A_Comparison_of_Corona-Treated_and_Flame-Treated_P (1)(1)

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/228879300
A Comparison of Corona-Treated and Flame-Treated

Polypropylene Films
Article in Plasmas and Polymers · April 2003

DOI: 10.1023/A:1022817909276
CITATIONS READS
183 2,819
7 authors, including:
Michael Ulsh Melvyn Branch

National Renewable Energy Laboratory University of Colorado Boulder
119 PUBLICATIONS 1,765 CITATIONS 57 PUBLICATIONS 1,058 CITATIONS
SEE PROFILE SEE PROFILE
All content following this page was uploaded by Michael Ulsh on 15 April 2014.
The user has requested enhancement of the downloaded file.

P1: GCR
Plasmas and Polymers [papo] pp778-papo-461817 March 4, 2003 10:55 Style file version Nov. 19th, 1999
Plasmas and Polymers, Vol. 8, No. 1, March 2003 (°

C 2003)
A Comparison of Corona-Treated and Flame-Treated

Polypropylene Films
Mark Strobel,1,5 Viv Jones,1 Christopher S. Lyons,1
Michael Ulsh,1 Mark J. Kushner,2 Rajesh Dorai,3
and Melvyn C. Branch4
Received October 1, 2002; accepted October 29, 2002
The comparison of corona-treated and flame-treated polypropylene (PP) films provides

insight into the mechanism of these surface-oxidation processes. Atomic force mi-
croscopy (AFM), contact-angle measurements, and X-ray photoelectron spectroscopy
(XPS or ESCA) were used to characterize surface-treated biaxially oriented PP. While
both processes oxidize the PP surface, corona treatment leads to the formation of water-
soluble low-molecular-weight oxidized materials (LMWOM), while flame treatment
does not. Computational modeling of the gas-phase chemistry in an air corona was
performed using a zero-dimensional plasma-chemistry model. The modeling results
indicate that the ratio of O to OH is much higher in a corona discharge than in a flame.
Chain scission and the formation of LMWOM are associated with reactions involving
O atoms. The higher ratios of O to OH in a corona are more conducive to LMWOM
production. Surface-oxidized PP exhibits considerable thermodynamic contact-angle
hysteresis that is primarily caused by microscopic chemical heterogeneity.
KEY WORDS: corona surface modification; flame surface modification; atomic force
microscopy; electrical discharges; contact-angle measurements; low-molecular-weight
oxidized materials.
1. INTRODUCTION
Polypropylene (PP) films generally require some type of surface modifica-
tion prior to commercial use. Surface oxidation is the most common modification
performed to PP. Oxidized PP surfaces tend to have improved wetting and adhe-
sion properties when compared with the unmodified polymer. In industry, the two
1 3M Company, 3M Center, St. Paul, Minnesota.
2 Department of Electrical and Computer Engineering, University of Illinois, Urbana, Illinois.
3 Department of Chemical Engineering, University of Illinois, Urbana, Illinois.
4 Department of Mechanical Engineering, Center for Combustion and Environmental Research,
University of Colorado, Boulder, Colorado.

5 To whom correspondence should be addressed at 3M Company, 3M Center, Building 208-1-01, St.
Paul, Minnesota 55144-1000; e-mail: [email protected].

61
1084-0184/03/0300-0061/0 °
C 2003 Plenum Publishing Corporation
P1: GCR
62 Strobel, Jones, Lyons, Ulsh, Kushner, Dorai, and Branch
processes most widely used to oxidize PP are corona discharges and flame. Al-
though both processes oxidize a thin surface region of the PP, there are significant
differences between the surfaces generated by the two techniques. In an earlier
paper,(1) we compared corona discharge, flame, air plasma, ozone, and combined
UV/ozone (UVO) treatments. That work focused on the extent and depth of oxi-
dation. In this work, we use atomic force microscopy (AFM), X-ray photoelectron
spectroscopy (XPS or ESCA), and contact-angle measurements to investigate the
surface properties of PP modified by corona and flame processes. The detailed
comparison of corona and flame treatments leads to further insights into the mech-
anism of these important surface-oxidation processes.
Although difficult to investigate directly, “surface molecular weight” has a
critical role in determining the wettability, adhesion properties, and permanence
(aging behavior) of surface-modified polymers. In our opinion, surface molecular
weight is comparably important in the determination of practical performance
properties as the extent of oxidation and the types of oxidized functionalities
found on a modified polymer.
In the past,(2–4) we tried to investigate surface molecular weight through the
detection of water-soluble low-molecular-weight oxidized material (LMWOM).
The presence of LMWOM on polymers can be detected by a number of analytical
techniques performed in conjunction with washing of the treated surface. LMWOM
is characteristically soluble in “polar” solvents such as water, alcohols, and ketones,
but remains insoluble in nonpolar solvents such as the alkanes and aromatics.
More recently, we used AFM(5) to directly examine the surface topography of
flame-treated PP. We found that flame treatment creates a “nodular” surface topog-
raphy that is caused by the oxidation of the PP rather than by any heating of the PP
surface that might occur in the flame. The surface topography, wettability, and ex-
tent of oxidation of flame-treated PP were not affected by washing with any type of
solvent, so we concluded that no water-soluble LMWOM was formed. We believe
that the nodular topography of flame-treated PP is formed by the agglomeration
and rearrangement of intermediate-molecular-weight oxidized products.
Our recent study of flame surface treatments(5) benefited from the use of
the SPIN computational model to examine the gas-phase chemistry of impinging
flames. The identification and quantification of the gas-phase species in the flame
yielded a number of unexpected insights into the mechanism of flame treatment.
A computational model that provides similarly detailed gas-phase information for
the type of discharge used in industrial corona treating has not been previously
available. In this study, we present our first results from the computational modeling
of an industrial corona-treating system. The use of this new model helps to explain
a number of the characteristics of corona-treated PP.
The nomenclature of atmospheric-pressure electrical discharges can be some-
what confusing. The widely used process for polymer surface modification has long
been termed a “corona discharge.” However, because of the presence of a dielectric
P1: GCR
Comparison of Corona-Treated and Flame-Treated Polypropylene 63
material on at least one of the discharge electrodes, industrial corona discharges

are more properly termed “alternating-current, atmospheric-pressure dielectric-
barrier discharges.” In this paper, we retain the historic “corona” terminology that
is still overwhelmingly used in the polymer industry.
2. EXPERIMENTAL
2.1. Corona and Flame Treatments
Biaxially oriented 30-cm wide isotactic PP film was the substrate used in this
study. The 0.05-mm thick PP film was produced using a homopolymer resin with a
weight-average molecular weight of 3.6 × 105 , a polydispersity index of 4.0, and
a peak melting temperature of 163◦ C. The base resin contained ca. 200 ppm of
an inorganic acid scavenger and ca. 1000 ppm of high-molecular-weight phenolic
antioxidants. The additives are not detectable by either ESCA or time-of-flight
static secondary ion mass spectrometry (SSIMS). This is the same clean, well-
characterized commodity PP film that we have used in studies dating back to the
late 1980s.
Corona treatment was performed using industrial equipment manufactured by
Pillar Technologies (Hartland, WI). During treatment, the PP film was in contact
with a 25-cm diameter, 50-cm face-width steel ground roll coated with a 2.0-mm
thick layer of American Roller (Union Grove, WI) # CL500 ceramic dielectric.
In this so-called “covered-roll” electrode configuration, the powered electrode
consisted of two 200 cm2 , 33-cm face-width stainless steel “shoes” separated
from the ground roll by an electrode gap of 1.5 mm. The treater housing was
continually flushed with ca. 5000 cm3 /s of either dried or humidified air at 25◦ C to
maintain relative humidities in the discharge region of either 2–5% or 95–100%.
The net power dissipated in the corona was measured with the power me-
ter incorporated into the Pillar P1000 (3.5 kW) power supply. The power meter
was carefully calibrated using voltage–charge parallelograms (Lissajous figures),
a method first described by Manley(6) and refined by the manufacturers of indus-
trial corona equipment. (C. R. Ballard, Pillar Technologies, Hartland, WI, personal
communication, 1988 and 1993). The discharge frequency was 9.6 kHz. The nor-
malized energy of the corona treatment was calculated from the net power and the
film speed:
Normalized energy, E = P/ws
where P is the net power (in W), w is the electrode face width (in cm), and s is the
film speed (in cm/s). Typical units are J/cm2 . To obtain a range of surface oxidation
on the PP, we varied the corona energy from 0.1 to 17 J/cm2 by using film speeds
from 1 to 150 m/min at powers from 300 to 2000 W. At a given normalized energy,
all combinations of corona power and film speed yielded similar results.
P1: GCR
Flame treaters ordinarily employ laminar-premixed flames in which the fuel

and the oxidizer are thoroughly mixed prior to combustion. In a laminar-premixed
flame, the luminous region is that portion of the flame where the temperature rise
is the greatest and where much of the reaction and heat release occur. The exact
ratio of oxidizer to fuel needed for complete combustion is known as the stoi-
chiometric ratio. Flame properties are commonly correlated with the equivalence
ratio, which is defined as the stoichiometric oxidizer:fuel ratio divided by the ac-
tual oxidizer:fuel ratio. Throughout this work we used a fuel-lean flame with an
equivalence ratio of 0.94, which is the optimal condition for oxidizing PP film.(5)
Flame treatment was performed using equipment built by 3M Company. A
schematic diagram of the equipment and a description of the flame treatment were
published previously.(7) Dust-filtered, 25◦ C compressed air with a dew point of <
−20◦ C was premixed with the natural gas fuel in a venturi mixer located upstream
of a 30 × 1-cm ribbon burner supplied by the Flynn Burner Corporation (New
Rochelle, NY). The natural gas had a specific gravity of 0.577, a stoichiometric
ratio with air of 9.6:1, and an average heat content of 37.8 J/cm3 . The normalized
flame power used in these experiments was 500 W/cm2 of burner surface area. At
this normalized flame power, the tips of the luminous flame cones were ca. 3 mm
above the topmost surface of the ribbon burner and about 4 mm from the surface
of the PP film.
To generate a range in the extent of PP surface oxidation, we varied the film
speed from 100 to 300 m/min while leaving all of the other flame conditions
constant. As discussed previously,(8) the total downstream dimension of the zone
of reactive product gases is ca. 15 cm so that the exposure time of the film to the
active species of the flame varied from 0.03 to 0.09 s.
2.2. Surface Analysis

