Langmuir ****, **, 851*-****-****
Anisotropic Structure and Transport in Self-Assembled Layered
Polymer-Clay Nanocomposites
Jodie L. Lutkenhaus, Elsa A. Olivetti, Eric A. Verploegen, Bryan M. Cord,
Donald R. Sadoway, and Paula T. Hammond*,
Departments of Chemical Engineering, Materials Science and Engineering, and Electrical Engineering,
Massachusetts Institute of Technology, 77 Massachusetts AVenue, Cambridge, Massachusetts 02139
ReceiVed February 13, 2007. In Final Form: May 20, 2007
Using the layer-by-layer (LbL) assembly technique, we create a polymer-clay structure from a unique combination
of LbL materials: poly(ethylene imine), Laponite clay, and poly(ethylene oxide). This trilayer LbL structure is
assembled using a combination of hydrogen bonding and electrostatic interactions. The films were characterized using
ellipsometry, profilometry, X-ray photon spectroscopy, atomic force microscopy, scanning electron microscopy, wide-
angle X-ray diffraction, grazing-incidence small-angle X-ray scattering, and electrochemical impedance spectroscopy
(EIS). We observe a layered, anisotropic structure, which resulted in in-plane ion transport 100 times faster than
cross-plane at 0% relative humidity. This study represents a first application of EIS in determining anisotropic ion
transport in LbL assemblies and its correlation to structural anisotropy.
Introduction ion conducting electrolyte owing to recent advances that
demonstrate high conductivity and transference numbers near
Synthetic clays are of interest for tuning bulk properties
unity.12-14 PEO complexes with and intercalates layered smectite
(rheological, mechanical, transport) at the nanoscale in the design
clays through the competition of PEO and water binding to the
of composite materials because of clay s unique materials
clay platelet.21,22 A well-studied polymer electrolyte, PEO
properties (e.g., negative charge, silicate surface) and dimensions
associates with alkali cations through ion-dipole interactions,
(e.g., nanoscale, platelet-shaped).1 The high aspect ratio of the
and cation mobility is influenced by local relaxations and
clay platelet is thought to yield superior transport barrier
segmental motion of the polymer backbone.23,24 Early work with
properties, particularly when oriented in layers.2,3 Clay composites
blended composites of PEO and montmorillonite demonstrated
or clay-modified materials are often produced using mechanical
ionic conductivities of 10-9 to 10-7 S cm-1 at 425 K,4,6,8
pressure,4 controlled drying from dilute solution,5 or simple
values much higher than montmorillonite alone. Adding a lithium
blending. Such systems may be of interest for diffusion blocking
salt such as lithium perchlorate can improve the room-temperature
layers, mechanical modifiers, coatings, dielectrics, and so forth.
conductivity ( 10-5 S cm-1).18 Addition of a plasticizer or
Montmorillonite and hectorite, both charged layer silicates
small molecule such as ethylene carbonate can also improve
(smectite clays), intercalated with poly(ethylene oxide) (PEO)
performance ( 10-4 S cm-1),12-14 but mechanical properties
and its derivatives have received much attention4,6-20 as a single-
may suffer.
The use of ultrathin electrolytes allows reduction in overall
* Corresponding author. E-mail: abqq6d@r.postjobfree.com.
Department of Chemical Engineering.
electrolyte resistance, R, which scales with film thickness, L (R
Department of Materials Science and Engineering.
) L/area ); thus, a film with a lower conductivity may provide
Department of Electrical Engineering.
a small resistance if made sufficiently thin. A simple and elegant
(1) Pinnavaia, T. J.; Beall, G. W. Polymer-Clay Nanocomposites; Wiley &
way to construct ultrathin, mechanically cohesive polymer-
Sons: Chichester, 2000.
(2) Struth, B.; Eckle, M.; Decher, G.; Oeser, R.; Simon, P.; Schubert, D. W.; clay nanocomposites is the layer-by-layer (LbL) technique.25
Schmitt, J. Eur. Phys. J. E 2001, 6 (5), 351-358.
Molecular species of opposite charge26,27 (or hydrogen-bonding
(3) Kim, D. W.; Choi, H.-S.; Lee, C.; Blumstein, A.; Kang, Y. Electrochim.
Acta 2004, 50 (2-3), 659-662. functionality)28,29 are alternately adsorbed on a substrate from
(4) Aranda, P.; Galvan, J. C.; Casal, B.; Ruiz-Hitsky, E. Electrochim. Acta aqueous solution to form thin films of tunable thickness, structure,
1992, 37 (9), 1573-1577.
and properties.26-29 Multilayers from positively charged poly-
(5) Ghosh, P. K.; Bard, A. J. J. Am. Chem. Soc. 1983, 105 (17), 5691-5693.
(6) Ruiz-Hitsky, E.; Aranda, P. AdV. Mater. 1990, 2 (11), 545-547. electrolytes and negatively charged clays have been studied as
(7) Aranda, P.; Ruiz-Hitsky, E. Chem. Mater. 1992, 4 (6), 1395-1403.
