Cm 9605577

Published on January 2017 | Categories: Documents | Downloads: 13 | Comments: 0 | Views: 150
of 6
Download PDF   Embed   Report

Comments

Content


Sol-Gel Template Synthesis of Semiconductor
Nanostructures
Bri nda B. Lakshmi , Peter K. Dorhout, and Charl es R. Marti n*
Department of Chemistry, Colorado StateUniversity, Fort Collins, Colorado 80523
Received October 31, 1996. Revised Manuscript Received J anuary 17, 1997
X
The templ ate method for prepari ng nanostructures entai l s synthesi s of the desi red materi al
wi thi n the pores of a nanoporous membrane or other sol i d. A nanofi bri l or tubul e of the
desi red materi al i s obtai ned wi thi n each pore. Methods used previ ousl y to deposi t materi al s
wi thi n the pores of such membranes i ncl ude el ectrochemi cal and el ectrol ess deposi ti on and
i n si tu pol ymeri zati on. Thi s paper descri bes the fi rst use of sol -gel chemi stry to prepare
semi conductor nanofi bri l s and tubul es wi thi n the pores of an al umi na templ ate membrane.
Ti O
2
, WO
3
, and ZnO nanostructures have been prepared. Ti O
2
nanofi bri l s wi th di ameters
of 22 nm were found to be si ngl e crystal s of anatase wi th the c-axi s ori ented al ong the fi bri l
axi s. Bundl es of these fi bri l s were al so found to be si ngl e crystal l i ne, suggesti ng that the
i ndi vi ual fi bri l s are arranged i n a hi ghl y organi zed fashi on wi thi n the bundl e. Fi nal l y, 200
nm di ameter Ti O
2
fi bri l s were used as photocatal ysts for the decomposi ti on of sal i cyl i c aci d.
Introduction
We have been expl ori ng a general method for prepar-
i ng nanomateri al s that entai l s synthesi s of the desi red
materi al wi thi n the pores of a nanoporous membrane.
1-3
The membranes used contai n cyl i ndri cal pores wi th
monodi sperse di ameters, and a nanoscopi c fi bri l or
tubul e of the desi red materi al i s synthesi zed wi thi n each
pore. Thi s method has been used to make tubul es and
fi bri l s composed of pol ymers, metal s, semi conductors,
carbon, and Li i on i ntercal ati on materi al s.
1-3
Methods
used to synthesi ze such materi al s wi thi n the pores of
the templ ate membranes i ncl ude el ectrol ess metal depo-
si ti on, el ectrochemi cal methods, and i n si tu pol ym-
eri zati on.
1-3
Sol -gel chemi stry has recentl y evol ved i nto a general
and powerful approach for prepari ng i norgani c materi -
al s.
4,5
Thi s method typi cal l y entai l s hydrol ysi s of a
sol uti on of a precursor mol ecul e to obtai n fi rst a suspen-
si on of col l oi dal parti cl es (the sol ) and then a gel
composed of aggregated sol parti cl es. The gel i s then
thermal l y treated to yi el d the desi red materi al . I t
occurred to us that sol -gel chemi stry coul d be done
wi thi n the pores of the nanoporous templ ate membranes
to obtai n tubul es and fi bri l s of a vari ety of i norgani c
materi al s. We have recentl y used thi s combi nati on of
sol -gel and templ ate methods to prepare fi bri l s and
tubul es of a vari ety of i norgani c semi conducti ng materi -
al s i ncl udi ng Ti O
2
, ZnO, and WO
3
. We have found that
si ngl e-crystal anatase-phase Ti O
2
nanostructures can
be obtai ned vi a thi s approach and that these nanostruc-
tures can be used as effi ci ent photocatal ysts. The
resul ts of these i nvesti gati ons are descri bed here.
Experimental Section
Materials. Ti tani um i sopropoxi de, zi nc acetate, Li OH‚H2O,
WCl 6, 2,4-pentanedi one, and sal i cyl i c aci d (al l from Al dri ch),
ethanol (McCormi ck Di sti l l i ng Co.), and concentrated HCl
(Mal l i nckrodt) were used as recei ved. Puri fi ed water was
obtai ned by passi ng house-di sti l l ed water through a Mi l l i -Q
(Mi l l i pore) water puri fi cati on system. The al umi na templ ate
membranes were ei ther obtai ned commerci al l y (Whatman
Anopore fi l ters, Fi sher) or prepared i n-house.
