Detection of Oligonucleotide – Gold Nanoparticle conjugates Using Cantilever Arrays Operated in Dynamic Mode Larry O’Connell 08390860 School of Physics Trinity College Dublin Supervisors: Prof. Martin Hegner Ph.D. Student Jason Jensen Nanobio-Nanomechanics Group Centre for Research on Adaptive Nanostructures and Nanodevices Trinity College, Dublin December 2011
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Detection of Oligonucleotide – Gold Nanoparticle conjugates Using Cantilever Arrays Operated in
Dynamic Mode
Larry O’Connell 08390860
School of Physics
Trinity College Dublin
Supervisors: Prof. Martin Hegner
Ph.D. Student Jason Jensen
Nanobio-Nanomechanics Group Centre for Research on Adaptive Nanostructures and Nanodevices
Trinity College, Dublin
December 2011
i
Abstract
This project endeavoured to demonstrate the ability of a micromechanical cantilever array-
based device, operating in dynamic mode, to detect binding to the cantilever surface of 12-
mer oligonucleotides in solution. The oligonucleotides were attached with a thiol bond to 50
nm diameter gold nanoparticles, while the complimentary oligonucleotide sequence was
similarly attached to the cantilever surface. Only non-specific binding was detected. The
sensitivity of the device was found to be approximately 3.4 pg/Hz.
ii
Preface
This final year project was conducted over a 2 month period from September 26th
until
November 25th
, with the Nanobio-Nanomechanics research group in the Centre for Research
on Adaptive Nanostructures and Nanodevices (CRANN), at Trinity College, Dublin (TCD),
under the supervision of Prof. Martin Hegner.
iii
Acknowledgments
I would like to thank my supervisor Prof. Martin Hegner for the opportunity to work on this
project. I would also like to thank all staff in the Nanobio-Nanomechanics group for creating
a welcoming atmosphere. Finally, I would like to thank Jason Jensen for his patience and
guidance, without which this project would not have been possible.
All image and photo credits are to the author unless otherwise stated.
Contents
Abstract ................................................................................................................................... i
Preface .................................................................................................................................... ii
Acknowledgments ................................................................................................................ iii
A typical functionalization pattern is shown in Table 1. The cantilever is allowed to
incubate for 20 minutes, allowing the probe molecules to self-assemble into a monolayer on
the cantilever’s upper and lower surfaces. The arrays are then placed in 10 mM sodium
phosphate buffer in a refrigerator.
2.2 Spectrophotometry
All oligonucleotide solutions bought from Microsynth were analysed using a
NanoDrop 1000 Spectrophotometer before being used for conjugation. The procedure
involved first calibrating the spectrophotometer before each measurement using nanopure
water. Samples on the scale of 4µl of each solution were used. Using a nearest-neighbour
model from Tataurov, You, and Owczarzy [2008] we can calculate ελ=260nm for arbitrary
sequences, and we arrive at the values in Table 2.38
2.3 Oligonucleotide-nanoparticle conjugation
Table 2 - Oligonucleotide sequences used in this experiment. The ‘SH’
denotes the thiol group bound to the 5’ end of each nucleotide species.
Name Sequence Calculated extinction coefficient at 260 nm
(L mol−1
cm−1
)
Bio B2 SH-5’-TGC TGT TTG AAG-3’ 113500
Bio B2 Compliment SH-5’-CTT CAA ACA GCA-3’ 117700
Bio B3 SH-5’-CCG GAA GAT TGC-3’ 116200
Bio B3 Compliment SH-5’-GCA ATC TTC CGG-3’ 109600
Unspecific 12 SH-5’-ACA CAC ACA CAC-3’ 119200
Bio B4 SH-5’-GGA AGC CGA GCG-3’ 120800
The oligonucleotides are mixed and agitated with DEE to remove the DTT from the
thiol group. DEE and the water are immiscible and so the DEE and oligonucleotide
suspension separate within seconds with a visible meniscus. A micropipette is used to remove
the DEE and the process is repeated 6 times. To completely remove the DEE, the
oligonucleotide solution is placed in a SpeedVac concentrator for 5 minutes. The
oligonucleotides are then ready for conjugation.