(8) Wu, J.; Lerner, M. M. Chem. Mater. 1993, 5 (6), 835-838. (19) Loyens, W.; Maurer, F. H. J.; Jannasch, P. Polymer 2005, 46 (18), 7334-
(9) Aranda, P.; Ruiz-Hitsky, E. Acta Polym. 1994, 45 (2), 59-67. 7335.
(10) Doeff, M. M.; Reed, J. S. Solid State Ionics 1998, 113-115, 109-115. (20) Manoratne, C. H.; Rajapakse, R. M. G.; Dissanayake, M. A. K. L. Int.
(11) Chen, H.-W.; Chang, F.-C. Polymer 2001, 42 (24), 9763-9769. J. Electrochem. Sci. 2006, 1 (1), 32-46.
(21) Parfitt, R. L.; Greenland, D. J. Clay Miner. 1970, 8, 305-315.
(12) Riley, M.; Fedkiw, P. S.; Khan, S. A. J. Electrochem. Soc. 2002, 149 (6),
A667-A674. (22) Aray, Y.; Marquez, M.; Rodriguez, J.; Vega, D.; Simon-Manso, Y.; Coll,
S.; Gonzalez, C.; Weita, D. A. J. Phys. Chem. B 2004, 108 (7), 2418-2424.
(13) Walls, H. J.; Riley, M. W.; Singhal, R. R.; Spontak, R. J.; Fedkiw, P. S.;
Khan, S. A. AdV. Funct. Mater. 2003, 13 (9), 710-717. (23) Tarascon, J.; Armand, M. Nature 2001, 414, 359-367.
(14) Singhal, R. G.; Capracotta, M. D.; Martin, J. D.; Khan, S. A.; Fedkiw, (24) Gray, F. M. Polymer Electrolytes; Royal Society of Chemistry: Cambridge,
P. S. J. Power Sources 2004, 128 (2), 247-255. 1997.
(25) Kleinfeld, E. R.; Ferguson, G. S. Science 1994, 265 (5170), 370-373.
(15) Kurian, M.; Galvin, M.; Trapa, P.; Sadoway, D. R.; Mayes, A. M.
Electrochim. Acta 2005, 50 (10), 2125-2134. (26) Decher, G.; Hong, J. D.; Schmitt, J. Thin Solid Films 1992, 210-211 (2),
831-835.
(16) Wong, S.; Vasudevan, S.; Vaia, R. A.; Giannelis, E. P.; Zax, D. B. J. Am.
Chem. Soc. 1995, 117 (28), 7568-7569. (27) Decher, G. Science 1997, 277 (5330), 1232-1237.
(28) Stockton, W. B.; Rubner, M. F. Macromolecules 1997, 30 (9), 2717-
(17) Vaia, R. A.; Vasudevan, S.; Krawiec, W.; Scanlon, L. G.; Giannelis, E.
P. AdV. Mater. 1995, 7 (2), 154-156. 2725.
(18) Chen, H.-W.; Chiu, C.-Y.; Chang, F.-C. J. Polym. Sci., Part B: Polym. (29) Wang, L.; Wang, Z.; Zhang, X.; Shen, J.; Chi, L.; Fuchs, H. Macromol.
Phys. 2002, 40 (13), 1342-1353. Rapid Commun. 1997, 18 (6), 509-514.
10.1021/la700432p CCC: $37.00 2007 American Chemical Society
Published on Web 06/29/2007
8516 Langmuir, Vol. 23, No. 16, 2007 Lutkenhaus et al.
surface modifiers, coatings, sensors, permeation barriers, and Layer-by-Layer Film Assembly. Films were constructed using
artificial nacre .3,25,30-43 In general, these layered organic- a modified programmable Carl Zeiss HMS slide stainer. Substrates
used were silicon wafer and ITO-coated glass. Si wafers were cleaned
inorganic composites form a highly stratified two-dimensional
using piranha solution of 70% sulfuric acid and 30% hydrogen
structure,25 which may be capable of blocking the diffusion of
peroxide. CAUTION: Piranha solution is extremely corrosiVe. ITO-
ions2 or the permeation of gas.3,35,42
coated glass substrates were cleaned by sequential sonication in
However, to explore applications in which the LbL polymer- dichloromethane, acetone, methanol, and Milli-Q water for 15 min
clay composites may be part of an electrochemical device, we each. Immediately before LbL assembly, the substrate was oxygen
need to introduce and understand ionic conductivity. In this work, plasma-treated for 2 min. After plasma treatment, the substrate was
layered polymer-clay structures from LbL assembly are char- first dipped in PEI solution for 10 min, rinsed with agitation in
Milli-Q water for 2 min, followed by an additional 1-min rinse.
acterized and created using a unique combination of materials:
Second, the substrate was exposed to the Laponite dispersion for 10
PEO, linear poly(ethylene imine) (PEI), and neat or lithium-
min and rinsed as before. Finally, the substrate was exposed to PEO
exchanged Laponite clay (designated as clay and Li-clay,
solution for 10 min and rinsed as before. These three exposures
respectively). Here, hydrogen bonding is used to introduce PEO
comprise one trilayer of PEI/clay/PEO. The procedure can be repeated
into the multilayer film while using a polycation, PEI, to stabilize
n times to give a film of n trilayers denoted by (PEI/clay/PEO)n.
