Department of Electronics and Communication

Government Engineering College, Thrissur

Seminar Report
2004

Carbon Nanotube Flow Sensors

Presented by

Ravi Krishnan Unni

S7 ECE
Roll No: 01-641
Govt. Engg. College, Thrissur Seminar Report ‘04

ACKNOWLEDGMENT

I would like to thank everyone who helped to see this seminar to

completion. In particular, I would like to thank my seminar coordinator
Mrs. Muneera.C.R for her moral support and guidance to complete my
seminar on time. Also I would like to thank Mr. C. D. Anil Kumar for
his invaluable help and support.

I would like to take this opportunity to thank Prof. Indiradevi,

Head of the Department, Electronics & Communication Engineering for
her support and encouragement.

I express my gratitude to all my friends and classmates for their

support and help in this seminar.

Carbon Nanotube Flow Sensors 2 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

CONTENTS

Abstract 4
1. Introduction 5
2. CNT structures and General properties 6
3. CNT electronic properties 9
4. Fluid flow through CNTs 11
4.1 Effect of Liquid Flow through Carbon Nanotubes 11
4.1.1 Verifications of Theoretical Predictions 12
4.1.2 Specifications of the materials
and Instruments used 12
4.1.3 The Experiment 13
4.1.4 Observations and Inferences 14
4.2 Applications 16
4.3 Effect of Gas flow through Carbon Nanotubes 17
4.3.1 Specifications of the materials
and Instruments used 18
4.3.2 Experimental setup, observations
and inferences 18
4.4 Applications 24
5. Conclusions 25
6. References 26

Carbon Nanotube Flow Sensors 3 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

ABSTRACT

Direct generation of measurable voltages and currents is possible when a fluids

flows over a variety of solids even at the modest speed of a few meters per second.
In case of gases underlying mechanism is an interesting interplay of Bernoulli's
principle and the Seebeck effect: Pressure differences along streamlines give rise to
temperature differences across the sample; these in turn produce the measured
voltage. The electrical signal is quadratically dependent on the Mach number M and
proportional to the Seebeck coefficient of the solids. This discovery was made by
professor Ajay sood and his student Shankar Gosh of IISC Bangalore, they had
previously discovered that the flow of liquids, even at low speeds ranging from 10-1
metre/second to 10-7 m/s (that is, over six orders of magnitude), through bundles of
atomic-scale straw-like tubes of carbon known as nanotubes, generated tens of
micro volts across the tubes in the direction of the flow of the liquid.
Results of experiment done by Professor Sood and Ghosh show that gas flaw
sensors and energy conversion devices can be constructed based on direct
generation of electrical signals. The experiment was done on single wall carbon
naontubes (SWNT).These effect is not confined to naotubes alone these are also
observed in doped semiconductors and metals.
The observed effect immediately suggests the following technology application,
namely gas flow sensors to measure gas velocities from the electrical signal
generated. Unlike the existing gas flow sensors, which are based on heat transfer
mechanisms from an electrically heated sensor to the fluid, a device based on this
newly discovered effect would be an active gas flow sensor that gives a direct
electrical response to the gas flow. One of the possible applications can be in the
field of aerodynamics; several local sensors could be mounted on the aircraft body
or aerofoil to measure streamline velocities and the effect of drag forces.
Energy conversion devices can be constructed based on direct generation of
electrical signals i.e. if one is able to cascade millions these tubes electric energy
can be produced.

Carbon Nanotube Flow Sensors 4 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

1. INTRODUCTION

As the state of art moves towards the atomic scales, sensing presents a major
hurdle. The discovery of carbon nanotubes by Sujio Iijima at NEC, Japan in 1991
has provided new channels towards this end.
A carbon nanotube (CNT) is a sheet of graphene which has been rolled up and
capped with fullerenes at the end. The nanotubes are exceptionally strong, have
excellent thermal conductivity, are chemically inert and have interesting electronic
properties which depend on its chirality.
The main reason for the popularity of the CNTs is their unique properties.
Nanotubes are very strong, mechanically robust, and have a high Young’s modulus
and aspect ratio. These properties have been studied experimentally as well as using
numerical tools. Bandgap of CNTs is in the range of 0~100 meV, and hence they
can behave as both metals and semiconductors. A lot of factors like the presence of
a chemical species, mechanical deformation and magnetic field can cause
significant changes in the band gap, which consequently affect the conductance of
the CNTs. Its unique electronic properties coupled with its strong mechanical
strength are exploited as various sensors. And now with the discovery of a new
property of flow induced voltage exhibited by nanotubes discovered by two Indian
scientists recently, has added another dimension to micro sensing devices.