The PP film samples were analyzed by ESCA, contact-angle measurements,
and AFM. ESCA spectra were obtained on a Hewlett-Packard Model 5950A spec-
trometer using a monochromatic Al Kα photon source at an electron take-off
angle with respect to the surface of 38◦ . Spectra were referenced with respect to
the 285.0 eV carbon 1s level observed for hydrocarbon. No oxygen was detected
on the untreated PP. No elements other than oxygen and carbon were detected on
any treated PP. Based on repetitive analysis of flame- and corona-treated PP films,
we estimate that the standard deviations of the O/C atomic ratios obtained from
ESCA are 0.02. This standard deviation indicates the reproducibility of both the
treatments and the ESCA technique. ESCA was conducted within several hours
of treatment and occasionally thereafter for 180 days.
Measurements of the advancing and receding contact angles in air of deion-
ized, filtered water were made using the Wilhelmy plate method on a Cahn
DCA-322 dynamic contact-angle instrument. Using the microbalance, the sur-
P1: GCR
face tension of the water was measured as 72.6 mN/m at 20◦ C. A three-layer
laminate was prepared using 3M Scotch Brand #666 double-coated tape to mount
the treated sides of the film outward. To prevent contamination during the prepa-
ration of this laminate, the treated surfaces only contacted untreated PP film. This
situation is analogous to the common practice of winding modified film into roll
form after treatment. ESCA, contact-angle, and AFM analysis confirm that no
LMWOM is transferred to the untreated side of the PP film during the preparation
of the Wilhelmy laminates or during the aging of the treated films. The laminate
was cut into a 2.54 × 2.54-cm square for analysis. The stage speed was 50 µm/s
with a travel distance of ca. 1 cm. At the position of maximum travel, the motion of
the stage was halted for 2 min prior to starting the retraction phase of the Wilhelmy
cycle, thereby soaking the sample in water for a period of time. The volume of
water used for the contact-angle measurements was ca. 50 cm3 , with a fresh vol-
ume of water used for each sample analyzed. The advancing and receding contact
angles were calculated using a software routine supplied with the Cahn instrument
that uses linear regression for the buoyancy correction. The untreated PP film had
advancing and receding water contact angles of 109◦ and 87◦ , respectively. Typi-
cal standard deviations for the contact-angle measurements were 2◦ , except for the
advancing angles on LMWOM-containing surfaces, which typically had standard
deviations of 4◦ .
Measurements of the water contact angles were made within an hour after
treatment of the PP and occasionally thereafter for a period of 180 days. The
treated samples were stored in roll form at ambient conditions (18–22◦ C, 20–50%
RH). In addition, the treated films were washed with deionized, filtered water in a
controlled fashion using the Cahn instrument. During the measurement of contact
angles by the Wilhelmy plate method, washing of the surface is accomplished by
the cycle of forcing the PP film into the water bath, holding it for 2 min in the water,
and then retracting the film. Approximately 2 min after this washing, we performed
a second measurement of the advancing and receding contact angles. This second
cycle generated the water contact angles for the water-washed materials. Note that
the receding contact angles are always measured on “washed” surfaces because the
PP is immersed in water during the measurement of the advancing angle. Hence,
no difference in the receding angle for unwashed and washed samples should be
expected. This is the behavior that we observe for these flame-treated PP films.
Using the Cahn microbalance, we found that the surface tension of the water did
not change after these two cycles of immersion. This means that any material that
might have dissolved into the water from the PP film did not detectably alter the
properties of the water. After the two cycles of washing, the PP samples were
analyzed by ESCA or AFM, thereby generating the data for the water-washed
materials.
Virtually all surface-treated polymers exhibit contact-angle hysteresis, a mea-
sured difference between the advancing and receding contact angles. It is important
P1: GCR
to distinguish kinetic hysteresis, which is time or frequency dependent, from ther-

modynamic hysteresis, which is an invariant property of the surface. In kinetic
hysteresis, changes occur to the liquid–solid system on a time scale comparable to
the time of measurement so that the observed contact angles appear to change with
time. In thermodynamic hysteresis, there are no changes in the observed contact
angles during the course of the measurements.
The contact angles that we measured were completely unaffected by up to
six additional Wilhelmy immersion cycles performed subsequent to the initial
washing of the samples. In addition, the measured contact angles were not affected
by the delay time in air between successive cycles (from no delay to 24 h) or
by the time of submersion in the water during washing (from seconds to 24 h).
The measured water contact angles were also unaffected by the humidity in the
environmental chamber of the Cahn DCA during the measurement process, over
the range from 5 to 90% RH. Finally, over a range of Wilhelmy stage speeds from
20 to 265 µm/s, all of the contact angles varied by <3◦ from those reported for
the 50 µm/s stage speed. This variation in contact angle as a function of stage
speed is extremely small relative to the observed hysteresis on the surface-treated
films, which, as will be shown below, can range up to 50◦ or more. Taken together,
these observations are a strong indication that the contact angles measured on the
untreated and surface-treated PP are not time dependent, that is, they do not exhibit
a kinetic hysteresis.(9) Therefore, the observed contact-angle hysteresis must be
thermodynamic.
The untreated and surface-treated PP films were also examined using a Digital
Instruments Dimension 5000 SPM atomic force microscope operated in the tap-
ping mode using Olympus Company OMCL-AC160TS etched-silicon cantilever
probes, which have nominal spring constants of ca. 42 N/m and resonant frequen-
cies of ca. 270 kHz. The radii of curvature of the probes were estimated to be
<10 nm. The typical Asp /Ao setpoint ratio used for imaging was 1.5/2.0, where
Asp is the amplitude of the cantilever oscillation (in volts) at the imaging setpoint
and Ao is the amplitude of the cantilever oscillation (in volts) in free space. Scan
rates were typically 1.2 Hz. For each image, root mean square (rms) roughness val-
ues were calculated using Digital Instruments software version 4.42r4. To include
those surface structures of primary interest, the rms roughness was calculated for
features of less than 250 nm.
To insure that the information that we obtained from AFM accurately repre-
sents the topography of our PP films, numerous examinations were made of each
pertinent sample. Characteristic images are shown in this paper. Nearly identical
images of these same PP surfaces were obtained at 3M with contact-mode AFM
using a Nanoscope II microscope and at the University of Western Ontario (M. J.
Walzak, Surface Science Western, The University of Western Ontario, London,
Ontario, personal communication, 1999) with both non–contact- and contact-mode
AFM using a Topometrix Explorer AFM.
P1: GCR
Fig. 1. Gas-phase reaction mechanism for air-corona discharges.
2.3. Gas-Phase Corona Reaction Mechanisms

The simplified reaction mechanism for dry- and humid-air coronas that was
used in our computational model is shown in Fig. 1. Rate constants for the heavy
particle gas-phase reactions have been obtained from the NIST Chemical Kinetics
Database.(10) The electron avalanche that initiates the discharge is caused by ion-
ization of the background gases:
e + N2 → N2 + + e + e (1.1)
+
e + O2 → O2 + e + e (1.2)
e + H2 O → H2 O+ + e + e (1.3)
After the discharge pulse, the electrons are lost primarily by dissociative at-
tachment to O2 and H2 O to form O− and OH− , respectively. These ionic species
later undergo recombination reactions with positive ions to form neutral species.
The chemistry of interest to surface modification usually occurs on longer time
scales than these ionic reactions through the various free-radical reactions
described below.
P1: GCR
In dry air, the initiating free-radical reactions are the electron-impact disso-
ciations of the background gases, N2 and O2 , that produce N and O atoms:
e + N2 → N + N + e (2)
e + O2 → O + O + e (3)
Oxygen atoms are also produced by the quenching of excited states of N2 , princi-
pally N2 (A), by O2 :
e + N2 → N2 (A) + e (4.1)
−12 3 −1
N2 (A) + O2 → O + O + N2 k = 1.5 × 10 cm s (4.2)
Subsequent reactions of the O and N radicals include
N + O2 → NO + O k = 4.4 × 10−12 exp(−3220/T ) cm3 s−1 (5)
O + O 2 + M → O3 + M k = 6.9 × 10−34 (T /300)−1.25 cm6 s−1 (6)
where M is any nonreacting gas-phase species. In a steady-state discharge condi-
tion, ozone can be dissociated by reaction with O2 (1 1) to form O and O2 :
e + O2 → O2 (1 1) + e (7.1)
−11 3 −1
O2 ( 1) + O3 → O + O2 + O2
1
k = 9.96 × 10 exp(−3050/T ) cm s
(7.2)
In practice, no air corona is truly dry; the 2–5% RH discharges used in this
study are as low as are encountered in most industrial operations. When water
vapor is present in the corona, electron-impact dissociation of H2 O produces H
atoms and OH radicals:
e + H2 O → H + OH + e (8)
The H atoms react with O2 to form HO2 radicals:
H + O2 + M → HO2 + M k = 6.14 × 10−32 (T /300)−1.6 cm6 s−1 (9)
The OH radicals can react with N to form NO or with HO2 to form H2 O and O2 :
OH + N → NO + H k = 3.8 × 10−11 exp(+85/T ) cm3 s−1 (10)
−11 3 −1
OH + HO2 → H2 O + O2 k = 8 × 10 cm s (11)
Most of the NO formed in the corona is converted to NO2 by reactions with O3
and HO2 radicals:
NO + O3 → NO2 + O2 k = 4.3 × 10−12 exp(−1598/T ) cm3 s−1 (12.1)
−12 3 −1
NO + HO2 → NO2 + OH k = 3.7 × 10 exp(+240/T ) cm s (12.2)
P1: GCR
Hydrogen peroxide (H2 O2 ) is formed by the reaction of HO2 with HO2 :

