Author: Gasparyan, Ferdinand
Date published: November 1, 2010
The functionalization of electrolyte-gate field-effect devices (FED), like ion-sensitive field-effect transistors (FET), capacitive electrolyte-insulator-semiconductor (EIS) structures and light-addressable potentiometric sensors (LAPS), with the nano- and biomaterials is one of the most attractive approaches for the development of (bio-) chemical sensors and biochips. Since FEDs are charge sensitive devices, each (bio-) chemical reaction leading to chemical or electrical changes at the gate insulator/electrolyte interface can be detected by coupling the gate with respective chemical or biological recognition elements. These devices were shown to be versatile tools for detecting pH, ion concentrations, enzymatic reactions, charged macromolecules (DNA-deoxyribonucleic acid, proteins, polyelectrolytes), cellular metabolism and action potentials of living cells, as well as for biocomputing logic gates (see e.g., [1-4]). More recently, nanoparticles, silicon nanowires (NAV) and carbon nanotubes (NT) have attracted significant interest as promising materials for novel nanoscale bioelectronic devices and gas sensors [5-12], due to their excellent mechanical, electronic and chemical properties, leading to an enhanced sensor performance. Single-walled carbon nanotubes (SWCNT) employed in a FET layout can be used as sensors for molecular adsorption. ChemFETs based on Si NWs, CNTs, and other nanomaterials have already proved to be very useful in detection of macromolecular complexes such as proteins, biomolecules and toxic gases [13-18]. Network behavior in SWNTs is examined by polymer electrolyte gating. ChemFETs enable direct, label-free, real-time, and continuous sensing of ionic or dipolar molecules, and are suitable for big molecular complexes. These nanosensors can also be useful in nanomedicine, which refers to highly specific medical inventions at the molecular scale, enabling detection of proteins and other biomolecules released by cancer cells. This is crucial from therapeutic point of view to diagnose cancer at its early-stage of development [19, 2O].
One potential approach to minimize undesirable NT-substrate interactions was used to employ an electrolyte gate. Liquid gating of the NT devices was first demonstrated by Kruger et al.  with multi-walled CNTs and later by Rosenblatt et al.  with single-walled CNT. Since solvent molecules are in intimate contact with NTs, the influence of the substrate on the NT is expected to be reduced. In  fabricated high performance FETs made from semiconducting SWNTs. Using chemical vapor deposition to grow the tubes, annealing to improve the contacts, and an electrolyte as a gate, obtain very high device mobilities and transconductances. These measurements demonstrate that SWNTs are attractive for both electronic applications and for chemical and biological sensing.
The invention  relates to sensors which are used for detecting analytes, in a gaseous or liquid environment, and which use, as transducer, one or more elements of the semiconductor NT or NW types. It applies to the detection of particular species present in a liquid medium (biochemical detection), etc. The analytes detected may be particular chemical or biological species, proteins, enzymes or other species, the harmfulness (virus, toxic gas, etc.) of which is generally desired to detected. In the field of the detection of biochemical or bacteriological species, functionalization of the NTs is fundamental for protein molecule recognition in order to overcome the nonselective binding already mentioned, which has the effect that protein molecules tend to become attached to the walls of the NT because of their hydrophobicity. This phenomenon definitely changes the response of the transistor. The same transistor, if it is not functionalized, cannot therefore be reused several times for detecting proteins. In addition, a reproducible result cannot be obtained in the case of a mixture of several molecules. Functionalization therefore has the function of acting as an inhibitor for the reaction of certain proteins with the NT, in order to allow the reaction of other proteins which, by interacting with the NT, change the characteristics of the transistor, and therefore enable them to be detected. Ion channels and nanopores in the plasma membrane of living organisms are excellent electrical nanodevices, commonly used as biosensors sensitive to a wide range of chemical and electrical stimuli.
In  reported CNT FETs that function as selective detectors of DNA immobilization and hybridization, presented recent innovations in CNT-assisted biosensing technologies. CNT FETs with immobilized synthetic oligonucleotides have been shown to specifically recognize target DNA sequences, including H63D single-nucleotide polymorphism discrimination in the HFE gene, responsible for hereditary hemochromatosis. The electronic responses of CNT FETs upon singlestranded DNA immobilization and subsequent DNA hybridization events were confirmed by using fluorescence-labeled oligonucleotides and then were further explored for label-free DNA detection at picomolar to micromolar concentrations . It is also observed a strong effect of DNA counterions on the electronic response, thus suggesting a charge-based mechanism of DNA detection using CNT FET devices. Implementation of label-free electronic detection assays using CNT FETs constitutes an important step toward low-cost, low-complexity, highly sensitive and accurate molecular diagnostics.
