Note: Descriptions are shown in the official language in which they were submitted.
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
PHOTONIC BANDGAP MATERIALS BASED ON SILICON
FIELD OF THE INVENTION
The present invention relates to a method of synthesis of periodic
composite materials of silicon and another material with a dielectric constant
less
than silicon, and more particularly the invention relates to photonic band gap
(PBG) materials based on silicon having complete photonic bandgaps.
BACKGROUND OF THE INVENTION
Photonics is the science of molding the flow of light. Photonic band gap
(PBG) materials, as disclosed in S. John, Phys. Rev. Left. 58, 2486 (1987),
and
E. Yablonovitch, Phys. Rev. Left. 58, 2059 (1987), are a new class of
dielectrics
which carry the concept of molding the flow of light to its ultimate level,
namely by
facilitating the coherent localization of light, see S. John, Phys. Rev. Lett.
53,
2169 (1984), P. W. Anderson, Phil. Mag. B 52, 505 (1985), S. John, Physics
Today 44, no. 5, 32 (1991), and D. Wiersma, D. Bartolini, A. Lagendijk and R.
Righini, Nature 390, 671 (1997). This provides a mechanism for the control and
inhibition of spontaneous emission of light from atoms and molecules forming
the
active region of the PBG materials, and offers a basis for low threshold micro-
lasers and novel nonlinear optical phenomena. Light localization within a PBG
facilitates the realization of high quality factor micro-cavity devices and
the
integration of such devices through a network of microscopic wave-guide
channels (see J. D. Joannopoulos, P.R. Villeneuve and S. Fan, Nature 386, 143
(1998)) within a single all-optical microchip. Since light is caged within the
dielectric microstructure, it cannot scatter into unwanted modes of free
propagation and is forced to flow along engineered defect channels between the
desired circuit elements. PBG materials, infiltrated with suitable liquid
crystals,
may exhibit fully tunable photonic band structures [see K. Busch and S. John,
Phys. Rev. Lett. 83, 967 (1999) and E. Yablonovitch, Nature 401, 539 (1999)]
enabling the steering of light flow by an external voltage. These
possibilities
suggest that PBG materials may play a role in photonics; analogous to the role
of
semiconductors in conventional microelectronics. As pointed out by Sir John
Maddox, "If only it were possible to make dielectric materials in which
electromagnetic waves cannot propagate at certain frequencies, all kinds of
almost magical things would be possible." John Maddox, Nature 348, 481
(1990).
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
The single biggest obstacle to the realization of these photonic capabilities
is the lack of a proven route for synthesis of high quality, very large-scale
PBG
materials with significant electromagnetic gaps at micron and sub-micron
wavelengths. The method of micro-fabrication must also allow the controlled
incorporation of line and point defects, for optical circuitry, during the
synthetic
process.
One very promising material for use in producing photonic devices is
silicon. Producing photonic devices from silicon-based photonic crystals would
be
a very significant commercial advantage since methods of fabricating such
materials could be readily retrofitted into existing silicon chip fabrication
facilities.
Nature produces optically unique materials based on silica. Specifically,
opals are semiprecious stones used in jewellery and decoration. The structure
of
naturally occuring opals was discovered for the first time in 1964 [J.V.
Sanders,
Nature 1964]. They are macroporous materials made by a periodic distribution
of
silica submicrometer spheres embedded in a silica medium with a slightly
different refractive index. They present iridescent colors due to Bragg
diffraction
of light as a consequence of the three dimensional periodic modulation of the
dielectric contrast in the structure. Owing to their potential technologic
applications, the fabrication of artificial opals has become a significant
goal in the
field of optics.
It is very advantageous to use artificial opals as a template from which to
produce inverse opals of pure silicon. In this way the periodicity of the self-
assembling opal template is transferred to the inverse opal. A large scale
periodic microstructure is thereby produced efficiently and at low cost,
without
recourse to time consuming and expensive photolithograghy (see S. John and K.
Busch, Journal of Lightwave Technology IEEE, volume 17, number 11, pages
1931-1943, (1999)). Up to this point in time, conventional photolithography
has
produced only very small scale structures, with a very small number of
repeating
unit cells (see S.Y. Lin and J.G. Fleming, J. of Lightwave Technology IEEE,
17,
no.11, 1944 (1999) and S. Noda et al. ibid, 1948 (1999)). This method is
effective
for creating two-dimensional patterns, but does not readily lend itself to the
production of large scale three-dimensional periodic structures.
It is particularly advantageous to provide a method which can produce
inverse silicon opals with lattice constants spanning the range from which
useful
photonic devices could be produced and which at the same time is scalable to a
very large number of repeating unit cells. With such a silicon inverse opal, a
large
2
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
number of photonic devices can be integrated into a single three-dimensional
optical chip.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a method for synthesizing
periodic silicon composite materials having unique optical properties, one
being a
complete photonic bandgap.
The synthesis and characterization of high quality, very large scale, face
centered cubic photonic band gap (PBG) material comprising a composite of
pure silicon-air, exhibiting a complete three-dimensional PBG centered on a
wavelength of 1.46 ,um is disclosed.
The present invention provides a method of producing artificial silica opals
with high optical quality, which can be made by microspheres in a wide range
of
diameters from 0.22 to 1.3 microns. A long-range face centered cubic (fcc)
ordering of the spheres in air medium has been achieved. The porous lattice of
these materials confers upon them the possibility to be employed as templates,
in which different materials can be infiltrated. Hence, they inherit the fcc
order of
the template.
Infiltration of these templates by silicon followed by removal of the silica
provides inverse silicon opals. This is obtained by chemical vapor deposition
and
anchoring of disilane into a self-assembling silica opal template, wetting of
a thick
silicon layer on the interior surfaces of the template, and subsequent removal
of
the template. This achievement realizes a long standing goal in photonic
materials and opens a new door for complete control of radiative emission from
atoms and molecules, ,light localization and the integration of micron scale
photonic devices into a three-dimensional all-optical micro-chip.
More particularly, the present invention provides a method for the
synthesis of a 0.1 mm to 1.0 cro scale single crystal of a face centered cubic
(fcc)
PBG material, comprising a close packed .870 micron diameter air spheres in
pure silicon. This silicon PBG material has a complete three-dimensional PBG
centered in the range of 1.3 to 1.7 microns, the wavelength range of choice
for
fiber optic telecommunication systems. The self-assembly synthetic approach
that we employ is straightforward, mild, inexpensive, accurate, and yields
single
crystal, inverse opal structures made of silicon comprising up to 10,000 x
10,000
x 10,000 unit cells into which various defect network architectures can be
imprinted during the initial stage of synthesis. The methodology is compatible
3
CA 02398632 2008-02-11
with, and can be easily integrated into, existing silicon fabrication
manufacturing
facilities.
In one aspect of the invention there is provided an inverted colloidal
photonic crystal including silicon and at least one other dielectric component
having an effective dielectric constant smaller than a dielectric constant of
silicon
and having a lattice periodicity ranging from about 0.28 microns to about 1.8
microns, wherein said inverted colloidal photonic crystal is an inverse
silicon opal
including air voids, said inverse silicon opal being characterized by having
at
least one complete photonic band gap in a total photon density of states
spanning at least 5% of a center frequency of the at least one complete
photonic
band gap.
In another aspect of the invention there is provided an inverted colloidal
photonic crystal including silicon and at least one other dielectric component
having an effective dielectric constant smaller than a dielectric constant of
silicon
and having a lattice periodicity ranging from about 0.28 microns to about 1.8
microns, wherein said inverted colloidal photonic crystal is an inverse
silicon opal
including air voids, said inverse silicon opal being characterized by having
at
least one complete photonic band gap in a local density of states, and wherein
a
ratio of said complete photonic band gap to a center frequency in the local
density of states ranges from 0% to about 20%.
