WO2002093145A1 - Thin-film bio-sensor, and method of operating same - Google Patents

Abstract

Apparatus (10) for determining the presence of a conditioning film as a precursor indicator to bio-contamination of fluid, includes: a substream (14) having an optically-reflective surface (16); a body (11) having a flow passage (18) for said fluid and having an optical passage (19), the optically-reflective substratum surface being arranged in said flow and optical passages, the body having two windows (20, 21) arranged perpendicularly to the axis of said optical passage; and means (22) for determining the presence of said conditioning film on said optically-reflective surface.

Description

THIN-FILM BIO-SENSOR, AND METHOD OF OPERATING SAME

Cross-Reference to Related Application Any U.S. corresponding patent application filed under the national stage of the Patent Cooperation Treaty claims the benefit of the earlier filing date of provisional U.S. Patent Application Serial No. 60/204,925, filed 17 May 2000.

Technical Field The present invention relates generally to devices for sensing the presence of a contaminant, and, more particularly, to an improved bio-sensor for sensing a bio- contaminating film that may be a precursor to bio-contamination.

Background Art Biological contamination of pure water systems threatens many bio-medical and industrial processes. It can cause human disease, equipment damage and reduced product yields. [Nivens, D. E., Chambers, J. Q., Anderson, T. R., and White, D. C, "Long-Term, On-Line Monitoring of Microbial Biofilms Using a Quartz Crystal Microbalance", Analytical Chemistry, 65:56-69 (1993).] Despite use of expensive and numerous water purification methods, the sustained absence of bacterial species in supply waters is often difficult to achieve. This failure to provide bacteria-free waters is acutely realized in bio-medical and micro-electronics manufacturing, where product defects can be caused from the presence of a single bacterium. [Mittelman, M. W., in Microbial Biofilms, Lappin-Scott, H. M., and Costerton, J. W., eds., Cambridge University Press, pp.133-147 (1995).] Despite advanced water purification technologies, relatively crude techniques, such as visual microscopy and cell culture assays, are used for bacterial detection. [Martyak, J. E., Carmody, J. C, and Husted, G. R., "Characterizing Biofilm Growth in Deionized Ultrapure Water Piping Systems", Microcontamination, 11 :39-44 (1993).] These techniques are labor-intensive and prevent real-time bacterial quantification, resulting in time delays and costly product defects before bio-contamination can be detected and the problem corrected. Adding to these detection deficiencies is the emphasis placed on quantifying plank- tonic bacteria, rather than monitoring the far greater number that are adhered to, and episodically released from, wetted surfaces. [Mittelman, M. W., Microbial Biofilms, supra.] A real-time micro-biocontamination monitoring device should be designed to detect and monitor not only the presence of an attached bacterial biofilm, but also a thinner conditioning film as well. Any substratum exposed to a biologically-active solution, whether it be the human vascular system, oral cavity, or a marine environment, will adsorb a proteinaceous "conditioning film" prior to any macroscopic fouling. [Baier, R. E., (1970), "Surface Properties Influencing Biological Adhesion", in: R. S. Manley, ed., Adhesion in Biological Systems, New York: Academic Press, pp.15-48 (1970).] Detection of such a precursor film on ultra-pure water- wetted surfaces should indicate the event of impending bio-contamination and allow the system's operator to take necessary actions against feed water contamination. Several real-time analytical techniques have been proposed to monitor the formation of biological molecules at liquid/solid interphases, but limited literature is available to demonstrate the sensitivity of these techniques in detecting changes in the thin conditioning film that forms prior to bacterial attachment in ultra-pure water systems. "Ultra-pure water" is a relative term used for purity standards set by particular industrial end-use requirements. The micro-electronics manufacturing industry is continuously striving for purer media to provide cleansing of surface processing residues and contaminants from memory circuits. As the size of micro-circuits continues to decrease, so too does the size of contaminants that will inevitably dam- age the circuit. Packing density, device speed, stability and power reduction are the strategic goals sought in developing a more functional computing device. [Seshan, K., Maloney, T. J., and Wu, K. J., "The Quality and Reliability of Intel's Quarter Micron Process" , Intel Technology Journal, Q3, pp.l-ll (1998).] Photolithography advances in the past decade have led to reduced line widths, and increased die sizes, resulting in an increase of the average number of transistors per micro-chip, or grid density, from 10⁶to 10⁸. [Intel Corporation, "Intel Museum Home Page", Internet Web Site: www.intel.com (1998).] Current 0.25 μm circuits have 0.45 μm junction depths, and 0.25 μm gate lengths, either of which can easily be compromised by a single bacterium that is 1 μm in length. In addition to their small size, modern microchips have multiple patterning levels, such that an attached bacterium at one pattern- ing level may micro-mask subsequent levels, resulting in a parasitic conduction path between levels. [Perera, A. H. and Satterfield, M. J., "Micromasking of Plasma Etching Due to Bacteria: A Yield Detractor for ULSI", IEEE Transactions on Semiconductor Manufacturing, 9:577-580 (1996).]

Current standards for semiconductor-grade ultra-pure water require supply waters to have less than one bacterium per liter. Supply waters must also be ionically pure, and have a resistivity of at least 18.2 Mohm/cm at 25 °C. Total organic carbon must be less than 1 part per billion. [Balazs Analytical Laboratory, "Ultra Pure Water Analysis Specifications", Internet Web Site: www.balazs.com (1998).] The problem of achieving such standards becomes evident when one considers that the EPA stan- dard for pre-treated drinking water is 500,000 bacteria per liter. [Environmental Protection Agency (EPA), Safe Drinking Water Act: 40 C.F.R. § 141.72 (b)(3)(i), Office of Ground Water and Drinking Water Internet Web Site: www.epa.gov/safe- water (1996).] In addition to ultra-purity requirements, micro-electronics manufacturing also requires significant volumes of ultra-pure water with consumption rates of more than 2000 gallons per minute. [Martyak, J. E., Carmody, J. C, and Husted, G. R, "Characterizing Biofilm Growth in Deionized Ultrapure Water Piping Systems", Microcontamination, 11 :39-44 (1993).] Recycling technology does exist, but is poorly implemented in the U.S. for fears of ultra-pure water system contamination which could lead to costly shut-downs. Similarly, pharmaceutical and medical device manufacturing industries must guarantee the sterility of their aqueous-based products. Microbial contamination of aqueous-based treatments, including respiratory care solutions, eye care solutions and dialysis waters have resulted in FDA recalls of these products. [Mittelman, M. W. and Geesey, G. G., "Biological Fouling of Industrial Water Systems: a Problem Solving Approach", SanDiego, California: Water Micro Associates (1987).] Microbial biofilm contamination of dental unit supply lines reportedly has led to oral infec- tions in immuno-compromised patients. [Whitehorse, R. L. S., Peters, E._s Lawsuit, J., and Liege, C, "Influence of Biofilms on Microbial Contamination in Dental Unit Water", Journal of Dentistry, 19:290-295 (1991); Martin, M., "The Significance of the Bacterial Contamination of Dental Unit Water Systems", British Dental Journal, 163:152- 154 (1987).] Corneal ulcers have been linked with bacterial contamination of externally-applied contact lens saline solutions. [Wilson, L. A., Schlitzes, R. L., Hearn, D. G., "Pseudomonas Corneal Ulcers Associated With Soft Contact-Lens Wear", American Journal of Ophthalmology, 92:546-554 (1981).] In such cases, human exposure to contaminating microorganisms has resulted in diminished patient care and increased risk of disease.

On-site production of ultra-pure water at semiconductor fabrication facilities is a complex and expensive process consisting of continuous filtration, ion exchange, ultraviolet sterilization, and degasifϊcation. Municipal feed waters are first filtered by carbon and diatomaceous earth beds, which mechanically retain many contami- nants. Reverse-osmosis membranes provide continued contaminant removal by utilizing high pressure differentials between permeate and concentrate membrane phases, enhancing the rate of fine particulate filtration. Ultraviolet (UN) radiation is used in two different wavelengths, 254 nm and 185 nm, to provide sterilization and oxidation, respectively, of remaining organic impurities. Oxidation and decomposi- tion of organics by UN- 185 nm results from the formation of hydroxyl and hydroper- oxyl free radicals, which act as active intermediate-reactant "oxidative" species. [Governal, R. A., Bonner, A., and Shadman, F., "Effect of Component Interactions on the Removal of Organic Inpurities in Ultrapure Water Systems", IEEE Transactions on Semiconductor Manufacturing, 4:298-303 (1991).] Ion exchange units and degassing membranes are required after such UV- 185 nm treatment to remove ions and carbon dioxide gas produced during the oxidation of organic constituents. After final degassing, waters are temporarily stored in a reservoir capped with nitrogen gas, preventing the reintroduction of viable life-sustaining gases. Despite research that suggests that biofilms form independently of flow [Martyak, J. E., Carmody, J. C, and Husted, G. R., "Characterizing Biofilm Growth in Deionized Ultrapure Water Piping Systems", Microcontamination, supra , residence times are minimized in all legs of the processing system, resulting in high interfacial shears that may reduce biofilm formation.

Due to the aggressive chemical nature of ultra-pure water [Mittelman, M. W., in: Microbial Biofilms, supra], all distribution lines and supply reservoirs are con- structed from special polymeric materials that are designed to provide pure exudate- free barriers. Halar® (a registered trademark of Ausimont USA, Inc., and a thermoplastic copolymer of ethylene and chlorotrifluoroetheylene) and Kynar® (a registered trademark of Elf Autochem, and a thermoplastic polyvinylidenefluoride) are the two most commonly-used materials. Given the high surface area-to-volume ratios found in ultra-pure water systems [Mittelman, M. W., (1995), in: Microbial Biofilms, supra], it should not be surprising that many areas are susceptible to colonization by micro-biological entities. In general, the rate and extent of biofilm formation is affected by the fluid environment, velocity, turbulence, temperature, and shear rate, and by characteristics of the substratum, such as chemical composition, roughness, and surface energy. [Meyer, A. E., Baier, R. E., and King, R. W., "Initial Fouling of Nontoxic Coatings in Fresh, Brackish, and Sea Water" , The Canadian Journal of Chemical Engineering, 66:55-62 (1988).] The same sequence of events has been reported for the formation of biofilms in all types of systems, regardless of environmental conditions. Initially, water is displaced from the wetted surface by a precursor film of depositing and adsorbing bio-exopolymers, followed by contact, attachment and proliferation of pioneering cells. [Goupil, D. W., DePalma, V. A., and Baier, R. E., "Physical/Chemical Characteristics of the Macromolecular Conditioning Film in Biological Fouling", Proceedings, 5^th International Congress on Marine Corrosion and Fouling, Spain, pp.401-410 (1980); Baier, R.E., "Surface Properties Influencing Biological Adhesion", in: R. S. Manley, ed., Adhesion in Biological Systems, New York: Academic Press, pp. 15-48 (1977); Baier, R. E., "On the Formation Of Biological Films", SwedDent , 1 -.261-211 (1977).] Subsequent retention of the adsorbed biological molecules has been correlated with the substratum's surface free energy, or, more precisely, critical surface tension as shown in Fig. 1. [Gucinski, H., "Correlation of Biophysical Surface Characteristics With Hydrodynamic Properties of Adhesive Biofilms", Ph.D. Dissertation, State University of New York at Buffalo (1984).] According to Mittelman [Mittelman, M. W., (1995), in: Microbial Biofilms, supra], adhesion to surfaces offers bacteria many advantages: (1) organic nutrients accumulate at surfaces because this is more favorable thermodynamically than to stay in solution or suspension; (2) attached sessile bacteria produce extra-cellular polymeric substances which further concentrate and trap nutrients; (3) non- viable surface- attached cells can be a nutrient source for neighboring bacteria; and (4) a multi- layered cellular film shields interior bacteria from biocidal entities, thereby allowing repopulation of the system after conventional sterilization. As biofilms mature, some attached basal cells become nutrient-limited and eventually die. This results in sloughing episodes that release many bacteria, as well as large amounts of extra-cellular polymers into the water phase. [Martyak, J. E., Carmody, J. C, and Husted, G. R., "Characterizing Biofilm Growth in Deionized Ultrapure Water Piping Systems", Microcontamination, 11 :39-44 (1993).] In ultra- pure water systems, the sloughing event can have tremendous impacts on potential product contamination.

Not all ultra-pure water system components are constructed from Halar® or Kynar® fluoropolymers. Many different materials are used in constructing various filtration membranes, ion exchange resin units, and degassing membranes used to purify these waters. Like the polymer distribution lines and storage reservoirs, these units are susceptible to bacterial attachment and biofilm formation. [Mittelman, M. W., in: Microbial Biofilms, supra; Mittelman, M. W. and Geesey, G. G., "Biological Fouling of Industrial Water Systems: a Problem Solving Approach", San Diego, California: Water Micro Associates (1987).] Current ultra-pure water treatment technologies incorporate biocidal agents, such as ultraviolet irradiation, chlorination, and hydrogen peroxide exposure to reduce the number of viable biofilm cells. [Mittelman, M. W., in: Microbial Biofilms, supra.] Despite these measures, many bacteria existing in biofilms are protected from the biocide effects by their extracellular slime matrices, and often survive the treatment allowing for repopulation of the system. [Mittelman, M. W., in: Microbial Biofilms, supra.] Once formed, bacterial biofilms in pure water systems can cause continued system contamination, leading to costly product defects. [Balazs Analytical Laboratory, "Ultra Pure Water Analysis Specifications", supra.]

Unfortunately, micro-biological contamination of ultra-pure water systems is often detected too late by excessive back pressures in reverse-osmosis membranes and filters due to the physical obstruction and degradation of membrane surfaces. [Mittelman, M. W., in: Microbial Biofilms, supra.] Preventive maintenance considerations, such as biofilm monitoring, are required to better forecast and predict fouling events. [Martyak, J. E., Carmody, J. C, and Husted, G. R., (1993), "Characterizing Biofilm Growth in Deionized Ultrapure Water Piping Systems", Microcontamination, supra.] Cell culture assays and light microscopy techniques are time-consuming, only estimate planktonic bacterial levels, and underestimate the potential for system biocontamination and product defect by the inherent inability to predict sloughing events in real-time. By monitoring the events of conditioning film adsorption, which precedes any bacterial attachment, a true indication of system cleanliness can be determined. Detection of such a film on ultra-pure water- wetted surfaces should indicate the event of impending bio-contamination and allow the system's operator to take necessary actions against supply water contamination.

Historically, flow cell sensors for biological substances have contained test substrata that must be removed for comprehensive analytical analysis, adding possi- ble artifact contamination and preventing in-situ real-time monitoring. [Nivens, D. E., Chambers, J. Q., Anderson, T. R., and White, D. C, "Long-Term, On-Line Monitoring of Microbial Biofilms Using a Quartz Crystal Microbalance", Analytical Chemistry, 65:56-69 (1993).] Despite these drawbacks, flow cells with removable substrata have been the only devices with sufficient sensitivity to monitor the events of conditioning film adsorption and subsequent cellular adhesion. Such devices were developed and used to monitor adsorption events in many diverse biological media, providing essential information about the dynamics of bio-fouling of engineering materials. [King, R. W., Meyer, A. E., Ziegler, R. C, and Baier, R. E.,"New Flow Cell Technology for Assessing Primary Biofouling in Oceanic Heat Exchangers", Proceedings of the 8^th Ocean Energy Conference, (June 1981).]

Laboratory modeling systems have been used to study biofilm mass, thick- ness, microbial activity, and effects on transport properties. These studies utilized a variety of techniques to quantify biofilms, such as the rotating annular reactor (which measures the viscous drag associated with the presence of a biofilm), and the test heat exchanger (which measures the change in heat transfer associated with the presence of a biofilm). Both techniques have reported biofilm thickness sensitivities of 285 μm with precisions of ± 256 μm and mass sensitivities of 1.1 g/m² (± 0.1 g/m²). [Characklis, W. G. and Marshall, K. C, Biofilms, John Wiley & Sons, Inc., pp. 55-89 (1990).]

A novel detection method for real-time in-situ biofilm monitoring that does not require substratum removal for analysis, providing continuous feed back on the kinetics of biofouling is desirable. Various technologies have been proposed.

