Adsorbtion of polyelectrolyte multilayers on plasma-modified

porous polyethylene

George Greenea,b,c, Rina Tannenbauma,*

aSchool of Materials Science and Engineering, Georgia Institute of Technology, Atlanta, GA 30332, USAbMaterials Department, University of California at Santa Barbara, Santa Barbara, CA 93106, USA

cDepartment of Chemical Engineering, University of California at Santa Barbara, Santa Barbara, CA 93106, USA

Available online 11 September 2004

Abstract

Hydrophilic and chemically reactive porous media was prepared by adsorbing functional polymers at the surface of sintered

polyethylene membranes. Modification of the membrane was accomplished by first exposing the membrane to an oxygen glow

discharge gas plasma to render electrostatic charge at the membrane surfaces. A cationic polyelectrolyte was adsorbed from

solution to the anionically charged surface to form an adsorbed monolayer. The adsorption of a second anionic polyelectrolyte

allowed further modification of the membrane surface with a polyelectrolyte bilayer complex. In this paper we probe the effect of

polymer structure on the conformation and stability of the adsorbed polyelectrolyte monolayers and bilayers on the modified

polyethylene surface. Using the wicking rate of deionized, distilled water through the porous membrane to gauge the interfacial

energy of the modified surface, we show that the wicking rate of the multilayer membrane can be controlled by varying the

chemistry of the adsorbing polyelectrolytes and their molecular weights.

# 2004 Elsevier B.V. All rights reserved.

Keywords: Polyelectrolyte; Polyethylene; Molecular weight

1. Introduction

The ability to modify the chemistry of polyethylene

surfaces by immobilizing a self-assembled polyelec-

trolyte multilayer complex provides a straightforward

mechanism for controlling surface chemistry and

therefore the properties of the modified surface. Tra-

ditional polymer surface modification techniques con-

sisting of reactive gas plasmas, corona discharge, and

electron beam, have generated modified surfaces with

intrinsic instability and poor uniformity of surface

properties. Therefore, most porous membrane manu-

factures have subscribed to the belief that the only

practical method to control the wetting properties and

migration rates of fluid through a membrane is to alter

its porosity, i.e. control and manipulate the pore size

[1,2]. However, as evidenced by the Lucas–Washburn

equation, the migration rate is fourth order with

respect to pore radius while only first order with

respect to surface energy [3,4]:

V

t¼ pr4

8Zrg þ 2gðlvÞ cos y

rL

� �(1)

Hence, minor variations in pore size both within and

between membranes can have significant impact upon

migration rate. Thus, controlling migration rate by

Applied Surface Science 238 (2004) 101–107

* Corresponding author. Tel.: þ1 404 3851235;

fax: þ1 404 8949140.

E-mail address: rina.tannenbaum@mse.gatech.edu

(R. Tannenbaum).

0169-4332/$ – see front matter # 2004 Elsevier B.V. All rights reserved.

doi:10.1016/j.apsusc.2004.05.190

manipulation of the membrane pore structure is

often a challenging undertaking that requires precise

control of every variable of the manufacturing

process. As the casting processes employed in the

manufacturing of the majority of synthetic porous

polymer membranes are, by nature, difficult to control,

considerable variability in the wicking rate is typically

observed not only between different production lots

but also between different regions within a single

membrane.

Greater consistency and uniformity of wetting prop-

erties can be achieved by keeping the physical and

geometric structure of the membrane constant while

manipulating the chemistry of the surface to achieve

different migration rates. Since fluid wicking rate is

only first order with respect to surface energy, migra-

tion rate consistency and uniformity of the membrane

are not compromised by minor variations in the che-

mical composition of the surface [5,6]. The method for

the modification of the surface properties of porous

polymer membranes has to encompass not only the

variations in the chemical composition of the surface,

but also has to demonstrate a far greater stability that

the conventional surface activation methods have

exhibited so far. The stability of surface properties

of these modified membranes is a fundamental

requirement for their potential use in a variety of

applications involving lateral flow and binding media

for bioassays. Hence, activation methods such as

plasma activation had to be coupled with additional

surface manipulations in order to achieve the desired

functionality and desired stability [7–12].