HO2 + HO2 + M → H2 O2 + O2 + M k = 1.9 × 10−33 exp(+980/T ) cm6 s−1
(13)
Thus, in industrial coronas in air, in which water vapor is always present to some
extent, the active gas-phase species are O, OH, O2 , O3 , HO2 , H2 O2 , N, and NOx .
The corona-discharge devices that are typically used for the surface modifi-
cation of polymers are repetitively pulsed by applications of a sinusoidal voltage
waveform, usually at frequencies of 5–50 kHz. The electric discharge occurs on
the leading edges of the positive and negative voltage waveforms (the “voltage
pulse”), producing current pulses of tens of nanoseconds. The charging of the se-
ries capacitance in the discharge circuit terminates the current pulse. The plasma
formed in the volume of the corona is typically in the form of arrays of microdis-
charges that are tens to hundreds of micrometers in diameter with area densities of
hundreds of microdischarges per square centimeter. Although the charged particles
are produced in an inhomogeneous manner, the contribution of ions to the surface
treatment is negligible, as discussed below in Section 4. The neutral radicals in the
microdischarges have longer lifetimes and larger mobilities than the ions. Thus, to
a great degree, diffusion homogenizes the initially nonuniform generation of these
species. The end result is that the surface treatment is fairly homogeneous and uni-
form. There are very few, if any, instances in which industrial corona treatments
are considered to be objectionably nonuniform.
2.4. Corona Computational Model

GLOBAL KIN is a zero-dimensional model used to investigate complex
plasma chemistries. It consists of circuit and plasma-chemistry modules.(11) The
model has been validated by comparison to experimental data from studies of
the plasma remediation of toxic compounds, such as sulfur dioxide and the nitro-
gen oxides, from effluent streams.(11–13) GLOBAL KIN models the chemistry in a
well-stirred batch reactor based on a user-defined reaction mechanism. The circuit
module of the model provides the E/N ratio (electric field/number density) in the
plasma, which is then used to obtain the electron–impact reaction-rate coefficients.
For this, a table of rate coefficients as a function of electron temperature is gen-
erated by solving the Boltzmann equation for electron-energy distribution using a
two-term spherical harmonic expansion for several values of the E/N ratio. The
equations are integrated as functions of time using VODE, an ordinary-differential-
equation solver.(14)
In corona treating, the PP film typically moves at speeds that generate a
boundary layer of gas that is drawn through the discharge by the film. The residence
time of the gas in the discharge is therefore approximated as being the same as
P1: GCR
that of the PP. In this residence time, a given surface area of the polymer and the
gas lying above it are exposed to many discharge pulses.
We used computational parameters that best simulated the corona operating
conditions described in Section 2.1. The number of voltage pulses that the PP is
exposed to is determined by multiplying the voltage frequency by the residence
time of the film in the discharge. To model the range of residence times encoun-
tered in our experiments, the number of voltage pulses studied was varied from
480 to 67,200. For most simulations of a typical air corona, we used a film speed of
150 m/min for the full downstream electrode length of 12 cm, which corresponds
to 921 voltage pulses and a treatment time of ca. 0.05 s. For most of these simula-
tions, we used an inlet gas composition of 79% N2 , 20% O2 , and 1% H2 O, which
corresponds to an RH of ca. 28% at 300 K. In a separate series of simulations, the
RH in the discharge was varied from 1 to 100%. For an applied voltage of 15 kV,
the corona energy was approximately 1.7 J/cm2 .
3. EXPERIMENTAL RESULTS
3.1. LMWOM
Table I shows the surface properties of a number of corona- and flame-treated
PP films. The corona and flame treatment conditions were varied to generate a series
of PP samples with varying levels of oxidation. All of the treated surfaces have
considerable hysteresis ranging from 35 to 60◦ . The most interesting difference
between the two types of treated surfaces is the effect of water washing. In the
case of flame-treated PP, washing with water has no effect on the wettability or on
Table I. Surface Propertiesa of Modified Polypropylene Films Before and After Washing With Water
ESCA O/C atomic ratio θa (◦ )b θr (◦ )b
Treatment Unwashed Washed Unwashed Washed Unwashed Washed
None 0.0 0.0 110 109 83 83
Flame (0.03 s) 0.07 0.07 97 96 41 42
Flame (0.06 s) 0.12 0.12 94 94 29 29
Flame (0.09 s) 0.18 0.18 84 85 21 21
For RH = 2–5%
Corona (0.17 J/cm2 ) 0.07 0.04 93 97 56 58
Corona (1.7 J/cm2 ) 0.12 0.06 76 96 51 53
Corona (17 J/cm2 ) 0.23 0.08 59 88 35 36
For RH = 95–100%
Corona (0.17 J/cm2 ) 0.08 0.06 91 97 54 54
Corona (1.7 J/cm2 ) 0.12 0.07 75 96 50 51
Corona (17 J/cm2 ) 0.23 0.09 54 90 31 30
a Properties measured within a few hours of surface treatment.
bθ and θr are the advancing and receding contact angles of water in air, respectively.
a
P1: GCR
the ESCA O/C atomic ratio.(5) As expected,(2–4) however, the surface chemistry of
corona-treated PP is strongly affected by water washing, with a significant loss of
surface oxidation and a noticeable increase in the advancing contact angle of water.
This is clear evidence of the presence of water-soluble LMWOM on the corona-
treated PP. Note that flame-treated PP, having an ESCA O/C atomic ratio of 0.18,
has no detectable LMWOM, whereas corona-treated polypropylene, having an
O/C ratio of only 0.07–0.08, has clearly evident LMWOM.
Using deionized, filtered water, we carefully washed the surface of a large
quantity of LMWOM-containing corona-treated PP, collected the solution, passed
the liquid through a 1-µm ceramic filter to remove any insoluble contaminants,
and then evaporated the water at room temperature in a dry, particulate-filtered
air stream. After the water was evaporated, the remaining LMWOM was an
oily, highly viscous liquid. We then analyzed this extract by laser-desorption
mass spectrometry (LDMS) (L. Hanley, University of Illinois at Chicago, per-
sonal communication, 1993–1996). In a typical LDMS experiment, LMWOM
was placed onto either clean gold or stainless steel probe tips. Pulsed 1064-nm
radiation from a Nb:YAG laser was used to desorb the molecules of LMWOM
from the probe tip; the desorbed molecules were then analyzed using an ion-trap
mass spectrometer. The LDMS technique enables a quantitative determination of
the weight-average molecular weight of the extracted LMWOM. LDMS analysis
of the LMWOM from a variety of corona-treated PP films yielded a molecular
weight of 400 ± 80 D, with some fragments having molecular weights as low as
100 D. This molecular weight agrees closely with the gel-permeation chromatog-
raphy results reported by Overney et al.(15) for the water-soluble extract from
corona-treated PP.
Analysis of the collected LMWOM using gas chromatography, 1 H NMR,
IR, and LDMS indicates that the LMWOM arises from cleavage of the PP chain
into a series of oligomers containing oxidized groups such as COOH, CHO, or
CH2 OH. Evidence of some unsaturation was also observed by NMR. Within the
error of the LDMS experiments, the molecular weight of the LMWOM was largely
independent of the corona energy or the humidity in the discharge over the range
of 2–100%.
Figure 2 shows an AFM image of untreated biaxially oriented PP, show-
ing the characteristic fine fibrillar network that has been observed by numerous
researchers.(5,16–21) The diameters of the fibrils are typically <10 nm. The untreated
PP has the finest structural features that we have observed with AFM. The fibrillar
structure is a result of the quenching and subsequent orientation of the PP.(19)
Figure 3 is a representative AFM image of flame-treated PP having an ESCA
O/C atomic ratio of 0.18. Flame treatment alters the fibrils by forming small
nodules, which have the appearance of a “bead-like” structure. This structure is
much less distinct and of larger characteristic dimensions than the fibrils of the
untreated PP. The characteristic size of the nodules is in the submicrometer range.
P1: GCR
Fig. 2. Tapping-mode AFM topographic image of an untreated biaxially oriented polypropylene

film. From the darkest to the lightest tone represents 50 nm of height in the z-direction. The rms
roughness of this sample is 0.67 nm.
As discussed in detail in our earlier report on flame-treated PP,(5) the formation

of the bead-like structure is not a result of surface heating but is directly asso-
ciated with the oxidation of the PP by the flame. The nodular topography of the
flame-treated materials prepared for this study is not removed or visibly altered
by vigorous extraction with water, methanol, acetone, or xylene.(5) The combined
contact-angle, ESCA, and AFM results demonstrate that there is no water-soluble
material formed on our PP as a result of exposure to the flames used in this study.
By contrast, corona-treated PP shows clear evidence of LMWOM in the AFM
images. Figure 4 shows a characteristic image of unwashed PP that was corona-
treated at an RH of 95–100% and a corona energy of 1.7 J/cm2 . This material
had an O/C atomic ratio of 0.12. The dominant topographical features are smooth,
P1: GCR
Fig. 3. Tapping-mode AFM topographic image of a flame-treated polypropylene film having an

ESCA O/C atomic ratio of 0.18. Washing of the film with water has no detectable effect on surface
topography. From the darkest to the lightest tone represents 50 nm of height in the z-direction. The
rms roughness of this sample is 0.35 nm.
globular “mounds” that are both much larger and less distinct than either the fibrils
of the untreated PP or the nodules formed by flame treatment. The height of the
mounds is >50 nm, while the area of a typical mound is ca. 15,000 nm2 . At
higher O/C ratios (higher corona energies), still larger mounds are observed (AFM
images not shown). Also visible is a smaller secondary structure located amidst
the mounds that appears similar to the nodules observed on flame-treated PP.
At lower relative humidities, the mounds are smaller. Figure 5 shows a char-
acteristic AFM image of unwashed PP corona-treated at an RH of 2–5% and a
corona energy of 1.7 J/cm2 . This treated material also had an O/C ratio of 0.12.
The height of the mounds is ca. 30 nm and the typical area is only ca. 500 nm2 .
P1: GCR
Fig. 4. Tapping-mode AFM topographic image of a polypropylene film treated in an air corona
discharge at 95–100% RH and 1.7 J/cm2 . The treated film has an ESCA O/C atomic ratio of 0.12.
From the darkest to the lightest tone represents 50 nm of height in the z-direction. The rms roughness
of this sample is 0.48 nm.
As shown by Fig. 6, upon washing the PP with water or other polar solvents,
the globular mounds are completely removed, revealing a nodular structure that
is similar to the topography of flame-treated PP. However, the nodules present on
water-washed, corona-treated PP appear more varied in size and are somewhat
larger than the nodules formed by flame treatment. The topography of corona-
treated films after washing with water appears to be independent of the discharge
RH. Obviously, the mounds are water-soluble LMWOM that has agglomerated
on top of and amidst the nodular structure. The nodular topography of the water-
washed corona-treated PP is unaffected by subsequent 24-h extraction with 25◦ C
water or acetone.
P1: GCR
Fig. 5. Tapping-mode AFM topographic image of a polypropylene film treated in an air corona
discharge at 2–5% RH and 1.7 J/cm2 . The treated film has an ESCA O/C atomic ratio of 0.12. From
the darkest to the lightest tone represents 50 nm of height in the z-direction. The rms roughness of
this sample is 0.39 nm.
Our earlier studies of corona-treated PP(2) using scanning electron microscopy

