key: cord-0036059-abxhslwe authors: Olsen, Eric; Vainrub, Arnold; Vodyanoy, Vitaly title: Acoustic Wave (TSM) Biosensors: Weighing Bacteria date: 2008 journal: Principles of Bacterial Detection: Biosensors, Recognition Receptors and Microsystems DOI: 10.1007/978-0-387-75113-9_12 sha: 4f1d320c87c2eb5693b9b7ffa34aabfa4fc6c272 doc_id: 36059 cord_uid: abxhslwe This chapter is focused on the development and use of acoustic wave biosensor platforms for the detection of bacteria, specifically those based on the thickness shear mode (TSM) resonator. We demonstrated the mechanical and electrical implications of bacterial positioning at the solid-liquid interface of a TSM biosensor and presented a model of the TSM with bacteria attached operating as coupled oscillators. The experiments and model provide an understanding of the nature of the signals produced by acoustic wave devices when they are used for testing bacteria. The paradox of “negative mass” could be a real threat to the interpretation of experimental results related to the detection of bacteria. The knowledge of the true nature of “negative mass” linked to the strength of bacteria attachment will contribute significantly to our understanding of the results of “weighing bacteria.” The results of this work can be used for bacterial detection and control of processes of bacterial settlement, bacterial colonization, biofilm formation, and bacterial infection in which bacterial attachment plays a role. Rapid, specific, sensitive detection and enumeration methods for microbial pathogens have long been a subject of research. This is especially true in the area of food-related illness prevention, where it's estimated that over three million deaths occur annually worldwide at a cost of $6.5-34.9 billion (Buzby and Roberts 1997) due to the consumption of food products contaminated with bacteria, bacterial toxins, or viruses (Foodborne Diseases 1997). The perceived need for instantaneous detection of pathogenic biological agents in both simple and complex matrices has increased tremendously based on the advent of sensor technologies capable of detecting macromolecules in near instantaneous or real time. To this end, specific, selective miniaturized biosensor assays that combine reliability, speed, and portability while reducing sample size and assay costs are needed to replace conventional identification techniques. Thousands of papers have been published describing a myriad of engineering approaches for microbial biodetection since 1962, when Clark and Lyons (1962) first published their essay on a reusable enzymatic electrode. These approaches are sometimes broadly categorized into optical, calorimetric, biological-biochemical, electrochemical, and acoustic wave-mass change methods. Of these, optical methods (e.g., SPR) and acoustic wave-mass change methods appear to suitably combine speed, sensitivity, and portability for future development of rapid biosensors for microbial analyses (O'Sullivan and Guilbault 1999; Janshoff, Galla, and Steinem 2000; Skládal 2003 ). The term "biosensor" is used rather broadly these days. As defined here, acoustic wave biosensors consist of two main components: a biological receptor that possesses affinity for an analyte of interest, and a piezoelectric transducer to convert the chemical signal of sample-receptor coupling to an amplified signal output that provides qualitative or quantitative assessment of their interaction. Acoustic waves in piezoelectric substrates (e.g., quartz) used as sensor platforms are based on mechanical waves created by an applied electric field. These waves propagate through the substrate and are then transformed back to an electric field for measurement. These discoveries have helped lead to the development of a wide range of acoustic wave devices (Morgan 2000) for applications including sensing of bacterial cells in solution. In this chapter we will focus on the development and use of acoustic wave biosensor platforms for the detection of bacteria, specifically those based on the thickness shear mode (TSM) resonator. We also discuss the mechanical and electrical implications of bacterial positioning at the solid-liquid interface of a TSM biosensor, and present a model of the TSM with bacteria attached operating as coupled oscillators. Biosensors based on acoustic waves are rooted in numerous fundamental concepts, including the discovery of piezoelectricity by Pierre and Paul-Jacques Curie in 1880 (Curie and Curie 1880) , the theory of acoustic waves as predicted by Lord Rayleigh in his 1885 analysis of surface waves in solids (Rayleigh 1885) , and Augustus Love's work on acoustic waves published in 1911 (Love 1911) , which included a description of shear surface waves having motion perpendicular to the sagittal plane. Subsequently, surface elastic waves were first measured in piezoelectric transducers in the 1940s and 1950s (Victorov 1967) , and in 1965 White and Voltmer (1965) experimentally demonstrated direct piezoelectric coupling to surface elastic waves using an interdigital electrode transducer (IDT) on a piezoelectric plate. Piezoelectricity refers to the generation of electrical charges in response to an applied mechanical stress. The converse is also true; application of a suitable electric field to a piezoelectric material (substrate) creates a mechanical stress, or as the name implies "converse piezoelectricity." While there are many different types of acoustic wave devices, all use the converse piezoelectric effect to produce acoustic waves. These waves propagate through a substrate, and are then transformed back to an electric field for measurement. This interconnection between piezoelectricity and acoustic waves has led to the development of a wide range of acoustic wave device applications. The use of acoustic wave devices in electronics can be traced back more than 80 years (Morgan 2000; Gizeli 2002 ) and today includes timing and frequency control for applications that require extreme precision and stability such as mobile phones, satellite communications, and radio transmitters. Several of the emerging applications for these devices in the medical sciences (biological and chemical sensors) and industrial and commercial applications (vapor, humidity, temperature, and mass sensors) may eventually equal the demand of the telecommunications market. Acoustic devices are generally described by way of their wave distribution, either through or on the surface of the piezoelectric substrate. Basically, acoustic waves differ in velocities and directions of particle movement within the substrate. Depending on the material and boundary conditions there can be different variants. Fig. 12 .1 shows the configuration of typical acoustic wave devices. Transverse or shear waves have particle displacements that are normal to the direction of wave propagation and which can be polarized so that the particle displacements are either parallel or normal to the sensing surface. Shear horizontal wave motion signifies transverse displacements polarized parallel to the sensing surface; shear vertical motion indicates transverse displacements normal to the surface. Some properties of selected acoustic wave devices, including TSM, transverse shear mode (i.e., QCM, quartz crystal microbalance); SAW, surface acoustic wave; STW, surface transverse wave; SH-APM, shear horizontal acoustic plate mode; FPW, flexural plate wave; and TRAW, thin rod acoustic wave, are shown in Table 12 .1 (Rickert et al. 1999) . A wave propagating through the substrate is called a bulk wave. The most frequently used bulk acoustic wave devices are the TSM resonator and the SH-APM. A wave propagated on the surface of a substrate is known as a surface wave. The most broadly used surface wave devices are the SAW and shear-horizontal surface acoustic wave (SH-SAW) sensors, also recognized as the surface transverse wave (STW) sensor. The waves are guided by reflection from multiple surfaces. Typical representatives of plate wave devices are FPW and APM. All acoustic wave devices are sensitive to perturbations of many different physical parameters. The change in the properties of the pathway over which the acoustic wave propagates will result in a change in output signal. While all acoustic wave devices will function in gases or vacuum, only a few operate efficiently in liquids. Whether an acoustic wave device can operate in liquid is determined by the direction of the particle displacement at the surface of the device. TSM, SH-APM, and SH-SAW devices all generate waves that propagate primarily in a shear horizontal motion. A shear horizontal wave does not radiate appreciable energy into liquid, allowing functionality without excessive attenuation. Conversely, SAW devices utilizing Rayleigh waves have a substantial surface-normal displacement that radiates compression waves into the liquid and thus cannot be employed in the liquid phase (Grate and Frye 1996) . An exception to this rule occurs for devices using waves that propagate at a velocity lower than that of sound in liquid. Therefore, such modes do not couple to compressional waves in liquid and are thus relatively unattenuated (Ballantine et al. 1997 ). Acoustic wave devices such as the TSM are essentially highly sensitive analytical balances, capable of discriminating extremely small mass deposition events. This makes them excellent analytical tools for the study of specific molecular interactions at the solid-liquid interface in air, and under vacuum or aqueous conditions (Bunde, Jarvi, and Rosentreter 1998; Cavicacute, Hayward, and Thompson 1999; Ivnitski et al. 1999; O'Sullivan and Guilbault 1999; Kaspar et al. 2000; Stadler, Mondon, and Ziegler 2003; Yakhno et al. 2007 ). The TSM resonator may be better known as the quartz crystal microbalance (QCM), because its natural resonant properties are based on the piezoelectric properties of resonators prepared normally from quartz. The QCM usually consists of a thin, round AT-or BTcut (angular orientation in relation to internal crystallography) quartz crystal wafer with two metallic electrodes (e.g., gold, silver, or palladium) deposited uniformly onto both sides of the quartz (Grate and Frye 1996) . The quartz substrate can have varying dimensions and resonant frequencies, the most common being 100 kHz and 1, 2, 4, 5, 8, and 10 MHz (Scherz 2000) . In itself it comprises an oscillatory circuit that can be modeled as an extended Butterworthvan Dyke equivalent circuit depending upon load conditions ( Fig. 12. 