The physical properties limiting sensor devices in planar semiconductors can be readily overcome by exploiting nanoscale devices. Binding to the surface of a NT or NW can lead to depletion or accumulation of carriers in the bulk of the nanometer diameter structure and increasing sensitivity to the point that single molecule detection.
Noise investigations in functionalized FEDs are especially important in the case of an application of FEDs for the measurement of low analyte concentrations, for the detection of biomolecules by their intrinsic molecular charge, as well as for the monitoring of action potential of living cells, where the sensor output signal can be very small. Moreover, development of noise spectroscopy in functionalized FEDs can give additional insight into the detection mechanisms of the biosensors. On the other hand, noise measurements could be as very sensitive method for the analysis of a semiconductor/other media (metal, insulator, semiconductor, electrolyte, gaseous) interface quality as well as for the quantitative determination of an analyte concentration as it has been shown for H2 gas sensors in [26, 27].
The use of an electrolyte as a gate that is in direct contact with the SWNT allows the gating efficiency to approach the theoretical upper limit (60 mV/decade), indicating optimal sensitivity to the electrostatic environment. The combination of operation in aqueous solution, high sensitivity, and the unique ID geometry with a critical dimension in the size range of individual biomolecules makes CNTs outstanding biosensors. A next challenge in the use of NTs as biosensors is to push the sensitivity down to the lowest analyte concentrations [28, 29] and ultimately down to single molecule sensitivity. In this context, optimizing the signal-to-noise ratio (SNR) is a crucial aspect, which has been given little attention to date. Low frequency (LF) noise of CNT transistors, which exhibits a 1/ftype spectrum  reveals a strong gate dependence [31, 32] consistent with a recently proposed charge noise model [31, 33]. The well-established high sensitivity of SWNT FETs as bimolecular detectors demonstrates that various fluctuating entities in the environment lead to a high level of noise in these devices.
2. Excess Noises in Nanoscale Sensors
Experiments on various types of nanochannels and nanopores, both biological and artificial, showed random fluctuations of the channel conductivity. The fluctuations result from random transitions between different conductivity states of the channel. It is now widely accepted that LF noise is produced by fluctuations in sample conductance. In general, if diffusive transport is assumed, the conductance σ is proportional to both carrier mobility µ and carrier concentration n :
σ = e µn.
In principle, both n and µ can contribute to the 1/f-noise. Numerous LF noise models have been proposed for different types of materials [48, 49] and they usually divide into two general categories, namely, number (Δn) and mobility (Δµ) fluctuations models. Besides, a combination and modification of both models applicable for many types of conventional FET and other semiconductor devices was proposed (see e.g. [31, 50, 51]. The stochastic analysis of the process can provide information on dynamics of the channel and phenomena occurring when a molecule is transported through the pore. Potentially, the stochastic characteristics of the channel can be utilized as a tool for biosensor design. Characteristic quantities such as noise amplitude and corner frequency dividend white noise segment from Lorentzian spectrum depend on transport characteristics and reveal details on interactions between pore walls and transported molecule. Such characteristics can be observed when the pore has only two conductivity states and transition between them is a random process without any memory on the channel history. There are indications that the stochastic process of the pore fluctuations may provide information about the chemical environment of the pore .
Sensitivity of the noise level can be expressed by means of the Hooge parameter A . To characterize the absolute amplitude of the excess noise the current noise spectral density (NSD) S1 expressed as
The exponent ß could be considered as another informative quantity. It has been reported for bacterial porin that higher concentration of diffusing molecules can be associated with larger NSD amplitude and, therefore, the Hooge parameter. Note mat the Hooge parameter A very sensitive to ionic strength in electro-nanopores .
In  the mechanism responsible for the LF noise in liquid-gated SWNT-FETs and its scaling with the length of the NT channel down to the nanometer scale was investigated. It is shown that the gate dependence of the noise level provides strong evidence for charge-noise model. It is find that NSD of the charge noise scales as the inverse of the channel length of the SWNT-FET. Measurements show that the ionic strength of the surrounding electrolyte has a minimal effect on the noise magnitude in SWNT-FETs. The noise spectra of the source-drain current / follows according Eq.(l) with ß = 2 .