Resonant wave modulation has been used to monitor biofilm formation in highly-controlled bacteria-inoculated pure water systems. [Nivens, D. E., Chambers, J. Q., Anderson, T. R., and White, D. C, "Long-Term, On-Line Monitoring of Micro- bial Biofilms Using a Quartz Crystal Microbalance", Analytical Chemistry, 65 : 56-69 (1993).] This technique was used to monitor the adsorption kinetics of hemoglobin to gold and methyl-terminated surfaces. [Hook, F., Rodahl, M., Kasemo, B., and Brzezinski, P., "Structural Changes in Hemoglobin During Adsorption to Solid Surfaces: Effects of pH, Ionic Strength, and Ligand Binding", Proceedings of the National Academy of Sciences, 95:12271-12276 (1998).] This technique uses the piezoelectric effect, measuring the change in frequency of a resonating quartz crystal associated with the addition of deposited mass. As biological entities deposit and adsorb over time, the mass of the resonating sensor increases, causing damping and reduced oscillations. The frequency is continuously monitored, recording changes that can be associated with apparent film thicknesses. Unfortunately, the resonant frequency is also sensitive to fluctuations in water pressure, temperature, viscosity and density, which lead to variable results. Detection thresholds of 3 x 10⁵ cells/cm² have been reported. [Nivens, D. E., Chambers, J. Q., Anderson, T. R., and White, D. C, "Long-Term, On-Line Monitoring of Microbial Biofilms Using a Quartz Crystal Microbalance", Analytical Chemistry, 65:56-69 (1993).] Hook and colleagues did not report a detection limit for the hemoglobin experiments referenced above. Fourier transform infrared spectrometry (FTIR) has been proposed to monitor the events of biofilm formation in pure water systems. This technique, when operated using attenuated total-reflection optics with infrared-transparent germanium or zinc selenide internal-reflection elements, has the ability to detect and monitor bacterial biofilms in-situ. [Mittelman, M. W., in: Microbial Biofilms, supra; Nichols, P. D., Henson, J. M., Guckert, J.B., Nivens, D. E., and White, D. C, (1985), "Transform-Infrared Spectroscopic Methods for Microbial Ecology: Analysis of Bacteria, Bacteria-polymer Mixtures, and Biofilms", Journal of Microbiological Methods, 4:79-94 (1985).] In theory, chemical fingerprints of adsorbed biofilm species, such as protein amide groups and carbohydrates, can be identified, providing valuable information about the type of biomass present. A detection threshold of 5 x 10⁵ cells/cm² has been reported for in-situ flow cell configurations. [Mittelman, M. W., in: Microbial Biofilms, supra; Nivens, D. E., Chambers, J. Q., Anderson, T. R., Tunlid, A., Smit, J., and White, D. C, "Monitoring Microbial Adhesion and Biofilm Formation by Attenuated Total -reflection/Fourier Transform Infrared Spectroscopy", Journal of Microbiological Methods, 17:199-213 (1993).]

Optical interference devices also have been proposed for ultra-pure water bio- contaminant monitoring. [Sjogren, J., Ph.D., Research Engineer, Center for Microcontamination Control, University of Arizona, Personal Communication (1999).] These devices utilize changes in reflected light intensity to determine the presence of an adsorbed film. When a light source of known wavelength passes through a weakly-reflecting interference film, the beam is reflected, in part, at each of the interfaces. The differing optical path lengths result in constructive and destructive interference of the reflected beam which, when measured as light intensity per specific wavelength, corresponds to a previously-calibrated film thickness. It has been reported that changes in interference intensities correspond to changes in organic film thickness. Many types of reflecting interference films have been used to enhance sensitivity, including combinations of silicon-doped tantalum and boron. These techniques have been demonstrated to monitor 50 A changes in receptor ligand films 7000 A thick. . [Piehler, J., Brecht, A., and Gauglitz, G, "Affinity Detection of Low Molecular Weight Analytes", Analytical Chemistry, 68:139-142 (1996); and Lin, N. S.-Y., Motesharei, K., Dancil, K.-P. S., Sailor, M. J., and Ghadiri, M. R., "A Porous Silicon-Based Optical Interferometric Biosensor", Science, 278:840-843 (1997).] Changes in thicker 120,000 A antibody-type films have also been reported. [Lin, N. S.-Y., Motesharei, K., Dancil, K.-P. S., Sailor, M. J., and Ghadiri, M. R., "A Porous Silicon-Based Optical Interferometric Biosensor", Science, 278:840-843 (1997).]

Null ellipsometry measurements conducted on dry highly-reflective substrata, have been reported to have a film thickness detection threshold of 1 A. [Spanier, R. F., "Ellipsometry: A Century Old New Technique", Industrial Research (1975).] MAIR-IR spectrometry has been demonstrated to detect and identify layers as thin as 10 A, and to monitor changes in the composition, as well as the configuration within such layers. [Baier, R. E. and Meyer, A. E., Surface Analysis, in: Handbook of Biomaterials Evaluation, Scientific, Technical, and Clinical Testing of Implant Materials, A. F. von Recum, ed, Macmillan Publishing Company, pp. 97-108 (1986).]

In-situ wet-cell ellipsometry was selected to monitor its sensitivity to events of conditioning film formation in ultra-pure water systems. This technique measures phase and amplitude changes of an elliptically-polarized beam upon reflecting and refracting from a substratum's surface, as modified by the addition of a thin film. Real-time wet-cell ellipsometry has been used extensively by researchers in Europe and Scandinavia to study the kinetics of biofilm formation from saliva. Earlier work demonstrated the utility of ellipsometry for measuring absorption kinetics of proteins and their resulting optical properties. [Vassilakos, N., Arnebrant, T., and Glantz, P- O., "An in Vitro Study of Salivary Film Formation at Solid/liquid Interfaces", Scan- dinavian Journal of Dental Research, 101:133-137 (1993).] Later work demonstrated that ellipsometry could be used to monitor complex interactions of multilay- ered biological molecules, namely protein-lipid complexes, with detection thresholds of 50 A. [Cuypers, P. A., Corsel, J. W., Janssen, M. P., Kop, J. M. M., Hermens, W. T., and Hemker, H. C, "The Adsorption of Prothrombin to Phosphatidylserine Multilayers Quantified by Ellipsometry, The Journal of Biological Chemistry, 258:2426-2431 (1983).] A fused quartz cell, with specially-positioned optical sur- faces, is most commonly used for these wet-cell studies. Despite annealing treatments following cell fabrication, these cells are susceptible to lens strains induced by liquid temperature fluctuations encountered during experimentation, as well as cell cleaning. [Arnebrant, T., Associate Professor and Section Manager, Ytkemiska Institutet, Institute for Surface Chemistry, Sweden, Personal Communication (1999).] A temperature-controlled water bath must be incorporated into the flow circuit, reducing temperature variations. In the study of salivary protein films greater than 100 A thick, the quartz cell's use is suitable, as minor errors caused by lens stresses and subsequent beam perturbation have a reduced effect on final results for these thicker films. Additionally, these used changes in ellipsometer-recorded values to calculate estimated film masses, which were reported to be less sensitive to errors than calculated film thicknesses. [Cuypers, P. A., Corsel, J. W., Janssen, M. P., Kop, J. M. M., Hermens, W. T., and Hemker, H. C, "The Adsorption of Prothrombin to Phosphatidylserine Multilayers Quantified by Ellipsometry, The Journal ofBiologi- cal Chemistry, 258:2426-2431 (1983).]

The present invention examined the question of whether, by designing a liquid cell (bio-sensor) with minimized lens-induced strain, one would be able to accurately and precisely monitor adsorption events for films with thicknesses less than 100 A with a detection threshold and resolution better than the 50 A values hitherto attained.

There is a lack of available literature describing the use of any of the above mentioned real-time techniques or devices for on-line ultra-pure water system monitoring. The following description of device placement and "fit" is, therefore, theorized. To determine contaminant levels and treatment efficiencies, it would seem that monitoring devices should be distributed throughout the ultra-pure water system, with additional units placed both before and after filters, reverse-osmosis membranes, ion- exchange units, and ultraviolet lamp sources.

Resonant wave modulation devices require an external cell which houses, actuates, and monitors the sensor. [Q-sense AB, (2000), Internet Address: www.q- sense . com] , so placement would be limited to external applications . A special supply tube is required to take waters off-line and into the resonating sensor's cell. Waters exiting from the cell would simply be drained off into waste, or reintroduced via another flow tube into the supply line downstream. Pressure, flow, and temperature regulators may be necessary to reduce input fluctuations. Q-sense, a Swedish company, currently manufactures and markets a quartz crystal microbalance system for (US) $67,000.00. A computer-interfaced multi-channel input system, which has the ability to monitor several resonating cells at once would be the most economical.

Fourier Transform Infrared Spectrometry (FTIR) sensors require a special union incorporating the infrared transparent "window" into the in-line water supply. Specially-positioned fiber-optic couplers can be mounted to miniaturized attenuated total-reflection optics at the ends of each window, allowing infrared energy to be received from, and transmitted back to, a single infrared spectrometer. Many "windows" could be monitored through coupled fiber-optic bundles. A high-speed switching device momentarily sends and receives infrared transmissions to and from each optically-coupled window. When the transmission percentage deviated from a predetermined amount, an alert would be posted for that particular window /location. The current cost of an "unmodified" FTIR system is (US) $20,000.00 to (US) $50,000.00 [Perkin-Elmer Corporation (2000)]. Special fiber-optic couplers, lines and switches, as well as infrared-transparent windows, attenuated total-reflection optics and unions, would be extra. Optical interference devices could use an approach similar to FTIR in-line monitoring. Due to the small sensor size available [Piehler, J., Brecht, A., and Gauglitz, G. , "Affinity Detection of Low Molecular Weight Analytes" , Analytical Chemistry, 68:139-142 (1996)], sensor surfaces could easily be placed in-line throughout the ultra-pure water system. A high-speed switching device could monitor wave- length shifts associated with possible deposits for each individual sensor. The estimated cost for each sensor is approximately (US) $4,500.00. [Sjogren, J., Ph.D., Research Engineer, Center for Microcontamination Control, University of Arizona, Personal Communication (1999).] The power source, high-speed switching device, and optic lines are additional. Wet ellipsometry bio-sensors could be incorporated into in-line distribution tubes by the addition of an external supply tube, distributing waters off-line and into the flow cell. Waters exiting from the cell would be drained off as waste, or reintro- duced via another flow tube into the supply line located downstream. A simplified ellipsometer-type optical system comprising a laser source, polarizing lenses, and a photodiode detector could be mounted to the flow cell body. A pre-set deviation in phase and amplitude of the elliptically-polarized beam caused by the addition of material to the sample surface would result in an alarm. Based on current materials and fabrication requirements, the approximate cost of the combined optical cell and mini-ellipsometer nears (US) $5,000.00.

Disclosure of the Invention The objective of this invention was to design, fabricate and evaluate a biosensor capable of detecting and monitoring the formation and retention of organic films as they deposit from trace amounts in ultra-pure, as well as distilled water. To determine the bio-sensor's accuracy and precision, calibration measurements were made using adsorbed and transferred mono-molecular films of reference materials. Real-time in-situ measurements were completed to determine the effectiveness of this technique for monitoring adsorption events in distilled, ultra-pure, and non-pure water supplies. The goal was to develop an advanced in-situ thin-film bio-sensor that could both accurately and precisely monitor the spontaneous adsorption of conditioning film constituents. With parenthetical reference to the corresponding parts, portions or surfaces of the disclosed embodiment, merely for purposes of illustration and not by way of limitation, the present invention provides apparatus (20) for determining the presence of a conditioning film as a precursor indicator to bio-contamination of a fluid (i.e., air or gas), comprising: a substratum (14) having an optically-reflective surface (16); a body (11) having a flow passage (18) for the fluid and having an optical passage (19), the optically-reflective substratum surface being arranged in the flow and optical passages, the body having two windows (20, 21) arranged perpendicularly to the axis of the optical passage; and means (22) for determining the presence of the conditioning film on the optically-reflective surface. In the preferred embodiment, the substratum is formed of a material, such as germanium or silicon, that is transparent to infrared energy, and the apparatus further includes a substratum holder. The means may include a helium-neon laser light source adapted to provide a collimated light beam. The windows are preferably formed of fused silica or quartz, and are adhesively secured to the body by means of an adhesive that includes barium and sulfur.

In use, the apparatus performs the method of determining the presence of a conditioning film as a precursor indicator to bio- , micro- or nano-contamination, of a fluid. The method has the ability to detect in real time the smallest quantities of any contaminant deposited from the fluid phase. The device has a sensitivity of 0.1 micrograms per square centimeter, or 100 nanograms per square centimeter, these being an indicator of the monolayer range of the detector. The improved method comprises the steps of: providing a substratum having an optically-reflective surface; providing a body having a flow passage for the fluid and having an optical passage, the optically-reflecting surface being arranged in the flow and optical pas- sages, the body having two windows arranged perpendicular to the axis of the optical passage; causing a beam of light to enter the body along the optical passage and to reflect and refract from the optically-reflective surface; and measuring parameters of the entering and exiting light beams; thereby to indicate the presence of the conditioning film on the optically-reflective surface. The entering light beam may be elliptically polarized.

Accordingly, the general object of the present invention is to provide an improved apparatus for determining the presence of a conditioning film as aprecursor indicator to bio-contamination of a fluid (i.e., a liquid or a gas).

Another object is to provide an improved method of determining the presence of a conditioning film as a precursor to bio-contamination of a fluid.

These and other objects and advantages will become apparent from the foregoing and ongoing written specification, the drawings, and the appended claims.

Brief Description of the Drawings Fig. 1 is a plot of biological fouling as a function of surface energy. Fig. 2 is an exploded perspective view of the inventive thin-film bio-sensor. Fig. 2 A is a schematic of the improved bio-sensor.

Fig. 3 is a view of the null ellipsometer components.

Fig. 4 is a MAIR-IR spectroscopy calibration of beta-lactoglobulin on Ge.

Fig. 5 is a Zisman plot for clean Halar® block stock. Fig. 6 is a Zisman plot for RFGPT Halar® block stock

Fig. 7 is a Zisman plot for PTFE sheathed silicone O-ring.

Fig. 8 is a MAIR infrared spectrum for Halar® fluoropolymer.

Fig. 9 is a MAIR infrared spectrum for Halar® fluoropolymer residue.

Fig. 10 is an EDX-ray spectrum for Halar® fluoropolymer. Fig. 11 is a MAIR infrared spectrum for methacrylate resin.

Fig. 12 is an EDX-ray spectrum for methacrylate resin.

Fig. 13 is a MAIR infrared baseline spectrum for germanium IRE # 526 using the Perkin-Elmer substratum locator.

Fig. 14 is a MAIR infrared baseline spectrum for germanium IRE # 526 using the bio-sensor.

Fig. 15 is a MAIR infrared spectrum for adsorbed Fibrinogen Film 1 using the Perkin-Elmer substratum locator.

Fig. 16 is a MAIR infrared spectrum for adsorbed Fibrinogen Film 1 using the bio-sensor. Fig. 17 is a MAIR infrared spectrum for transferred Octadecanol Film 1 using the Perkin-Elmer substratum locator.

Fig. 18 is a MAIR infrared spectrum for transferred Octadecanol Film 1 using the bio-sensor.

Fig. 19 is a MAIR infrared spectrum for transferred Octadecanoic Acid Film 1 using the Perkin-Elmer substratum locator. '

Fig. 20 is a MAIR infrared spectrum for transferred Octadecanoic Acid Film 1 using the bio-sensor.

Fig. 21 is a MAIR infrared spectrum for transferred Octadecylamine Film 1 using the Perkin-Elmer substratum locator. Fig. 22 is a MAIR infrared spectrum for transferred Octadecylamine Film 1 using the bio-sensor. Fig. 23 is a MAIR infrared spectrum for transferred Octadecylamin Film 2 using the bio-sensor filled with distilled water.

Fig. 24 is a MAIR infrared spectrum for distilled water supply residue.

Fig. 25 is a MAIR infrared spectrum for ultra-pure water supply residue. Fig. 26 is a MAIR infrared spectrum for aquarium water supply residue.

Fig.27 is a MAIR infrared spectrum for 24-hour distilled water, real-time run 1 A, using the bio-sensor.

Fig.28 is a MAIR infrared spectrum for 24-hour distilled water, real-time run IB using the bio-sensor. Fig.29 is a MAIR infrared spectrum for 1 -hour aquarium water, real-time run

5 A using the bio-sensor.

Fig.30 is a MAIR infrared spectrum for 1 -hour aquarium water, real-time run 5B using the bio-sensor.

Fig.31 is a plot of 24-hour distilled water, real-time run 1 A, ellipsometry film thickness values.

Fig.32 is aplot of 24-hour distilled water, real-time run IB, ellipsometry film thickness values.

Fig.33 is a plot of 88-hour distilled water, real-time run 2 A, ellipsometry film thickness values. Fig. 34 is a plot of 89-hour distilled water, real-time run 2B,. ellipsometry film thickness values.

Fig. 35 is a plot of 24-hour ultra-pure water, real-time run 3 A, ellipsometry film thickness values.

Fig. 36 is a plot of 24-hour ultra-pure water, real-time run 3B, ellipsometry film thickness values.

Fig. 37 is a plot of 94-hour ultra-pure water, real-time run 4 A, ellipsometry film thickness values.

