Nanoengineering mechanically robust aerogels via control of foam morphology

Article abstract of DOI:10.1063/1.2236222

Potential of aerogels for technological applications is often limited by their poor mechanical properties. Here, we demonstrate that alumina aerogel monoliths with excellent mechanical properties can be made by controlling the crystallographic phase, shape, and size of nanoligaments. In particular, we show that thermal processing of aerogels with a morphology of interconnected nanoleaflets causes dehydration and associated curling of the nanoleaflets, resulting in a dramatic improvement of mechanical properties. This study shows an effective way to control mechanical properties of the nanoporous solids that can be synthesized with ligaments having a quasi-two-dimensional shape, such as platelets, ribbons, or leaflets.

Full text of DOI:10.1063/1.2236222

Nanoengineering mechanically robust aerogels via control of foam morphology
S. O. Kucheyev, T. F. Baumann, C. A. Cox, Y. M. Wang, J. H. Satcher Jr., A. V. Hamza, and J. E. Bradby
Citation: Applied Physics Letters 89, 041911 (2006); doi: 10.1063/1.2236222
View online: http://dx.doi.org/10.1063/1.2236222
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/89/4?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Controlled nanoporous Pt morphologies by varying deposition parameters
Appl. Phys. Lett. 95, 073116 (2009); 10.1063/1.3194109
Thermal, Morphological and Rheological Properties of Rigid Polyurethane Foams as Thermal Insulating
Materials
AIP Conf. Proc. 1027, 905 (2008); 10.1063/1.2964889
Reduction in helium thermal conductivity by 1 mg ∕ cc silica aerogel foam
Appl. Phys. Lett. 92, 221913 (2008); 10.1063/1.2938883
Morphology of Si Nanowire Grown via SolidLiquidSolid Mechanism
AIP Conf. Proc. 893, 47 (2007); 10.1063/1.2729763
Controlled morphology of biologically derived metal nanopatterns
Appl. Phys. Lett. 71, 1264 (1997); 10.1063/1.119868
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
155.33.148.196 On: Mon, 01 Dec 2014 20:37:37

APPLIED PHYSICS LETTERS 89, 041911 ͑2006͒
Nanoengineering mechanically robust aerogels via control
of foam morphology
S. O. Kucheyev,a͒ T. F. Baumann, C. A. Cox, Y. M. Wang,
J. H. Satcher, Jr., and A. V. Hamza
Lawrence Livermore National Laboratory, Livermore, California 94550
J. E. Bradby
Department of Electronic Materials Engineering, The Australian National University,
Canberra ACT 0200, Australia
͑Received 20 March 2006; accepted 5 June 2006; published online 26 July 2006͒
Potential of aerogels for technological applications is often limited by their poor mechanical
properties. Here, we demonstrate that alumina aerogel monoliths with excellent mechanical
properties can be made by controlling the crystallographic phase, shape, and size of nanoligaments.
In particular, we show that thermal processing of aerogels with a morphology of interconnected
nanoleaflets causes dehydration and associated curling of the nanoleaflets, resulting in a dramatic
improvement of mechanical properties. This study shows an effective way to control mechanical
properties of the nanoporous solids that can be synthesized with ligaments having a
quasi-two-dimensional shape, such as platelets, ribbons, or leaflets. © 2006 American Institute of
Physics. ͓DOI: 10.1063/1.2236222͔
Aerogels ͑AGs͒ are ultralow-density open-cell nano-
foams with typical porosities of տ90%.1 In AGs, the mor-
phology of a sol-gel-derived wet gel is maintained after the
liquid filling the pores is replaced with air during supercriti-
cal drying.1 With very high surface areas, ultralow densities,
and control over the elemental composition of nanoliga-
ments, potential applications of AGs are numerous. These
include advanced gas- and biosensors, batteries, catalytic de-
vices, thermal insulators, and low dielectric constant materi-
als for integrated circuits.1–7 For example, alumina AGs, the
subject of this study, are attractive for high performance cata-
lysts, ceramics, and separators,3,6,7 and alumina AG mono-
liths are desirable as high-temperature thermal insulators for
alloy melt processing and in space applications.6 However,
AGs are notoriously known for their poor mechanical prop-
erties ͑see, for example, Refs. 1 and 6–15͒. In fact, most
non-silica-based AG monoliths with porosities տ95% are
very fragile and difficult to handle, which limits their current
applications.6,7
Although we are not aware of any previous attempts to
improve mechanical properties of inorganic non-silica-based
AGs, there have been several such studies for silica
AGs.9,11,13–16 One approach involved soaking silica gels in
monomers which condense into the necks between the
particles.9 Prakash et al.16 modified silica AGs by covering
ligaments with silyl groups, which are chemically inert com-
pared to the hydroxyl groups passivating the surface of typi-
cal silica AGs.10 Other attempts to toughen silica AGs in-
volved adding organic polymer molecules,13 silica and
alumina microfibers,11 or activated carbon powders.14 How-
ever, in most cases, the apparent improvement of mechanical
properties could be largely attributed to an increase in the
monolith density as a result of adding cross-linking
material.11,13,14
with excellent mechanical properties can be achieved by syn-
thesizing foams with a morphology of interconnected
nanoleaflets. Thermal annealing of such AGs causes con-
trolled material dehydration and associated curling of
nanoleaflets, resulting in a significant improvement of me-
chanical properties.