The conjugation protocol that was followed necessitated Dynamic Light Scattering
analysis of the nanoparticles to ascertain their mean diameter and the extent of polydispersity,
prior to conjugation. A Malvern Zetasizer Nano ZS dynamic light scattering apparatus was
used. After carrying out the DLS measurement of the mean diameter, we carry out
21
spectrophotometric analysis of the oligonucleotide solution to determine an exact figure for
the molar concentration of the oligonucleotides (Table 3.5). We use these values to calculate
the amount of oligonucleotide solution needed to conjugate the particles according to Eq. 5.22
An example calculation of oligonucleotide necessary for conjugation is as follows:
The specification sheet for our gold nanoparticles indicates a value of 7.473 108
or
equivalently 4.49 1010
nanoparticles per ml. Using Eq. 5 we calculate the required amount of
oligonucleotides to conjugate 1 ml of nanoparticles:
[20]
Adding the 50% molar excess as recommended by the protocol, we obtain a value of 0.1855
nmol needed to ensure good coverage. Using the oligonucleotide concentration values from
spectroscopic analysis, we can calculate the volume of oligonucleotide solution needed. We
produced separate solutions of Bio B2-functionalized nanoparticles, and Bio B3-
functionalized nanoparticles. Both our solutions were at 100 µM concentration, thus we
needed 1.855 µl of the solution to add to the 1 ml of gold nanoparticles.
The mixture is placed in a glass vial which is covered in tin foil and agitated on a
linear shaker at ~1 Hz for 16 hours. The tin foil serves to prevent exposure to light which
hinders the reaction.24
The mixture is then brought to a 10 mM Sodium Phosphate concentration which acts
as a pH 7 buffer. The addition of Sodium Phosphate buffer is split into 5 smaller additions
and gradually added over 5 hours as is recommended for nanoparticles larger than 20 nm.24
22
The sodium phosphate serves as a pH7 buffer to facilitate DNA binding.
The solution was then centrifuged at 4000rpm for 15 mins in low-adhesion Eppendorf
tubes wherein they form a crimson oil of nanoparticles beneath a clear supernatant of excess
oligonucleotide in solution. The supernatant was removed and retained for analysis. This is
done to remove the free oligonucleotides from the suspension which would otherwise
hybridize with the cantilever-bound complimentary strands, thus preventing those sites from
binding with the nanoparticle-bound oligonucleotides. The nanoparticle oil was then
resuspended in the same volume of identical molar concentration of 10 mM sodium
phosphate buffer. This solution was then centrifuged again and the process was repeated 6
times, with the supernatant retained each time.
The final solution should be virtually devoid of free oligonucleotides at this point.
Spectrophotometric analysis is then carried out on the final solution and each of the
supernatants. This data is used to quantitate the successful binding of the oligonucleotides to
22
the nanoparticles. A drop in the total number of free oligonucleotides in the supernatant
indicates successful conjugation with the gold nanoparticles.
2.4 Dip test
This experiment relies on the successful conjugation of the nanoparticles with the
oligonucleotides. Running the experiment involves a significant time investment. If the
procedure fails to produce an obvious resonance frequency shift, it is important to know
whether this is due to a measurement error in the device or a failure to conjugate the
nanoparticles and/or bind them to the cantilever surface. To test this, we functionalized
several cantilevers using the standard protocol outlined in Section 2.1.4. Rather than attempt
to use a frequency shift measurement in the full apparatus to detect binding, we bathed the
functionalized arrays in low-adhesion Eppendorf tubes, each containing one of several
solutions (either Bio B2, Bio B3, or bare gold nanoparticles) and then imaged them under a
Scanning Electron Microscope (SEM). This is known as a preliminary “dip test”.
