 |
Description  |
|
|
BACKGROUND OF THE
INVENTION
1. Field of the Invention
This invention relates to novel biochips that combine integrated circuit elements, electro-optical excitation and detection systems, and molecular receptor probes in a self-contained integrated microdevice.
2. Description of Related Art
Much interest has centered on the development of DNA chips based on high density oligonucleotide arrays and fluorescence analysis such as described by Hacia et al. (J. G. Hacia, L. C. Brody, M. S. Chee. S. P. A. Fodor F. S. Collins in Nature
Genetics Dec. 14, 1996). This principle has been commercialized in the Affymetrix.RTM. GeneChip.RTM. which was developed to process large amounts of genetic information. GeneChip.RTM. probe arrays are arranged on single chips in the form of tens of
thousands of DNA probes that are designed to fluorescence when hybridized to their targets. The light is scanned with laser light and the light intensity stored for later computations (Jul. 23, 1997, Affymetrix http://www.affymetrix.com/).
Unfortunately, the DNA chips, while much like the microprocessor chips that currently run today's technology, have yet to be successfully developed into integrated systems that conveniently interpret what information can be captured by DNA chips. Thus, an Affymetrix.RTM. chip that is stated to detect HIV mutations still requires an external scanning and interpretation of the signals that are generated by a DNA-captured nucleic acid.
The detection of biological species in complex systems is important for many biomedical and environmental applications. In particular, there is a strong interest in developing detection techniques and sensors for use in such applications as
infectious disease identification, medical diagnostics and therapy, as well as biotechnology and environmental bioremediation. An objective in developing new techniques and sensors is not only to be able to selectively identify target compounds but to
be able to assay large numbers of samples. Yet, there remain problems in reproducibly detecting and measuring low levels of biological compounds conveniently, safely and quickly.
A basic interest has been in the development of inexpensive biosensors for environmental and biomedical diagnostics. Biosensors have been investigated, mostly based on DNA probes and on various systems for analysis of oligonucleotide arrays, but
there appears to be limited consideration and development of integrated circuit (IC) gene probe-based biosensors on microchips. Existing systems typically employ photomultipliers or 2-dimensional detectors such as charge-coupled device (CCD) systems
which require bulky electronic and data conditioning accessories (Affymetrix.RTM. http, 1997; Schena, et al., 1995; Piunno, et al., 1995; Kumar, et al, 1994; Eggars, et al., 1994; and, Graham, et al., 1992).
There are several methods for selectively identifying biological species, including antibody detection and assay as in the well-known Enzyme-linked Immunosuppresent Assays (ELISA) employing molecular hybridization techniques. Generally speaking,
it is possible to identify sequence-specific nucleic acid segments, and to design sequences complementary to those segments, thereby creating a specific probe for a target cell, such as different pathogen cells or even mammalian cells that have mutated
from their normal counterparts. In principle, one can design complementary sequences to any identified nucleic acid segment. In many instances, unique sequences specific to an organism may be used as probes for a particular organism or cell type. The
quantitative phenotypic analysis of yeast deletion mutants, for example, has utilized unique nucleic acid sequence identifiers to analyze deletion strains by hybridization with tagged probes using a high-density parallel array (Shoemaker et al., 1996).
Hybridization involves joining a single strand of nucleic acid with a complementary probe sequence. Hybridization of a nucleic acid probe to nucleic acid sequences such as gene sequences from bacteria, or viral DNA offers a very high degree of
accuracy for identifying nucleic acid sequences complementary to that of the probe. Nucleic acid strands tend to be paired to their complements in double-stranded structures. Thus, a single-stranded DNA molecule will seek out its complement in a
complex mixture of DNA containing large numbers of other nucleic acid molecules. Hence, nucleic acid probe (e.g., gene probe) detection methods are very specific to DNA sequences. Factors affecting the hybridization or reassociation of two
complementary DNA strands include temperature, contact time, salt concentration, the degree of mismatch between the base pairs, and the length and concentration of the target and probe sequences. In perhaps the simplest procedure, hybridization is
performed on an immobilized target or a probe molecule attached on a solid surface such as a nitrocellulose or nylon membrane or a glass plate.