the composite. The resulting structure, studied using atomic force The film thickness was measured using either ellipsometry or
microscopy (AFM), scanning electron microscopy (SEM), wide- profilometry depending on film thickness. Film thicknesses less
angle X-ray diffraction (WAXD), and grazing-incidence small- than 150 nm were measured using a Gaertner ellipsometer. Film
angle X-ray scattering (GI-SAXS), suggests lateral orientation thicknesses greater than 150 nm were measured using a Tencor P-10
over large areas (>4 cm2). We demonstrate and characterize the profilometer. The thickness was recorded three times on two different
degree of anisotropic ion transport using electrochemical samples to give one data point.
X-ray Photon Spectroscopy. Surface characterization and
impedance spectroscopy (EIS), and it was found that dry-state
in-plane ionic conductivity (7.2 10-8 S cm-1 at 401 K) is 100 elemental analysis were performed using a Kratos AXIS Ultra
times higher than cross-plane conductivity (6.8 10-10 S cm-1 Imaging X-ray photoelectron spectrometer at 0.5 eV/step and 80 eV
pass energy.
at 405 K), a result of the layered structure within the film. Thus,
AFM. A Dimension 3100 AFM by DI Instruments with a
structural anisotropy within LbL polymer-clay composite films
Nanoscope 3A Controller in tapping mode was used to investigate
is correlated to anisotropic ion transport within the same film. surface morphology of LbL films assembled on silicon. NCH
To the best of our knowledge, this study represents the first Pointprobe AFM cantilevers were purchased from Pacific Nano-
application of EIS in determining anisotropic ion transport in technologies.
LbL assemblies. SEM. Images were captured using a Carl Zeiss LEO field-emission
SEM system operating between 1 and 5 keV. Two nanometers of
Experimental Section Au-Pd was sputter-deposited on the samples before imaging to
suppress charging. Cross section images were taken from samples
Solution Preparation. Poly(ethylene oxide) of 4 000 000 mo-
cleaved using a diamond scribe.
lecular weight (MW) and linear poly(ethylene imine) of 25 000
WAXD. A Rigaku RU300 X-ray diffractometer (Cu KR, )
MW were purchased from Polysciences. Polymer solutions of PEO
1.541 ) was used for both powder diffraction and glancing angle
and PEI were separately made using polymer and Milli-Q water.
WAXD. Powder diffraction of Laponite clay and thin film diffraction
The concentration of polymer was 0.02 M based upon monomer
unit. The pH of PEI solution was adjusted to 5.00 ( 0.01 using of the LbL assembly on silicon were conducted in ambient conditions
(25 C and 30% relative humidity). Scans were conducted from 2
hydrochloric acid and a Beckman Coulter 390 pH meter.
) 3 to 50 at a rate of 0.01 /s.
Laponite RD, a synthetic hectorite, was purchased from EECS
GI-SAXS. Experiments were performed at the G1 beamline at
Cosmetics, and the manufacturer-reported diameter and thickness
were 30 and 1 nm, respectively.44 A dispersion of 0.5 wt % the Cornell High Energy Synchrotron Source. The wavelength of
the incident beam was 1.239 with a sample to detector distance
Laponite in Milli-Q water was made and stirred overnight. Laponite
of 1752 mm, and a 2-D area detector was used for data collection.46
purchased from the manufacturer contained (exchangeable) sodium
Electrochemical Impedance Spectroscopy. Cross-plane imped-
cations.45
ance measurements were conducted using a cell described by
DeLongchamp and Hammond.47 Briefly, patterned ITO-coated glass
(30) Ferguson, G. S.; Kleinfeld, E. R. AdV. Mater. 1995, 7 (4), 414-416.
(Donnelly and DCI) was used as the substrate for LbL assembly.
(31) Kleinfeld, E. R.; Ferguson, G. S. Chem. Mater. 1995, 7 (12), 2327-2331.
(32) Kleinfeld, E. R.; Ferguson, G. S. Chem. Mater. 1996, 8 (8), 1575-1578. Following LbL assembly atop the ITO-coated glass, gold electrodes
(33) Lvov, Y.; Ariga, K.; Ichinose, I.; Kunitake, T. Langmuir 1996, 12 (12), (100-nm thick and 2-mm wide) were thermally evaporated using an
3038-3044.
Edwards Auto 306. Copper tape from 3 M was applied to the gold
(34) Kotov, N. A.; Haraszti, T.; Turi, L.; Zavala, G.; Geer, R. E.; Dekany, I.;
to form a contact pad. The active area was 6 mm2.
Fendler, J. H. J. Am. Chem. Soc. 1997, 119 (29), 6821-6832.
(35) Kotov, N. A.; Magonov, S.; Tropsha, E. Chem. Mater. 1998, 10 (3), In-plane conductivity measurements were performed on LbL films
886-895. deposited on independently addressable microband electrodes
(36) MacNeill, B. A.; Simmons, G. W.; Ferguson, G. S. Mater. Res. Bull.