2.5 Dynamic mode measurement
The sensor flow cell and supplying circuit were flushed with ethanol at a rate of 225
µl/minute for ~90 minutes. The sensor flow cell was then flushed with nanopure water at the
same rate for ~90 minutes. For all subsequent solutions, a flow rate of 18.2 µl/min was used,
corresponding to a bulk velocity in the circuit of 1.2 m/s . As a cleaning process, the circuit is
filled with 10 mM Sodium Phosphate pH7 buffer for 2.5 mins and then left static for 42.5
minutes. The Bio B2 nanoparticle solution was then injected into the analyte storage loop and
flowed into the sensor flow cell at 18.2 µl/min for 2.5 mins after which the flow was stopped
and left static for 42.5 mins. Following this 10 mM Sodium Phosphate was flowed through
the device at 18.2 µl/min for 10 mins and the flow was left static for 30 mins. The Bio B3
nanoparticle solution was then injected into the storage loop and flowed through the sensor
flow cell at 18.2 µl/min for 2.5 mins after which the flow was left static for 42.5 mins. Finally
10 mM Sodium phosphate buffer was again flowed through the device at 18.2 µl/min for 10
mins, followed by a final 30mins of static flow.
2.6 Data analysis
All data analysis was carried out using data analysis software NOSEtools. This
software runs in the IGOR Pro environment. The model used in this software is described in
Braun et al. [2005]36
and is outlined in Section 1.10.
23
The 7th
and 8th
resonant modes were monitored. 1000 data points in the frequency
interval 120 kHz – 270 kHz were taken giving a frequency resolution of ~150 Hz. Each
frequency is excited for 1ms and the response rate was sampled at a rate of 107 samples per
second. The peaks were fitted with a amplitude spectrum of a simple harmonic oscillator, and
the time evolution of the centre frequency of the peaks was used to calculate mass uptake.
The peak centre frequency ( ) and width ( ) were taken from the fit and used to calculate
the quality factor for each peak ( ). The standard error was also calculated using
the statistics function in OriginPro 8.
The data were baseline corrected by calculating the overall linear drift of the
cantilever resonance frequencies over the baseline period (the initial 45 mins of the
experiment, see Fig. 14a). This was done for each cantilever individually. The resonant
frequency trace for each cantilever was saved in a time-stamped file and analysed in
post-processing using NOSEtools. The data was then normalized such that only relative
frequency shifts are apparent (Fig. 14b). A median filter (box size 7) was applied to the data
to reduce noise. Finally the mass uptake was calculated from this frequency trace by fitting
each frequency spectra with the model outlined below. Plots were made of bound mass vs.
time to determine the binding behaviour during the experiment.
Figure 11 – The amplitude spectrum for a simple harmonic oscillator, fitted to the 7th and 8th resonant modes
24
Chapter 3 Results & Discussion
3.1 Dip test verification of compliment-specific binding
Figure 12 – A typical SEM indicating a positive result in a dip-test. The cantilever shows bound nanoparticles of
mean diameter 50 nm. The visible halo surrounding this cluster is due to local charge concentration. Note that this
image is from an earlier run of the experiment which used the same equipment and fabrication techniques. This
particular image shows non-specific binding of bare gold nanoparticles to an unfunctionalized cantilever.
A typical SEM image showing successful binding will exhibit randomly distributed
particles with a mean diameter of approximately 50 nm (Fig. 13), while the reference
cantilever will not. The SEM images of our cantilever showed low levels of non-specific
binding of nanoparticles to the upper surfaces of all cantilevers. There was no obvious target
specific binding of the nanoparticles, and no apparent correlation between specificity of a
cantilever’s functionalization and the observed binding to that cantilever.
