cantilever sensing used for
biological and chemical species
Department of Mechanical Engineering (REU), Iowa State University, Ames, Iowa
In this report, we present a cantilever sensor capable of static and dynamic sensing with
high sensitivity and one that can be used in a liquid environment without
altering the setup of the system. A Mach
Zehnder type interferometer is combine
d in the
device to allow alteration of the divided light waves
used in determining the
interference signal. Detection of biological and chemical species in a liquid medium was
determined using the
deformation of the sensing cantilever
corresponding relationship among
path length difference
was also verified.
Static deformation was also
the concepts of optics and
Utilizing the concepts of cantilever deflection and interferometry,
we demonstrate sensing capabilities of surface stress sensing
in and out of a liquid
The deformation of the sensing cantilever with respect to the reference
cantilever creates a m
easured response which is insensitive to most environmental
for enhanced accuracy
ng and the search for the ideal
cantilever sensor have been an important
topic among researchers.
sensitive and simple means of measuring cantilever deflections
because of its applications. Cantilever deflections are used
Scanning Tunneling Microscopy (
, various drug detections, DNA
on, as well as the indicat
ion of proteins, antibodies,
, and cancerous cells
sensing innovations generally
lead to an increase in sensitivit
and a smaller means of measuring
cantilever beam deflection.
an interferometry system th
of various optical
, a sensing and referencing pair. This
of static and dynamic sensing
in and out of a liquid medium.
Currently, there are dozens of ways to appl
cantilevers as transducers to
Lavrik et al 
a single cantilever
could be used where a
single beam of light is reflected and measured.
Figure 1 provides a visual depiction of the setup.
: Single cantilever beam deflection 
The single cantilever sensor consists of a cantilever beam and a means of recording the
deformation that occurs within the cantilever. To determine that deformation, a
a position sen
sitive photodetector are used. Essentially, the laser shoots a beam of light toward
the reflective coated cantilever where that light then reflects toward the photodetector. As the
cantilever deforms, the light
positions from where the photodetec
tor initially sensed the
beam of light. That change
needed results which can then be used to determine the
change in deflection of the cantilever beam. The advantages of this single cantilever and optical
am deflection technique are a
precision, as well as a
reliability. Additionally, there is an
readout efficiency and reduction of size of the
system. On the contrary, the disadvantages are that the system may easily be affected by
ectromagnetic, or acoustic fields such as a person talking. Likewise, the system may
also be affected by thermally induced noise and temperature
. The other problem with
the single cantilever
technique is that the distance between the cantileve
r and position sensitive
todetector is relatively large.
Since the sensitivity is proportional to the distance between the
photodetector and the cantilever components, the greater the distance the more sensitive the
system is to nanoscale cantilever de
The system is becoming smalle
r, but still not
a great amount of sensitivity
Although the cantilever and its
components are small, the m
eans of measuring the
deformation of the beam is not. This large
distance also c
ontributes to a loss of intensity and directionality of the reflected optical beams,
Second, a single cantilever could be used with a reference and sensing beam of light as
presented in Cunningham et al .
fter the coherent beam of light is split in two,
one beam reflects off the base of the cantilever and the other reflects
off the tip of the cantilever.
Dual cantilever beam deflection 
On the cantilever base, there will be no measurable deflection which establishes a reference
point. However, on the cantilever tip, there will be measurable deflection which establishes a
sensing point. Through interferometry, the measurable phase change
between the reflected
reference beam and the sensing beam can be determined and results can be
advantages of this technique include less sensitivity to outside noise, an overall smaller device,
as well as a resolution increase.
, the primary disadvantage is that this concept is
hard to apply to commercial microcantilevers. This set up works well with larger cantilevers, but
the micromachined, commercially
be used where two beams of light are aimed at a
reference and sensing cantilever pair as explained by Shrotriya et al . The differential surface
stress sensor consists of two similar cantilevers: a reference cantilever and a sensing cantile
of the setup
can be seen in Fig
Optical differential surface stress sensor 
This technique requires full understanding of path length difference as well as phase changes and
interference. This technique measures the cantilever deflection using
fiber optic interferometry
advantages of using dual cantilever beams essentially
the results. It decreases accuracy among unwanted factors and yet increases
among the wanted, sought
after, results. Basically the main improvements eliminate
and disturbance, any changes in pH, and atmospheric fluctuations
such as temperature change. Sensitivity and resolution are also incre
ased with this technique.