(IAMEs) from Abtech Scientific. Each ITO band was 3-mm long
1999, 34 (3), 455-461.
with 5- m width and spacing. The active area was given by 3 mm
(37) van Duffel, B.; Schoonheydt, R. A.; Grim, C. P. M.; De Schryver, F. C.
Langmuir 1999, 15 (22), 7520-7529. times the thickness of the LbL film.
(38) Glinel, K.; Laschewsky, A.; Jonas, A. M. Macromolecules 2001, 34 (15), Impedance measurements were performed using a Solartron 1260.
5267-5274.
The ac amplitude was 100 mV to improve the signal-to-noise ratio
(39) Glinel, K.; Laschewsky, A.; Jonas, A. M. J. Phys. Chem. B 2001, 106
(43), 112**-*****. at high impedance. A linear sweep of the cross-plane and in-plane
cells from -100 to 100 mV gave a linear current response, confirming
(40) Tang, Z.; Kotov, N. A.; Magonov, S.; Ozturk, B. Nat. Mater. 2003, 2 (6),
413-418. that impedance measurements at this amplitude are appropriate.
(41) Vuillaume, P. Y.; Glinel, K.; Jonas, A. M.; Laschewsky, A. Chem. Mater.
2003, 15 (19), 3625-3631.
(42) Ku, B.-C.; Froio, D.; Steeves, D.; Kim, D. W.; Ahn, H.; Ratto, J. A.; (45) Southern Clay Products, Inc. Laponite RD Product Bulletin; Southern
Blumstein, A.; Kumar, J.; Samuelson, L. A. J. Macromol. Sci., Pure Appl. Chem. Clay Products: Gonzales, TX; http://www.scprod.com/product_bulletins/
2004, 41 (12), 1401-1410. PB%20Laponite%20RD.pdf.
(43) Kim, D. W.; Blumstein, A.; Kumar, J.; Samuelson, L. A.; Kang, B.; Sung, (46) Roe, R. J. Methods of X-ray and Neutron Scattering in Polymer Science;
C. Chem. Mater. 2002, 14 (9), 3925-3929. Oxford University Press: New York, 2000.
(44) Porion, P.; Faugere, A.-M.; Delville, A. Eur. Phys. J. Special Topics (47) DeLongchamp, D. M.; Hammond, P. T. Chem. Mater. 2003, 15 (5),
2007, 141, 281-284. 1165-1173.
Structure and Transport in Polymer-Clay Nanocomposites Langmuir, Vol. 23, No. 16, 2007 8517
Table 1. Thickness per LbL Cycle
LbL system per cycle
54 ( 4
PEI/clay/PEO
47 ( 3
PEI/Li-clay/PEO
11 ( 2
PEI/clay
-
PEO/clay
-
PEI/PEO
observed because initial layers of polymer and clay form
nucleation islands until a uniform coating covers the substrate.49
After 10 trilayers, the substrate no longer affects polymer and
clay adsorption, and film growth proceeds uniformly. The
observed linear growth profile suggests that each cycle results
in the deposition of the same amount of material on the substrate.
The cycle thickness for each of the three systems implies that
the clay platelets adsorb in flat or slightly tilted layers. Given
Figure 1. Growth profiles of PEI/clay, PEI/clay/PEO, and PEI/ a cycle thickness of 54, 47, or 11 (Table 1) and a manufacturer-
Li-clay/PEO from ellipsometry and profilometry show that films reported platelet diameter of 30 nm, we calculated using
grow linearly with 11, 54, and 47 of material per trilayer,
Pythagorean theorem that a single clay platelet may tilt as much
respectively.
as 10, 9, or 2, respectively, relative to the substrate (Supporting
Information and Figure 1). This behavior can be explained by
Samples probed at 53% humidity were enclosed in a sealed
the negative charge associated with the platelet face.50 By aligning
chamber with magnesium nitrate hexahydrate salts to maintain
face-down, platelets can maximize interaction with the positively
humidity. For dry measurements, an argon-filled glovebox with 2
charged PEI layer. In contrast, the platelet edge has a positive
ppm water was used. A home-built Faraday cage and hotplate system
polarity50 that discourages edge-up adsorption.
allowed control of cell temperature within the glovebox. In both
cases, cells were allowed to equilibrate with the box environment Both neat and lithium-exchanged clay composites produced
for 24 h before electrochemical measurements. smooth films, as measured by profilometry. A film of (PEI/
Lithium-Exchanged Clay Preparation. Lithium-exchanged clay clay/PEO)40 exhibited a root-mean-square (rms) roughness of
(Li-clay) was prepared according to Khan et al.12 Briefly, Laponite 2.4 nm, and a film of (PEI/Li-clay/PEO)40 exhibited an rms
RD clay was dissolved in Milli-Q water. Excess lithium chloride roughness of 3.0 nm by profilometry.
was added to the dispersion to facilitate ion exchange. The clay
Because films of only PEO and clay were gel-like and
suspension was centrifuged, and the opaque gel was retained. The
unprocessable, we sought to include a third component that would
dissolution and exchange process was repeated twice more. The
stabilize film formation via mutual interactions between PEO
final gel was heated at 100 C and then washed with methanol until
and clay. PEI, partially charged at pH 5,47 was chosen as the
a drop of silver nitrate in the effluent remained clear to confirm the
stabilizing component because of its ability to interact with
complete removal of chloride ions. The resulting white powder was
dried at 80 C to give the final product, Li-clay, with a calculated Laponite via electrostatic interactions and PEO via ion-dipole
yield of 60 wt %. interactions and hydrogen bonding. For comparison, cycle
thickness increases from 11 for (PEI/clay) to 54 for (PEI/
Results and Discussion clay/PEO), which is indicative of incorporation of PEO in the
stabilized LbL film.