Despite significant strides in developing DNA chips, detection and analysis methods have not been well developed to take advantage of the amount of information that such chips can obtain in a short period of time. A common technique for
detecting DNA probes involves labeling the probe with radioactive tags and detecting the probe target hybrids by autoradiography. Phosphorous-32 (.sup.32 P) is the most common radioactive label used because of its high-energy emission and, consequently
short exposure time. Radioactive label techniques, however, suffer several disadvantages, such as limited shelf life. For example, .sup.32 P has a limited shelf life because it has a 14-day half-life.
Several optical detection systems based on surface-enhanced Raman fluorescence of visible and near-infrared (NIR) dye probe labels have been investigated (Vo-Dinh, et al., 1987 and Isola, et al., 1996) for non-radioactive detection of tagged gene
probes. Fluorescence detection is extremely sensitive when the target compounds or labeled systems are appropriately selected. Indeed, a zeptomole (10.sup.-21 mole) detection limit has been achieved using fluorescence detection of dyes with laser
excitation (Stevenson, et al., 1994). Even so, detection systems are macro compared to the micro world of DNA arrays, as many detection/analysis methods are mere adaptations from other systems. This means that analysis is relatively slow, compared to
data accumulation.
There is therefore a distinct need for development of systems that will allow rapid, large-scale and cost effective use of recently developed DNA biochips.
SUMMARY OF THE INVENTION
The invention in its broadest aspect comprises an integrated microchip biosensor device. Such a device employs multiple optical sensing elements and microelectronics on a single integrated chip combined with one or more nucleic acid-based
bioreceptors designed to detect sequence specific genetic constituents in complex samples. The microchips combine integrated circuit elements, electrooptics, excitation/detection systems and nucleic acid-based receptor probes in a self-contained and
integrated microdevice. A basic biochip, for example, may include: (1) an excitation light source; (2) a bioreceptor probe; (3) a sampling element; (4) a detector; and (5) a signal amplification/treatment system.
The integrated circuit biomicrochips of the present invention comprise an integrated circuit that includes an optical transducer and associated optics and circuitry for generating an electrical signal in response to light or other radiation
indicative of the presence of a target biological species, particularly a nucleic acid. The chip may also include a support for immobilizing a bioprobe, which is preferably a nucleic acid. In particular embodiments, a target nucleic acid may be tagged
or labeled with a substance that emits a detectable signal; for example, luminescence. Alternatively, the bioprobe attached to the immobilized bioprobe may be tagged or labeled with a substance that emits a detectable or altered signal when combined
with the target nucleic acid. The tagged or labeled species may be fluorescent, phosphorescent, or otherwise luminescent, or it may emit Raman energy or it may absorb energy.
The highly integrated biosensors of the present invention are advantageous in part because of fabricating multiple optical sensing elements and microelectronics on a single integrated circuit, and further combining the chip in preferred
embodiments with a plurality of molecular hybridization probes (Geiger, et al., 1990 and Aubert, et al., 1988). When the probes selectively bind to a targeted species, a signal is generated that is picked up by the chip. The signal may then be
processed in several ways, depending on the nature of the signal.
In one aspect, the present invention concerns an integrated system that includes (1) a targeted nucleic acid sequence in combination with a biological probe which is modified to receive light or other radiation of a first frequency and thereby to
emit light or other radiation of a different frequency than the first frequency, and (2) to detect the emitted radiation by means of a phototransducer. The target nucleic acid is typically a uniquely characteristic gene sequence of a pathogen such as a
fungus, bacteria, or virus, or other distinct nucleic acid species such as may be found in mutant mammalian cells or in individuals with inherited errors of metabolism. The target nucleic acid is modified or labeled to include a tag or label that emits
a signal upon exposure to an incident light or other radiation.
The target nucleic acid may be immobilized onto the integrated microchip that also supports a phototransducer and related detection circuitry. Alternatively, a gene probe may be immobilized onto a membrane or filter which is then attached to the
microchip or to the detector surface itself, such as the transducer detector described herein. This approach avoids the need to bind the bioreceptor directly to the transducer and thus is attractive for simplifying large-scale production
In one preferred embodiment of the invention, light of a highly directional or focused nature is impinged on a target nucleic that inherently or by virtue of an appropriate tag or label will emit a detectable signal upon irradiation. The
irradiation may be provided by a suitable light source, such as a laser beam or a light-emitting diode (LED). With the Raman, fluorescence and phosphorescence detection modes, the incident light is further kept separate from the emitted light using
different light paths and/or appropriate optical filters to block the incident light from the detector.