Four sets of alumina AGs were synthesized, labeled
CWE-1, CWE-2, CE, and NE ͑Table I͒.20 Details of the AG
synthesis can be found in Ref. 7. The average bulk densities
of AGs were determined by measuring the dimensions and
mass of monolithic samples. Selected AGs were annealed for
2 h in air at temperatures ͑Tann͒ of 500 and 800 °C and
remained monolithic and apparently crack-free. The AGs
were studied by x-ray diffraction ͑XRD͒ and transmission
electron microscopy ͑TEM͒. All four sets of AGs had
Brunauer-Emmett-Teller surface areas of ϳ700 m2 g−1, and
annealing at 800 °C resulted in an expected reduction in
surface areas to ϳ300–400 m2 g−1.7
All AGs were subjected to indentation using a UMIS-
2000 nanoindentation system with an ϳ1-mm-radius sap-
phire spherical indenter. A series of both continuous and par-
tial load-unload indents ͑with ϳ5–10 indents for each
loading mode for every sample͒ was carried out at room
temperature. The loading rates were ϳ1–3 mN/s. Indenta-
tion curves for different indents on the same sample were
generally reproducible within ϳ10%. Such relatively large
error bars do not affect any of the conclusions of this work
and are attributed to effects of surface roughness, which is
difficult to measure and control for ultralow-density nano-
porous solids. Indentation data were analyzed using the
method of Field and Swain21 to extract the ͑Meyer͒ hardness
H and elastic modulus E. To verify the E values obtained
from nanoindentation, selected samples were also studied
with Panametrics Ultrasonic analyzer ͑model 5052 UA͒ with
180 kHz center frequency transducers. For such measure-
ments, crack-free AGs were cast in shapes of right cylinders,
ϳ1.2 cm in diameter and ϳ1.5–3.0 cm in length, with par-
allel cylinder faces. A procedure from Ref. 8 was used to
calculate E from the sound velocities measured.
Here, we report on an entirely different approach to de-
sign mechanically robust AGs by controlling their morphol-
ogy. We demonstrate that crack-free, alumina AG monoliths
a͒
Electronic mail: kucheyev@llnl.gov
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
0003-6951/2006/89͑4͒/041911/3/$23.00 89, 041911-1 © 2006 American Institute of Physics
155.33.148.196 On: Mon, 01 Dec 2014 20:37:37

041911-2
Kucheyev et al.
Appl. Phys. Lett. 89, 041911 ͑2006͒
TABLE I. Properties of full-density alumina ͑Refs. 17–19͒ and alumina aerogels with nanoleaflet ͑NL͒ and string-of-pearls ͑SP͒ morphologies. Effective
hardness H and elastic modulus E were determined from nanoindentation data for a plastic penetration depth of 5 ␮m. Values of E given in parentheses were
obtained from acoustic velocity measurements.