2) (Janshoff, Steinem, and Wegener 2004) . The piezoelectric properties of the quartz result in deformation of the crystal when an electrical potential is created across the electrodes, which in turn induces a transverse, standing wave of resonance oscillation in the quartz at a fundamental frequency (Babacan et al. 2000) . AT-cut crystals displace the oscillation parallel to the resonator surface and are utilized predominantly in liquids, due to their temperature stability. Any changes in the resonance frequency of the crystal are usually attributed to the effect of added mass due to binding at the active (overlapping) area of the electrodes. Theoretical modeling of the TSM response to mass accumulation has been demonstrated under various loading conditions, including ideal mass layers (thin layers of Au and SiO 2 ), a semi-infinite fluid (glycerol in water), and a viscoelastic layer represented by thin layers of oil (Martin, Granstaff, and Frye 1991; Bandey et al. 1999) . According to theory (Sauerbrey 1959) , when a mass, m, binds at the surface of the sensor, a corresponding proportional decrease of the resonator's oscillation frequency occurs, the total quantity of which can be solved for using Sauerbrey's (1959) equation as follows, provided that the mass creates a rigid, uniform film that does not slip and has the same acousto-elastic properties as quartz: where f is the observed change in frequency (Hz) of the resonator under oscillation at its fundamental frequency due to mass loading, Cf = sensitivity factor of the resonator in Hz/ng/cm 2 , and m = change in mass per unit area in g/cm 2 . Traditionally, the TSM has served as a mass-sensitive monitor for commercial applications such as thin-film deposition under vacuum, and electroless and electroplating processes. Sauerbrey's (1959) calculations were originally described for depositions under vacuum conditions but his theory has been extended to liquid application, as proof in concept development of sensors for biological analysis have increased dramatically in the past decade. Acoustic wave biosensors in general have been the subject of intense research since the first analytical application reported by King (1964) . As a solitary mass-sensitive transducer the device is non-specific. However, when the electrode is coated with a high affinity receptor or biorecognition component through a reliable deposition process, sample coupling between the receptor and its complementary analyte at the sensor surface can be attributed to a mass change (Rickert et al. 1999 ) that can be converted to a signal output, amplified, and processed to provide specific, sensitive qualitative or quantitative measurement of their interaction. Thus, a biosensor is the spatial unity of a physical transducer and a complementary biological recognition component such as an antibody, bacteriophage, DNA, or enzyme. For more in-depth information regarding theory, the reader can consult numerous excellent references including Janshoff and Steinem (2001) , and Ballantine et al. (1997) . Improved characterization and modeling (Martin, Granstaff, and Frye 1991; Bandey et al. 1999 ) of TSM responses and functionality under liquid loading conditions have hastened development of rapid bacterial biosensors, because for the most part bacteria are naturally found under liquid conditions. There are numerous proposed applications, including use in the food industry (Leonard et al. 2003) , water and environmental monitoring (Kurosawa et al. 2006) , pharmaceutical sciences (Pavey 2002) , bio-threat defense (Ivnitski et al. 1999; Petrenko and Vodyanoy 2003) , and clinical diagnostics (Lazcka, Campo, and Munoz 2007) . The specificity of any TSM sensor is wholly dependent upon a complementary, immobilized bioreceptor. Bioreceptors for whole cell analysis generally correspond to some outside portion of the cell wall such as proteins, or possibly lipopolysaccharides or some other cell wall-associated structure . Direct application for the detection of whole bacterial cells includes food pathogens such as Salmonella spp., Escherichia coli, and Listeria monocytogenes; as well as other human pathogens such as Chlamydia trachomatis, Vibrio cholerae, Staphylococcus aureus, Pseudomonas aeruginosa, Mycobacterium tuberculosis, Fransicella tularensis, Legionella, and Bacillus anthracis spores. A comprehensive list of selectively identified or quantitated bacterial organisms (e.g., biofilm formation in selective culturing media) using acoustic wave devices is given in Table 12 .2. Also, acoustic wave biosensors have been used for the direct detection of other microorganisms including human, plant, and bacterial viruses such as coronavirus (Zuo et al. 2004 ), tobacco mosaic virus (Dickert et al. 2004) , dengue virus , hepatitis A and B (Konig and Gratzel 1995) , rotavirus and adenovirus (Konig and Gratzel 1993) , cymbidium mosaic potexvirus and odontoglossum ringspot tobamovirus (Eun et al. 2002) , and M-13 phage (Uttenthaler et al. 2001) ; yeast (Muramatsu et al. 1986; Hayden and Dickert 2001; Hayden, Bindeus, and Dickert 2003) ; and even algae (Nakanishi et al. 1996) . Acoustic wave biosensors have also been used for indirect detection of microorganisms through the detection of corresponding: DNA from E. coli O157:H7 (Deisingh and Thompson 2001; Mo et al. 2002; Mao et al. 2006) , hepatitis A virus , and human papilloma virus ; specific bacterial protein products for E. coli (Nanduri et al. 2007 ); antigenic proteins from dengue virus Tai et al. 2006 ); (Carter et al. 1995) Ab = corresponding antibody; NS = not selective (e.g. for biofilm monitoring) or some other means of selectivity (e.g. specific culture media) was used other than attached bioreceptor; PM = piezoelectric membranes; SIP = surface imprinted polymer layer; SSBW = surface skimming bulk wave; * = non-specific detection of bacterial growth in milk; * * = non-specific detection of bacterial growth in blood culture bottles. antibodies from bacteria including Helicobacter pylori , Treponema palladium (Aizawa et al. 2001) , Salmonella enteritidis , Francisella tularensis (Pohanka and Skládal 2005) , and Staphylococcus epidermidis (Pavey et al. 1999) , the helminth Schistosoma japonicum (Wu et al. 1999 , and African swine fever virus (Uttenthaler, Kolinger, and Drost 1998) ; and bacterial toxins from E. coli including LT (Spangler et al. 2001) , Stx (Uzawa et al. 2002) , and an unidentified endotoxin (Qu et al. 1998) , and Staphylococcus including SEB and C2 toxins (Harteveld, Nieuwenhuizen, and Wils 1997; Gao, Tao, and Li 1998; Lin and Tsai 2003) . Additionally, new innovations such as gas chromatography coupled to SAW technology have been used for indirect detection of Klebsiella pneumoniae, Pseudomonas aeruginosa, Escherichia coli, and two Candida albicans yeast strains (Casalinuovo et al. 2005) . Ion chromatography combined with TSM has been used to monitor Lactobacillus fermentation through lactic acid production (Zhang et al. 2001) . As shown in Table 12 .2, the overwhelming majority of acoustic wave biosensors described in the literature for direct detection of whole bacterial cells is based on the TSM platform, with the most frequently targeted organism being Salmonella, specifically S. typhimurium. Salmonella is a leading etiology of foodborne illness and death in the U.S. (Mead et al. 1999) . Prominent acoustic wave sensors for Salmonella include those of Prusak-Sochaczewski and Luong (1990) , who reported the first QCM assay for Salmonella with an assay time of 50-60 s, a lower detection limit of 10 5 cells/ml, and 0.5-5 hour incubation period, depending on the concentration of the microbial suspension; Park and Kim (1998) , whose thiolated immunosensor possessed an assay time of 30-90 minutes, a lower detection limit of 9 9 × 10 5 cells/ml, and a detection range up to 1 8 × 10 8 cells/ml; Ye, Letcher, and Rand (1997) , whose linear (R = 0.942) biosensor assay for S. typhimurium had a 25 min response time, a lower detection limit of 5 3 × 10 5 CFU/ml, and a range up to 1 2 × 10 9 CFU/ml; Pathirana et al. (2000) , who developed an antibody-based TSM sensor to detect Salmonella typhimurium in poultry that possessed rapid analytical response times of 79 ± 20 seconds, linear (R > 0 98, p < 0 01) dose-response over 5 decades (10 2 to 10 7 cells/ml) of bacterial concentration, sensitivity of 18 ± 5 mV/decade of S. typhimurium concentration, and a detection range of 350 ± 150 to 10 10 cells/ml; and the sensors of Babacan et al. (2000 Babacan et al. ( , 2002 , Park, Kim and Kim (2000) , Su and Li (2005) , and Kim, Rand, and Letcher (2003) . TSM functionality in liquids is complex. Influences from numerous non-gravimetric contributions include liquid viscosity and density (Bandey et al. 1999) ; surface free energy (Thompson et al. 1991) ; roughness, surface charge density, and water content of biomolecules (Janshoff and Steinem 2001) ; pressure and temperature (Niven et al. 1993) ; and the viscoelasticity and interfacial effects (Lucklum 2005) of thin films deposited in the form of bioreceptors. Therefore, the use of Sauerbrey's (1959) equation to strictly quantitate mass deposited to the solid-liquid interface under liquid conditions is controversial. Sauerbrey's equation was developed based on the oscillation of TSM in vacuum and only applies to thin, uniform, rigid masses attached tightly to the crystal. Thus, frequency response under liquid conditions cannot solely be attributed to mass deposition (Gizeli 2002; Lucklum 2005) . For example, TSM sensors exposed to relatively large protein and polysaccharide molecules in solution have also been shown to give responses that did not correlate with mass changes at the solid-liquid interface (Ghafouri and Thompson 1999) . The authors ascribed this phenomenon to viscoelastic and acoustic coupling at the interface. One could expect especially complicated interfacial properties when the TSM sensor is exposed to larger biological entities such as bacterial cells. Electromechanical forces created by live and moving organisms may contribute to the apparent mass of binding bacteria. Additionally, factors such as nutrition, growth, differentiation, chemical signaling, and mutagenic exposure may also factor in controlling the physiological state of binding bacteria. A bacterial cell (e.g., E. coli) can possess a mass of approximately 665 fg, making it one million times heavier than a typical (150 kD) antibody molecule (Neidhardt 1987) used as a bioreceptor. Bacteria carry out or are involved with various movements including flagellation, Brownian motion, chemotaxis, swimming behavior, adaptation, and other cell phenomena (Alberts et al. 1994) . Bacterial binding on sensor surfaces may also depend on the presence of fimbriae (Otto, Elwing, and Hermansson 1999) , flagella , or other surface-associated adhesion factors, as well as the ability of single cells to associate and form colonies. Bacterial interaction with a biosensor may also be highly dependant upon environmental conditions (Olsen et al. 2003) . Notwithstanding, the ability to function in liquid environments conducive to bacterial growth and the fact that mass can be sensitively and specifically differentiated as a molecular recognition/binding event at the solid-liquid interface are two good reasons the TSM is being developed as a rapid detection tool. Normally, the TSM sensor is enclosed within a cell into which fluids are injected ("flow injection analysis") or flowed via a peristaltic pump. Numerous examples are available by reviewing the references in Table 12 .2. While "closed systems" are prevalent and rather simple in operation, solutions can also be simply applied by pipette directly to the surface of the TSM, or what can be contrasted as an "open system," where fluids are directly applied to the sensor surface (Olsen et al. 2003 (Olsen et al. , 2006 . Systems have also been devised for air-borne sample-to-liquid transfer (Frisk et al. 2006) to facilitate acquisition of airborne threat agents such as Bacillus anthracis spores. Bacterial binding as the signal output of the transducer has been measured and analyzed using many different formats to give a detailed analysis of surface/interface changes, including fundamental resonance and/or overtone frequency shift, frequency shift with dissipation, voltage, resistance and capacitance, and acoustic impedance. Absolute or differential ( f frequency changes alone due to binding are given by most authors and appear to be acceptable and sensitive as a measurement of sensor functionality. For example, Fig. 12 .3a shows typical frequency response curves of a prepared (phage) biosensor tested with logarithmic concentrations of S. typhimurium ranging from 0 (PBS)-10 7 cells/ml. For each concentration, the sensor quickly comes to steady-state equilibrium within several hundred seconds following specific phage-bacteria binding at the solid-liquid interface. Plotting the mean values of steady-state frequency readings as a function of bacterial concentration ( Fig. 12 .3b) gives a high dose-response relationship (R = −0 98, p < 0 001), small signal to noise ratio (-10.9 Hz) measured as the slope of the linear portion of the dose-response, linearity over six decades of bacterial concentration, and a lower limit of detection at 100 cells/ml, well below the infectious dosage of Salmonella. Other authors (Otto, Elwing, and Hermansson 1999) attempt to determine dissipation in the system ( D) as a quantitative measure