Fig. 38 is a plot of 90-hour ultra-pure water, real-time run 4B, ellipsometry film thickness values. Fig.39 is a plot of 1 -hour aquarium water, real-time run 5 A, ellipsometry film thickness values. Fig.40 is a plot of 1 -hour aquarium water, real-time run 5B, ellipsometry film thickness values.

Fig.41 is a plot of 2-hour aquarium water, real-time run 6A, ellipsometry film thickness values. Fig.42 is a plot of 2-hour aquarium water, real-time run 6B, ellipsometry film thickness values.

Fig.43 is an SEM micrograph and EDX-ray spectrum for pre-exposed germanium substratum flow surface.

Fig. 44 is an SEM micrograph and EDX-ray spectrum for pre-exposed stain- less steel substratum flow surface.

Fig. 45 is an SEM micrograph and EDX-ray spectrum of 24-hour distilled water, real-time run 1 A, germanium substratum flow surface.

Fig. 46 is an SEM micrograph and EDX-ray spectrum of 24-hour distilled water, real-time run IB, germanium oxide surface deposits. Fig. 47 is an SEM micrograph and EDX-ray spectrum of 24-hour distilled water, real-time run IB, germanium oxide deposits, low mag.

Fig. 48 is an SEM micrograph and EDX-ray spectrum of 88 -hour distilled water, real-time run 2A, stainless steel substratum flow.

Fig. 49 is an SEM micrograph and EDX-ray spectrum of 24-hour ultra-pure, water, real-time run 2A, stainless steel substratum flow.

Fig. 50 is an SEM micrograph and EDX-ray spectrum of 24-hour ultra-pure water, real-time run 4A, stainless steel substratum flow surface.

Fig. 51 is a SEM micrograph and EDX-ray spectrum of 1-hour aquarium water, real-time run 5 A, germanium substratum flow surface. Fig. 52 is an SEM micrograph and EDX-ray spectrum of 2-hour aquarium water, real-time run 6A, stainless steel substratum flow surface.

Description of the Preferred Embodiments At the outset, it should be clearly understood that like reference numerals are intended to identify the same structural elements, portions or surfaces consistently throughout the several drawing figures, as such elements, portions or surfaces may be further described or explained by the entire written specification, of which this detailed description is an integral part. Unless otherwise indicated, the drawings are intended to be read (e.g., cross-hatching, arrangement of parts, proportion, degree, etc.) together with the specification, and are to be considered a portion of the entire written description of this invention. As used in the following description, the terms "horizontal", "vertical", "left", "right", "up" and "down", as well as adjectival and adverbial derivatives thereof (e.g., "horizontally", "rightwardly", "upwardly", etc.), simply refer to the orientation of the illustrated structure as the particular drawing figure faces the reader. Similarly, the terms "inwardly" and "outwardly" generally refer to the orientation of a surface relative to its axis of elongation, or axis of rotation, as appropriate.

The following sections describe the design of, as well as the materials and techniques used to fabricate and evaluate, the improved thin-film bio-sensor for use with any film and any fluid.. Bio-Sensor Design and Fabrication

Referring now to the drawings, and, more particularly, to Fig. 2 thereof, the improved bio-sensor, generally indicated at 10, is designed to monitor the formation of biological films (not-shown) as they deposit in ultra-pure and distilled waters. The improved bio-sensor combines two non-destructive surface-specific analytical tech- niques: "null ellipsometry" and "internal-reflection infrared spectrometry".

Referring now to Fig. 2 A, the improved bio-sensor 10 is shown as broadly including an uppermost optical cell or body 11 , a substratum holder or locator 12, a substratum 14 mounted on holder 12, and a lowermost mounting plate 13 adapted to be positioned on top of an ellipsometer 15. The substratum is shown as being an elongated member having a rectangular transverse cross-section, and has an uppermost polished or optically-reflective surface 16. The uppermost cell or body 11 has a flow passage 18 for a fluid-to-be-tested, and has a V-shaped optical passage 19. The optical passage has two fused silica or quartz windows that are adhesively secured to the body in positions that are perpendicular to the axis of the optical passage. The polished surface on the substratum, the fluid passage and the optical path are exposed to one another in the vicinity of the optically-reflective surface. The device further includes means, such as an elhpsometer, for determining the presence of a conditioning film on the optically-reflective surface as a function of a sensed parameter. Thus, light from a helium neon laser light source is passed along the optical path, and reflects off the surface of 16 to the receiver portion of the ellipsometer. By comparing a change in parameters of the reflected light with the transmitted light, one can infer the presence of a conditioning film on the optical surface.

Physical dimensions of all three components were optimized to accommodate an internal-reflection element (50 x 20 mm x 1 mm) [Harrick Scientific Corp., Ossining, NY], a multiple-attenuated internal -reflection infrared spectrometer with an attenuated total-reflection optical platform [Perkin-Elmer Model 1420 Ratio Recording Infrared Spectrophotometer, Norwalk, CT], and a Rudolph thin-film null ellipsometer stage. [Type 43702-200E, Rudolph Research Corp., Flanders, NJ.]. Other manufacturers of spectrophotometers and ellipsometers may have different spacing requirements which would require modification to the bio-sensor design. This design accommodated a Perkin-Elmer optical platform requiring a sample holder footprint with maximum dimensions of 2.25" (5.7 cm) in width, 1.5" (3.8 cm) in depth, and unrestricted height. Dimensions for positioning the internal-reflection element were taken from the purchased substrata locating mount. Spacing limits for the ellipsometer stage were less than 3" (7.6 cm) in width, 3" (7.6 cm) in length, and 1.5" (3.8 cm) in height.

The optical cell was designed to provide a rigid dimensionally-stable inert- liquid flow reservoir with accurately-positioned optical windows and a precision mounting system. In order to obtain ellipsometer measurements of films in real-time on test substrata underwater, two fused silica optical windows were used to provide an interference-free laser light beam pathway between the external ambient air and internal cell liquid phase. A 5 mW helium neon laser of wavelength 632.8 nm [Mel- les Griot Laser Division, Model 3222H-PC, Carlsbad, CA], was used as the light beam source for all ellipsometry experiments, as lasers have been shown to be more stable and have higher signal outputs, improving the signal-to-noise ratio over previously-used mono-chromatic light sources. [Cuypers, P. A., Hermens, W. Th., and Hemker, H. C, "Ellipsometry as a Tool to Study Protein Films at Liquid-Solid Interfaces", Analytical Biochemistry, 84:56-67 (1978).] It has been shown that for measuring films with thicknesses less than 700 A, maximum sensitivity can be obtained by setting the ellipsometer to a 70° angle of incidence. [McCrackin, F. L., Passaglia, E. , Stromberg, R. R. , and Steinberg, H. L . , "Measurement of the Thickness and Refractive Index of Very Thin Films and the Optical Properties of Surfaces by Ellipsometry", Journal of Research of the National Bureau ofStandards-A. Physics and Chemistry, 67A:363-377 (1963).] With the ellipsometer set at a 70° angle of incidence, both optical windows were inclined at a 70° angle of incidence with re- spect to the substratum base, so that the laser would pass through each window at normal incidence. To maintain optical alignment, the windows were located on machined shoulders positioned centrally on opposite outside faces of the cell. This window positioning allowed 1 cm of ellipsometer X-axis travel, ensuring access to multiple sample surface areas for analysis. The nominal spot size of the laser beam as it reflects and refracts from the substratum's surface, is about 1 mm². A total sample area of 1 cm² can be analyzed by multiple measurements. If the bio-sensor is moved beyond the 1 cm X-axis travel limit in either direction, the laser beam reflects off the bottom or top of the optical cell's window flange. Ellipsometry stage Y-axis travel was limited by the width of the window port, 0.25" (0.6 cm). For all experiments performed, the variable Y-axis was kept static, positioning the laser so that its beam was centered in the window port, which minimized possible reflections and interference from the cell body. Ellipsometer stage Z-axis travel was adjusted when maximizing laser beam throughput, and remained constant as long as the same substratum was used. Variation in substrata thickness required Z-axis adjustment. Materials for constructing the optical cell's structure included 6061 -T6 aluminum alloy and Halar® stock, which provided a rigid scaffold for window mounting, preventing movement or translation of stresses from upsetting the state of beam polarization. Halar® also was chosen for having superior inertness and purity in corrosive environments. [MatWeb, "The Online Materials Information Resource", Internet Web Site: www.matweb.com (2000).] Because of the need to transport the bio-sensor between the ellipsometer and infrared spectrophotometer, an alignment system was designed using a pair of precision-ground guide dowels. Guide dowels allowed the optical cell to be mounted and dismounted from the substratum locator without positioning error. Four spring-loaded bolts retained the substratum between the optical cell base and the substratum locator. Supply and return tubes were posi- tioned at the ends of the cell to assure laminar flow across the substratum surface at flow rates of 1 ml/min.

The substratum locator with MAIR-IR spectrophotometer and ellipsometer positioning guides was designed to provide a rigid substratum positioning platform and to maintain positioning guide tolerance, while preventing optical alignment errors. MAIR-IR spectra could be obtained when the optical cell and substratum locator were assembled. Positioning receptacles were machined into the bottom of the substratum locator to provide proper alignment when mounted to the MAIR-IR spectrophotometer optical stage. The substratum locator also provided the bolting threads needed for attachment of the optical cell. The ellipsometer mounting plate was designed to provide a precise mounting receptacle for the assembled optical cell and substratum locator, eliminating re-alignment when the optical cell is mounted.

Details on how the bio-sensor was fabricated are listed below. Bio-sensor Fabrication

The liquid flow cell bio-sensor was fabricated using standard machine tool fabrication techniques. The bio-sensor (Fig. 2) comprises three main structural components: an optical cell, substratum locator for multiple-attenuated internal- reflection infrared (MAIR-IR) spectrophotometer and ellipsometer positioning guides, and an ellipsometer mounting plate. All three structural components were fabricated from billet block 6061-T6 aluminum alloy for ease of machineability and good dimensional stability. A replicate optical cell was fabricated from Halar® 6014 chlorofluoropolymer sheet stock [Ausimont USA, Inc., Thorofare NJ] for purity and inertness in corrosive environments.

To achieve tolerances acceptable for constructing the optical cell and precision mounting system, a 1937 M-Head Bridgeport vertical milling machine was reconditioned with four new quill bearings and a 6" Kirt Angle Lock vise. Quill spindle-to-table skew was minimized by shimming out the head until a tolerance of (+ 0.00025"), 1" off -spindle center, was achieved, as measured with a chucked Starrett depth micrometer (± 0.00005") with a 3/8" shank rotating against an AA grade Starrett Precision Granite block. Three-axis vice alignment was achieved by again indicating with the chucked Starrett dial indicator, from the fixed jaw's top and face, as well as along the vice bed. Product surface finish and cross-feed rates were optimized by maintaining a high spindle speed of 3450 revolutions per minute. With the mill properly tuned, both aluminum alloy and Halar® block materials could be precisely and accurately machined to designed dimensions.

After rough layout, three block sections, one at (2.25"J x 3.0" Wx 1.0"H) and two at (2.25"Z x 3.0"^x 0.5"H) were cut from 1.0" and 0.5" 6061-T6 aluminum alloy plate stock using a worm-drive rotary cut-off saw with a carbide-tipped blade. All edges were then deburred using a stainless steel knife edge, and inspected for carbide impregnation. Using the Bridgeport milling machine with a new Do All 7/16" two-flute center-cutting carbide end mill, each block was individually surfaced to provide a square block, from which fine layout measurements were transferred. The best surface of each block was placed down and locked into the vice atop 1.5" Brown and Sharpe Ultra Precision parallels, and then surfaced. A maximum cut of 0.050" was required to remove all imperfections and provide a uniform straight surface. The block was then dismounted, deburred and re-mounted to extend 0.5" beyond the vise j aws with the newly-surfaced face now against the parallels. With the second block top face surfaced, the end mill's two side cutters were used to square and surface the extended side face. The block was then dismounted, deburred, remounted with the squared- and surfaced-face placed against 0.5" Brown and Sharpe parallels, tightened in place, and surfaced. After dismounting and deburring, the block was positioned atop 1.5" parallels with 0.5" vise overhang to surface the last two remaining opposite side faces. The squared- and surfaced-block was again dismounted and deburred. Using a Starrett 1 " outside micrometer, a Starrett 6"caliper, and a 6" precision square, the block was inspected for height variation and squareness. This procedure was repeated for two aluminum and one Ηalar® block stocks. A stainless steel scribe was used to transfer layout patterns to each block. The optical cell was fabricated first to minimize the chance of guide and bolt patterning errors. With the optical cell block positioned vertically on 0.5 " parallels, the side face was relieved 0.1875" to accommodate an average internal-reflection element which measures 50 mm in length. The opposite block side was also relieved 0.1875". With maximum dimensional stability in mind, side relief was limited to 0.3750" from the top edge. From the top view, the block resembles a "T" in shape. With burrs removed, the optical cell block was then placed on 0.5" parallels in the horizontal position. A 7/16" cobalt end mill was plunged 0.125" into block center and cross-fed to form a (1 3/8"Z x 1/16"W x 1/8"H) groove. With the y-table feed locked, a 1/4" DoAll center-cutting two-flute carbide end mill was chucked and plunged 0.75 " and cross-fed 0.875" on-center, forming an internal cavity necessary for fluid containment. With the y-table feed still locked, the 1/4" end mill was removed and replaced with a 7/16" end mill. The cell block was positioned and locked down with the open cavity against a 30° Brown and Sharpe Ultra Precision angle block. The 7/16" end mill was plunged 0.050" per cross-fed stroke until a 0.75" long window shoulder had been formed. A 1/4" end mill was now chucked and plunged into the window flange center and cross-fed 0.25" on center, forming a 0.5" long x 0.25" wide window with a 0.125" wide window shoulder. The optical block was then rotated 180° and locked against the 30° angle block. Identical materials and methods were used to fabricate the second window opening.

Optical cell inlet and outlet supply tubes were fabricated from two 304L stainless steel (2.0" x 16 gauge) syringe needles cut down to 1.0" and 1.5" in length, and ends beveled using 600 grit sand paper. To provide a flat drilling surface, the optical cell was positioned at 50° nominal to the table, a 1/8" end mill was plunged 0.125" into the top left hand corner 0.125" from the block bottom. This procedure was repeated for the lower right hand optical cell corner. The optical cell was again positioned at 50° nominal to the drill table, where a series of four K-Swiss twist drills (0.0520", 0.0550", 0.0595", 0.0635") and two DoAll straight-shank reamers, (0.0645" and 0.0650") were individually chucked to machine out both inlet/outlet bores to 0.0650". Both stainless steel inlet and outlet tubes were tapped into place following substrata guide fabrication using two small pieces of red oak, a 5 lb. hammer, and a drill vise.

Fabrication of the optical cell's locator guides, bolt bores, and bolt counter bores was undertaken after the substratum holder was fabricated to allow for simultaneous error-free machining. Bonding of the optical lenses was performed last. The substratum locator block was positioned atop 1.5" parallels and cut down to 0.3125" thick using a 7/16" end mill. The substratum locating groove was formed using a 7/16" end mill, centered 0.5" from block bottom and plunged 0.0625". The block was cross-fed until a (1.0" L x 3.0'Wx 1/16" H) recessed notch was formed. Centered in the notch, the 7/16" end mill was plunged through the block and cross- fed 0.75" on-center, forming a (1.5" L x 7/16" W) access window. The block was dismounted, deburred, and placed vertically on the vice bed to machine MAIR-IR spectrophotometer stage mount receptacles. With both X and Y cross-feeds locked into position, a new 1/8" DoAll center-cutting two-flute long-cut carbide end mill was plunged 7/16". The block was then positioned to the second MAIR-IR spectrophotometer-stage dowel location and plunged. A 0.625 " K-Swiss twist drill was chucked and used to counter-drill a substratum stop bore 0.125" above the edge of the bottom guide. A small (0.625" O.D. x 0.5") 316L stainless steel dowel was then inserted and peened to provide a substratum stop peg.

Both optical cell and substratum locator were again deburred, washed, and measured for proper tolerances. The optical cell and substratum locator were aligned in parallel, and locked into a high-speed drill vice atop 0.25" parallels. A K-Swiss wire gage #15-size twist drill was used to simultaneously counter-drill through both blocks. Two guide bores and four bolt bores were drilled out to 0.1800" O.D. With the optical sensor removed, the substratum locator's two dowel bores were reamed using a 3/16" undersize 0.1865 " DoAll straight shank reamer. Two center punches, 0.1800" and 0.1850", respectively, were used to transfer bore location to the ellipsometer mount block. Four bolt bores were tapped using a NAPA 3/16" x 24 pitch tap and Tap Free cutting fluid. With the substratum locator remo ed, the optical cell was remounted into position and all six bores were reamed using a 3/16" oversized 0.1885" DoAll straight-shank reamer. Four bolt bores were opened up an additional 0.0459" using a letter "A" K-Swiss twist drill. To accommodate four compression springs (0.250" ID. x 0.250" L), the optical cell's four bolt bores were counter-bored 0.3125" using a 3/8" end mill. Both blocks were again deburred and washed. Prior to dowel guide press-fitting, an oxy-acetylene torch with a #2 tip was used to preheat the substratum holder to approximately 660°F. When heated, a 10" bench vise and a (0.25" ID. x 0.375"Z) brass bushing were used to press-fit both (0.1875" O.D. x 1.5"J) 316L stainless steel surface ground dowels through the undersized substratum holder's dowel bores. To inspect component fit and correct dowel extension length, the optical cell was mounted to the substratum holder using four stainless steel (0.1 S15"O.D. x 1.5"Z) hex head cap screws. When assembled and bolted with a 1 mm substratum, both guide dowels extend 0.125" beyond the optical cell's surface to accommodate substrata and O-rings of varying thickness. Both guide dowels also extend 0.3750" beneath the substratum holder to accept an ellipsometer mounting plate.