Carbon Nanotube Flow Sensors 5 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

2. CNT STRUCTURE AND GENERAL

PROPERTIES

There are two main types of nanotubes that can have high structural perfection.
Single walled nanotubes (SWNTs) consist of a single graphite sheet seamlessly
wrapped into a cylindrical tube (Fig. 1, A to D). Multiwalled nanotubes (MWNTs)
comprise an array of such nanotubes that are concentrically nested like rings of a
tree trunk (Fig. 1, E).

Fig.1 : Schematic representation of structures of carbon nanotubes (A, A1)

armchair, (B, B1) Zigzag and (C, C1) Chiral SWNTs. Projections normal to the
tube axis and perspective views along the tube axis are on the top and bottom,
respectively. (D) Tunneling electron microscope image. (E) Transmission electron
microscope (TEM) image of a MWNT containing a concentrically nested array of
nine SWNTs.

The structure of a nanotube can be defined using a roll-up vector r and/or a chiral
angle θ as shown in Fig. 2. The rollup vector can be defined as a linear combination
of base vectors a and b of the basic hexagon r = na + mb with m and n being
integers. The roll-up vector is perpendicular to the axis of the nanotube. In Fig. 2,

Carbon Nanotube Flow Sensors 6 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

the shaded zone is the area which is rolled up along an axis perpendicular to the
roll-up vector. Different types of nanotubes are defined by the values of m and n
thus chosen. Three major categories of nanotube structures can be identified based
on the values of m and n:
m=n “Armchair”
m = 0 or
n=0 “Zigzag”
m≠n “Chiral”

C.

Fig 2: Relation between the hexagonal carbon lattice and the chirality of Carbon
Nanotubes.
(A) The construction of a carbon nanotube from a single graphene sheet. By rolling
up the sheet along the wrapping vector C, that is, such that the origin (0, 0)
coincides with point C, nanotube is formed. Wrapping vectors along the dotted lines
lead to tubes that are zigzag or armchair. All other wrapping angles lead to chiral
tubes whose wrapping angle is specified relative to either the zigzag direction (q )
or to the armchair direction (F= 30 ْ - q ). Dashed lines are perpendicular to C and
run in the direction of the tube axis indicated by vector T. The solid vector H is
perpendicular to the armchair direction and specifies the direction of nearest-

Carbon Nanotube Flow Sensors 7 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

neighbor hexagon rows indicated by the black dots. The angle between T and H is
the chiral angle F.
( B ) Schematic of a two-dimensional graphene sheet illustrating lattice vectors a1
and a2, and the roll-up vector Ch = na + mb. The limiting cases of (n, 0) zigzag and
(n, n) armchair tubes are indicated with dashed lines. As represented here, the angle
between the zigzag configuration and Ch is negative.
(C) Chiral nanotube showing angle F.

The electronic properties of perfect MWNTs are rather similar to those of

perfect SWNTs, because the coupling between the cylinders is weak in MWNTs.
Because of the nearly one-dimensional electronic structure, electronic transport in
metallic SWNTs and MWNTs occurs ballistically (i.e., without scattering) over
long nanotube lengths, enabling them to carry high currents with essentially no
heating. Phonons also propagate easily along the nanotube: The measured room
temperature thermal conductivity for an individual MWNT (3000 W/m K) is greater
than that of natural diamond and the basal plane of graphite (both 2000 W/m K).
Superconductivity has also been observed, but only at low temperatures, with
transition temperatures of 0.55 K for 1.4-nm-diameter SWNTs and 5 K for 0.5-nm-
diameter SWNTs grown in zeolites. Small-diameter SWNTs are quite stiff and
exceptionally strong, meaning that they have a high Young’s modulus and high
tensile strength. With the total area per nanotube in a nanotube bundle for
normalizing the applied force to obtain the applied stress, the calculated Young’s
modulus for an individual nanotube is 0.64 TPa, which is consistent with
measurements. Because small-diameter nanotube ropes have been extended
elastically by5.8% before breaking, the SWNT strength calculated from the product
of this strain and modulus is 37 GPa, which is close to the maximum strength of
silicon carbide nanorods (53 GPa) This modulus of 0.64 TPa is about the same as
that of silicon carbide nanofibers (0.66 TPa) but lower than that of highly oriented
pyrolytic graphite (1.06 TPa).
More impressive and important for applications needing light structural
materials, the density-normalized modulus and strength of this typical SWNT are,
respectively,19 and 56 times that of steel wire, respectively, and 1.7 times that of
silicon carbide nanorods. The challenge is to achieve these properties of individual
SWNTs in nanotube assemblies found in sheets and continuous fibers.