A target nucleic acid sequence is preferably hybridized with a nucleic acid sequence that is selected for that purpose (bioprobe). As stated earlier, the selected bioprobe is immobilized on a suitable substrate, either on the biochip itself or
on a membrane type material that is then contacted or attached to the chip surface. The bioprobe may be labeled with a tag that is capable of emitting light or other non-radioactive energy. Upon hybridization with a target nucleic acid sequence, the
hybrid product can be irradiated with light of suitable wavelength and will emit a signal in proportion to the amount of target nucleic acid hybridized, see FIG. 20. The labeled bioprobe may comprise a labeled molecular bioreceptor. Known receptors are
advantageous to use because of their known ability to selectively bind with the target nucleic acid sequence. In certain particular examples, the bioreceptor itself may exhibit changes in light emission when its cognate is bound.
In certain applications, it may be desirable to increase the amount of biotarget when only trace quantities are present in a sample. The present invention is compatible with polymerase chain reaction (PCR), which is a technique to amplify DNA
sequences.
BRIEF DESCRIPTION OF THE DRAWINGS
For a more complete understanding of the present invention and the advantages thereof, reference is now made to the following description taken in conjunction with the accompanying drawings in which like reference numbers indicate like features
and wherein:
FIG. 1 illustrates a schematic, exploded view of an example of a DNA biochip of the invention.
FIG. 2 illustrates a schematic diagram of one possible optical detector and amplifier circuit that may be implemented on an integrated circuit in order to convert an optical signal into an electrical signal suitable for data digitization and
capture by a computer.
FIG. 3 is a perspective, partially enlarged and exploded schematic view of a biochip having multiple arrays of exciting light sources and detectors.
FIG. 4 is a schematic, sectional view of an integrated circuit microchip system for a biochip with integrated light emitting diode excitation sources.
FIG. 5 illustrates a schematic circuit diagram for an integrated light emitting diode (LED) light source and phototransistor detection device.
FIG. 6 illustrates a schematic layout of an integrated circuit that implements the light source and detection device illustrated in FIG. 5.
FIG. 7 illustrates a physical layout of a large-area, 4.times.4 n-well integrated amplifier-photodiode array designed as a single, custom integrated circuit.
FIG. 8 illustrates a schematic cross-section of an n-well photodiode used in the photodiode array of FIG. 7.
FIG. 9 illustrates a detection circuit for use in conjunction with the photodiode array of FIG. 7.
FIG. 10 illustrates an alternative detection circuit for use in conjunction with the photodiode array of FIG. 7.
FIG. 11 illustrates schematically an analog multiplexer that enables any element in the photodiode array of FIG. 7 to be connected to an amplifier.
FIG. 12 illustrates one possible schematic implementation of the multiplexer in FIG. 11 for use with 16 cells.
FIG. 13 illustrates schematically a partially-parallel system that may be used to obtain data from the photodiode array shown in FIG. 7. The partially-parallel system has only a readout system for every row of photodetectors.
FIG. 14 illustrates a fully-parallel system that may be used to obtain data from the photodiode array shown in FIG. 7. The fully-parallel device has a read-out system (amplifier, electronics) for every photodetector.
FIG. 15 shows a calibration curve for an NIR dye-labeled single-stranded DNA (sequence: 5'-CCTCCTCCTTCCCAGCAGGG-3'; SEQ ID NO:1) over a concentration range from 1 pmol/.mu.L to 3 fmol/.mu.L.
FIG. 16 illustrates the results of measurements of gene probes tagged with fluorescein, a dye label emitting in the visible range.
FIG. 17 illustrates the performance of an integrated circuit microchip (ICM) phototransistor and amplifier circuit consisting of a 2 .mu.m, p-well CMOS process occupying an area of 160,000 square microns and 220 phototransistor cells connected in
parallel by showing the signal output response for various concentrations of the dye label Rhodamine-6G excited with a small helium-cadmium laser (8 mW, 325 nm).