(SEM) did not detect any topographical features at low discharge RH, but the
mounds of LMWOM detected by AFM at low RH are sufficiently small that we
would not have been able to easily observe them using SEM. At a discharge RH of
95–100%, the mounds of LMWOM are much larger than at low RH, thus reaffirm-
ing our SEM results from the 1980s.(2) In that earlier work, we used gravimetric
methods to prove that more LMWOM was actually formed at low RH than at
high RH. It thus appears that large amounts of water vapor may not be necessary
for the formation of agglomerated mounds of LMWOM, but that the presence of
water aids in the aggregation of LMWOM into larger topographical features. In
P1: GCR
Fig. 6. Characteristic tapping-mode AFM topographic image of a corona-treated polypropylene film

following washing with water. There is no detectable effect of discharge RH on the topography of
the treated films following washing. The washed film has an ESCA O/C atomic ratio of 0.06. From
the darkest to the lightest tone represents 50 nm of height in the z-direction. The rms roughness of
this sample of washed PP is 0.36 nm.
summary, the amount of LMWOM formed is greater at low RH, but the size of the
agglomerated features increases as the RH in the discharge increases.
For both flame and corona treatments, the fine fibrillar structure of the un-
treated PP is not restored by any type of washing that we have performed. In
addition, we found that annealing of the treated PP for up to 24 h at 100◦ C in a
vacuum oven does not restore the fine topographical features of the untreated PP.
Other researchers have previously observed similar structures on corona-
treated PP. In 1993, Overney et al.(15) were the first to publish AFM images of
corona-treated PP. They observed “droplets” that could be removed by gentle
wiping. Overney et al. did not generate any images of washed materials. O’Hare
P1: GCR
et al.(20) , Greenwood et al.,(22) and Boyd et al.(23) observed large, micron-sized

globular mounds on corona-treated PP that expanded in size as the corona energy
was increased. Both O’Hare et al. and Boyd et al. also reported that, although the
mounds disappeared after washing with polar solvents, the fine structure of the
untreated PP was not restored. Instead, small topographic features in the 0.1-µm
size range remained on the washed surface. Finally, Vancso et al.(24) reported
“droplet-like” features on corona-treated PP, which, they stated, confirmed the
results of Overney et al. These AFM studies lead to the conclusion that the globular
mounds are LMWOM.
All of the recent AFM investigations of corona-treated PP confirm earlier
observations made using SEM.(2,25,26) The researchers who used SEM typically
found “bumps,” “mounds,” or a “frosted” surface on PP that was treated in a corona
at high energies and/or high humidities. Among the many advantages of AFM is
the ability to detect the mounds of LMWOM on PP at the much lower corona
energies typically used in industrial treatments. For example, we(2) obtained high
quality SEM images of LMWOM mounds only for treatments at corona energies
>10 J/cm2 while O’Hare et al.(20) found that AFM clearly detects the mounds at
corona energies as low as 0.5 J/cm2 .
LMWOM generated using other surface-oxidation processes forms topog-
raphy similar to that produced by corona treatment. Both Greenwood et al.(22)
and Mahlberg et al.(27) showed that mounds of LMWOM are formed by the low-
pressure oxygen-plasma treatment of PP. Using 3M PP film identical to that used in
this work, Nie et al.(18) found that combined UV and ozone (UVO) treatment yields
a surface consisting of larger globular mounds and smaller nodules, which corre-
spond to the topographical features that we observe on corona-treated PP. Since
Nie et al. found that washing with water removed the larger droplet-like mounds
but retained the smaller nodules, they also ascribed the mounds to agglomerated
LMWOM.
3.2. Treatment Stability

Table II shows the changes in the properties of representative surface-oxidized
PP films after 6 months of ambient aging in roll form. Films treated at conditions
Table II. Changes in the Surface Properties of Modified Polypropylene Films After 6 Months
of Ambient Aging
Initial O/C Change in O/C Change in Change in
Treatment atomic ratio atomic ratio θa (◦ )a θr (◦ )a
Flame 0.18 0.00 0 +3
Unwashed corona (for RH = 2–5%) 0.12 −0.01 +9 +9
Washed corona 0.06 0.00 +4 +9
aθ and θr are the advancing and receding contact angles of water in air, respectively.
a
P1: GCR
that yield other O/C atomic ratios show similar trends. Long-term aging of flame-
treated PP does not cause a significant loss in wettability or in the extent of ox-
idation. Only the receding contact angle of water decreases over a 180-day pe-
riod, and then by only 3◦ . In contrast, corona-treated PP exhibits more significant
changes upon aging. On unwashed corona-treated PP, there are modest increases
in both the advancing and receding contact angles that are accompanied by a small
decrease in the O/C atomic ratio, which is of the order of the standard deviation
in the atomic ratios. For corona-treated PP that is washed with water and then
subsequently aged, there is no change in the O/C atomic ratio, but there are still
detectable increases in the advancing and receding contact angles.(3) While both
treated surfaces are quite stable when compared with, for example, corona-treated
poly(ethylene terephthalate) (PET),(4) flame-treated PP is definitely more stable
than corona-treated PP.
4. COMPUTATIONAL RESULTS
The electron number densities (n e ) for the first, last, and selected inter-
mediate modeled pulses are shown in Fig. 7(a). The peak n e is approximately
6 × 1014 cm−3 . The electron density during the first pulse is smaller than the den-
sities at later pulses because the dielectric has yet to be charged. During later
pulses, the electron density attains a periodic steady state (PSS); the higher densi-
ties are the result of the charged dielectric contributing voltage to the gap. Because
the concentrations of the major background gases N2 , O2 , and H2 O remain nearly
constant over the series of discharge pulses, electron production in the later pulses
remains essentially unchanged.
The corresponding electron temperature (Te ) during the pulses is plotted in
Fig. 7(b). The peak Te is approximately 9.5 eV. As in the case of n e , the difference
in Te between the initial and later pulses occurs because of the incomplete charging
of the dielectric. Because of the residual voltage on the dielectric in the later pulses,
the E/N ratio is higher, producing a higher Te and, hence, a more rapid electron
avalanche.
The concentrations of O, N, and OH for the first, last, and selected intermediate
pulses are shown in Fig. 8(a) and (b). The concentrations of these primary radicals
reach an asymptote after hundreds of discharge pulses. Any H atoms formed from
water in the discharge are quickly consumed by reactions with O2 to form HO2 so
that both the lifetime (∼µs) and the concentration of H in the discharge are much
less than that of O and OH.
The temporal evolution of the densities of the ionic species O2 + and N4 +
is shown in Fig. 8(c). O2 + is produced by the electron impact ionization of O2
(Eq. (1.2)), while N4 + is formed by the reaction of N2 with N2 + ,
N2 + + N2 + M → N4 + + M. (14)
P1: GCR
Fig. 7. Discharge properties for the first, the last, and se-
lected intermediate pulses: (a) n e ; (b) Te . The gas mixture is
N2 /O2 /H2 O=79/20/1 at 300 K and 1 atm pressure. Electron
densities during the first pulse differ from those of the later
pulses because of the smaller voltage across the gap resulting
from the incomplete charging of the dielectric. Correspond-
ingly, Te is also smaller during the first pulse. After 10 s of
pulses, the electron density and temperature reach a periodic
steady-state (PSS). The electron density and temperature re-
main the same during the later pulses because the density of
the background gases does not change significantly. The ap-
plied voltage is 15 kV and the corona energy is 1.7 J/cm2 .
In general, the ions are consumed faster than the radicals are. As a result, the time-
averaged concentrations of ions are much smaller than that of the radicals. Hence,
the ion chemistry at the surface can be safely neglected. An additional reason for
neglecting ion chemistry in corona treatment is that the ions arriving at the PP
P1: GCR
80
P1: GCR
surface are thermal. This is a result of two facts. Firstly, for most of the time that
the polymer is in the discharge, the voltage across the electrode gap is nearly zero.
Hence, there is no external electrical field to accelerate the ions toward the surface.
Secondly, many collisions occur in the sheath region of an atmospheric-pressure
discharge so that those ions that do arrive at the surface are thermal.
4.1. Effect of Corona Energy