Multilayer Growth Mechanism and Rate. In a desire to
We hypothesize the following mechanism for film formation
produce an LbL film composed of a polymer electrolyte and a
in the PEI/clay/PEO trilayer system. Positively charged PEI
single-ion conductor, negatively charged Laponite clay was
adsorbs from solution to a negatively charged silicon substrate
selected as a well-investigated and robust single-ion conductor.12
to yield a positive substrate surface charge. Negatively charged
In this study, neutral PEO and positively charged linear PEI
Laponite then adsorbs to the PEI-coated substrate, reversing the
were both selected as candidate polymer electrolytes.24 PEI47
surface charge. Third, PEO adsorbs to the Laponite-coated surface
and PEO,48 owing to their polar backbones, have demonstrated
from solution. It is believed that PEO and clay associate through
promising ionic conductivities when used as a component in
hydrogen bonding and the desorption of water along the platelet
electrostatic47 and hydrogen bonding48 LbL electrolyte films.
surface.21,22 These three steps comprise a single deposition cycle,
Attempts to create LbL structures from neutral PEO and negatively
resulting in a single trilayer of PEI/clay/PEO. Weak association
charged Laponite clay were unsuccessful, owing to the formation
between PEO (deposited during the nth deposition cycle) and
of a thixotropic gel during deposition. Also, multilayer formation
PEI (deposited during the n + 1 cycle) through ion-dipole
from positively charged PEI and neutral PEO was unsuccessful.
interactions and hydrogen bonding ensures adhesion between
Film thickness as a function of cycle number n was investigated
successive trilayers. Thus, positively charged PEI is used to
using profilometry and ellipsometry for three systems: (PEI/
associate with both negatively charged Laponite and hydrogen-
clay), (PEI/clay/PEO), and (PEI/Li-clay/PEO). Figure 1 dem-
bonding PEO to create stable and cohesive thin films.
onstrates a representative growth profile for these three systems,
Film Characterization. X-ray photon spectroscopy (XPS) of
where each film thickness was measured after drying. In each
(PEI/Li-clay/PEO)60 was used to quantify the composition of the
case, a linear slope was obtained, where the thickness per cycle
LbL assembly. On the basis of the relative XPS signals of
was taken as the slope of the growth profile (Table 1).
magnesium from the clay and carbon and nitrogen from the
The shape of the growth profile of each composite resembled
polymers, the LbL assembly contained 66, 30, and 4 wt % Li-
previously reported curves for polymer-clay layer-by-layer
assemblies.33,41 At early deposition cycles, little growth was
(49) Jeon, J.; Panchagnula, V.; Pan, J.; Dobrynin, A. V. Langmuir 2006, 22
(10), 4629-4637.
(48) DeLongchamp, D. M.; Hammond, P. T. Langmuir 2004, 20 (13), 5403- (50) Baghdadi, H. A.; Sardinha, H.; Bhatia, S. R. J. Polym. Sci., Part B:
Polym. Phys. 2005, 43 (2), 233-240.
5411.
8518 Langmuir, Vol. 23, No. 16, 2007 Lutkenhaus et al.
clay, PEO, and PEI, respectively, when clay was the topmost
layer as well as when PEO was the topmost layer. Lithium atoms
could not be detected owing to their low concentration and the
weak XPS signal of the Li 1s orbital. Of note, sodium was present
in low concentrations, 0.04 wt %, and chlorine was undetectable.
Tapping mode AFM characterized the surface features of an
LbL film of (PEI/clay/PEO)60 in which clay was the topmost
layer. Circular and oblong features were observed in both the
height and phase images (Figure 2a,b, respectively). The diameters
of these features (40 to 60 nm) are slightly larger than the
manufacturer s reported diameter of the clay platelet (30 nm),
though this difference may be attributed to artifacts from the
cantilever tip. Rms roughness from an 800-nm square height
image was 3.5 nm, whereas profilometry gave a roughness of
2.4 nm.
Cross-sectional SEM (Figure 2c) of a (PEI/Li-clay/PEO)200
assembly further suggests our proposed layered structure. Bright
regions are associated with clay platelets, while dark regions are
likely polymer. In the micrograph, the edges of individual clay
platelets appear to lay parallel to the silicon substrate, while the
top of the multilayer film appeared edge-on as a smooth surface.