25
3.2 Dynamic-mode measurements
Our data shows no discernable specificity to the mass uptake during the course of the
experiment. This suggests non-specific binding of Bio B2 nanoparticle conjugates to all
cantilevers, followed by a universal drop in mass-uptake during the period that buffer was
flowed through the cell (90- 95 mins) following the Bio B2 stop-flow period. This is possibly
due to a rinsing effect of the buffer, removing loosely bound nanoparticles. Following this we
see a common mass uptake across all cantilevers during the end of the buffer flow-through
period (95 – 100 mins). The trace from this period (90 – 100 mins) could be an artefact of the
change in flow conditions in the device. Following this, during the stop-flow buffer period
(100 – 130 mins), we see common drift across all cantilevers which seems to continue at the
same rate during the introduction of the Bio B3 functionalized nanoparticles (130 – 175 mins)
and the period of buffer flow-through (175 - 215 mins).
Figure 13 – Left: Non baseline-corrected data. Right: Baseline-corrected data. The hatched regions indicate the
period of liquid flow through the sensor flow cell. The unhatched regions indicate the stop-flow period during which
no liquid flowed through the sensor flow cell. The red region indicates the period during which Bio B2 conjugates
were present in the bulk liquid in the cell, and similarly the blue region indicates the period during which Bio B3
conjugates were present in the bulk liquid. The white regions indicate the period during which 10 mM sodium
phosphate buffer was flowed through the cell. The traces of each of the compliment-functionalized cantilevers are
shown, with their respective functionalizations indicated in the legend. The trace of mass uptake has been
normalized, to show relative mass uptake; and baseline-corrected, to disregard drift due to extraneous factors
(temperature drift, loosening of the clamp against the array due to vibration, etc.).
(a) (b)
26
3.3 Dynamic light scattering: nanoparticle size distribution
Functionalization
Mean Diameter
(nm)
FWHM (nm)
None (Bare gold) 50.52 22.02
Bio B2 57.83 30.53
Bio B3 65.19 45.31
Figure 15 – Dynamic Light Scattering measurements for unfunctionalized, Bio B2-functionalized, and Bio B3-
functionalized nanoparticles.
The DLS results are shown in Fig. 15. The nanoparticles bought from BBI Life
Sciences were found to be sufficiently monodisperse, having a mean diameter of 50.52 nm
with a full-width half-maximum (FWHM) of the intensity trace of 22.02 nm. The Bio B2
functionalized nanoparticles were found to have a mean diameter of 57.83 nm, which
conforms to our expectation of the size increase. Intuitively, we would expect the diameter of
the nanoparticles to increase by double the length of the attached oligonucleotides. The
oligonucleotides have a length of approximately ~4 nm, and so the observed diameter
increases (after conjugation) of 7.8 nm and 15.2 nm are reasonable.
However, the intensity trace shows a leg on the left hand side for both conjugated
nanoparticles, indicating significant scattering around the 6-8 nm mark.
0
5
10
15
20
1 10 100
Intensity (%)
Radius (nm)
Bare gold nanoparticles
0
5
10
15
1 10 100
Intensity (%)
Radius (nm)
Bio B2 Conjugates
0
5
10
15
1 10 100
Intensity (%)
Radius(nm)
Bio B3 Conjugates
27
3.4 ζ-potential measurements
Functionalization ζ-potential (mV)
None (Bare gold)
-0.535
Bio B2 -0.393
Bio B3 0.488
Figure 16 – ζ-Potential measurements for unfunctionalized, Bio B2-functionalized, and Bio B3-functionalized
nanoparticles.
The ζ-potential measurements are shown in Fig. 16. The results gave nonsensical
values for the ζ-potential of all particle solutions. We would expect to see a value of -30mV
or less, since gold exhibits a negative surface charge25
and the colloid is empirically observed
to be stable. The ζ-potential appears to vary around 0mV. This value is not possible given the
observed stability of the colloids.
3.5 Spectrophotometry
Spectrophotometric analysis of the oligonucleotide solutions allowed us to calculate
the 10mm absorbance of the oligonucleotide solutions (Fig. 17). The calculated
oligonucleotide concentration was 258.6 ng/µl for the Bio B2 solution, and 156.2 ng/µl for
the Bio B3 solution. Dividing these values by the molecular weight of each oligonucleotide
species, we can calculate the molar concentration of their respective solutions.
