Macromolecules 2010, 43, 9087–9093 9087

DOI: 10.1021/ma1013429

Effect of Molecular Weight and Arm Number on the Growth

and pH-Dependent Morphology of Star Poly[2-(dimethylamino)ethyl
methacrylate]/Poly(styrenesulfonate) Multilayer Films
Zhizhang Guo, Xingyu Chen, Jianyu Xin, Duo Wu, Jianshu Li,* and Chenlong Xu
College of Polymer Science and Engineering, Sichuan University, Chengdu 610065, China
Received June 17, 2010; Revised Manuscript Received September 25, 2010

ABSTRACT: The influence of molecular weight and arm number of star polymers on the buildup and
pH-dependent morphology changes of multilayer films assembled from star poly[2-(dimethylamino)
ethyl methacrylate] (PDMAEMA) and linear poly(styrenesulfonate) (PSS) was systematically investigated.
Well-defined star PDMAEMA, consisting of β-cyclodextrin (CD) core and different arm number (21, 14,
and 7) were successfully synthesized via atom transfer radical polymerization (ATRP). It is found that star
polymers with certain arm number deposit more mass with the increase of molecular weight. As for the star
polymers with almost the same arm length but different arm number, film assembled from 14-arm star
PDMAEMA exhibits the fastest growth and the most significant morphology changes upon the pH 2.75
treatment. It is assumed that the interdiffusion of PSS controlled by star PDMAEMA with different struc-
tures should be responsible for the different buildup and pH-dependent morphology changes of the multi-
layer films.

Introduction polymerization, which has been applied successfully to acquire

The layer-by-layer (LbL) deposition technique initially intro- well-defined star polymers with precise arm numbers. Meanwhile,
duced by Decher et al. in the early 1990s has been the prime choice controlled/living radical polymerization,21 including atom trans-
for constructing nanoscale films on a surface.1-3 As a versatile fer radical polymerization (ATRP), has been demonstrated to be
and tunable method, it has rapidly attracted the most attention an efficient technique for synthesizing star polymers with well-
over a wide range of materials science. Besides conventional poly- defined structure,22-24 arm chain length in particular. Further, star
electrolytes, inorganic nanoparticles,4 dendrimers,5 carbon nano- polymers could be easily modified to have temperature-, pH-, or
tubes,6 and biological substances such as DNA,7 proteins,8 and other stimuli-responsive functionalities, which would be applied
viruses9 have been successfully used as building blocks for LbL to modulate the assembly and mechanical properties of LbL multi-
multilayer films. The increasing kinds of LbL component materials layer films by varying the nanoenvironment conditions. Connal
based on various driving forces further enriched their potential et al.19 have utilized poly(allylamine hydrochloride) (PAH) and
application in several areas, including controlled drug release,10 pH-responsive poly(acrylic acid) (PAA) core cross-linked star
solar cells,11 biomaterial coatings,12 catalysts,13 and biosensors.14 polymers to prepare LbL films. The star PAA/PAH films showed
In recent years, much effort has been made to gain further under- distinct morphology changes in response to post-treatment with
standing of the factors that govern the multilayer buildup pro- different pH solutions. The assembly behavior of pH-responsive
cess.15,16 Hyperbranched polymers, especially dendrimers5,17 and all-star polymer multilayer films was also investigated and com-
star polymers,18-20 have been introduced into multilayer thin films pared to linear polyelectrolytes by Kim et al.20
due to their unique conformation and nanostructural function- The above studies have fully demonstrated the promising appli-
ality. Some precursory researches demonstrated that LbL depos- cations of star polymers in multilayer films. However, the effects
ited films constructed by dendrimers or star polymers behaved of molecular weight (MW) and arm number of star polymers on
quite differently from those of their linear counterparts with equiv- the formation of LbL films have not been discussed yet. In this
alent compositions. It was concluded that the star structure of the work, a series of well-defined star poly[2-(dimethylamino)ethyl
polyelectrolytes play a crucial role in controlling the formation and methacrylate] (PDMAEMA) with different molecular weight
the resulting properties of LbL multilayer films.19,20 and arm number were synthesized via ATRP through “core-
Star polymers are characterized as a globular shape with a first” method. The ATRP initiators with designed initiation sites
number of arms radiating from a central core. The starlike poly- (21, 14, and 7) were prepared according to our previous work.25
mers can basically be synthesized using two different strategies: The multilayer thin films are prepared by the alternate LbL
“arm-first” and “core-first”. The “arm-first” approach consists deposition based on electrostatic interactions between positively
of reacting preformed living linear polymers (arms) with a multi- charged star PDMAEMA and negatively charged linear poly-
functional quencher molecule that forms the core of the star. Star (styrenesulfonate) (PSS). We report, to the best of our knowl-
polymers can also be prepared via the “arm-first” method if edge, the first example of investigating the effects of star polymer
the arms and multifunctional coupling reagent contain comple- architecture on the assembly, morphology, and pH-responsivity
mentary functionalities. In contrast, the alternative “core-first” of LbL membrane. The growth approaches and pH-dependent
method involves using a multifunctional initiator to initiate morphology changes of these films are demonstrated and com-
pared, which will be helpful for further understanding the internal
arrangement of the star-polymer-based LbL films during their
*Corresponding author. E-mail: [email protected]. buildup and post-treatments.
r 2010 American Chemical Society Published on Web 10/14/2010 pubs.acs.org/Macromolecules
9088 Macromolecules, Vol. 43, No. 21, 2010 Guo et al.