6,7
The com-
merci al membranes had 200 nm di ameter pores, and the i n-
house-prepared membranes had 22 nm di ameter pores. The
pore di ameter of the i n-house prepared membranes was
obtai ned from a cal i brati on curve of pore di ameter vs vol tage
used duri ng membrane synthesi s.
7
Synthetic Methods. Ti O2 tubul es and fi bri l s were pre-
pared usi ng a sol -gel method si mi l ar to that descri bed by
Hamasaki et al .
8
Ti tani um i sopropoxi de (5 mL) was added to
25 mL of ethanol , and the resul ti ng sol uti on was sti rred i n an
i ce bath. To a second 25 mL porti on of ethanol were added
0.5 mL of water and 0.5 mL of 0.1 M HCl . The ti tani um
i sopropoxi de sol uti on was removed from the i ce bath and the
ethanol /HCl /water sol uti on was sl owl y added; unl ess otherwi se
noted, the temperature was mai ntai ned at 15 °C.
After ca. 60 s the resul ti ng mi xture turned mi l ky whi te (sol
formati on). The al umi na templ ate membrane was i mmedi -
atel y di pped i nto thi s sol uti on for an i mmersi on ti me that was
vari ed between 5 and 60 s. After the desi red i mmersi on ti me,
the membrane was removed from the sol and dri ed i n ai r for
30 mi n at room temperature. The membranes were then
pl aced i n a tube furnace (i n ai r), and the temperature was
ramped (50 °C h
-1
) to 400 °C. The membranes were heated
at thi s temperature for 6 h, and the temperature was ramped
back down (30 °C h
-1
) to room temperature. A l i terature
search produced one recent reference to the producti on of Ti O2
fi bri l s i n membranes of thi s type; however, a much more
di ffi cul t el ectrosyntheti c method was used.
9
ZnO fi bers were prepared usi ng a method si mi l ar to that of
Sakhora et al .
10
To 20 mL of ethanol was added 0.35 g of zi nc
acetate, and the resul ti ng mi xture was boi l ed unti l a cl ear
* Correspondi ng author. E-mai l : crmarti n@l amar.col ostate.edu.
X
Abstract publ i shed i n AdvanceACS Abstracts, February 15, 1997.
(1) Marti n, C. R. Science1994, 266, 1961.
(2) Marti n, C. R. Acc. Chem. Res. 1995, 28, 61.
(3) Marti n, C. R. Chem. Mater. 1996, 8, 1739.
(4) Bri nker, C. J.; Scherer, G. W. Sol-gel Science; Academi c Press,
I nc.: New York, 1990. Hench, L. L.; West, J. K. Chem. Rev. (Wash-
ington, D.C.) 1990, 90, 33.
(5) O’Regan, B.; Moser, J.; Anderson, M.; Gra¨tzel , M. J . Phys. Chem.
1990, 94, 8720; Nature1991, 335, 737.
(6) Foss, C. A.; Ti erney, M. J.; Marti n, C. R. J . Phys. Chem. 1992,
96, 9001.
(7) Hornyak, G. L.; Marti n, C. R. J . Phys. Chem., submi tted.
(8) Hamasaki , Y.; Ohkubo, S.; Murakami , K.; Sei , H.; Nogami , G.
J . Electrochem. Soc. 1994, 141, 660.
(9) Hoyer, P. Langmuir 1996, 12, 1411.
(10) Sakhora, S.; Ti ckanen, L. D.; Anderson, M. A. J . Phys. Chem.
1992, 96, 11087.
857 Chem. Mater. 1997, 9, 857-862
S0897-4756(96)00557-1 CCC: $14.00 ©1997 Ameri can Chemi cal Soci ety
sol uti on was obtai ned (ca. 30 mi n). The vol ume was returned
to 20 mL wi th ethanol , and 0.06 g of Li OH‚H2O was added.
The resul ti ng sol uti on was ul trasoni cated unti l a whi te
suspensi on was obtai ned (ca. 1 h). The al umi na membrane
was i mmersed i nto thi s sol for 1 mi n, removed, and al l owed
to dry i n ai r at room temperature for 30 mi n. The membrane
was then heated i n ai r at 120 °C for 6 h.
WO3 fi bers were prepared usi ng a method si mi l ar to that
of Ni shi de et al .