0
50000
100000
-50 0 50
Total Counts
ζ-potential (mV)
Bare gold nanoparticles
0
50000
100000
-50 0 50
Total Counts
ζ-potential (mV)
Bio B2 conjugates
0 40000 80000
120000 160000 200000
-50 0 50
Total Counts
ζ-potential (mV)
Bio B3 conjugates
28
Figure 147 – Spectrophotometry results for Bio B2 and Bio B3 solutions
Analysis of the conjugates solutions yielded oligonucleotide concentrations which were too
low to be measured. Successful conjugation is suggested, however, by the observed diametric
size increase of the nanoparticles after conjugation (Section 3.3).
3.6 Discussion
Selective metallization of the cantilever surface has been shown to significantly
improve the Q-factor of single-crystal silicon cantilevers.40
A 2005 study18
looking at gold-
coated silicon cantilevers, observed a severe degradation in Q factor compared to bare
cantilevers. Their work suggested confining the metalized layer to the tip of the cantilever as
a method of reducing dissipation. To illustrate the potential improvement from selective
metallization, the damping caused by metallization of the hinge accounted for ~60% of the
total damping caused by a full coat.40
The signal to noise ratio in the cantilever response
could possibly be improved by refraining from metalizing the cantilever hinge. This
improvement may be negligible, however, as the viscous damping contribution to the Q-
factor typically dominates at atmospheric pressure.18
We know from Rayleigh’s approximation that the intensity of scattering of a particle
is proportional to the 6th
power of its diameter29
(see Section 1.7). The observed leg on the
intensity traces for the conjugated nanoparticles (Fig. 16) could be an obscured “peak”
indicating a significant presence of 6-8 nm scale scatterers. Since this is size measurement is
consistent with expected oligonucleotide length, this could indicate the presence of free
oligonucleotides that either came loose after conjugation with the nanoparticles, or perhaps
were never removed during the conjugation protocol. Alternatively, the observed small scale
0 1 2 3 4 5 6 7 8 9
220 240 260 280 300 320 340
Ab
sorb
ance
Wavelength (nm)
10mm Absorbance vs Wavelength
Bio B2 solution (66.2 µM 258.585 ng/µl)
Bio B3 solution (40.4 µM 156.19 ng/µl)
29
scattering could be explained as simply arising from contamination. The conjugated
nanoparticles are put through several processes that the bare gold nanoparticles are not,
exposing them to air and passing them to and from different containers, and this could result
in an unknown contaminant. However, the small scale of the apparent contaminant scatterers
(6-8 nm) makes it unlikely that it is due to dust particles (on the order of 500 µm) or even
bacteria (on the order of 500 nm). Thus it is reasonable to assume the observed small-scale
scattering is due to scattering from free oligonucleotides. This could explain the absence of a
clear mass-uptake trend in the dynamic mode cantilever experiment, since free
oligonucleotides would hybridize with their cantilever-bound compliments, thus passivating
potential binding sites for the mass-tagged conjugates.
30
Chapter 4 Conclusion
In the Dynamic mode experiment, we would have expected to see a larger mass-
uptake in the Bio B2c functionalized cantilevers during the period that the cantilever is
bathed in Bio B2 conjugates, followed by a plateau in mass uptake while buffer and then Bio
B3 conjugates were flowed through the sensor flow cell. Concurrently, the Bio B3
functionalized cantilevers should have shown no uptake during the same Bio B2 bathing
period. We then should have seen a larger mass-uptake in the Bio B3c functionalized
cantilevers during the period that the cantilever is bathed in Bio B3 conjugates. We would
expect to see some binding of nanoparticles to reference cantilevers, and slightly less to the
control cantilevers (Bio B4 functionalization) due to a passivation effect of having a mono-
layer of oligonucleotides that would not hybridize with nanoparticle-bound oligonucleotides.
The results did not conform to expectations. We can conclude after baseline correction and
normalization that all cantilevers showed mass binding.