LF noise measurements on individual SWNT transistors exhibiting ambipolar characteristics. With a polymer electrolyte as gate medium, LF noise can be monitored in both p- and ?-channel operation of the same NT under the same chemical environment. 1/f-noise in the p-channel of polymer electrolyte gated NT transistor is similar to that of back gate operation. Devices exhibit significantly larger noise amplitude in the ?-channel operation that has a distinct dependence on the threshold voltage. A nonuniform energy distribution of carrier trapping/scattering sites is considered to explain this behavior . In  reported the observation of 1/f-noise in CNTFETs as a function of gate potential. In order to avoid additional LF noise contribution from SÌO2-NT interactions, the measurements were carried out in conducting liquid. In  demonstrated the drastic improvement of sensitivity (or SNR) in CNTFET sensors. The a.c. measurement with a lock-in amplifier was adopted to suppress the fluctuations in drain current of CNTFETs without attenuating the signal level. The noise level of CNTFETs incubated in phosphate buffer solution was highly suppressed by the a.c. measurement. Authors also investigated the sensing operations of CNTFET pH sensors and biosensors. Sensing performances in CNTFET sensors were dramatically improved. The SNR of pH sensors measured by a.c. was six times higher than that by d.c. measurement. A small amount of bovine serum albumin of 250 pM was effectively detected by CNTFET biosensors using a.c. measurement . In  show that the relative 1/f-noise amplitude of CNT films may depend strongly on device dimensions and on the film resistivity, following a power-law relationship with resistivity near the percolation threshold after properly removing the effect of device dimensions.
The LF noise of liquid-gated SWNT-transistors reveals a 1/f-type spectrum, where the noise power is inversely proportional to the length, L , of the SWNT [31, 32, 38].
As shown by Tersoff  the LF noise is well described by an augmented charge-noise model, which predicts that
Here the first term describes the effect of fluctuating charges in the vicinity of the SWNT that induce current noise through a field-effect. For this charge-noise component
where S^sup input^ is a proportionality constant, I^sub sd^ is the source-drain current, V^sub ig^ is the gate voltage . The second term in Eq.(2) describes additional gate-independent current-noise, which becomes apparent in ON-state and is modeled as a series resistor that exhibits LF noise . Here A8 represents the resistance-noise amplitude of this series resistor, and R^sub S^ and R^sub tot^ are the series resistance and total device resistance, respectively. The validity of this charge-noise model for liquid-gated SWNT-devices was confirmed by Mannik et al. . The augmented charge-noise model is in good agreement with the data for both bare and contact protected device (polymethil methacrilate device).
The SNR for real-time biosensing with liquid-gated GSiT transistors is crucial for exploring the limits of their sensitivity [4O]. Although biosensing is often performed at high transconductance where the device displays the largest gate response. In  show that the maximum SNR is actually obtained when the device is operated in the subthreshold regime. In the ON-state, additional contributions to the noise can lead to a reduction of the SNR by up to a factor of 5. For devices with passivated contact regions, the SNR in ON-state is even further reduced than for bare devices. It is show that when the conductivity of the contact regions can be increased using a conventional back gate, the SNR in the ON-state can be improved. These results demonstrate that biosensing experiments are best performed in the subthreshold regime for optimal SNR. Investigated SWNT transistors were fabricated on thermally oxidized Si wafers and shown that the electrostatic gating effect of biomolecules is unaffected by contact protection, and that the real-time signal is highly dependent on the liquid gate [4O]. Interestingly, the sensor response can be enhanced by proper device precoating. The noise parameters, A and ? of a device with 10 µ?? separation between source and drain were measured in buffers with different ionic strengths and pH values. To compare this results with data of [30,41] in Fig. l(a), one normalized 1/f magnitude A by the device resistance at the corresponding gate voltage and plotted empirical estimates from  ^ = HT11A and A = W11R/ E'3 , where R is the device's resistance and L is the separation between source and drain in microns. It should be noticed that device to device variation of A/ R in both of these expressions is approximately one order of magnitude, thus data are consistent with other observations. Not shown in the figure, parameter / turns out to be 1.05±0.07 and essentially does not depend on the gate voltage. As it is seen from Fig. 1, varying the electrolyte concentration from 50 to 500 mM while keeping the pH value constant leads to some 10-30 mV shift of the device threshold voltage to the left, which may be attributed to changes of buffer capacitance or additional electron doping of the NT when the buffer concentration is increased .
The normalized noise amplitudes for all concentrations show similar gate voltage dependence suggesting that ionic strength of the buffer does not noticeably affect noise characteristics of the device. The source of 1/f-noise can be diagnosed by looking at the dependence of the normalized noise power of the drain current on the gate potential. As it concluded by Ghibaudo and Boutchacha  if Δµ fluctuations have the main contribution in 1/f-noise, then the noise parameter A should be inversely proportional to the drain current, i.e., A/ R is independent of the gate voltage, which is in contradiction with the data in Fig. l(a).