Density
Tann
E
H
͑kPa͒
Sample
CWE-1
Precursor
Solvent
͑°C͒ Morphology
Phase
͑mg cm−3
͒
͑%͒
͑MPa͒
AlCl3 ·6H2O
AlCl3 ·6H2O
AlCl3 ·6H2O
AlCl3 ·6H2O
AlCl3 ·6H2O
AlCl3 ·6H2O
AlCl3 ·6H2O
Al͑NO3͒3 ·9H2O
Al͑NO3͒3 ·9H2O
¯
Water/ethanol
Water/ethanol
Water/ethanol
Water/ethanol
Ethanol
20 NL
500 NL
800 NL
20 NL
20 NL
500 NL
800 NL
20 SP
Boehmite
48
55
1.41 1.75
1.62 3.35
39.10
CWE-1
CWE-1
CWE-2
CE
␥+ boehmite
144.00
227.00
158.00
87.80
␥
80
2.35 7.18 ͑12.4͒
1.77 4.14 ͑3.2͒
3.82 5.70
Boehmite
60
Boehmite
130
150
240
135
200
CE
Ethanol
␥+ boehmite
4.41 9.64
132.00
1170.00
103.00
60.20
CE
Ethanol
␥
7.06 80.20
3.97 4.37
NE
Ethanol
Diaspore
NE
Ethanol
800 SP
␥
5.88 4.89
Alumina
¯
¯
full density
¯
3400 100.00 2ϫ105
1ϫ107
A comparison of Figs. 1͑a͒ and 1͑d͒ illustrates that
changing the precursor from Al͑NO3͒3 to AlCl3 dramatically
changes the morphology of resultant AGs from the well
known sting-of-pearls morphology ͓Fig. 1͑d͔͒ to a morphol-
ogy for which the solid network is made from interconnected
nanoleaflets ͓Fig. 1͑a͔͒. The nanoparticles forming the AG
with the string-of-pearls morphology shown in Fig. 1͑d͒ are
ϳ5–10 nm in size, whereas the nanoleaflets shown in
Fig. 1͑a͒ are ϳ10–15 nm wide and Շ100 nm long. This
nanoleaflet morphology is common for all our AGs synthe-
sized with the AlCl3 precursor ͑i.e., CWE-1, CWE-2, and CE
AGs͒. However, our TEM study of CE AGs, for which ab-
solute ethanol was used as a solvent, revealed narrower
nanoleaflets ͑ϳ5 nm wide͒ than for the case of CWE-1 and
CWE-2 AGs ͓shown in Fig. 1͑a͔͒ prepared with a water and
ethanol mixture as a solvent ͑see Table I͒.
Figures 1͑a͒–1͑c͒ illustrate that thermal annealing at 500
and 800 °C of CWE-1 AGs ͑with the nanoleaflet morphol-
ogy͒ results in the curling of nanoleaflets. In contrast, our
TEM study of NE AGs ͑with the string-of-pearls morphol-
ogy͒ annealed at 800 °C has revealed a negligible change in
the size of nanoparticles forming the AG skeleton. In addi-
tion, our XRD study has shown that the nanoligaments in all
as-prepared AGs with the nanoleaflet morphology ͑CWE-1,
CWE-2, and CE͒ have the ␥-boehmite structure, while AGs
with the string-of-pearls morphology ͑NE͒ exhibit the
␣-diaspore structure. Both boehmite and diaspore are struc-
tures of AlO͑OH͒,3 indicating that as-synthesized AGs are
monohydroxides. After annealing at 800 °C, ligaments in all
AGs dehydrate and recrystallize to the ␥-Al2O3 structure, as
revealed by XRD ͑see Table I͒. The crystallographic phase of
nanoligaments ͑studied by XRD͒ can also explain their shape
revealed by TEM ͑Fig. 1͒. Indeed, ␥-AlO͑OH͒ ͑boehmite͒,
the phase of struts forming AGs with the nanoleaflet mor-
phology, is a layered structure with the layers held together
by hydrogen bonds, while ␣-AlO͑OH͒ ͑diaspore͒ in AGs
with the string-of-pearls morphology is not a layered
structure.3
To illustrate the dramatic effects of the monolith density
and foam morphology on mechanical properties of AGs, Fig.
2 shows curves of H ͑a͒ and E ͑b͒ for four representative
AGs as a function of indenter penetration below the circle of
contact. Values of E and H for all the samples are given in
Table I. In addition, the inset of Fig. 2͑a͒ shows typical con-
tinuous load-unload curves for the same four AGs.