FIG. 18 illustrates schematically an experimental setup that may be used to evaluate a 4.times.4 array amplifier/phototransistor integrated circuit microchip device.
FIG. 19 illustrates the results when four sample spots of 1 .mu.L of fluorescein labeled DNA were placed on a nitrocellulose membrane that was translated over a detection channel of the device shown in FIG. 18.
FIG. 20 illustrates the calibration curve of fluorescein labeled DNA using the device shown in FIG. 18.
FIG. 21 illustrates an embodiment of the present invention that is used for absorption and reflection measurements.
FIG. 22 is a schematic vertical section of a portion of a biochip of the invention that may operate with either luminescent or Raman radiation.
FIG. 23 is a schematic vertical section of a portion of a biochip that may employ luminescent or Raman energy for detection, or may operate by detecting the degree of absorption of a light beam.
FIG. 24 illustrates a plan view of one embodiment of the present invention that is used to detect samples from a microfluidic device via a plurality of cells.
FIG. 25 illustrates a schematic vertical section of a microfluidic injection system for liquid and gas sample in and out of the ICM chip used in the embodiment illustrated by FIG. 24.
FIG. 26 illustrates an embodiment of the present invention that is used to detect samples from a microfluidic device using an imaging lens, binary optics, or a lens array to image each microfluidic channel onto a detector element of the ICM.
FIG. 27 illustrates a micro-electromechanical system (MEMS) that is used to construct an ICM with Random Access Microsensing.
FIG. 28 illustrates an overview of an ICM system that uses the MEMS depicted in FIG. 27.
FIG. 29 illustrates an embodiment of the present invention that uses individually addressable, integrated vertical-cavity surface-emitting lasers (VCSEL) and on-axis and/or off-axis diffractive lenses.
FIG. 30 is an LED driver circuit where V.sub.1 is a DC voltage (5V, for example) and V.sub.2 is a DC voltage (5V to turn on LED and 0V to turn off LED) or a pulse train with 5V levels turning on the LED and 0V turning off the LED.
DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS
Definitions
An integrated circuit (IC) is a circuit comprised of elements such as transistors, resistors and capacitors fabricated in a single piece of semiconducting material, usually silicon or gallium arsenide. As used herein, "integrated circuit" not
only refers to the common definition but also to highly integrated structures including, for example:
1) multichip modules where several IC's and other circuit elements including molecular target probes may be combined compactly on a polymer, quartz, glass, sliver, ceramic or other substrates. In some cases, one IC may be the substrate with
other components, such as photodiodes or LEDs mounted on it;
2) Hybrid microcircuits where one or more IC's and other circuit elements are mounted on or several substrate(s); and,
3) Other compact electromechanical arrangements of a circuit comprising primarily one but possibly more IC's and other electronic components and microelectromechanised systems (MEMs).
Biosensor Probes
The development of biosensor technologies for detection of trace quantities of biological species in complex systems is important for many biomedical and environmental applications. Spectroscopic chemical sensors and biosensors have been
developed using laser induced fluorescence, room temperature phosphorescence, surface enhanced Raman spectroscopy, antibody based immunofluorescence and gene probe Raman sensing methods, including gene probes having surface-enhanced Raman scattering
labels to enhance the selectivity and sensitivity of chemical sensors and biosensors (Vo-Dinh et al., 1994).
The present invention uses spectroscopic techniques such as luminescence with visible and NIR labels is a useful detection scheme for gene biosensors without having the limitation of radioactive methods.
Non-radioactive probes, particularly gene probes, are desirable because of their selectivity in addition to avoiding the hazards involved with radioactive materials. Recognition and detection of biological species is based on the principle that
cell specific nucleic acid sequences can be specifically recognized and can be combined with a receptor that specifically binds with that species. Such receptors include, for example, antibodies, enzymes, cells, bacterial probes, complementary nucleic
acids, or nucleic acids that selectively hybridize with a cell-specific nucleic acid sequence. Receptors may be found and employed in the form of organelles, tissue components, chemoreceptors or even whole cells or microorganisms. Other types of
receptors may include biomimetic materials such as cyclodextrins, molecular imprint materials, etc.