On the contrary, the disadvantages of using this two beam technique with fiber optics are
lens alignment factors and an overall more complicated system. Measurements
this technique and
DNA hybridization, however, are not dependent of the
distance between the cantilever components and the photodetector. This provides an overall
simplified mircocantilever sensing device applicable to read world situations.
There are two main
types of cantilever sensing: static sensing and dynamic sensing. Surface
stress sensing results in a static or fixed deflection of the cantilever beam. The cantilever is
s that are
number of molecules on the cantilever increases, and whether or not the molecules share an
attractive force, the cantilever will deflect. The coated surface may decrease or increase in
due to the binding of chemicals or biological sp
and cause the measurable fixed
Tensile surface stress
leads to upward
the cantilever beam
and can be seen in Fig. 4A
ompressive surface stress
leads to downward def
of the cantilever beam
provided in Fig. 4B
Regardless of the deflection of the cantilever beam, the deformation
c mode de
flection (A) Tensile
Mass sensing results in a dynamic or vibrant deflection of the cantilever beam. Similar to the
static mode, the cantilever during the dyn
amic mode is again coated with
Here the beam vibrates at its resonant frequency
and while molecules attach themselves to the cantilever, the frequency will decrease.
schematic of the dynamic mode is provided in Fig. 5.
That decrease in
the frequency is
to the addition of the mass of the molecules. If performed in a liquid medium, the
frequency will be dampened and the sensitivity may decrease as compared to an air medium.
Although the frequency is significantly dampened, t
he differential bending of the sensing and
reference cantilevers are expected to overcome this problem.
Dynamic mode deflection 
The two novel
types of interferometry include Michelson and Mach Zehnder interferometry
Michelson, as provided in Fig.6
, consists of a coherent light s
ource, a half
mirrors, and a photodetector.
Basically, the light waves split, recombine with a
phase change, and reach the photodetector for sensing.
en the two beams reach the
interfere and create a fringe pattern of light and dark rings or, in some cases,
. This fringe pattern correlates precisely with the additional wavelength created by
moving mirror A or mirror B forward or backward.
Constructive interference will occur when
the separated light waves differ by a whole number of wavelengths
and will res
ult in the bright
. In contrast, destructive interference will occur when the two waves differ by a half
and will show up as the dark areas on the fringe
. For most purposes, the geometry of
Zehnder interferometer is more usefu
l and more applicable to experiments.
Michelson type interferometer [
For the Mach
Zehnder interferometer, the collimated light source is split and two beams are
sent to two detectors. Unlike the Michelson technique,
Zehnder is not folded upon
itself and each individual be
am may be manipulated however
. This manipulation can
exist as light wave plane changes as well as a change in speed of the wave
and is an important
factor for our technique. The in
terferometer provided in Fig. 7, is used to measure the phase
shift of a thin opaque sample such as glass. The glass would be placed before either mirror and a
relative amount of light entering the detectors would allow a calculation of the resulting phas
shift. This technique requires precise alignment and understanding of
refraction indices as well
Our setup essentially uses one mirror or the reflective surface of the cantilevers,
but the concepts of the device remain identical.
Zehnder type interferometer [9
provided in Fig. 8
along with a simplified schematic
Zehnder type interferometer
(L. Magnan 2011)
The first step in the alignment process was to secure the
fiber coupled laser source (Thorlabs: S1FC635)
for a consistent
Once secure, and the placement of the light beam
was marked on the back of the bookshelf
the polarized lens
and the beam displacer
were added. The purpose of the polarized lens is to polarize the light into one plane while the
ates the incoming beam into two beams on different planes 90 degrees
The next step was to cancel out one of the two beams produced by
. This was
accomplished by manipulating and rotating the polarized lens.