This micrograph is similar to images reported for layered
montmorillonite/PDAC LbL structures.40
Structural Analysis Using WAXD and GI-SAXS. WAXD
was performed on neat Laponite powder, an LbL film of (PEI/
Clay/PEO)60, and an LbL film of (PEI/Li-clay/PEO)60 in which
clay was the topmost layer for both films. Figure 3a shows the
-2 plot obtained from WAXD. Neat Laponite powder
exhibited one shoulder and three distinct peaks in the scan range
shown, consistent with previous reports of Laponite.51 The low-
angle shoulder at 6.8 corresponds to a basal (001) spacing of
13.0, which is considered the periodic distance from platelet
to platelet. For example, if the platelets are 1-nm thick, as reported
by the manufacturer, the gallery spacing (or the distance between
stacked platelets) is 3 . The Scherrer equation,52 which estimates
crystallite size or range of order, could not be used here because
the shape of the basal reflections was not well-defined.
In the (PEI/clay/PEO)60 film, the low-angle peak shifted to
6.3 with a basal spacing of 14.0, where the distance between
platelets increased slightly to 4 . Two low-intensity higher-
angle peaks (19.3 and 26.2 ) appear at angles similar to those
observed in neat Laponite (20.0 and 28.0 ). Multilayers containing
lithium-exchanged Laponite, (PEI/Li-clay/PEO), exhibited peaks
identical to multilayers containing unlithiated clay, (PEI/clay/
PEO).
Further evidence of the (PEI/Li-clay/PEO) structure was
obtained using GI-SAXS (Figure 3b), which measures the
orientation of periodic structure within a thin film. The off-
specular scattering can be analyzed for incidence angles close
to the critical angle of total external reflection of the composite,
revealing both lateral structure within the film and structure normal
to the substrate.53 The peak scattering intensity was observed at
q ) 4.48 nm-1, corresponding to a basal spacing of 14, which
Figure 2. AFM height (a) and phase (b) images of a (PEI/clay/
is similar to our WAXD observation. Because the observed
PEO)60 film where clay is the top most layer, 800 nm square with
scattering was preferentially along the film normal, the results 30 nm and 30 scale. (c) Cross-sectional SEM of (PEI/Li-clay/
suggest that Laponite platelets are oriented parallel to the substrate. PEO)200.
The Hermans orientation parameter (f)46,54 was used to quantify
completely random distribution of orientations. When f is 1 or
the degree of orientation within the LbL assembly. This parameter
-1/2, the system is completely aligned parallel or perpendicular,
ranges from 1 to -1/2, in which a value of zero indicates a
respectively, to the chosen reference direction (in this case, normal
(51) Le Luyer, C.; Lou, L.; Bovier, C.; Plenet, J. C.; Dumas, J. G.; Mugnier,
to the substrate). The intensity of the scattering at the scattering
J. Opt. Mater. 2001, 18 (2), 211-217.
vector q in question was analyzed, and the Hermans orientation
(52) Warren, B. E. X-Ray Diffraction; Addison-Wesley: Reading, MA, 1969.
(53) Busch, P.; Rauscher, M.; Smilgies, D.-M.; Posselt, D.; Papadakis, C. M. parameter was found to be 0.7, indicating that the platelets within
J. Appl. Crystallogr. 2006, 39 (3), 433-442.
the LbL assembly have significant, but imperfect, orientation
(54) Finnigan, B.; Jack, K.; Campbell, K.; Halley, P.; Truss, R.; Casey, P.;
Cookson, D.; King, S.; Martin, D. Macromolecules 2005, 38 (17), 7386-7396. parallel to the silicon substrate.
Structure and Transport in Polymer-Clay Nanocomposites Langmuir, Vol. 23, No. 16, 2007 8519
Figure 4. Representative Nyquist and Bode (inset) plot of (PEI/
Li-clay/PEO)60 at 170 C. Data fit to above model give an electrolyte
resistance (R2) of 62 000 . With a cell constant of (L/A) ) 0.0045
cm-1, the resulting conductivity is 7.3 10-9 S cm-1.
cycle (from WAXD and GI-SAXS), we propose that the
multilayer structure of PEI/Li-clay/PEO consists of alternate,
stratified layers of polymer and clay. From XPS, it is believed
that the majority of the polymer content to be PEO (as stated
earlier, multilayers contained 66, 30, and 4 wt % Li-clay, PEO,
and PEI, respectively). We hypothesize that anisotropic structure
of the LbL film, as confirmed by AFM, SEM, WAXD, GI-
SASXS, and growth profiles, influences ionic conductivity with
respect to orientation. This was investigated using EIS, detailed
below.
Ionic Conductivity and Electrochemical Impedance Spec-
troscopy. EIS is a useful tool for investigating the movement
and transport of ions (e.g., conductivity) within an electrolyte.
Multilayered composites of (PEI/Li-clay/PEO)60 were probed
using EIS to measure ionic conductivity as a function of
temperature, humidity, and orientation. Both cross-plane and
in-plane conductivities were investigated at 53% and 0% relative
humidity (RH).
The impedance response of (PEI/Li-clay/PEO)60 was measured
in two different cells to isolate the cross- and in-plane directions.