As discussed above, as the corona energy increases, the extent of oxidation
and the amount of LMWOM produced also increases. To examine the effect of
corona energy on the gas-phase species, the time-averaged number concentrations
of O and OH were calculated for selected energies.
The PSS electron densities are shown in Fig. 9(a) for normalized energies
of 0.9, 1.3, 1.7, 2.0, and 2.4 J/cm2 , which correspond to applied voltages of 10.0,
12.5, 15.0, 17.5, and 20.0 kV, respectively. The RH for these cases is ca. 28%. At
the higher applied voltages, the electron-impact rate coefficients are larger thereby
initiating a more rapid avalanche. The PSS densities of O atoms and OH radicals
resulting from this avalanche are shown in Fig. 9(b) and (c). With increasing corona
energy, the increased electron density produces more dissociation of O2 and H2 O,
resulting in the increased production of O and OH. The primary loss process for
O atoms is by reaction with O2 to form O3 (Eq. (6)). OH radicals are mainly
consumed by reactions with N and HO2 (Eq. (10) and (11)) to form NO and H2 O.
For our conditions, NO concentrations are a few ppm while NO2 concentrations
are a few tenths of a ppm.
The time-averaged concentrations of selected species were calculated accord-
ing to the equation:
Rτ
Ni dt
< Ni > = 0
τ
where < Ni > is the time-averaged concentration, Ni is the instantaneous concen-
tration of species i, and τ is the residence time of the PP in the discharge. These
time-averaged values provide a sense of the species concentration that affects the
PP surface during passage through the discharge.
The ratio of the time-averaged concentrations of O, denoted by <O>, to OH,
denoted by <OH>, as a function of corona energy is shown in Fig. 10(a). As the
←
Fig. 8. Densities of primary radicals and ions for the first, the last, and selected intermediate pulses:
(a) N; (b) O and OH; (c) O2 + and N4 + . Conditions are the same as described for Fig. 7. The difference
in the densities of the ions and radicals between the first and later pulses is due to the difference in the
production of electrons arising from the incomplete charging of the dielectric. After 100s of discharge
pulses, the radicals attain a PSS. Ions attain a periodic steady-state (PSS) in densities more quickly, in
a few discharge pulses. Ions are consumed much faster as compared with the radicals.
P1: GCR
Fig. 9. Periodic steady-state (PSS) densities of electrons

and radicals as a function of energy deposition: (a) n e ;
(b) O; (c) OH. Conditions are the same as described for
Fig. 7. Corona energies of 0.9, 1.3, 1.7, 2.0, and 2.4 J/cm2
correspond to applied voltages of 10.0, 12.5, 15.0, 17.5,
and 20.0 kV, respectively. Higher applied voltages pro-
duce hotter electrons resulting in higher electron densi-
ties and a more rapid avalanche. This results in increased
rates of dissociation of the background gases, thereby
producing more O and OH. For a given energy depo-
sition, peak OH concentrations are smaller than peak
O-atom concentrations.
corona energy increases, the <O>/<OH> ratio also increases. The proportion of
the corona energy channeled into the reactions that lead to the direct and indirect
production of O and OH as a function of the E/N ratio is shown in Fig. 10(b).
For applied voltages of 10–20 kV, the E/N ratio is 270–540 Td (1 Td = 1 ×
10−17 V cm2 ). In this range, most of the energy deposition is channeled into the
excitation of N2 and O2 . Through the reactions with O2 and O3 (Eqs. (4.2) and
(7.2)), the increased densities of N2 (A) and O2 (1 1) produce more O. Although
the fractional energy expended for the dissociation of H2 O to produce OH also
increases with increasing E/N ratio, the absolute increase is small.
P1: GCR
Fig. 10. Effect of corona energy on the time-averaged concen-

trations of O and OH: (a) <O>/<OH> as a function of corona
energy; (b) fractional energy deposition into the electron-impact
reactions resulting in the direct and indirect production of O and
OH. At the E/N ratios of interest (100s of Td; 1 Td = 1 × 10−17
V cm2 ), most of the input energy goes into the vibrational ex-
citation of N2 (producing N2 (A)) and the electronic excitation
of O2 (producing O2 (1 1)). Both N2 (A) and O2 (1 1) produce O
by reaction with O2 and O3 (Eqs. (4.2) and (7.2)). Hence, as the
E/N ratio increases, O-atom production increases. Conditions
are the same as in Fig. 7.
P1: GCR
The increase in <O>/<OH> with energy is also caused, in part, by the higher
gas-phase reactivity of OH relative to O, which results in a higher consumption of
OH in the corona. For example, the rate coefficients at 298 K for the reactions of
O and OH with various reactants are:
OH + N → NO + H k1 = 5.0 × 10−11 cm3 s−1 (15.1)
O + N → NO k2 = 2.2 × 10−13 cm3 s−1 (15.2)
−11 3 −1
OH + NO2 → HNO3 k1 = 5.4 × 10 cm s (16.1)
O + NO2 → NO + O2 k2 = 9.7 × 10−12 cm3 s−1 (16.2)
−11 3 −1
OH + HO2 → H2 O + O2 k1 = 8.0 × 10 cm s (17.1)
O + HO2 → OH + O2 k2 = 1.5 × 10−11 cm3 s−1 (17.2)
−14 3 −1
OH + O3 → HO2 + O2 k1 = 6.8 × 10 cm s (18.1)
−15 3 −1
O + O3 → O2 + O2 k2 = 8.4 × 10 cm s (18.2)
In all cases, OH is more reactive than O. Thus, not only is a smaller amount of
OH produced as compared with O, but the OH that is produced is then consumed
more rapidly than the O. The net result is a lower time-averaged concentration of
OH as compared with O.
Some of the O-atom consuming reactions are reversible. For example, the
consumption of O by O2 to form O3 (Eq. (6)) is followed by O3 conversion back
to O by the reaction with O2 (1 1) (Eq. (7.2)). With this cyclic process, higher O
atom concentrations can be maintained, which also leads to higher <O>/<OH>
ratios.
One of the notable gaseous products of air-corona processing is O3 . The time-
averaged concentration of O3 as a function of corona energy is shown in Fig. 11.
The primary loss processes for O3 are the reactions with O2 (1 1) (Eq. (7.2)) and H:
H + O3 → OH + O2 k = 1.4 × 10−10 exp(−470/T ) cm3 s−1 (19)
The modeling results shown in Fig. 11 agree with the observation(28) that ozone
production in a corona treater increases relatively linearly with corona power at
low power loadings.
Another product of an air corona is H2 O2 . The time-averaged concentrations
of H2 O2 and HO2 as a function of corona energy are also shown in Fig. 11. With
increasing energy, more H is produced by the electron impact dissociation of H2 O
(Eq. (8)), which then reacts with O2 to form HO2 (Eq. (9)). HO2 then reacts with
itself to produce H2 O2 (Eq. (13)).
4.2. Effect of Relative Humidity

The ratio of the time-averaged concentrations of O and OH as a function of
RH is shown in Fig. 12 for a corona energy of 1.7 J/cm2 . With increasing RH,
P1: GCR
Fig. 11. Time-averaged concentrations of O3 , H2 O2 , and

HO2 as a function of corona energy. O3 is produced by the
reaction of O atoms with O2 (Eq. (6)). H2 O2 is produced by
the reaction of HO2 with itself (Eq. (13)). With increasing
energy deposition, larger amounts of O and H atoms are pro-
duced by the electron-impact dissociation of O2 and H2 O,
which then produce more O3 and HO2 (Eqs. (6) and (9)).
Increased HO2 formation leads to the increased production
of H2 O2 . Conditions are the same as in Fig. 7.
<O>/<OH> decreases due to the increase in <OH> that results from the higher
rates of electron-impact dissociation of H2 O. As a disproportionate amount of
power is dissipated into the H2 O, less power is available for O-atom production.
Hence <O> decreases and as a result, <O3 > also decreases as RH is increased
(Fig. 12). This result again agrees with the observation that ozone production in
coronas and ozonizers decreases as RH increases.(28) The qualitative agreement
between experimentally determined ozone production rates, both as a function
of energy and of RH, and our modeling data for ozone production provides a
meaningful validation of the computational model used in this study.
4.3. Comparison of Corona Gas-Phase Chemistry to Flame

Gas-Phase Chemistry
In 2001, we reported(5) on the use of SPIN to model the chemistry of impinging
air–hydrocarbon flames. The SPIN modeling results allowed us to relate the flame
equivalence ratio to the flame chemistry, which was then correlated to the surface
oxidation generated by the flame treatment. For PP, optimal oxidation occurs at an
equivalence ratio of ca. 0.94, which is the equivalence ratio used to prepare all of the
flame-treated PP films used in this study. As calculated by SPIN, the mole fractions
of the important gas-phase species at the surface of the PP at an equivalence ratio
of 0.94 are listed in Table III. In addition, in a fuel-lean flame, there is considerable
P1: GCR
Fig. 12. Variation of <O>/<OH> and <O3 > as a

function of RH at a corona energy of 1.7 J/cm2 . At
higher RH, more energy is channeled into OH pro-
duction resulting in a decreased <O>/<OH>. <O>
decreases with increasing RH because of the decreased
rate of dissociation of O2 . As a result, <O3 > decreases.
free O2 present from the excess air, but no ozone. Thus, the optimal flame conditions
for oxidizing PP are characterized by high concentrations of the oxidizing species
OH, HO2 , H2 O2 , and O2 coupled with relatively low concentrations of H and O
atoms.(5) The ratio of [O]:[OH] in this type of flame is ca. 0.006.
Table III also lists the time-averaged concentrations of the same gas-phase
species in the corona at low and high RH. An air corona is characterized by high
concentrations of O, HO2 , H2 O2 , and O3 , but a relatively low concentration of OH
and virtually no H atoms. In a corona, the <O>:<OH> ratio decreases from ca.
Table III. Concentrations and Ratios of Some of the Important