The Meyer hardness, defined as the average contact
pressure, has different physical meanings for low-density
FIG. 2. ͑Color online͒ Curves of hardness ͑a͒ and elastic modulus ͑b͒ as a
function of indenter penetration depth, as determined from partial load-
unload data, demonstrating the effects of the monolith density, foam mor-
phology, and the phase of nanoligaments on properties of four representative
alumina aerogels. Typical continuous load-unload curves for the same four
FIG. 1. Bright-field TEM images of CWE-1 ͓͑a͒–͑c͔͒ and NE ͑d͒ aerogels
annealed at 20 ͓͑a͒ and ͑d͔͒, 500 ͑b͒, and 800 °C ͑c͒. All four images are of
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
the same magnification.
samples are shown in the inset.
155.33.148.196 On: Mon, 01 Dec 2014 20:37:37

041911-3
Kucheyev et al.
Appl. Phys. Lett. 89, 041911 ͑2006͒
only slightly increases E but, at the same time, results in a
slight reduction in H. Finally, the choice of the solvent dur-
ing the sol-gel synthesis ͑which, as discussed above, affects
the average width of the nanoleaflets͒ also affects AG me-
chanical properties ͑compare circles and up triangles in Fig.
3͒. Hence, it would not be appropriate to describe data from
Fig. 3 with scaling law models for mechanical properties of
porous solids12,22 since, in addition to the monolith density,
other important parameters are variable and strongly affect
mechanical characteristics.
The above results demonstrate that both the framework
connectivity and the mechanical properties of individual
nanoligaments in foams with a nanoleaflet morphology can
be improved by thermal processing, which causes dehydra-
tion, phase transformation, and associated curling of leaflets.
FIG. 3. ͑Color online͒ Dependencies of contact pressure ͑H, closed sym-
bols͒ and Young’s modulus ͑E, open symbols͒ on the density of the follow-
ing sets of aerogels: CWE-1 ͑circles͒, CWE-2 ͑stars͒, CE ͑up triangles͒, and
NE ͑down triangles͒. Data for full-density alumina are shown by diamonds.
brittle nanoporous solids than for full-density materials. In-
deed, the hardness for full-density ͑nonviscoelastic͒ solids is
defined by fundamental physical processes such as the gen-
eration and mobility of dislocations, slip, and/or pressure-
induced phase transformations. In contrast, deformation
mechanisms in brittle nanofoams ͑such as alumina AGs͒ are
currently poorly understood6,8–15 but likely involve elastic
bending and brittle fracture of struts, accompanied by the
collapse of pores and foam densification. In this case, hard-
ness is determined by the fracture toughness of struts rather
than by the dislocation and slip activities or phase transfor-
mations as for full-density solids.
It is seen from Fig. 2 that for AGs with the nanoleaflet
morphology, both E and H increase with increasing monolith
density. More importantly, for a given monolith density, AGs
with the nanoleaflet morphology are significantly stiffer and
have larger H than AGs with the string-of-pearls morphol-
ogy. Figure 2͑b͒ also shows that, except for indenter penetra-
tion depths Շ1 ␮m affected by surface roughness, E is es-
sentially independent of depth. However, H increases with
depth for all the samples ͓Fig. 2͑a͔͒, which we attribute to
the effects of strut fracture, associated collapse of pores, and
accumulation of the strut material beneath the indentation
contact with increasing load.
Table I also shows that our alumina AGs have
exceptional mechanical properties. For example, cross-
linked “strong” silica-based AGs with densities of
ϳ250–300 mg cm−3, recently reported by Leventis et al.13,
have E of ϳ5–12 MPa, which are comparable to values for
our alumina AGs with the string-of-pearls morphology of
similar densities, but which are an order of magnitude
smaller than E for our AGs with the nanoleaflet morphology.
The dependence of H and E on material parameters is
more clearly illustrated in Fig. 3. It is seen that both H and E
decrease superlinearly with decreasing relative density of
AGs, with slopes of 2.9 0.3 and 2.8 0.2, respectively.
However, it is clear that data for neither H nor E closely
follow a simple power law scaling with a constant exponent.
Hence, in addition to the monolith density, other parameters
͑in particular, the crystallographic phase and shape of liga-
ments͒ strongly influence the mechanical properties of the
nanofoam. For example, data from Fig. 3 and Table I clearly
show that thermal annealing of AGs with the nanoleaflet
morphology which causes curling of the nanoleaflets, results
in a dramatic improvement of mechanical properties. In con-
trast, annealing of AGs with the string-of-pearls morphology
Work at LLNL was performed under the auspices
of the U.S. DOE by the UC, LLNL under Contract No.