Cross-plane (z direction) ion transport was measured in a cell
consisting of multilayers deposited on patterned ITO glass and
Figure 3. (a) WAXD of Laponite clay powder in blue (bottom), gold electrodes evaporated atop the multilayer film; in-plane
PEI/Li-clay/PEO in green (middle), and PEI/clay/PEO in pink (top).
(x-y direction) ion transport was measured using independently
(b) The shape of the GI-SAXS pattern of PEI/Li-clay/PEO indicates
addressable ITO microband electrodes (IAMEs). A typical
orientation parallel to the substrate surface. (c) Proposed structure
Nyquist plot for a cross-plane cell (Figure 4) gave a depressed
of PEI/Li-clay/PEO LbL assembly. The trilayer thickness is 47
(from growth profile), the basal spacing is 14 (from GI-SAXS and semicircle at high frequency and a near-vertical line at lower
frequency, which is similar to previous reports of PEO-clay
WAXD), and the gallery spacing is 4 (basal spacing minus clay
platelet thickness, 14 - 10 ) 4 ). composites;7 in-plane measurements gave a similar response.
This high frequency behavior, previously described,4,12 is best
The peaks from the LbL assemblies observed in WAXD and modeled using a resistor and constant-phase element (CPE) in
GI-SAXS suggest periodic structure within the film (i.e., the parallel, preceded by a resistor and CPE in series to capture low
clay is not exfoliated). The small increase in gallery spacing frequency domains (equivalent circuit in Figure 4). R1 and R2
from 3 to 4 (neat Laponite and PEI/Li-clay/PEO, respectively) represent the electrode resistance and multilayer assembly
does not indicate complete intercalation of polymer between resistance, respectively. CPE1 describes the nonideal capacitive
double layer, most likely caused by a rough electrode-electrolyte
individual platelets; however, the presence of the low-angle peaks
indicates that, for each clay-deposition step, clay platelets are interface, and CPE2 corresponds to bulk polarization of the LbL
adsorbed in multiple layers, not a monolayer. Given an LbL film. To check cell design and self-consistency, samples of varying
thickness (200-300 nm) were constructed; electrolyte resistance,
assembly growth rate of 47 per trilayer (from ellipsometry and
profilometry), a periodic length scale of 14 from clay platelet- R2, scaled linearly with thickness, as expected.
to-platelet (from WAXD and GI-SAXS), lateral orientation (from Given the equivalent circuit, described above, and the
GI-SAXS), and at least two layers of clay adsorbed per trilayer impedance response of our multilayers at 53% and 0% RH and
8520 Langmuir, Vol. 23, No. 16, 2007 Lutkenhaus et al.
activation energies, which are related to the slope of log vs
1/T, for in- and cross- plane conductivities were 0.37 and 0.35
eV (36 and 33 kJ/mol), respectively; these numbers compare
well with the activation energy of Li+ in PEO, which ranges
from 0.2 to 0.3 eV.56,57 Alternatively, the activation energy for
ion transport in Li+-montmorillonite is 1 eV.7 Given an
observed activation energy of 0.35-0.37 eV and assuming
Laponite behaves similarly to montmorillonite, we propose that
ion transport in (PEI/Li-clay/PEO) mirrors that of Li+ transport
in PEO. Because both in- and cross-plane activation energies are
similar to that of Li+-PEO, we propose that chain segments of
PEO may participate in the cross-plane ion transport process as
PEO serves to bridge between clay platelets.
Despite the similar in- and cross-plane conduction activation
energies, cross-plane conductivity was 100 times less than in-
plane conductivity (Figure 5a). This is explained by the tortuous
path the small lithium ion (r ) 0.68 ) must travel to migrate
in the z-direction (Figure 5b), weaving around oriented clay
platelets (r ) 15 nm). Further evidence of a tortuous path is
present in differences observed in R of the CPE2, ZCPE ) 1/[Q
(j )R]. The CPE represents a distribution of time constants for
ion transport, and the resistive or capacitive character of the
response is described by R, which ranges between 0 and 1.58
Figure 5. (a) Arrhenius plot of the variation of conductivity with
In-plane measurements were nearly capacitive with R ) 0.97,
temperature from 30 to 200 C of (PEI/Li-clay/PEO) assemblies in
meaning there is one mode of ion transport. Cross-plane
a dry argon glovebox. In-plane conductivity (pink *) is consistently
measurements, R ) 0.7 to 0.8, were less capacitive in character
100 times higher than cross-plane conductivity (blue x). The similar
and pointed to mixed time constants of ion transport. Indeed, in
slopes (dashed lines) indicate comparable activation energies of
0.35 and 0.37 eV for cross- and in-plane conductivities, respectively. the cross-plane, multiple time constants are possible because an
(b) Cross-plane ion conduction is hindered by the presence of ordered ion has many tortuous paths to choose from, whereas with in-
clay nanoplatelets, while in-plane ion conduction is unhindered.
plane conduction, ion transport occurs relatively uninterrupted
in one direction.