Gas-Phase Reactive Species in a Flame and an Air Corona
Time-averaged concentrationsb
in an air corona
Species Mole fractions in
a fuel-lean flamea at RH = 1% at RH = 100%
OH 51.4 × 10−6 0.18 × 1014 1.34 × 1014
HO2 30.4 × 10−6 1.46 × 1014 5.00 × 1014
H2 O2 18.7 × 10−6 4.98 × 1014 16.0 × 1014
H 15.9 × 10−6 0.002 × 1014 0.016 × 1014
O 0.3 × 10−6 7.57 × 1014 4.88 × 1014
O3 <10−12 4700 × 1014 1800 × 1014
O/OH 0.006 42.1 3.63
a At an equivalence ratio of 0.94.
b Concentrations in species per cm3 .
P1: GCR
42 to 4 as the RH increases from 1 to 100%.
5. DISCUSSION
5.1. Oxidation and Chain Scission
On PP, LMWOM is much less likely to form during flame treatment than dur-
ing corona treatment. With coronas, LMWOM is easily detected at an O/C atomic
ratio as low as 0.06, while with flame treatment, no LMWOM is observed at an
O/C = 0.18. A high level of oxidation alone is clearly not sufficient to form water-
soluble LMWOM. Corona treatment must also cause a much greater reduction in
the molecular weight of the PP during oxidation than does flame treatment.
For polypropylene, it is widely accepted(29–32) that chain scission and the
lowering of the molecular weight of oxidized PP occur primarily through the
β-scission reaction:
---- C ---- C(O•) ---- C → ---- C ----
----O + •C ----
Polymer alkoxy radicals are, therefore, the principal precursor to LMWOM. There
are several routes to alkoxy-radical formation in PP oxidation mechanisms:
1. the direct formation of alkoxy radicals from the reaction of polymer alkyl
radicals with atomic oxygen:
R• + O → RO•
2. the reaction of polymer peroxy groups with atomic O:
ROO• + O → RO• + O2
3. the reaction of peroxy groups to form hydroperoxy groups, which subse-
quently decompose to form alkoxy groups:
ROO• + R0 H → ROOH + R0 •
ROOH → RO• + OH
4. and the reaction of ozone with alkyl radicals:(33)
R• + O3 → RO• + O2
Polymer peroxy or hydroperoxy groups can be formed by the reactions of O2 ,
HO2 , or H2 O2 with alkyl radicals:
R• + O2 → ROO•
R• + HO2 → ROOH
R• + H2 O2 → ROO• + H2 or ROH + OH
P1: GCR
However, Hansen et al.(34) state that, when atomic oxygen is present, the per-
oxy/hydroperoxy route to chain scission is unimportant relative to the reactions
involving atomic oxygen. The hydroperoxy route to scission is simply too slow to
account for the formation of LMWOM in the short exposure times that are typi-
cally encountered in industrial surface-oxidation processes.(5) Therefore, extensive
chain scission and the formation of LMWOM should be associated with reactions
involving atomic oxygen. As we discuss below, a major difference between the
gas-phase chemistry of corona discharges and that of flames is the relative amounts
of ozone, O atoms, and OH radicals. In a corona, ozone and O atoms are predom-
inant, while in a flame, the OH radical is the dominant oxygen-containing species
in the gas phase.
The influence of UV or vacuum UV radiation on polymer surface modifica-
tion by plasmas has been investigated by a number of researchers,(35,36) who have
concluded that the oxidation of polymer surfaces by O atoms is greatly enhanced
when the surface is concurrently illuminated by UV photons. Although UV illumi-
nation may play an important role in low pressure plasmas, the contribution of UV
radiation in corona treatment is likely to be small. The UV radiation is produced
only during the discharge pulse from excited states that have short lifetimes. The
duty cycle of illumination is therefore small so that the opportunities for syner-
gistic effects are also small as compared to continuously operated low-pressure
plasmas. Even with higher duty cycles, the contributions of UV radiation to surface
modification in corona discharges sustained at atmospheric pressure are likely to
be small because of the short mean-free-path for absorption of the radiation by
gas-phase species.
As noted above, computational modeling of flame treatment(5) showed that
the concentration of atomic oxygen at the PP surface was relatively low. The ratio
of O to OH at the PP surface was ca. 0.006. We then hypothesized that the lack of
detectable LMWOM on flame-treated PP was a result of this low concentration
of atomic oxygen in the flame. In flames, oxidation of the PP occurs through
the reactions of H, OH, HO2 , H2 O2 , and O2 , without the LMWOM formation
that is associated with extensive O-atom reactions. This explained the lack of
LMWOM formation, but did not explain the formation of the nodules observed
by AFM. During flame treatment, we believe that scission of the PP occurs only
to a limited extent through alkoxy radicals formed by the small numbers of O
atoms present in the flame coupled with the slower hydroperoxy route to scission
described above. This limited scission of the PP lowers the molecular weight of
the surface region, but does not quickly generate oxidized chains of sufficiently
low molecular weight to be water soluble. These non–water-soluble, intermediate-
molecular-weight products may be tethered to the unmodified PP chains by only
a few bonds, thereby enabling increased mobility relative to the bulk unoxidized
PP. This increased mobility enables the oxidation products to agglomerate to a
certain extent in a fashion analogous to the formation of the globular mounds of
P1: GCR
LMWOM observed by AFM on corona-treated PP. The net result of limited chain
scission and this modest agglomeration is the formation of the nodules observed
on flame-treated PP.
An alternate hypothesis for the lack of observable LMWOM on flame-treated
PP is that LMWOM is formed by flame oxidation, but is then quickly lost by
migration or evaporation. To determine the likelihood of this happening in the
flame, we exposed LMWOM-containing corona-treated PP (O/C = 0.12) to 150–
160◦ C for 10 s in a convection oven. AFM and ESCA examinations before and
after this heat treatment showed no change to the surface. This strongly implies
that, even when present in large quantities, LMWOM does not evaporate or migrate
away from the surface in less than a few seconds, even when heated to near the
bulk melting point of the PP. So, if the flame were to form any LMWOM, it should
not disappear from the surface in the <0.1 s that the film is exposed to the high
temperatures of the flame. We therefore conclude that LMWOM is simply not
formed to any great extent in flame treatments.
In contrast to a flame, O atoms are a dominant species in the gas phase
of a corona; OH radicals are much less common in coronas relative to flames.
As shown in Fig. 12 and Table III, in industrial coronas, O concentrations are
significantly higher than OH concentrations. Our computational results compare
favorably with other studies. Rosocha and Korzekwa(37) report that the ratio of the
concentration of O(3 P) atoms to the concentration of OH radicals in a “humid”
corona discharge is about 10:1. Gentile and Kushner(11) modeled a dielectric-
barrier (corona) discharge at “high RH,” finding that the ratio of <O>:<OH> was
about 8:1.
There is a significant concentration of N atoms in a corona discharge, yet
no nitrogen is detected by ESCA on the treated PP surface, even at the high
O/C ratios of 0.23. In their recent careful ESCA study, O’Hare et al.(20) also
found only oxidation occurring during the corona treatment of PP. Analysis of
corona-treated PP by static secondary ion mass spectrometry (SSIMS) detects
trace amounts of oxidized nitrogen species on the surface at negative ion masses
of 46 and 62 D. No other nitrogen-containing ion clusters are detected. These trace
amounts of oxidized nitrogen can be readily assigned to the reactions of the small
quantities of NO and NO2 in the corona with the PP.(8) Despite a relatively high
gas-phase concentration, N atoms do not appear to add to PP surfaces. The reaction
mechanism appears to be exclusively dominated by the oxygen-containing species
in the corona.
A possible explanation for this lack of affixed nitrogen is that atomic nitrogen
reacts with organic molecules primarily by abstraction or scission, rather than by
insertion.(38,39) Thus, reactions such as
R + N → R• + HCN
R• + N → R0 • + HCN
P1: GCR
may be most likely reactions of the N formed in a corona with PP. These types of
reactions form nitrogen-containing gas-phase products rather than affix nitrogen
to polymer surfaces.
Both Kill et al.(40) and Moss et al.(31) showed that the concentration of
O atoms is high in air or oxygen glow discharges. These plasma treatments
also form detectable, water-soluble LMWOM, as documented by a number of
researchers.(21,25,41–43) In UVO treating, oxygen atoms and molecules are the only
significant gas-phase reactive species.(44,45) As noted above, UVO treatment of PP
yields LMWOM and surface topographical structures nearly identical to corona
treatment. These observations strongly support the contention that O atoms are
primarily responsible for LMWOM formation.
Additionally, our previous study(44) showed that the oxidation of PP with
ozone only does not lead to the formation of detectable amounts of LMWOM. In
ozone treatments, there is little, if any, atomic oxygen. The key reaction in ozone
treatments is
RH + O3 → ROO• + OH
This reaction starts the hydroperoxy route to chain scission, which is too slow
to generate detectable LMWOM in surface-modification reactions. Our work on
ozone treatments also supports the contention that O-atom reactions are needed
for LMWOM formation.
Thus, in air-corona, UVO, and oxidizing-plasma treatments, atomic oxygen
leads to rapid chain scission. Widespread chain scission quickly leads to the for-
mation of water-soluble LMWOM. Obviously, if LMWOM is formed from PP,
there should also be scission products with intermediate molecular weights. As
noted above, these non-water-soluble scission fragments will be more mobile than
the bulk polymer. This increased mobility should enable the intermediate-weight
oxidation products to agglomerate to some extent, thereby forming the nodules that
are observed in conjunction with the globular LMWOM mounds on corona- and
UVO-treated PP. Thus, for all types of surface-oxidation processes on PP, it seems
likely that nodular topography is generated by the agglomeration and ordering of
non–water-soluble intermediate-molecular-weight scission products.
Because nodules form in low-temperature corona, UVO, and plasma treat-
ments, heating of the surface is not required for the formation of the nodules.
This supports our earlier contention(5) that the topographical changes observed
on flame-treated PP are the result of oxidation rather than of heating. The ef-
fects of relative humidity on the topography of corona-treated PP imply that water
vapor adsorbed on the PP surface may enhance the mobility of both LMWOM
and any intermediate-molecular-weight scission products. In flame treatment,
the moist, elevated-temperature environment of the flame should enhance the
mobility of the oxidized scission products, even though they are not soluble
in water.
P1: GCR
During the corona treatment of PP, increasing the RH in the discharge seems
to enhance the agglomeration of LMWOM into mounds, but actually decreases the
total amount of LMWOM formed. This behavior is readily explained by the fact
that the ratio of O to OH decreases as the amount of water vapor in the discharge
increases. As the concentration of O atoms declines, less scission occurs, so that
there is less formation of LMWOM. However, the LMWOM that is formed is more
likely to agglomerate because of the added mobility imparted by adsorbed water.
As we explained previously,(5) the nodules are not likely to be PP crystalline
lamella revealed by the oxidative “etching” of amorphous PP material. Any crys-
talline structure exposed by the removal of amorphous material should be smaller
than the fibrils observed on the untreated film. Instead, we observe a gradual in-
crease in the characteristic size of the features formed by surface oxidation. This
contention is strongly supported by the earlier work of Mahlberg et al.,(27) who
state that etching cannot be the explanation for the topographical changes observed
on oxygen-plasma-treated unoriented PP because the size of the surface features
increases with increasing extent of oxidation.
5.2. Treatment Stability