The unique and powerful electrostatic interactions

between oppositely charged polyelectrolytes may be

utilized to construct membrane-like, alternating poly-

electrolyte layers [13–26]. Modifying the surface of

sintered porous polymer membranes, having a rela-

tively uniform pore size, with a self-assembled poly-

electrolyte bilayer, through sequential adsorption of

cationic and anionic polyelectrolytes, promises a

greater versatility and stability in the control of the

modified surface properties, as compared to any other

traditional stand-alone surface activation method [13–

17,19–26]. The high charge density of these polyelec-

trolytes, combined with the large macromolecular

structure of the polymer chains, generate a massive

counter ion, enabling these polymers to rapidly adsorb

to even weakly charged surfaces [16–26]. The macro-

molecular structure gives adsorbing polyelectrolyte

chains the ability to form a multiplicity of bonds with

a surface, generating highly stable monolayers even

on surfaces with diffuse concentrations of charge

centers and weak electrostatic charge [16,17,19–26].

Charge overcompensation by adsorbing polyelectro-

lytes, which occurs when more polyelectrolyte mole-

cules adsorb than is required to satisfy the electrostatic

charge of the surface, is often observed when the

adsorbing polyelectrolyte possesses a higher density

of charge than the substrate. This process has the

potential to greatly amplify the electrostatic potential

and charge density of weakly charged surfaces [16–

26]. The adsorption dynamics and structure of the

adsorbed polyelectrolyte complex are governed by

electrostatic effects and are thus sensitive to the charge

density and polarity of both the surface and the

adsorbing polyelectrolytes [27–32].

In this paper we describe the construction of a

surface-active porous polyethylene membrane that

has been modified by plasma treatment followed

by the adsorption of a monolayer of positively

charged polyethylenimine (PEI) and the subsequent

adsorption of a variety of negatively charged poly-

electrolytes. The surface wetting characteristics were

tested by the measurement of the migration rates of

deionized water over fixed distances in the mem-

branes. The molecular weights of the outer layer

polyelectrolytes were also varied in order to probe

the effect of charge overcompensation on the wetting

properties of the surface.

2. Experimental

Polyethylenimine (PEI) was obtained by BASF

Corporation (various �Mn), poly(acrylic acid) and addi-

tional polyelectrolytes were purchased from Sigma–

Aldrich (various �Mn). Porous polyethylene (PPE)

membranes (UHMWPE, 40.0 mm thickness), having

40–45% pore volume fractions with an average pore

diameter of 7 mm were obtained from the Porex

Corporation (Fairburn, GA).

PPE membranes were activated by gas plasma at a

chamber pressure of 120 mTorr of argon, an Rf

power density of 200 W/m2, and an exposure time

of 1 min. Subsequently the membranes were exposed

for 15 min to a reactive oxygen plasma generated at a

102 G. Greene, R. Tannenbaum / Applied Surface Science 238 (2004) 101–107

chamber pressure of 120 mTorr and an Rf power

density of 200 W/m2 in order to chemically modify

the properties of the surface. Details of the procedure

are described elsewhere [8].

Immediately following plasma exposure, the mem-

branes were immersed in a 0.1% PEI/ethanol solution

at room temperature for 10 min. After solution treat-

ment, the membranes were washed thoroughly with

deionized water before being submerged in a deio-

nized water ultrasonic bath for 2 min to expel any

retained alcohol solution from the pores and remove

any physically bound but unadsorbed PEI from the

surface. The treated membranes were then dried in air

overnight before testing.

In order to further modify the polymer surface, the

dry PEI modified membranes were submerged in a

0.1% PAA/ethanol solution at room temperature for

10 min. After solution treatment, the membranes were

washed and dried similarly to the protocol described

earlier. Similar experiments were conducted with

various PEI and PAA molecular weight polyelectro-

lytes, and with other polyelectrolytes with various

chemistries.

Wicking properties were used as a basis for com-

paring the extent of chemical modification in PPE

membranes exposed to reactive oxygen plasmas.