Gene probes operate on a hybridization process. Hybridization involves joining of a single strand of nucleic acid with a complementary probe sequence. Hybridization of a nucleic acid probe to a biotarget such as bacterial or viral DNA or RNA or
selected gene segments, offers a high degree of accuracy for identifying nucleic acid sequences complementary to the probe. Nucleic acid strands tend to be paired with complementary strands, such as is typically found in double-stranded DNA structures.
Therefore, a single-stranded DNA (or RNA) will seek out its complement in a complex mixture of DNA containing large numbers of other nucleic acid molecules. Nucleic acid probe or gene probe detection methods are specific to DNA sequences. Factors
affecting the hybridization or reassociation of two complementary DNA strands include temperature, contact time, salt concentration, the degree of mismatch between the base pairs, and the length and concentration of the target and probe sequences.
Immobilization Techniques
Biologically active DNA probes may be directly or indirectly immobilized onto a transducer detection surface to ensure optimal contact and maximum detection. When immobilized onto a substrate, the gene probes are stabilized and therefore may be
used repetitively. In general terms, hybridization is performed on an immobilized nucleic acid target or a probe molecule is attached to a solid surface such as nitrocellulose, nylon membrane or glass. Numerous other matrix materials may be used,
including reinforced nitrocellulose membrane, activated quartz, activated glass, polyvinylidene difluoride (PVDF) membrane, polystyrene substrates, polyacrylamide-based substrate, other polymers such as poly(vinyl chloride), poly(methyl methacrylate),
poly(dimethyl siloxane), photopolymers (which contain photoreactive species such as nitrenes, carbenes and ketyl radicals capable of forming covalent links with target molecules (Saiki, et al., 1994).
Binding of the bioprobe to a selected support may be accomplished by any of several means. For example, DNA is commonly bound to glass by first silanizing the glass surface, then activating with carbodimide or glutaraldehyde. Alternative
procedures may use reagents such as 3-glycidoxypropyltrimethoxysilane (GOP) or aminopropyltrimethoxysilane (APTS) with DNA linked via amino linkers incorporated either at the 3' or 5' end of the molecule during DNA synthesis. DNA may be bound directly
to membranes using ultraviolet radiation. With nitrocellous membranes, the DNA probes are spotted onto the membranes. A UV light source (Stratalinker, from Stratagene, La Jolla, Calif.) is used to irradiate DNA spots and induce cross-linking. An
alternative method for cross-linking involves baking the spotted membranes at 80.degree. C. for two hours in vacuum.
Gene bioprobes may first be immobilized onto a membrane and then attached to a membrane in contact with a transducer detection surface. This method avoids binding the bioprobe onto the transducer and may be desirable for large-scale production.
Membranes particularly suitable for this application include nitrocellulose membrane (e.g. from BioRad, Hercules, Calif.) or polyvinylidene difluoride (PVDF) (BioRad, Hercules, Calif.) or nylon membrane (Zeta-Probe, BioRad) or polystyrene base substrates
(DNA.BIND.TM. Costar, Cambridge, Mass.).
Development of an Integrated Circuit Microchip (ICM)
ICM System and Design
An important component of an ICM is the detection method and spectral range of sensing system. In the present work, several optical detection systems based on fluorescence of visible and near-infrared (NIR) dyes have been investigated for
non-radioactive detection of tagged gene probes. Fluorescence detection means employing the integrated microchips of the present invention have been shown to selectively detect hybridized nucleic acids bound to the chip.. As shown previously, zeptomole
(10.sup.-21 mole) detection limit using fluorescence detection of dyes with laser excitation is possible (Vo-Dinh et al., 1994). The miniaturization of optical biosensors is facilitated by the versatility of waveguide configurations. FIG. 3A shows one
configuration of a nucleic acid detection biochip. Various configurations and uses of the DNA biochips are illustrated in FIG. 3B.
ICM System Using Integrated Phototransistors
The instrumental system discussed herein includes the design of integrated electrooptic sensing photodetectors for the biosensor microchips. Highly integrated biosensors are made possible partly through the capability of fabricating multiple
optical sensing elements and microelectronics on a single integrated circuit (IC).