In another aspect of the invention there is provided an inverted colloidal
photonic crystal including silicon and at least one other dielectric component
having an effective dielectric constant smaller than a dielectric constant of
silicon
and having a lattice periodicity ranging from about 0.28 microns to about 1.8
microns, wherein said inverted colloidal photonic crystal is an inverse
silicon opal
including air voids having a diameter between about 0.8 to about 0.9 microns,
and said inverse silicon opal having a complete photonic bandgap centered on a
wavelength in a range from about 1.3 to about 1.7 microns.
In another aspect of the invention there is provided an inverted colloidal
photonic crystal comprising an inverse silicon opal including close packed
spherical air voids in silicon, the spherical air voids having a diameter in a
range
from about 0.2 to about 1.3 microns, the spherical air voids being
monodisperse
with a size distribution of the spherical air voids having a standard
deviation in a
range between about 2% to about 5%, said inverse silicon opal characterized by
a complete photonic bandgap.
4
CA 02398632 2008-02-11
In another aspect of the invention there is provided an inverted colloidal
photonic crystal comprising an inverse silicon opal including close packed
spherical air voids in silicon, the spherical air voids having a diameter in a
range from about 0.8 to about 0.9 microns and exhibiting a complete photonic
bandgap centered on a wavelength in a range from about 1.3 to about 1.7
microns, the spherical air voids being monodisperse with a size distribution
of
the spherical air voids having a standard deviation in a range between about
2% to about 5%.
In another aspect of the invention there is provided an inverted colloidal
photonic crystal comprising an inverse silicon opal including close packed
spherical air voids in silicon, the spherical air voids having a diameter in a
range from about 0.55 to about 1.3 microns, the spherical air voids being
monodisperse with a size distribution of the spherical air voids having a
standard deviation in a range between about 2% to about 5%, said inverse
silicon opal exhibiting a complete photonic bandgap.
In another aspect of the invention there is provided a method of
growing an inverse silicon opal, comprising:
providing a three dimensional opal template comprising particles
having an effective geometry and composition;
infiltrating the opal template with an effective amount of silicon into
voids between said particles; and
etching out the particles to produce an inverse silicon opal.
The present invention also provides a method of growing an inverse
silicon opal, comprising:
providing a three dimensional silica opal template made of silica
spheres;
infiltrating voids in the silica opal template with enough silicon to fill
between about 80% to about 100% of said voids; and
etching the silica spheres out of the template to produce an inverse
silicon opal.
The present invention also provides a method of growing an inverse
silicon opal with a complete three dimensional photonic bandgap, comprising:
providing a three dimensional silica opal template including
substantially mono-disperse silica spheres having a diameter in a range from
about 0.55 to about 1.3 microns;
infiltrating voids in the silica opal template with enough silicon to fill
4a
CA 02398632 2008-02-11
between about 80% to about 100% of said voids; and
etching all the silica out of the template to produce an inverse silicon
opal.
In another aspect of the invention there is provided a method of
growing silica spheres having a diameter between about 0.55 microns to
about 1.3 microns, comprising:
growing silica seed particles by rapidly adding a first amount of
20
30
4b
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
tetraethylorthosilicate (TEOS) to a stirred alcohol solution comprising water
and
aqueous ammonia to form a suspension of silica seed particles;
after a first effective period of time of stirring enlarging the silica seed
particles to silica spheres with a preselected diameter by slowly adding a
second
amount of tetraethylorthosilicate (TEOS) with stirring and thereafter stirring
the
suspension for a second effective period of time; and
collecting the silica spheres with a diameter between about 0.6 microns to
about 1.3 microns from said suspension.
The present invention provides method of synthesizing an opal from silica
spheres, comprising;
providing a suspension of silica spheres in a liquid, the silica spheres
having an effective diameter and the liquid having an effective viscosity and
density so said silica spheres settle with an effective velocity;
settling the silica spheres from said suspension at a first effective
temperature to form a sediment of preselected dimensions; and
drying the sediment at a second effective temperature.
BRIEF DESCRIPTION OF THE DRAWINGS
The method of synthesis of silicon-based photonic band gap materials
according to the present invention will now be described, by way of example
only, reference being made to the accompanying drawings, in which:
Figure 1 shows a transmission electron micrograph (TEM) image of silica
spheres (left) and the corresponding size distribution (right);
Figure 2 shows a scanning electron micrograph (SEM) image of silica
spheres made by a re-growth process on seeds having a diameter of 0.853
0.012 microns (left) and the corresponding size distribution (right);
Figure 3 shows an SEM image of a cleft edge of a crystallized sediment of
0.448 micron diameter silica spheres;
Figure 4 shows an SEM image of a cleft edge of the crystallized sediment
of 0.853 microns diameter silica spheres;
Figure 5 is a vertical view of an electrophoretic cell used to grow silica
opals;
Figure 6 are plots of settling times versus height for sedimentation of Si02
spheres of 0.500 micron diameter settling in the presence and absence of an
electric field;
Figure 7 shows SEM images of a cleaved edge of a silica opal produced
5
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
using 0.870 micron diameter Si02 spheres according to the present invention,
a)
the spheres settled without an electric field, its Fourier transform (inset)
showing
the absence of order while the opal shown in b) was settled using an electric
field, its Fourier transform (inset) showing the presence of periodicity;
Figure 8 shows Bragg diffraction from two different silica opals at different
temperatures, (a) silica opal produced by sintering 0.870 micron diameter Si02
spheres according to the present invention whose sedimentation was slowed and
(b) silica opal produced from as grown 0.205 micron diameter Si02 spheres
settled under acceleration;
Figure 9 shows scanning electron micrographs for a silica opal template
sintered at 950 C for 3 hours (left) and sintered at 1025 C 12 hours (right);
Figure 10(a) is a scanning electron micrograph (SEM) of an internal [113]
facet of a Si infiltrated opal produced in accordance with the present
invention;
Figure 10(b) is an atomic force microscopy (AFM) image of a surface
showing smooth growth of silicon on an opal template;
Figure 11 a shows an SEM image of an internal [110] facet of a silicon
inverse opal;
Figure 11b shows an SEM image of an internal [111] facet of a silicon
inverse opal;
Figure 12 shows the photonic band structure of a silicon inverse opal, with
88 % infiltration of silicon into the opal template voids, the complete
bandgap
being shown by the crosshatched region; and
Figure 13 shows the measured reflectivity spectrum of silicon inverse opal,
the shaded regions centered around 2.5 /,zm and 1.46 ,um show the calculated
positions of the first stop band and the complete photonic bandgap.
DETAILED DESCRIPTION OF THE INVENTION
In a preferred embodiment of the present invention there is provided a
three dimensional periodic composite material comprising silicon and a
dielectric
component having a dielectric constant smaller than the dielectric constant of
silicon. The periodic composite material has a cubic lattice periodicity
(center to
center distance between adjacent cubic repeating units) ranging from about
0.28
microns to about 1.8 microns.
In a more preferred embodiment the dielectric constant of the lower
dielectric component is in a range from about 1 to about 2.1 and said
composite
6
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
material is characterized by at least one complete photonic bandgap centered
in
the range of 1.3 to 1.7 microns. A preferred method of producing this
silicon/dielectric material composite involves producing an inverse silicon
opal
from a silica opal with the silica opal produced using monodisperse silica
spheres
of selected diameter. A major advantage obtained by producing an inverse
silicon opal in this way is that composites with complete photonic bandgaps
can
be economically synthesized which heretofore has not been realized.