Decreases in the extent of oxidation or losses in wettability as a function of
aging are usually attributed to the reorientation or localized migration of oxidized
functional groups located near the surface of a modified PP. As the molecular
weight of the surface region decreases, the tendency for functional-group reorien-
tation or migration should increase. The molecular weight of a corona-treated PP
surface is lower than that of a flame-treated PP surface. Therefore, corona-treated
PP is less stable than flame-treated PP. The AFM images of washed corona-treated
PP (Fig. 6) suggest that the molecular weight of the washed surface may be lower
than that of the washed flame-treated PP (Fig. 2). Treatment stability is an exam-
ple of how the surface molecular weight can be more important than the extent of
oxidation in determining a practical performance property.
5.3. Contact-Angle Measurements

Surface roughness or surface chemical heterogeneity(9,46,47) causes thermo-
dynamic contact-angle hysteresis, such as that observed on the surface-oxidized
PP produced for this study. On chemically heterogeneous surfaces, the advancing
contact angle is more sensitive to the low energy, unmodified components of the
surface, while the receding angle is more sensitive to the high energy, oxidized
groups introduced by treatment. Therefore, the receding contact angle is actually
the measurement most characteristic of the modified component of a surface. To
best characterize any surface-oxidized polymer, both the advancing and receding
contact angles should be examined, with the receding angle actually being the
more important measurement.
P1: GCR
Our untreated PP exhibits considerable contact-angle hysteresis (22◦ ). We

believe that this hysteresis is not caused by chemical heterogeneity because the
film is produced from a homopolymer PP resin and careful ESCA or SSIMS do not
detect any additives at the film surface. Thus, the hysteresis must either be caused by
surface roughness or by domains of different polymer morphology (crystallinity)
on the PP surface or by some combination of the two. Past consensus has been
that surface heterogeneity needs to be of the order of 0.1–0.5 µm for roughness
and 0.1 µm for chemistry(9,46,47) to affect contact-angle measurements. Some
researchers, however, report a limit as low as 0.05 µm(47) for roughness-induced
hysteresis. More recently, Youngblood and McCarthy(48) showed that roughness
on the nanometer scale can contribute to contact-angle hysteresis. Our untreated
PP has considerable structure on the submicrometer scale, with individual fibrils
having dimensions of 10 nm or less. It is therefore probable that this topography
is the major contributor to contact-angle hysteresis on untreated PP films.
At low levels of flame-generated oxidation (O/C < ca. 0.07), there is no
AFM-detectable change in surface topography.(5) Nevertheless, there is a marked
increase in the extent of contact-angle hysteresis, from 22◦ for the untreated PP
to ca. 55◦ for flame-treated PP with an O/C = 0.07. Relative to the untreated
PP, the changes in the advancing and receding contact angles and the further
increase in the hysteresis must therefore be caused by heterogeneous oxidation
rather than by topographical changes. In semicrystalline polymers such as PP, it is
widely believed(49–52) that oxidation occurs mainly in the amorphous and interphase
regions. The amorphous regions of the PP are more accessible to the diffusion of
the gas-phase reactive species. Gaseous species will not be able to quickly diffuse
into the PP lamellae, so that those portions of the PP will be more resistant to
oxidation. This heterogeneity in the physical structure of semicrystalline polymers
leads to heterogeneity in the surface chemistry of those polymers. In addition,
further studies(49,53) suggest that inhomogeneity in the extent of oxidation may be
an inherent property of PP molecules. An inhomogeneous distribution of oxidized
functional groups can lead to a mixture of highly oxidized small molecules together
with relatively intact macromolecules. This type of surface structure would cause
the increased contact-angle hysteresis observed for flame-treated PP at low levels of
oxidation.
For all of the corona-treated PP that we studied, as well as for the more
highly oxidized flame-treated PP (O/C > ca. 0.10), surface topographical changes
accompany oxidation. On some of these surfaces, the contact-angle hysteresis
can approach 65◦ . For these films, the increase in the amount of hysteresis could
be caused by both chemical heterogeneity and by the topographical changes to
the surface. In the extreme case of LMWOM formation, the advancing angles on
unwashed surfaces are affected by the dissolution of the LMWOM into the water,
as evidenced by the major increase in the advancing angle on the washed materials
and the greater scatter (higher standard deviation) in the data.
P1: GCR
While microscopic chemical heterogeneity seems to be the most likely cause

for the contact-angle hysteresis on surface-oxidized polymers, there is an alter-
native explanation. Yasuda et al.(54,55) suggest that contact-angle hysteresis can
be associated with the rapid reorientation of hydrophilic functionalities. During
the measurement of the advancing contact angle, the water front encounters a
polymer–air interface where the concentration of hydrophilic groups is low; hence
the measured advancing angle is large. During the measurement of the receding
angle, the water front traverses a polymer-water interface where the concentration
of hydrophilic groups is high; hence the measured receding angle is low. Chan(47)
believes that this type of behavior can only occur on the surface of a polymer with
extremely high chain mobility, such as a hydrogel material. However, it is con-
ceivable that surface-modified polymers might have sufficient mobility to exhibit
similar behavior. Earlier, we described a series of contact-angle experiments that
seemingly shows that the hysteresis on our surface-modified PP films is thermody-
namic, rather than kinetic. That data would argue against the type of reorientation
proposed by Yasuda et al. However, if the reorientation were nearly instantaneous,
it might be difficult to experimentally detect it. Thus, it is possible that an ex-
tremely rapid reorientation of oxidized scission products could account for some
of the hysteresis that we observe.
6. CONCLUSIONS
Although both corona and flame treatments of polypropylene are surface-
oxidation processes, the two methods differ considerably in the mechanism of
oxidation. Corona discharges are dominated by O-atom reactions that lead to scis-
sion of the PP and to the formation of water-soluble LMWOM. The LMWOM
agglomerates into mounds that are plainly visible in AFM images. The extensive
scission associated with corona processes affects the wettability and stability of
the treated PP films. By contrast, flame surface oxidations are dominated by OH-
radical reactions that lead to oxidation without LMWOM formation. The more
limited chain scission associated with flame treatment leads to the formation of
topographic nodules of intermediate-molecular-weight materials that are much
smaller than the mounds of LMWOM. Because of the lack of LMWOM forma-
tion, flame-treated PP is more highly oxidized after washing with water, more
wettable, and more stable than corona-treated PP.
The effect of increasing relative humidity in a corona is to decrease the O/OH
ratio in the gas phase. This leads to less formation of LMWOM at higher RH.
However, at higher humidities, any LMWOM that is formed is more likely to
agglomerate because of the enhanced mobility imparted by adsorbed water vapor.
Untreated, corona-treated, and flame-treated PP all exhibit considerable ther-
modynamic contact-angle hysteresis. While topographic features are the likely
cause of the hysteresis on untreated PP, microscopic chemical heterogeneity is the
P1: GCR
major contributor to the hysteresis on the surface-oxidized PP. The importance

