4, and

6min plasma treated samples, respectively. For the

untreated plasma sample, several characteristic peaks

wereattributed to the imide structure, suchas: the carbonyl

group of the imide ring at 1 770–1 780 and 1 710–

1 720 cm�1, and the characteristic band for the C—N

vibration at 1 360–1375 and 720–730 cm�1.[6,27] The

absorption peaks at 2 940 and 2870 cm�1 are attributed

to the methyl and methylene groups. The characteristic

absorption bands of hexafluoroisopropylidene are present

at 1 260 and 1 210 cm�1. For the water plasma treated

samples (II–IV), shift and broadening of the infrared

absorption lines could be observed in the spectra. The

C55O band of the imide group shifted gradually to the

lesser wavenumbers (from 1730 cm�1 for the 2min

plasma treated sample to 1 670 cm�1 for the 6min plasma

treated sample). The shift and broadening of the infrared

absorption bands are caused by the collision and scattering

between phonons.[28] This implies a scission of the

imide group. It was previously shown that the generation

of the reactive species formed in the plasmawas enhanced

with the discharge time (Figure 4). It may be assumed

that highly energetic excited species may reach the


polymer surface by means of bubbles and streamers,

producing the scission of the PI structure into its

component groups, i.e., aromatic dianhydride (hexafluor-

oisopropylidendiphthalic dianhydride-6FDA) and aromatic

diamine (4,40-diamino-3,30-dimethyl diphenylmethane-

MMDA). The increase in the temperature may determine

thebroadeningof spectral lineandadecrease in the relative

intensity. The added thermal energy may gradually

overcome the intermolecular and intramolecular forces,

and less energy is required to excite the vibrational dipoles.

Thus, the spectral peaks will appear at lesser wavelengths

corresponding to an increasing structural material disor-

der.[29] For the 6min plasma treated sample (IV) could be

observed a splitting in the C55O of the imide ring in two

well-defined peaks (1 745 and 1 788 cm�1). The same

behavior was observed by authors in the case of the

(CO)2NC imide group which were separated into two

different peaks due to the increase in the cure temperature.

The electron temperature increased (up to 2.5 eV) with the

increase of the water plasma treatment time. Thus, an

increase in the local surface temperature may induce the

formation of more carbonyl groups (C55O) on the polymer

film surface.

Theabsorptionbandat1 860 cm�1 (CO—O—CO)wasalso

identified inthespectraonly for theplasmatreatedsamples

(II–IV) and indicates the presence of the unreacted

anhydride. The band at 1 540 cm�1 (y amide C—N) is also

present in all the FTIR spectra (I–IV). The fact that only the

1 860 cm�1 band exists in the spectra of the plasma treated

samplesmaysuggest a scissionof thePI structure in its own

components, namely the aromatic dianhydride and aro-

matic diamine. The imidic bond may be broken after the

interactionwith the reactive species formed in plasma and

the components used initially for the PI synthesis are now

present on the film surface.

DOI: 10.1002/ppap.201300015

Table 1. The element relative atomic concentration of the C, O, N,and F (in atom%) of the PI surface and the rms surface roughnessbefore and after the water plasma treatment.

Treatment

time [min] rms [nm] C [%] O [%] N [%] F [%]

– 2 73.7 21.3 5.0 –


The characteristic absorption band of the amide group

are evidenced at 3 170–3 870 cm�1 (COOH and NH2

vibrations) only in the case of the plasma treated samples

(II–IV).[30] The band was also formed in the spectra as a

consequenceof thePI separation intoaromatic dianhydride

and diamine which contain the functional groups respon-

sible for this absorption region.

2 14 69.8 17.7 3.9 6.0

4 18 64.8 17.8 3.5 6.6

6 32 64.0 18.2 4.3 8.1

3.3. Atomic Force Microscopy (AFM)

The surface morphology of the plasma treated PI was

investigated by AFM (Ntegra Spectra, NT-MDT). An area of

10� 10mm2 was scanned for each sample (Figure 6). The

surface roughness of the PI film was evaluated in terms of

rms of the roughness. The untreated PI films are very

smooth with an rms of only 2nm, while after 2min

plasma treatment the surfaces showed a rms of 14nm. For

the 4min treatment time, the rmswas 18nm,while for the

6min treatment time, the roughness increased to 32nm. It

may be assumed that the excited species generated in the

plasma-induced changes in the surface morphology due to

the interaction with the PI functional groups.