of system damping, usually due to lossy or viscoelastic films or near surface interaction of the bacteria. Resistance (R), capacitance (C), and/or impendence (L) measurements are sometimes determined (He et al. 2003; Kim, Rand, and Letcher 2003; Su and Li 2005) through a high frequency impedance analyzer based on the TSM as an RLC series equivalent circuit. There are many other characteristics of TSM platforms that require consideration when developing and testing biosensors, including: • Specificity: the strength of the interaction between a molecular probe (e.g., antibody) and an antigen (target analyte) as estimated by the dissociation constant K d . The smaller the K d the higher the specificity of binding. The free energy of dissociation ( G d ) of a ligand-receptor complex is related to its equilibrium dissociation constant K d by the equation: where k is a Boltzmann constant and T is a temperature in K. The equation refers to a standard reference state where all chemical species are 1 M (i.e., K 0 ∼ 0 6 molecules/nm 3 ) and attributes a free energy of zero to a complex with a dissociation constant of 1 M (Chothia and Janin 1975) . • Binding selectivity: is defined by a selectivity coefficient (K). Binding selectivity can be estimated from dose responses of a biosensor to different analytes (e.g., bacteria). The signal response V as a function of the primary analyte (e.g., bacteria) concentration (C) can be represented by the following empirical equation: where C is the primary analyte concentration, A is the constant, and S is the slope of the dose response dependence, defined as the sensitivity of the sensor (Pathirana et al. 1996) . The selectivity coefficient for any other analyte to the primary analyte (e.g., bacteria) (K) can be determined from the signal responses at different concentrations using a method similar to the matched potential method (Pathirana et al. 1996; Umezawa 1996) . The selectivity coefficient is defined as the concentration ratio (R) of primary to interfering species [ C p / C i ], which gives the same response change at the same condition. Using the definition of the selectivity coefficient and Eq. 12.2, the following is derived: where S p and S i are slopes of signal responses to primary and interfering species (other bacteria), respectively. • Sensitivity: The change of the biosensor's output signal when the analyte content (total quantity or concentration) changes by one unit. For non-linear sensors, the sensitivity depends on the analyte level and is given by the slope of the sensor's output curve versus the analyte content. • Detection threshold. The ability of the biosensor to discriminate an analyte (e.g., bacteria) from background at the lowest quantity of analyte in the testing solution. • Dynamic or Working Range: The range of the analyte content over which the sensor can perform qualitative or quantitative detection. • Linear range: That part of the dynamic range where the sensor's output is a linear function of the analyte content. • Saturation: The level at which the sensor no longer functions correctly. For biosensors, this is usually the point where the bioreceptor has been saturated with analyte and reaches a peak signal. • Response Time: The amount of time required to detect the analyte as given by the signal output. • Accuracy: Closeness of the sensor measurement result to the actual quantity of cells in solution. Actual quantity of cells (usually stated in reports as cells/ml) is found from traditional plate culture of the organism. Optical counting methods are also possible (Olsen 2000) . • Stability: The ability of the sensor signal to give a constant, steady output signal when measuring a steady input, such as a load of cells. Traditionally, the TSM has served as a mass-sensitive monitor in commercial applications such as thickness monitoring and deposition rate control for thin films under vacuum, and for electroless and electroplating processes (Krause 1993) . The functionality of the TSM under liquid conditions has increasingly driven adaptation to the development of extremely sensitive biosensors in the past decade. Total QCM systems are relatively inexpensive and simple in operation, requiring for the most part only the resonator crystal, external oscillatory circuit, and frequency counter. Many of the systems described in the literature for sensor developments are pieced together or custom built and may additionally include impedance analyzers, thermostatic jacketing for temperature control, and pump or flow injection equipment. With the advent of the Internet, numerous commercial QCM products including crystal resonators and holders, frequency monitors, flow cells, and even entire systems are now easily available throughout the world, making entry into this field reasonable in terms of cost and availability. A recent review of the Internet yielded numerous larger manufacturers and suppliers of complete QCM systems (Table 12. 3). One such commercially available microbalance produced by Maxtek Inc. can be used for both biosensor preparation and testing and consists of a 50 cm sensor probe connected by a tri-axial cable to a precalibrated plating monitor ( Fig. 12.4) . This system is often used in electroplating processes within vats, necessitating the long probe and open face exposed to solution. The plating monitor has a frequency resolution of 0.03 Hz and mass resolution of 0 375 ng/cm 2 at 5 MHz. TSM transducers are precleaned AT-cut plano-plano quartz liquidplating resonators possessing a 5 MHz nominal resonant frequency. Resonators (2.54 cm diameter, 333 m thickness) have gold plated electrodes evaporated onto titanium adhesion layers on both the top and bottom (Fig. 12.5 ). The electrodes are polished to an average surface roughness of approximately 50 Å. This minimizes liquid entrapment within the pores at the crystal surface, reducing the creation of apparent mass loadings under liquid measurement conditions. Also, resonators are pretested to assure conformance to critical accuracy specifications required for reproducibility, and rate and thickness measurements (PM-740 series operation and service manual 1996). Both the bioreceptor, during sensor preparation (Fig. 12.6a) , and the analyte, during sensor testing, can be directly applied to the surface of the sensor by pipette (Fig. 12.7) . Absolute frequency readings from the sensor are transferred to a PC directly from the plating monitor or via a multimeter, in which case voltage readings can be captured (Pathirana et al. 2000) . The sensor probe, attached to the stand, and all necessary components of the experiment can be contained at room temperature within an Atmosbag™ gloved isolation chamber (Sigma-Aldrich, Milwaukee, WI) inflated with inert nitrogen gas during bioreceptor deposition studies to prevent possible contamination of the resonator by particulate matter. In addition to some of the previously mentioned TSM characteristics, some additional factors to consider before purchasing commercially available equipment include cost, resolution, reproducibility, reliability, ruggedness, analytical range, speed, noise, cost, power requirements, space limitations, availability, technical servicing/maintenance, life expectancy, data capture capabilities, ease of use, and other analytical capabilities and adaptabilities such as use under differing temperatures, pressures, or other environmental conditions, and adaptability to peripheral devices such as voltmeters, PC, peristaltic pumps, and thermostatic jacketing. Like all other sensory devices, the TSM as a sensor platform has its advantages and drawbacks. In addition to addressing necessary characteristics of TSM biosensors such as speed, accuracy, precision, sensitivity, and specificity, several other factors should be considered, including incubation time of analyte, numerous steps including application of analyte and washing and drying, regeneration of the sensor surface if reusability is a factor, and total cost of assay to include resonators, reagents, bioreceptors, etc. A major drawback to the TSM as a sensor is its non-specificity. Anything that can and will attach to it under liquid loading conditions can be recognized as a molecular binding event. Therefore, application of bioreceptors is necessary in order to affect specificity towards the analyte of choice (bacteria, bacteria components, toxins, or complementary DNA, etc.). The sensing properties of a sensor depend on the physical-chemical environment of antibody and antigen-antibody complex, which are in turn determined by antibody immobilization techniques (Ahluwalia et al. 1992; Storri, Santoni, and Mascini 1998) . While the TSM can be very quick in its measurement, building the sensor with bioreceptors can be a tedious, multi-step process that can take numerous hours or even days. Additionally, and possibly the greatest consideration, is the reproducibility of the bioreceptor immobilization process. An in-depth analysis of the techniques of probe immobilization onto sensor surfaces is presented in the following sections. The most common techniques involve direct bonding of an antibody receptor to a reactive group coupled to the surface. The coupling agent and reactive group are generally selected to match the chemistry of the specific antibody. However, the adsorption process is difficult to control and the amount of protein adsorbed to most solid surfaces is usually below that which would correspond to a close-packed monolayer. Further, during the adsorption, the exposure of internal hydrophilic groups of proteins to hydrophobic surfaces causes a decrease in the activity and specificity of the protein/target interactions. In spite of these shortcomings of this method, direct physical adsorption is the simplest way of antibody immobilization on the sensor surface. This method has been successfully employed for immobilization of a wide range of biological elements directly onto piezoelectric electrodes, including anti-human serum albumin (Muratsugu et al. 1993) , IgG (Minunni, Skladal, and Mascini 1994) , goat anti-ricin antibody (Carter et al. 1995a ), anti-Vibrio Cholera (Carter et al. 1995a) , African swine fever virus protein (Uttenthaler, Kolinger, and Drost 1998) , recombinant protein fragments of HIV specific antibodies (Aberl and Wolf 1993) , filamentous phage (Sykora 2003; Olsen et al. 2006; Nanduri et al. 2007; Olsen et al. 2007 ), lytic phage (Balasubramanian et al. 2007) , and designer peptides (Selz et al. 2006) . Protein molecules adsorb strongly and irreversibly on gold surfaces due to hydrophobic actions (Horisberger 1984 (Horisberger , 1992 . Quantitative "dip and dry" deposition experiments can be used to monitor physical adsorption of bioreceptors in the preparation of biosensors (Olsen et al. 2006) . Dip and dry, as described by Prusak-Sochaczewski and Luong (1990) , is the change in the resonant frequency, f , of a dry TSM resonator prior to and after mass deposition. Using Sauerbrey's (1959) equation, the physical adsorption of phage as a function of time can be determined. For example, Fig. 12 .6a shows steady-state oscillation of a dry, clean resonator prior to the application of 1 ml of diluted stock phage E2 in suspension (6 7 × 10 10 virions/ml) (f S ), followed by an 18 h incubation period at room temperature, removal (f R ) of the phage suspension and washing with degassed water, and finally drying, with a subsequent return to steady-state resonance (f E ). The resulting frequency change, f , measured as a decrease, f S − f E , was -161 Hz, indicating that phage adsorbed to the resonator. This can be contrasted to a control (Fig. 12.6b) consisting of a clean TSM resonator tested with degassed water only, which indicated no frequency change (f S − f E = 0 Hz). Resonance frequency changes due to phage adsorption were determined for periods up to 24 h (1440 min) (Table 12 .4). When the quantity of adsorbed phage is graphed as a function of time the majority of phage appeared to adsorb