Carbon Nanotube Flow Sensors 8 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

3. CNT ELECTRONIC PROPERTIES

Electrically CNTs are both semiconductor and metallic in nature which is

determined by the type of nanotube, its chiral angle, diameter, relation between the
tube indices etc. The electronic properties structure and properties is based on the
two dimensional structure of Graphene. For instance if the tube indices, n and m,
satisfies the condition n-m=3q where q is and integer it behaves as a metal. Metal,
in the sense that it has zero band gap energy. But in case of armchair (where n=m)
the Fermi level crosses i.e. the band gap energy merges. Otherwise it is expected the
properties of tube will be that of semiconductor. The table below (Table 1) is the
observations of experiments done on nanotubes by Scanning tunneling microscope
(STM) and Scanning tunneling spectroscopes (STS). The dependency of energy gap
or band gap energy on chiral angle and diameter is clear from the observation
below.

Table 1. Here d is the nanotube diameter; F is the chiral angle; E gap is the apparent
band gap in the STS I–V spectra and dE is the shift of the Fermi energy due to
doping of the tube by the substrate. Note that a chiral angle 0ْ denotes armchair
nanotube and an angle of 30ْ a zigzag tube. The flat Au surface allowed the
diameter d of the nanotubes to be determined with an accuracy of 0.1nm by
measuring the tube heights relative to the surface. A possible systematic uncertainty
in determining the diameter is due to a difference in barrier heights for the gold
substrate and the tubes. The wrapping angle F can be determined with an accuracy
of 1ْ. Accuracy in F is limited by the curvature of the tubes. A combination of high
accuracy in both F and d (0.05 nm) is required for an unambiguous identification of
the n, m indices. Accuracy in Egap and dE is 0.05–0.1 eV. For this sample a shift

Carbon Nanotube Flow Sensors 9 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

in the Fermi energy towards the conduction band was observed, instead of a shift
towards the valence band as observed in the other samples.

From the table above, two categories can be distinguished: one with gap values
around 0.5–0.6 eV, the other with significantly larger gap values. The first category
of tubes is identified as the semi conducting type, the second as metallic tubes.
Another noted property of nanotube is that oscillation of Band Gap energy when
magnetic field is applied parallel to the tube axis. It has been showed theoretically
that the band gap energy would oscillate with increasing magnetic field, so a metallic
tube is converted to a semiconductor and then metallic again, with a period depending
on the magnetic field strength.

Fig 3 : Electronic properties of single-

walled carbon nanotubes.
(A). Current–voltage curves obtained by
tunneling spectroscopy on various
individual nanotubes is shown here. Tubes
no. 1 to 6 are chiral, no. 7 is zigzag and no.
8 is armchair. The bias voltage is applied
to the sample, which means that the sign of
Vbias corresponds to that of the energy
relative to the tube Fermi level. Curves no.
1–7 show a low conductance at low bias,
followed by several kinks at larger bias
voltages. The armchair tube does not show
clear kinks in the range 1 to 1V
(B) Energy gap Egap versus diameter d for
semiconducting chiral tubes. The data
points correspond quite well to the
theoretical predicted values. The solid line
denotes a fit of:

Egap = 2g0aC–C /d with g0 = 2:7 eV.

Where, g0 is the c-c tight binding energy.

aC–C, the nearest neighbour distance
(.142nm).
Tunnel currents I as a function of the bias
voltage V applied to the sample were
recorded with a home-built STM while
scanning and feedback were switched off.

Carbon Nanotube Flow Sensors 10 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

4. FLUID FLOW THROUGH CARBON

NANOTUBE

Recently there has been extensive study on the effect of fluid flow through
nanotubes,which is a part of an ongoing effort worldwide to have a representative in
the microscopic nano-world of all the sensing elements in our present macroscopic
world. Indian Institute of Science has a major contribution in this regard. It was
theorotically predicted that flow of liquid medium would lead to generation of flow-
induced voltage. This was experimetally established by two Indian scientist at IISc.
Only effect of liquid was theorotically investigated and established experimentally,
but effect of gas flow over nanotubes were not investigated, until A.K Sood and
Shankar Ghosh of IISc investigated it experimentally and provided theorotical
explanation for it. The same effect as in case of liquid was observed, but for entirely
different reason.
These results have interesting application in biotechnology and can be used in
sensing application. Micro devices can be powered by exploiting these properties.