The fabrication of artificial opals requires several stages including 1)
synthesis of monodisperse silica spheres with diameter between 0.2 and 1.3
microns; 2) growth of silica opal template; and 3) sintering the three
dimensional
periodic structure to increase the mechanical stability and control the volume
filling fraction. After synthesis of the artificial opal, silicon (silicon
alloys) is
infiltrated into the opal template followed by etching to remove the silica.
1) Growth Of Silica Spheres
The method of synthesis begins with the highly controlled formation of a
silica opal template comprising a weakly sintered face centered cubic (fcc)
lattice
of monodisperse silica (SiO2) spheres having a diameter chosen between 0.6
microns and 1.3 microns. The inverse opal is produced by infiltration of the
template with the desired amount of silicon then etching away the silica,
described more fully hereinafter. The choice of large silica spheres with
diameters in the range between 0.6 microns and 1.3 microns ensures that the
final reverse opal structure will have a complete PBG in a frequency range
below
the optical absorption edge of bulk silicon which makes it a building block
for
silicon-based optical circuit elements.
11) Growth of Spheres With Diameter Between 0.2 to 0.6 Microns
In one aspect of this invention there is provided a method of synthesizing
suspensions of silica colloidal spheres of diameters in the range 0.2-1.3
microns
which are monodisperse with a narrow size distribution (standard deviation <
5%)
in such a way as to reduce formation of defects in the spheres, which
advantageously reduces imperfections in opal structures produced from the
silica
spheres. The inventors have discovered that the synthesis of monodisperse
(typically less than 5% variation in diameter) large silica spheres may be
achieved by a modified Stober method [W. Stober, A. Fink, E. Bohn; J. of
Colloid and Interface Science, Vol. 26, pp. 62-69, 1968].
Generally, smaller silica spheres (0.2-0.6 micron diameter) were grown by
mixing two different solutions, one containing a mixture of water, ammonia and
7
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
ethanol and the other containing a mixture of tetraethylorthosilicate (TEOS)
and
ethanol were mixed. The concentrations employed are shown in Table 1. Water
was used as a varying parameter to control the sphere size. The solution was
thoroughly agitated and the temperature kept constant by using a thermally
stabilised bath at.27 C. This was done in order to prevent lack of homogeneity
in
the solution during particle growth. By this procedure suspensions of
spherically
shaped, well dispersed silica particles of diameters between 0.2 and 0.55
microns were obtained. The size distribution was very - narrow, the standard
deviation being between 2% and 5% in all cases. After this, the suspensions
were centrifuged in alcohol several times, the supernatant liquid being
removed
each time. This was done in order to completely clean the suspensions from the
ammonia remains of the synthesis process. Following collection and cleaning of
the colloidal spheres they are dispersed in water. Example 1 gives an
illustrative,
non-limiting example of growth of silica spheres smaller than 0.6 microns in
diameter.
EXAMPLE 1
Synthesis of 0.448 0.006 Micron Diameter Silica Spheres.
Two different solutions were prepared, the first solution contained 0.727
ml of tetraethylorthosilicate (TEOS) and 4.5 ml of ethanol. The second
solution
included 1.219 ml (28% weight in water) of NH3, and 0.864 ml of double
distilled
water and 4.69 ml of ethanol. The solutions were kept at 27 C in a thermally
stabilised bath for 1 hour. The solutions were then mixed and stirred and the
reaction allowed to proceed for two hours. When all the TEOS had reacted the
colloidal spherical particles of the suspension were analysed by scanning
electron microscopy (SEM) and an image of some spheres and their size
distributions determined by light scattering may be seen in Figure 1. The
different
concentrations employed for different batches of different sphere samples are
shown in Table 1.
35
8
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
Table 1. Concentrations of the different chemicals employed in the sphere
synthesis process.
TEOS (M) [NH3] (M) [H20] (M) ~ (pm) a (nm)
0.266 1.45 4.6 0.361 0.013
0.266 1.45 5.6 0.426 0.014
0.266 1.45 7.6 0.448 0.006
0.266 1.45 9.6 0.485 0.009
0.266 1.45 11.6 0.555 0.015
0.266 1.45 13.6 0.581 0.013
0.266 1.45 15.6 0.502 0.033
0.266 1.45 17.6 0.431 0.024
0.266 1.45 19.6 0.332 0.049
0.266 1.45 21.6 0.313 0.031
0.266 1.45 23.6 0.0272 0.022
1 ii) Growth of Spheres With Diameter Between 0.55 to 1.3 Microns
Suspensions of colloidal silica spheres with diameters in the range 0.55-
1.3 microns were produced starting with suspensions of 0.55 micron diameter
spheres grown according to Example 1 above. The 0.55 micron silica spheres
were used as seeds on which a continuous silica growth process was carried
out.
Due to the larger sphere size of the seeds, a smaller number of re-growth
cycles
are needed advantageously providing a synthesis process much faster than has
been available to date. Monodisperse spheres having a diameter of 1.3 microns
were grown using three re-growth cycles. The particles were allowed to settle
under natural sedimentation (1 g) in water which facilitated obtaining
monodisperse particles. 'By doing this, it was possible to separate the
smaller
spheres, those remaining from the original seed suspension or resulting from a
thinner silica recovering, from the larger ones. Once the smaller spheres were
removed from the suspension the standard deviation was determined to be
between 2 and 5%. Illustrative and non-limiting examples of growth of silica
spheres between 0.6 to 1.3 micron diameter are given herebelow.
9
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
EXAMPLE 2
Synthesis Of 0.863# 0.025 Micron Diameter Silica Spheres
Silica seed particles were grown to diameters of about 0.55 microns by
mixing 74 ml of absolute ethanol, 10 ml of aqueous ammonia (32% wt) and 4 ml
of double distilled water and stirring the mixture in a flask with a magnetic
stirrer.
Then, 5 ml of tetraethylorthosilane (TEOS) was added rapidly with stirring to
get
the seeds in suspension. After 2 hours the first re-growth cycle was initiated
by
adding 10 ml of TEOS drop by drop over a period of 30 minutes while the
suspension was stirred with a magnetic stirrer. For 3 hours after addition of
this
TEOS the stirring was maintained after which the colloidal silica is washed
with
ethanol three times and then with water three times again. The colloidal
spherical
particles of the suspension were analysed by scanning electron microscopy and
an image of the same spheres and their size distribution determined by light
scattering may be seen in Figure 2. Table 2 summarizes the amounts of TEOS
added along with the diameters of the spheres obtained and the standard
deviation 6.
Table 2. TEOS concentration employed in the sphere re-growth process.
Seed concentration: 0.64% volume. Seeds: 0.55 microns diameter spheres.
[TEOS] is given in volume percentage.
1St 2nd Diameter (pm) [TEOS] % Vol. [TEOS] % Vol. (%)
5.38 No 0.740 2.8
7.53 No 0.770 2.5
10.75 No 0.863 2.9
5.38 5.38 0.886 3.7
EXAMPLE 3
Synthesis of 1.090+ 0.021 Micron Diameter Silica Spheres
Silica seeds of diameter of 0.55 microns were grown as described above
in Example 2. To a stirred suspension of these seed particles 5 ml of TEOS was
added rapidly. After 2 hours of stirring the seed suspension, 74 ml of
absolute
ethanol, 10 ml of aqueous ammonia (32% wt) and 4 ml of double distilled water
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
were added (0.32% volume of seeds) while the suspension was stirred. The first
re-growth cycle was initiated by adding 10 ml of TEOS (5.52% volume) drop by
drop, over a period of 90 minutes while the suspension was stirred. After all
the
TEOS was added the stirring was continued for another 2 hours. The second re-
growth step was initiated by transferring 100 ml of this suspension to another
flask and adding 10 ml of TEOS (10% volume) drop by drop, to the suspension
over a period of 90 minutes with stirring. After stirring the suspension for 3
hours
the colloidal suspension was washed as described above in Example 2.