11
Ethanol (10 mL) was purged of ai r usi ng
an Ar bubbl er, and 1 g of WCl 6 was added wi th conti nued Ar
bubbl i ng. 2,4-Pentanedi one (2 mL) and H2O (0.05 mL) were
added, and the resul ti ng mi xture turned bl ue i n col or. The
al umi na membrane was i mmersed i nto thi s mi xture for 1 mi n,
removed and dri ed i n ai r at room temperature for 30 mi n. The
membrane was then pl aced i n a furnace and the temperature
was ramped (as before) to 550 °C. The membrane was heated
at thi s temperature for 6 h, after whi ch the temperature was
ramped back down to room temperature.
These sol -gel syntheti c methods yi el ded ei ther tubul es or
fi bri l s of the desi red semi conductor wi thi n the pores of the
templ ate membrane. I n addi ti on, thi n fi l ms of the materi al
were deposi ted on both faces of the membrane. One or both
of these surface fi l ms were removed pri or to anal ysi s of
the tubul es or fi bri l s. Thi s was accompl i shed by si mpl y
pol i shi ng the surface(s) of the membrane wi th sand paper
(1500 gri t).
Electron Microscopy. Scanni ng el ectron mi croscopi c
(SEM) i mages of the 200 nm di ameter tubul es and fi bri l s were
obtai ned as fol l ows: One surface l ayer was removed, and the
membrane was gl ued (usi ng Torr-Seal Epoxy, Vari an) to a
pi ece of paper towel . The membrane was gl ued wi th the
pol i shed face up. The resul ti ng composi te was i mmersed i nto
6 M aqueous NaOH for 10 mi n i n order to di ssol ve the
al umi na. Thi s yi el ded an ensembl e of semi conductor tubul es
or fi bri l s that protruded from the epoxy surface l i ke the bri stl es
of a brush. Thi s sampl e was attached to an SEM sampl e stub
wi th si l ver pai nt (Spi ), and 10 nm of Au-Pd was sputtered
onto the surface usi ng an Anatech sputter coater. The
resul ti ng sampl e was i maged usi ng a Phi l l i ps 505 mi croscope.
Transmi ssi on el ectron mi croscopi c (TEM) i mages of the
semi conductor nanostructures were obtai ned as fol l ows: Both
surface l ayers were removed, and a pi ece of the resul ti ng
membrane was pl aced onto a carbon-fi l m-coated TEM gri d.
The 6 M aqueous NaOH sol uti on was then appl i ed to the
membrane i n order to di ssol ve the al umi na. The freed
nanostructures adhered to the TEM gri d and were i maged
usi ng a JEOL 2000 mi croscope. El ectron di ffracti on data were
al so obtai ned usi ng thi s i nstrument. The accel erati ng vol tage
of the el ectron beam was 100 kV and the camera l ength was
120 cm. A gol d si ngl e crystal was used as a standard to check
the camera l ength.
UV-Visible Spectroscopy. Both surface l ayers were
removed, and the al umi na membrane wi th the semi conductor
nanostructures wi thi n the pores was pl aced i n a custom-made
sampl e hol der. The sampl e hol der was pl aced i n the sampl e
chamber of an Hi tachi spectrophotometer. A si mpl e absorp-
ti on mode spectrum was obtai ned usi ng a bare al umi na
membrane as the reference. Because the al umi na templ ate
membranes are transparent throughout the vi si bl e and UV
regi ons, these spectra showed the absorpti on band edge of the
semi conductor nanostructures wi thi n the pores.
Photocatalysis. Ti O2 i s known to be a good catal yst for
photodecomposi ti on of organi c mol ecul es.
12-19
The photocata-
l yti c acti vi ty of the sol -gel -synthesi zed Ti O2 fi bri l s was
eval uated usi ng sal i cyl i c aci d as the organi c mol ecul e to be
decomposed. Sampl es i n whi ch the Ti O2 fi bri l s protruded from
an epoxy surface were prepared as descri bed for the SEM
anal yses. The area of epoxy surface used for these photoca-
tal ysi s studi es was 1 cm
2
; thi s area contai ned ca. 10
9
200 nm
di ameter Ti O2 fi bri l s. After removal of the al umi na templ ate
membrane i n NaOH, the Ti O2 fi bri l s were ri nsed i n fi ve
porti ons of water to ensure compl ete removal of the NaOH.
The fi bri l s were then i mmersed (overni ght and i n the dark)
i n 15 mL of 0.5 mM aqueous sal i cyl i c aci d i n order to al l ow
organi cs present i n the epoxy to l each out pri or to the
photocatal ysi s experi ments.