All wicking measurements of PPE media were con-

ducted immediately after removal from the plasma

chamber. Wicking measurements were conducted

with deionized water at room temperature (27 8C)

on 1 cm � 5 cm samples cut from the plasma treated

membranes. Deionized water was chosen as the

wicking fluid due to its inability to wet unmodified

PPE, supporting the assumption that the observed

wicking properties are the sole product of chemical

modification and that variations in the wicking

properties are a reflection of the extent of chemical

modification [3,6]. Testing of the wicking properties

was carried out on PPE samples mounted vertically

to a stationary bar held above a shallow dish of

deionized water. Using an adjustable platform, the

dish of deionized water was raised to a height

where the surface of the water makes initial contact

with the end of the suspended sample. A stopwatch

was used to measure the time required for the water to

wick through the porous sample to a height of 4 cm

after the initial contact between the sheet and the

water.

3. Results and discussion

3.1. Variation of polyelectrolyte chemistry

The plasma activation of PPE membranes followed

by adsorption of a monolayer of the positively charged

polyelectrolyte PEI was ascertained via XPS analysis,

as shown in Fig. 1. The appearance of the N 1s core

electron peak demonstrates the presence of the various

amine functionalities on the polyethylene surface. In

order to demonstrate the principle that wetting proper-

ties may best be modified by the control of the

chemistry of the outer polyelectrolyte chemistry, a

variety of anionic polyelectrolytes were adsorbed to

form a bilayer on the PPE membrane. Table 1 lists the

chemical names and acronyms of the anionic poly-

Fig. 1. High resolution X-ray photoemission spectra of N 1s core

electrons of polyethylene films after plasma activation and

adsorption of a PEI monolayer on the activated polymer.

Table 1

Chemical formulae and acronyms of the various polyelectrolytes

Polyelectrolyte name Acronym

Unmodified polyethylene Untreated

Polyethylenimine PEI

Polyacrylic acid PAA

Poly(styrenesulfonic acid-co-malaic acid) PSSA-co-MA

Poly(methyl vinyl ether-alt-maleic acid) PMVE-alt-MA

Poly(styrene-alt-maleic acid) PS-alt-MA

Poly(sodium-4-styrenesulfonate) PS-4-SS

Poly(vinylalcohol-co-vinylacetate-co-

itaconic acid)

PVA-co-VA-co-IA

Poly(vinylsulfonic acid) PVSA

Poly(methyl vinyl ether-alt-maleic acid

mono-ethyl ester)

PMVE-alt-MAME

Poly(styrene-co-maleic acid) partial

isobutyl/methyl mixed ester

PS-co-MA (mixed)

G. Greene, R. Tannenbaum / Applied Surface Science 238 (2004) 101–107 103

electrolytes used in this experiment, and Table 2 lists

their characteristic chemical functionalities, and their

molecular weights.

The average wicking rate of deionized water mea-

sured in the various modified membranes is shown in

Fig. 2. By simply varying the properties of the termi-

nating polyelectrolyte layer it is possible to achieve

modified surfaces displaying a wide range of wetting

properties without altering the physical structure and

geometry of the porous matrix. More importantly, the

data in Fig. 2 refutes the popular belief of the depen-

dency of migration rate solely on pore size. The

wetting properties and stability properties of a bilayer

consisting of polyethylenimine and polyacrylic acid

(PEI–PAA) has been thoroughly investigated [33,34],

and hence, the migration rate of such a membrane was

used as a reference. As it is clear from the data, the

migration rates can be either increased or decreased by

substituting an alternative anionic polyelectrolyte for

PAA. The versatility of this approach for the control of

the migration rates is evident by the breadth of varia-

tion in rates from 0.134 cm/s measured for the PMVE-

alt-MA layer to 0.0 cm/s (non-wetting) measured for

PS-co-MA (mixed) layer.