FIG. 3A, FIG. 3B, and FIG. 4A show an example of such integration. This figure schematically shows a two-dimensional array of optical detector-amplifiers integrated on a single IC chip. The insert in this figure shows that each optical detector
is a phototransistor coupled to a transimpedence amplifier followed by an amplifier. This block is repeated several times on the IC chip and combined with other electronic elements such as filters and amplifiers, which can also be integrated on the same
IC.
The operational amplifier used with the phototransistor is a two-stage, unbuffered amplifier. The circuit is compact, occupying an area of only 185 .mu.m.times.200 .mu.m, yet has moderately high performance. It was designed to be useful for
wide-band amplification and low-level signals. The gain-bandwidth product is 70 Mhz and the amplifier is stable for gains greater than 10. Other typical characteristics include: input offset voltage less than 5 mV, DC gain of 220, positive slew rate of
80 V/.mu.s and negative slew rate of 9V/.mu.s. The circuit requires 2.5 mW from a single 5-V supply. In the preferred embodiement this IC chip performed the complete conversion from an optical signal to an electrical signal suitable for data
digitization and capture by a computer.
FIG. 4B shows the physical layout of the phototransistor and amplifier circuit. The circuit was fabricated in a 2-.mu.m, p-well CMOS process and occupied an area of 160,000 square microns. The phototransistor is composed of 220 phototransistor
cells connected in parallel. An individual phototransistor cell occupied 760 square microns. The transimpedence amplifier had a gain of 100 kV/A. As the phototransistors were coupled with the 10-fold amplifier gain, the resulting gain was 10.sup.6 V/A.
The phototransistors had a conversion gain on the order of 10 .mu.A/.mu.W, so the entire chain had an approximate conversion gain of 10 V/.mu.W. The exact gain generally depends on the spectral region of interest and, to some extent, on the signal level
being monitored.
The above described elements may be modified to tailor the devices to specific applications. Since the phototransistor is made from basic photocell elements, it can be connected to as many cells as needed to create the desired geometry or
required number of channels to adapt the detector to a specific application. Similarly, the gain and bandwidth of the amplifiers can be adjusted using simple resistor or capacitor changes as the application requires. Other light sensing structures in
addition to the phototransistor may be fabricated using standard Complementary Metal Oxide Semiconductor (CMOS) processing steps. Several photodiode structures are possible using the p-n junctions that would ordinarily form wells or transistors.
ICM System with Integrated Phototransistors
A biochip having multiple arrays of exciting light sources and detectors is shown in FIG. 3. A design of an ICM system for such a biochip with integrated light emitting diodes (LED) excitation sources is illustrated in FIG. 4. This ICM system
contains both GaAs chips used as light sources and the phototransistor as the detector. The electrooptic circuit and layout of this device shown in FIG. 5 and FIG. 6 FIG. 5 and FIG. 6 show the schematic diagram of the IC circuit and layout for the IR
light source and the analog signal detection, respectively.
ICM System with 4.times.4 N-Well Photodiode Array
A second ICM system includes large-area, 4.times.4 n-well integrated amplifier-photodiode array that has been designed as a single, custom integrated circuit (IC), fabricated for the biochip. This IC device is coupled to the biosensor and is
designed for monitoring very low light levels.
The physical layout of the IC array is illustrated in FIG. 7 shows the individual photodiodes have 0.9-mm square size and are arrayed on a 1-mm grid. The photodiodes and the accompanying electronic circuitry were fabricated using a standard
1.2-micron n-well CMOS process from Orbit Semiconductor (Sunnyvale, Calif.). The use of this type of standard process allows the production of photodiodes, phototransistors as well as other numerous types of analog and digital circuitry in a single IC
chip. The photodiodes themselves are produced using the n-well structure that is generally used to make resistors or as the body material for a PMOS transistor. FIG. 8 shows a schematic cross-sectional drawing of the n-well photodiode. Since the anode
of the diode is the p-type substrate material, which is common to every circuit on the IC chip, only the cathode is available for monitoring the photocurrent and the photodiode is constrained to operate with a reverse bias.