On the other hand, if the number fluctuations is more dominant in the appearance of the LF noise, then A is proportional to g^sup 2^^sub m^/I^sup 2^ , where g^sub m^ = ∂I/∂V^sub g^ is the device transconductance . The latter case was observed for most devices, which suggests that An fluctuations make the major contribution to the 1/f-noise parameter A . To get the first order approximation for the number fluctuations per unit area of the NT network, in  simplify the analysis presented in , and make the assumption that Δn fluctuations are the only factor producing 1/f-noise. In this case, the fluctuations of the number of carriers lead to the equivalent fluctuations of gate potential
with N being the number of carriers, and C being the total NT surface capacitance . Also, the fluctuation of source-drain current can be represented as
where S^sub N^ is the power fluctuation per hertz of the total number of carriers.
In principle, for uniform networks, the slope in Fig. 2(a) should be inversely proportional to the total number of carbon atoms, 10, 25 so to BL^sup 2^/R , where B is the slope, A is the sample resistance in air at zero gate voltage, and L is the channel length, should be independent of device geometry at fixed electrolyte concentration and pH. This is demonstrated by Fig. 2(b) with an average value for all devices equal to approximately 6x10^sup -23^ V^sup 2^m^sup 2^/n . In this case, Eq. (6) can be rewritten as
where C [asymptotically =]16µF/cm^sup 2^ is the total capacitance per unit area  so #χ [asymptotically =]6x10^sup 13^ Q'W2 for 150 mM PBS (phosphate buffered saline) at pH 7.4 and 7=25 °C. The parameter a characterizes the number fluctuations per unit area in the NT network and depends on the density of the NT network on the substrate. Since the device resistance in NT networks is often dominated by the contact and intertube resistances, the parameter a should also on such network properties as lengths of NTs, NT diameter distribution, metal type used for forming source-drain leads, etc. In  note that it is more probable that 1/f-noise is produced by thermally activated trapping or diffusive processes that have a uniform distribution of energies within a broad range . Each process with a certain characteristic energy forms Lorentzian noise spectra, and their assembly appears as 1/f-noise.
In order to quantify the gate dependence of parameter A , in  estimate the total number of carriers as
where C is the total gate capacitance, V^sub g^ is the applied gate bias, V^sub th^ is the threshold voltage, L is the channel length. For polymer electrolyte gating, the quantum capacitance of the SWNT (C^sub Q^ 10^sup -10^ F/m) dominates C whereas the geometric capacitance (estimated as C^sub bg^ « 2pe^sub 0^e/ln(2t/a) - 10^sup ~11^ F/m, where e^sub 0^ is the permittivity of free space, e and t are the dielectric constant and the thickness of the SiO^sub 2^ gate dielectric, respectively, and d is the diameter of the SWNT) dominates in the back-gate operation.
With these values for the capacitances, the inverse noise amplitude is fitted with equation
resulting in α^sub H^ - 0.009 for p-channel polymer electrolyte gating and 0.01 for back-gating. Seven other examined devices exhibit α^sub H^ values of 0.006, 0.01, 0.011, 0.013, 0.014, 0.021, and 0.18 for p-channel operation under PEG-based polymer electrolyte gate. The same device with the same NT operating with exactly the same polymer electrolyte gate medium exhibits 2 orders of magnitude larger α^sub H^ value for the ?-channel ( α^sub H^ =0.53) than the p-channel (α^sub H^ = 0.006). Furthermore, there appears to be a distinct trend in the distribution of α^sub H^ with respect to V^sub th^ (i.e., a maximum near KA ~ 0.2 V and decreasing away from this value). Since the SWNTs are in an electrochemical environment, it is necessary to consider potential contributions from the surrounding electrolyte solution in explaining the observed trend. The fact that 1/f-noise in the p-channel shows negligible difference between electrolyte gate and back-gate operations suggest that electrochemical environment may not be an important factor here. However, there is still a possibility that the electrochemical environment selectively affects the 1/f noise in the ?-channel. Such a situation might arise if there were redox processes at positive gate voltages near the ?-channel threshold voltage or if the anions and the cations had different affinities for the SWNT. In  consider an energy distribution of traps/scattering sites (e.g., in the oxide substrate immediately surrounding the SWNT or strongly adsorbed species directly on SWNT). If the distribution of trapping/scattering sites were uniform, one can expect aH to be constant with V^sub th^ . It is considered that the number of sites decays exponentially from a maximum at some fixed potential (V^sub t/s^ ). Then α^sub H^ can be expressed as