Such a wide range of migration rates are made

possible by the unique sandwich construction of the

polyelectrolyte bilayer in which the chemical compo-

sition of the modified surface are nearly identical to

that of terminating polyelectrolyte layer. Because the

modified surface chemistry is a reflection of the

terminating polyelectrolyte chemistry [19–26], con-

trolling such properties of the modified surface as

wettability, reactivity, and functional group density

merely involves choosing a polyelectrolyte with the

appropriate chemical composition. In this manner, the

chemistry of membrane surfaces can be modified with

an assortment of types and variety of functional

groups. In this experiment, adsorbing polyelectrolytes

having different chemistries enabled the creation of

modified membrane surfaces with a variety of chemi-

cal functionalities that include amino, carboxyl,

hydroxyl, maleic acid, sulfonic acid, sulfonate, and

itaconic acid. Likewise, different wetting properties

were achieved by varying the lengths, types, polariza-

tion energies, and concentrations of hydrophobic and

hydrophilic segments forming the molecular chains of

the terminating polyelectrolyte. Varying the composi-

tion and properties of hydrophobic segments in the

polyelectrolyte chain allows membrane surfaces to be

modified in order to achieve both desired chemical

functionality and wetting property.

Table 2

Summary of polyelectrolyte functionalities and molecular weights

Polymer Primary functionalities �Mn

Untreated None

PEI Primary, secondary,

tertiary, amine

750000

PAA Carboxyl 250000

PSSA-co-MA Sulfonic acid, maleic acid 20000

PMVE-alt-MA Maleic acid 1980000

PS-alt-MA Maleic acid 120000

PS-4-SS Sulfonate 1000000

PVA-co-VA-co-IA Hydroxyl, itaconic acid NA

PVSA Sulfonic acid NA

PMVE-alt-MAME Maleic acid, ester NA

PS-co-MA (mixed) Maleic acid, ester 225000

Fig. 2. Migration rates of DI water in membranes with various capping polymers adsorbed on the PEI monolayer.

104 G. Greene, R. Tannenbaum / Applied Surface Science 238 (2004) 101–107

To illustrate this concept, consider the surfaces that

were modified with maleic acid functionalized poly-

electrolytes and the impact of different non-functional

chain segments on the observed average migration

rates. Schematic description of these three different

co-polymers is shown in Fig. 3. In the case of PMVE-

alt-MA, the non-functional methyl vinyl ether chain

segments are, themselves, fairly hydrophilic and con-

sequently the observed migration rate of the modified

membrane is extremely high at 0.134 cm/s. The pre-

sence of chain segments composed of bulky and

highly non-polar styrene monomers as found in

PSSA-co-MA and PS-alt-MA greatly slows the mea-

sured average migration rate respectively to 0.082 and

0.059 cm/s. The addition of esters into the polyelec-

trolyte chains reduces the migration rate further. In the

case of PMVE-alt-MAME where the ester is a mono-

ethyl ester, the measured migration rate is 0.026 cm/s.

In contrast, in PS-co-MA (mixed ester) where styrene

and ester groups are present, the modified membrane

surface is unable to be wetted by deionized water.

3.2. Variation of polyelectrolyte molecular weight

The molecular weight of the capping polymer

adsorbed on the PEI monolayer is an important vari-

able as well. The anionic polyelectrolyte interacts with

the cationic centers on the PEI, and conform itself on

the PEI surface to maximize the exposure of its

anionic center, generating charge overcompensation.

Clearly, as the molecular weight of the capping poly-

mer is larger, it is expected that the charge concentra-

tion on the surface will be larger and hence the wetting

properties will be enhanced. Fig. 4 shows the wetting

properties of several experiments conducted with

three different molecular weights of PEI, and of

experiments conducted with three different molecular

weights of PAA adsorbed on PEI with molecular

weight of 750,000 g/mol. In the case of the adsorption

of PEI onto the plasma-activated porous polyethylene,

the wetting properties scale linearly with the polymer

molecular weight. However, the enhancement of these

properties becomes less pronounced at very high

molecular weights probably due to the very high

viscosity of the polymers and their inability to undergo

the appropriate conformational changes for maximal

surface exposure. The same general trend can also be

observed with the PAA capping layer. In this case,

however, the higher molecular weight actually results

in a decrease in the wetting properties of the poly-

electrolyte bilayer. While the increased viscosity and

the resulting slower dynamics of higher molecular

weight polymers may play a crucial role in the

observed reduction of surface wetting properties, it

Fig. 3. Chemical structures of some selected co-polymers having

maleic acid as one of the building blocks.