This experiment was unsuccessful in establishing the efficacy of dynamic mode
cantilever array detection of gold nanoparticle-bound oligonucleotides. Though we observed
a negative result for the dynamic mode experiment, we cannot ascertain why this was
unsuccessful. The problem may lie in a failure to conjugate the nanoparticles with the
oligonucleotides, or alternatively the problem may be a failure of the nanoparticles to retain
their oligonucleotide covering. Indeed, due to the wealth of existing research done using
nanosensing cantilevers, we conclude that the observed results are a consequence of the
conjugation protocol followed, rather than the device design itself. We arrive at this
conclusion since our device successfully detected mass-uptake during the course of the
dynamic-mode experiment, (albeit it due to non-specific binding) while exhibiting sensitivity
on the scale of ~4 pg/Hz with very low noise of approximately ±0.5 ng,
It stands as a testament to the precision of the device’s design, that we can detect such
small mass uptake since very few methods allow such a degree of sensitivity. Furthermore,
even fewer technologies allow such sensitivity in a liquid that resembles physiological
environments, and it is this prerequisite which places cantilever arrays in a crucial position in
the field of probing biological processes. Further work is needed to prove that this method is
as promising as many similar methods being investigated in this exciting sub-field of
biophysics.
31
References
1 Joseph Wafula Ndieyira, Moyu Watari, Alejandra Donoso Barrera, Dejian Zhou, Manuel Vogtli, Matthew Batchelor, Matthew A. Cooper, Torsten Strunz, Mike A. Horton, Chris Abell, Trevor Rayment, Gabriel Aeppli, and Rachel A. McKendry, Nat Nano 3 (11), 691 (2008).
2 David Maraldo, Fernando U. Garcia, and Raj Mutharasan, Analytical Chemistry 79 (20), 7683 (2007).
3 Sen Xu and Raj Mutharasan, in Environmental Microbiology - Current Technology and Water Application, edited by Nicholas J. Ashbolt Keya Sen (Caister Academic Press, 2011), pp. 65.
4 Natalia Nugaeva, Karin Y. Gfeller, Natalia Backmann, Marcel Düggelin, Hans Peter Lang, Hans-Joachim Güntherodt, and Martin Hegner, Microscopy and Microanalysis 13 (01), 13 (2007).
5 J. Zhang, Hans-Peter Lang, F. Huber, A. Bietsch, W. Grange, U. Certa, R. McKendry, H.J. Guntherodt, Martin Hegner, and Ch. Gerber, Nat Nano 1 (3), 214 (2006).
6 Kathrin Gruber, Tim Horlacher, Riccardo Castelli, Andreas Mader, Peter H. Seeberger, and Bianca A. Hermann, ACS nano 5 (5), 3670 (2011).
7 Rüdiger Berger, Emmanuel Delamarche, Hans Peter Lang, Christoph Gerber, James K. Gimzewski, Ernst Meyer, and Hans-Joachim Güntherodt, Science 276 (5321), 2021 (1997).