4.1 Effect Of Liquid Flow Through Nanotube

P král and M shapiro published a paper in Physical review letters [ Vol.86

No.131 (2001)], that dealt with development of voltage / current when liquid flows
through CNTs. Generally an electric current in a material is produced when flow of
free charge carriers is induced in the material. According to Král and Shapiro, the
generation of an electric current in a nanotube is essentially due to transfer of
momentum from the flowing liquid molecules to acoustic phonons in nanotube so
as to have a dragging effect on the free charge carriers in the nanotube. The
outcome, according to these workers, is a linear dependence of the induced electric
current on the flow velocity.
Another mechanism involved, as per these authors involves a direct scattering
of the free carriers from the fluctuating Coulombic fields of the ions or polar

Carbon Nanotube Flow Sensors 11 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

molecules in the flowing liquid. They argued, however, that the latter mechanism
creates a current that is five orders of magnitude smaller than the current that results
from the phonon-induced electron drag.
These predictions were experimentally verified by A K Sood and Shankar
Ghosh of IISc. In sharp contrast, Sood and coworkers have found that the behaviour
is highly sub linear where the induced voltage fits a logarithmic velocity
dependence over nearly six decades of velocity. Strong dependence of induced
voltage on polarity and ionic nature of the liquid and relatively weak dependence on
its viscosity was revealed from this experiment.

4.1.1 Verification on Theorotical Prediction:

Experimental setup, observed results and hence the inference gathered from the
observations, dependencies of voltage induced on various factors that has been
observed are explained in following sections

4.1.2 Specification of materials and instuments used:

The SWNT bundles used for the experiment were prepared by electric arc
method, followed by purification process. The dimensions and physical properties
of samples are listed below:

Diameter of nanotube 1.5nm

Resistivity 0.02Ω-m

Sensor using this tube was prepared by densly packing nanotubes between two
metal electordes. The dimensions of Sensor:

Along the flow 1mm

Thickness 0.2mm
Width 0.2mm

Carbon Nanotube Flow Sensors 12 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

The specimen was placed inside glass tube of innner dia 0.03 and length 0.9m ,
kept vertically. Voltage mesurment were used by using KEITHLEY 2000
multimeter.
Hydrocloric and Glycerine to increase polarity and viscocity of fluid media
respectively.

4.1.3 The Experiment:

The scematic representation of experiment is shown in figure 4.To avoid any

turbulent flow that may have been caused by the expansion of the flow at the inlet
of the flow chamber it was placed at the center of glass tube. The velocity u of the
liquid flowing past the sample was measured from the bulk flow rate, which was
varied by adjusting the height of the reservoir and the opening of the valve. The
liquid enters the flow chamber against gravity such that the formation of air
pockets is avoided. The flow at the center of the cell is expected to be laminar
because the Reynolds number (Re) (~300 for a velocity of 10-2 m/s) was much
less than the critical value of 2000 for the onset of turbulent flow in a pipe. The
measurements shown here were taken after the transients had subsided.

Fig 4 :
Experimental setup where R is the
reservoir, L is the valve controlling
the liquid flow, S is the cylindrical
glass flow chamber, and G is the
Voltmeter.

Carbon Nanotube Flow Sensors 13 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

4.1.4 Observations and Inferences:

When the sample was immersed in water at rest, a voltage (~1 mV) developed
along the sample as an electrochemical potential difference at the interface of the
SWNT bundle with the metal electrodes. This offset voltage was subtracted from
the voltage measured at finite velocities of the liquid to obtain the flow-induced
voltage. The voltage developed only along the flow direction; no voltage was
generated in the direction perpendicular to the flow (i.e., along the width).
Figure 5 shows the observed
dependence of the flow
induced voltage on flow
velocity of water. Even at a
very low speed of 5x10-6 m/s, a
voltage of 0.65 mV was
generated. The solid line in
Fig5 is a fit to empirical
relation:
V= a log(ub+1)

Where a and b are parameters, a= 0.6mV and b= 6.5 x 106 s/m.