FIG. 9 shows a one-diode version of the circuit. The operational amplifier and feedback resistor R.sub.1 (227) form a transimpedence amplifier that is used to convert the photocurrent into a voltage. The conversion gain (V/A) is determined by
the value of rasistor 227. The feedback capacitor performs two functions: (a) it prevents the amplifier circuit from oscillating, and (b) it limits the bandwidth of the circuit. Generally, the bandwidth should be no more than that necessary for the
desired measurement. Reduction of the bandwidth decreases the noise present at the amplifier's output, so if this reduction can be accomplished without attenuating the signal, a net gain in signal-to-noise ratio is achieved. For this circuit, the
signal bandwidth is given by f=1/(2.pi.R.sub.1 C.sub.1), where R.sub.1 is the resistance and C.sub.1 is the capacitance.
The voltage applied to the non-inverting input of the operational amplifier determines the reverse bias applied to the photodiode. The IC can be operated with a single %--v supply--For example, if 2 V is applied to the non-inverting input, then
the dc level of the other input and the output will also be 2 V, so the reverse bias on the diode will be 2 V. Photocurrents flowing into the diode will also flow through the feedback resistor, causing the amplifier output to become more positive. As
the operational amplifier output cannot exceed the positive supply, the maximum output will be approximately 5 V, so the maximum signal excursion is 3 V, which corresponds to a maximum current of 3 V/R1.
Alternative photodiode and integrated amplifier circuit block diagrams are shown in FIG. 10. Alternate to the transimpedence amplifier plus low-pass filter readout method is the use of an integrating amplifier as shown in FIG. 10. In this case,
the amplifier integrates the current from the photodiode until the signal is converted to digital format. After conversion, the integrator is reset to its initial state and is capable of starting another measurement. This scheme has the advantage that
the integration time can be controlled to allow adapting to various light (and therefore current) levels. The disadvantage of this scheme is that it requires coordination between the analog-to-digital conversion process and the integrator.
An analog multiplexer is designed to allow any of the elements in the array to be connected to an amplifier. In a specific embodiment, each photodiode could be supplied with its own amplifier. For this many applications, this feature is not
necessary unless the additional data acquisition speed due to having parallel channels is required. The multiplexer is made from 16 cells as illustrated in FIG. 11. Each cell has two CMOS switches that are controlled by the output of the address
decoder cell. Each cell has a unique 4-bit address. One switch is closed only when that particular cell is the one that is being addressed while the other switch is closed, except when that cell is the one being addressed. This process connects the
addressed diode to one amplifier while all the others are connected in parallel to the other amplifier as shown in FIG. 12.
This arrangement allows connecting a 4.times.4 array of light sources (different fluorescent probes, for example) to the photodiode array and reading out the signal levels sequentially. With some modification, a parallel reading system can be
designed. Using a single photodiode detector would require mechanical motion to scan the source array. The additional switches and amplifier serve to correctly bias and capture the charge generated by the other photodiodes. Failure to do this would
result in erroneous measurements due to the addressed photodiode collecting current generated from elsewhere in the IC. Additionally, the additional amplifier and switches allow using the IC as a single, large area (nearly 4 mm square) photodetector.
Alternative Photodetectors
An avalanche photodiode (APD) provides an alternative solid-state method of detecting low light levels. Advantages of APD arrays over ordinarily used photodiode arrays include electron multiplication obtained by the avalanche process. This may
improve the signal-to-noise ratio so that lower light levels are detected.
The process of making an APD and/or APD arrays from silicon includes fabricating structures that are not compatible with steps used in standard CMOS processing. A fully integrated microchip including avalanche photodiodes requires special
semiconductor fabricating processes. Such processes are known in the art (Geiger, et al., 1990 and Aubert, et al., 1988).
Alternate Photodiode Array Readout Schemes
There are several alternate possibilities for reading out an array of photodiodes in addition to the multiplexed scheme that was implemented. For many applications, the low pass filter time constant is set to a large value to improve the
signal-to-noise ratio. This requires a long time to acquire data for the whole array if each diode is read sequentially. For example, if the time constant is set to 1 second and the array is 4.times.4, then the minimum time to acquire data would be
something 4.times.4.times.1 second.times.5=80 seconds. The last factor of 5 allows for the amplifier and low pass filter to settle to better than 1% accuracy after its input is switched to another photodiode.