With the ellipsometer mount block stock squared , surfaced and scribed, both transferred guide dowel positions were drilled and then reamed using the K-Swiss "#15" wire gage twist drill and a oversized DoAll (0.1885") reamer. Four transferred bolt center-punch marks were counter-milled with a 3/8" end mill to provide bolt clearance when both optical cell and substratum holder were mounted.

A high level of precision was required to ensure optical alignment for both ellipsometry and MAIR-IR spectrometry techniques. Poor optical energy transfer resulting from misalignment, stresses or positioning errors would reduce the level of sensitivity required when monitoring events at, or near, the detection limits of these techniques.

Because aluminum, even anodized aluminum with a thick passive oxide layer, is susceptible to corrosion in environments with a high or low pH, a barrier coating was applied to prevent ion migration and material degradation. Halar® 6014 chlorofluoropolymer [AusimontUSA, Inc., Thorofare, NJ], a copolymer of ethylene, chlorotrifluroethylene, and hexafluoroisobutylene with excellent barrier properties [MatWeb, (2000), "The Online Materials Information Resource", Internet Web Site: www.matweb.com] was applied using standard electrostatic spraying methods. The aluminum optical cell was washed with detergent [Sparkleen, Fisher Scientific Co., Pittsburgh, PA], and rinsed with distilled water. To increase the surface energy of the already-high energy aluminum oxide surface, the optical cell was subjected to radio frequency gas plasma treatment (RFGPT) for three minutes. [Plasma Cleaner/Sterilizer, Model PDC-32G, Harrick Scientific Corp., Ossining, NY.] The optical cell was then preheated in a ceramic tissue burner to 280 °C. [Jelenko Accu-Therm 250, Armonk, NY.] A topcoat of Halar® was applied using a KH electrostatic spray gun with a pneumatic box feeder at 30 psi. The powder-coated optical cell was then placed into the oven until "flow-out" of the first topcoat was observed. "Flow-out" refers to the phase transition of the applied "dry" polymer solid into a viscous "liqui- fied" polymer through the addition of heat. With oven and part temperatures reduced by 15 °C, a second Halar® topcoat was applied and allowed to flow out into a glossy transparent film. When cooled to room temperature, the Halar®-encapsulated optical cell was inspected for film porosity and general film integrity, with no such defects found. The chemical composition of Halar® was documented using MAIR infrared spectroscopy. A sample was taken of the powder and clamped to a KRS internal- reflection element (IRE). [Harrick Scientific Corp., Ossining, NY.] A KRS internal- reflection element was selected for the ability to clamp samples tightly to its surfaces without the possibility of fracture, which could happen if a germanium IRE were used. A residue spectrum also was obtained to identify any residual polymer remaining on the IRE surface after removal of the powder from the (IRE). An energy- dispersive X-ray spectrum was obtained of the Halar® to document elemental composition. Fused silica was chosen as the optical window material, providing superior purity and thermal stability as compared with other crystalline quartz window materi- als. Optical cell windows were cut to rough dimension using a diamond scribe and then brought to tolerance [7/8" (2.2 cm)Zx 3/8" (0.9 cm) W] with a wet bench finishing sander using 600 grit silicon carbide paper.

Radio frequency gas plasma treatment (RFGPT) was used to enhance adhesive resin wettability and window adhesive bonding to both Halar®-coated aluminum and Halar® block optical cells. The RFGPT process, in air, has been shown to modify the polymer surface to form strong covalent carbon-oxygen bonds, which are more polar and reactive than the intrinsic carbon-hydrogen bonds. Increased surface polarity allows for better resin wettability and added interfacial covalent bonding. [Scheppele, J., "Surface Modification of Plastics for Medical Device Applications", Internet Web Site: www. a2c2.com/archive/699surfacemod. htm (2000).] Wetting liquids of known liquid- vapor surface tensions (Table 1), and a contact angle goniometer [Model NRL-100, Rame-Hart, Inc., Mountain Lakes, NJ)], were used to document the change in critical surface tension of the Halar® surfaces associated with RFGPT.

Each optical cell was individually RFGPT' d for three minutes prior to resin application and lens bonding. A conservative film (as thin as possible) of UV-curable methacrylate resin [Clear Glass Adhesive, Loctite Corporation. , Rocky Hill, CT] was applied to each window flange surface prior to window placement. A UV lamp was placed 6" (15.2 cm) from both window surfaces and illuminated for thirty seconds. [Lamp from Second Wind Model lOOOka, Derksen Air Products International Inc., Fort Erie, Ontario, Canada.] The optical cell then was placed in a 500 ml beaker and purged with overflowing hot tap water for 30 minutes to remove any uncured resin. To identify the chemical fingerprint of the cured adhesive resin, a MAIR-IR spectrum was taken of an appropriately-sized (1 cm² x 0.25 mm) cured resin sample using a germanium IRE. [Harrick Scientific Corp., Ossining, NY.] With the cured resin sample removed from the IRE, a residue MAIR-IR spectroscopy spectrum was obtained to identify any exudates. An energy-dispersive X-ray spectrum was also obtained to document elemental composition.

The optical sensor and substratum holder were assembled and flow tested to check for leaks using a 50 ml syringe filled with distilled water. Input and output lines [0.125" (0.32 cm) ID. x 15" (38 cm) L] were coupled using 2 mm (diameter) Masterflex microbore tubing sections to provide leak-proof connections. The biosensor remained leak-free at flow rates up to 300 ml/min. Substrata with poor dimensional parallelism with the mounting base required the use of a gasket or O-ring, which compensated for the uneven surface. Because Halar® and other mono- fluoropolymers are incompressible and tend to be difficult to displace, one ultrasonically-welded Teflon® (polytetrafluoroethylene)-sheathed silicone O-ring [Creavy Seal Co., Olyphany, PA] was used to provide the needed dimensional compliance, as well as to maintain cell content purity. Contact angle measurements were performed on the O-ring's outer sheath to estimate critical surface tension, thereby confirming outermost chemical makeup. [Baier, R. E., Shafrin, E. G., and Zismsn, W. A., "Adhesion: Mechanisms That Assist or Impede It", Science, 162:1360-1368 (1968).] Silicone or fluorinated Viton® O-rings commonly used in laboratory devices were not used, as they failed to provide a suitable level of chemical purity when measuring events in pure environmental systems. Silicone oils leach and creep from the material, contaminating surrounding surfaces. Surface Analysis

The following subsections describe methods used to characterize surfaces by criteria of outermost chemical make-up, optical film thickness, surface potential, morphology, and elemental composition. A. Null Ellipsometry

Null ellipsometry was chosen for this study as the in-situ real-time monitoring technique for optically measuring the film thickness. Dry films have previously been measured with reported precisions approaching± 1 A. [Spanier, R. F., "Ellipsometry A Century Old New Technique", Industrial Research (1975.] Null ellipsometry involves measuring the change in reflection of a mono-chromatic, collimated, elliptically-polarized light beam. When appropriately-selected elliptically-polarized light is reflected, it can become linearly- or plane-polarized. Phase differences and amplitude ratios between the two axes of the "optical" ellipse, as measured by the ellipsometer, can be directly related to the index of refraction and the extinction coefficient of the reflecting surface. [James, W. M., "Use Ellipsometry to Measure Thin Films", Research/Development, pp.67-70 (1977).] When the reflecting substratum is covered with a thin film of differing refractive index, the plane of polarization of the reflected light changes and the measured angles of extinction or "null" also change. Using computer software, the angle change in degrees can be directly related to the index of refraction and the thickness of the film. [McCrackin, F. L., "A Fortran Program for Analysis of Ellipsometer Measurements", National Bureau of Standards Technical Note 479, Washington, D.C. (1969).] Usually, a given index ofrefraction is assumed, and the film thickness is calculated for the assumed value.

Referring now to Fig.3, the components of the null ellipsometer, in direction of energy travel, were a helium-neon laser (632.8 nm), abeam-depolarizer/collimator, an adjustable quarter-wave polarizer, an adjustable quarter-wave compensator, a reflective substratum, an adjustable quarter- wave analyzer, a filter, a hotodiode, and an extinction meter. The 632.8 nm wavelength He-Ne laser provides a stable collimated polarized energy source. To provide superior beam performance, a depolarizer and collimator head are mounted to the laser body to depolarize, further colli- mate, and filter the laser output beam before entering the polarizer. Once the laser beam has been transmitted through the adjustable polarizer, it becomes linearly- or plane-polarized. The compensator, also adjustable, consists of a quarter-wave plate. It is used to convert the linearly-polarized light into elliptically-polarized light. A fixed compensator value of 45° is used for most measurements. The adjustable analyzer is used to extinguish the reflecting linearly-polarized beam, as monitored by the photodiode and displayed on the null meter. If some ellipticity exists, as detected by the photodiode, the polarizer and analyzer are adjusted until a "null" is reached. For maximum precision, both the polarizer and analyzer are repetitively "nulled" until maximum extinction has been achieved. Polarizer and analyzer extinction measurements can be obtained in each of the four quadrants. Typically, only the first (0°-90°) and second (90°- 180°) quadrants are measured. A Rudolph thin-film null ellipsometer [Type 43702-200E, Rudolph Research Corp., Flanders, NJ] was used.

B. Multiple- Attenuated Internal-Reflection Infrared Spectroscopy The technique of multiple-attenuated internal-reflection infrared (MAIR-IR) spectroscopy was incorporated into the bio-sensor design to provide chemical identity, as well as mass per unit absorbance, of deposited films. It has been demonstrated that a linear relationship exists between the deposited film's mass and IR absorbance values for films having thicknesses up to 1/100^th of the absorbed wavelength. [Fornalik, M. S., Meyer, A. E., and Baier R. E., "Experimental Determination of the Information Depth (Di) for Strongly Absorbing Species Assayed by Internal Reflec- tion Spectroscopy", Abstracts, Federation of Analytical Chemistry and Spectroscopic Societies, Abstract No. 342 (1983).] An example of a linear mass/absorbance relationship for the protein β-lactoglobulin is provided (Fig.4). All calibration and realtime films investigated in this recent study were less than 100 A in thickness. For films investigated in this research, it has been assumed that the mass per unit absop- tions show similar linear relationships.

MAIR-IR spectroscopy operates by measuring the frequency at which infrared energy is absorbed and emitted by resonating covalent bonds at frequencies between 4000 cm"¹ and 400 cm^"1. A spectrum is then generated, from which the types of bonds present in the film can be identified. Sample absorption area is maximized by the use of attenuated total-reflection optics and an infrared transparent internal-reflection element (IRE). Large germanium IRE's (50 mm x 20 mm x 1 mm) with 45° bevels provide for approximately eleven internal-reflections per side, for a total of twenty-two reflections. [Mattson, J. S. and Smith, C. A., "Enhanced Protein Adsorption at the Solid-Solution Interface: Dependence on Surface Charge", Science, 181:1055-1057 (1973).] A Perkin-Elmer Model 1420 Ratio Recording Infrared Spectrophotometer [Norwalk, CT], was used. C. Surface Contact Potential

Surface contact potential measurements were obtained to monitor the changes in surface potential of a substratum associated with the addition of a thin film. This technique is useful when measuring dipolar hydrocarbon-based films, as the differ- ence in surface potential values should indicate the degree of orientation, for example, and molecular orientation of the film in relation to the substratum, provided that the film's molecular dipoles are positioned perpendicularly to the substratum's surface. [Adamson, A. W., Physical Chemistry of Surfaces, 2nd Ed., John Wiley and Sons, Inc, pp. 127-129, 150-159 (1967).] The technique of contact potential used here is based on the vibrating electrode method, where an audio-frequency current drives a loudspeaker magnet that, in turn, vibrates a small metallic disk mounted 0.5 mm. above, and parallel to, the sample surface. Disk vibration causes a corresponding variation on the capacity across the air gap, setting up an alternating current in the second circuit, whose mag- nitude depends on the voltage difference across the gap. [Adamson, A. W., Physical Chemistry of Surfaces, supra, at pp. 127-129, 150-159.] The null current value or the voltage is then recorded and later compared to the modified sample surface to determine the change in polarity associated with the thin deposited film. The contact potentiometer used in this study was custom-made. D. Scanning Electron Microscopy/Energy-Dispersive X-ray Spectrometry

Scanning electron microscopy with energy-dispersive X-ray spectrometry was used to provide morphological, as well as elemental, identification of adsorbed deposits. The technique of SEM uses the interaction of an electron beam with a conductive sample surface to emit secondary electrons, producing a signal from which a three-dimensional image of the sample surface is formed. X-ray species are also emitted from the near sample surface during electron beam bombardment, producing a signal from which elemental compositions are obtained. [Goldstein, J. I., Newbury, D. E., Echlin, P., Joy, D. C, Fiori, C, and Lifshin, E., Scanning Electron Microscopy andX-Ray Microanalysis, A Text for Biologists, Material Scientists, and Geolo- gists., Plenum Press, New York and London (1981).]

For greater contrast, samples analyzed by SEM and EDX-ray were not over- coated with any conductive layers. The thin films on intrinsically-conductive substrata sufficiently suppressed the substratum's emitted secondary electrons, such that excellent contrast was developed for the overlying organic deposits . [Fornalik, M.S., "Biophysical Models of Acquired Pellicle Formation on Substrata of Varying Surface Energy", Masters Thesis, State University of New York at Buffalo, USA (1982).] A Hitachi Model S4000 Scanning Electron Microscope. [San Jose, C A.] with energy-dispersive X-ray spectrometer [Model OS26-H009, Princeton Gamma-Tech Spectrometer, Princeton, NJ] was used during the course of this study. E. Comprehensive Contact Angle Analysis Contact angle analysis was performed to determine the critical surface tension and to infer the surface energy of material surfaces used in optical cell construction. In 1805, Thomas Young described the relationship of a contact angle (θ) of a liquid drop at equilibrium on a plane solid surface, relating the equilibrium contact angle to the interfacial tension between the solid and liquid phase (γ^sl), the liquid-vapor surface tension (γlv), and the surface- vapor surface tension (γ^sv) as follows:

An empirical value, the critical surface tension (γ^c), which is conceptually related to, but not equal to, the surface free energy (γ^s) has been defined from the contact angle data in the form of a Zisman plot. A Zisman plot is the rectilinear relationship between the cosine of the contact angle and the known liquid-vapor surface tension (γ^lv). [Baier, R. E., Shafrin, E. G., and Zismsn, W. A., "Adhesion: Mechanisms That Assist or Impede It", defence, 162:1360-1368 (1968).] The CST is defined as the intercept where cosine θ = 1, of the best-fit line plot of cosine θ vs. γ^lv, for all liquids with finite (non-zero) contact angles. [Baier, R. E., Surface Properties Influencing Biological Adhesion, in: R. S. Manley, ed., Adhesion in Biological Systems, New York: Academic Press, 15-48 (1970).]

Theoretical surface energy values, also can be derived from comprehensive contact angle analysis. The polar component of the surface free energy of a solid (γ^p) and the dispersive component of the surface free energy of a solid (γ^d) can be calcu- lated from contact angle data using the D. H. Kaelble approach. [Gucinski, H., "Cor- relation of Biophysical Surface Characteristics With Hydrodynamic Properties of Adhesive Biofilms", Ph.D. Dissertation, State University of New York at Buffalo (1984).] Both components contribute to the theoretical surface free energy as follows: γP + γ^d = γ^s The surface free energy can be attributed to the work of cohesion (W_c ), or the energy required to divide a material into two new surfaces. The surface energy generally correlates with the hardness of the material. Hard solid materials have higher surface energies than softer solid materials. Because most liquids are in the low energy range, it can be expected that spreading events will occur on high energy surfaces only. [Baier, R. E., "On the Formation Of Biological Films", Swed. Dent. , 1:267-271 (1977).] Spreading is the result of inter-molecular interactions between solid and liquid phases, leading to adhesion through overcoming the cohesive forces of the liquid.