EXAMPLE 4
Synthesis Of 1.360+ 0.039 Micron Diameter Silica Spheres
Silica seeds of diameter of 0.55 microns were grown as described above
in Example 2. After 2 hours of stirring the suspension of 0.55 micron diameter
seeds a solution comprising 74 ml of absolute ethanol, 10 ml of aqueous
ammonia (32% wt ) and 4 ml of double distilled water was added (0.32% volume
of seeds) while the suspension was continuously stirred. The first re-growth
process was initiated by adding 10 ml of TEOS (5.52% of volume) drop by drop
over a period of 70 minutes while the suspension was stirred. After this the
stirring was maintained for 2 hours. The second re-growth process was
initiated
by transferring 98 ml of this suspension to another flask to which was added
15
ml of TEOS (15.31 % volume) drop by drop over a period of 140 minutes while
the suspension was stirred with a magnetic stirrer. The resulting mixture was
stirred for 4 hours. The third regrowth step was initiated by,adding 15 ml of
TEOS
(15.31 % volume) drop by drop to the suspension over a period of about 150
minutes. The stirring was maintained for a further 4 hours and then the
colloidal
suspension is washed as in the above Example 2. Table 3 gives the TEOS
concentrations used in the re-growth process.
35
11
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
Table 3. TEOS concentration employed in the sphere re-growth process.
Seed concentration: 0.32%. Seeds: 0.55 microns diameter spheres. [TEOS]
is given in volume percentage.
1 st 2"d 3th Diameter
[TEOS] % Vol. [TEOS] % Vol. [TEOS] % Vol. (pm) (%)
5.52 No No 0.832 2.5
5.52 5.81 No 1.000 2.2
5.52 10 No 1.090 1.9
5.52 15.31 15.31 1.360 2.9
The method of silica sphere growth disclosed herein very advantageously
provides monodisperse spheres with a dispersity less than 5%. These are the
essential building-blocks needed to produce the silica opaltemplates from
which
the inverted silicon opals are produced.
2) Growth Of Silica Opal Template
The next step in the fabrication of an artificial opal is the crystallization
of
the silica spheres into a three dimensional periodic structure or template.
The
inventors have discovered that different methods for settling silica spheres
are
needed depending on the sphere diameter.
2i) Crystallisation Of Spheres Of Diameters Between 0.2 And 0.55 Microns
In A Face Centred Cubic Structure
In this range, natural sedimentation (under 1 gravity) in an aqueous
solution was used to crystallize the opal. There was dispersed 175 mg of
spheres
in 180 cm3 of water. The silica spheres were allowed to settle on a circular
polished poly(methylacrylate) substrate (mean rugosity < 50 nm) having a 2 cm
diameter. The sediment was completely formed after several days, depending on
sphere size (larger spheres sedimented faster than the smaller ones). Once the
sediment was formed, the supernatant liquid was removed and the sedimentation
tube placed in an oven at 60 C until the water was fully evaporated.
Afterwards,
the sediment was carefully removed from the substrate and its structure was
analyzed.
Studies of the growing surface confirmed that the spheres arrange in a
close packed structure, which grows close to the equilibrium following the
Edwards-Wilkinson equation. This implies the particles behave as effective
hard
12
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
spheres. This conclusion is supported by the fact that no ordering was
observed
in the suspension even at high concentrations. Also, the sedimentation
velocity
followed Stokes law. Three-dimensional order was analysed by SEM and optical
transmission spectroscopy. Samples were fractured and the internal free
surfaces observed. Cleft edges show long range face centred cubic domains, no
facets belonging to any other type of periodic structure being observable.
Domain
size ranges from 20 to >100 microns.
EXAMPLE 5
Crystallization Of 0.448 0.006 Micron Diameter Silica Spheres
Initially 175 mg of such spheres were dispersed in 180 cm3 of double
distilled water. Spheres were let to settle during 7 days on the mentioned
above
substrate. The supernatant liquid was then removed until a 2 mm high liquid
column was left above the sediment. The sedimentation tube was then placed in
an oven at 60 C until the whole liquid evaporated (1 day). The sediment was
then
carefully removed from the substrate and its internal structure analyzed. An
example of a fractured edge is shown in Figure 3, in which a long range fcc
domain can be observed.
2ii) Crystallisation Of Spheres Of Diameters Between 0.55 And 1.3 Microns
In A Face Centred Cubic Structure By Using Different Solvents, Co-
solvents And Temperatures
In order to obtain opals made of large spheres different organic solvents
were employed as a sedimentation medium. This was done to change the falling
velocity of the particles as well as the interactions between them.
Ethyleneglycol,
glycerol, acetone and ethanol and their aqueous mixtures at several different
concentrations were used as settling media. Spheres were then allowed to
settle.
When the sediment was formed, the supernatant liquid was removed until a 2
mm height liquid column was left in the sedimentation tube. Then, the sediment
was dried at different temperatures in an oven, ranging between 60 C and
120 C. Temperature plays an important role in the crystallization process.
Excellent results were obtained. SEM and optical characterization show that
fcc
optical quality opals were obtained by this procedure.
13
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
EXAMPLE 6
Crystallization Of 0.853 0.012 Microns Diameter Silica Spheres.
About 179 mg of spheres having a diameter of 0.853 0.012 microns were
dispersed in 180 cm3 of a mixture of 40% weight of etyleneglycol and 60% of
double distilled water. Spheres were allowed to settle during 4 days on the
above
mentioned substrate. Then the supernatant liquid was removed until a 2 mm
height liquid column was left above the sediment. The sedimentation tube was
then placed in an oven at 60 C during 1 day and later at 100 C during 5 days.
When the sediment was dry, it was carefully removed from'the substrate and its
internal structure analyzed. An example of a fractured edge is shown in Figure
4,
in which a long range fcc domain can be observed.
2iii) Crystallisation Of Spheres Of Diameters Between 0.55 And 0.9 Microns
In A Face Centred Cubic Structure By Electrophoretic Deposition To
Control Artificial Opal Growth
Natural sedimentation presents two problems. The first one is the time
required to obtain an opal. If the silica spheres are too small (under 0.30
microns
of diameter), several weeks are needed or even they may not settle at all
because thermal agitation compensates gravitational forces. The other
difficulty
that has been observed is related to heavy spheres which are over 0.56 microns
in diameter. In this case the sedimentation velocity is such that it is
difficult to
achieve an ordered array and it becomes completely impossible if the diameter
is
further increased. In this situation, the electrophoretic phenomena offers a
method for overcoming these two problems. Using the electric field to drive
the
sedimentation velocity and keep it around 0.4 mm/hour would solve the
difficulties mentioned before. The model of constant velocity particle packing
is
based on the interaction of gravitational (Fg 1/6Trpsgd3), Archimedes
(FA=1/67upWgd3) and frictional forces (Ft=37rrivd). Where ps and pW are the
spheres
and water mass densities, g is the gravity acceleration, rj is the viscosity
of water,
d is the spheres diameter and v is their velocity. When all forces are
balanced,
the Stokes law is obtained.
It is well known that Si02 particles in a colloidal suspension have a surface
charge density when they are away from the point of zero charge (PZC), in
which
case the electric charge is null. Taking into consideration the force produced
by
an electric field E parallel to all other forces, the following equation is
obtained for
the velocity:
u= [(d2 (p5 -pW)g)/18,9] + uE
14
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
where the first part of this equation is the classical Stokes law and the
second
part corresponds to the contribution of the electric field to the
sedimentation
velocity, related to the mobility of the spheres u. Now, the main problem is
how to
calculate the particle's *mobility. The application of the electrophoretic
concept
can solve it. Provided that Stokes velocity without electric field is
calculated with
great accuracy, the electrophoretic mobility can be obtained in a
straighiforward
manner if Stokes velocity is subtracted from the experimental velocity of the
sample under a known electric field. Once the mobility is determined, the
electric
field necessary to achieve a given velocity can be stated beforehand.