3.4. Surface Analysis by XPS

XPS analysis was used to determine the elements present

on surface, their chemical states and, after quantitative

analysis, to finding out both the element and the chemical

state relative concentrations. XPS survey scanned for the

four PI film samples: untreated PI and plasma treated PI for

2, 4, and 6min. The carbon, oxygen, and nitrogen peaks

Figure 6. 3D AFM images of (a) the untreated PI film and of the plasmfor (b) 2min, (c) 4min, and (d) 6min.


were present in all spectra. Traces of silicon impurity from

residuals in the chamber were also observed without any

effectonthesurfacechemistry. Theatomic concentrationof

C, O, N, and F on the PI surface calculated from the XPS

spectra and the rms of the surface roughness are presented

in the Table 1. The carbon signal decreased gradually with

the increase of the plasma treatment time, from 73.7% in

the untreated sample to 64.0% after 6min of plasma

discharge. The oxygen concentration also decreased after

the plasma treatment, the minimum value being obtained

for the 2min of discharge (17.7%) (Table 1). Nitrogen

concentration slightly decreased for all the plasma treated

PI, from 5% (untreated sample) to 3.5% for 4min plasma

treated PI (Table 1). Taking into account the errors in the

quantitative data (�10%), it may be concluded that the

surface chemistry of the PI film was slightly modified,

depending on the exposure time of the film to the water

plasma. The most obvious change in the surface chemical

a-treated PI films

statewasgivenby thefluorinespecies.On

the surface of the untreated PI film could

not be detected the fluorine species, but

its concentration increased gradually

with the discharge time, from 6.0% for

the 2min plasma treated sample to 8.1%

in the case of the 6min plasma treated

PI. Thismaybe attributed to anadditional

fluorination of the surface induced

by the water plasma. The surface rough-

ness increased from 2 to 32nm after

the plasma treatment (Figure 6). Thus,

the water plasma yielded surfaces rich

in fluorine atoms and with higher

roughness.

The C1s spectrum of the untreated PI

sample could be resolved into four

component peaks which include the C—

C peak at 284.7 eV, the C—O peak at

286.6 eV, theC55Opeakat288.3 eV, thep–

p� shake-up satellite peak at 290.9 eV

(Figure 7a). The relative fractions of these

components are shown in the Table 2.


Figure 7. C1s XPS peak of the (a) untreated and (b) water plasma treated PI films.

Table 2. The relative fraction of the C—C, C55O, C—O, and CF3peaks of the C1s spectrum before and after the water plasmatreatment.

Treatment

time [min] C—C [%] C55O [%] C—O [%] CF3 [%]

– 76.3 9.0 12.6 –

2 71.7 8.3 17.0 3.0

4 72.5 9.0 12.8 4.1

6 70.8 9.1 15.6 4.5

C. Miron et al.

804

For the plasma-treated PI samples the percentage of the

C—Cbondinggraduallydecreased to70.8%.This showsthat

the cleavage of a low number of C—C bonds occurs during

the plasma treatment of the PI. The relative fraction of C55O

was slightly decreased (with�1%), while the C—O fraction

increased after the exposure to the plasma (from �12.6 to

17%). The dissociation energy of the C55Obond (11.09 eV) is

higher than that of the C—C bond (6.2 eV), explaining the

slight modification of the C55O fraction during the

discharge. The CF3 peak at 293.4 eV was identified only in

the plasma treated spectra (Figure 7b) and increased with

thedischarge time (Table2). Thismaybetheconsequenceof

a segregation process caused by the migration of the CF3groups at the polymer surface. This aspectwill be treated in

more detail later on in this paper.

The O1s peak exhibited three component peaks in the

reference sample: O55C at 531.9 eV, O—C at 533.2 eV, and

O2� at 530.6 eV (Figure 8a). The relative fractions of these

components are presented in the Table 3. The O55C peak

decreased, the lowest value obtained was for the 2min

plasma treatment (59.0%),while theO—Cpeak increased to

33.5% for the same treatment time (Figure 8b). The O—C

Figure 8. O1s XPS peak of the (a) untreated and (b) water plasma tr


groupmaybeaderivative of theO55Cgroupafter theaction

of the excited species with the polymer surface. It was

noticed just a tiny amount of oxygen bonded in the lattice

by mainly, Si, which was detected as residual component/

impurity on the surface’ sample (O2� at 530.6 eV).