Fig. 4. Dependence of the wetting properties of the PEI monolayer

on the polymer molecular weight, and of the PEI–PAA bilayer

(with constant �Mn of PEI of 750,000 g/mol) on the molecular

weight of PAA.

G. Greene, R. Tannenbaum / Applied Surface Science 238 (2004) 101–107 105

is important to take into account also how the adsorp-

tion of such large molecules impacts the physical

geometry of the pore. A bilayer made from a �Mn ¼750,000 g/mol PEI and a �Mn ¼ 750,000 g/mol PAA

should be very thick when fully hydrated. A bilayer

only 500 nm thick would shrink the size of a 10 mm

circular pore by 1 mm. Most likely, the observations

are a combination of the two factors [35,36].

The dependence of surface properties on the mole-

cular weight of the capping polyelectrolytes is also

evident for the polymers with various functionalities.

For example, the difference in the wicking rates of

deionized water between PMVE-alt-MA and PSSA-

co-MA may be attributed not only to the difference in

the chemistry of the polymer backbone groups, but

also to the very large difference in the molecular

weight of the two polyelectrolytes. As seen in

Table 2, the molecular weight of PMVE-alt-MA used

here was 1,980,000 g/mol as compared with 20,000 g/

mol for PSSA-co-MA. As expected, based on the

behavior of the PEI–PAA bilayer, the higher mole-

cular weight polyelectrolyte, having a high charge

density, exhibits higher surface activity. On the other

hand, molecular weight is only one of the variables

that determine the surface properties of the polyelec-

trolyte bilayer. The molecular weight of PS-alt-MA

was 120,000 g/mol, but the wicking rates on the sur-

face, when compared to those of PSSA-co-MA, are

lower, despite the larger polymer size. Hence, the

wetting properties of the various surfaces are depen-

dent on both the polarity and ionizability of the

polyelectrolyte core segments and on the charge den-

sity of the ionized centers, which, in turn, is dependent

on the molecular weight of the polyelectrolytes.

4. Conclusions

The wetting properties of polyelectrolyte multi-

layers adsorbed on plasma-activated porous polyethy-

lene membranes depend on the ionizability of the

reactive groups on the polyelectrolyte and the polarity

of the core segments of the polymers. Moreover, the

distribution of charge density throughout the polymer

chains, specifically in the outer-layer polyelectrolyte,

has a strong influence on the overall surface properties

of the membrane. The changes in the surface proper-

ties of the monolayers and bilayers as a function of an

increase in the molecular weight of the polyelectro-

lytes, is a combination of both an increased viscosity

and reduced dynamic capabilities of the polymers, and

an increase in the geometric restrictions on the poly-

ethylene pores themselves.

Acknowledgements

This research was supported by the Porex Corpora-

tion in Fairburn, Georgia. We would like to give

special thanks to Drs. George Yao and Gary Mao

from Porex and to Dr. Brent Carter form the School

of Materials Science and Engineering at Georgia Tech,

for their valuable discussions and advice.

References

[1] A. Della Martina, L. Caramszegi, J.G. Hilborn, J. Polym. Sci.,

Part A: Polym. Chem. 41 (2003) 2036.

[2] S.T. Wereley, A. Akonur, R.M. Lueptow, J. Membr. Sci. 209

(2002) 469.

[3] B.V. Zhmud, F. Tiberg, K. Hallstensson, J. Coll. Interf. Sci.

228 (2000) 263.

[4] E. Chilbowsky, R. Perea-Carpio, Adv. Coll. Surf. Sci. 98

(2002) 245.

[5] M. Ohring, The Material Science of Thin Films, Academic

Press, New York, NY, 1992, pp. 51–53.

[6] C.M. Chan, Polymer Surface Modification and Characteriza-

tion, Hanser/Gardner Publications, Inc., Cincinnati, OH,

1994, pp. 45–49.