Once one axis was cancelled out,
Babinet Compensator (
VIS) was placed on the rail. The rotation
locking thumb screw
dial was then
loosened and the SBC
VIS was turned
only one beam of light on the bookshelf.
This was performed to establish an
axis on the polarizer.
Finally, the polarized lens was turned 45 degrees from the original position
to allow equal
intensity beams out of the BD27
and the other parts were placed on the
. The liquid cell
factured at ISU)
, the Wollaston Prism (
the concave lens (Model
and the photodetector
were added respectively.
components, among the aforementioned parts, were all
post attachments were modified for this setup
was oriented the same as the polarized lens, 45 degrees with respect to the SBC
. The purpose of this orientation is to take the
ight waves and separate
them into two
interfered beams which reach each photodetector
. Since the tilt of the prism is 45
degrees, it will take the
horizontal and vertical components of the incoming beam
which is where the interfere
nce takes place.
The main significance of the Soleil
Babinet Compensator is to change the plane of the light
waves as well as create a slow wave axis with respect to the fast axis. Additionally, the
rotational dial on the end may
be manipulated to determ
ine the length of the wave length of the
The alignment procedure for our experiments consisted of several steps. With the orientation
of the optical components correct, the first step was to insert the cantilevers and align the two
s on the tip of the cantilevers.
This was performed with the laser intensity at
. Figure 9
shows the alignment of the cantilevers.
: 250x magnification of cantilever alignment
(L. Magnan 2011)
If the canti
levers were not centered on the light beam and roughly
apart, the cantilevers
were pulled out and remounted on the cantilever holder.
Once in place, the cantilevers were
to the cantilever base and the laser intensity was turned to approximately
Next, a concave lens (Thorlabs: LMR05) was placed in one of the detector posts to truly align
both reflected beams for a fringe pattern.
To find the fringe pattern, the
cantilever mounts had to
be moved side to side and forward and backward respectively until
the desired fringe appeared
step consisted of r
ising the cantilev
ers back onto the tips and, with the same intensity as
the previous step, the SBC
turned approximately one wavelength and verified with
For the final alignment step, the buffer was added into the liquid cell and the
wavelength was again checked with the SBC
The rest of the experimental
procedure is explained in
the second validation.
To validate this
were performed in a laboratory setting.
to test the SBC
one wavelength, as determined
steps followed in the manual (Thorlabs: SBC User Guide)
corresponds to one w
avelength on the computer imaging program
To begin the
the beam was focused
on a flat mirror, repres
enting the reflective surface of
n, in .
starting from 0.000mm,
rotated outward until 19.000mm. By recording the SBC values with respect to the values on the
computer, a graph of intensity vs. SBC movement could be constructed
shown in Fig. 10
experiment was performed 3
19.000mm and 3 times from 19.000
Wavelength verification of SBC
VIS movement with respect to intensity output
A second validation was performed usi
ng the cantilevers, rather than a mirror
and the buffer
was inserted in the liquid cell
Again, in .500mm increments, the SBC rotational dial was rotated
outward until 19.000mm.
It was also performed once rotating the dial inward from 19.000mm to
so an average could be determined. Fig
is the average of the two
Wavelength verification of SBC
VIS movement with respect
to intensity output in buffer
With the reflection being off
the cantilevers, the unfiltered intens
ity reading on LabV
fluctuated so hand recordings were not possible. Rather, a different technique was attempted
where the SBC value was changed quickly and held constant for 5 seconds to establish a pattern
e computerized graph. From the graph, the results could be picked out and a new graph
could be constructed.
Intensity Output (V)
VIS Dial Movement (mm)
Intensity Output (V)
VIS Dial Movement (mm)
Once all necessary results were obtained and recor
ded, the SBC was rotated to a
intensity of the both waves was
equal. At this
point, the cantilever deflection will be
the most sensitive and responsive
to the sensing cantilever’s deflection.