The sampl es pretreated i n thi s way were then i mmersed
i nto a beaker contai ni ng 15 mL of fresh 0.5 mM aqueous
sal i cyl i c aci d. A quartz l i d was pl aced over the top of the
beaker, and the beaker was pl aced on the roof of the chemi stry
bui l di ng i n di rect Col orado sunl i ght. Two control s were run
at the same ti me and pl ace as the fi bri l l ar Ti O2 sampl es. The
fi rst consi sted of an i denti cal beaker wi th the same amount
of sal i cyl i c aci d sol uti on; however, a thi n fi l m (200 nm thi ck)
of Ti O2, prepared under i denti cal condi ti ons as the fi bri l l ar
sampl e, was used as the photocatal ysts. The second control
was al so i denti cal but contai ned no photocatal yst.
The concentrati on of sal i cyl i c aci d i n the three sol uti ons was
determi ned every 5 mi n for a total of 30 mi n. The sal i cyl i c
aci d concentrati on was determi ned from i ts characteri sti c UV
absorbance (296 nm), usi ng a cal i brati on curve obtai ned from
sol uti ons of known concentrati on.
Results and Discussion
TiO
2
Tubules and Fibrils. Fi gure 1 shows SEM
i mages of Ti O
2
tubul es and fi bri l s prepared i n the
al umi na membrane wi th 200 nm di ameter pores. Tu-
bul es were obtai ned i f the membrane was i mmersed i nto
the sol for a bri ef peri od (5 s, Fi gure 1a), whereas sol i d
Ti O
2
fi bri l s were obtai ned after l ong i mmersi on ti mes
(11) Ni shi de, T.; Yamaguchi , H.; Mi zukami , F. J . Mater. Sci. 1995,
30, 4946.
(12) Fuji shi ma, A.; Honda, K. Nature1972, 37, 238.
(13) Matthew, R. W. J . Phys. Chem. 1987, 91, 3328.
(14) Kraeutl er, B.; Bard, A. J. J . Am. Chem. Soc. 1978, 100, 5985.
(15) Ol l i s, D. F.; Hsi ao, C.; Budi man, L.; Lee, C. J . Catal. 1984, 88,
89.
(16) Okamoto, K.; Yamamoto, Y.; Tanaka, H.; Tanaka, M.; I taya,
A. Bull. Chem. Soc. J pn. 1985, 58, 2015.
(17) Mathews, R. W. J . Catal. 1987, 97, 565.
(18) Spanel , L.; Anderson, M. A. J . Am. Chem. Soc. 1991, 113, 2826.
(19) Gol dstei n, S.; Czapski , G.; Rabani , J. J . Phys. Chem. 1994, 98,
6586.
Figure1. SEM i mages of Ti O2 tubul es and fi bri l s prepared i n the al umi na membrane wi th 200 nm di ameter pores. The sol was
mai ntai ned at 15 °C, and the i mmersi on ti me was vari ed from 5 to 60 s. (a) I mmersi on ti me ) 5 s; remnants of the Ti O2 surface
l ayer can be seen i n thi s i mage. (b) I mmersi on ti me ) 25 s. (c) I mmersi on ti me ) 60 s.
858 Chem. Mater., Vol. 9, No. 3, 1997 Lakshmi et al.
(60 s, Fi gure 1c). I ntermedi ate i mmersi on ti mes yi el d
tubul es wi th very thi ck wal l s (25 s, Fi gure 1b). I n al l
cases the tubul es and fi bri l s were 50 µm l ong (the
thi ckness of the templ ate membrane) and had outsi de
di ameters of 200 nm (the di ameter of the pores).
Whether tubul es or fi bri l s are obtai ned i s al so deter-
mi ned by the temperature of the sol . The structures
shown i n Fi gure 1 were obtai ned by i mmersi on of the
al umi na templ ate membrane i nto a sol mai ntai ned at
15 °C. When the sol was 5 °C, thi n-wal l ed tubul es were
obtai ned even at l ong i mmersi on ti mes (1 mi n). I n
contrast, when the sol was mai ntai ned at 20 °C, sol i d
Ti O
2
fi bri l s were obtai ned even after very bri ef (5 s)
i mmersi on ti mes. These resul ts show that through
control of temperature and i mmersi on ti me, both tu-
bul es and fi bri l s can be prepared; i n addi ti on, the wal l
thi ckness of the tubul es can be vari ed at wi l l (e.g.,
Fi gure 1a,b).
The mechani sm of formati on of Ti O
2
from aci di fi ed
ti tani um al koxi de sol uti ons i s wel l documented.