The contact angles of eleven test liquids (Table 1) were measured using a contact angle goniometer [Model NRL-100, Rame-Hart, Inc., Mountain Lakes, NJ], as individually applied to the sample surfaces with a thin platinum wire, flamed prior to each droplet application. A Zisman plot was constructed for each sample, plotting the average cosine contact angle vs. the liquid-vapor surface tension for each test liquid. A line was drawn through the data points, forming a line of best-fit for the liquids with non-zero contact angles. The intercept of the line, where cosine θ = 1, determined the critical surface tension of the solid. [Baier, R. E., "Surface Properties Influencing Biological Adhesion", in: R. S. Manley, ed., Adhesion in Biological Systems, supra at pp.15-48.] Zisman plots were constructed for the Halar® surfaces before and after RFGPT, as well as for the Teflon® O-ring. Bio-Sensor Calibration

Calibration measurements were necessary to determine the accuracy and precision of the optical system, as well as to test the overall design and function prior to real-time monitoring.

Calibration measurements were conducted using germanium internal-reflection elements with adsorbed and transferred films of reference human fibrinogen, octadecanoic acid, octadecanol, and octadecylamine. Both the custom-made sensor and the purchased Perkin-Elmer optical mounting systems, were used on the spectrophotometer to record the chemical identity and absorbance of each film. Spectral absorbances were calculated from absorption band data:

Absorbance = log (T₀/T_s) where T₀ is the percent transmittance of the baseline spectrum at the wavelength of interest, and T_s is the percent transmittance of the sample at the same wavelength (i. e. at the minimum transmission for the absorption band of interest). [Meyer, A. E., "Dynamics of 'Conditioning' Film Formation on Biomaterials", Ph.D. Dissertation, University of Lund, Sweden (1990).] The thin-film null ellipsometer was used to record null angles, which, when entered into a computer [McCrackin, F. L., "A Fortran Program for Analysis of Ellipsometer Measurements", National Bureau of Standards Technical Note 479, Washington, D.C. (1969)], provided a refractive index or optical thickness of each film in the dry (i.e., outside the optical cell), dry optical cell (i.e., inside the optical cell), and wet optical cell (i. e., inside water-filled optical cell) environments. Knowing the mass (m) per unit area (a) provided by MAIR-IR spectrometry and apparent geometric thickness (h) from ellipsometry, relative dry and wet bio-sensor film density indexes were derived and compared. The relative density index values calculated are not true densities in terms of g/cm³, but are indexes used to compare films. [Meyer, A. E.,"Dynamics of 'Conditioning' Film Formation on Biomaterials", Ph.D. Dissertation, University of Lund, Sweden (1990).]

D.I. = m/v where D.I. = density index, m = index of mass from unit absorbance values, and v = (film area from X and Y dimensions of prism face(s)) x (film geometric thickness at presumed refractive index).

Contact potential measurements were taken at baseline and after film adsorption or transfer to determine the change in surface dipole moment associated with the film. Baseline measurements refer to the pre-modified substratum condition, or in the clean deposit-free state. A. Adsorbed Human Protein Calibration Series

In preparation for MAIR-IR spectroscopy, ellipsometry and contact potential baseline measurements, a germanium internal-reflection element (IRE) was detergent- washed [Sparkleen, Fisher Scientific Co., Pittsburgh, PA], rinsed in distilled water, and RFGPT' d in air for three minutes. All components of the bio-sensor, including the optical cell, substratum holder, supply lines, and 50 ml syringe reservoir were detergent-washed and rinsed in distilled water both before and after each calibration measurement. Individual baseline, or "clean" , MAIR-IR spectra were taken using both the Perkin-Elmer substratum holder and the bio-sensor. For bio-sensor ellipsometer mount alignment, a black fine-tip marker was used to place a vertical line directly on the center of each window. With the IRE still mounted, the bio-sensor was fitted into the ellipsometer mount, and aligned by rotating the ellipsometer base mount until both lines were illuminated by the laser beam. To prevent misalignment, double-sided sticky tape [3M Corporation., St. Paul, MN] was used to secure the bio-sensor's ellipsometer mount to the ellipsometer stage. To ensure sharp null signals, laser beam throughput was maximized by adjusting the ellipsometer stage Y, Z, pitch and roll adjustments. These position measurements were recorded to ensure substratum relocation if accidental stage movement occurred. With the bio-sensor aligned and its signal maximized, ellipsometer baseline measurements were recorded in the first and second quadrants for two x-axis locations, analyzing two separate 1 mm² regions of surface area. Measurements were recorded, first on the IRE using only the bio-sensor substratum holder to maintain position and optical alignment, eliminating the need for laser beam throughput maximization once the optical cell had been remounted. To determine the influence of the optical cell windows, the optical cell was mounted and a second set of "ambient" dry angle measurements was recorded. The previously- cleaned 50 ml syringe was used to infuse 20 ml of distilled water into the bio-sensor before a set of "wet" ellipsometer angles were recorded. The bio-sensor was then drained, disassembled, and the IRE allowed to air-dry while tilted at 70° to assist drainage. Once dry, the IRE was placed into the contact potentiometer and the surface potential recorded.

A 40 μg/ml solution of human fibrinogen [Cat # 850-80, Lot # 97F-6171, Sigma Diagnostics., St. Louis, MO] in distilled water was puddled on the IRE surface for 30 minutes at 69 °F. The IRE then was tilted to remove excess fibrinogen solution and leached in distilled water for 15 seconds to remove any loosely-bound non-adsorbed protein. Leaching is the process of gently water- wetting the substratum surface, minimizing interfacial shear, thus preserving adsorbed species, while removing, when the IRE is tilted, any dissolved and non-adsorbed species. Once dry, the IRE was again leached in distilled water for 15 seconds a second time.

MAIR-IR spectroscopy, dry, dry-cell, and wet-cell ellipsometry, and contact potential measurements were conducted on the protein-modified IRE. This experimental procedure was replicated for one additional fibrinogen experiment. For ease of discussion, the first calibration film referenced will be designated as film 1. The replicate film will be designated as film 2. For example, the first fibrinogen film referenced above will be denoted as fibrinogen film 1 in the discussion section that follows. The second replicate fibrinogen film will be denoted as fibrinogen film 2. This numerical designation is repeated for all primary and replicate transferred calibration films. B. Hydrocarbon Calibration Series

Hydrocarbon calibration films were analyzed using the analytical procedures for germanium IRE baseline and film analysis described above. However, the method of film preparation differed. A custom-built automated Langmuir-Adam trough was used to individually prepare model mono-molecular films of octadecanol [Cat #25, 876-8, Lot #0441 IHT, Aldrich Chemical Co., Inc., Milwaukee, WI], octadecanoic acid [Cat #S-4751, Lot #37F-8365, Sigma Chemical Company., St. Louis, MO], and octadecylamine. [Cat #S-9273, Lot #70H3728, Sigma Chemical Co., St. Louis, MO.] The Langmuir-Adam trough allowed a compressed mono-molecular film of each material to be transferred to the solid IRE substratum. A vertical dipping device provided constant upward substratum movement of 1 mm min, while a trough-mounted tensiometer drove a servo motor that maintained apparent film-packing pressure (surface tension reduction from the pure-water value) at 16 dynes/cm, and automatically compensated for loss of film during substratum removal, ensuring uniform film transfers. Hydrocarbon and chloroform [Cat #9180-01, Lot #N10A10, J.T. Baker.,

Phillipsburg, NJ] solutions (500 μg/ml) were individually prepared prior to each film transfer experiment. The Langmuir-Adam trough was filled with distilled water, the water surface swept with a Teflon® bar to remove any surface contaminants, and the tensiometer nulled. A 0.15 ml aliquot containing 75 μm of hydrocarbon, was applied to the air/water interface by using a detergent-washed, rinsed in distilled water, glass microscope slide as a ramp, down which the solution was trickled. The film was allowed to stabilize twenty minutes for chloroform evaporation and film distribution. The tensiometer was preloaded over 1 hour, at approximately 4 dynes/cm every 15 minutes, to achieve an apparent film packing pressure (surface tension reduction) of approximately 16 dynes/cm. [Baier, R. E., Ph.D., PE, Professor and Director, Center for Biosurfaces, State University of New York at Buffalo, Personal Communication (1999).] Using the dipping device, the clean germanium IRE substratum was lowered and raised perpendicularly to the water's surface through the air/film interface at 1 mrn/min. Because hydrocarbon species used were amphipative, the polar ends of the molecules oriented toward the water's surface, while the hydrocarbon tails oriented toward the air phase. Film transfer did not occur during the lowering of a polar substratum, clean germanium, through the air/film interface because the non-polar tails did not interact with the polar substratum. Only upon a change in substratum direction (in this case, removal), did the polar ends of the molecules become attached and transferred to the polar substratum. A single mono-layer of hydrocarbon was expected to be transferred to the clean polar substratum during the first removal from the air/film interface. The substratum was removed from the dipping device and both faces leached in distilled water for 15 seconds. Under normal circumstances, transferred mono-layers are not water-leached or rinsed, but were leached in these experiments in order to reduce the chance of film migration during wet-cell bio-sensor ellipsometry calibration measurements.

After the first hydrocarbon film had been analyzed using identical techniques and methods as described for the fibrinogen film series, a second hydrocarbon film was transferred onto the first, using the Langmuir-Adam trough method described above. Film transfer was expected during both substratum insertion and removal in this instance because the previously-film-modified substratum now had a non-polar surface allowing for interaction with, and transfer of, the trough's air-phase-oriented non-polar surface film during insertion, as well as interaction with, and transfer of the polar entities during extraction. Two additional mono-layers of hydrocarbon should have been transferred to the substratum during the second dipping series. The substratum was removed from the dipping device and leached in distilled water for 15 sec- onds to remove any loosely-bound poorly-transferred hydrocarbon species, reducing the potential of film migration during wet-cell bio-sensor ellipsometry calibration measurements.

Substratum analysis for the second transferred hydrocarbon film series was performed using the materials and methods described for the above fibrinogen film series. This experimental procedure was repeated for all additional hydrocarbon calibration film measurements. Real Time In-Situ Monitoring

Real-time in-situ measurements were performed to determine the effectiveness of the bio-sensor for monitoring adsorption events in ultra-pure, as well as distilled, and non-pure water systems. Real-time measurements were conducted using germanium internal-reflection elements and similarly-sized 300-series stainless steel alloy coupons over 24-and 88-hour time periods, respectively. Because previous experiments have demonstrated that ultra-pure, as well as distilled, waters corrode germanium surfaces (data not shown), exposure times were limited to 24-hours to prevent excessive IRE damage. Stainless steel, noted for its good corrosion resistance, was exposed for the longer-term 88-hour measurements. In addition to MAIR-IR spectroscopy, ellipsometry and contact potential, scanning electron microscopy with energy- dispersive X-ray spectrometry was used to provide morphological, as well as elemental identification of deposits. The referenced magnification provided on each SEM micrograph represents the nominal original magnification used when recording the image. The scale bar represents the true unit of scaled dimension, calibrated using a standard diffraction grating.

A. Distilled Water Series

In preparation for baseline measurements, a germanium IRE was detergent- washed, rinsed in distilled water, and exposed to RFGPT for two minutes. One SEM micrograph and an EDX-ray elemental spectrum of the IRE central flow surface were obtained to ascertain pre-exposure conditions. The IRE was again detergent- washed, rinsed in distilled water, and exposed to RFGPT for three minutes, removing any residual vacuum pump oil that may have been electron-beam-polymerized during SEM/EDX-ray analysis [Little, 1991]. Individual pre-exposure MAIR-IR spectra were taken using both the Perkin-Elmer substratum holder and bio-sensor.

With the bio-sensor positioned on the ellipsometer mounting block, laser beam throughput was maximized by adjusting the ellipsometer stage Y, Z, pitch, and roll adjustments. Ellipsometer-stage position measurements were recorded to ensure substratum relocation if stage movement occurred. With the bio-sensor aligned and maximized, ellipsometer baseline measurements were recorded in the first and second quadrants for two x-axis areas on the IRE. Ellipsometer baseline measurements were recorded first on the IRE using only the bio-sensor substratum holder to maintain position and optical alignment, eliminating the need for laser beam throughput maximization once the optical cell had been remounted. To monitor optical cell window performance, the optical cell was mounted on the ellipsometry and "dry-cell" measurements were recorded.

A detergent- washed, distilled water-rinsed polyethylene bottle was filled with 250 ml of distilled water to provide a fluid reservoir from which a flow-calibrated peristaltic pump [Masterflex®, Cole-Parmer Instrument Company, Vernon Hills, IL] supplied 1 ml/min through 0.125" (0.3 cm) I.D. PTFE tubing [Cole-Parmer Instrument Company, Vernon Hills, IL] to the bio-sensor. A 0.125" (0.3 cm) I.D. PTFE return line provided fluid transfer back to the reservoir. Once filled, the bio-sensor was positioned onto the ellipsometer mount and "wet" baseline angles were recorded for the same two previously-characterized surface areas. Parafilm® tape (Type "M", American National Can™, Chicago, II) was placed over the reservoir's top to secure and seal both PTFE lines, preventing evaporation. Five individual ellipsometer measurements were taken over a 24-hour time frame (10 min, 1 , 6, 15, 24 hrs) for the same two previously-examined surface areas. After 24 hours of continuous distilled water flow, the bio-sensor was drained and disassembled, and the IRE air-dried at a 70° tilt to ensure complete drainage. After each experiment, the optical cell was detergent- washed and rinsed in distilled water to remove any dehydrated residues that may have formed on the windows. In sequence, MAIR-IR spectroscopy, ellipsometry (dry and dry-cell), and SEM/EDX-ray spectrometry analysis were performed on the post-exposure IRE surface. This experimental procedure was replicated for one additional 24- hour distilled water run. For ease of discussion, the first 24-hour distilled water experiment will be designated as distilled water run 1 A. The replicate 24-hour experiment will be designated as distilled water run IB.

All distilled water supplies were sourced from a Corning Mega-Pure™ System, Model MP-1 [Corning Glass Works, Corning, NY] distillery known to be in good working condition. Characterization of distilled water aliquots included pH using a calibrated pH meter [pH tester 1 ™, by Oakton] contact angle measurement of individual drops of the distilled water on lab-standard polytetrafluoroethylene (PTFE) calibration film [Baier, R. E. and Meyer, A. E., Surface Analysis, in: Handbook of Biomaterials Evaluation, Scientific, Technical, and Clinical Testing of Implant Materials, A. F. von Recum, ed, Macmillan Publishing Company, 97-108 (1986)], and chemical identification of dried-down deposits from 1 ml of water evaporated on a germanium IRE using MAIR-IR spectrometry. Contact angle and pH measurements were made in triplicate.

For long-term (88-hour) real-time distilled water measurements, a 300-series stainless steel substratum was used. Stainless steel was chosen for having good corrosion resistance in water and high reflectivity when polished, as required for ellipsometry. Because stainless steel is not infrared transparent, internal-reflection MAIR- IR spectroscopy was not possible. With the exception of MAIR-IR spectroscopy analysis, identical experimental procedures for bio-sensor baseline and real-time flow measurements were used as discussed in Section 3.4.1. Onereplicate 88-hour distilled water experiment was performed.

For ease of discussion, the first 88 -hour distilled water experiment will be designated as distilled water run 2A. The replicate (89-hour) experiment will be designated as distilled water run 2B. B. Ultra-pure Water Series

Real-time short-and long-term ultra-pure water experiments were conducted on germanium and stainless steel substrata using identical techniques and methods as described above. One gallon of ultra-pure water was provided by Dr. Jon Sjogren from the University of Arizona's Center for Microcontamination Control ultra-pure water system. A standard polyethylene laboratory jug was used to transport and store the ultra-pure water prior to use. Ultra-pure water aliquots were characterized using the methods and materials described above. Because ultra-pure water was shown to corrode the surface of germanium prisms in preliminary studies (data not shown), IRE' s used for replicate 24-hour UP runs 3 A and 3B were not examined by MAIR-IR. The stainless steel substratum used in the distilled water experiments was used for replicate 88+ hour UP trials 4A and 4B. C. Non-Pure Water Series

To monitor adsorption events in a non-pure water system, aquarium water was selected to represent a biologically-active source. The water was supplied from a freshwater tank of approximately 30 gallons containing two different species offish and some zebra mussels. The tank had been in continuous use for seven years prior to this experiment. A mature biofilm was observed on the tank surfaces. Water aliquots were characterized using the methods and materials described above. Replicate 1-hour experiments, 5 A and 5B were performed on germanium IRE's and replicate 2-hour experiments, 6A and 6B with stainless steel substrata. Because the mea- surements of aquarium water was productive of biological deposits as determined by MAIR-IR spectrometry, pH, and droplet contact angle measurements, ellipsometry (wet-cell) values were obtained every 10 minutes. Substratum baseline and film analysis techniques and methods were the same as described above. Results The following sections present the results of the various experiments and measurements performed during the course of this study. A. Design and Fabrication

Design of the thin-film bio-sensor was determined to be satisfactory in providing a dimensionally-stable optical platform. Leaching from optical cell materials, such as the lens-bonding adhesive, PTFE-sheathed silicone O-ring, and Halar®, into the flow chamber and onto the test substratum (even when exposed to harsh ultra-pure water) was not observed. Optical cell rigidity was maintained, providing alignment of the windows and preventing translational stresses from altering the state of laser beam polarization during ellipsometry measurements. The precision guide dowel mounting system provided precise relocation of all bio-sensor components during assembly/disassembly, as well as during transfer and remounting to other instruments. Optical realignment of the ellipsometer stage mount was not required throughout the duration of calibration and real-time measurements. Bio-sensor MAIR-IR spectrometer positioning receptacles provided optimal alignment with the MAIR-IR spectrometer' s optical platform, allowing greater beam transmission than the original manufac- turer' s locator mount.