Figure 6 compares the voltage developed along nanotubes as a function of the flow

Carbon Nanotube Flow Sensors 14 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

velocity for a series of fluid mixtures. A notable effect is the strong dependence of
the flow-induced voltage on the ionic strength of the flowing liquid. It can be seen
that for u = 10-4 m/s, the voltage for 1.2 M HCl is about five times that for water.
A comparison of expected results as per Král and shapirov and experimental results
are compared in following paragraph,
They considered metallic nanotubes that were densely packed in two layers
separated by a layer of water of thickness d (~10 nm). In their mechanism, the
force exerted by the liquid with a flow velocity u on the surface of the nanotubes
is:
F = (hu/d) X interface periphery X interface length.
This force results in the steady-state quasi-momentum (p) transfer to the
phonon bath was calculated using p ~ Ft where h is the shear viscosity of the

liquid, t known as phonon umklapp time. Using this value of p, the resulting

current was calculated, the results showed that current depends on viscosity of
fluid and flow velocity Vs flow induced voltage would be linear in nature. Using
glycerol mixed with water at ratios 88:12 and 75:25 having h= 113 and h= 234
mPa s respectively revealed the weak dependence of induced voltage on h. The
induced voltage decreased with increase in shear viscosity yet the effect was
negligible.
The experiment also revealed the dependence of flow induced voltage on
polarity of liquid which is clear from figure 6, there is a steep increase in flow
induced voltage with increase in polarity of liquid. 0.6 M and 0.12 M HCL and
methanol were used for this purpose. Thus, for u =6 x10-4 m/s, Vwater = 2.1 mV,
whereas Vmethanol = 0.2 mV. Thus, data set reveals three features: the sublinear
(almost logarithmic) voltage response to flow velocity, its strong dependence on
the ionic and polar nature of the flowing liquid, and its relatively weak
dependence on the liquid viscosity—the latter, in fact, acting to reduce the flow-
induced voltage.
The predominant mechanism here that results in the flow induced voltage is
fluctuating columbic fields.

Carbon Nanotube Flow Sensors 15 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

As ions and polar molecules have fluctuating Coulomb fields, Sood and
coworkers have assigned the observed behaviour to scattering of free carriers by
these fluctuating fields. The exact model is termed as ‘the pulsating asymmetrical
ratchet model’. The terms describe the phenomenon itself, whereby ions flowing
past a certain point in the nanotube produce Coulomb pulses while the asymmetry
is provided by the velocity gradient at the liquid–solid interface.
An important point must be made regarding the direction (polarity) of the
voltage and current in relation to the fluid flow direction. In the experiments, it
was observed that the electric current and the fluid were flowing in the same
direction. In a linear response theory based on the viscous drag (the phonon wind),
the particle current has the same direction as the fluid flow velocity, and,
therefore, the sign of the electric voltage and current is determined by the sign of
the charge carriers (electron or hole). Thus, for the hole like carriers, the induced
voltage and current have the same direction as the flow velocity. This, however, is
not the case for a pulsating ratchet model in general. For a given direction of the
asymmetry, the direction of the voltage is independent of the sign of the charge on
the carriers. It will, of course, depend on the sense of the bias (asymmetry) of the
ratchet potential. In a polar fluid of a given ionic strength, such as concentrations
of H+ and OH– ions (along with their hydration shells), it is expected to have
ratchet potentials of either sign (bias). The net effect is then determined by the
dominant ratchet. What is important here, however, is that there is no a priori
(symmetry) reason to expect the oppositely biased ratchets to give an exact
cancellation.

4.2 Applications

Flow sensor that is based on SWNT directly produces an electrical signal in

response to a fluid flow. This sensor can be scaled down to length dimensions of
the order of micrometers—i.e., the length of the individual nanotubes—making it
usable in very small liquid volumes. The sensor also has high sensitivity at low
velocities and a fast response time (better than 1 ms). The nanotubes also could be

Carbon Nanotube Flow Sensors 16 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

used to make a voltage and current source in a flowing liquid environment, which
may have interesting biomedical applications.
In bio medical applications it is speculated that carbon nanotubes can be used
to power coronary heart pace makers using flowing blood as source of power this
would make the use of batteries in coronary heart pace makers obsolete. The
results using blood have shown encouraging results.
Another application would be as a chemical sensor which can be used to check
the polarity, acidity and detecting any impurities (that are polar in nature) present
in the liquid media.
If CNTs become cheaper and economically viable then it can be used in bulk
to produce electricity in small scale using flowing water as source of active media.