20,21
I n
the earl y stages, sol parti cl es hel d together by a network
of -Ti -O- bonds are obtai ned. These parti cl es ul ti -
matel y coal esce to form a three-di mensi onal i nfi ni te
network, the gel . That tubul es are i ni ti al l y obtai ned
when thi s process i s done i n the al umi na membrane
i ndi cates that the sol parti cl es adsorb to the pore wal l s.
Thi s i s not surpri si ng si nce the pore wal l s are negati vel y
charged and the parti cl es are posi ti vel y charged.
22
An
anal ogous si tuati on occurs when posi ti vel y charged
electronically conductive polymers are synthesized within
membranes contai ni ng ani oni c si tes on thei r pore wal l s.
2
I n the conducti ve pol ymer case,
2
the rate of pol ym-
eri zati on wi thi n the pore i s faster than i n bul k sol uti on
due to enhanced l ocal concentrati on of ol i gomers ad-
sorbed to the pore wal l . A si mi l ar si tuati on occurs i n
the sol -gel Ti O
2
case. When a l ower concentrati on of
ti tani um i sopropoxi de (6 v/v % as opposed to the 20 v/v
% used i n Fi gure 1) was used to make the sol , gel ati on
i n bul k sol uti on was extremel y sl ow, even at room
temperature. However, when the al umi na membrane
was di pped i nto thi s sol , sol i d fi bri l s of Ti O
2
are obtai ned
i n the pores, even at short (5 s) i mmersi on ti mes. These
resul ts show that gel ati on occurs wi thi n the pores under
condi ti ons where gel ati on i n bul k sol uti on i s negl i gi bl y
sl ow. By anal ogy to the conducti ve pol ymer case
2
thi s
i s undoubtedl y due to the enhancement i n the l ocal
concentrati on of the sol parti cl es due to adsorpti on on
the pore wal l .
Fi gure 2 shows an absorpti on spectrum for an al u-
mi na membrane contai ni ng 200 nm di ameter Ti O
2
fi bri l s wi thi n the pores. An abrupt i ncrease i n absor-
bance i s observed at wavel engths bel ow 389 nm. Thi s
corresponds to the bandgap of the Ti O
2
, and bul k
sampl es of Ti O
2
show an anal ogous absorpti on spec-
trum.
23
These fi bri l s (and even the 22 nm di ameter
fi bri l s to be di scussed bel ow) are too l arge i n di ameter
to expect to see evi dence for quantum confi nement i n
the absorpti on spectrum.
23
However, prepari ng fi bri l s
wi th di ameters smal l enough to see evi dence for quan-
tum confi nement i s a goal of thi s research effort.
Electron Diffraction Data for TiO
2
Fibrils Pre-
pared in the Membrane with 22 nm Diameter
Pores. Fi gure 3a shows a TEM i mage of the Ti O
2
nanostructures obtai ned after di ssol vi ng away the 22
nm pore di ameter membrane. Numerous i mages of thi s
type were obtai ned, and i n al l cases bundl es of the Ti O
2
nanostructures were observed. The bundl e si zes ob-
served ranged from as smal l as 2-4 fi bri l s to as l arge
as 10 or more fi bri l s. The mai n feature i n Fi gure 3a
that runs di agonal l y across the i mage consi sts of two
bundl es of fi bri l s, one on the ri ght edge of the mai n
feature and one on the l eft edge. I n thi s case the
bundl es consi st of approxi matel y 3-4 fi bri l s. The
el ectron di ffracti on data to be di scussed bel ow (Fi gure
3b) were obtai ned from a porti on of the bundl e on the
l eft si de of thi s mai n feature. A second set of two
bundl es i s observed bel ow thi s mai n feature; thi s second
set of bundl es al so proceeds di agonal l y across the i mage
but at a smal l er angl e.
Fi gure 3b shows the i ndexed el ectron di ffracti on
pattern obtai ned from the center of the fi bri l bundl e on
the l eft si de of the mai n feature i n Fi gure 3a. The
ori entati on of the i mages i n Fi gure 3a,b i s the same;
that i s, the c* axi s i n Fi gure 3b i s paral l el to the fi bri l
bundl e axi s i n Fi gure 3a. These data show that the
fi bri l s are hi ghl y crystal l i ne anatase-phase Ti O
2
, wi th
the c* axi s of the anatase ori ented al ong the l ong axi s
of the fi bri l . Smal l fi bri l bundl es throughout the sampl e
di spl ayed the same crystal l i ne ori entati on; i .e., the
reci procal l atti ce di recti on [110] i s al most al ways paral -
l el to the el ectron beam, and the c* axi s i s al ong the
fi bri l axi s. We can concl ude from these observati ons
that the fi bri l s crystal l i ze as l ong, pri smati c crystal s
wi th the rare, and metastabl e, anatase mi neral ogi cal
ori entati on [001] wi th {110}.