Partially Parallel Readout Scheme. According to one embodiement of the invention, the row of diodes are multiplexed into one amplifier/low pass filter circuit. Columns or other subunits of the array could be used as well. The output of each
filter is input to a multi-channel analog-to-digital converter (ADC), which can be implemented on the same IC as the photodiode array, amplifier and filter. Compared to the sequential or serial readout case, the time required is reduced by a factor of n
for an n.times.n array. The schematic block diagram of this partially parallel readout system is shown in FIG. 13.
Fully Parallel Readout Scheme. This provides the fastest readout speed. Each diode is provided with its own amplifier, low-pass filter and ADC input channel. Compared to the serial readout case, the time required is reduced by a factor of
n.sup.2 for an n.times.n array. The block diagram at a fully parallel readout system is shown in FIG. 14.
Photodetector Array Sizes
One advantage of a custom biosensor IC is that the photodiodes can be made to physically match the probe. The prototype uses 0.9-mm square photodiodes on a 1 mm.times.1 mm square grid, but arrays with a larger number of smaller photodiodes can
be made. Using readily available 1.2-micron technology, photodiodes on a 20-micron grid could be made for a 10.times.10 array. This would give 100 photodiodes in an area only 0.04 mm.sup.2, i.e. 2500 photodiodes per mm.sup.2. Even more density is
possible using 0.5-micron processes that are commercially available.
Using the serial readout scheme allows most of the chip area to be used for the photodiode array. It is likely that a 1000 element array will fit on a 5 mm.times.6 min IC, which would be considered a medium-sized die format. For large arrays
using the fully parallel readout scheme, the limiting factor on IC area is most likely the area required by the wiring and the readout electronics. Future advances in IC technology could increase the density of elements on the chip.
The partially parallel readout scheme is a compromise between the other two cases. Compared with the serial case for a large array, the die size might double to allow partially parallel readout.
Alternative Detection Circuits
Previous examples have shown use of a transimpedence amplifier to convert the current to a voltage and low-pass filters to improve signal-to-noise ratio for very low frequency or dc signals. Such filters may be implemented as digital filters or
for different types of signals, such as a fluorescence or phosphorescence decay. A filter matched to the signal could be used to give optimal identification of the signal.
In another embodiment according to the invention, present low current levels may also be measured by other means such as integrating the current for a fixed time prior to measuring the voltage. Alternatively, one may integrate the current until
a predetermined voltage level is reached and measure the time required for the integration. These methods use the following relations allowing the determination of current from measured voltages, capacitances and times.
charge=current.times.time
charge (on a capacitor)=capacitance.times.voltage
The circuit used to integrate current is similar to the transimpedence amplifier shown in FIG. 9, except that the feedback element of the amplifier is a capacitor shown as 241 in FIG. 10 instead of a resistor.
Another integrating amplifier possibility is to make an amplifier that amplifies the voltage developed across a capacitor which receives the charge.
In yet another embodiment according to the present invention, another method of using an integrating amplifier is to employ an oscillator. The integrator integrates the unknown current to a preset voltage and then is reset by discharging the
capacitor with a switch and the integrator is allowed to start again. This then cycles repeatedly. The frequency of oscillation is proportional to the input current. This common technique in circuit design may be used in the systems disclosed herein.
Several methods may be employed to implement the integrating capacitor for the described detection circuits. One may use a capacitor such as polysilicon that is compatible with CMOS technology. Or one may take advantage of the capacitance of
the photodiode itself, using the amplifier to amplify the voltage across the photodiode. This has the advantage that fewer components are needed compared to using separate integrating capacitors. For an array of photodiodes, all photodiodes may be
integrated simultaneously and one (fast) amplifier used to multiplex the outputs without depriving the circuit of significant measurement time.
Excitation Light Sources
Light sources such as light-emitting diodes (LEDs) and semiconductor lasers may be used in connection with the integrated microchips herein described. One may also choose to employ alternative microlaser systems such as edge-emitting lasers and
surface emitting lasers. Vertical-cavity surface-emitting (VCSELs) are particularly suitable light sources for integrated microchips. The linearity of laser arrays makes them ideal for compact 2-dimensional and 3-dimensional configurations in ICM
systems. Quantum Cavity (CQ) lasers can also be used due to their small sizes. The QC lasers provide powerful mid-infrared sem | | |