within the first few hours after deposition was started (Olsen et al. 2006) . The quantity of phage in virions can be calculated from the total adsorbed biomass, m, by estimating the mass of a single recombinant fd-tet phage at 2 66 × 10 7 daltons, based on 4000 pVIII outer coat proteins, each containing 55 amino acids with a total molecular weight of 2 35 × 10 7 (Kouzmitcheva 2005) , and DNA with a molecular weight of 3 04 × 10 6 . As shown in Table 12 .4, the total number of phage particles deposited to the TSM resonator ranged from 1 8 × 10 10 − 6 5 × 10 10 virions as a function of exposure time (20 min -24 h, respectively) to phage in solution. Phage deposition to the TSM was confirmed in real time by fluorescence microscopy for a period of two consecutive hours and characterized by strong, non-reversible binding under aqueous conditions (Olsen et al. 2006 ). To overcome disadvantages of a direct physical adsorption method, a range of immobilization methods have been suggested, including lipid bilayer entrapment (Ramsden 1997a (Ramsden , 1997b (Ramsden , 1998 (Ramsden , 1999 (Ramsden , 2001 , thiol/disulfide exchange, aldehyde and biotinavidin coupling (Mittler-Neher et al. 1995) , photo-immobilization to photolinker-polymerprecoated surfaces (Gao et al. 1994) , molecular imprinted polymer layers Dickert et al. 2004) , and site-specific immobilization of streptavidin (Tiefenauer et al. 1997) . A more advanced approach for the immobilization of antibodies for the immunosensor coatings is through the combined Langmuir-Blodgett (LB)/molecular assembling method (Samoylov et al. 2002a (Samoylov et al. , 2002b . The method includes LB deposition (Petty 1991; Pathirana et al. 1992 ; Barraud et al. 1993; Pathirana 1993; Vodyanoy 1994; Bykov 1996; Pathirana et al. 1996; Sukhorukov et al. 1996; Pathirana, Neely, and Vodyanoy 1998; Olsen 2000; Pathirana et al. 2000; Olsen et al. 2003; Olsen 2005; Olsen et al. 2007 ) of a biotinylated monolayer onto a sensor surface and non-LB, molecular self-assembling of a probe layer using biotin/streptavidin coupling (Furch et al. 1996; Volker and Siegmund 1997) . The combined LB/molecular assembling method has been demonstrated with biosensors based on phage display-derived peptides as biorecognition molecules (Samoylov et al. 2002a (Samoylov et al. , 2002b . Schematic design of the peptide sensor is shown in Fig. 12 The biosensor consists of four components: 1 -quartz crystal; 2 -biotinylated phospholipid; 3 -streptavidin; 4 -biotinylated peptide; 5 is a tissue vesicle. (b) Calibration of acoustic wave device with stearic acid monolayers. (c) Validation of peptide sensor preparation. ST -sensor was covered with biotinylated phospholipid and exposed to streptavidin. The bar represents the change of mass due to binding of streptavidin to biotinylated phospholipid. -gal -sensor covered with streptavidin was exposed to -gal solution. The bar represents the change of mass due to binding of -gal to streptavidin. P/ -galcompleted sensor, covered with peptide, was exposed to -gal. The bar represents the change of mass due to -gal binding to the sensor. P/Mcompleted sensor, covered with peptide, was exposed to murine muscle homogenate. The bar represents the change of mass due to binding of the component of the tissue homogenate (3.8 mg/ml protein) to the peptide. Langmuir-Blodgett (LB) monolayers. First, a monolayer is formed on a subphase surface and compressed to a desired surface pressure. A solid substrate is then moved through the monolayer vertically, so that it is dipped into and withdrawn completely out of subphase. The first monolayer is transferred to the substrate like a carpet with the tail groups toward the substrate surface during the downward movement of the substrate through the condensed monolayer. A monolayer is transferred to the substrate both when it is raised through and when it is lowered into the subphase through a compressed monolayer (reprinted from Yilma et al. (2007) , with permission of Elsevier). containing phospholipid, N-(biotinoyl)-1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine (2), were transferred onto the gold surface of an acoustic wave sensor (1) using the Langmuir-Blodgett technique. Multilayers were obtained by successive dipping of the sensors through the monomolecular film deposited at a water-air interface ( Fig. 12.9 ). Biotinylated peptide (4) was coupled with the phospholipid via streptavidin intermediates (3) by molecular self-assembly. Measurements of binding of target vesicles were carried out using a PM-700 Maxtek plating monitor with a frequency resolution of 0.5 Hz at 5 MHz. The device was calibrated with stearic acid monolayers. The deposition of increasing numbers of stearic acid monolayers on the surface of acoustic wave crystal resulted in linear increase of the mass (Fig 12.8b) . The deposition of a single monolayer of stearic acid on the crystal resulted in additional mass of 2 6 × 10 −7 g cm −2 . This agrees well with the theoretical estimate based on the molecular area of stearic monolayer in the condensed state (Davies and Rideal 1963) . Binding of streptavidin to biotinylated phospholipid is an important step in immobilization because concentration and orientation of streptavidin molecules determine the properties of a bound molecular probe. The change of mass due to streptavidin binding normally reached 80 ng cm −2 , or 8 × 10 11 molecules cm −2 (Fig. 12. 8c-ST). When the samples were exposed to 500 nM biotinylated -gal for 2 h the apparent mass change was at the level of 80 ng cm −2 , or 3 4 × 10 11 molecules cm −2 (Fig. 12.8c--gal) . The completed biosensor, covered with the biotinylated peptide, no longer bound biotinylated -gal ( Fig. 12 .8c-P/ -gal), but strongly bound target vesicles (Fig. 12 .8c-P/M). The combined LB/molecular assembling method was also exercised in the immobilization of filamentous phage onto the surface of thickness shear mode (TSM) quartz sensors Olsen et al. 2006 Olsen et al. , 2007 . Monolayers containing biotinylated phospholipid were transferred onto the gold surface of the sensor using the Langmuir-Blodgett technique (Fig. 12.9 ). Biotinylated phage was coupled with the phospholipid via streptavidin intermediates by molecular self-assembling. The dissociation constant of 0.6 nM found by this method compares well with one found for antibodies isolated from a phage display library (Vaughan et al. 1996) . An important aspect of sensor preparation is defining the conditions under which monolayers prepared with bioreceptors can be successfully formed on a liquid/gas interface and then optimized in terms of sensitivity, reliability, and useful lifetime. Although some effects of pH, ionic strength, and oriented coupling on the immunosensor performance have been examined (Barraud et al. 1993; Ahmad and Ahmad 1996) , detailed information about the influence of physical, chemical, and molecular environments on the antigen-antibody system remains largely unknown. Traditional methods for forming LB films (Gaines 1966 ) require dissolution of monolayer forming compounds into a volatile organic solvent. As a separate phase, the organic solvent functions to prevent dissolution of the monolayer components in the aqueous phase. When the mixture is spread onto an aqueous subphase solution at the air-liquid interface, the solvent evaporates, leaving a monolayer at the interface. Unfortunately, the organic solvent often damages the monolayer components and leaves an undesirable residue (Sykora, Neely, and Vodyanoy 2004) . LB films formed from such monolayers may also possess unacceptable levels of nonspecific binding (Ahluwalia et al. 1992) , which is non-saturable and hampers quantitative measurement of specific binding. These problems can be solved using methods of monolayer formation that don't require use of an organic solvent (Trurnit 1960; Sobotka and Trurnit 1961; Pattus, Desnuelle, and Verger 1978; ) and have been demonstrated by immobilizing polyvalent somatic O antibodies specific for most Salmonella serovars onto gold electrodes of TSM resonators using the LB method (Pathirana et al. 2000) . Many features of antibody immobilization originate from the very nature of the antibody itself. Typical antibodies are Y-shaped molecules (2 Fab plus Fc immunoglobulin structure) with two antigen binding sites located on the variable region of the Fab fragments. All classes of antibody produced by B lymphocytes can be made in a membrane-bound form and in a soluble secreted form (Alberts et al. 1994 ). The two forms differ only in their carboxyl terminals; the membrane-bound form has a hydrophobic tail (Fc) that anchors it in the lipid bilayer of the B cell membrane, whereas the secreted form has a hydrophilic tail, which allows it to escape from the cell. Of these, only the form with a hydrophobic tail is capable of being held by the monolayers. Thus, it is uniquely qualified for use in the Langmuir-Blodgett technique. This form also renders it suitable for proper alignment and orientation in sensor membranes. Antibodies derived from immunized animals in the form of antisera or purified protein preparations can be present in both membrane-bound and soluble form and may contain impurities. Organic solvents used as a spreading carrier in LB monolayer preparation may drag these impurities and both forms of antibodies into the monolayer. Furthermore, these methods may produce monolayers with high densities of antibodies but also with residuals of organic solvent, impurities, and entrapped hydrophilic antibodies that destabilize the monolayer and modulate antigen-antibody interactions. A monolayer with no solvent can be formed on the air-liquid interface by allowing the spreading solution to run down an inclined wetted planar surface that is partially submersed into subphase (Fig. 12.10) (Pathirana et al. 2000) . Membrane vesicles (natural components of serum, or the artificial lipid vesicles) are positioned on a wet slide at the edge of a positive meniscus of liquid, at the liquid-air interface. The hydrophobic antibodies are bound to the vesicular membrane; hydrophilic antibodies and some impurities are suspended inside the vesicle. When surface forces rupture the vesicle, it splits into a monolayer and purification occurs. Membrane-bound antibodies are left bound to the newly created monolayer, but soluble antibodies and impurities dissipate into the subphase beneath the monolayer. Only membrane-bound antibodies surrounded by compatible lipids are left when the monolayer is compressed and transferred onto a sensor surface. Alternatively, probes can be conjugated with vesicles by covalent binding (Betageri et al. 1993) . Lipid vesicles containing whole antibodies or Fab fragments can also be constructed. Large, unilamellar liposomes are prepared from synthetic L-(1,2-dipalmitoyl-sn-glycero-3-phosphocholine) (DPPC) and maleimido phenyl butyrate phosphatidylethanolamine (MPB-PE). Monoclonal antibodies (or Fabs) with specificity for S. typhimurium or E. coli O157:H7 are modified by the heterobifunctional reagent N-Succinimydyl-3(2-pyridyldithio)propionate (SPDP) in the presence of dicetyl phosphate and dithiothreitol (DTT) and conjugated to liposomes. The liposomes, containing specific antibodies, are then converted into monolayers and deposited on the sensor surface by the LB method (unpublished results). A large number of bio-assays and biosensors depend on highly specialized, sensitive, and selective antibodies as recognition reagents (Goodchild et al. 2006) . While antibodies frequently have the desired sensitivity and selectivity, their use is limited by many factors. For example, the binding properties of antibodies may be lost due to unfavorable environmental conditions (Olsen et al. 2003) . This factor can be especially important in dealing with environmental applications, where organic solvents must be used for extraction of compounds (Ahmad and Ahmad 1996) . Also, production of polyclonal antibodies requires a process that is very