8 Gossett A. Campbell and Raj Mutharasan, Analytical Sciences 21 (4), 355 (2005). 9 Gossett A.Campbell and Raj Mutharasan, Biosensors and Bioelectronics 21 (9), 1684 (2006). 10 Qing Zhu, Wan Y. Shih, and Wei-Heng Shih, Biosensors and Bioelectronics 22 (12), 3132
(2007). 11 Gossett A.Campbell and Raj Mutharasan, Biosensors and Bioelectronics 23 (7), 1039 (2008). 12 Gossett Campbell, David Maraldo, and Raj Mutharasan, in Nanoscience and Nanotechnology
for Chemical and Biological Defense (American Chemical Society, 2009), Vol. 1016, pp. 25. 13 Nathaniel Rosi and Chad Mirkin, Chemical Reviews 105, 1547 (2005). 14 J. A. Dougan, C. Karlsson, W. E. Smith, and D. Graham, Nucleic acids research 35 (11), 3668
(2007). 15 L. M. Prescott, J. P. Harley, and D.A. Klein, Microbiology, 6 ed. (McGraw-Hill Education,
Boston, 2005). 16 Daniel Thevenot, Klara Toth, Richard Durst, and George Wilson, Biosensors \&
Bioelectronics 16, 121 (2001). 17 Murali Krishna Ghatkesar, Thomas Braun, Viola Barwich, Jean-Pierre Ramseyer, Christoph
Gerber, Martin Hegner, and Hans Peter Lang, Applied Physics Letters 92 (4), 043106 (2008). 18 R. Sandberg, K. Mølhave, A. Boisen, and W. Svendsen, J. Micromech. Microeng. 15 (12),
2249 (2005). 19 Madeleine Price Ball, (Wikimedia Commons, 27 March 2007). 20 Bharat Bhushan, Springer Handbook of Nanotechnology, 2 ed. (Springer, 2006). 21 Hans Peter Lang, Martin Hegner, and Christoph Gerber, Materials Today 8 (4), 30 (2005). 22 T. Andrew Taton, in Current Protocols in Nucleic Acid Chemistry (John Wiley & Sons, Inc.,
2001). 23 Sigma-Aldrich, (Sigma-Aldrich, 2011). 24 Linette M. Demers, Chad A. Mirkin, Robert C. Mucic, Robert A. Reynolds, Robert L. Letsinger,
Robert Elghanian, and Garimella Viswanadham, Anal. Chemistry 72 (22), 5535 (2000). 25 Douglas H. Everett, Basic Principles of Colloid Science. (The Royal Society of Chemistry,
1988). 26 B.V. Derjaguin, L.D. Landau, E. J. W. Verwey, and J. T. G. Overbeek, Acta Physiochim. 14
(633) (1941).
32
27 Zetasizer Nano User Manual. (Malvern, Malvern Instruments Ltd., Worcestershire, United Kingdom, 2009), p.302.
28 Bruce J. Berne, Dynamic light scattering : with applications to chemistry, biology, and physics / Bruce J. Berne and Robert Pecora. (Wiley, New York :, 1976).
29 Masuo Hosokawa, Kiyoshi Nogi, Makio Naito, and Toyokazu Yokoyama, (Elsevier Science, 2007), pp. 372.
30 (Thermo Fischer Scientific, Wilmington, U.S.A, 2010). 31 (Ocean optics, 2012). 32 K. Y. Yasumura, T. D. Stowe, E. M. Chow, T. Pfafman, T. W. Kenny, B. C. Stipe, and D. Rugar,
Microelectromechanical Systems, Journal of 9 (1), 117 (2000). 33 John Elie Sader, Journal of Applied Physics 84 (1), 64 (1998). 34 Cornelis A. Van Eysden and John E. Sader, Journal of Applied Physics 106 (9), 094904 (2009). 35 Jason Jensen and Martin Hegner, Journal of Sensors 2012 (2011). 36 Thomas Braun, Viola Barwich, Murali Krishna Ghatkesar, Adriaan H. Bredekamp, Christoph
Gerber, Martin Hegner, and Hans Peter Lang, Physical Review E 72 (3), 031907 (2005). 37 Jean Aigueperse, Paul Mollard, Didier Devilliers, Marius Chemla, Robert Faron, René
Romano, and Jean Pierre Cuer, in Ullmann's Encyclopedia of Industrial Chemistry (Wiley-VCH Verlag GmbH & Co. KGaA, 2000).
38 Andrey V. Tataurov, Yong You, and Richard Owczarzy, Biophysical Chemistry 133 (1-3), 66 (2008).
39 Thomas Braun, Murali Krishna Ghatkesar, Natalija Backmann, Wilfried Grange, Pascale Boulanger, Lucienne Letellier, Hans-Peter Lang, Alex Bietsch, Christoph Gerber, and Martin Hegner, Nat Nano 4 (3), 179 (2009).
40 G. Sosale, K. Das, L. Frechette, and S. Vengallatore, J. Micromech. Microeng. 21 (10) (2011).