Results of flow cell materials characterization by contact angle analysis, MAIR-IR spectrometry, and energy-dispersive X-ray spectrometry provided useful information in identifying outermost chemical compositions. Contact angle measurements were made on bulk Halar®, RFGPT Halar® and PTFE O-ring materials, and are represented graphically in the form of Zisman plots, (Figs. 5, 6, and 7). Table 2 lists the values of the extrapolated critical surface tension γ^c, the calculated polar and dispersion components, γ^p and γ^d, the total surface energy, γ^s, and the ratio of polar component to total surface energy, percent γ^p. For increased adhesive bonding of the fused silica windows, the Halar® optical cell was RFGPT'd. As indicated by the Zisman plots for stock Halar® (Fig. 5) and RFGPT'd Halar® (Fig. 6), where cosine θ = 1, the RFGPT process increased Halar® 's critical surface tension 3 dynes/cm, from 27 dynes/cm to 30 dynes/cm indicating a slightly more oxidized surface. The polarity of the optical cell's Halar® surface was substantially increased by 12 mN/m following RFGPT, providing a more polar surface for better resin wetting and bond- ing. A Zisman plot for the PTFE-sheathed silicone O-ring (Fig. 7) indicates a critical surface tension of 19 dynes/cm, confirming the manufacturer's claim of a Teflon® outermost composition.

Table 2

Contact Angle Results

The MAIR-IR spectrum for Halar®, obtained using a KRS internal-reflection element, is given in Fig. 8. Using an infrared correlation chart [Barnes Engineering Company, Stamford, CT], absorption bands were observed for carbon-fluorine at 1230 cm^"1 and 1150 cm^"1, CH=CH₂ bonds at 1000 cm^"1, and C-Cl bonds at 770 cm^"1, 720 cm^"1, and 635 cm^"1. A residue spectrum was obtained after removing the Halar® sample from the prism (Fig. 9), to determine if any exudates had been transferred to the prism during clamping. No residual Halar® residues were detected above the baseline spectrum (not shown), suggesting that Halar® is resistant to migration under the conditions of this experiment. The energy-dispersive X-ray spectrum obtained from the Halar® sample (Fig. 10) revealed large carbon, fluorine and chlorine peaks, confirming Halar®' s elemental composition.

It was also of interest to identify the chemical fingerprint of the UV-cured window bonding adhesive. A MAIR-IR spectrum for the methacrylate adhesive was obtained using a germanium IRE. Using the infrared correlation chart, the resin's spectrum (Fig. 11) revealed absorptions associated with C-O bonds at 1730cm^"1, C-O- C bonds at 1220 cm^"1 and 1160 cm^"1, and C=O bonds at 1450 cm"¹. The energy- dispersive X-ray spectrometry spectrum obtained from the resin (Fig. 12) revealed one large carbon peak, a modest oxygen peak, and a small silicon peak. It is possible that the silicon is present as SiO₂ in the resin as a filler, or as a silicone component. Surface Analysis of Adsorbed and Transferred Calibration Films

The following four subsections report the results of calibration experiments performed, designed to measure the accuracy, precision, and detection thresholds of the bio-sensor system.

A. Multiple-Attenuated Internal Reflection-Infrared Spectroscopy MAIR-IR spectroscopy was a useful surface analytical technique, confirming adsorbed and transferred calibration film presence and chemical composition. Baseline MAIR-IR spectra obtained with a germanium IRE using both the Perkin-Elmer optical mount (Fig. 13), and the bio-sensor (Fig. 14), were recorded prior to each calibration measurement. Spectra recorded using the mounted bio-sensor revealed small absorptions at 1200 cm^"1 and 1140 cm^"1, representative of the PTFE O-ring used to seal the optical cell to the germanium substratum.

IR spectra obtained for fibrinogen film 1 using the Perkin-Elmer mount (Fig. 15) and the bio-sensor (Fig. 16) revealed absorptions at the 3300 cm^"1, 1650 cm^"1, and 1540 cm"¹ regions, representing nitrogen-hydrogen and hydroxyl groups, protein amide I and protein amide II groups, respectively [Meyer, 1990]. Using the method described above for determining film mass, the amide I plus amide II absorbance values were calculated. Absorbance totals (amide I plus amide II) of fibrinogen film 1 using both the Perkin-Elmer mount and bio-sensor are compared in Table 3. The ratios of amide I to amide II absorbance of fibrinogen film 1 using both the Perkin- Elmer mount and the bio-sensor are compared in Table 4. The results for fibrinogen film 1 indicate that only slight differences in absorbance values are evident when comparing spectra obtained using the Perkin-Elmer mount and the bio-sensor. Transmission of infrared energy through the prism is greater when using the bio-sensor by as much as 16%. Starting at the same percent transmission, the bio-sensor showed a 10% gain from 2000 cm^"1 to 1480 cm^"1 and a 12-16% gain from 1480 cm^"1 to 880 cm^"1. A possible explanation for increased transmission could be that, because one side on the prism contacts the aluminum (infrared opaque) bio-sensor surface, some infrared energy is reflected back into the prism. Another possibility is that the bio-sensor maintained better optical alignment with the optical platform's mirror array.

Table 3

Multiple- Attenuated Internal-Reflection Infrared (MAIR-IR)

Spectrophotometry Absorbance Values for Adsorbed and

Transferred Calibration Films

Calibration Optical Log Sum ToMEAN S.D.± Film Mount tal

Adsorbed Fibrinogen Films

Spectra obtained for fibrinogen film 2 using the Perkin-Elmer mount and the bio-sensor revealed nearly identical absorptions inthe 3300 cm^"1, 1650 cm^"1, and 1540 cm^"1 regions. Total absorption, however, for film 2 was only about half that of film 1, indicating that there was significantly less mass present for film 2. Amide I and amide II absorbance ratios (Table 4) for both fibrinogen films 1 and 2, show similar percentages, indicating that the two films are nearly identical in chemical composition. Starting at the same percent transmission, the bio-sensor showed a 10% increase in transmission from 1480 cm"¹ to 980 cm^"1.

Spectra obtained for octadecanol film 1 using the Perkin-Elmer mount (Fig. 17) and the bio-sensor (Fig. 18) revealed absorptions at the 2890 cm^"!and 2820 cm^"1 regions, representing CH₂ hydrocarbon species. Absorbance totals of octadecanol film 1 using both the Perkin-Elmer mount and bio-sensor, compared in Table 3, show identical absorptions. When using the bio-sensor, transmission of infrared energy through the prism was about 12% greater from 1480 cm^"1 to 980 cm^"1.

Spectra obtained for octadecanol film 2 using both mounts, revealed absorptions at identical regions as observed in octadecanol film 1. Mean absorption for film 2, however, was greater than for film 1 , indicating that film 2 had approximately 63 % more mass. When using the bio-sensor, transmission of infrared energy through the prism was about 10% greater from 1480 cm^"1 to 980 cm^"1.

Spectra obtained for octadecanoic acid film 1 using the Perkin-Elmer mount (Fig. 19) and the bio-sensor (Fig. 20) revealed CH₂ hydrocarbon absorptions at the 2920 cm^"1 and 2840 cm^"1 regions, and OH-C=O carboxylic acid groups in the 1700 cm^"1 region. Absorbance totals of octadecanoic acid film 1 using both the Perkin- Elmer mount and bio-sensor, compared in Table 3, are substantially identical. When using the bio-sensor, transmission of infrared energy through the prism was about 13% greater from 1480 cm^"1 to 980 cm^"1.

Spectra obtained for octadecanoic acid film 2 using both mounts revealed absorptions at identical regions as observed in octadecanoic acid film 1. Mean absorbance for film 2, however, was greater than for film 1 , indicating that film 2 had ap- proximately three times the relative mass. When using the bio-sensor, transmission of infrared energy through the prism was about 5% greater from 2000 cm"¹ to 980 cm^"

1

Spectra obtained for octadecylamine film 1 using the Perkin-Elmer mount (Fig. 21) and the bio-sensor (Fig. 22) revealed absorptions at the 2900 cm^"1 and 2840 cm^"1 regions, representing CH₂ hydrocarbon species. Absorbance totals of octadecylamine film 1 using both the Perkin-Elmer mount and bio-sensor, compared in Table 3, are identical. When using the bio-sensor, transmission of infrared energy through the prism was about 5% greater from 1400 cm^"1 to 980 cm^"1.

Spectra obtained for octadecylamine film 2 using both mounts, revealed ab- sorptions at identical regions as observed in stearylamine film 1. Mean absorbance for film 2, however, was slightly greater than for film 1, indicating that film 2 had about 20% more mass. When using the bio-sensor, transmission of infrared energy through the prism was about 6% greater from 1400 cm^"1 to 980 cm^"1.

For demonstration purposes, the bio-sensor was filled with distilled water to determine the "wet" MAIR-IR spectral absorbance of octadecylamine film 2 (Fig.23). Large OH peaks at 3350 cm^"1 and 1640 cm^"1 dominated the spectrum, obscuring most of the CH₂ hydrocarbon peaks. For this reason, all further MAIR-IR spectra for this study were obtained in the dry bio-sensor cell environment. It has been demonstrated, however, that with sufficient computational power, useful spectra could be recorded from fully-hydrated specimens in wet-cells. [Baier, R. E. and Meyer, A. E., Surface Analysis, in: Handbook of Biomaterials Evaluation, Scientific, Technical, and Clinical Testing of Implant Materials, A. F. von Recum, ed, Macmillan Publishing Company, pp. 97-108 (1986).]

B. Thin-Film Ellipsometry Thin-film ellipsometry was used to determine the optical thickness of adsorbed and transferred calibration films in the dry ambient air phase, as well as in the "dry- cell" ambient air, and liquid "wet-cell" phases using the bio-sensor. Calibration was necessary to determine variability between dry ambient, and "wet-cell" bio-sensor values. Dry-cell bio-sensor thickness values also were recorded and compared to monitor the optical cell's window performance, to detect any residual stresses. Ellipsometry measurements were recorded for baseline and film-modified substrata in two geometric quadrants on two different substratum locations. Because all calibration films used in this exercise were hydrocarbon based, the refractive index value of 1.5 was used when obtaining "apparent" geometric film thickness values. [Meyer, A. E., "Dynamics of 'Conditioning' Film Formation on Biomaterials", Ph.D. Disserta- tion, University of Lund, Sweden (1990).] Changing the film's assumed refractive index when calculating the geometric film thickness from McCrackin's program [McCrackin, F. L., "A Fortran Program for Analysis of Ellipsometer Measurements", National Bureau of Standards Technical Note 479, Washington, D.C. (1969)] did lead to higher or lower values of the associated geometric film thickness but, not the com- parative trend between "dry" ambient and bio-sensor "wet-cell" values. When computing ellipsometry Del (change in beam phase upon reflection) and Psi (change in beam amplitude upon reflection) values, the refractive index of the phase surrounding the laser beam's refraction point must be known. For "dry" and "dry-cell" ambient measurements, the surrounding phase is air. Therefore, a refractive index value of 1.0 was used. For bio-sensor "wet-cell" measurements, the surrounding phase is distilled water. Therefore, the refractive index value should be between 1.32 and 1.33. After a series of trial and error calculations, it was determined that refractive index values of 1.33, 1.32, 1.25 and 1.22 would be individually entered into the program when calculating the wet bio-sensor optical thicknesses for each calibration film. A "wet- cell" phase refractive index of 1.22 yielded the closest values to those of the "dry" and "dry-cell" thickness values; all "wet-cell" bio-sensor ellipsometry thickness values reported here were calculated using a refractive index phase value of 1.22. Using a refractive index of 1.33 raised the apparent "wet-cell" bio-sensor thickness values by up to 10%. The optical path through the fused silica windows may have altered the ellipse of the beam, resulting in modified Del and Psi values.

Comparative geometric film thickness values for all calibration films for dry, dry-cell, and wet-cell bio-sensor conditions are reported in Table 5. Means are given for each film (Table 5), derived using film thickness values from two sample surface locations on the specimen. In theory, thickness values for adsorbed fibrinogen films should have been identical, and "second-dip" hydrocarbon transferred films should have been three times the thickness of the first transferred film.

Fibrinogen film 1 had an average film thickness of 89 A with only a small difference in apparent film thickness values as evident by comparing dry, dry-cell, and wet-cell bio-sensor film thickness values for both substratum locations. Fibrinogen film 2 had an average film thickness of 40 A with greater differences in apparent film thickness values as indicated by a standard deviation value of 4. Octadecanol film 1 had an average film thickness of 27 A with slight differences in apparent film thickness values as evident by comparing dry, dry-cell, and wet-cell bio-sensor film thickness values. Octadecanol film 2 had an average film thickness of 47 A with a similarly-small standard deviation. Octadecanoic acid film 1 had an average film thickness of 22 A with slight differences in apparent film thickness values. Octadecanoic acid film 2 had an average film thickness of 48 A with greater thickness value variability as indicated by the standard deviation value of 4.0. Octadecylamine film 1 had an average film thickness of 36 A with slight differences in apparent film thickness values as evident by comparing dry, dry-cell, and wet-cell bio-sensor film thick- ness values. Octadecylamine film 2 had an average film thickness of 33 A with a similarly small standard deviation. Octadecylamine was the only hydrocarbon film that failed to retain additional material thickness after leaching with distilled water. C. Surface Contact Potential Contact potential measurements were obtained to monitor the changes in surface potential of a substratum associated with the addition of a thin film. This technique was expected to be useful when measuring amphipative hydrocarbon films, as the difference in surface potential values should indicate the molecular orientation of the film in relation to the substratum, provided that the film's molecules are positioned perpendicularly to the substratum's surface. Contact potential values for base- line and calibration films are listed in Table 6. Changes in substratum surface potential associated with film depositions are also listed, as well as percent change in potential, and associated film thickness.

The results indicate that all primary calibration films had a negative surface potential value ranging from -0.407 mv. for octadecylamie to -0.604 mv. for octadecanol. Only small differences in surface polarity resulted from the addition of a second hydrocarbon film to the substratum. All secondary films became slightly more negative, with values ranging from -0.513 mv. for octadecanoic acid, to -0.654 mv. for octadecanol, suggesting that additional film further reduces the contact potential. However, surface polarity for the secondary octadecylamine film became more negative despite having less film present.

D. Calibration Film Density ^Relative Values) Relative density values of calibration films were calculated for each film using thickness data provided by ellipsometry and absorbance values (relative mass) determined by MAIR-IR spectroscopy as described above. Mean relative density values and standard deviations for each film using dry and wet bio-sensor ellipsometry thickness values, as well as Perkin-Elmer and bio-sensor MAIR-IR absorbance values, are presented in Table 7. Only slight variations were observed when comparing phase and optical mount relative density values for each film, illustrating that both precise and accurate thin film measurements were obtained using the bio-sensor. When comparing density index values of replicate films, slightly larger variations occurred, showing that the method of film adsorption or transfer had some degree of variability.

Table 7

Calibration Film Density Index Values

Surface Analysis of Short- and Long-Term Real-Time Experiments The following subsections discuss the results of the bio-sensor's ability to monitor film adsorption events in real-time, using distilled, ultra-pure, and non-pure water supplies.

A. Water Supply Characterization

To document water quality, all three water supplies were individually charac- terized using MAIR-IR spectroscopy, contact angle measurements on PTFE film, and pH measurements. Spectra obtained for distilled water (Fig.24), and ultra-pure water (Fig. 25), revealed only small absorptions at 1430 cm^"1, suggesting that some carbonate-forming materials were present in the water supplies during evaporation. Using the Barnes infrared correlation chart, the infrared spectrum obtained for aquar- ium water (Fig.26) revealed large absorptions at 3350 cm^'1, 1400 cm"¹, 1110 cm^"1, and at 1000 cm^"1, illustrating that the aquarium water had more hygroscopic salts and biological materials present, as contrasted with distilled and ultra-pure waters.