Scheme 1. Synthesis of Star PDMAEMA via ATRP from Per-2,3-di- with 0.45 μm microporous filter, 0.2 mL of star polymer solution
O-chloroacetyl-β-cyclodextrin (14Cl-β-CD) (0.5 mg/mL, pH = 6.0) was added into the sample cell and diluted
with water to 10 mL. The titration was carried out with a standard
anionic polyelectrolyte (potassium poly(vinyl sulfate) (PVSK))
solution (concentration = 1 mM). Three repeats were performed
to get the average value for each sample.
Assembly of Star PDMAEMA/PSS Multilayer Films. The
aqueous solutions of star PDMAEMA (0.5 mg/mL, 0.1 M NaCl,
pH = 6.0) and PSS (0.5 mg/mL, 0.1 M NaCl, pH = 6.0) were
prepared and filtrated with 0.45 μm Nylon Cameo filtersyringe
prior to use. The substrate of cleaned quartz was initially im-
mersed into the PDMAEMA solution for 10 min and rinsed by
consecutively three dips in Milli-Q water for 1 min each, followed
Experimental Section by drying with nitrogen airgun before performing data collection.
Materials. Poly(styrenesulfonic acid) sodium salt (PSS, Mw Then the substrate was dipped into PSS solution for 10 min and
70 000 g/mol) was purchased from Alfa Aesar. 2-(Dimethylamino) followed with the same rinsing and drying steps as described above.
ethyl methacrylate (DMAEMA, 99%, Aladdin Reagent Co., Ltd., Film with certain number of layers was obtained by alternat-
China) was purified by passage through a column of activated basic ing sequential adsorption in PDMAEMA and PSS solutions.
alumina to remove inhibitor. Chloroacetylated β-cyclodextrins (21-, A bilayer is defined as a layer of polycation and polyanion,
14-, and 7Cl-β-CD, Supporting Information) were prepared as and the half-integer number of bilayer means that the film ends
multifunctional ATRP initiators according to our previous work.25 with star PDMAEMA as the outermost layer. The films prepared
N,N,N0 ,N00 ,N00 -Pentamethyldiethylenetriamine (PMDETA; 98%) for QCM analysis were manufactured by dropping and extending
was purchased from J&K Chemical Ltd. (Beijing, China). Copper- 200 μL solution onto a certain area of the QCM electrode using
(I) chloride (CuCl; 97%) was stirred in glacial acetic acid, filtered, a pipet instead of immersing it into the polyelectrolyte solution.
then washed promptly with absolute ethanol, and dried under All experiments were carried out at room temperature in an air-
vacuum. All other reagents were of analytical grade or higher and conditioned room.
used as received unless stated otherwise. Quartz crystal micro- Characterizations of LbL Multilayer Film. UV/vis spectra was
balance (QCM) electrodes (titanium/gold crystals, resonant fre- performed with a Mapada UV-1800PC spectrophotometer. The
quency = 5 MHz) were of Stanford Research System Corp. prod- absorbance of PSS at 262 nm was recorded after each adsorption
ucts. A home-built QCM, constructed using a model QCM 100 step. Assembly steps were also monitored by a home-built QCM
analogue controller (Stanford Research System products Corp.) device with a frequency counter. An Inspect F scanning electron
and an Agilent 53131A universal counter, was employed for the microscopy (SEM) was used to observe the surface and cross-
microgravimetric experiments. The quartz wafers (10  30 mm2) section morphology, operating at 10.0 keV. For the atomic force
were cleaned following the “RCA clean” protocol to hydrophilize microscopy (AFM) measurements, a SPA400 Digital Instrument
the surfaces using the method described by Connal et al.19 Ultra- equipped with SI-DF20 Micro cautilever was used. All AFM
pure water produced from a Milli-pore system with a resistivity images of multilayer films were captured in tapping mode (TM-
higher than 18.2 MΩ 3 cm was used in all the experiments. AFM) with spring constant of 12 N/m and resonant frequency of