24
Just as i nteresti ng, such smal l fi bri l bundl es do not
demonstrate a great deal of mosai c spread i n the
di ffracted el ectron beam, as one woul d expect for pol y-
crystal l i ne materi al s or ori ented pol ymer fi ber sampl es.
25
Larger bundl es of crystal l i ne fi bri l s do show mosai c
spreadi ng of the refl ecti ons, but they remai n ori ented
(20) Lee, B. I .; Pope, E. J. A. Chemical processing of ceramics;
Marcel Dekker, I nc.: New York, 1994.
(21) Li vage, J.; Henry, M.; Sanchez, C. Prog. Solid. State Chem.
1988, 18, 259.
(22) Bi schoff, B. L.; Anderson, M. A. Chem. Mater. 1995, 7, 1772.
(23) Enri ght, B.; Fi tzmauri ce, D. J . Phys. Chem. 1996, 100, 1027.
(24) Dana, J. D. (rewri tten by Pol ache, C.; Berman, H.; Frondel ,
C.) Thesystemof mineralogy; John Wi l ey: New York, 1955.
(25) Kl ug, H. P.; Al exander, L. E. X-ray Diffraction procedures for
polycrystalline and amorphous materials; Wi l ey I ntersci ence: New
York, 1974.
Figure2. UV-vi si bl e spectrum of the 200 nm di ameter Ti O2
fi bri l s.
Synthesis of Semiconductor Nanostructures Chem. Mater., Vol. 9, No. 3, 1997 859
al ong thei r [001] fi ber axi s. I n addi ti on, dark-fi el d
i mages of the fi bri l bundl es i ndi cate that the crystal l i ne
domai ns are very l arge. For exampl e, bundl es wi th
di ameters as l arge as 1 µm and l engths correspondi ng
to the enti re thi ckness of the templ ate membrane (30
µm) are si ngl e crystal s.
These data suggest that after the al umi na templ ate
i s removed, the i ndi vi dual fi bri l s are drawn together i n
an ordered fashi on, l i ke stacked 4 × 4 l umber. Thi s
stacki ng may be faci l i tated by vacant or l abi l e l i gand
si tes on the Ti atoms, whi ch si t near the (110) faces
(Fi gure 4). The Lewi s aci di c surface si tes may be
attracted to the oxi de (or hydroxi de) si tes on other
fi bri l s; these i nteracti ons coul d hol d the fi bri l s i n an
ori ented bundl e that di ffracts l i ke a l arger, si ngl e crystal
(Fi gure 3b).
Hi ghl y ori ented crystal l i ne domai ns of other meta-
stabl e phases have been shown to grow i n an “i nternal ”
matri x, i ndi cati ng that al though our observati ons of
ori ented Ti O
2
are unusual , they are not wi thout prece-
dent.
26
Furthermore, i t i s known that sol -gel syntheses
can resul t i n crystal l i zati on of metastabl e phases be-
cause of the l ower temperatures often empl oyed i n such
syntheses.
27
I n addi ti on, i t has been previ ousl y shown
that sol -gel synthesi s can yi el d si ngl e-crystal l i ne mate-
ri al , i f the crystal s are grown at very l ow tempera-
tures.
28
Si ngl e-crystal l i ne ori ented MgO nanorods have
al so recentl y been reported, al though thi s was not a sol -
gel synthesi s.
29
Ori ented carbon nanofi bers have been
prepared by templ ate synthesi s.
30
Fi nal l y, el ectron di ffracti on data from the 200 nm
di ameter Ti O
2
fi bri l s show that these fi bri l s are al so
hi ghl y crystal l i ne anatase. However, the data suggest
that these l arger-di ameter fi bri l s are not si ngl e crystal s.
Photocatalysis with the TiO
2
Fibrils. Ti O
2
can
be used as a photocatal yst to decompose or gani c
mol ecul es.
12-16
The mechani sm i s bel i eved to i nvol ve
absorpti on of a UV photon by Ti O
2
to produce an
el ectron-hol e pai r. These react wi th water to yi el d
hydroxyl and superoxi de radi cal s whi ch oxi di ze the
organi c mol ecul e. The ul ti mate products of these reac-
ti ons are CO
2
and water.