Table 8 lists the mean contact angle on PTFE film for each water supply. All contact angle measurements were made on lab-standard polytetrafluoroethylene (PTFE). These results indicate that, as water purity increases, mean contact angles increase and standard deviations decrease. This trend was likely related to varying amounts of surface-active substances dissolved in each water supply. The presence of such an impurity in the liquid, would decrease its liquid-vapor surface tension, resulting in a lower contact angle.

Table 9 presents mean pH for each water supply. All pH values were recorded using a calibrated pH meter as described above. Standard deviation results suggest that as water quality increases, difficulty in measuring pH increases. This trend was likely due to varying buffering capacities of each water supply. The slightly alkaline aquarium water showed the least pH measurement variability, suggesting a well- buffered fluid.

B. Multiple- Attenuated Internal-Reflection-Infrared Spectroscopy

MAIR-IR spectroscopy was used to confirm and identify films adsorbed during real-time bio-sensor measurements. Baseline MAIR-IR spectroscopy spectra were obtained with a germanium IRE using both the Perkin-Elmer substratum holder and the bio-sensor prior to each measurement. All baseline spectra obtained indicated that the prism was clean and free of surface contaminants.

Spectra obtained for distilled water run 1 A using the Perkin-Elmer mount and the bio-sensor (Fig. 27), revealed no absorption different from their corresponding baseline, indicating a film-free surface. Similarly, spectra obtained for replicate distilled water run IB using the Perkin-Elmer mount and the bio-sensor (Fig. 28), revealed no additional absorptions, indicating the absence of any adsorbed film constituents.

Spectra obtained for aquarium water run 5 A using the Perkin-Elmer mount and the bio-sensor (Fig. 29), revealed small absorptions in the 1650 cm^"1, and 1540 cm^'1 regions, representing amide I and amide II groups, respectively, from the first deposited "conditioning film" . Similarly, spectra obtained for aquarium water run 5B using the Perkin-Elmer mount and the bio-sensor (Fig. 30), revealed absorptions associated with these same protein bands at 1650 cm^"1 and 1540 cm^"1. C Thin-film Ellipsometry

Ellipsometry measurements were obtained in-situ using the bio-sensor for each water supply, determining the effectiveness of this technique for monitoring interfacial surface events in real-time. Short-term measurements were made on germanium internal-reflection elements allowing for simultaneous MAIR-IR spectroscopy analy- sis for distilled and non-pure water supplies. Long-term real-time in-situ measurements were individually performed using a polished 300-series stainless steel coupon. Apparent film thickness results for 24-hour distilled water real-time run 1 A are presented graphically in Fig. 31. Because there was no apparent deposition of organic molecules, as independently determined by MAIR-IR spectroscopy and SEM/EDX- ray analysis (instead, a loss of high refractive index material was probable), in-situ thickness values were calculated using refractive indexes of 3.9, 4.0, and 4.1. Error bars represent the difference in film thickness values calculated using these three refractive indexes. These were the highest refractive index values found to be consistent with the calculations yielding physically realistic values for film thickness. Only one positive film thickness value was observed, 10 minutes after initiating water flow. All subsequent measurements indicated a continuous loss of material from the germanium surface, for a total loss of 7 A after 24-hours. Dry and dry-cell ellipsometry measurements independently confirmed material losses of 7 A and 6 A, respectively. Fig. 32 represents the apparent film thickness results for 24-hour distilled water real-time run IB. Vertical bars represent the difference in film thickness values calculated using refractive indicies of 3.9, 4.0, and 4.1. Again, only one positive film thickness value was recorded 10 minutes after baseline (initiation of water contact). Subsequent in-situ material losses totaled 10 A of thickness loss (germanium plate corrosion). Dry and dry-cell thickness values confirmed in-situ material losses with corresponding values of 10 A and 9 A, respectively. Fig. 33 represents the apparent film thickness results for 88-hour distilled water real-time run 2 A (stainless steel substratum). There was no discernible deposition of organic molecules when inspected by SEM/EDX-ray analysis. A loss of high refractive index material was noted, with apparent thickness values calculated using the refractive indexes of 2.6 and 2.7. These are refractive index values for chromium oxide, the native oxide of 300-series stainless steel. Vertical bars in Fig.33 represent the difference in film thickness values calculated using the two refractive indexes. A total material thickness loss of 4 A was measured in-situ, 13 hours from the beginning of the experiment, after which a passivating film formed protecting the surface from further degradation. Dry and dry-cell thickness values showed apparent thickness gains of 5 A and 4 A, respectively, indicating that some additional material may have deposited during substratum removal and residual water evaporation. Alternatively, the passivating surface oxide film may have become thicker, or more dense and refractive after water evaporation.

Fig. 34 represents the apparent film thickness results for 89-hour distilled water real-time run 2B . Vertical bars represent the difference in film thickness values calculated using the 2.6 and 2.7 refractive index values. Material losses or gains were not observed in-situ over the 89-hour exposure, suggesting that the passivation film formed during run 2A was sufficient to provide corrosion protection in the distilled water environments. Dry and dry-cell thickness values showed gains of 2 A and 4 A, respectively, again resulting from possible buildup of the superficial oxide layer.

Fig. 35 represents the apparent film thickness results for 24-hour ultra-pure water real-time run 3 A (germanium substratum). Vertical bars represent the range in film thickness values calculated using refractive indicies of 3.9, 4.0, and 4.1. Two positive film thickness values were recorded at 10 and 60 minutes after baseline (initial exposure to ultra-pure water). All subsequent measurements demonstrated in- situ material losses totaling 12 A. Dry and dry-cell thickness values confmned mate- rial losses with corresponding losses of 12 A and 15 A, respectively.

Fig. 36 represents the apparent film thickness results for 24-hour ultra-pure water real-time run 3B. Vertical bars represent the range in film thickness values calculated using refractive indicies of 3.9, 4.0 and 4.1. Unlike the first exposure real- time film thickness results for the germanium substratum, no initial increase in film thickness was observed. All in-situ measurements resulted in a continuous loss on material, for a total loss of 18 A. Dry and dry-cell thickness values confirmed material losses with corresponding values of 18 A and 17 A, respectively.

Fig. 37 represents the apparent film thickness results for 94-hour ultra-pure water real-time run 4 A (stainless steel substratum). Vertical bars represent the difference in film thickness values calculated using the 2.6 and 2.7 refractive index values. A slight material loss of 2 A was observed in-situ over the 94 hour exposure, suggesting that the passivating film formed during prior distilled water exposures was not as resistant to the more aggressive ultra-pure water. Dry and dry-cell thickness values showed losses of only 1 A and 2 A, respectively, indicating a substantial resistance of the surface oxide to further losses or growth.

Fig. 38 represents the apparent film thickness results for 90-hour ultra-pure water real-time run 4B. Vertical bars represent the difference in film thickness values calculated using the 2.6 and 2.7 refractive index values. Only a slight material loss of 1 A was observed in-situ over the 94-hour exposure, confirming that the passivating film was still intact from previous ultra-pure water exposure. Dry and dry-cell thickness values showed neither material loss, nor gain, over the 90-hour exposure, again indicating stability of the passivating oxide layer.

Film thickness results for 1-hour aquarium water real-time run 5 A are pre- sented graphically in Fig. 39 (germanium substratum). Because deposited and retained film molecules were determined by MAIR-IR spectroscopy to be proteinaceous type species, thickness values were calculated using a refractive index of 1.5. [Meyer, A. E., "Dynamics of 'Conditioning' Film Formation on Biomaterials", Ph.D. Dissertation, University of Lund, Sweden (1990).] A continuous gain in film thickness was observed, resulting in a total adsorbed film thickness of 28 A in just 1 hour of exposure, to a total volume of only 250 ml of water having total dissolved organic compo- nents of only a few parts per million (see Fig. 29). Dry and dry-cell ellipsometry measurements independently confirmed material gains of 38 A and 37 A, respectively, showing some possible addition of material from viscous "carry-out". [Baier, R. E. and Glantz, P.O., "Characterization of Oral in Vivo Films Formed on Different Types of Solid Surfaces", Acta Odontol. Scand, 36:289-301 (1978).] That is, the increased film thickness may have been associated with the observation that, following biosensor disassembly and substratum removal, the tilted substratum' s bound protein film retained a viscous water layer even when "wicked" onto filter paper from the bottom. This "bulk" water phase probably contained additional material which deposited upon dehydration, causing the noted 10 A gain in final film thickness.

Film thickness results for 1-hour aquarium water real-time run 5B are presented graphically in Fig. 40. Thickness values were calculated using a refractive index of 1.5. A continuous gain in film thickness was observed, resulting in a total adsorbed film thickness of 26 A during 1-hour of exposure. Dry and dry-cell ellipso- metry measurements independently confirmed material gains of 34 A and 34 A, respectively, again showing some carry out. Additional wicking was utilized prior to dry down, resulting in 4 A less "carry-out" than in run 5A.

Film thickness results for 2-hour aquarium water real-time run 6A are presented graphically in Fig. 41 (stainless steel substratum). Thickness values were calculated using a refractive index of 1.5. A continuous gain in film thickness was observed, resulting in an apparent total adsorbed film thickness of only 8 A during 2 hours of exposure. Dry and dry-cell ellipsometry measurements independently confirmed material gains of 40 A and 44 A, respectively, well beyond the extra amount attributable to "carry-out". One explanation for the excessive dry/wet film "apparent" thickness is that a lack of a refractive contrast was associated with relatively close refractive index values for the hydrated adsorbed film and hydrated stainless steel passivating layer, resulted inthe inability to follow film growth as sensitively as with the germanium substratum. In fact, the initial wet film was probably much thicker on stainless steel than on germanium, since wicking of water from the film after disas- sembly was not possible, as the very viscous bound film restricted any capillary draining action. Film thickness results for 2-hour aquarium water real-time run 6B are presented graphically in Fig. 42. Thickness values were calculated using a refractive index of 1.5. A continuous gain in film thickness was observed, resulting in a total adsorbed film thickness of 8 A during 2 hours of exposure. Dry and dry-cell ellipso- metry measurements independently confirmed material gains of 50 A and 52 A, respectively, showing more substantial wet-to-dry discrepancies than in above film 6A. It is a matter of concern that the most corrosion-resistant stainless steels show the least ellipsometric sensitivity, underwater, to the events of "conditioning film" formation that one hopes to monitor with the utmost sensitivity in ultra-pure water systems. D. Scanning Electron Microscopy/Energy-Dispersive X-ray Spectroscopy

Scanning electron micrographs and energy-dispersive X-ray spectrometry spectra were obtained for each pre- and post-exposed substratum surface, documenting changes in morphological and elemental compositions associated with real-time exposures. An example of baseline SEM micrographs with complementary EDX-ray spectrometry spectra for pre-exposed germanium (Fig.43) and stainless steel (Fig.44) surfaces, shows clean deposit-free surfaces.

SEM/EDX-ray spectrometry analysis of the exposed germanium flow surfaces for 24-hour distilled water real-time runs 1 A (Fig. 45) and IB revealed small white particles in the centers of etch pits evenly distributed over the surface. The particles were small and of limited surface area, so it was difficult for EDX-ray spectroscopy to confirm their elemental identity. Upon further examination, it was confirmed that these white deposits were germanium oxides (Fig. 46), probably deposited within the etch pits during dehydration of the water- wetted surfaces. When the bio-sensor was drained, it was tilted to a 70° angle and slightly rolled, allowing most of the liquid to exit through the bottom supply tube. However, when all of the liquid did not escape, it remained localized by capillary forces at the lowest optical cell/substratum interface, as well as in surface pits and defects. These are the only locations where germanium oxide deposits were found as represented in above figures and Fig.47. A "conditioning film" was not observed on either substratum within the 24-hour exposure period, as confirmed independently by ellipsometry and MAIR-IR spectroscopy. SEM/EDX- ray spectrometry analysis of the exposed stainless steel flow surface for distilled water real-time runs 2A (Fig.48) and 2B showed no differences in surface appearance from their respective baselines.

SEM/EDX-ray spectrometry analysis of the exposed germanium flow surface after 24-hour ultra-pure water real-time experiments 3A (Fig. 49) and 3B revealed small white particles in the centers of pits evenly distributed over the surface, identical to those found in the replicate 24-hour germanium distilled water experiments. EDX- ray spectroscopy was unable to detect any element other than germanium.

SEM/EDX-ray spectrometry analysis of the exposed stainless steel flow surface for ultra-pure water real-time runs 4A (Fig. 50) and 4B showed no differences in surface appearance from their respective baselines.

SEM/EDX-ray spectrometry analysis of the germanium flow surface for 1- hour aquarium water real-time experiments 5 A (Fig. 51) and 5B revealed dark surface deposits distributed in patchy areas over the whole surface area. For ran 5A, the deposits were concentrated in the center of the test substratum. For run 5B, faint deposits were seen to be orientated longitudinally, along the plate surface. EDX-ray spectroscopy spectra identified deposits as carbon-based, as represented by small carbon peaks.

SEM/EDX-ray spectrometry analysis of the stainless steel flow surface for aquarium water real-time runs 6A (Fig. 52) and 6B showed surface deposits of varying types. Run 6A's surface deposits were dark and patchy, while 6B's were small and are "charged" (appearing bright due to inability of electrons to escape). EDX-ray spectra revealed very slight carbon peaks.

E. Relative Density Values

Relative density values for sample films were calculated for all films using thickness data provided by ellipsometry and estimated mass determined by MAIR-IR spectroscopy as described in Section 3.3. Mean relative density values and standard deviations for each film using the last recorded in-situ bio-sensor ellipsometry thickness values, as well as Perkin-Elmer and bio-sensor MAIR-IR absorbance values, are presented in Table 10. Only slight variations were observed when comparing relative density values for all films.

Analysis

Spontaneous formation and retention of "conditioning films" on engineering materials exposed to biological fluids has been well established over the past 30 years .

[Meyer, A. E., (1990), "Dynamics of 'Conditioning' Film Formation onBiomaterials",

Ph.D. Dissertation, University of Lund, Sweden).] What has not been established, however, is a method for accurately and precisely measuring these events in-situ over extended periods in pure water environments. This research identifies a technology that will allow monitoring of conditioning films, as well as the further investigation of biological interfaces, allowing for more effective scheduling of water treatments in pure water systems. Current proposed technologies for monitoring pure water system contamination, such as resonant wave modulation and optical interference, have not yet been shown to detect conditioning film events in real time. By monitoring the events of conditioning film formation over time, an accurate estimate of bio-contamination can be established.

This work utilized a multi -method approach for the characterization of films adsorbed on and transferred to previously characterized surfaces. In addition to increasing scientific yields, the influences of testing environments on substrata surface properties, as well as the formation of adsorbed and deposited surface films were observed. Where one analytical technique becomes limited in terms of sensitivity or sampling depth, another can be used to continue characterizing the sample of interest. The use of multiple analytical techniques for comprehensive surface characterization of retained films also allowed for the direct comparison of results, not available through any single technique, providing an invaluable means for calibrating the biosensor system.

Calibration film experiments determined the bio-sensor' s accuracy and preci- sion as an optical cell allowing for "wet" ellipsometry in-situ film measurements. All ellipsometery film values demonstrated minimal variation, strongly supporting the bio-sensor's ability to allow film thickness measurements on pre-adsorbed and transferred films in an aqueous environment. Similar results were reported earlier [Cuy- pers, P. A., Hermens, W. Th., and Hemker, H. C, "Ellipsometry as a Tool to Study ProteinFilms at Liquid-Solid Interfaces", Analytical Biochemistry, 84:56-67 (1978)] , demonstrating the precision of "wet" ellipsometry, Del and Psi values using transferred layers of barium stearate.

Fibrinogen film thickness values observed in this research also were similar to those earlier reported. [Rothen, A., Ellipsometric Studies of Thin Films, in: Prog- ress in Surface and Membrane Science, D. A. Cadenhead, J. F. Danielli, M. D. and Rosenberg, eds., Academic Press, 8:81-116 (1974).] Differences in overall thickness between fibrinogen films 1 and 2, probably resulted from slightly different elapsed times prior to fibrinogen leaching. Fibrinogen film 1 was allowed to air dry before distilled- water leaching, thus permanently retaining more fibrinogen. Fibrinogen film 2 was distilled water-leached immediately following 30 minutes of fibrinogen adsorption, thus preventing dehydration and retention of additional fibrinogen.

A high bio-sensor dry-cell value for fibrinogen film 2, 8 A above the corresponding dry and wet bio-sensor values, probably resulted from residule moisture in the cell. Following all ellipsometer baseline measurements, the bio-sensor was detergent-washed and rinsed in distilled water. If residual water were not completely removed from the bio-sensor's supply tubes after baseline measurement cleaning and drying, the bio-sensor, when assembled and mounted, formed a condensate of water vapor inside the cell and on the windows, causing laser beam attenuation errors. Water presence in the supply tubes was not detected by the MAIR-IR spectroscopy technique, which is usually sensitive to the presence of water vapor because of the relatively cool room temperature near the instrument. Air circulating near the ellipso- metry work station is on average 10 °F warmer than that circulating by the MAIR-IR spectrometer, thus enhancing the formation of condensation inside the bio-sensor when positioned onto or near the cooler ellipsometer.