4.3 Effect of Gas Flow over Carbon Nanotubes

The effects of Gas flow over CNTs were not studied both theoretically and
experimentally, until A K Sood and his Co workers investigated on it. Papers
pertaining to these works were published in Physical Review letters on 20 Aug
2004. Research works on these findings and possible applications are being
investigated by Sood and coworkers.
The same effect of flow induced current/voltage observed with liquid media
was observed when gas medium was used. Although the effect is same, yet the
reason for it is completely different from that of liquid flow. Further this effect is
not confined to nanotubes, same effect is observed in wide range of solids
including semiconductors, metals for wide range of velocities.
The underlying principle is an interesting interplay of Bernoulli’s Principle
and Seebeck effect. Pressure differences along the streamlines give rise to
temperature difference along the tube which, in turn produce the measured
voltage. Unlike in case of liquid the flow induced voltage in gas, depends not only
on flow velocity but also on the orientation of specimen. The experimental setup,
observation, findings, inferences and theory behind this phenomenon is explained
in the following sections.

Carbon Nanotube Flow Sensors 17 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

4.3.1 Specifications of Specimen and Instrument:

To establish the ubiquity of this phenomenon variety of solids were used, they
include nanotubes, semiconductors and metal which are described below:

n-type Ge(Sb-doped) of s(conductivity)=100/Wcm, p-type Si s =100/Wcm,

n-type Si s =100/Wcm, SWNTs, MWNTs graphite, poly-crystalline cooper and

platinum were used. The sizes of the semiconducting samples and copper are
about 3 mm along the flow and 1mm perpendicular to the flow. The electrical
contacts for the semiconducting materials (Si, Ge) were made with copper leads of
1.3X10-4 m dia using silver emulsion [shown by shaded region in schematic
representation of experimental setup fig.7] The exposed (active) portion of the
sample, that is not covered by silver emulsion is about 2mm [marked as d in Fig
7b] along the flow and 1mm perpendicular to the flow. The samples of SWNTs,
MWNTs and graphite, prepared by densely packing the powder between two
metal electrodes, were about 1mm along the flow, 2mm wide and 0.2mm thick.
Proper electromagnetic interference shielded cables were used to measure the
signals using a KEITHLEY 2000 multimeter.
Active medium i.e. gases used were oxygen, argon, nitrogen and air. These
were contained in a gas chamber kept at a pressure of 150 bars. The flow velocity
was measured indirectly using a Rota meter.

4.3.2 Experimental setup, Observations and Inferences:

Figure 7, (next page) shows a schematic layout of experiment of the

experimental setup to achieve a calibrated gas flow velocity on the sample. The
gas from a compressed gas cylinder (maximum pressure of 150 bars) is let out at a
given pressure in a tube of diameter 7 X 10-3 m. The flow rate Q is measured
using a Rotameter as shown in Fig. 7(a). The average velocity u at the end of the
tube with cross-sectional area (Φ) is deduced from Q as u= Q ⁄Φ. The sample is

kept at an angle α=π⁄4 with respect to the horizontal axis to achieve the optimal

Carbon Nanotube Flow Sensors 18 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

signal. At an angle α= 0ْ would produce no effect since the pressure gradient is

zero, and α=π⁄2 too would give no signal due to symmetry. The results are similar
when the sample is kept inside the tube (at a distance of 2 X 10-2 m from the exit
point) or 1 X 10-2 m outside the tube.

Fig 7 : (a) Schematic of the experimental setup. The flow rate at the exit point is
deduced from the measured flow rate at the side port using the rotameter. (b)
Sample: shaded portions mark the electrodes. The positive terminal of the
voltmeter is connected to the right ® of the sample of active length d kept at and
angle α=π⁄4 with respect to the horizontal axis.
Figure 8 below shows the voltage across the n-Ge sample as a function of time
when the gas flow (u= 7 ms) over the sample is switched on and off. The gas used
here is argon. Data for the steady flow voltage V for a variety of samples and flow
velocities are shown in Figs. 9 and 10. It can be clearly seen that for nitrogen gas
over p-Si, n-Si, n-Ge, p-Ge, SWNT, MWNT and graphite, the voltage V