Table 2. In-Plane and Cross-Plane Conductivity at 0% and
As suggested by Ruiz-Hitsky and Aranda and others,6,13-15
53% RH and 25 C
polymer-clay composites are single-ion conductors. The rela-
(S cm-1) at 53% RH (S cm-1) at 0% RH
orientation
tively large anionic clay nanoplatelets are virtually immobile
2.6 ( 0.2 10-7
in-plane (z) a
compared to the facile lithium ion, yielding an ideal lithium
cross-plane (x-y) 4 ( 1 10-8 7 ( 1 10-13
transference number of unity. The LbL films discussed here are
a Exceeded limits of impedance analyzer. potential single-ion conductors, where the lithium cation is
solvated by PEI and PEO24 and charge balanced by negatively
25 C, we calculated in- and cross-plane conductivity (Table 2) charged Laponite clay. However, we were unable to measure the
using ) L/(R2 A), where L is the distance between electrodes transference number of the LbL system because a cell of Li LbL-
and A is the area between electrodes. The ratio of the in- and film Li, necessary for this measurement, could not be constructed
cross-plane measurements gives the anisotropy factor; films at owing to difficulties in isolating the LbL film. Future efforts aim
53% RH exhibited an anisotropy factor of 7 (from Table 2). The at resolving the challenge of lifting-off or isolating the film to
anisotropy factor for films at 0% RH and 25 C could not be allow the measurement of the transference number as well the
calculated because in-plane measurements at these conditions mechanical and transport properties.
exhibited impedance that exceeded the limits of the analyzer. Of
note, while electrode resistance (R1) remained constant with Conclusion
increasing humidity, the multilayer resistance (R2) dramatically
In summary, polymer-clay nanocomposites of PEI, Laponite
decreased (i.e., LbL conductivity increased). This behavior can
clay, and PEO were constructed using LbL assembly technique;
be explained by the presence of water within the LbL assembly:
each trilayer was 5 nm in thickness, and clay platelets appeared
at 53% RH, the ionic conductivity is expected to be primarily
to lay face-down relative to the substrate. Anisotropic structure
protonic because water adsorbed along the platelet faces is
of the films was confirmed using multiple techniques (GI-SAXS,
predominately acidic;55 however, in dry conditions, the solvated
WAXD, AFM, SEM). This system is thought to be built upon
cation, Li+, is considered the mobile species.6,7
hydrogen bonding and electrostatic interactions among the three
To further understand ion transport in (PEI/Li-clay/PEO)60 in
components. Anisotropic ion transport, resulting from anisotropic
the dry state (0% RH), in- and cross-plane conductivities were
structure, was investigated using EIS, which demonstrated in-
measured as a function of temperature (30-200 C). The
plane ionic conductivities 100 times faster than cross-plane
temperature response of both cross- and in- plane conductivities
conductivities (at 0% RH). The activation energy associated with
(Figure 5a) exhibited Arrhenius behavior. Results were repro-
ducible from sample to sample with no hysteresis from thermal
(56) Robitaille, C. D.; Fauteux, D. J. Electrochem. Soc. 1986, 133 (2), 315-
cycling. Below 115 C, the impedance of the IAMEs for in- 325.
plane measurements exceeded the limits of the analyzer. The (57) Chung, S. H.; Jeffrey, K. R.; Stevens, J. R. J. Chem. Phys. 1991, 94 (3),
1803-1811.
(58) Orazem, M. E. Impedance Spectroscopy Short Course, 209th Meeting of
(55) Slade, R. C. T.; Barker, J.; Hirst, P. R.; Halstead, T. K. Solid State Ionics the Electrochemical Society, Denver, CO, May 7, 2006; The Electrochemical
1987, 24 (4), 289-295. Society: Denver, CO, 2006; pp 3-9.
Structure and Transport in Polymer-Clay Nanocomposites Langmuir, Vol. 23, No. 16, 2007 8521
ion transport in (PEI/Li-clay/PEO) (0.35-0.37 eV) at 0% RH introduction of anisotropy in gas permeability or mechanical
was similar to that of lithium cations in PEO. When humidity properties for additional applications.
was increased from 0% RH to 53% RH, observed cross-plane
Acknowledgment. We thank the Dupont-MIT Alliance for
conductivity increased (from 7 10-13 to 4 10-8 S cm-1,
funding. We thank the Institute of Soldier Nanotechnology, Center
respectively) and the degree of anisotropic transport decreased
for Materials Science and Engineering, and Prof. Yang Shao-
(from 100 to 7, respectively) at 25 C. With regard to LbL
Horn for facilities. We thank Dr. Piljin Yoo and Kathleen McEnnis
assemblies, this study represents a first correlation of structural for their assistance. Dr. Lutkenhaus thanks the National Science
anisotropy to transport anisotropy using EIS. Foundation Graduate Fellowship for support. This work is based
Recommendations for refining the PEI/clay/PEO system for upon research conducted at the Cornell High Energy Synchrotron
practical use as an electrolyte include (1) adding a plasticizer Source, which is supported by the National Science Foundation
such as ethylene carbonate (increasing charge mobility), (2) and the National Institutes of Health/National Institute of General
adding a Li + salt to increase Li+ concentration, and (3) creating Medical Sciences under Award DMR-0225180.
exfoliated or disorganized LbL