Overall thicknesses of transferred C₁₈ hydrocarbon films as measured in the dry, dry-cell, and wet-cell environments were consistent with previously reported values. [Adamson, A. W., Physical Chemistry of Surfaces, Second Edition, John Wiley and Sons, Inc, 127-129, 150-159 (1967).] By monitoring each film in two sample locations, some added information on film thickness was established, but it would have been beneficial to examine additional surface sample locations. Spacing limitations for the bio-sensor prevented such further measurements.

Resulting wet-cell bio-sensor film thickness measurements for octadecanoic acid film 2, were on average, 5 A greater than either dry or dry-cell values. This error occurred due to the presence of air bubble formations at the film/water interfaces resulting from an abnormally-high dissolved air concentration in the distilled water supply reservoir. When held to room lighting, small bubbles in the reservoir's distilled water supply, as well as in the bio-sensor and on the film, became evident. After an elapsed time of 40 minutes, the micro-bubbles had merged, forming several large bubble pockets excluded to the wall of the syringe. The apparent surface roughness and hydrophobicity of the octadecanoic acid film, allowed micro-bubble retention, resulting in small ellipsometric errors even after the water supply had de-gassed.

Octadecylamine was the only hydrocarbon film that failed to retain material from a second Langmuir-Blodgett transfer after distilled water leaching. Therefore, only a single mono-layer of octadecylamine could be measured.

Dry-cell bio-sensor ellipsometry measurements were useful in determining the relative rate of beam perturbation associated with window strain. It was learned that careful attention was required when washing the optical cell, as large temperature variations in wash and rinse waters induced reversible window strains (data not shown).

All MAIR-IR spectrometry absorbance values demonstrated minimal variation, showing that the bio-sensor can additionally be used to provide substratum positioning with attenuated total-reflection optics. Spectral absorbances confirmed the presence of all adsorbed and transferred reference films and provided an essential reference for relative film mass used in calculating relative density values for each film.

Contact potential results obtained on all hydrocarbon films indicate discrepancies in defining the orientations of the amphipative hydrocarbon molecules. A single mono-layer of at least one of these differently terminated C _{t g} hydrocarbons transferred to a polar substratum, using the Langmuir-Adam trough, should have resulted in a positive surface potential, as the molecule's terminal dipole orientations would be toward or away from the substratum in the instance of octadecanoic acid and octadecylamine. Yet, the same effect was measured by the contact potentiometer for all primary films. Negative contact potential values recorded for the second transferred film may have resulted from the removal of the third, or outermost, mono-layer during distilled- water leaching. Three mono-layers of hydrocarbon should have been transferred to the polar substratum after two sequential dips. Ellipsometry thickness values indicate, with exception to octadecylamine, that only two mono-layers of hydrocarbon are present. The technique of contact potential was shown to be very sensitive to these differences, but lacking in interpretability.

For the most part, relative film density values were similar, indicating that replicate films were of mostly uniform film thickness and composition. Relative density variability of the fibrinogen films may be due to differing molecular orienta- tions associated with film dehydration. One explanation for variability in the hydrocarbon film density index values, is that poor film transfer or uneven film removal during distilled- water leaching resulted in a patchy loosely-organized film. Additionally, vibration of the dipping device used to lower and raise the substratum through the air/film interface on the Langmuir-Adam trough may have resulted in incomplete film transfer, as small water waves were observed emanating from the resonating substratum during immersion and extraction. Film thickness measurements were made only on one face of the germanium substratum, leaving the transferred film on the opposite face uncharacterized, resulting in a possible over- or under-represented film thickness value. MAIR-IR spectroscopy absorption values were derived from both sides of the germanium IRE. Therefore, the total film absorption was represented in calculating the relative density, even though the total film thickness value may not have been accurately defined.

Replicate real-time in-situ thickness measurements established the bio-sensor' s ability as an optical flow cell to allow accurate and precise ellipsometric monitoring of optical thickness changes associated with distilled, ultra-pure, and non-pure water exposures. No organically derived conditioning film components were detected on short-term, distilled and ultra-pure water exposed germanium substrata as confirmed by ellipsometry, MAIR-IR spectroscopy, and SEM/EDX-ray spectrometry. Due to the aggressive nature of distilled and ultra-pure waters, the germanium substrata's sensing surfaces were severely etched, reducing MAIR-IR spectrometry transmission. For this reason, possible use of an in-line FTIR monitored bio-sensing window in pure water systems [Mittelman, M. W., (1995), in: Microbial Biofilms, Lappin-Scott, H. M., and Costerton, J. W., eds., Cambridge University Press, 133-147; Nichols, P. D., Henson, J. M., Guckert, J.B., Nivens, D. E., and White, D. C, (1985), "Transform- infrared Spectroscopic Methods for Microbial Ecology: Analysis of Bacteria, Bacteria-polymer Mixtures, andBiofilms" , Journal of Microbiological Methods, 4:79- 94] would be unwise until more corrosion-resistant prisms are quantified. Recent experience with silicon IRE's (data not shown) suggests that silicon may serve better in this role. The extent of material removal was recorded in real-time for average removal rates of 0.4 A/hr for distilled and 0.6 A/hr for ultra-pure waters. Similar slow rates of chemical etching of germanium with water have been reported earlier. It can be concluded that the rate of germanium removal is directly related to the relative purity of the test waters, as correlated with pH, contact angle, and MAIR-IR spectrometry measurements. SEM/EDX-ray spectrometry analysis substantiated material loss by confirming the morphological and elemental identification of surface corrosion with germanium oxide deposits. Previous research has identified the "white", water- induced germanium corrosion deposits as hexagonal germanium oxides. [Rindner, W., and Ellis, Jr. R. C, "Electrolytic Etching of Germanium in Water", Journal of the Electrochemical Society, 109:537-539 (1962).] One explanation for the small gains in apparent film thicknesses measured at the first recording period after baseline, is that subsequent hydration of the native germanium oxides resulted in the transformation to thicker germanium hydroxides. [Baier, R. E., Ph.D., PE, Professor and Direc- tor, Center for Biosurfaces, State University of New York at Buffalo, Personal Communication (2000).] Another possible explanation is that the spontaneous attack on native oxides by pure waters resulted in an initially-patchy microscopically-rougher surface leading to increased Del and Psi ellipsometer values, as the surface roughness of a substratum has been shown to modify these values when measuring Del and Psi changes in air. [Tompkins, H. G., A User's Guide to Ellipsometry, Academic Press, Inc., pp. 95-108 (1993).]

Replicate long-term distilled and ultra-pure water exposures on stainless steel also failed to identify any conditioning film deposits. Post-exposure film thickness measurements and SEM/EDX-ray spectrometry measurements confirmed the lack of adsorbed deposits. It can be concluded that, due to the low concentration of materials present in these pure waters, insufficient time had elapsed to detect any structured conditioning film formations. Additionally, the relatively high surface area-to-volume ratio of the bio-sensor, supply tubing and 250 ml reservoir, may have reduced the potential for conditioning film formation on the surface under inspection. Similar results have been reported when using pure waters to temporarily store high surface energy substrata. [Meyer, A. E., "Dynamics of 'Conditioning' Film Formation on Biomaterials", Ph.D. Dissertation, University of Lund, Sweden (1990).]

Stainless steel substrata surfaces were shown to be significantly more resistant to corrosion than were the germanium substrata. Even so, before a passivating film was developed during the first exposure to distilled waters, there clearly was initial removal of surface material. A replicate distilled water exposure showed no additional signs of corrosion or passivation, concluding that the corrosion-resistant barrier formed during the first exposure was sufficient in preventing further material attack. Replicate ultra-pure water exposures, however, showed slight decreases in film thickness early in the exposure cycle, suggesting that the stainless steel's passivating film required, at least initially, a more complete and stable chromium oxide film to fully prevent corrosion in the more aggressive ultra-pure water. This is of importance to food and pharmaceutical industries which often use 300-series stainless steel supply tubing and reservoirs for ultra-pure water processes. [Mittelman, M. W., in: Microbial Biofilms, Lappin-Scott, H. M., and Costerton, J. W., eds., Cambridge University Press, pp. 33-147 (1995).]

In contrast to the pure waters tested, aquarium waters showed spontaneous adsorption of proteinaceous substances to all exposed substrata surfaces. Similar results have been reported in fresh and marine waters using conventional flow cell configurations, demonstrating deposited film thicknesses of less than 100 A over an elapsed time of 24-hours. [Meyer, A. E., Baier, R. E., and King, R. W., "Initial Fouling of Nontoxic Coatings in Fresh, Brackish, and Sea Water", 77ze Canadian Journal of Chemical Engineering, 66:55-62 (1988).] Replicate aquarium water exposures on germanium and stainless steel confirmed the ability of the bio-sensor to accurately and precisely monitor the events of conditioning film formation. Ellipsometry film values showed little variability between runs, strongly supporting the bio-sensor's ability to allow film thickness measurements in an aqueous environment, extending the previously demonstrated results. [Cuypers, P. A., Corsel, J. W., Janssen, M. P., Kop, J. M. M., Hermens, W. T., and Hemker, H. C, "The Adsorption of Prothrombin to Phosphatidylserine Multilayers Quantified by Ellipsometry, The Journal ofBiological Chemistry, 258:2426-2431 (1983).] MAIR-IR spectrometry and SEM/EDX-ray spectrometry confirmed and identified the organic deposits on the germanium substrata. SEM EDX-ray spectrometry confirmed the nature of the organic deposits on stainless steel substrata. Relative density values for the "naturally adsorbing" condi- tioning films from replicate exposures on germanium showed little deviation, indicating that the films were of more uniform thickness and mass than either the adsorbed or transferred calibration films. Additionally, the "naturally adsorbing" films were significantly more dense, by as much as 50%, than either adsorbed or transferred calibration films, demonstrating that there may be greater adhesive forces interacting between the substratum and the adsorbing molecules. An additional observation made during this research was the formation of small air bubbles on the bio-sensor's Halar® optical cell walls. After 20 hours of continuous flow, small bubbles appeared, which continued to enlarge throughout the duration of each run. Bubble production associated with surfaces is commonly called "heterogeneous nucleation", wherein bubbles form at sites of surface roughness and impurities. [Maris, H. and Balibar, S., "Negative Pressures and Cavitation in Liquid Helium", Physics Today, 53:29-34 (2000).] The bio-sensor's rough, as-machined internal surfaces may have provided small nucleation sites, allowing for the continued growth of surface bubbles. This may have significant implications in biofilm formation in Halar®-plumbed ultra-pure water systems, where bubble formation at unions and other incongruent surfaces may yield a mechanism by which nutrients and bacteria concentrate. It is well established that the air/water interface is one of the most biologically active interfaces. [Maclntyre, F., "The Top Millimeter of The Ocean", Scientific American, 230:62-77 (1974).] Perhaps degassing of waters prior to UV exposure, as well as after CO₂-evolving UV- induced "oxidation", will help reduce dissolved gas levels and the subsequent risk of surface bubble formation. Additionally, researchers have observed N₂ bubble formation under denitrifying biofilms, and reason that sloughing may result from pressure changes causing the bubbles to expand. [Characklis, W. G. and Marshall, K. C, Biofilms, John Wiley & Sons, Inc., 55-89 (1990); Jansen, J. and Marshall, K. C, "Fixed Film Kinetics: Denitrification in Fixed Films", Report 80-59, Department of Sanitary Engineering, Technical University of Denmark, in: Biofilms, Characklis, W. G, and Marshall, K. C, eds, John Wiley & Sons, Inc., pp. 341-394 (1980).]

Another observation made during real-time bio-sensor measurements was the sharpness of Del and Psi nulls as related to the apparent presence of a conditioning film. The "null width", defined as the amount of rotation in degrees the ellipsometer' s polarizer and analyzer can be moved without observing a change in meter extinction when "nulled", is a useful additional correlating parameter. In short- and long-term distilled and ultra-pure water tests, the nulls remained sharp, with average null widths of 0.05°. In non-pure aquarium water trials, the null widths continued to increase from baseline measurements, resulting in average null widths of 0.40°. This trend suggests that as the thickness of a hydrated film increases, so to, does null width. The continued null sharpness observed during distilled and ultra-pure water tests further substantiates the lack of any deposited film moieties. Additionally, null width variation can be incorporated into a bio-sensor controller, which would simply measure the null widths associated with a relative state of micro-biocontamination. Conclusions

The research performed in this study demonstrated the design, development, and evaluation of a novel thin-film bio-sensor. By using a multi-methodological approach to surface characterization, a more precise and accurate technique for monitoring the events of initial conditioning film formation was established. Experimental data strongly substantiate the precision and accuracy of this ellipsometry/bio-sensor system. By maintaining stress-free optics and precise optical alignment, ellipsometer measurements could be recorded with minimal interference or perturbation.

On an industrial level, the bio-sensor, when combined with simplified ellipsometer components, could be used as an in-line monitor to predict impending biocontamination in pure water systems, reducing the potential problems caused by unpredictable biofilm sloughing events. Pure water manufacturing systems could also benefit from recycling of waters that are currently overflowed and wasted while attempting to minimize the risk of bacterial contamination.

On an experimental level, the bio-sensor could be used to provide more accurate and precise thickness values for the kinetics of a variety of bio-molecular interac- tions, including antibody-antigen interactions, hormone-receptor interactions, and enzyme-substrate interactions. Corrosion rates and oxide stability of metallic biomaterials in physiologic solutions could also be investigated.

The technique of in-situ ellipsometery is not new. What is new is the ability to reliably monitor film thickness changes with resolutions and detection thresholds of 1 A, as demonstrated by the results of this research.

Further investigation is required to establish a range of in-situ monitoring sensitivities for materials of different refractive index values. By achieving a greater differential between the refractive index of the substratum and that of the film, sensitivity to Del and Psi changes should increase, yielding a more accurate measurement. This data will provide invaluable information on the optical interference properties as measured by the technique of in-situ ellipsometry. Additionally, further investigation is required to determine the maximum film thickness that can be reliably monitored by this technique.

Finally, for in-situ pure water monitoring, an automated elliptically polarized beam source with photodiode detector will need to be incorporated into existing biosensor technology, as the physical dimensions and costs of laboratory scale scientific ellipsometers would preclude their use. A retrofit of the current bio-sensor configuration could easily be incorporated into full size distribution piping, allowing for multiple internal line placements within the pure water system.

Modifications

The present invention contemplates that many changes and modifications may be made. For example, the particular substratum holder material may not be critical to the apparatus. Secondly, the body may be formed of many different materials. The substratum is preferably transparent to infrared energy, although this is not deemed critical. The windows are preferably formed of fused silica or quartz, and are prefera- bly adhesively secured to the body.

Therefore, while the presently-preferred form of the improved apparatus has been shown and described, and several modifications thereof discussed, persons skilled in this art will readily appreciate that various additional changes and modifications may be made without departing from the spirit of the invention, as defined and differentiated by the following claims.

Claims

ClaimsWhat is claimed is:

1. Apparatus for determining the presence of a conditioning film as a precursor indicator to bio-contamination of a fluid, comprising: a substratum having an optically-reflective surface; a body having a flow passage for said fluid and having an optical passage, said optically-reflective substratum surface being arranged in said flow and optical passages, said body having two windows arranged perpendicularly to the axis of said optical passage; and means for determining the presence of said conditioning film on said optically- reflective surface.

2. The apparatus as set forth in claim 1 wherein said substratum is formed of a material that is transparent to infrared energy.

3. The apparatus as set forth in claim 2 wherein said substratum is formed of germanium or silicon.

4. The apparatus as set forth in claim 1 and further comprising a substratum holder.

5. The apparatus as set forth in claim 1 wherein in said means includes a helium- neon laser light source adapted to provide a collimated light beam.

6. The apparatus as set forth in claim 1 wherein said windows are formed of fused silica or quartz.

7. The apparatus as set forth in claim 6 wherein said windows are adhesively secured to said body.

8. The apparatus as set forth in claim 7 wherein said adhesive includes barium and sulfur.

9. The method of determining the presence of a conditioning film as a precursor indicator to bio-contamination of a fluid, comprising the steps of: providing a substratum having an optically-reflective surface; providing a body having a flow passage for said fluid and having an optical passage, said optically-reflecting surface being arranged in said flow and optical passages, said body having two windows arranged perpendicular to the axis of said optical passage; causing a beam of light to enter said body along said optical passage and to reflect and refract from said optically-reflective surface; and measuring parameters of the entering and exiting light beams; thereby to indicate the presence of said conditioning film on said optically- reflective surface.

10. The method as set forth in claim 9 wherein said entering light beam is ellipti- cally polarized.