Finally, the setup was ready for introduction of
The first injection contained
ced a solution with a final
. The second injection contained
again and created a
solution with a final concentration of
and produced a final concentrated solution
mutant protein. 30
minutes were allotted after each injection to ensure a reaction if capable. See Fig. 12 and Fig. 13
for the results on the mutant protein and non
specific binding. The small surface
toward the end of the experiment is solely due to the negligible intermolecular reactions taking
place on the cantilevers surface.
, the sensing and reference cantilevers
were heated in d
eionized water at
to regenerate the protein sequence. The regeneration
allows the sensing cantilevers to be used a number of times (Baker, Lai et al. 2006) and each
cantilevers could be used for at least three sensing experiments.
experiment, we introduced a wild
type protein into the
. The first injection contained
and produced a solution with a final
protein. The second injection contained 100µL
created a solution with a final concentration of
injection contained 100µL
and produced a final concentrated solution
protein. 30 minutes were allotted after each injection to ensure a reacti
on if capable. See Fig. 12
and Fig. 13 for the results on the wild
type protein and specific binding. Notice the reaction and
cantilever deformation happens soon after the injection
and then remains relatively steady
each 30 minute interval
the SBC was rotated
back to the
position where the intensities
This was done to ensure maximum sensitivity of the sensing cantilever’s deformation.
From these results
, the path length difference, or th
e amount of beam deflection, that
the reaction can be determined. In addition, a relationship between surface stress and
concentration can be composed.
Experimental data for specific and non
specific protein binding
Experimental data for wild
type and mutant protein
In sum, the search for the ideal transducer continues.
ew innovations generally lead to
better sensitivity and a smaller means of measuring the beam defl
ection. In our case, static and
dynamic responses may be measurable with the same device. A measurement may also be taken
from a liquid medium where the dampening of the frequency has little to no effect.
two experiments in this report, we are
able to conclude that the injection of the mutant protein
caused little to no cantilever deformation. However, the injection of the wild
type protein caused
relatively a good amount of cantilever deformation. Also the specific binding of the wild
otein causes tensile and compressive stress on the cantilever surface.
interferometer keeps the system compact while the dual cantilevers hinder surrounding
vibrations and thermal noise. This system may also remain highly sensitive
and useful toward
and is amendable toward insertion in a MEMS device
Although the surface stress sensing was included in this report, mass detection
was not. In
ure, this device will test
mass detection from
namic standpoint. The
explaining mass detection
in a liquid environment due to the dampening effect. Our setup, composed of the differential
should gain back
some of the sensitivity
lost through dampening
substances and molecules may also be tested with this device such as DNA hybridization and
other protein sequences.
Throughout this research, there were many contemporary
ethical issues and obstacles
The novel principles such as voluntary participation, informed consent, and risk of
harm were taken into consideration.
Due to the equipment in the laboratory, such as the laser
used in this report, consideration f
or my safety as well as the safety of others was always in
Another important consideration during our experiments was computer usage. It would
be great if everyone could have their own computer, programs, and laboratory equipment. Since
ent is expensive, sharing was necessary
and sometimes difficult. During our
experiments, the lights in the laboratory had to be shut off to reduce error in the results. This
was also difficult because of the others in the lab. In the end, everyone worke
d out a system that
Since all research is unique, n
o set of ethical standards can be established or
every ethical implication.
That is why it is important to take all research
differently and consider all possible eth
ical implications in a serious manner
as done for this
This research was supported by the Iowa State University REU Program along with K. Kang
and P. Shrotriya. The author would like to thank the program and its members for
helpful discussions, and assistance in design.
1) N. V. Lavrik et al.: Rev. Sci. Instrum.
2) M. J. Cunningham et al.: Meas. Sci. Technol.
3) P. Shrotriya et al.: Appl. Phys. Lett.
(2008) 143107 1
4) C Schonenberger and S. F. Alvarado: Rev. Sci. Instrum.
5) V. K. Sandeep: Online Nanotech. Jour. AZojono
) B. Llic and H. G. Craighead et al.: Appl. Phys. Lett.
) J. Fritz
: Anal. Chem.
D. M. Harrison: Phys. Bookshelf