Here V^sub 0^ is the spread or the distribution parameter of the exponential. The choice of an exponential distribution is from  and may be somewhat arbitrary, but the key aspect was looking lies in Vtis (i.e., where the majority of the trapping/scattering sites are energetically, not in the details of the functional form of the spread of such sites). Most p-channel threshold voltages are sufficiently far away from V^sub t/s^ of -0.2 V such that aH is nearly constant. One device that exhibits an anomalously large α^sub H^ value of 0.18 is due to the p-channel threshold voltage being closer to V^sub t/s^. Since the threshold voltage is directly related to the relative band edge positions, this means that the valence band edge of this SWNT lies closer to V^sub t/s^ and therefore p-channel operation of this device leads to significantly larger noise levels than other devices. The ?-channel threshold voltages of all devices examined here with PEG-based electrolyte gate, on the other hand, lie very close to V^sub t/s^ unlike most p-channel threshold voltages. The conduction band edge lying close to V^sub t/s^ can then explain the relatively large 1/f-noise. In  note that when there is an energy distribution of traps/scattering sites, 1/f-noise can begin to deviate from the 1/f dependence. That is, ? in can deviate from 1 (γ [arrow right] 1 as V^sub 0^ [arrow right] ∞ , i.e. a uniform energy distribution of trapping/scattering sites). The n-channel operation of SWNT devices with relatively large aH values indeed exhibit this deviation with ? as small as -0.8.
Transfer characteristics of a PEG-based polymer electrolyte gated SWNT transistor, normalized current NSD, gate dependence of /¿ and comparison of inverse noise amplitude of the same SWNT operated with back gate and with polymer electrolyte gate are presented in Fig. 4. Fig. 4c shows the strong correlation between the gate dependence of the noise power S1 and that of /¿ operating with polymer electrolyte gate.
Low frequency noise in a capacitive field-effect Al-P-Si-SiO2-Ta2Os EIS structure functionalized with a polyaminoamide (PAMAM)/SWNT multilayer has been investigated and compared with the noise in a bare EIS device [1 1-12]. The noise spectral density exhibits 1/f dependence with the power factor of γ » 0.8 and γ = 0.8 - 1.8 for the bare and functionalized EIS sensor, respectively. The gate-voltage noise spectral density was practically independent on pH value of the solution, and is increased with increasing the gate voltage or gate-leakage current. It has been observed that the existence of the PAMAM/SWNTs multilayer leads to an essential reduction of 1/f noise. The gate-current noise behavior in bare and functionalized EIS devices was modeled using the flatband-voltage fluctuations concept. The experimentally observed gate-voltage dependence of the noise in capacitive EIS structures can be explained by the gate-voltage-dependent changes in the occupancy of the oxide trap levels resulting in a modulation of the conductivity of current paths or charge carriers passing through the EIS structure. The detailed mechanism of current transport in field-effect capacitive electrolyte(SWNTOAMAM)-Ta^sub 2^O^sub 5^-SiO^sub 2^-Si-Al and electrolyte-Ta^sub 2^O^sub 5^- S K)2-Si- Al structures as well as the microscopic origin of the noise-reduction effect requires further investigations.
Generally, noise determined by the modulation of semiconductor depletion region capacitance and surface potential due to charge fluctuation at the insulator/electrolyte interface. Modified charge fluctuation noise model successfully used for explanation of noise peculiarities of pSi/SiO^sub 2^/Ta^sub 2^Os/electrolyte and p-Si/SiO^sub 2^/Ta^sub 2^O^sub 5^/dendrimer/CNT/electrolyte (bio-) chemical sensors [11, 12, 47]. For noise spectral density received following expression
Here w and / are the gate width and length, N01 is the equivalent density of traps per unit area at the SiO^sub 2^/electrolyte interface. It should be emphasized, that an increase of the 1/f-noise level with decreasing gate area.
Thus, on the base of the analyses of electrical, physical, chemical and other characteristics of the nanoscale sensors based on the NTs and NWs we can have some conclusions:
* NTs and NWs are promising candidates for advanced nanoelectronic devices, and they have great potential in a wide range of applications, such as FETs, elementary logic circuits, bio- and chemical sensors, nanotechnology, biotechnology, electronics, memory devices, optics and other fields of materials science and architectural fields.