Syntheses of Star PDMAEMA Polymers via ATRP. A repre- 127 kHz. The root-mean-square (rms) of the surface roughness
sentative process for the polymerization of DMAEMA using the data were collected from three imaging area of 5  5 μm2.
14Cl-β-CD initiator was described as follows (Scheme 1). The
reaction was performed in a 50 mL three-necked flask with mag- Results and Discussion
netic stirring. Distilled DMAEMA (3.6 mL, 21.4 mmol), 14Cl-β-
CD (22 mg, 0.01 mmol), and PMDETA (100 μL, 0.42 mmol) were The controlled polymerization of DMAEMA monomer via
dissolved in 2.7 mL of a methanol/water (7/1, v/v) mixture. After ATRP in nonpolar solvents was first reported by Zhang et al.26
three freeze-pump-thaw cycles, the reaction mixture was further After then, Mao and co-workers27 have discussed the efficient poly-
degassed by bubbling nitrogen for 20 min. Then, CuCl (42 mg, merizations of DMAEMA via ATRP in various methanol and
0.42 mmol) was added into the flask under a nitrogen atmosphere. water mixtures at room temperature. The synthesis of PDMAEMA
As the CuCl dissolved gradually, the solution turned to brilliant with molar mass up to 1 million was also successfully achieved
blue. The reaction was then kept at 25 °C for 8 h. Sample was taken by Mao et al.28 The solvent mixture of methanol and water was
from the rubber stopper for 1H NMR analysis. The polymeriza- fixed at 7/1 (volume ratio) in this work. The well-defined star
tion was terminated by exposing to air and diluted with 50 mL of polymers, consisting of β-cyclodextrin (CD) core and different
THF. A colorless solution was obtained after catalyst complex number of PDMAEMA arms, were successfully synthesized via
was removed by passing through a short basic alumina column. ATRP using the “core-first” method. For example, the monomer
After being concentrated in a rotary evaporator, the solution was conversion and yields while preparing S5 was 65% and 45%,
poured into excess cold n-hexane for precipitation, yielding white respectively, as shown in Figure 2 of the Supporting Information.
solid. Then, star polymers were collected and dried under vacuum The reaction conditions and GPC results of PDMAEMA star
at 40 °C for 2 days. polymers are summarized in Table 1. The polydispersity index
Characterizations of Star PDMAEMA. Gel permeation chro-
(PDI) value of star polymers remains low, indicating the well-
matography (GPC) measurement was performed on a Waters
controlled polymerization (Supporting Information).
GPC system equipped with Ultrahydrogel columns and Waters
410 differential refractometer detector at 25 °C. Water was used as As shown in Table 1, star polymer S1-S4 have 21 arms but
the eluent, and the flow rate was 1.0 mL/min. Calibration was different arm chain length. Meanwhile, samples S2, S5, and S6 are
made using standard monodispersed poly(ethylene glycol) (PEG). designed to have the same number of DMAEMA repeat units per
Dynamic light scattering (DLS) analysis was carried out on a arm, which are 93, 87, and 88 calculated from Mn, respectively.
Brookhaven BI-200 SM system incorporating a He-Ne laser These star polymers have very similar arm length but different arm
(λ = 532 nm) at 25 °C. The hydrodynamic diameters of polymer number. The arms are connected to the CD core via ester linkages,
at a concentration of 10 mg/mL in water (pH varying from 10 which could be cleaved by hydrolyzation under basic conditions.
to 2) were collected three times at an angle of 90°. The apparent To determine the number of star polymer arms, star polymer S1
charge density of star PDMAEMA was determined using a M€ utek was chosen and hydrolyzed with a potassium hydroxide solution
particle charge detector (Herrsching, Germany). After filtrated in a mixture of THF and methanol at 40 °C for 2 days, at which
Article Macromolecules, Vol. 43, No. 21, 2010 9089