W-7405-Eng-48. Work at the ANU was supported by the
ARC.
1See, for example, a review by N. Hüsing and U. Schubert, Angew. Chem.,
Int. Ed. 37, 22 ͑1998͒.
2R. D. Miller, Science 286, 421 ͑1999͒.
3P. Euzen, P. Raybaud, X. Krokidis, H. Toulhoat, J. L. L. Loarer, J. P.
Jovilet, and C. Froidefont, in Handbook of Porous Solids, edited by F.
Schuth, K. S. W. Sing, and J. Weitkamp ͑Wiley-VCH, Weinheim, 2002͒,
p. 1591.
4D. R. Rolison, Science 299, 1698 ͑2003͒.
5J. L. Mohanan, I. U. Arachchige, and S. L. Brock, Science 307, 397
͑2005͒.
6J. F. Poco, J. H. Satcher, Jr., and L. W. Hrubesh, J. Non-Cryst. Solids 285,
57 ͑2001͒.
7T. F. Baumann, A. E. Gash, and J. H. Satcher, Jr., Chem. Mater. 17, 395
͑2005͒.
8J. Gross, G. Reichenauer, and J. Fricke, J. Phys. D 21, 1447 ͑1988͒.
9T. Woignier and J. Phalippou, J. Non-Cryst. Solids 100, 404 ͑1988͒.
10J. Fricke, Nature ͑London͒ 374, 409 ͑1995͒.
11K. E. Parmenter and F. Milstein, J. Non-Cryst. Solids 223, 179 ͑1998͒.
12H. S. Ma, A. P. Roberts, J. H. Prevost, R. Jullien, and G. W. Scherer, J.
Non-Cryst. Solids 227, 127 ͑2000͒.
13N. Leventis, C. Sotiriou-Leventis, G. Zhang, and A. M. M. Rawashdeh,
Nano Lett. 2, 957 ͑2002͒.
14M. Moner-Girona, E. Martinez, J. Esteve, A. Roig, R. Solanas, and E.
Molins, Appl. Phys. A: Mater. Sci. Process. 74, 119 ͑2002͒.
15E. M. Lucas, M. S. Doescher, D. M. Ebenstein, K. J. Wahl, and D. R.
Rolison, J. Non-Cryst. Solids 350, 244 ͑2004͒.
16S. S. Prakash, C. J. Brinker, A. J. Hurd, and S. M. Rao, Nature ͑London͒
374, 439 ͑1995͒.
17A skeletal density of alumina of 3.4 g cm−3 was used to estimate the rela-
tive density of AG monoliths since all the three phases of alumina studied
here have densities similar to that ͑3.4, 3.4, and 3.7 g cm−3 for boehmite,
diaspore, and ␥-Al2O3, respectively͒ ͑Ref. 18͒. We are also not aware of
any reports of E and H measurements for full-density boehmite and di-
aspore. In Table I, we use E=200 GPa and H=10 GPa for full-density
alumina since these values are representative of full-density ␥-Al2O3 and
amorphous Al2O3 ͑Ref. 19͒.
18CRC Handbook of Chemistry and Physics, 84th ed., edited by D. R. Lide
͑CRC, New York, 2003͒.
19See, for example, T. C. Chou, T. G. Neih, S. D. McAdams, and G. M.
Pharr, Scr. Metall. Mater. 25, 2203 ͑1991͒; J. C. Barbour, J. A. Knapp, D.
M. Follstaedt, T. M. Mayer, K. G. Minor, and D. L. Linam, Nucl. Instrum.
Methods Phys. Res. B 166–167, 140 ͑2000͒.
20Note that the only difference in preparation conditions of samples CWE-1
and CWE-2 in Table I was a different concentration of the precursor in the
starting solution, resulting in different densities of AG monoliths.
21J. S. Field and M. V. Swain, J. Mater. Res. 8, 297 ͑1993͒.
22See, for example, L. J. Gibson, and M. F. Ashby, Cellular Solids ͑Cam-
bridge University Press, Cambridge, 1997͒; D. J. Green and P. Colombo,
MRS Bull. 28, 296 ͑2003͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
155.33.148.196 On: Mon, 01 Dec 2014 20:37:37