The nitrogen core-level spectrum of the untreated

plasma sample exhibited two components: N—C55O peak

at 400.2 eV (75.3%) and N—C—C peak at 398.5 eV (24.7%)

(Table 4). The concentrationof theN—C55Ogroup increased

eated PI films.

DOI: 10.1002/ppap.201300015

Table 3. The relative fraction of the O55C, O—C, and O2� peaks ofthe O1s spectrum before and after the water plasma treatment.

Treatment

time [min] O55C [%] O—C [%] O2� [%]

– 78.9 12.3 8.8

2 59.0 33.5 7.6

4 75.9 19.9 4.2

6 74.3 19.8 5.9

Table 4. The relative fraction of the N—C—C and N—C55O peaksof the N1s spectrum before and after the water plasmatreatment.

Treatment

time [min] N—C—C [%] N-C55O [%]

– 24.7 75.3

2 – 100

4 – 100

6 – 100

Figure 10. Evolution of the liquid temperature with the dischargetime at 20mm from the core plasma.


up to 100% with the increasing of the plasma treatment

time, while the N—C—C group could not be identified in

any spectra of the plasma treated samples (Figure 9b). The

shift in the nitrogen core-level spectrum at 398.5 eV was

explained by some authors as the result of an incomplete

imidization of the polymer.[31] Therefore, it may be

assumed that the imidization was incomplete for the

untreated plasma film, process whichwas enhanced under

the water plasma action. Some authors stated that the

degree of imidization increases with the cure temperature

(more amide groups cyclodehydrate to imide rings) and

attain a maximum value at 240–250 8C.[30,32] Using a

digital thermometer Thermo Tech TT 0986, the water

temperaturewasmeasuredat20mmfromthecoreplasma,

Figure 9. N1s XPS peak of the (a) untreated and (b) water plasma tr


where the polymer filmwas placed during the experiments

(Figure 10). The water temperature was monitored for 10

minutes of discharge and it was observed that did not

increasemore than 70 8C in all this time. Itmay be assumed

that the imidization of the PI surface may be the

consequence of the high energy excited species which

escape from the plasma and reach the film surface. These

species might have a much higher temperature than that

measured in the surrounding water, inducing the surface

modifications mentioned above. The FTIR measurements

have shown the presence of characteristic band of the PI

6FDA-MMDA in the absorption spectra, indicating the fact

that the imidization occurred during the plasma discharge

determined the formation of PI having the same chemical

structure as that of the untreated sample (Figure 5).

Therefore, the both phenomena, the scission and the

imidization of the PIs, might occur on the film surface

under the water plasma treatment.

eated PI films.


Table 5. Contact angle values in (0) of the untreated and waterplasma treated PI samples.

Untreated

sample [0]

Plasma

treatment

time [min]

Plasma

treated

samples [0]

74 2 82

4 87

6 91

C. Miron et al.

806

The CF3 peak was identified at 291.6 eV only for the

plasma treated samples (Figure 7b). The concentrationwas

of 3.0% for the 2min plasma treated PI, for 4min treatment

increased to 4.1%, and after 6min, the CF3 concentration

was 4.5%. By angle-resolved XPS was investigated the

segregation of the CF3 group for a better understanding of

the surface composition versus the surface energy of the

samples. The segregation depths were evaluated starting

frominelasticmean freepath (imfp)by taking the sampling

depth:

Plasma

� 2013

d ¼ 3lcosu ð1Þ

where l is the imfp and u is the angle between the emitted

photoelectrons to the analyzer and the surface normal. The

errors are within the range (10–15%). The segregation

depthswere 4.2, 6, and 8.5 nm for the 2, 4, and 6minplasma

treated samples, respectively. The segregation process is

determined by the difference in the surface energy of the

polymers film components. The lower surface energy parts

will have the tendency to move to the PI surface/air

interface.[33]