timeand labor-intensive, and can produce a variable product. Production of monoclonal antibodies is often even more difficult and expensive. These limitations can be addressed in part by using bacteriophage or their coat proteins as recognition elements for biosensors (Goldman et al. 2000; Petrenko et al. 2005; Nanduri et al. 2007 ). Both lytic and filamentous phages present reach libraries to identify proteins interacting with molecular targets. In phage display, the phage filament serves as the framework for random peptides that are fused to the N-terminus of every copy of the major phage coat protein. These random peptides form the "active site" of the landscape phage and comprise up to 25% by weight of the particle and up to 50% of its surface area (an extraordinarily high fraction compared with natural proteins, including antibodies) (Nanduri et al. 2007) . A large mixture of such phages, displaying up to a billion different guest peptides, is called a "landscape library." From this library, phages can be affinity selected for specificity to a certain antigen, thus functionally mimicking antibodies. These phages can be efficiently and conveniently produced and are secreted from the cell nearly free of intracellular components in a yield of about 20 mg/ml (Nanduri et al. 2007) . The purification procedure is simple and does not differ dramatically from one clone to another. The surface density of the phage binding peptides is 300-400 m 2 /g, comparable to the best known absorbents and catalysts (Nanduri et al. 2007) , and with thousands of potential binding sites per particle, creates a multivalency. Other advantages of phages over antibodies include the extraordinary robustness of the phage particle. It is resistant to heat (up to 70 C), many organic solvents (such as acetonitrile), urea (up to 6M), acid, alkali, and many other stresses (Nanduri et al. 2007 ). Purified phages can be stored indefinitely at moderate temperatures without losing infectivity (Nanduri et al. 2007 ). Thus, phages may be viable as substitute antibodies in many applications such as biosensors, affinity sorbents, hemostatics, etc. Numerous examples of uses of both lytic and filamentous phages as probes for biological detection in biosensors have been reported in the literature (Chin et al. 1996; Ramirez et al. 1999; Goldman et al. 2000 Goldman et al. , 2002 Auner et al. 2003; Olsen et al. 2003; Sayler, Ripp, and Applegate 2003; Ozen et al. 2004; Tabacco, Qian, and Russo 2004; Chen et al. 2005; Nanduri 2005; Petrenko et al. 2005; Sorokulova et al. 2005; Wu et al. 2005; Lakshmanan et al. 2006; Olsen et al. 2006; Balasubramanian et al. 2007; Nanduri et al. 2007 ). A critical step in the use of whole phages or phage proteins as a bioreceptor is their immobilization to the platform area on the sensor where the analytes (bacteria, toxins, etc.) will bind. Immobilization of whole phage particles to a sensor surface presents certain difficulties with phage positioning. While multivalent, phage particles are typically assembled in bundles that may present obscurity of binding sites (Fig. 12.11) . Immobilization of proteins extracted by organic solvents may present difficulties in depositing a correctly orientated layer of proteins on the sensor surface. A better way of phage protein immobilization is to use the intact protein coat of the phage particle. For example, when T2 lytic bacteriophage was subjected to osmotic shock, it lost most of its DNA, but the protein coat or "ghost" of the phage was left intact (Herriott and Barlow 1957) . The coat retained the phage shape and some of the biological functions of the phage. Kleinschmidt and coworkers (Kleinschmidt et al. 1962 ) were able to convert T2 coat particles into protein monolayers by allowing the water suspension of coat particles to run down a wet glass rod glass surface that was partially submersed into subphase (similar to that shown in Fig. 12 .10) (Trurnit 1960; Sobotka and Trurnit 1961; Pattus, Desnuelle, and Verger 1978; Pathirana et al. 2000) . Monolayers made of the phage coats transferred to solid substrates were first described by Kleinschmidt et al. (1962) . A similar approach was applied to obtain monolayers of filamentous phages. Griffith and coworkers (Griffith, Manning, and Dunn 1981; Olsen 2005) demonstrated that filamentous bacteriophages transformed into hollow spherical particles upon exposure to a chloroform-water interface. These particles could then be converted into monolayers and deposited onto solid substrate by the LB method (Sykora 2003; Olsen 2005) . Thus, the technology of phage coat immobilization consists of three major steps: phages are first converted into spheroids, monolayers are formed from the spheroids, and finally the monolayers are deposited onto the sensor surface by the LB method. Phage coat monolayers made of coats of lytic and filamentous phages have been immobilized onto biosensor surfaces. When a suspension of filamentous phage protein streptavidin binder, 7b1, (Petrenko and Smith 2000) was vortexed with an equal volume of chloroform and the aqueous phase was examined by electron microscopy, spherical particles termed "spheroids" were observed along with other semicircular particles that may be intermediates in the filament to spheroid conversion (Griffith, Manning, and Dunn 1981) . Chloroform transforms the infectious phage filaments into non-infective hollow spheres. This drastically alters the surface architecture of the phage. As well, the -helix content of pVIII decreases from 90% to 50-60% (Griffith, Manning, and Dunn 1981; Roberts and Dunker 1993) . Spheroids are formed when the coat proteins contract into vesicle-like structures and two-thirds of the phage DNA is extruded (Griffith, Manning, and Dunn 1981) . An electron micrograph of this is shown in Fig. 12 .12 . Similar hollow spheroids can be obtained by the same method from phage f8-1 that bind Salmonella typhimurium (Olsen 2005) . Olsen formed LB monolayers from the spheroid suspension using a wetted glass rod that was partially submersed into the subphase (Trurnit 1960; Sobotka and Trurnit 1961; Kleinschmidt et al. 1962; Pathirana et al. 2000; Olsen 2005) . After the vesicle slid down the glass rod and reached the air-water interface, surface forces ruptured the vesicle and split it into a monolayer. Compression of an LB monolayer prepared from a spheroid suspension yields a pressure ( )-area (A) isotherm ( Fig. 12.13) (Sykora 2003) . The curve is biphasic, having a small "kink" Figure 12 .12. Electron micrographs of 7b1 filamentous bacteriophage following chloroform treatment. Sample was stained with 2% phosphotungistic acid. Spherical particles are called "spheroids." Mag., 302,500x (A) and 195,300x (B) ; reproduced by permission of The Electrochemical Society). around 20 mN/m, followed by a steep condensed region. A pressure of ∼ 50 mN/m was noted before the barrier reached the end of its stroke. This pressure is very high for protein monolayers, indicating a very stable system (Davies and Rideal 1963; Gaines 1966) . Fig. 12 .14 shows elasticity versus surface pressure for the monolayer. There are two maxima in elasticity separated by a minimum around 20 mN/m (from the "kink" in the isotherm). The largest maximum reach was ∼ 50 mN/m, very high for protein monolayers (Davies and Rideal 1963; Gaines 1966) , at a pressure of 30 mN/m. This pressure is optimal for transferring monolayers onto solid sensor substrates by the LB method . A similar technology for immobilization of a Salmonella binder phage E2 onto a surface of a QCM sensor was developed by Olsen et al. (2007) . Figure 12. 14. Graph of elasticity versus surface pressure ( ) for monolayer formed from spreading of spheroid suspension. Elasticity was calculated from the surface pressure-area isotherm (Fig. 8.13 ). The amino acid sequence of the pVIII coat protein from phage 7b1 with the foreign octapeptide insert is shown in Fig. 12.15 . Conformation of the pVIII binding peptide at both the air/water interface and on the sensor surface can be elucidated based on the amino acid sequence. The octapeptide insert consisting of residues number 2 through 9 is located at the N-terminal region. Based on the amino acid sequence, the peptide is expected to have threehelical regions according to both the Garnier-Robson (Garnier, Osguthorpe, and Robson 1978) and Chou-Fasman (Chou and Fasman 1974) calculation methods, as shown in Fig. 12.16 . The central region of the peptide is hydrophobic, while the N-terminal and C-terminal regions are somewhat hydrophilic. This hydrophobic region is the part of the peptide that spans the bacterial cell membrane during assembly (Bashtovyy et al. 2001; Houbiers et al. 2001; Branch and Brozik 2004; Houbiers and Hemminga 2004; Aisenbrey et al. 2006 ). Most of the amphipathic and flexible regions of the peptides correspond with the hydrophilic regions of the peptide. According to calculations of the antigenic index, the most probable antigen-binding region lies on the N-terminus, which is the region where the octapeptide insert is located. A hypothetical arrangement of the pVIII coat proteins at the air/water interface is shown in Fig. 12.17 . Here, the hydrophilic N-terminal and C-terminal -helices interact with the water phase while the central hydrophobic region remains at the interface. A hypothetical arrangement of these peptides on the sensor surface is subsequently shown in Fig. 12 .18 (Bashtovyy et al. 2001; Houbiers et al. 2001; Houbiers and Hemminga 2004; Im and Brooks 2004; Aisenbrey et al. 2006) . Here the peptides are suggested to be arranged in a conformation similar to that Hypothetical schematic of the arrangement of pVIII coat proteins at the air/water interface. N and C designate the N-terminus and C-terminus of the peptide, respectively, while the cylinder represents the hydrophobic domain. ; reproduced by permission of The Electrochemical Society). ; reproduced by permission of The Electrochemical Society). in the phage particle, where the positively charged lysine residues of the C-terminal region interact with the negatively charged gold surface, thus allowing the N-terminal region and the octapeptide insert to be exposed to solvent. Immobilization of large molecular probes (antibodies, proteins, DNA, etc.) requires a complex environment in order to maintain viability and functional activity of the probes. These conditions are difficult to meet on a continuous solid sensor surface. Under natural conditions biological receptors are supported by biological membranes that are interfaced with water solutions on both sides. The Australian group of Cornell (Cornell et al. 1997 (Cornell et al. , 2001 ) devised a multi-step assembly procedure to tether a lipid bilayer containing molecular probes linked to a gold surface. Such a tethered configuration is of interest in general for sensor technology because it creates a water reservoir between the sensor surface and membrane and serves to maintain the bilayer fluidity and facilitate the incorporation of molecular probes. Although this example clearly demonstrates the feasibility of an electrode-supported ion channel-based sensor, it suffers from several serious problems. First is the shear complexity of the synthetic approach; using thiol-and lipid-based self-assembly techniques, six different reagents are sequentially organized onto the gold surface. Second, because the tethers are randomly positioned on the electrode, weaker unsupported regions of the membrane could collapse. Finally, there appears to be no obvious patterning procedure. To overcome the above problems and in addition develop qualitatively new functionalities, thin porous substrates of silica have been developed (Fan et al. 2000; Jiang et al. 2006; Nishiyama et al. 2006; Xomeritakis et al. 2007 ). These films can be used as a new type of support for molecular probes in biosensors (Thust et al. 1999; Bessueille et al. 2005; Gawrisch et al. 2005; Yun et al. 2005; Dai, Baker, and Bruening 2006; Song et al. 