The electrophoresis cell shown in Figure 5 comprised a cylindrical tube (2
cm of diameter) of poly(methylacrylate) fixed to the base where the opal
should
settle, obtained from a standard silicon wafer sputtered with titanium or gold
(with
less than 0.1 nm of rugosity and thick enough to assure a good conductivity).
The
material used for the upper electrodes were platinum because it has the
highest
redox potential so that electrolysis is avoided. Both electrodes are connected
to a
dc source in order to develop an electrical field. With this method
sediments.with
thickness ranging between a few monolayers and 1 mm (depending on the
amount of silica spheres used) with surface areas about 3.1 cm2 are produced.
To measure the sedimentation velocity, the height descended by the
colloid/clear
water interface (setting 0 mm the initial height) was monitored with time.
The electrophoretically assisted sedimentation of Si02 spheres was
studied. An electric field was applied to colloidal suspensions of Si02
spheres in
which the original pH was varied by adding HCI to change the surface charge.
The point of zero charge, PZC, of silica occurs at a pH=2.5, so the pH values
of
the suspensions were chosen to be different enough without being close to the
PZC: pH=3.8 and the reference value (no acid added) of pH=8.4. The results of
the sedimentation velocities for silica spheres of 0.50 microns of diameter
are
graphically compared with the theoretical Stokes fall of a sample without
electric
field in the left panel of Figure 6. It can be clearly seen that, as the pH
moves
away from the PZC, the mobility increases and so does uE.
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
Table 4. Mobilities and velocities from Si02 spheres of 0.50 microns in
diameter at different pH and electric fields.
PH E(V/m) u(,um cmN s) v (mm/h)
3.8 -33 -2.0 2.9
8.4 -33 -3.9 5.2
8.4 0.5 -3.9 0.35
In order to study the effects of velocity variations on silica particle
ordering, two more sedimentations were prepared from the same sample. One of
the suspensions was left to settle in the absence of an electric field,
whereas in
the other one the electrodes were inverted to decrease'the sedimentation
velocity by opposing the gravity and the electric field. Since the mobility
can be
extracted from the previous experiment (u= -3.9 ,um cmN), the electric field
needed to get the desired velocity (0.4 mm/hour) was calculated to be 0.5 V/m.
The experimental value (v=0.35 mm/hour, see right graph in Figure 6) was close
to it. In Table 4 the results from this experiments are numerically compared.
Electronic and optical microscopy studies of all these samples were made
and it was observed that the sample in which sedimentation was slowed
electrophoretically demonstrated superior ordering than the one settled in the
absence of an electric field and while the accelerated samples from the
previous
experiment presented no order at all. Bragg diffraction was performed as well
showing that the opal grown with controlled sedimentation presented well-
defined
Bragg peaks.
For comparison, silica spheres with a diameter of 0.87 microns were
settled both in the presence and absence of an electric field. The scanning
electron microscopy (SEM) of a cleaved edge of the naturally settled opal (no
E
field) is shown in Figure 7A. A high velocity (1.54 mm/hour) was obtained for
these large spheres and no long-range order was achieved as evidence by the
Fourier transformed image shown in the inset of Figure 7A. A colloidal
suspension of silica spheres of the same diameter was settled under a
retarding
electric field, in which the velocity was kept close to 0.35 mm/hour. Figure
7B
shows that large ordered domains are obtained when sedimentation is performed
under an appropriate electric,field. Confirmation of this is evident from the
Fourier
transforms of both images, the opal settled under electric field presents a
clear
pattern that is not present in the naturally settled opal.
16
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
A Bragg diffraction study from the opal grown under slowed sedimentation
conditions was performed after sintering and very clear peaks were measured
as shown in Figure 8a while the other sample did not present any kind of peak
as
a result of the lack of large enough ordered domains. In addition, a little
percentage of small spheres was present in this sample. They were observed in
SEM images of the naturally settled sample but they were not present in the
other one because the electric force compensated the gravity force. This
suggests that the electrophoretic concept could be used to control the
presence
of small spheres in sedimentation when monodispersity is not granted.
Normally a suspension containing silica spheres of small diameter, (e.g.
0.205 microns of diameter) would take up to two months to settle to produce
the
sediment. The settling rate using electrophoretic assisted sedimentation was
accelerated from 0.09 mm/hour (natural velocity) to 0.35 mm/hour so that
complete sedimentation was achieved in less than two weeks without decreasing
the optical quality. Diffraction studies of the as-grown opal showed Bragg
peaks
as shown in Figure 8b, which denoted the presence of order within the opal.
The results disclosed herein demonstrate the importance of using
electrophoretic deposition for opal sedimentation. With this method it is
possible
to assemble opals comprising ordered arrays of spheres with diameters greater
than 0.55 microns which has heretofore been a major limitation.
Electrophoretic
assisted deposition has been shown to be an efficient way to control the
sedimentation velocity of silica spheres over a wide range of diameters.
EXAMPLE 7
300 mg of Si02 spheres with a diameter of 0.87 microns were suspended
in 30 ml of double distilled water. An electric field value of E= -8.3 V/m was
applied across a column of 8 cms in height containing the suspension. The
sedimentation velocity was 0.35 mm/hour, and the mobility of these spheres was
4.0 mm cmNs. Six days were needed to perform the whole sedimentation and
two more days to dry the samples at 60 C in an oven.
3) Sintering The Three dimensional Periodic Silica Opal
Crystalline sediments of silica spheres suffer from low mechanical stability
which makes them difficult to handle. In order to solve this, as-grown samples
were sintered at different final temperatures. The sintering process leads to
the
necking, or the formation of small necks, between neighboring silica spheres.
Necking is the thermally induced softening and flow of silica into the regions
17
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
defined by the touching of silica spheres in the colloidal silica crystal to
create a
silica neck with a diameter that facilitates infiltration of silicon into the
voids of the
silica opal and etching of silica from the infiltrated opal to create the
inverse
silicon opal.
Another extremely important parameter of the opals when used as
matrices for other compounds, is the filling fraction (ratio between the
volume
occupied for each compound and the total volume of the structure). Sintering
provides an accurate way to control the filling fraction between 74% and 100%
of
silica in opals. The process of necking allows tuning of the dimensions of the
silica opal and the resulting inverse silicon opal. The process of necking
also
provides mechanical stability to the template in addition to providing a
control
over the opal void volume for subsequent synthesis and providing the connected
network topology for removal of the template by an etching process. Studies
have shown that silica opals sintered at 950 C for 3 hours have a mechanically
stabilized compact face centered cubic (fcc) structure with a silica filling
factor of
74%. Further, sintering the opals at different temperatures between 950 C and
1100 C for different periods of time provided a method of controlling or
tuning the
optical properties and the free volume in the opals. Figure 9a shows an SEM of
a
silica opal sintered at 950 C for 3 hours compared to a silica opal sintered
at
1025 C for 12 hours, Figure 9b.
Example 8 below provides an illustrative, non-limiting example of use of
sintering temperature for tuning the optical and physical properties of a
silica
opal.
EXAMPLE 8
Pieces of an opal synthesized from 0.426 micron diameter spheres were
sintered at 1025 C for different periods of time. One piece of the opal was
placed
in an oven and heated up to 70 C employing a temperature gradient of 1 /min.