18,19
Most i nvesti gati ons of the
photocatal yti c acti vi ty of Ti O
2
have used UV l amps as
the source and Ti O
2
thi n fi l ms or powders as the
catal yst. We report here the use of sunl i ght as the
source and immobilized template-synthesized TiO
2
fibrils
as the catal yst.
Fi gure 5 shows a pl ot of sal i cyl i c aci d concentrati on
vs ti me of exposure of the sol uti on to sunl i ght. The
upper curve i s for the sol uti on that contai ned no Ti O
2
photocatal yst. No l oss of sal i cyl i c aci d from the sol uti on
was observed. I ndeed, there was a sl i ght i ncrease i n
concentrati on wi th ti me, due to evaporati on of water;
that evaporati on occurred i s proven by the appearance
of water drops on the quartz l i d. The l ower curve i n
(26) (a) Li n, J.; Cates, E.; Bi anconi , P. A. J . Am. Chem. Soc. 1994,
116, 4738. (b) Bi anconi , P. A.; Li n, J.; Strzel ecki , A. Nature1991, 349,
315.
(27) Lange, F. F. Science1996, 273, 903.
(28) Spanhel , L.; Anderson, M. A. J . Am. Chem. Soc. 1991, 113,
2826.
(29) (a) Yang, P.; Lei ber, C. M. Science 1996, 273, 1836. (b) I toh,
H.; Utamapanya, S.; Stark, J. V.; Kl abunde, K. J.; Schl up, J. R. Chem.
Mater. 1993, 5, 71.
(30) Kyotani , T.; Tsai , L.; Tomi ta, A. Chem. Mater. 1996, 8, 2109.
Figure3. (a, top) TEM i mage of a bundl e of 15 nm di ameter
Ti O2 fi bri l s. (b, bottom) Correspondi ng el ectron di ffracti on
pattern.
Figure 4. Vi ew down the c-axi s of anatase showi ng the
possi bl e l abi l e oxygen si tes on the (110) face. Ti O6 pol yhedra
wi th fi l l ed ci rcl es (Ti ) and open ci rcl es (O) are represented.
860 Chem. Mater., Vol. 9, No. 3, 1997 Lakshmi et al.
Fi gure 5 i s for the sol uti on that contai ned the fi bri l l ar
Ti O
2
photocatal yst. A rapi d (rel ati ve to the thi n-fi l m
catal yst, mi ddl e curve) decrease i n sal i cyl i c aci d con-
centrati on was observed. I ndeed, the rate of photode-
composi ti on for the fi bri l l ar catal yst i s an order of
magni tude hi gher than for the thi n-fi l m catal yst (see
ki neti c anal ysi s bel ow).
Both the thi n fi l m and fi bri l l ar catal yst were sup-
ported on an epoxy surface wi th an area of 1 cm
2
. The
enhancement i n rate for the fi bri l l ar catal yst i s due to
the hi gher total surface area of Ti O
2
exposed to sol uti on.
I ndeed, i f i t i s assumed that the surface area of the thi n-
fi l m catal yst i s 1 cm
2
and that al l of the surfaces of the
fi bri l s are catal yti cal l y acti ve, an enhancement i n rate
of 315 woul d be predi cted for the fi bri l l ar catal yst. Thi s
cal cul ated enhancement ari ses because the total cal cu-
l ated surface area of the fi bri l s i s 315 cm
2
of Ti O
2
surface
area/cm
2
of substrate epoxy surface area. Thi s corre-
sponds to 5.1 m
2
of Ti O
2
surface area/g of anatase fi bri l s.
Even hi gher surface areas woul d be expected for tubul es
or smal l er di ameter fi bri l s.
That the experi mental enhancement i n rate i s l ess
i s not surpri si ng. Fi rst, the surface of the Ti O
2
thi n-
fi l m catal yst i s undoubtedl y not atomi cal l y smooth,
meani ng that i ts true catal yti c surface area i s greater
than the assumed 1 cm
2
. Second, SEM i mages show
that the fi bri l s do not stand strai ght up but rather “l ean”
agai nst each other to form cl umps or bunches. As a
resul t, l arge porti ons of the fi bri l surfaces are shaded
from the sunl i ght. Hence, thi s prototype fi bri l l ar cata-
l yst i s not opti mal . Vari abl es to be consi dered for
opti mi zati on i ncl ude di ameter and aspect rati o of the
fi bri l s (l ength di vi ded by di ameter) and di stance be-
tween fi bri l s.