* Mainly LF noise spectral density expressed as
* Parameter ? generally reflects the sample quality and increases with decreasing device size and depends on many parameters of material, its structure, sizes, NTs bulk and surface physical and chemical conditions, from its fabrication method. The noise amplitude
R is the device resistance, N is the number of atoms or carriers in the system, L is the sample length (N ∞ L). A = 1.0 x 10^sup -11^R. Parameter A vary within 10^sup ~13^ up to 4 x 10^sup -4^.
* The size scaling is incorporated in Hooge's empirical law
* In the linear regime, 1/ A ∝ |V^sub g^-V^sub th^| if noise is due to mobility fluctuations and 1/ A ∝ |V^sub g^-V^sub th^|^sup 2^ oc noise is due to number fluctuations.
* Parameter β = Q is expected for pure resistance fluctuation in ohmic conductors. The γ ≠ 1 behavior is associated with nonlinear characteristics. Usually \β\ « 1, ?/| « 1 are material dependent numbers (γ = 1 + Δγ ).
* Excess noise with a slope different from unity (γ ≠ 1) can be explained by a superposition of a few Lorentzians. The change in slope with respect to temperature is thus explained by the variation of trap activities.
. C. -S. Lee, S. K. Kim, and M. Kim, Ion-sensitive field-effect transistors for biological sensing, Sensors, Vol. 9, 2009, pp. 7111-7131.
.M. J. Schöning and A. Poghossian, BioFEDs (field-effect devices): State-of-the-art and new directions, Electroanalysis,Vol 18, 2006, pp. 1893-1900.
. P. Fromherz, Three levels of nanoelectronic interfacing: Silicon chips with ion channels, nerve cells, and brain tissue, Ann, NYAcad, ScL, Vol. 1093, 2006, pp. 143-160.
. M. Krämer, M. Pita, J. Zhou, M. Ornatska, A. Poghossian, M. J. Schöning, and E. Katz, Coupling of biocomputing systems with electronic chips: electronic interface for transduction of biochemical information, J. Phys. Chem. C, Vol. 113, 2009, pp. 2573-2579.
. E. Katz, I. Willner, and J. Wang, Electroanalytical and bioanalytical systems based on metal and semiconductor nanoparticles, Electroanalysis, Vol. 16, 2004, pp. 19-44.
. Special issue on nanosensors for defense and security, IEEE Sensors Journal, Vol. 8, No. 6, 2008.
. J. Gun, M. J. Schöning, M. H. Abouzar, A. Poghossian, and E. Katz, Field-effect nanoparticle-based glucose sensor on a chip: Amplification effect of co-immobilised redox species, Electroanalysis, Vol. 20, 2008, pp. 1748-1753.
. F. Patolsky, G. Zheng, and C. M. Lieber, Nanowire-based biosensors, Anal. Chem., Vol. 78, 2006, pp. 4260-4269.
. B. P. Timko, T. Cohen-Karni, G. Yu, Q. Qing, B. Tian, and C. M. Lieber, Electrical recording from heart with flexible nanowire device arrays, Nano Lett., Vol. 9, 2009, pp. 914-918.
.K. Balasubramanian, and M. Burghard, Biosensors based on carbon nanotubes, Anal. Bioanal. Chem., Vol. 385, 2006, pp. 452-468.
.F. V. Gasparyan, A. Poghossian, S. A. Vitusevich, M. V. Petrychuk, V. A. Sydoruk, A. V. Surmalyan, J. R. Siqueira Jr., O. N. Oliveira Jr., A. Offenhäusser, M. J. Schöning, Low frequency noise in electrolytegate field-effect devices functionalized with dendrimer/carbon-nanotube multilayers, in Proc. of the 2Cf Int Conf. Fluctuation and Noise, 14-19 June, Pisa, Italy. 2009, pp. 133-136.
.F. V. Gasparyan, A. Poghossian, S. A. Vitusevich, M. V. Petrychuk, V. A. Sydoruk, J. R. Siqueira Jr., O. N. Oliveira Jr., A. Offenhäusser, M. J. Schöning, IEEE Sensors Journal, 2010.
. Yi Cui, Q. Wei, H. Park, C. M. Lieber, Nanowire nanosensors for highly sensitive and selective detection of biological and chemical species, Science, Vol. 293, 2001, pp. 1289-1292.
.R. J. Chen, S. Bangsaruntip, K. A. Drouvalakis, N. W. S. Kam, M. Shim, Y. Li, W. Kirn, P. J. Utz, and H. Dai, Monovalent functionalization of carbon nanotubes for highly specific electronic biosensors, Appi. Phys. Sci., Vol. 100, No. 9, 2003, pp. 4984-4989.
