Electro-optic characteristics of a transparent nanophotonic device based on carbon nanotubes and liquid crystals.

Abstract

We present electro-optic characteristics of a transparent nanophotonic device fabricated on quartz substrate based on multiwall carbon nanotubes and nematic liquid crystals (LCs). The nanotube electrodes spawn a Gaussian electric field to three dimensionally address the LC molecules. The electro-optic characteristics of the device were investigated to optimize the device performance and it was found that lower driving voltages were suitable for microlens array and phase modulation applications, while higher driving voltages with a holding voltage can be used for display-related applications.

Full text

Page 1

Electro-optic characteristics of a transparent
nanophotonic device based on carbon
nanotubes and liquid crystals
Ranjith Rajasekharan, Qing Dai, and Timothy D. Wilkinson*
Department of Engineering, Centre of Molecular Materials for Photonics and Electronics,
University of Cambridge, 9 J. J. Thomson Avenue, Cambridge CB3 0FA,
United Kingdom
*Corresponding author: tdw13@cam.ac.uk
Received 3 November 2009; revised 25 February 2010; accepted 14 March 2010;
posted 15 March 2010 (Doc. ID 119488); published 5 April 2010
We present electro-optic characteristics of a transparent nanophotonic device fabricated on quartz
substrate based on multiwall carbon nanotubes and nematic liquid crystals (LCs). The nanotube elec-
trodes spawn a Gaussian electric field to three dimensionally address the LC molecules. The electro-
optic characteristics of the device were investigated to optimize the device performance and it was found
that lower driving voltages were suitable for microlens array and phase modulation applications, while
higher driving voltages with a holding voltage can be used for display-related applications.
Optical Society of America
OCIS codes:
230.0230, 230.3720, 350.4238, 350.4600.
© 2010
1.
The electro-optic characteristics of carbon nanotubes
(CNTs) doped in liquid-crystal (LC) cells have been
studied [1–6]. A change in threshold voltage and re-
sponse time was observed due to suppression of field
screening effect and increase in LC dielectric aniso-
tropy induced by incorporating CNTs [6]. We have
demonstrated nanotubes as an electrode site to three
dimensionally address the LC molecules and, hence,
an electrically switchable nanophotonic device based
on a sparse array of multiwall CNTs (MWCNTs)
grown on a silicon surface [7]. The nanotube electro-
des were grown vertically on silicon substrate. (per-
pendicular to the plane of silicon substrate). The
device operated in reflective mode. The Gaussian
electric-field profile from nanotube electrodes [8] dic-
tated the refractive index profile across the device.
The refractive index profile then acted as a series
of graded index profiles, which form a simple phase
modulating element [9]. By varying the external
Introduction
electric field applied, it was possible to tune the
properties of this graded index structure. In this
paper, we report electro-optic characterization and
optimization of a new nanophotonic device where
the nanotubes were grown vertically on a quartz sub-
strate. The device operated in transmission mode.
The transparent device has an advantage of making
compact electro-optic devices with less alignment
complexity. The response time, contrast ratio, and
phase modulating power of the nanophotonic device
were studied at different operating voltages. A step-
voltage driving scheme was used to eliminate the op-
tical bounce and effectively improve the rise time and
fall time of the device. The device applications were
proposed based on the variation of applied electric
field, which showed that lower driving voltages were
suitable for microlens array and phase modulation
applications, while higher driving voltages with a
holding voltage can be used for display applications.
Figure 1 shows a transparent nanophotonic device
with nanotubes attached vertically to the lower elec-
trode and an upper electrode acting as an earth plane
for the electric field. The lower electrode was made of
quartz substrate with a 100nm thick titanium
0003-6935/10/112099-06$15.00/0
© 2010 Optical Society of America
10 April 2010 / Vol. 49, No. 11 / APPLIED OPTICS2099