Once the temperature reached 70 C it was kept constant at 70 C for 3 hours to
prevent rapid or abrupt water de-sorption from the opal. After this, the
temperature was increased up to 1025 C employing a temperature gradient of
1 /min. The opal was maintained at this temperature for 3 hours. Two other
pieces of the starting opal were sintered using the same procedure but one
piece
was sintered for 6 hours and the other for 12 hours. Characterization of the
optical properties of the differently sintered opals reveal the free volume of
the
three opal pieces were different, decreasing with increasing temperature.
18
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
4) Infiltration of Silicon Into The Silica Opal
Silicon was grown inside the void spaces of the silica opal template by
means of chemical vapor deposition (CVD) using disilane (Si2H6) gas as a
precursor. The temperature during infiltration may be in the range from 100 C
to
about 500 C, but preferably the temperature is varied from 250 C for low in-
filling
samples to 350 C for high in-filling ones.
Example 9 below provides illustrative, non-limiting examples of use of
silicon infiltration into the silica opal template and annealing of the
silicon in the
template.
EXAMPLE 9
The silica opal was placed in a quartz cell and dried under vacuum for
about 5 hours. Disilane gas was added to the cell to raise the pressure to
about
200 torr, but the pressure may be in the range from 0.1 to about 760 Torr. The
cell was heated at 350 C for different periods of time hours, Table 5. The
cell
was evacuated by vacuum to remove disilane that remained unreacted and
annealed to 500 C for 30 minutes. (Disilane deposition: Dag 0; Ozin GA; Yang
H; Reber C; Bussiere G; Photoluminescent silicon clusters in oriented
hexagonal
mesoporous silica film, Advanced Materials 1999, Vol 11, Iss 6, pp 474-482.
Chomski E; Dag 0; Kuperman A; Coombs N; Ozin GA; New forms of
luminescent silicon: Silicon-silica composite mesostructures, Advanced
Materials: Chemical Vapor Deposition 1996, Vol 2, Iss 1, pp 8-15. Silicon
annealing: Temple-Boyer P; Scheid E; Faugere G; Rousset B; Residual stress in
silicon films deposited by CVD from disilane, Thin Solid Films 1997, Vol 310,
Iss
1-2, pp 234-237).
Temperature ( C) Time (hours) Degree of filling (%)
350 12 88
350 24 100
335 44 100
335 24 50
335+350* 36+24 65
335+335** 24+24 80
335 48 90
350 48 90
Table 5. Conditions required for different degrees of silicon infiltration in
the silica opal. * This sample was heated twice. ** After the initial heating,
the cell was pumped out and fresh disilane added and the cell was heated
19
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
again.
Theory predicts that the maximum PBG is obtained with a 90% to 97% in-
filling of the opal voids in the form of a uniform, thick, wetting layer on
the silica
surfaces. The reaction time was typically 24 hours and the disilane pressure
was
about 200 torr. After anchoring and growth of silicon, the samples are
annealed
or heated to 500 C to assist diffusion of silicon into the voids in the
template to
provide substantially uniform spatial distribution of silicon in the voids.
The
annealing temperature is varied depending on whether crystallization of the
silicon is required. The silica-silicon composite may be annealed in the
temperature range from about 400 C to 950 C.
The silica template is subsequently removed using fluoride-based etching
procedures designed to minimize the dissolution of the macroporous silicon
framework. The inverse silicon opal may be annealed in the temperature range
from about 400 C to 1100 C.
Examples below provide illustrative, non-limiting examples of silica opal
removal from the composite silicon-silica opal material.
EXAMPLE 10
From the silicon photolithography literature, fluoride based etches have
the best selectivity for silica in the presence of silicon. In this example,
the silicon
infiltrated silica opal template was placed in a 1 or 2% aqueous HF solution
for
about 8 hours. To those skilled in the art of etching silica from silicon it
is well
known that buffered 10: 1 NH4F/HF solution can be usefully employed to keep a
constant level of fluoride ions. This is common in small photolithography
setups
to dissolve silica in the presence of silicon. (Silicon Etchants: Ghandhi, SK;
VLSI
Fabrication Principles, Wiley, 1983). Thermal reductive-elimination of surface
hydride from the inverse silicon opal is employed to control the ultimate
hydrogen
content of the silicon layer.
In addition to disilane, other precursors for silicon that could easily be
infiltrated into silica colloidal crystals (opals) followed by sacrificial
etching of the
silica template include the following. Molecular beam and laser ablation of Si
atoms followed by thermal post treatment in a controlled atmosphere to control
the amorphous and crystalline silicon content. Capped and uncapped colloidal
and molecular cluster forms of silicon using vapor, melt and solution-phase
techniques followed by thermal post treatment. Infiltration of silane-based
polymers using solution and melt impregnation and thermal post-treatment
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
techniques. Other volatile CVD silane-based precursors may be taken from the
homologous series SinH2n+2 where n = 1,2,3 etc.
Examples of other silicon precursors, other deposition techniques, other
forms of silicon for synthesizing the inverse silicon opal comprise, but are
not
limited by, the following. Capped silicon clusters like octasilacubanes
(R8Si8)
could be used as a Si source for CVD. Octa-tert-butyloctasilacubane vaporizes
around 200 C and decomposes to silicon from 350-450 C. Furukawa K; Fujino
M; Matsumoto N; Superlattice structure of octa-tert butylpentacyclo-
[4.2Ø0(2,5).0(3,8).0(4,7)]octasilane found by reinvestigation of X-ray
structure
analysis, Journal Of Organometallic Chemistry 1996, Vol 515, Iss 1-2, pp 37-
41. Yang CS; Bley RA; Kauzlarich SM; Lee HWH; Delgado GR; Synthesis of
alkyl-terminated silicon nanoclusters by a solution route, Journal Of The
American Chemical Society 1999, Vol 121, pp 5191-5195. Silicon
nanocrystallites could be used to infiltrate the silica opal. Sweryda-Krawiec
B;
Cassagnneau T; Fendler JH; Ultrathin electroactive junctions assembled from
silicon nanocrystallites and polypyrrole, Advanced Materials 1999, Vol 11, pp
644-659. Kanemitsu Y; Silicon and germanium nanoparticies, Light Emission in
Silicon From Physics to Devices, Semiconductors and Semimetals, Academic
Press, San Diego 1998, pp 157-202. Brus L; Silicon polymers and nanocrystals,
Light Emission in Silicon From Physics to Devices, Semiconductors and
Semimetals, Academic Press, San Diego 1998, pp 303-326. Abelson JR;
Plasma deposition of hydrogenated amorphous silicon, studies of the growth
surface, Applied Physics A, Materials Science & Processing 1993 Vol 56, pp
493-512. Yoon JH; Lim SH; Moon BY; Jang J; Polycrystalline silicon film
deposited at 300 C, Journal Of The Korean Physical Society 1999 Vol 35
S1017-S1020, Suppl. S. Bhat KN; Ramesh MC; Rao PRS; Ganesh B; Polysilicon
technology, IETE Journal Of Research 1997 Vol 43, pp 143-154. Porous silicon
could also be used, Cullis, AG; Canham LT; Calcott PDJ; The structural and
luminescence properties of porous silicon, Journal Of Applied Physics 1997
Vol 82, pp 909-965.
Characterization of Inverse Silicon Opal
Micro-Raman spectroscopy (MRS) was used to ascertain the nature and
quality of the sample. The silicon phonon peak observed in the inverse silicon
opal samples was narrow and centered at 515 cm"', suggesting the presence of
crystalline silicon. Scanning electron microscopy (SEM) and atomic force
21
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
microscopy (AFM) were used to characterize the silicon growth. Figure 10(a)
shows an internal [113] face of the silicon, infiltrated opal. The SEM picture
reveals a large single domain of fcc order and a thick, uniform layer of
silicon
surrounding the silica spheres, indicating a high degree of infiltration.