It is well known that fluorine-containing species on a

polymer surface determine low surface tension, due to the

physical forces which held together the molecules. In this

regard, we further measured the water static contact angle

between the liquid meniscus and the polymer surface, by

means of the sessile drop technique. The measurements

were performed before and after the water plasma

treatment of the polymer films using a goniometer KSV

CAM 101. Uniform distilled water drop of 1ml volume was

placed on the polymer surface and the contact angle was

measured for 20 s with a video-based optical contact angle

measuring device. The contact angle gradually increased

with the increase of the time exposure of the film to the

plasma, from 740 for the untreated sample to 820, 870, and

910 for the 2, 4, and 6min plasma treated samples (Table 5).

Therefore, the energetic particles of the water plasma

induced modifications of the surface properties. It may be

assumed that the surface segregation of the CF3 groups

determined an increase in the surface hydrophobicity of

the polymer. The increase in the water contact angle may

be also attributed to the increase in the degree of

imidization of PI discussed above. During the imidization,

more polar carboxylic (COOH) and amide (CONH) groups

become less polar imide groups, leading to a significant

decrease in the strength of the polar part free energy. As a

result, the surface free energy of the resultant PI film is

diminished.[5]

4. Conclusion

Surface modification of the PI films (6FDA-MMDA) was

realized by water plasma treatment for 2, 4, and 6min.

Process. Polym. 2013, 10, 798–807WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Optical emission spectroscopy has shown that the reactive

species identified in the spectra were formed due to the

water dissociation and ionization. No other excited species

belonging to the polymer chemical structure could be

observed in the emission spectra.

FTIR measurements have shown the presence of

characteristic PI bands. In the FTIR spectra were also

identified absorption bands corresponding to the aromatic

dianhydride and aromatic diamine used in the PI synthesis,

suggesting a scission of the PI structure into its initial

forming components. With the increase of the discharge

time, broadening and shifts of the absorption bands were

seen in the spectra due to the highly energetic and

collisional species which interacted with the polymer

surface. These results were in concordance with the AFM

measurements,whichhave shownsurfacemodificationsof

the PI films. The surface roughness increased with the

increase of the discharge time.

XPSmeasurementswere used to determine the chemical

composition of the PI surface after the water plasma

treatment. It was observed that the imidization of PIs

occurred during the plasma treatment, accompanied by a

segregation of the CF3 groups at the polymer film surface.

These modifications have led to an increase in the surface

hydrophobicity of the PI samples, confirmed by the static

contact angle measurements, which have shown an

increase in the contact angle values with the increase of

the discharge time.

Therefore, the water plasma treatment induced mod-

ifications of the chemical composition of the PIs surface by

breaking the PI structure into its forming components,

followed by an imidization process of the PI. These

preliminary results have shown that the electrical dis-

charge in water may be used for a controlled and efficient

surface modification method of this type of polymer.

Further work will be devoted to the investigation of the PI

surface modification mechanism by the water plasma

technique.

Acknowledgments: The research leading to these results hasreceived funding from the European Union’s Seventh Framework

DOI: 10.1002/ppap.201300015


Programme (FP7/2007-2013) under grant agreement no 264115 –STREAM.

Received: January 30, 2013; Revised: March 25, 2013; Accepted:May 14, 2013; DOI: 10.1002/ppap.201300015

Keywords: plasma treatment; polyimides; surface modification;water plasma

[1] J.-Y. Jeon, T.-M. Tak, J. Appl. Polym. Sci. 1996, 61, 529.[2] A. V. Gribanov, Yu. N. Sazanov, M. Ya. Goikhman, M. V.

Mokeev, G. N. Fedorova, V. L. Ugolkov, Russ. J. Appl. Chem.2000, 73, 2089.

[3] H. N. Isfahani, K. Faghihi, M. Hajibeygi, M. Bokaei, Polym. Bull.2010, 64, 633.

[4] E. M. Maya, J. Benavente, J. de Abajoa, Mater. Chem. Phys.2012, 131, 581.