Page 2

nitride layer for electric contact, and the upper elec-
trode is indium tin oxide on 0:5mm thick borosilicate
glass. The cell gap wasset to 20μm using spacer balls
in UV glue and was filled with nematic LC BL048
from Merck. The top electrode alone was given a
planar alignment by rubbing a thin film of polyimide
(AM4276) and, hence, the resultant alignment of the
device was hybrid. In the current device, the nano-
tubes were patterned in small groups of six with a
1μm spacing between the nanotubes and 10μm spa-
cing between the groups to increase the resultant
electric field and, hence, the diameter of each phase
modulating element, as shown in Fig. 2.
2.
Device
Electro-Optic Characteristics of the Nanophotonic
A.
A He–Ne laser (633nm) was used to characterize
electro-optic properties of the nanophotonic device.
The device was placed between a crossed polarizer
Transmission Voltage Characteristics
and analyzer with a rubbing direction 45° to the
polarizer axis. Figure 3 shows the transmission (T)
versus voltage (V) curve for the device, along with
images of device switching at 0, 3, and 4:5Vrms.
The T–V curve goes through several maxima and
minima as the applied voltage increased from 0 to
15Vrms, and it attained a lower limiting value at
higher voltages. This is because the nanophotonic de-
vice acted as a phase retardation plate. The maxi-
mum value of measured phase retardation from
the T–V characteristics is 16π for the transparent de-
vice. This can be further explained using voltage-
dependent director orientation of LC molecules. The
rodlike nematic LC molecules in the device align
parallel to a preferred direction specified by the
director vector. When the laser beam propagates
through the planar aligned LC cell device with its
polarization axis at an angle of 45° to the director or-
ientation (rubbing direction), ordinary rays (no) and
extraordinary rays (ne) in the outgoing beam experi-
ence a phase difference δ [10]:
δ ¼2πdΔn
λ
sin2θ;
ð1Þ
where d is the cell thickness, Δnðne− noÞ is the bire-
fringence of the LC, λ is the wavelength, and θ is the
angle between the optic axis of the LC and the light
propagation direction. From the optical phase differ-
ence, the transmitted intensity through the device
between the crossed polarizer and analyzer can be
represented as I⊥:
I⊥¼ Ioexpð−α0dÞsin2δ
2;
ð2Þ
where Iois the incident beam intensity and α0is the
LC absorption coefficient for the ordinary ray. The
reason for the maxima and minima in the transmis-
sion curve is that the LC molecules are oriented by
the applied electric field, which changes the phase of
the outgoing ordinary and extraordinary rays and,
hence, creates changes in optical retardation, as gi-
ven in Eqs. (1) and (2). The transmitted intensity was
minimum for the device at higher voltages because of
homeotropic alignment of LC molecules at nanotube
positions. The alignment of the device was studied
further by rotating the device from 0° to 360° be-
tween the crossed polarizer and analyzer. The align-
Fig. 1.
photonic device.
(Color online) CNT and LC based transparent nano-
Fig. 2.
CNT electrodes on quartz.
Transmission electron microscope image of multiwall
Fig. 3.
photonic device along with an image of the device switching at 0, 3,
and 4:5Vrms.
(Color online) T–V characteristic of the transparent nano-
2100APPLIED OPTICS / Vol. 49, No. 11 / 10 April 2010