Table 1. Polymerization Conditions, GPC, and Charge Density

Results of Star PDMAEMA Samples
arm charge density
PDMAEMA numbera [M]0/[I]0/ time (10-5
samples (narm) [Cu]0/[L]0b (h) Mn Mw/Mn equiv/g)

S1 21 230/1/3/3 8 393 500 1.19 74.1 ( 0.6

S2 21 160/1/3/3 8 310 900 1.13 85.8 ( 1.7
S3 21 110/1/3/3 8 227 400 1.10 86.9 ( 1.1
S4 21 55/1/3/3 8 95 100 1.17 92.1 ( 2.1
S5 14 160/1/3/3 8 194 000 1.09 39.7 ( 2.4
S6 7 160/1/3/3 8 98 800 1.12 43.4 ( 1.5
a
Determined by ATRP initiators with designed initiation sites (21-,
14-, and 7Cl-β-CD). b [M]0, [I]0, [Cu]0, and [L]0 represent the initial malor
ratio of monomer, total initiation sites of different initiators, Cu(I)Cl,
and ligand (PMDETA), respectively.

Figure 2. Absorbance of PSS at 262 nm versus the number of star

PDMAEMA/PSS bilayers. The absorbance of the first layer (star S1)
deposited on quartz substrate corresponds to zero. All of the polymer
solutions were made with 0.1 M NaCl. The dotted line indicates the
stepwise PSS adsorption/extraction process.
polymer become more stretched due to the electrostatic repulsion
between the chains.
The exponential growth mode of polyelectrolyte multilayers
has been studied by several groups,30,31 and then the so-called “in
and out” diffusion buildup mechanism has been widely accepted.
The thickness of star PDMAEMA and linear PSS LbL films
investigated by QCM measurement also increases exponentially
with the bilayer number, which would be related to the diffusion
ability of at least one of the polyelectrolytes. Because of its glob-
ular structure and higher molecular weight, star polymer is
unlikely to be the diffusion species in the multilayer film.18
The layer-by-layer sequential depositions of the two polyelectro-
Figure 1. Hydrodynamic diameters (DH) of star PDMAEMA at different
pH conditions. The concentration of star polymers S1 was 10.0 mg/mL.
lytes are demonstrated both by UV/vis adsorption spectroscopy
and QCM. It has been reported by Kim et al. that PDMAEMA
condition hydrolysis is restricted to the esters joining the arms to could be well-covered on silicon wafer.20 Therefore, star poly-
the core and does not affect the main polymer chains (see Figure 4 mers, not the branched PEI, were deposited on the quartz surface
of the Supporting Information). After hydrolysis, the Mn and PDI as the first layer in this work. The adsorption and desorption
of the polymer arms measured by GPC were 17 400 and 1.47, behavior of dye molecules in alternating deposited LbL films has
respectively. Taking into account of the Mn of the original star poly- been reported by Tedeschi et al.32 The highly identical trend was
mer S1 and CD core initiator molar mass (Mn 1135), the number also found in our experiments as presented in Figure 2. The
of arms is estimated to be 22.5, which is rather close to the designed maximum absorption of PSS at 262 nm was chosen in order to
21 arms. Meanwhile, the absolute Mw of star polymer S1 and avoid the interferer adsorption of PDMAEMA between 200 and
linear arm polymer obtained by multiangle laser light scattering 240 nm. The absorbance decreased after each star PDMAEMA
(MALLS) detector (Supporting Information) were 490 100 and layer was deposited. Meanwhile, the star PDMAEMA aqueous
20 600 g/mol, respectively. Thus, the number of arms is estimated solution turned to be whitish after manufacturing several layers.
to be 23.7. The disparity of Mw obtained from MALLS and GPC It would be the electrostatic complexes formed between the star
measurement may due to the dense structure of star polymer polymer in solution and the PSS extracted from the LbL films.
compared to the linear polymer of similar molecular weight. In the QCM experiment, the crystal resonance frequency
The effective pKa of the DMAEMA star polymers shows changes are recorded to reflect the mass of adsorbed polyelec-
slightly disparity from their linear counterparts. For example, the trolyte on the crystal surface. As depicted in Figure 3, the linear
pKa of star S1 shifts from 7.5 (linear29) to 6.5 (measured by PDMAEMA/PSS exhibited a regular linear growth trend at 2
potentiometric titration, data not shown). This is because of the mg/mL with 0.1 M NaCl. However, a wavelike growth region
higher inner osmotic pressure caused by the more compact was observed for star polymer at the same conditions after two
charged groups due to the unique architecture of star polymers.20 bilayers. It seems that star polymers at higher concentration
That is, to reach the same degree of ionization, more protons are (2 mg/mL) could not be stably deposited onto the quartz crystal
needed to overcome the higher pressure and penetrate the con- microbalance. The frequency increased after each deposition
gested structure to access the inner of star polymers. The pH- of star PDMAEMA, which indicates that the deposited mass
responsive behavior of star PDMAEMA is demonstrated by was unexpectantly lost in the bilayer manufacturing cycle. No
dynamic light scattering as shown in Figure 1. The average hydro- wavelike growth was found while the concentration was changed
dynamic diameters (DH) of star polymer change along with the to 0.5 mg/mL (Figure 4). Thus, it is important to note that the
varying of pH. With the decrease of pH from 10 to 2, the DH of polymer solutions should be prepared in certain concentration
star PDMAEMA S1 increases from 14 to 26 nm. The star poly- ranges for LbL assembly. A higher polymer concentration would
mers are in a constrained coiled conformation at high pH. While not necessarily result in greater deposited mass. Also, the LbL
the protonation degree of DMAEMA arms increases, the star thin films exhibit pH tunable growth due to the protonation
9090 Macromolecules, Vol. 43, No. 21, 2010 Guo et al.