2006) . In our laboratory we have immobilized the antimycotic agent amphotericin B onto the porous silicon surface and observed ion currents associated with ion conductance of amphotericin B ion channels connected to internal and external reservoirs of liquids separated by a membrane (Yilma et al. 2007a (Yilma et al. , 2007b . The thickness shear mode quartz crystal resonator (QCM) is often considered a masssensitive sensor. During the sensing process, it's expected that the response (frequency change, f ) of the sensor is directly related to any additional mass that adheres to the resonator, usually resulting in a resonance frequency decrease f < 0 . Sauerbrey (1959) demonstrated this for thin, rigid layers (like metal film). For mass m the frequency decrease is where, f 0 is the resonance frequency and M q is the mass of a quartz oscillator. However, use of QCM for in situ bacteria and cell detection in fluids has revealed more complex sensor responses. For example, the observed value of f differs from that predicted by Sauerbrey's (1959) relation, and the signal is often small (Thompson et al. 1991; Voinova, Jonson, and Kasemo 2002) . Observed deviations from Sauerbrey's (1959) predicted mass theory have been noted during sensor testing under liquid conditions using both antibodies (Olsen et al. 2003 ) and phages (Olsen et al. 2006 ) as bioreceptors. The most peculiar results show that under certain conditions there is an appearance of a negative apparent mass; i.e., with increasing bacterial concentration there can be a dose-dependent decrease of the apparent mass. It is possible that the bacterial microenvironment and location of the antigen on the surface of a bacterium can determine the value and sign of the analytical signal generated by the acoustic wave device (Olsen et al. 2003) . Bacterial positioning and binding may be very important at the solid/liquid interface of the sensor and factors such as viscoelasticity, shear forces and damping. For example, in our studies using Salmonella and E. coli antibodies as receptors (Olsen et al. 2003) , when attachment between bacteria and bacteria-specific somatic O antibodies at the solidliquid interface of a TSM resonator was rigid and strong, the sensor's output was directly proportional to the logarithmic concentration of free bacteria in suspension, and the sensor's behavior could be described as that predicted by mass theory ( f < 0) (Fig. 12.19a) . Conversely, flexible binding observed for bacteria attached by flagella to immobilized flagellar H antibodies resulted in inversely proportional sensor signals f > 0 ( Fig. 12.19b ). This premise was affirmed by studying the responses of environmentally aged sensors. Sensor responses and binding efficiency, confirmed by dark-field microscopy, decreased as the duration of sensor environmental aging under differing conditions of temperature increased (Fig. 12.20) . Viscoelastic properties of the bacterial layer attached to the surface are anticipated to be different depending on the mechanism of binding-somatic or flagellar. Also, the viscous shear and viscous drag forces of the attached bacteria are very different. Clearly, bacteria rigidly or flexibly attached (Fig. 12.21 ) take different roles in the oscillation of the whole system. When binding is rigid, bacteria oscillate in unison with the sensor and therefore contribute to the effective oscillating mass of the system. This is shown by the increase of the apparent Olsen et al. (2003) , with permission of Elsevier). mass when bacterial binding concentration is increased. Conversely, in the case of flexible attachment, the oscillation of the bacteria may be not in phase with the oscillation of the sensor, resulting in a decrease in the apparent mass even when concentration of binding bacteria is increased. Additionally, we propose that the electrically charged bacterium on the surface of an acoustic wave sensor is not only engaged in the mechanical oscillations of the crystal but also directly interacts with the electric field driving the sensor crystal. This field drives the piezoelectric quartz crystal and at the same time creates an electrophoretic force applied to the electrically charged bacterium. The piezoelectric and electrophoretic forces can be of different values and directions, depending on the positioning of bacteria by the O antigen ( Fig. 12. 21afirm positioning) or the H antigen ( Fig. 12.21a-flexible attachment) , and their combination may contribute to the change of the apparent mass of the bacteria as measured by the acoustic wave device. Obstruction of antibodies by the buildup of a biofilm during aging may cause decreased accessibility to bacterial targets (Fig. 12.21a-weak binding) . In contrast with Sauerbrey's (1959) observations for a thin, firmly attached film, this seeming contradiction of normal mass loading theory is consistent with the observations of Olsen et al. (2003) , with permission of Elsevier). Dybwad (1985) , who first described mass-dependent frequency increases in conjunction with particulates, such as small (10-50 m diameter) Au spheres, under normal atmospheric conditions while loosely attached to a horizontally positioned QCM resonator. Dybwad's (1985) proposed equivalent mechanical model (Fig 12.21b ) of a loosely bound particle as a coupled mass-spring system corresponds exactly with the flexible attachment of our simplified model shown in Fig. 12 .21a (Olsen et al. 2003) depicting bacterial positioning at the sensor surface as the determinant factor of the sensor's analytical response. Dybwad's (1985) results were affirmed by Vig and Ballato (1998), who stated: "Significant deviation from the Sauerbrey equation will also occur when the mass is not rigidly coupled to the QCM surfaces. The effects of liquids have been discussed in the sensor literature [references given]; however, the effects of nonrigid coupling of solids do not seem to be well- Olsen et al. (2003) that shows bacterial binding positions (p) at the solid/liquid interface of the quartz resonator (QR). When binding is firm between bacteria and receptor (left), the natural frequency of the cell as an independent mass-spring system equals the frequency of the resonator, forming a composite unit that produces expected mass loading effect with corresponding frequency decrease. When binding is flexible or weak (center and right, respectively) between analyte and receptor a coupled oscillator is formed, the frequency of which is dictated by the difference in the spring constants between the oscillator and bacteria. (B) Coupled oscillator model of Dybwad (1985) depicting quartz resonator (QR) as one mass (M) spring (K) system, and a loosely attached particle (p) as a second mass (m) spring (k) system. Attachment of the loose particle causes QR to oscillate at a new, higher frequency when k < K. When k = K, a composite system is formed that produces expected mass loading effect with corresponding frequency decrease. known. For example, when a particle is placed on an electrode of a QCM, the Sauerbrey equation predicts a decrease in the frequency of the QCM, but the frequency actually increases. When the particle on the resonator is modeled as a coupled oscillator, the model correctly predicts a frequency increase [as verified by Dybwad (1985) ]." This type of behavior has been documented in the literature on several occasions. Berg, Johannsmann, and Ruths (2002) used a single asperity contact to show that frequency shifts associated with a quartz resonator operating in shear mode increased linearly with increasing contact radius. Borovsky et al. (2001) used a nanoindenter probe in conjunction with a QCM to elicit positive frequency shifts characteristic of the contact stiffness. Sorial and Lec (2004) experienced size-dependent frequency increases with polystyrene spheres using a QCM under aqueous conditions. Otto et al. (1999) observed diminished, f < 0, response for weaker bacteria attachment using an E. Coli QCM sensor. Other documented reports of negative apparent mass using QCM-based platforms include Hayden et al. (2003) and , in response to loose binding of bacterial cells in yeast imprinted layers; Dickert et al. (2004) , due to loose binding of non-specific compounds to tobacco mosaic virus imprinted polymers; Marxer et al. (2003) , who attributed frequency increases to alterations in cytosolic viscosity of adsorbed epithelial cells; Thompson, Arthur, and Dhaliwal (1986) , due to immunochemical interactions resulting in decreased acoustic transmission at the liquid/solid interface; and Pereira de Jesus, Naves, and Lucia do Lago (2002), as a result of polymeric film stiffness in the determination of boron. More recently, Lucklum (2005) described non-gravimetric contributions of viscoelastic films at the solid/liquid interface, resulting in positive frequency shifts. He clearly showed that typical elastic and energy dissipation properties are as important to frequency response, both positive and negative, as the layer's mass and therefore in many cases the added mass cannot be determined simply from the QCM response alone. The author correctly notes that the traditional name "quartz crystal microbalance (QCM)" as a technique is misleading. Collectively, these reports support our hypothesis that positive frequency shifts can be observed under certain conditions as a result of viscoelastic changes at the solid/liquid interface from surface films, bioreceptor layers, and bacterial attachment. We contend that the continuous model of the viscoelastic layer is not directly applicable to the bacterial sensing. This is a consideration for researchers in terms of the importance of bacterial attachment schemes (bioreceptors) that result in high-affinity and multiple binding valences. In the next section, we present a coupled oscillators model that explicitly accounts for discrete events of bacteria attachment in terms of the elastic constant and the dissipation of the bond between bacteria and the sensor surface. This model is in agreement with our previous model (Olsen et al. 2003) , since the parameters of an LRC series equivalent circuit model of the QCM are perfectly analogous to a damped harmonic oscillator system (Table 12 .5). The continuous layer model does not elucidate total understanding of the bacterial sensing process when discrete bacteria are bonded to the sensor surface. We present here a simple coupled oscillators model, depicted in Fig. 12.22 . The unloaded quartz oscillator is described by the oscillator of mass M connected to the spring with the force constant K and moving in the fluid with the viscous friction force v, where v is velocity. This oscillator models realistic assay conditions where the bacterial bioreceptors (e.g., antibodies or phages) are deposited on the surface of an immersed TSM transducer, but no bacteria are present. The bonded bacterium of mass m is connected to the oscillator by an elastic bond with the force constant k and experiences a viscous friction with coefficient . The equations of the motion for the quartz oscillator and each of the oscillating bacteria numbered by the index i (i= 1, 2…n) are: Here F 0 sin t = ImF 0 e iwt is the periodic external force driving the oscillator. We want to find a stationary solution of Eqs. (12.6) when all the oscillators move with the frequency of the external force: Substitution of Eq. (12.7) into Eqs. (12.6) gives a system of linear algebraic equations: To solve Eqs. (12.8) with respect to X 0 , the variables x i0 are expressed as: Substitution of Eq. (12.9) into the first part of Eq. (12.8) gives: Here, 0 2 = K/M 0 2 = k/m (12.11) denotes the resonance frequencies for uncoupled oscillators. The resonance frequencies of coupled oscillators correspond to maxima of the amplitude X 0 as a function of . Fig. 12 .23 presents an example of the resonance curve for 0 /2 = 5 · 10 6 Hz, /2 M = 1Hz, 0 /2 = 3 · 10 6 Hz, /2 m = 1Hz, n = 100, and m/M = 2 · 10 −7 . As expected, the main resonance is near the quartz oscillation