Page 3

ment at 0 and 3Vrmsis shown in Fig. 4. It was clear
from the analysis that the Gaussian electric field
from the CNT gave a more horizontal component
of electric field at lower voltages and, hence, planar
alignment predominated over the device, which re-
sulted in an increased transmission at lower voltages
compared to decreased transmission at higher vol-
tages because homeotropic alignment predominated
[10,11].
B.
The response time of a LC device is important for
making applications. When an external electric field
is applied, the field induces external torque to each
molecule, which leads to fast switching. But when
the voltage is removed or reduced, interaction be-
tween LC molecules provides the major restoring
forces and these forces are much weaker than the ex-
ternal-field-induced torque and hence slow switching
time. The response time varied from 1:57s to 125ms
for different voltages for BL048 nematic LCs with a
cell gap of 20μm [12]. The nanophotonic device was
characterized using a He–Ne laser of wavelength
633nm to measure the response time. Driving elec-
tronics were developed to make a holding voltage
(bias voltage) to drive the device from different vol-
tage levels using a counter, an analog switch, and lo-
gic gates. The device was placed between the crossed
polarizer and analyzer with the rubbing direction 45°
to the transmission axis. A voltage pulse of 3:12Vrms
without any holding voltage was used here to drive
the device, as shown in Fig. 5. The molecules were
allowed to relax after switching OFF the driving vol-
tage pulse, as shown in Fig. 5(a). After the field was
switched OFF, the retardance changed by a half-
wave in about 34ms (fall time), which is the time
to reach first minima in Fig. 5(a). The time required
for the consecutive half-waves of the stroke was in-
creased. The molecules near the surface switched
over faster than those in the rest of the device.
The final half-wave of the stroke required 1:4s [final
fall time to low state (OFF)] to relax. The rise time
was measured after driving the device from the
low state to the high state, as shown in Fig. 5(b).
The field was switched ON to 3:12Vrmsfrom 0Vrms
Response Time of the Device
using the driver electronics without any holding vol-
tage. The time required for the retardance to change
to half-wave was 36ms (rise time) immediately after
the device was switched ON. The consecutive rise
time increased with a final value about 335ms [final
rise time to high state (ON)].
The rise time and fall time were measured to calcu-
lateresponsetime(rise time þ fall time)ofthedevice
aftergivingaholdingvoltageusingthedriverelectro-
nics. The nanophotonic device was switched from vol-
tage levels corresponding to different maxima and
minima in the T–V characteristics (Fig. 3) to study
variationintheresponsetime.Wethenusedaholding
voltage and a transition voltage to switch the device
ON and OFF, corresponding to different voltage le-
vels, to improve the response time. Figure 6 shows
a transition from 4.7 to 8:5Vrms, where 4:7Vrmswas
used as the holding voltage. Figure 7 shows the re-
sponse time of the device for different holding and
transition voltages. Response time of the device var-
iedfrom3:1sto43ms.Theminimumresponsetimeis
43msforthe4.7to8:5Vrmstransition,where4:7Vrms
is the holding voltage. The response time was also
measured for a LC cell (without CNTelectrodes) with
the same thickness (20μm) as the nanophotonic de-
vice to find the effect of nanotube electrodes in the
switching speed of LCs. Response time of the LC cell
Fig. 4.
angle of rotation at two different voltages: 0 and 3Vrms.
(Color online) Transmission of the device with respect to
Fig. 5.
and (b) switched ON from 0 to 3:12Vrms(upper trace, optical
response; lower trace, electrical signal).
(Color online) (a) Device switched OFF from 3.12 to 0Vrms
Fig. 6.
with 4:7Vrmsas the holding voltage (upper trace, optical response;
lower trace, electrical signal).
(Color online) Device switched ON from 4.7 to 8:5Vrms,
10 April 2010 / Vol. 49, No. 11 / APPLIED OPTICS 2101

Page 4

varied from 3:8sto 100ms for different transitions. It
was observed that the nanotube electrodes slightly
improve the response time of LC molecules because
the nanotube electrodes reduce effective thickness
and also spawn more electric field due to a large as-
pect ratio. The difference in light intensity was mea-
sured when the device was ON and OFF to measure
contrast ratio. Figure 8 shows the contrast ratio mea-
sured for the different holding and transition vol-
tages. The contrast ratio is 5.4 for the transition 4.7
to 8:5Vrms. This transition, which has improved re-
sponse time and contrast ratio, can be used for dis-
play-related applications. The response time can be
further enhanced by tuning the thickness of the de-
vice and the contrast ratio by giving an alignment
to the bottom layer because the current alignment
of the device is hybrid. The fabricated device has a re-
solution of over 1000 × 1000 lenslets and a size of
10mm × 10mm. The nanotube electrodes can be rea-
lized as pixels to make high-density pixel arrays for
variousphotonicapplications.Thefringingfieldeffect
in such systems can be reduced with a suitably pat-
terned top electrode along with vertically grown
CNT bottom electrodes.
3.
The phase profile of the device was retrieved using a
Fourier transform technique [13–15]. An interfer-
ence setup attached to an optical microscope was
used to obtain interference fringes from the transpar-
ent nanophotonic device. Figure 9 shows interference
fringes from four lenslets and the two-dimensional
unwrapped phase of a single lenslet at 0.98, 2, and
3:5Vrms. The fringes were observed at 0Vrmsdue
to the alignment of LC molecules by CNT, even
though the top electrode was given a planar align-
ment. The interference fringes were more or less
circular and clear at lower voltages [9,16]. As the
Optical Phase Profile of the Device
Fig. 7. (Color online) Response time with respect to different voltage transitions.
Fig. 8.(Color online) Contrast ratio of the device with respect to different voltage transitions.
2102APPLIED OPTICS / Vol. 49, No. 11 / 10 April 2010