Furthermore, the small necks between neighboring silica spheres, which appear
white in the image, clearly demonstrate the connectivity of the lattice. In
Figure
10(b) an AFM image of a local area of the infiltrated opal surface is shown,
highlighting the smoothness of the silicon coating. From the AFM measurements,
the surface roughness was estimated to be 2 nm. The growth of the silicon-
wetting layer is quite homogeneous and is independent of the local
characteristics of the opal template. The nearly complete and homogeneous
infiltration of silicon occurs throughout the depth of the sample.
Figure 11 a is an SEM image of an internal [110] face of the inverse silicon
opal taken after etching and Figure 11 b shows an internal [111] facet of an
inverse opal structure. These images clearly show an infiltrated structure
having
an interconnected network of air spheres surrounded by thin silicon shells,
inheriting the face centered cubic structure of the opal template. The
adjacent air
spheres are connected via windows, defining the neck regions which result from
the sintering process.
Unlike earlier studies of inverse opal structures made of Ti02 [see B. T.
Holland, C. F. Blanford and A. Stein , Science 281, 538 (1998): J. Wijnhoven
and W. L. Vos, Science 281, 802 (1998)], graphitic carbon [see A. A. Zakhidov
et. al. Science 282, 897 (1998)], CdSe [see D. Norris, et. al. Adv. Mater. 11,
165
(1999)], and CdSe and CdS [see P.V. Braun and P. Wiltsius, Nature 402, 603
(1999)], the silicon inverse opal synthesized according to the present
invention
simultaneously satisfies the two essential criteria for complete PBG
formation.
First, the refractive index of silicon is 3.5, well above the theoretically
determined
threshold of 2.8 for a PBG in a fcc lattice of air spheres disclosed in K.
Busch and
S. John, Phys. Rev. E 58, 3896 (1998). Secondly, the optical absorption edge
of
the silicon backbone occurs at a frequency well above the frequency range of
the
PBG, thereby allowing coherent localization of light within the material, with
minimal absorptive losses. This is an essential feature for future PBG device
applications.
The photonic band structure of a silicon inverse opal, with 88 % infiltration
of silicon into the opal template voids, is shown in Figure 12. The hatched
region
highlights a complete PBG with a gap to mid-gap ratio of 5.1 %. The
calculations
22
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
were performed using the plane wave expansion method, following the model of
K. M. Ho, C. T. Chan and C. M. Soukoulis, Phys. Rev. Lett. 65, 3152 (1990),
using a basis of 2662 plane waves.
The optical properties of the inverse opal were characterized by
measuring the reflection spectrum and comparing the spectral positions of the
observed stop bands with predictions from band structure. A Bohmen Fourier
transform infrared (FTIR) spectrometer was used to measure the specular
reflectance spectrum. In order to accurately fit the observed reflection
spectrum
of the inverse opal to the results from band structure calculations it was
necessary to independently determine the lattice constant and the degree of
silicon infiltration. The lattice constant is related to the sphere size of
the original
silica opal template. It was. obtained by fitting the spectral positions of
the first
stop band edges in the P-L direction of the bare opal to the positions
predicted by
band structure calculations (the band edges were obtained by measuring the 3dB
points of the reflectance peak). The refractive index of the silica spheres
was
measured to be 1.456 using index matching experiments.
The degree of silicon infiltration is determined by both direct and indirect
means. In the direct method, the SEM image is analysed by a computer
graphics program. The graphics program provides a means of identifying image
pixels in a 2d coordinate system. The resolution of the SEM picture (measured
in
nanometeres/pixel) is obtained by measuring the pixel extension of the ruler
drawn at the bottom of the SEM picture. The thickness of the silicon coating
layer
on the silica sphere is measured at a large number of points at locations on
the
picture where the thickness is clearly visible and the angle of viewing is
known.
The average value and the standard deviation is recorded. The (cubic) lattice
constant can also be obtained from the center to center distance between
adjacent spheres and multiplying by 1.4142. Using a formula which gives the
degree of infiltration as a function of coating thickness, the degree of
infiltration is
evaluated. This formula is based on a model of the structure in which the
silica
spheres are in a close packed fcc (or other as the case may be) lattice and
the
silicon uniformly coats all exposed silica surfaces in the form of a spherical
shell.
In the indirect method, the photonic band structure associated with the
mathematical model described above is computed. This determines the precise
frequency ranges spanned by all of the photonic stop bands (in specific
directions) as well as the complete photonic band gap (spanning all
directions).
The optical reflectivity from the sample (at normal incidence to the sample)
23
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
probing the lowest frequency stop gap is then fitted to the photonic band
structure calculated for different silicon coating layer thicknesses. The best
fit yields the actual coating thickness, and hence the degree of infiltration.
It has
been found that both the direct and the indirect methods yield the same result
for
the degree of silicon infiltration.
In characterizing the present inverse opals, after the infiltration of
silicon,
the reflectance spectrum in the I'-L direction was again measured. The
spectrum
changed dramatically from that of the bare opal, with the first stop band
shifting
by 0.97 ,um. With the sphere size fixed from the bare opal measurements, the
degree of infiltration was determined to be 88% by fitting the positions of
the first
stop band edges to the calculated band structure.
In order to calculate the band structure of the silicon inverse opal and
compare the measured spectrum, it was necessary to examine the lattice
constant. The lattice constant, which is preserved after infiltration and
inversion,
was independently determined from reflectivity measurements of the bare silica
opal at normal incidence (the L-point). A (cubic) lattice constant of 1.23
microns
was obtained by fitting the spectral positions of the first stop band edges to
those
predicted by band structure calculations (using 1.45 as a refractive index for
silica). This corresponds to center-to-center distance between adjacent
spheres
of 0.87 microns. (The cubic lattice constant is 1.4142 x (the center to center
distance) for the fcc lattice.)
The silicon inverse opal crystals obtained after etching were then
measured. A microscope coupled to the FTIR was used to probe a single crystal
domain and also cover a wide range of angles in a single measurement. The
microscope produced a spot size of approximately 20 x 20 ,um2 and an incident
cone of wave vectors with an angular bandwidth spanning 15-35 from normal
incidence. The measured spectrum, shown in Figure 13, exhibits a broad peak
with a center wavelength of 2.5 ,um followed by a series of three peaks in the
near-IR regime. One of these latter peaks is centered at 1.46 microns with a
width of 5.1 % and corresponds to 88% silicon infiltration. Calculations show
that
this gap is sensitive to percent silicon, for example with 90 % silicon
infiltration
the gap center moves to 1.5 microns.
The band structure also reveals something very surprising and
unexpected. Namely, that as the degree of silicon infiltration is increased
gradually from 88% to about 97%, somewhere in between, there will be observed
a full PBG as large as 9% rather than 5%. The exact position of the optimum
24
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
depends somewhat on the details of the sintering. Nevertheless an optimum of
roughly 9% does appear in almost all of the models that were studied. At 100%
infiltration the gap again reduces to about 5%.
Table 6: PBG Position And Relative Width For A Photonic Crystal Made
From Spheres of Diameter 0.870 Microns And Index of Refraction 3.45
Degree of Si infiltration (%) PBG Center (microns) PBG Width to Center
Fre cy Ratio %
79 1.414 0.5
83 1.434 2.4
84 1.441 3.0
86 1.447 3.8
88 1.461 5.1
90 1.470 6.0
92 1.477 6.7
93 1.484 7.5
95 1.491 8.2
97 1.497 8.9
100 1.546 4.5
The synthesis of a very large scale, silicon based PBG material offers a
number of imminent possibilities, involving further infiltration of this
highly open
structure with light sensitive (i.e. light emitting) molecules or atoms,
magnetically
sensitive dopants and electrically sensitive dopants. Preferred dopants for
these
silicon based photonic crystals are those that luminescence in a wavelength
range located in or near the photonic bandgap. These luminescent dopants
include, but are not restricted to, rare earth atoms such as erbium, organic
dyes,
inorganic dyes, organic polymers and inorganic polymers which luminesce.