frequency 0 , because the effect of bacteria binding is small. The second weak resonance occurs near the bacterium oscillation frequency 0 , but it is too weak to be registered experimentally. When damping does not occur ( = 0, = 0), complete consideration of the resonances is feasible in closed analytical form. At resonance the oscillation amplitude given by Eq. (12.10) grows infinitely because the denominator is zero: This is a square equation for 2 and it has two positive solutions, − and + . Fig. 12 .24 shows − and + as functions of the bacteria elastic binding strength 0 and 0 /2 = 5 · 10 6 Hz, n = 100, and m/M = 2 · 10 −7 . As discussed above and shown in Fig. 12 .23, only the resonance near 0 is strong and observable in a realistic case when damping occurs. Thus, in Fig. 12 .24, the observable resonance corresponds to + branch when the bacteria binding is weak ( 0 < 0 , but is given by − branch when the binding is strong ( 0 > 0 ). Hence, there are two systems corresponding to the two marked rectangular areas in Fig. 12 .24. First, for weak bacteria binding ( 0 < 0 ) the resonance frequency increases in contradiction to the intuitive expectation that the addition of the bacterial mass to the oscillator will decrease the frequency. The positive frequency shift grows when the bacteria binding 0 increases and approaches the frequency 0 . The second system corresponds to strong bacteria binding ( 0 < 0 ) when the resonance frequency decreases as expected for mass added to the oscillator. In this case For weak attachment ( 0 < 0 ), the real part of this complex number is negative. This means that for small the bacteria and quartz surface oscillate with the phase shift close to 180 , i.e., anti-phase. Hence, the bacterial oscillator pushes the quartz oscillator towards the equilibrium position (Fig. 12.25 ). This increases the restoring force and thus also increases the resonance frequency of the quartz oscillator. In the case of strong bacteria attachment ( 0 > 0 ), the bacterial and quartz oscillators move approximately in the same phase and the frequency decreases. Therefore, the model of coupled oscillators clearly reveals the mechanism of the apparent "negative mass" effect. To discuss the performance and rational design of the bacterial sensor we calculated the frequency response as a function of the number of attached bacteria. Fig. 12 .26 shows the results for different strengths of the elastic attachment bond. The output signal (frequency shift) is considerably stronger for strong attachment. This suggests using a strong attachment method to achieve a low bacteria detection threshold. Importantly, a substantial linear range of detection occurs that is especially broad for weak attachment. This allows finding the number of attached bacteria by calibration measurement of only the initial slope of the response curve. Additionally, we modeled the effect of inhomogeneous bacterial binding strength. Fig. 12 .27 shows that for two types of binding sites with 2 and 10 MHz attachment strength, the response of a 5 MHz sensor crucially depends on the distribution of bacteria between the sites. In particular, at about 0.8/0.2 distribution, the signal is very small and thus a false negative result will be measured. This underlines the importance of using the bioreceptors with a narrow distribution of strong binding strengths. In particular, at about 0.8/0.2 distribution, the signal is very small and thus a false negative result will be measured. Even though acoustic wave technologies such as the TSM are not new, their adaptation to biological analysis has flourished mainly in the past decade. Most likely the next ten years will see an even larger contingent of researchers developing biosensors based on these platforms as affordability, access, sensitivity, and technical understanding increase. The most remarkable property of acoustic wave devices is their relative simplicity. The number of different devices adapted to work in the biological/medical environment is likely to expand in the future. It is therefore vital to establish a good understanding of the nature of the signals produced by acoustic wave devices when they are used for testing bacteria. The paradox of "negative mass" is a real threat to the interpretation of experimental results related to the detection of bacteria. Knowledge of the true nature of "negative mass" linked to the strength of bacteria attachment will contribute significantly to our understanding of the results of "weighing bacteria." We hope it may stimulate increased interest in the technology and motivate new experiments with a variety of microorganisms. The impact of these studies may extend beyond an appreciation of bacterial detection. One may now begin to conceive of strategies for the study and control of processes of bacterial settlement, bacterial colonization, biofilm formation, and bacterial infection in which bacterial attachment plays a role. Present trends in immunosensors. LaborPraxis 70-74 A comparative study of protein immobilization techniques for optical immunosensors Solvent effect on antibody antigen interaction Proton-decoupled 15N and 31P solid-state NMR investigations of the Pf3 coat protein in oriented phospholipid bilayers Rapid diagnosis of Treponema pallidum in serum using latex piezoelectric immunoassay Molecular biology of the cell Dual-mode acoustic wave biosensors microarrays Evaluation of antibody immobilization methods for piezoelectric biosensor application Piezoelectric flow injector analysis biosensor for the detection of Salmonella typhimurium Thickness shear mode (TSM) resonators used for biosensing Lytic phage as a specific and selective probe for detection of Staphylococcus aureus-A surface plasmon resonance spectroscopic study Modeling the responses of thickness-shear mode resonators under various loading conditions Determination of microorganisms with a quartz crystal microbalance sensor A rapid method for determination of Proteus vulgaris with a piezoelectric quartz crystal sensor coated with a thin liquid film Study of immunoglobulin G thin layers obtained by the Langmuir-Blodgett method: application to immunosensors Constrained modeling of spin-labeled major coat protein mutants from M13 bacteriophage in a phospholipid bilayer Piezoelectric immunosensors for urine specimens of Chlamydia trachomatis employing quartz crystal microbalance microgravimetric analyses Frequency response of quartz crystal shear-resonator during an adhesive, elastic contact in a surface forces apparatus Detection of Escherichia coli O157:H7 with langasite pure shear horizontal surface acoustic wave sensors Assessment of porous silicon substrate for well-characterized sensitive DNA chip implement Fc-receptor-mediated targeting of antibodybearing liposomes containing dideoxycytidine triphosphate to human monocyte/macrophages Measuring nanomechanical properties of a dynamic contact using an indenter probe and quartz crystal microbalance Low-level detection of a Bacillus anthracis simulant using love-wave biosensors on 36 YX LiTaO3 Piezoelectric quartz crystal biosensors Economic costs and trade impacts of microbial food-borne illness Langmuir-Blodgett films and nanotechnology Piezoelectric detection of ricin and affinity-purified goat anti-ricin antibody Quartz crystal microbalance detection of Vibrio cholerae O139 serotype Experimental use of a new surface acoustic wave sensor for the rapid identification of bacteria and yeasts Acoustic waves and the study of biochemical macromolecules and cells at the sensor-liquid interface Series quartz crystal sensor for remote bacteria population monitoring in raw milk via the Internet Construction of a novel peptide nucleic acid piezoelectric gene sensor microarray detection system Development of a bacteriophage displayed peptide library and biosensor Principles of protein-protein recognition Prediction of protein conformation Electrode systems for continuous monitoring in cardiovascular surgery A biosensor that uses ion-channel switches Tethered-bilayer lipid membranes as a support for membrane-active peptides Use of porous membranes modified with polyelectrolyte multilayers as substrates for protein arrays with low nonspecific adsorption Sequences of E. coli O157:H7 detected by a PCR-acoustic wave sensor combination On-line rapid detection of urease-producing bacteria with a novel bulk acoustic wave ammonia sensor Acoustic wave immunosensing of Escherichia coli in water Borderline applications of QCM-devices: synthetic antibodies for analytes in both nm-and um-dimensions Bioimprinted QCM sensors for virus detection-screening of plant sap Sensor strategies for microorganism detection-from physical principles to imprinting procedures A sensitive new method for the determination of adhesive bonding between a particle and a substrate Detection of two orchid viruses using quartz crystal microbalance (QCM) immunosensors Rapid prototyping of patterned functional nanostructures World Health Organization (1997) Foodborne diseases-possibly 350 times more frequent than reported A micromachined interface for airborne sample-toliquid transfer and its application in a biosensor system Self-assembled monolayers as the coating in a quartz piezoelectric crystal immunosensor to detect Salmonella in aqueous solution Ultrathin oligonucleotide layers for fluorescence based DNA-sensors Photolinker-polymer-mediated immobilization of monoclonal antibodies, F(ab')2 and F(ab') fragments Research on detection of type C2 staphylococcus enterotoxin in food with piezoelectric immunosensors Analysis of the accuracy and implications of simple methods for predicting the secondary structure of globular proteins Solid-supported membranes inside porous substrates and their use in biosensors Interfacial properties of biotin conjugate avidin complexes studied by acoustic wave sensor Biomolecular Sensors Selection of phage displayed peptides for the detection of 2,4,6-trinitrotoluene in seawater Phage-displayed peptides as biosensor reagents Engineering antibodies for biosensor technologies Acoustic wave sensors Filamentous bacteriophage contract into hollow spherical particles upon exposure to a chloroform-water interface Detection of staphylococcal enterotoxin B employing a piezoelectric crystal immunosensor Combining atomic force microscope and quartz crystal microbalance studies for cell detection Selective microorganism detection with cell surface imprinted polymers Rapid detection of Escherichia coli using a separated electrode piezoelectric crystal sensor Rapid diagnosis of M. tuberculosis using a piezoelectric immunosensor A new MSPQC system for rapid detection of pathogens in clinical samples A rapid method for determining Mycobacterium tuberculosis based on a bulk acoustic wave impedance biosensor A new antimicrobial susceptibility testing method of Escherichia coli against ampicillin by MSPQC The protein coats or "ghosts" of Escherichia coli phage T2. Preparation, assay, and some chemical properties Electron-opaque markers: a review Colloidal gold and its application in cell biology Protein-lipid interactions of bacteriophage M13 gene 9 minor coat protein Conformation and orientation of the gene 9 minor coat protein of bacteriophage M13 in phospholipid bilayers A comparison of protocols for the optimisation of detection of bacteria using a surface acoustic wave (SAW) biosensor De novo folding of membrane proteins: an exploration of the structure and NMR properties of the fd coat protein Determination of microbial populations with piezoelectric membranes Biosensors for detection of pathogenic bacteria Piezoelectric mass-sensing devices as biosensors-an alternative to optical biosensors Quartz crystal microbalance for bioanalytical applications Nanometer-thick conformal pore sealing of selfassembled mesoporous silica by plasma-assisted atomic layer deposition Thickness shear mode resonators ("mass sensitive devices") in bioanalysis Impedance characterization of a piezoelectric immunosensor