Carbon Nanotube Flow Sensors 19 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

generated varies as u2 over a wide range of u ( as does the current, not shown).
Figure 9 shows the fit to V= Du2 with fit parameter D given in Table2. This is
even clearer in Fig 9 which shows the same data plotted versus M2, where M is the
Mach number given by
M=u / c, c being the sound
velocity (=333 m/s for
nitrogen and 323 m/s for
argon at 300K). The solid
lines fit to curve V=AM2,
where A is a fitting
parameter also given in
Table 2. Experiments
using polycrystalline
copper sheet for which the
slope A is very small. It
can be seen that the sign of P-type Si and SWNT is opposite to that for n-type Si,
n-type Ge, graphite and copper. SWNT samples are usually unintentionally p-
doped, which can explain the sign of the flow-induced voltage which was found to
be the same for SWNT
and p-Si. The inset in
fig10 shows the slope A
versus the known
Seebeck coefficient S of
the samples of the same
d, as given in Table 2.
The coefficient depends
linearly on S, as shown
by fitted line in the inset,
with slope=60K.
Figure 10 shows the
generated voltage V over

Carbon Nanotube Flow Sensors 20 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

a large range of values of M2 for the flow of argon (solid squares) and nitrogen
gases (open circles) over n-Ge. A change in slope is clearly seen around M2≤.05
(Called regime 1) is higher for argon than for nitrogen.

The ratio of slopes [A (argon) / A (nitrogen) = 1.2]. That there are actually two
M2 regimes is clear from our theoretical analysis of the mechanism behind the
generation of electrical signal induced by the flow of gases over the solids, and is
discussed below. For the adiabatic steady inviscid flow of a gas, Bernoulli’s
equation gives the pressure difference along a streamline in terms of Mach
number M as:

CP (CV) being the specific heat at a constant pressure (volume). The values of
g for argon and nitrogen are 1.667 and 1.404, respectively. In eq1 P0 is the

maximum pressure at a point on the streamline where velocity is zero . Such a

point is the leading edge on the surface of the sample past which the gas is moving
and is called the stagnation point. For the sample geometry shown in Fig. 7 (b),

Carbon Nanotube Flow Sensors 21 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

the pressure difference between the two ends of the active sample exposed to the
gas flow (i.e., without the electrodes) is hence:

The subscripts L(R) denote the left (right) of the active sample when the gas
flows from left to right. From the ideal gas law, the fractional temperature
difference DT/T is related to the pressure difference DP/P and the density
difference Dr/r as
DT/T=DP/P-Dr/r. When M<<1, the change in density of the gas is
negligible, i.e., the fluid is essentially incompressible and hence DT/T=DP/P.
Therefore, the temperature difference along a streamline between two points
separated by distance d for M<<1 (called regime I) is:

Where, DT= TL - TR > 0. The gas flowing past the sample kept at an angle a
with respect to the horizontal axis corresponds to an accelerating flow and hence
MR>ML.The tangential component of the velocity u of the outer flow depends on
the streamline distance x measured along the flat boundary as:

For this geometry a= pi/4 and hence u(x) .The temperature difference along
the streamline in the gas flow will induce a temperature difference in the solid
along the flow direction. The temperature difference, in turn, will result in a
voltage difference V, defined as VL – VR, due to the Seebeck effect. Thus:

Carbon Nanotube Flow Sensors 22 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

Where, S is the Seebeck coefficient of the solids,

positive for p-type and negative for n-type materials. The factor k depends on the
specific interactions between the gas and the solid surface as well as on the
boundary conditions of the temperature difference between the gas and the solid.
In nonrarefied gases, the boundary conditions at the surface of a solid is that the
temperatures of the gas and solid are equal, in which case k=1. But this boundary
condition applies only if the mean free path of the gas molecules is vanishingly
small. The factor k takes into account all the differences from this ideal boundary
condition. The present phenomenon of voltage generation by flow of gases is not
applicable to the flow of liquids where the viscous drag dominates.
Beyond a certain value of M (~0.2), called regime II the density changes of the gas
should be taken into account, which gives:

And hence:

On comparing the predictions of this model with the experiments:

1. From equation 4 that it can be deduced that induced voltage V is directly

proportional to M2 and the experimental results show that the curve fits to
V=A M2 .

2. From eq 5 the slope A should depend linearly on S in regime 1 where :

The inset in figure 10 agrees with the result obtained theoretically.

3. Another property can be deduced from the equations above that for
substance that has S=0 will have no induced voltage. Results with platinum
as specimen have showed no induced voltage (for Pt S~0).