[5] I. Sava, S. Chisca, Mater. Chem. Phys. 2012, 134, 116.[6] I. Sava, S. Chisca, M. Bruma, G. Lisa, Polym. Bull. 2010, 65,

363.[7] G. Hougham, J. G. Tesoro, J. Shaw, Macromolecules 1994, 27,

3642.[8] C.-Y. Yang, J. S. Chen, S. L.-C. Hsu, J. Electrochem. Soc. 2006, 153,

F120.[9] J. B. Park, J. S. Oh, E. L. Gil, S. J. Kyoung, J. T. Lim, G. Y. Yeom, J.

Electrochem. Soc. 2010, 157, D614.[10] S. J. Park, H. J. Sohn, S. K. Hong, G. S. Shin, J. Colloid Interface

Sci. 2009, 332, 246.[11] J. F. Friedrich, R. Mix, R.-D. Schulze, A. Meyer-Plath, R. Joshi, S.

Wettmarshausen, Plasma Process. Polym. 2008, 5, 407.[12] R. Joshi, R.-D. Schulze, A. Meyer-Plath, J. F. Friedrich, Plasma

Process. Polym. 2008, 5, 695.[13] R. Joshi, R.-D. Schulze, A. Meyer-Plath, M. H. Wagner, J. F.

Friedrich, Plasma Process. Polym. 2009, 6, S218.[14] K.-Y. Shih, B. R. Locke, IEEE Trans. Plasma Sci. 2011, 39,

883.


[15] F. Liu, W. Wang, S. Wang, W. Zheng, Y. Wang, J. Electrostat.2007, 65, 445.

[16] C. P. Lungu, A. M. Lungu, P. Chiru, O. Pompilian, N. Georgescu,M. Magureanu, I. Mustata, T. Velea, D. Stanciu, V. Predica,Rom. J. Phys. 2009, 54, 369.

[17] C. Miron, M. A. Bratescu, N. Saito, O. Takai, Plasma Chem.Plasma Process. 2010, 30, 619.

[18] J. F. Kolb, R. P. Joshi, S. Xiao, K. H. Schoenbach, J. Phys. D: Appl.Phys. 2008, 41, 234007.

[19] U. Gangal, M. Srivastava, S. K. Sen Gupta, Plasma Chem.Plasma Process. 2010, 30, 299.

[20] B. Sun, M. Sato, A. Harano, J. S. Clements, J. Electrostat. 1998,43, 115.

[21] P. Lukes, M. Clupek, V. Babicky, P. Sunka, Plasma Sources Sci.Technol. 2008, 17, 024012.

[22] M. Sahni, B. R. Locke, Plasma Process. Polym. 2006, 3, 668.[23] Y. Yang, H. Kim, A. Starikovskiy, A. Fridman, Y. I. Cho, Water

Res. 2010, 44, 3659.[24] R. W. B. Pearse, A. G. Gaydon, The Identification of Molecular

Spectra, Chapman and Hall, London 1976.[25] J. Guan, F. Wang, T. Ziegler, H. Cox, J. Chem. Phys. 2006, 125,

044314.[26] A. Descoeudres, Ch. Hollenstein, G. Walder, R. Demellayer

Perez, Plasma Sources Sci. Technol. 2008, 17, 024008.[27] S.-J. Park, E.-J. Lee, S.-H. Kwon, Bull. Korean Chem. Soc. 2007,

28, 188.[28] E. Balan, S. Delattre, D. Roche, L. Segalen, G. Morin, M.

Guillaumet, M. Blanchard, M. Lazzeri, C. Brouder, E. K. H. Salje,Phys. Chem. Miner. 2010, 1, 12.

[29] M. S. Sevegney, R. M. Kannan, A. R. Siedle, P. A. Percha, J.Polym. Sci. Part B: Polym. Phys. 2005, 43, 439.

[30] I. Sava, S. Chisca, M. Bruma, G. Lisa, J. Therm. Anal. Calorim.2011, 104, 1135.

[31] D. Bhusari, H. Hayden, R. Tanikella, S. A. Bidstrup Allen, P. A.Kohl, J. Electrochem. Soc. 2005, 152, F162.

[32] J.-H. Jou, P.-T. Huang, Macromolecules 1991, 24, 3796.[33] T. Dikic, S. J. F. Erich, W. Ming, H. P. Huinink, P. C. Thune,

R. A. T. M. van Benthem, G. de With, Polymer 2007, 48,4063.


Documents

Surface Modification of the Polyimide Films by Electrical Discharges in Water