Figure 3. Frequency shifts on QCM electrodes as a function of the

number of layers for star PDMAEMA and linear PDMAEMA
(hydrolysis from star polymer) multilayer films. All of the polymer
solutions were prepared at 2 mg/mL at pH 6 with 0.1 M NaCl.

Figure 5. (a) QCM frequency shift versus the number of 21-arm star
PDMAEMA/PSS bilayers prepared at pH 6/6. (b) QCM frequency
shift versus the number of different arm star PDMAEMA/PSS bilayers
prepared at pH 6/6.

Figure 4. Exponential growth of S1/PSS at 0.5 mg/mL against the molecular weight on several properties of multilayer films. It is
bilayer number of LbL films fabricating under different pH. found that 21-arm star polymer with longer chain length got more
deposited mass for the same bilayers, calculating from the QCM
process of DMAEMA units on the star polymer chains. The frequency changes for films fabricated by S1-S4 (Figure 5a).
pH of PSS solution was fixed at 6.0, while the S1 solutions were Such difference in film growth is considered to be related to the
prepared at pH 4.0 and 6.0, respectively. As shown in Figure 4, diffusion movements within the films.36 By using dye-conjugated
the film fabricated under pH 6 was thicker than that prepared polyelectrolytes, Picart et al.30 have proved that the thickness of
under pH 4. The following experiential equation is often used to the new forming layer on top of the film is proportional to the
estimate the film thickness from the frequency shift (ΔF ).33 amount of the free diffusion specie during the buildup process. In
this work, it seems that the diffusion of PSS could significantly
d ðnmÞ  - 0:017ΔF ðHzÞ affect the thickness of multilayer films. In order to fully illustrate
it, the charge density of each star polymer was measured as shown
Thus, the film thicknesses of nine bilayers calculated from the in Table 1. The higher apparent charge density would restrict the
total QCM frequency changes are 31.2 nm (pH = 6) and 21.3 nm diffusion of PSS and decrease the amount of free PSS at the outer
(pH = 4). On the other hand, the pH-dependent growth agrees layer due to the higher degree of ionic cross-linking. Meanwhile,
well with the conclusion drawn in previous work,20,34 which have the lower extent of chain interpenetration between S1 due to its
demonstrated that increasing the degree of ionization of a weak higher molecular weight and more congested architecture would
polyelectrolyte by controlling the dipping pH condition could be benefit to the diffusion of PSS and result in thicker films.
result in thinner absorbed layers over a pH range. In this work, the investigation is focused on the influence of arm
The influence of the polyelectrolyte molecular weight on the number of the star polyelectrolytes. The star PDMAEMA (S2, S5,
exponential growth and morphology of multilayers has already and S6) with almost the same arm length but different arm num-
been investigated with natural polysaccharides,31 the hyaluronic ber (21, 14, and 7, respectively) are chosen to carry out the QCM
acid/poly(L-lysine) (HA/PLL) and poly(L-glutamic acid)/poly- dissipation monitoring, together with AFM measurement. The
(allylamine) (PGA/PAH) systems.35 Recent advances of con- frequency changes of each bilayer (pH 6/6) were recorded and are
trolled/living polymerization techniques make it feasible to syn- plotted in Figure 5b. As shown in Figure 5, the LbL film thickness
thesize polymers with narrow polydispersity (PDI < 1.2), which calculated from QCM frequency changes (pH = 6) for S1-S6
is of great benefit to systematically investigate the influence of the was 31.2, 18.6, 17.4, 14.6, 31.7, and 24.7 nm, respectively. LbL
Article Macromolecules, Vol. 43, No. 21, 2010 9091

Figure 6. AFM images of different star PDMAEMA/PSS multilayer films before and after post-treatments. All the images areas are 5  5 μm2.

assemblies of all the films follow the exponential growth pattern.