Carbon Nanotube Flow Sensors 23 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

4. From equation 5 in regime 1 ratio of voltage generated for argon and

nitrogen with same velocities for same samples were 1.2 which is the ratio
of g (argon) / g (nitrogen), hence consistent with theoretical predictions.

Hence, the theoretical predictions deduced are consistent with the experiment.

4.4 Applications

The experiments clearly suggest that a sensor to measure the flow velocity of
the gases can be made based on the generated electrical signal. It is an active
sensor which gives direct electrical response to the gas flow. This should be
compared with the widely used gas flow sensor based on thermal anemometry,
wherein the fluid velocity is sensed by measuring changes in heat transfer from a
small, electrically heated sensor (wire or thin film) exposed to the fluid. Thermal
anemometry works on heat balance equations and hence any small changes in the
temperature, pressure, or composition of the gas can cause erroneous readings.
Such effects are minimum or can be easily taken into account in the sensors based
on the direct generation of flow-induced voltage or current in the sensor material.
The magnitude of generated voltage can be easily scaled up by using series and
parallel and connections of sensing elements. This suggests that flow energy can
be directly converted into electrical signal without any moving part, thus having a
potential for application in generating electricity.
Ajay Sood and his coworkers have shown that this property can be exploited
for construction of nanotube based vibrational sensor in liquids and as
accelerometer. The SWNT based accelerometer has so far measured the frequency
range 0.5 Hz to 1 kHz and the minimum detectable acceleration by the
accelerometer is 10-3 g.

Carbon Nanotube Flow Sensors 24 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

5. CONCLUSION

The Past as Indication of the Future: The exponential increase in patent

filings and publications on carbon nanotubes indicates growing industrial
interest that parallels academic interest.
Nanotubes electronic devices might be the most promising field. Impressive
advances have been made in demonstrating nanotube electronic device concepts,
but a decade or more of additional progress is likely required to reliably assess if
and when these breakthroughs will reach commercial application, So will the
nanotube sensing applications as flow sensors / electrical signal generators. As
for now control on their electronic properties which is not so impressive where
large scale production is considered, poses a major hurdle in this regard.
Possible chemical sensing applications of nonmetallic nanotubes are interesting,
because nanotube electronic transport and thermo power (voltages between
junctions caused by inter-junction temperature differences) are very sensitive to
substances that affect the amount of injected charge. The main advantages are
the minute size of the nanotube sensing element and the correspondingly small
amount of material required for a response. The mechanical robustness of the
nanotubes dramatically increase probe life and minimize sample damage during
repeated hard crashes into. These uses may not have the business impact of
other applications, but they increase the value of measurement systems for
characterization and manipulation on the nanometer scale. With the addition of
this new phenomenon i.e. flow induced voltage nanotube seems to head for
ubiquity in nanoworld as sensor, actuator probes, energy storage and generating
devices.
Independent of the outcome of the ongoing races to exploit nanotubes in
applications, carbon nanotubes have provided possibilities in nanotechnology
that were not conceived in the past. Nanotechnologies of the future in many
areas will build on the advances that have been made for carbon nanotubes.

Carbon Nanotube Flow Sensors 25 Dept. of Electronics and Communication

Govt. Engg. College, Thrissur Seminar Report ‘04

6. REFERENCES

1. Carbon nanotube flow sensors Vol.299 Science, by A.K. Sood

Shankar ghosh and N.Kumar, Dated: Feb 14 2003
2. Direct generation of electricity by gas flow through CNTs, Physical
Review Letters No. 8 by A K Sood and Shankar Ghosh, Dated:
Aug 20 2004
3. Science compass Carbon Nanotubes-The route toward applications
SCIENCE Vol.297, Dated Aug 2 2004
4. Carbon Nanotubes-The next generation. Final report
University of California Dated March 22 2004
5. Electronic Structures of Atomically resolved CNTs by Dept. of
Applied physics Delft University
NATURE Vol.391, Dated Jan 1 1988
6. Frontline Vol.20 issue 03, Dated Feb 01 - 14 2003
7. Frontline Vol.21 issue 18, Dated Aug 28 – Sept 10
8. http://www.nanotech.org, Dated Jan 17 2003
9. http://www.economictimes.com, Dated Sep 23 2004
10. http://www.indianexpresss.com, Dated Jan 17 2003
11. http://www.aip.org/pnu/subpnu.htm, No.627, Dated Aug 19 2004
12. http://www.sciencemag.org, Dated Sept 22 2004

Carbon Nanotube Flow Sensors 26 Dept. of Electronics and Communication