of using AFM to assess surface topography is quite obvious. Without AFM im-
ages, it is impossible to determine if there is a possible effect of roughness on the
contact-angle measurements of oriented polymers.
ACKNOWLEDGMENTS
The authors thank 3M employees Rebecka Wheaton and Joel Getschel for
their assistance in obtaining the contact-angle data, Steve Pachuta for his SSIMS
analysis, and Mike Prokosch for his ESCA work.
REFERENCES
1. M. Strobel, M. J. Walzak, J. M. Hill, A. Lin, E. Karbashewski, and C. S. Lyons, J. Adhes. Sci.
Technol. 9, 365 (1995).
2. M. Strobel, C. Dunatov, J. M. Strobel, C. S. Lyons, S. J. Perron, and M. C. Morgen, J. Adhes. Sci.
Technol. 3, 321 (1989).
3. J. M. Strobel, M. Strobel, C. S. Lyons, C. Dunatov, and S. J. Perron, J. Adhes. Sci. Technol. 5, 119
(1991).
4. M. Strobel, C. S. Lyons, J. M. Strobel, and R. S. Kapaun, J. Adhes. Sci. and Technol. 6, 429 (1992).
5. M. Strobel, N. Sullivan, M. C. Branch, V. Jones, J. Park, M. Ulsh, J. M. Strobel, and C. S. Lyons,
J. Adhes. Sci. Technol. 15, 1 (2001).
6. T. C. Manley, Trans. Electrochem. Soc. 84, 83 (1943).
7. M. C. Branch, N. Sullivan, M. Ulsh, and M. Strobel, in Twenty-Seventh Symposium (International)
on Combustion, The Combustion Institute, Pittsburgh, PA (1998), pp. 2807.
8. M. Strobel, N. Sullivan, M. C. Branch, J. Park, M. Ulsh, R. S. Kapaun, and B. Leys, J. Adhes. Sci.
Technol. 14, 1243 (2000).
9. F. Garbassi, M. Morra, and E. Occhiello, Polymer Surfaces, Wiley, Chichester, England (1994),
Ch. 4.
10. NIST Chemical Kinetics Database, version 2Q98 (1998).
11. A. C. Gentile and M. J. Kushner, J. Appl. Phys. 78, 2074 (1995).
12. M. B. Chang, J. H. Balbach, M. J. Rood, and M. J. Kushner, J. Appl. Phys. 69, 4409 (1991).
13. R. Dorai and M. J. Kushner, J. Phys. D: Appl. Phys. 34, 574 (2001).
14. P. N. Brown, A. C. Hindmarsh, and G. D. Byrne, Variable-Coefficient Ordinary Differential Equa-
tion Solver, Lawrence Livermore Laboratories, Livermore, CA (1989).
15. R. M. Overney, R. Lüthi, H. Haefke, J. Frommer, E. Meyer, H.-J. Güntherodt, S. Hild, and J.
Fuhrmann, Appl. Surf. Sci. 64, 197 (1993).
16. G. J. Vancso, T. D. Allston, I. Chun, L.-S. Johansson, G. Liu, and P. F. Smith, Int. J. Polym. Anal.
Characteriz. 3, 89 (1996).
17. S. Süzer, A. Argun, O. Vatansever, and O. Aral, J. Appl. Polym. Sci. 74, 1846 (1999).
18. H.-Y. Nie, M. J. Walzak, B. Berno, and N. S. McIntyre, Appl. Surf. Sci. 144/145, 627 (1999).
19. H.-Y. Nie, M. J. Walzak, and N. S. McIntyre, Polymer 41, 2213 (2000).
20. L.-A. O’Hare, S. Leadley, and B. Parbhoo, Surf. Interface Anal. 33, 335 (2002).
21. S. Guimond, I. Radu, G. Czeremuszkin, D. J. Carlsson, and M. R. Wertheimer, Plasmas Polym. 7,
71 (2002).
22. O. D. Greenwood, R. D. Boyd, J. Hopkins, and J. P. S. Badyal, J. Adhes. Sci. Technol. 9, 311
(1995).
P1: GCR
23. R. D. Boyd, A. M. Kenwright, J. P. S. Badyal, and D. Briggs, Macromolecules 30, 5429 (1997).
24. G. J. Vancso, T. D. Allston, I. Chun, L.-S. Johansson, G. Liu, and P. F. Smith, Int. J. Polym. Anal.
Characteriz. 3, 89 (1996).
25. D. Briggs, C. R. Kendall, A. R. Blythe, and A. B. Wootton, Polymer 24, 47 (1983).
26. C. Y. Kim and D. A. I. Goring, J. Appl. Polym. Sci. 15, 1357 (1971).
27. R. Mahlberg, H. E.-M. Niemi, F. S. Denes, and R. M. Rowell, Langmuir 15, 2985 (1999).
28. H. F. Mark (executive ed.), Encyclopedia of Chemical Technology, 3rd ed., Vol. 16, Wiley-
Interscience, New York (1981), p. 689.
29. J. F. Rabek, Polymer Photodegradation Mechanisms and Experimental Methods, Chapman and
Hall, London (1995), Chs. 2, 3, 6, and 7.
30. D. J. Carlsson and D. M. Wiles, J. Macromol. Sci., – Rev. Macromol. Chem. C14, 65 (1976).
31. S. J. Moss, A. M. Jolly, and B. J. Tighe, Plasma Chem. Plasma Process. 6, 401 (1986).
32. F. D. Egitto, V. Vukanovic, and G. N. Taylor, Plasma Etching of Organic Polymers, in Plasma
Deposition, Treatment, and Etching of Polymers, R. d’Agostino, ed., Academic Press, San Diego,
CA (1990), p. 339.
33. R. Paltenghi, E. A. Ogryzlo, and K. D. Bates, J. Phys. Chem. 88, 2595 (1984).
34. R. H. Hansen, J. V. Pascale, R. De Benedictis, and P. M. Rentzepis, J. Polym. Sci. A 3, 2205 (1965).
35. A. Holländer, J. E. Klemberg-Sapieha, and M. R. Wertheimer, Macromolecules 27, 2893 (1994).
36. A. C. Fozza, J. E. Klemberg-Sapieha, and M. R. Wertheimer, Plasmas Polym. 4, 183 (1999).
37. L. A. Rosocha and R. A. Korzekwa, J. Adv. Oxid. Technol. 4, 247 (1999).
38. O. K. Fomin, Russ. Chem. Rev. 36, 725 (1967).
39. R. Brown and C. A. Winkler, Angew. Chem. Int. Ed. 9, 181 (1970).
40. G. Kill, D. H. Hunter, and N. S. McIntyre, J. Polym. Sci. A: Polym. Chem. 34, 2299 (1996).
41. E. M. Liston, L. Martinu, and M. R. Wertheimer, J. Adhes. Sci. Technol. 7, 1091 (1993).
42. R. Foerch, G. Kill, and M. J. Walzak, J. Adhes. Sci. Technol. 7, 1077 (1993).
43. G. Z. Xio, J. Adhes. Sci. Technol. 11, 655 (1997).
44. M. J. Walzak, S. Flynn, R. Foerch, J. M. Hill, E. Karbashewski, A. Lin, and M. Strobel, J. Adhes.
Sci. Technol. 9, 1229 (1995).
45. L. F. Macmanus, M. J. Walzak, and N. S. McIntyre, J. Polym. Sci. A: Polym. Chem. 37, 2489
(1999).
46. M. Morra, E. Occhiello, and F. Garbassi, Adv. Colloid Interface Sci. 32, 79 (1990).
47. C.-M. Chan, Polymer Surface Modification and Characterization, Hanser/Gardner, Cincinnati, OH
(1994), Ch. 2.
48. J. P. Youngblood and T. J. McCarthy, Macromolecules 32, 6800 (1999).
49. M. Iring and F. Tudos, Prog. Polym. Sci. 15, 217 (1990).
50. A. Holmström and E. Sörvik, J. Polym. Sci., Polym. Chem. Ed. 16, 2555 (1978).
51. D. J. Carlsson and S. Chmela, Polymers and High-Energy Irradiation: Degradation and Stabiliza-
tion, in Mechanisms of Polymer Degradation and Stabilization, G. Scott, ed., Elsevier Applied
Science, London (1990), Ch. 4, p. 113.
52. N. S. Allen and M. Edge, Fundamentals of Polymer Degradation and Stabilisation, Elsevier
Applied Science, London (1992), p. 14.
53. A. Holländer, J. E. Klemberg-Sapieha, and M. R. Wertheimer, J. Polym. Sci. A: Polym. Chem. 33,
2013 (1995).
54. H. Yasuda, A. K. Sharma, and T. Yasuda, J. Polym. Sci., Polym. Phys. Ed. 19, 1285 (1981).
55. C. W. Weikart and H. K. Yasuda, J. Polym. Sci. A: Polym. Chem. 38, 3028 (2000).
View publication stats

A_Comparison_of_Corona-Treated_and_Flame-Treated_P (1)(1)

Uploaded by

Copyright:

Available Formats

A_Comparison_of_Corona-Treated_and_Flame-Treated_P (1)(1)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A_Comparison_of_Corona-Treated_and_Flame-Treated_P (1)(1)

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

A Comparison of Corona-Treated and Flame-Treated

Article in Plasmas and Polymers · April 2003

Michael Ulsh Melvyn Branch

SEE PROFILE SEE PROFILE

The user has requested enhancement of the downloaded file.

Plasmas and Polymers, Vol. 8, No. 1, March 2003 (°

A Comparison of Corona-Treated and Flame-Treated

Received October 1, 2002; accepted October 29, 2002

The comparison of corona-treated and flame-treated polypropylene (PP) films provides

University of Colorado, Boulder, Colorado.

Paul, Minnesota 55144-1000; e-mail: [email protected].

62 Strobel, Jones, Lyons, Ulsh, Kushner, Dorai, and Branch

Comparison of Corona-Treated and Flame-Treated Polypropylene 63

material on at least one of the discharge electrodes, industrial corona discharges

64 Strobel, Jones, Lyons, Ulsh, Kushner, Dorai, and Branch

Flame treaters ordinarily employ laminar-premixed flames in which the fuel

2.2. Surface Analysis

Comparison of Corona-Treated and Flame-Treated Polypropylene 65

66 Strobel, Jones, Lyons, Ulsh, Kushner, Dorai, and Branch

to distinguish kinetic hysteresis, which is time or frequency dependent, from ther-

Comparison of Corona-Treated and Flame-Treated Polypropylene 67

Fig. 1. Gas-phase reaction mechanism for air-corona discharges.

2.3. Gas-Phase Corona Reaction Mechanisms

68 Strobel, Jones, Lyons, Ulsh, Kushner, Dorai, and Branch

Comparison of Corona-Treated and Flame-Treated Polypropylene 69

Hydrogen peroxide (H2 O2 ) is formed by the reaction of HO2 with HO2 :

2.4. Corona Computational Model

70 Strobel, Jones, Lyons, Ulsh, Kushner, Dorai, and Branch

Comparison of Corona-Treated and Flame-Treated Polypropylene 71

72 Strobel, Jones, Lyons, Ulsh, Kushner, Dorai, and Branch

Fig. 2. Tapping-mode AFM topographic image of an untreated biaxially oriented polypropylene

As discussed in detail in our earlier report on flame-treated PP,(5) the formation

Comparison of Corona-Treated and Flame-Treated Polypropylene 73

Fig. 3. Tapping-mode AFM topographic image of a flame-treated polypropylene film having an

74 Strobel, Jones, Lyons, Ulsh, Kushner, Dorai, and Branch

Comparison of Corona-Treated and Flame-Treated Polypropylene 75

Our earlier studies of corona-treated PP(2) using scanning electron microscopy

76 Strobel, Jones, Lyons, Ulsh, Kushner, Dorai, and Branch

Fig. 6. Characteristic tapping-mode AFM topographic image of a corona-treated polypropylene film

Comparison of Corona-Treated and Flame-Treated Polypropylene 77

et al.(20) , Greenwood et al.,(22) and Boyd et al.(23) observed large, micron-sized

3.2. Treatment Stability

78 Strobel, Jones, Lyons, Ulsh, Kushner, Dorai, and Branch

Comparison of Corona-Treated and Flame-Treated Polypropylene 79

Comparison of Corona-Treated and Flame-Treated Polypropylene 81

4.1. Effect of Corona Energy

82 Strobel, Jones, Lyons, Ulsh, Kushner, Dorai, and Branch

Fig. 9. Periodic steady-state (PSS) densities of electrons

Comparison of Corona-Treated and Flame-Treated Polypropylene 83

Fig. 10. Effect of corona energy on the time-averaged concen-

84 Strobel, Jones, Lyons, Ulsh, Kushner, Dorai, and Branch

4.2. Effect of Relative Humidity

Comparison of Corona-Treated and Flame-Treated Polypropylene 85

Fig. 11. Time-averaged concentrations of O3 , H2 O2 , and

4.3. Comparison of Corona Gas-Phase Chemistry to Flame

86 Strobel, Jones, Lyons, Ulsh, Kushner, Dorai, and Branch

Fig. 12. Variation of <O>/<OH> and <O3 > as a

Table III. Concentrations and Ratios of Some of the Important

Comparison of Corona-Treated and Flame-Treated Polypropylene 87