Fourteen-arm star polymer (S5) shows the most amount of de-
posited mass for a given number of bilayers. Meanwhile, 21-arm
star PDMAEMA (S2) gets the least deposited mass. It is obvious
that the growth speed of star PDMAEMA/PSS multilayer films
does not only depend on the molecular weight but also on other
factors. Star PDMAEMA polymers with various arm number
would achieve different protonation degree at certain pH/con-
centration due to their architectural difference. It is found that the
sequence of apparent charge density under assembly condition
is S2>S6>S5 (Table 1), which is reasonable to explain the de-
posited mass sequence of S2 < S6 < S5 (Figure 5) as the diffusion
of negatively charged PSS is restricted by positively charged star
polymer. However, the charge density might not be the only effect
factor while we compare S1 with S5 as they exhibit close deposited
mass although S1 has much higher charge density than S5. The
higher molecular weight and more congested architecture of S1
could be another reason, which cause less polymer chain inter- Figure 7. Surface roughness information on the films before and after
penetration and make the diffusion of PSS to be easier. Thus, post-treatments. The rms values are obtained from three different AFM
both molecular weight and arm number of star polymers, which images of 5  5 μm2.
will result in different charge density and chain interpenetration
at the pH/concentration condition of LbL assembly, may affect
the buildup process and properties of LbL films.
The surface morphologies of multilayer films assembled from
different star polymers are observed by AFM measurement
(Figure 6). Also, the rms roughness data are presented in Figure 7.
The films assembled from 21- and 7-arm star polymers have almost
the same rms roughness (18.78 and 19.04, respectively). As the
fastest growth system (see Figure 5), the film assembled from
14-arm star polymer (S5) exhibits the smoothest surface morphol-
ogy. It seems that star polymer with lower charge density may result
in the deposition of thicker and smoother film. The discrepancy
may be caused by the different degree of protonation and the Figure 8. Large-scale AFM images of multilayer film assembled with
conformation of star polymers under the assembly conditions. 14-arm star polymer after being post-treated in pH 2.75 acidic solution
In general, the pH-driven thickness change and formation of for 30 min.
large-scale porous morphology within the membrane are ex-
plained to be caused by structural rearrangement and phase sep- the star polyelectrolyte’s compact structure can reduce the extent
aration of weak polyelectrolytes.34,37,38 For star polymer, it has of interpenetration and the degree of ionic cross-linking between
been reported that LbL multilayer film assembled with all star the multilayer assembly components.20 The AFM measurement
polymer exhibited more extensive pH-responsiveness upon the was performed to further investigate the pH-induced phase sep-
same post-treatment conditions compared with that assembled aration of multilayer films assembled from star polymer with
with their linear counterpart.20 A reasonable explanation is that different arm number. Each film undergoes remarkable thickness
9092 Macromolecules, Vol. 43, No. 21, 2010 Guo et al.

Figure 9. Morphology changes of star PDMAEMA/PSS film before (A, C) and after (B, D) post-treatment at pH 2.75. SEM cross section (A, B)
and top-down (C, D) images are presented. The representative pores are labeled by red circles in (D).

swelling and surface morphology changes after being immersed were then heated at 120 °C for 2 h to lock the structure prior to the
into pH 2.75 acidic solution for 30 min and dried up at 120 °C SEM observation. However, the thickness of 9.5 bilayers film
for 2 h. At lower pH, the tertiary amine groups on the star (pH = 6) measured from SEM cross-section image is 150.3 nm
PDMAEMA chains would become protonated gradually, which (Figure 9A). It appears that the film thicknesses obtained from
will disrupt a part of the original ionic cross-links and induce the the two measurements (SEM and QCM) are in quite disagree-
as described spinodal-type decomposition.40 The surface roughness ment. Since SEM cross-section thickness is visible and more cred-
of films assembled from 21- and 7-arm star polymers dereased after ible, the experiential equation mentioned above may not suitable
acidic treatment (Figure 7). The films were subsequently exposure for the star polymer system. The film thickness measured from the
to neutral water bath overnight to investigate the possible second- scratched step height by AFM was about 135 nm, which is in good
ary reorganization. It is confirmed that the surface become rougher agreement with the thickness measured by SEM (see Figure 6 of
upon the neutral water treatment. These phenomena are consistent Supporting Information). The average thickness of untreated film
with the finding of previous work that smoother films are accom- is estimated at about 150 nm, and the surface exhibits a rather
panied by the more stretched conformation of star polymers or vice dense texture (Figure 9A,C). For the post-pH-treated film, the
versa.19 As shown in column 2 of Figure 6, ununiform pores are film changes dramatically to a nanoscale or even microporous
observed within multilayer thin films deposited from 21-arm and structure, with pores in diameter ranging from approximately
14-arm star polymers. It seems that star polymers with more arms 300 nm to 1.8 μm (Figure 9D). The microporous surface morphol-
can induce more notable morphology change upon the post- ogy is quite similar to that of post-pH-treated (pH 3.0) layer-by-
treatment due to the enlarged effect of the conformation transition layer multilayers consisting of linear poly(ethylenimine) (LPEI)
from a swollen coil to a globule. In the AFM images, the smaller and poly(acrylic acid) (PAA), which has been described as
pores may indicate the initial stage of porosity transformation and isolated craters.38 The thickness of the post-treated porous film is
pore development. It is interesting to find that the film fabricated more than double of the original film, reaching up to 328 nm
with 14-arm star PDMAEMA exhibits the most significant mor- (Figure 9B). Meanwhile, it is interesting to find that the film in
phology change with pore diameter up to 2 μm, and the surface SEM images turns to be brighter after the porosity transforma-
turns to rougher upon the post-acidic treatment. Large-scale AFM tion at pH 2.75, which may due to the changing of pore conduc-
images of 10  10 μm2 were captured, and several round pores of tive surface on the cross section of the film, hence changing of the
about 80 nm height were observed (Figure 8). It is hypothesized electrons in SEM measurement. Given the conservation of mass
that star polymers with the lower charge density have less ionic in the films during the porosity transformation undergoing the pH
binding sites when assembled into LbL films. Thus, the fracture of 2.75 immersion, the pore volume of 21-arm star PDMAEMA/
ionic cross-links and the structural rearrangement upon the post- PSS film system (assembled at pH 6/6) is estimated at 53%
acidic treatment may lead to more dramatic morphology change according to the previous reports.39,40
for 14-arm star polymer within the multilayer film.
The large area surface morphology and thickness change of
assembled multilayer film (9.5 bilayers) were also examined by Conclusions
SEM. The quartz substrate was cleaved and separated into two The synthesis of controlled star PDMAEMA with different
parts after a “freeze-fractured” process. One part of the substrate molecular weight and arm number (21, 14, and 7) was achieved by
remained untreated as a control while the other part was im- ATRP through the “core-first” method. The pH-responsive
mersed into a bath of acidic solution (pH 2.75 diluted HCl) for behavior of star PDMAEMA was demonstrated by DLS in
30 min followed by a 20 s rinse with ultrapure water. The films solution. It is anticipated that the LbL assembled thin films could
Article Macromolecules, Vol. 43, No. 21, 2010 9093