Page 5

voltage increased, the circular fringes became rec-
tangular in shape due to repulsion of electric field
from each CNT groups. The fringes started to disap-
pear around 2:5Vrmsand we observed a black spot at
each lenslet position. The phase modulation is 2:86π
at 0:98Vrmswith distorted phase profile. The phase
profile becomes smooth and parabolic, like at 2Vrms,
with phase modulation of 3:5π. The phase modula-
tion decreases and the phase profile starts to distort
as the voltage increases. The focal length of each
lenslet was calculated using the equation [9,16]
f ¼
Γ2
2OPD;
where OPD is the peak-to-valley optical path differ-
ence from the center to the edge of the hybrid grating
lenslet and Γ is the radius of the test area (5μm). The
focal length was 10μm at 0Vrmsand increased to
35μm at 3Vrms. Further increase in the voltage dis-
torted the orientation of LC molecules in the device
and, hence, no focusing was observed. The nanopho-
tonic device acts like a voltage reconfigurable micro-
lens array at lower voltages. The phase profile across
the device wasvoltage reconfigurable and that can be
used for hologram applications. A complex hologram
can be displayed in the device due to the complex vol-
tage reconfigurable refractive index profiles attain-
able in the device. From the electro-optic analyses
and interference experiments it was concluded that
the device’s applications can be divide into two re-
gimes, namely, low voltage and high voltage regimes.
The response time and contrast ratio of the nanopho-
tonic device is better at higher voltages and can be
used for display applications. But the phase modula-
tion is found maximum with a smooth paraboliclike
phase profile at lower voltages, which could find ap-
plication as a voltage reconfigurable microlens array,
inholograms,and in
applications.
other phase-modulating
4.
This paper presented the electro-optic characteris-
tics of a transparent nanophotonic device fabricated
with multiwall CNTs as electrode sites on a quartz
substrate covered with nematic LCs. The multiwall
CNTs acted as individual electrode sites that spawn
an electric-field profile, dictating the refractive index
profile within the LC and, hence, creating a voltage
reconfigurable optical element. The transmission
voltage characteristic, response time, and the con-
trast ratio were studied to optimize the device’s per-
formance. The analyses show that the device has
applications as a microlens array, a reconfigurable
hologram, and a grating at lower voltages, in addi-
tion to the suitability of the device for display appli-
cations at higher voltages. The phase profile of the
device was retrieved to study the voltage-dependent
phase modulation capabilities of the device. The de-
vice modulated light up to 3:5π.
The nanotube electrodes can be realized as pixels
to make submicrometer pixels with matrix addres-
sing for high-resolution display applications using
the nanophotonic device, because a device with a size
of 10mm × 10mm has over 1000 × 1000 phase mod-
ulating elements. The density of nanotube electrodes
can be further increased by growing nanotubes in
Future Device Developments and Conclusions
Fig. 9.
(a) 0.98, (b) 2, and (c) 2:5Vrms.
(Color online) Unwrapped phase of a single lenslet and interference fringes from the nanophotonic device (four lenslets) at
10 April 2010 / Vol. 49, No. 11 / APPLIED OPTICS2103