.E. Souteyrand, J. P. Cloarec, J. R. Martin, C. Wilson, I. Lawrence, S. Mikkelsen, and M. F. Lawrence, Direct Detection of the Hybridization of Synthetic Homo-Oligomer DNA Sequences by Field Effect, J. Phys. Chem. B, Vol. 101, 1997, pp. 2980-2985.
. S. Park, T. A. Taton, C. A. Mirkin, Array-Based Electrical Detection of DNA with Nanoparticle Probes, Science, Vol. 295, 2002, pp. 1503-1506.
. A. Star, J.-C. Gabriel, K. Bradley, and G. Grüner, Electronic Detection of Specific Protein Binding Using Nanotube FET Devices, Nano Lett., Vol. 3, 2003, pp. 459-463.
.T. Tang, X. Liu, C. Li, B. Lei, D. Zhang, M. Rouhanizadeh, T. Hsiai, and C. Zhou, Complementary response of In2O3 nanowires and carbon nanotubes to low-density lipoprotein chemical gating, Appi. Phys. Lett., Vol. 86, art. 103903, 2005.
. G. Zheng, F. Patolsky, Y. Cui, W. U. Wang, and C. M. Lieber, Multiplexed electrical detection of cancer arkers with nanowire sensor arrays, Nature Biotechnology, Vol. 23, 2005, pp. 1294-1301.
A. H. Ting, K. M. McGarvey, and S. B. Baylin, The cancer epigenome-components and functional correlates, Genes Dev., Vol. 20, 2006, pp. 3215-3231.
.M. Kruger, M. R. Buitelaar, T. Nassbaumer, and C. Schonenberger, Electrochemical carbon nanotube fieldeffect transistor, Appi. Phys. Lett., Vol. 78, No. 9, 2001, pp. 1291-1293.
. S. Rossenblatt, Y. Yaish, J. Park, J. Gore, V. Sazonova and P. L. McEuen, High performance electrolyte gated carbon nanotube transistors, Nano Lett., Vol. 2, No. 8, 2002, pp. 869-872.
.P. Bondavalli, P. Legagneux, P. L. Barny, D. Pribat, J. Nagle, Array of Fet Transistors Having a Nanotube or Nanowire Semiconductor Element and Corresponding Electronic Device, For the Detection of Analytes, Patent US, IPC8 Class: AH01L2978FI USPC Class: 257253.
. S. N. Kim, J. F. Rusling, and F. Papadimitrakopoulos, Carbon nanotubes for electronic and electrochemical detection of biomolecules, Adv. Mater Deerßeld, Vol. 19, 20, 2008, pp. 3214-3228.
.A. Star, E. Tu, J. Niemann, J.-C. P. Gabriel, C. S. Joiner, and C. Valcke, Label-free detection of DNA hybridization using carbon nanotube network field-effect transistors, PNAS, Biophysics, Vol. 103, No. 4, 2006, pp. 921-926.
 .V. M. Aroutiounian, Z. H. Mkhitaryan, A. A. Shatveryan, F. V. Gasparyan, M. Zh. Ghulinyan, L. Pavesi, L. B. Kish, and C. -G. Granquist, Noise spectroscopy of gas sensors, IEEE Sensors J., Vol. 8, 2008, pp. 786-790.
.Z. Mkhitaryan, F. Gasparyan, and A. Surmalyan, Low frequency noises of hydrogen sensors, Sensors & Transducers, Vol. 104, 2009, pp. 58-67.
.P. E. Sheehan, L. J. Whitman, Detection limits for nanoscale biosensors, Nano Lett., Vol. 5, 2005, pp. 803-807.
. T. M. Squires, R. J. Messinger, S. R. Manalis, Making it stick: convection, reaction and diffusion in surface-based biosensors, Nat. Biotechnol., Vol. 26, 2008, pp. 417-426.
.P. G. Collins, M. S. Fuhrer, and A. Zettl, 1/f noise in carbon nanotubes, Appi. Phys. Lett., Vol. 76, No. 7, 2000, pp. 894-896.
J. Mannik, I. Heller, A. M. Janssens, Serge G. Lemay, and C. Dekker, Charge noise in liquid-gated singlewall carbon nanotube transistors, Nano Lett., Vol. 8, No. 2, 2008, pp. 685-688.
.M. Briman, K. Bradley, G. Grüner, Source of 1/f noise in carbon nanotube devices, J. of Appi. Phys., Vol. 100, 2006, pp. 013505 (1-5).
.J. Tersoff, Low-frequency noise in nanoscale ballistic transistors, Nano Lett., Vol. 7, No. 1, 2007, pp. 194-198.