preserve the pH-responsivity when the star polymers are con- (12) Zhang, J.; Senger, B.; Vautier, D.; Picart, C.; Schaaf, P.; Voegel,
structed into the films, which will be of great benefit for potential J. C.; Lavalle, P. Biomaterials 2005, 26, 3353–3361.
biomedical applications, especially for LbL films containing biol- (13) Liu, J.; Cheng, L.; Song, Y.; Liu, B.; Dong, S. Langmuir 2001, 17,
6747–6750.
ogical components. Herein, the effects of molecular weight and (14) Calvo, E. J.; Danilowicz, C.; Lagier, C. M.; Manrique, J.; Otero, M.
arm number of star polymers on the formation and morphology Biosens. Bioelectron. 2004, 19, 1219–1228.
changes in response to the post-pH treatment of LbL films were (15) Messina, R. Macromolecules 2004, 37, 621–629.
investigated. It is found that star polymers with certain arm (16) Schlnoff, J. B.; Dubas, S. T. Macromolecules 2001, 34, 592–598.
number deposit more mass with the increase of molecular weight. (17) Blasini, D. R.; Flores-Torres, S.; Smilgies, D. M.; Abru~ na, H. D.
The film assembled from 14-arm star PDMAEMA exhibits the Langmuir 2006, 22, 2082–2089.
(18) Yang, S.; Zhang, Y.; Wang, C.; Hong, S.; Xu, J.; Chen, Y. Langmuir
fastest growth and the most dramatic pH-dependent morphology 2006, 22, 338–343.
change with pore diameter up to 2 μm upon the post-acidic (19) Connal, L. A.; Li, Q.; Quinn, J. F.; Tjipto, E.; Caruso, F.; Qiao, G. G.
treatment. The different charge density and the degree of inter- Macromolecules 2008, 41, 2620–2626.
mingling between star polymers within the film are considered to (20) Kim, B.-S.; Gao, H.; Argun, A. A.; Matyjaszewski, K.; Hammond,
be the main factor which determines the formation and the pH- P. T. Macromolecules 2009, 42, 368–375.
dependent properties of multilayer films. Current work is under- (21) Braunecker, W. A.; Matyjaszewski, K. Prog. Polym. Sci. 2007, 32,
93–146.
way to investigate self-regulated drug delivery systems based on (22) Gao, H.; Matyjaszewski, K. Macromolecules 2006, 39, 3154–3160.
LbL films fabricated with star polymers, insulin, and glucose (23) Li, J.; Xiao, H.; Kim, Y. S.; Lowe, T. L. J. Polym. Sci., Part A: Polym.
oxidase. Chem. 2005, 43, 6345–6354.
(24) Li, J.; Guo, Z.; Xin, J.; Zhao, G.; Xiao, H. Carbohydr. Polym. 2010,
Acknowledgment. We thank Prof. Hua Ai for technical sup- 79, 277–283.
(25) Guo, Z.; Chen, X.; Zhang, X.; Xin, J.; Li, J.; Xiao, H. Tetrahedron
port and helpful suggestions. The financial support from the
Lett. 2010, 51, 2351–2353.
National Natural Science Foundation of China (20704026) (26) Zhang, X.; Xia, J.; Matyjaszewski, K. Macromolecules 1998, 31,