[7] H. Wheale, C.P. Barker, J.P.S. Badyal, Langmuir 14 (1998)

6699.

[8] G. Greene, G. Yao, R. Tannenbaum, Langmuir 19 (2003) 5869.

[9] J.G.A. Terlingen, G.A.J. Takens, F.J. Van Der Gaag, A.S.

Hoffman, J. Feijen, J. Appl. Polym. Sci. 52 (1994) 39.

[10] J. Behnisch, A. Hollaender, H. Zimmermann, Surf. Coat.

Technol. 59 (1993) 356.

[11] D. Andelman, J.F. Joanny, Polym. Interf. 4 (2000) 1153.

[12] Y. Shin, J.E. Roberts, M.M. Santore, J. Coll. Interf. Sci. 247

(2000) 220.

[13] T.G. Vargo, J.M. Calvert, K.J. Wynne, J.K. Avlyanov, A.G.

MacDiarmid, M.F. Rubner, Supramol. Sci. 2 (1996) 169.

[14] H. Fukumoto, Y. Yonezawa, Thin Solid Films 327–329 (1998)

748.

[15] S. Chibowski, M. Wisniewska, Coll. Surf. A 208 (2002) 131.

[16] R. Bijlsma, A.A. van Well, M.A. Cohen Stuart, Physica B

234–236 (1997) 254.

[17] M. Muller, M. Brissova, T. Rieser, A.C. Powers, K. Lunkwitz,

Mater. Sci. Eng. C 8/9 (1999) 163.

[18] P. Schuetz, F. Caruso, Coll. Surf. A 207 (2002) 33.

[19] T. Radeva, Coll. Surf. A 209 (2002) 219.

[20] W. Chen, T. McCarthy, J. Macromol. 30 (1997) 78.

106 G. Greene, R. Tannenbaum / Applied Surface Science 238 (2004) 101–107

[21] J.-M. Levasalmi, T. McCarthy, J. Macromol. 30 (1997) 1752.

[22] M.C. Hsieh, R.J. Farris, T. McCarthy, J. Macromol. 30 (1997)

8453.

[23] G. Decher, Science 277 (1997) 1232.

[24] A.V. Nabok, A.K. Hassan, A.K. Ray, Mater. Sci. Eng. C 8/9

(1999) 505.

[25] F. Caruso, K. Niikura, D.N. Furlong, Y. Okahata, Langmuir

13 (1997) 3427.

[26] G. Ladam, P. Schaaf, F.J.G. Cuisinier, G. Decher, J.C. Voegel,

Langmuir 17 (2001) 878.

[27] J.D. Mendelsohn, C.J. Barrett, V.V. Chan, A.J. Pal, A.M.

Mayes, M.F. Rubner, Langmuir 16 (2000) 5017.

[28] S.Y. Park, C.J. Barrett, M.F. Rubner, A.M. Mayes, Macro-

molecules 34 (2001) 3384.

[29] S.Y. Park, M.F. Rubner, A.M. Mayes, Langmuir 18 (2002)

9600.

[30] A.J. Chung, M.F. Rubner, Langmuir 18 (2002) 1176.

[31] G. Decher, J.B. Schlenoff, Multilayer Thin Films: Sequential

Assembly of Nanocomposite Materials, Wiley/VCH Verlag

GmbH & Co. KGaA, Weinheim, Germany, 2003.

[32] A.V. Dobrynin, M. Rubinstein, J. Phys. Chem. B 107 (2003)

8260.

[33] G. Greene, R. Tannenbaum, J. Disper. Sci. Technol, in press.

[34] G. Greene, G. Yao, R. Tannenbaum, Langmuir, in press.

[35] R.J. Stokes, D.F. Evans, Fundamentals of Interfacial En-

gineering, Wiley/VCH, New York, NY, 1997.

[36] T. Radeva, V. Milkova, I. Petkanchin, J. Coll. Interf. Sci. 244

(2001) 24.

G. Greene, R. Tannenbaum / Applied Surface Science 238 (2004) 101–107 107