.M. Kotulska, l/fa noise in nanopores and its potential application in biosensors, Elektronika, Vol. 8, No. 9, 2004, pp. 130-132.
.J. H. Back, S. Kim, S. Mohammadi, and M. Shim, Low-frequency noise in ambipolar carbon nanotube transistors, Nano Lett., Vol. 8, No. 4, 2008, pp. 1090-1094.
 .Y. Yamamoto, Y. Ohno, K. Maehashi, and K. Matsumoto, Carbon nanotube field-effect transistor biosensors with high signal-to-noise ratio using alternating current measurement, Nanotechnology and Applications (NANA 2008).
 .A. Behnam, A. Ural, ad G. Bosnian, Modeling and measurements of low frequency noise in single-walled carbon nanotube films with bulk and percolation configurations, in Proc. of the 2(f* Int. Conf on Noise and Fluctuations, Pisa, Italy, 14-19 June, 2009, pp. 79-84.
.Y.-M. Lin, J. Appenzeller, J. Knoch, Z. Chen, and P. Avouris, Low-frequency current fluctuations in individual semiconducting single-wall carbon nanotubes, Nano Lett., Vol. 6, No. 5, 2006, pp. 930-936.
.Z. Wang and X. Zu, Atomistic simulation of the size and orientation dependencies of thermal conductivity in GaN nanowires, Appi. Phys. Lett., Vol. 90, 2007, pp. 161923 (1-3).
. I. Heller, J. Mannik, S. G. Lemay, and C. Dekker, Optimizing the signal-to-noise ratio for biosensing with carbon nanotube transistors, Nano Lett., Vol. 9, No. 1, 2009, pp. 377-382.
.E. S. Snow, J. P. Novak, M. D. Lay, and F. K. Perkins, 1/f noise in single-walled carbon nanotube devices. Appl. Phys. Lett., Vol. 85, No. 18, 2004, pp. 4172-4174.
.M. Kruger, M. R. Buitelaar, T. Nassbaumer, and C. Schonenberger, Electrochemical carbon nanotube fieldeffect transistor, Appi. Phys. Lett., Vol. 78, No. 9, 2001, pp. 1291-1293.
. G. Ghibaudo and T. Boutchacha, Rical noise and RTS fluctuations in advanced CMOS devices. Microelectron. Reliab., Vol. 42, 2002, pp. 573-582.
 .K. Bradley, J. Gabriel, M. Briman, A. Star and G. Grüner, Charge Transfer from Ammonia Physisorbed on Nanotubes, Phys. Rev. Lett., Vol. 91, 2003, pp. 218301-218304.
.J. H. Back, S. Kim, S. Mohammadi, and M. Shim, Low-frequency noise in ambipolar carbon nanotube transistors. Nano Lett., Vol. 8, No. 4, 2008, pp. 1090-1094.
.J. I. Lee, J. Brini, A. Chovet, C. A. Dimitriadis, On 1/f noise in semiconductor devices, Solid-State Electron., Vol. 43, 1999, pp. 2181-2183.
.F. V. Gasparyan, Z. H. Mkhitaryan, A. V. Surmalyan, Basis structure of (bio-)chemical sensors: Comparative analysis of CVC and noises, in Proc. of the 7th Int. Conf. Semicond. Micro. & Nanoelectronics, Tsakhcadzor, Armenia, July 3-5, 2009, pp. 101-104.
.F. N. Hooge, T. G. M. Kleinpenning, L. K. J. Vandamme, Experimental studies on 1/f noise, Rep. Prog. Phys., Vol. 44, 1981, pp. 479-532.
.F. N. Hooge, 1/f noise sources, IEEE Trans. Electron Devices, Vol. 41, 1994, pp. 1926-1935.
.K. Hung, P. Ko, C. Hu, and Y. Cheng, A unified model for the flicker noise in metal-oxide-semiconductor field-effect transistors, IEEE Trans. Electron Devices, Vol. 37, 1990, pp. 654-665.
J1. V. Gasparyan, H. V. Asriyan, S. V. Melkonyan C. E. Comían, Method of 1/f noise reduction and Noise Level Manipulation in semiconductor based devices, May 7, 2010, as U. S. Patent Application for Letters Patent of the United States, No. 61/332, 408.
Department of Physics of Semiconductors & Microelectronics
Yerevan State University,
I Alex Manoogian St., 0025 Yerevan, Armenia
Received: 16 July 2010 /Accepted: 18 November 2010 /Published: 30 November 2010