part II: Salmonella typhimurium detection using magnetic enhancement Application of a flow-type antibody sensor to the detection of Escherichia coli in various foods Characteristics of a label-free piezoelectric immunosensor detecting Pseudomonas aeruginosa Piezoelectric sorption detector Preparation and length measurements of the total deoxyribonucleic acid (DNA) content of T2 bacteriophages Detection of viruses and bacteria with piezoelectric immunosensors A piezoelectric immunosensor for hepatitis viruses Process control for Ni/Au plating with QCM technology Quantitative analyses of Streptococcus mutans biofilms with quartz crystal microbalance, microjet impingement and confocal microscopy Quartz crystal microbalance immunosensors for environmental monitoring Detection of Salmonella typhimurium using phage based magnetostrictive sensor Pathogen detection: A perspective of traditional methods and biosensors A goat-anti-human IgG modified piezoimmunosensor for Staphylococcus aureus detection Advances in biosensors for detection of pathogens in food and water Piezoelectric crystal immunosensor for the detection of staphylococcal enterotoxin B Some problems of geodynamics. Cambridge: Cambridge Lucklum R (2005) Non-gravimetric contributions to QCR sensor response A nanoparticle amplification based quartz crystal microbalance DNA sensor for detection of Escherichia coli O157:H7 Characterization of quartz crystal microbalance with simultaneous mass and liquid loading Cell spreading on quartz crystal microbalance elicits positive frequency shifts indicative of viscosity changes Food-related illness and death in the United States A piezoelectric quartz crystal biosensor as a direct affinity sensor Spectroscopic and surface-analytical characterization of self-assembled layers on Au Microbalance-DNA probe method for the detection of specific bacteria in water A Love wave immunosensor for whole E. coli bacteria detection using an innovative two-step immobilisation approach A history of surface acoustic wave devices Piezoelectric immunosensor for the detection of Candida albicans microbes Quartz crystal microbalance for the detection of microgram quantities of human serum albumin: relationship between the frequency change and the mass of protein adsorbed Detection of the red tide-causing plankton Chattonella marina using a piezoelectric immunosensor Phage at the air-liquid interface for the fabrication of biosensors Phage as a molecular recognition element in biosensors immobilized by physical adsorption Synthesis of ordered mesoporous zirconium phosphate films by spin coating and vapor treatments Long-term, on-line monitoring of microbial biofilms using a quartz crystal microbalance Commercial quartz crystal microbalances-theory and applications Functional durability of a quartz crystal microbalance sensor for the rapid detection of Salmonella in liquids from poultry packaging Phage-coupled piezoelectric biodetector for Salmonella typhimurium Specific and selective biosensor for Salmonella and its detection in the environment Affinity-selected filamentous bacteriophage as a probe for acoustic wave biodetectors of Salmonella typhimurium Phage fusion proteins as bioselective receptors for piezoelectric sensors Effect of ionic strength on initial interactions of Escherichia coli with surfaces, studied on-line by a novel quartz crystal microbalance technique Development of bacteriophage-based bioluminescent bioreporters for monitoring of microbial pathogens Thiolated Salmonella antibody immobilization onto the gold surface of piezoelectric quartz crystal Operational characteristics of an antibody-immobilized QCM system detecting Salmonella spp Assembly of cadmium stearate and valinomycin molecules assists complexing of K+ in mixed Langmuir-Blodgett films Interaction of valinomycin and stearic acid in monolayers Condensing and expanding the effects of the odorants (+)-and (−)-carvone on phospholipid monolayers Interaction of valinomycin and stearic acid in monolayers at the air/water interface Rapid and sensitive biosensor for Salmonella Spreading of liposomes at the air/water interface Lipid-protein interactions in monolayers at the air-water interface Further studies on the spreading of biomembranes at the air/water interface. Structure, composition, enzymatic activities of human erythrocyte and sarcoplasmic reticulum membrane films Quartz crystal analytical sensors: the future of label-free, real-time diagnostics? Application of the quartz crystal microbalance to the monitoring of Staphylococcus epidermidis antigen-antibody agglutination A rapid, non-destructive method for the determination of Staphylococcus epidermidis adhesion to surfaces using quartz crystal resonant sensor technology Determination of boron by using a quartz crystal resonator coated with N-Methyl-D-glucamine-modified poly(epichlorohydrin) Phages from landscape libraries as substitute antibodies Biospecific peptide probes against Salmonella isolated from phage display libraries, and their diagnostic and drug delivery uses. US Patent Application Phage display for detection of biological threat agents Methods of forming monolayers of phage-derived products and uses thereof Application of multilayer films to molecular sensors: some examples of bioengineering at the molecular level Series Operation and Service Manual Piezoelectric immunosensor for Francisella tularensis detection using immunoglobulin M in a limiting dilution Development of a piezoelectric immunosensor for the detection of Salmonella typhimurium Development of a biosensor for E. coli based on a flexural plate wave (FPW) transducer Rapid detection of Escherichia coliform with a bulk acoustic wave sensor based on the gelation of Tachypleus amebocyte lysate Detection of molecular interactions by using a new peptide-displaying bacteriophage Dynamics of protein adsorption at the solid/liquid interface Protein adsorption at the solid/liquid interface Biomimetic protein immobilization using lipid bilayers On protein-lipid membrane interactions Multiple interactions in protein-membrane binding On waves propagating along the plane surface of an elastic solid Long-term monitoring of biofilm growth and disinfection using a quartz crystal microbalance and reflectance measurements Structural changes accompanying chloroform-induced contraction of the filamentous phage fd Recognition of cellspecific binding of phage display derived peptides using an acoustic wave sensor Peptide biosensor for recognition of cross-species cell surface markers Use of quartz vibrator for weighing thin films on a microbalance Bioluminescent biosensor device. US Patent Application Designing allosteric peptide ligands targeting a globular protein Rapid detection of Salmonella enteritidis by piezoelectric immunosensor Preparation of a piezoelectric immunosensor for the detection of Salmonella paratyphi A by immobilization of antibodies on electropolymerized films Piezoelectric quartz crystal sensors applied for bioanalytical assays and characterization of affinity interactions Need title of their article here Comparison of effective working electrode areas on planar and porous silicon substrates for cholesterol biosensor A piezoelectric interfacial phenomena biosensor Landscape phage probes for Salmonella typhimurium Comparison of the Spreeta®surface plasmon resonance sensor and a quartz crystal microbalance for detection of Escherichia coli heat-labile enterotoxin Protein adsorption on surfaces: dynamic contact angle (DCA) and quartz-crystal microbalance (QCM) measurements A piezoelectric biosensor for DNA hybridisation detection Development of immunochips for the detection of dengue viral antigens Serological determination of Helicobacter pylori infection using sandwiched and enzymatically amplified piezoelectric biosensor A self-assembled monolayer-based piezoelectric immunosensor for rapid detection of Escherichia coli O157:H7 A QCM immunosensor for Salmonella detection with simultaneous measurements of resonant frequency and motional resistance Piezoelectric quartz crystal based veterinary diagnosis for Salmonella enteritidis infection in chicken and egg Multilayer films containing immobilized nucleic acids. Their structure and possibilities in biosensor applications Monolayers of biomolecules for recognition and transduction in biosensors Solvent effects on amphotericin B monolayers Fluorescent virus probes for identification of bacteria. US Patent Application Artificial receptors in serologic tests for the early diagnosis of dengue virus infection Detection and analysis of the growth characteristics of Proteus vulgaris with a bulk acoustic wave ammonia sensor Liquid-phase piezoelectric and acoustic transmission studies of interfacial immunochemistry Thickness-shear-mode acoustic wave sensors in the liquid phase: a review Enzyme immobilization on planar and porous silicon substrates for biosensor applications Towards amperometric immunosensor devices The spreading of protein monolayers Quartz crystal biosensor for detection of the African Swine Fever disease Ultrasensitive quartz crystal microbalance sensors for detection of M13-Phages in liquids A quartz crystal microbalance method for rapid detection and differentiation of shiga toxins by applying a monoalkyl globobioside as the toxin ligand Development of a quartz crystal microbalance (QCM) immunosensor for the detection of Listeria monocytogenes Human antibodies with sub-nanomolar affinities isolated from a large non-immunized phage display library Comments about the effects of non-uniform mass loading on a quartz crystal microbalance Functional reconstitution of mammalian olfactory receptor Forster energy transfer in ultrathin polymer layers as a basis for biosensors Multichannel piezoelectric genesensor for the detection of human papilloma virus Direct piezoelectric coupling to surface elastic waves Immunosensor for the differentiation and detection of Salmonella species based on a quartz crystal microbalance Piezoelectric immunochip for the detection of dengue fever in viremia phase A direct immunoassay for Schistosoma japonium antibody (SjAb) in serum by piezoelectric body acoustic wave sensor An amplified mass piezoelectric immunosensor for Schistosoma japonicum Anodic alumina supported dual-layer microporous silica membranes Uncoated quartz resonator as a universal biosensor Bulk acoustic wave bacterial growth sensor applied to analysis of antimicrobial properties of tea Piezoelectric biosensor for the detection of Salmonella typhimurium Large-conductance cholesterol-amphotericin B channels in reconstituted lipid bilayers Amphotericin B channels in phospholipid membrane-coated nanoporous silicon surfaces: implications for photovoltaic driving of ions across membranes Piezoelectric crystal for sensing bacteria by immobilizing antibodies on divinylsulphone activated poly-m-aminophenol film Highly sensitive and renewable amperometric urea sensor based on self-assembled monolayer using porous silicon substrate Monitoring of Lactobacillus fermentation process by using ion chromatography with a series piezoelectric quartz crystal detector Effect of static magnetic field on growth of Escherichia coli and relative response model of series piezoelectric quartz crystal Rapidly determining E. coli and P. aeruginosa by an eight channels bulk acoustic wave impedance physical biosensor Detection of hepatitis B virus by piezoelectric biosensor Piezoelectric immunosensor for SARS-associated coronavirus in sputum Our work was supported by grants from Sigma Xi Grants-in-Aid of Research (10040088), DARPA (MDA972-00-1-0011), ARO/DARPA (DAAD 19-01-10454), NIH (R21 AI055645), USDA (99-34394-7546), and Aetos Technologies Inc. We are grateful to Dr. Bryan Chin for support and dialogue, and Oleg Pustovyy for technical assistance. The views expressed in this article are those of the authors, and do not reflect the official policy or position of the United States Air Force, Department of Defense, or the U.S. Government.