The data i n Fi gure 5 can be used to quanti tati vel y
eval uate the rate of photodecomposi ti on. The ki neti cs
of photodecomposi ti on of organi cs on Ti O
2
has been
expl ai ned i n terms of a Langmui r model .
13,15-17
At a
l ow concentrati on of substrate (i .e., the mol ecul e to be
decomposed) thi s model predi cts si mpl e pseudo-fi rst-
order ki neti cs wi th respect to the substrate concentra-
ti on. When the data i n Fi gure 5 are pl otted accrdi ng
to thi s model , strai ght-l i ne pl ots were obtai ned and the
decomposi ti on rate constant can be cal cul ated from the
sl ope. The fi bri l l ar catal yst shows a rate constant of
0.03 mi n
-1
, as opposed to a rate constant of 0.003 mi n
-1
for the thi n-fi l m catal yst.
ZnOandWO
3
Fibrils. Fi gure 6 shows SEM i mages
of sol i d ZnO fi bri l s obtai ned wi thi n the pores of the
200 nm di ameter al umi na templ ate membranes. The
surface l ayer i s not compl etel y removed whi ch makes
the fi bers sti ck together. As was the case for Ti O
2
,
these fi bri l s are 200 nm i n di ameter and 50 µm l ong.
These sol i d fi bri l s were obtai ned usi ng a sol at 20 °C
Figure5. (a) Photodecomposi ti on of sal i cyl i c aci d i n sunl i ght.
Data for no photocatal yst, the thi n-fi l m Ti O2 photocatal yst,
and the fi bri l l ar (200 nm) Ti O2 photocatal yst are shown.
Figure 6. SEM i mage of 200 nm di ameter ZnO fi bri l s.
Figure7. UV-vi si bl e spectrum of the 200 nm di ameter ZnO
fi bri l s.
Figure 8. SEM i mage of 200 nm di ameter WO3 fi bri l s.
Synthesis of Semiconductor Nanostructures Chem. Mater., Vol. 9, No. 3, 1997 861
and an i mmersi on ti me of 1 mi n. By anal ogy to the Ti O
2
case, hol l ow ZnO tubul es are obtai ned i f l ower temper-
atures and shorter i mmersi on ti mes are used. An
absorpti on spectrum for a composi te membrane con-
tai ni ng these ZnO fi bri l s i s shown i n Fi gure 7. Agai n,
the bandgap absorpti on for ZnO i s observed.
10,18
Fi -
nal l y, when exposed to 366 nm l i ght, these fi bri l s
showed the characteri sti c
10
greeni sh-yel l ow bul k l umi -
nescence of ZnO.
Fi gure 8 shows SEM i mages of sol i d WO
3
fi bri l s
obtai ned wi thi n the pores of the 200 nm di ameter
al umi na templ ate membrane. As was the case for the
other materi al s, these fi bri l s are 200 nm i n di ameter
and 50 µm l ong. The SEM pi cture shows that some of
the surface l ayer i s not removed as i n the case of ZnO.
These sol i d fi bri l s were obtai ned usi ng an i mmersi on
ti me of 1 mi n and temperature of 20 °C. We are
especi al l y i nterested i n expl ori ng the el ectrochemi stry
of these WO
3
fi bri l s, to obtai n the correspondi ng tung-
sten bronzes.
Conclusions
We have shown that sol -gel syntheti c methods can
be used wi th the templ ate-based approach for prepari ng
nanomateri al s to yi el d semi conductor nanostructures.
Thi s marri age of sol -gel and templ ate methods shoul d
be appl i cabl e to a l arge number of materi al s. Both
fi bri l l ar and tubul ar nanostructures are possi bl e, and
si ngl e-crystal l i ne nanostructures of the metastabl e ana-
tase phase can be obtai ned. We are currentl y expl ori ng
appl i cati ons of these nanostructures i n photocatal ysi s,
el ectrochemi stry, battery research and devel opment,
photoel ectrochemi stry, and enzyme i mmobi l i zati on.
Acknowledgment. Fi nanci al support for thi s work
was provi ded by the U.S. Department of Energy (DE-
FG03-95ER14576) and by the Nati onal Sci ence Founda-
ti on (CTS-9423090).
CM9605577
862 Chem. Mater., Vol. 9, No. 3, 1997 Lakshmi et al.

Sponsor Documents

Or use your account on DocShare.tips

Hide

Forgot your password?

Or register your new account on DocShare.tips

Hide

Lost your password? Please enter your email address. You will receive a link to create a new password.

Back to log-in

Close