and Fok Ying Tong Education Foundation (122034) is greatly 5167–5169.
appreciated. (27) Mao, B.; Gan, L.-H.; Gan, Y.-Y.; Li, X.; Ravi, P.; Tam, K.-C.
J. Polym. Sci., Part A: Polym. Chem. 2004, 42, 5161–5169.
Supporting Information Available: Additional details of the (28) Mao, B.; Gan, L.-H.; Gan, Y.-Y. Polymer 2006, 47, 3017–3020.
(29) van de Watering, P.; Moret, E. E.; Schuurmans-Nieuwenbroek,
synthesis of star polymers and the film thickness measure-
N. M. E.; van Steenbergen, M. J.; Hennink, W. E. Bioconjugate
ment. This material is available free of charge via the Internet at Chem. 1999, 10, 589–597.
http://pubs.acs.org. (30) Picart, C.; Mutterer, J.; Richert, L.; Luo, Y.; Prestwich, G. D.;
Schaaf, P.; Voegel, J.-C.; Lavalle, P. Proc. Natl. Acad. Sci. U.S.A.
References and Notes 2002, 99, 12531–12535.
(31) Kujawa, P.; Moraille, P.; Sanchez, J.; Badia, A.; Winnik, F. M.
(1) Decher, G. Science 1997, 277, 1232–1237. J. Am. Chem. Soc. 2005, 127, 9224–9234.
(2) Hammond, P. T. Adv. Mater. 2004, 16, 1271–1293. (32) Tedeschi, C.; Caruso, F.; M€ ohwald, H.; Kirstein, S. J. Am. Chem.
(3) Boudou, T.; Crouzier, T.; Ren, K.; Blin, G.; Picart, C. Adv. Mater. Soc. 2000, 122, 5841–5848.
2010, 22, 441–467. (33) Lvov, Y.; Ariga, K.; Ichinose, I.; Kunitake, K. J. Am. Chem. Soc.
(4) Kaschak, D. M.; Mallouk, T. E. J. Am. Chem. Soc. 1996, 118, 1995, 117, 6117–6123.
4222–4223. (34) Shiratori, S. S.; Rubner, M. F. Macromolecules 2000, 33, 4213–4219.
(5) Khopade, A. J.; Caruso, F. Nano Lett. 2002, 2, 415–418. (35) Porcel, C.; Lavalle, Ph.; Decher, G.; Senger, B.; Voegel, J.-C.;
(6) Mamedov, A. A.; Kotov, N. A.; Prato, M.; Guldi, D. M.; Wicksted, Schaaf, P. Langmuir 2007, 23, 1898–1904.
J. P.; Hirsch, A. Nature Mater. 2002, 1, 190–194. (36) Sun, B.; Jewell, C. M.; Fredin, N. J.; Lynn, D. M. Langmuir 2007,
(7) Lvov, Y.; Decher, G.; Sukhorukov, G. Macromolecules 1993, 26, 23, 8452–8459.
5396–5399. (37) Tokarev, I.; Minko, S. Adv. Mater. 2009, 21, 241–247.
(8) Caruso, F.; M€ ohwald, H. J. J. Am. Chem. Soc. 1999, 121, 6039– (38) Lutkenhaus, J. L.; McEnnis, K.; Hammond, P. T. Macromolecules
6046. 2008, 41, 6047–6054.
(9) Yoo, P. J.; Nam, K. T.; Qi, J.; Lee, S. K.; Park, J.; Belcher, A. M.; (39) Yoo, D.; Shiratori, S. S.; Rubner, M. F. Macromolecules 1998, 31,
Hammond, P. T. Nature Mater. 2006, 5, 234–240. 4309–4318.
(10) Khopade, A. J.; Caruso, F. Biomacromolecules 2002, 3, 1154–1162. (40) Mendelsohn, J. D.; Barrett, C. J.; Chan, V. V.; Pal, A. J.; Mayes,
(11) Lowman, G. M.; Hammond, P. T. Small 2005, 1, 1070–1073. A. M.; Rubner, M. F. Langmuir 2000, 16, 5017–5023.