Variations of the present invention comprise the inverse silicon opals
having optically sensitive molecules adsorbed or chemically bonded to the
surface of the silicon. Exemplary optically sensitive molecules include
luminescent dyes and luminescent polymers adsorbed or chemically anchored to
the surface in the form of monolayers or multilayers. The silicon surface may
also
be modified with physically or chemically anchored/adsorbed monolayers or
multilayers including hydrophobic and hydrophylic organic molecules that could
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
facilitate the infiltration of other optically, electrically, magnetically
interesting
species.
The infiltrated silicon may be in the form of single crystal silicon,
amorphous silicon, polycrystalline silicon, porous silicon and nanocrystalline
silicon. Literature examples cited above for different precursors and
different
deposition techniques could be used to create these different forms of silicon
which comprise the inverse silicon opal.
Further, alloys of silicon may be used to produce composite silicon-based
materials with different optical/electronic properties than those with pure
silicon-
air. For example, silicon alloys that may be used include, but are not
restricted to,
silicon-germanium alloys SiXGe,_,, 0< x <1, silicon-carbide alloys SiXC,_X,
0< x <1, silicon-tungsten alloys, silicon-nickel alloys, silicon-titanium
alloys,
silicon-chromium alloys, silicon-aluminum alloys and silicon-molybdenum
alloys.
These alloys facilitate changes in the electronic band gap as well as the
photonic
band gap of the periodic composite. In this way electrical and optical
properties
of the material can be tailored for specific device applications.
The resulting luminescence and lasing characteristics of light emitting
species near a complete three-dimensional photonic band edge are expected to
be quite striking (see S. John and T. Quang, Physical Review Letters 74, 3419
(1995)). Important low threshold all-optical switching effects (see S. John
and T,.
Quang, Physical Review Letters 78, 1888 (1997)), anomalous nonlinear optical
response (see S. John and T. Quang, Physical Review Letters 76, 2484
(1996)), and coherent control of radiative dynamics are contemplated (see M.
Woldeyohannes and S. John, Physical Review A 60, 5046 (1999)). In this
regard, it is useful to explore self-assembly synthetic methods for creating
diamond lattice templates from which a considerably larger PBG may be
achieved, see K. M. Ho, C. T. Chan and C. M. Soukoulis, Phys. Rev. Lett. 65,
3152 (1990). It is also of considerable importance to generalize the template
formation procedure, to engineer wave-guide channels and specified point
defects through which and between which light can flow. Methods of soft
lithography as disclosed in Y. Xia and G. M. Whitesides, Angew. Chem. Int. Ed.
Engl. 37, 550 (1998) coupled with self-assembly may prove effective in
realizing
such "circuits of light".
The achievement disclosed herein of a periodic silicon-air composite
material with a complete photonic bandgap realizes a long standing goal in
photonic materials research and opens a new door for complete control of
26
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
radiative emission from atoms and molecules, light localization and the
integration of micron scale photonic devices into a three-dimensional all-
optical
micro-chip. The inverse silicon opals grown by the method disclosed herein,
which may form the basis of photonic circuit elements, may be grown with a
variety of geometries, shapes or morphologies including fibers, films,
spheres,
lithographic patterns and monoliths from microsopic to macroscopic dimensions.
For example, the opals may be grown with dimensions in a range from 2 x 2 x 2
unit cells to a x b x c unit cells, wherein 2 < a < 10,000, 2 < b < 10,000, 2
< c <
10,000.
Three dimensional inverse silicon opals may be grown having a planar
thin film geometry with dimensions in a range from 1 x 2 x 2 unit cells to a x
b x c
unit cells, wherein 1< a < 100, 10 < b, c < 100,000.
The method of producing the periodic silicon-air composites starting with
silica opals and producing the inverse opals therefrom is a preferred or best
mode known at present since the periodicity of the opal can be efficiently
transferred to the inverse opal. However, those skilled in the art will
appreciate
that synthesis of periodic silicon-air composites or variants thereof as
disclosed
herein will not be restricted to conversion of silica opals. Other silica
templates
and non-silica templates may be employed. Silica templates involving lattice
structures other than the close packed face center cubic lattice may be used
and
templates using two or more different sphere sizes may be used. These include
for example the hexagonal close packed structure, the body center cubic
structure, the diamond lattice structure, the hexagonal AB2 structure. Non-
silica
templates include periodically arrrayed block co-polymers and other self-
assembling organic materials. In this case non-spherical, repeating units can
be
realized. Here a multi-stage infiltration process is required since the
polymeric
material may not withstand the high temperatures required for silicon CVD.
Therefore, a material such as silica would be infiltrated into the polymer
template
and the polymer template will be removed, prior to the final infiltration with
silicon
and the final removal of silica.
Those skilled in the art will understand that silicon photonic crystals grown
by the present method, not having a complete PBG but only a photonic pseudo-
gap, that is to say a material for which there is a large suppression in the
total
photon density of states (DOS) from what it would be in either air or in bulk
silicon, have important applications as well. From theoretical studies (see S.
John
and T. Quang, Physical Review Letters 78, 1888 (1997)), it is known that even
27
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
a sharp drop in the DOS by a factor of 2 over a small frequency range would
lead
to novel optical switching devices. The same holds true for materials grown by
the present method, which do not exhibit a complete photonic band gap in the
total density of states, but only a complete photonic band gap in the local
photon
density of states. In particular, the local density of states (LDOS) controls
the
rate of spontaneous emission of light from atoms and molecules at particular
locations in the photonic crystal, for lasing and optical switching
applications. The
pseudogap material encompasses a broader range of materials and composites
than the rather restricted set of materials which exhibit a complete PBG.
Likeswise, materials with a complete gap or pseudogap in the LDOS encompass
an even broader range of materials than those which exhibit corresponding gaps
in the total density of states.
The LDOS is the density of states as felt by an atom or molecule in a
particular position in the photonic crystal. As stated above, a gap in the
LDOS
may occur under less restrictive conditions than those required for a gap in
the
total DOS. For microlaser device applications, it is contemplated that low
threshold laser action may be achieved with a gap only in the LDOS where the
light emitting atoms are actually situated. The LDOS is what actually controls
the
radiative dynamics of individual atoms and molecules. Finally, it should be
noted
that whereas the total DOS may only have a gap of only 10% in a silicon
inverse
opal with a "complete 3-d PBG", the LDOS may exhibit a gap of up to 20% in the
same material.
Certain silicon-air composites comprising doped silicon are useful as
sensors. The silicon may be doped silicon, n-type by doping with phosphorus or
p-type silicon obtained by doping with boron. The dopant is incorporated by
infiltrating the silicon in the presence of gaseous phosphenes or boranes.
Such a
three dimensional periodic composite material comprising silicon and a
dielectric
component having a dielectric constant small than a dielectric constant of
silicon
is treated by anodic oxidation to render it luminescent. The doped macroporous
silicon crystal with controlled porosity silicon walls functions as a
chemoselective
sensor to discriminate optically between molecules in a mixture, depending on
the diameter of the pores that grown in the silicon walls.
The foregoing description of the preferred embodiments of the invention
has been presented to illustrate the principles of the invention and not to
limit the
invention to the particular embodiment illustrated. It is intended that the
scope of
the invention be defined by all of the embodiments encompassed within the
28
CA 02398632 2002-07-29
WO 01/55484 PCT/CA01/00049
following claims and their equivalents.
29
