The present invention relates to photosensing with photosensor
arrays on integrated circuits (ICs).
U.S. Patent No. 6,580,507
describes a muitiple-longitudinal flow cell channel system in which an
array detector is positioned to monitor radiation from at least two of multiple
flow cell channels, at separate groupings of pixels on the detector.
It would be advantageous to have improved techniques for
sensing light emanating from channels or moving objects.
In accordance with a first aspect of the present invention,
a sensor comprises a fluidic structure having defined therein a set of one or more
channel portions within each of which fluid can move in a respective flow direction
and photons can emanate from the moving fluid, with photons emanating from the moving
fluid in each channel portion having energies within a respective range of photon
an IC that includes a photosensor array; the array including, for at least one of
the channel portions in the set, a respective series of two or more sets of cells;
each series extending along the respective channel portion in its flow direction;
each set of cells in a channel portion's series being capable of photosensing photons
within a respective subrange of the channel portion's range of photon energies;
the respective subranges of at least two of the sets of cells being different from
In accordance with a second aspect of the present invention,
a method of detecting photon energies emanating from objects comprises causing one
or more objects to travel along a path while photons within a range of photon energies
emanate from the objects; and
along each of two or more segments of the path, photosensing quantity of a respective
subset of photons from the objects using a respective set of one or more cells in
a photosensor array included in an IC; the subset of photons of each segment being
within a respective subrange of the range of photon energies; the respective subranges
of at least two of the segments being different from each other.
These and other features and advantages of exemplary embodiments
of the invention are described below with reference to the accompanying drawings,
- Fig. 1 is a schematic diagram of an analyzer on a fluidic structure.
- Fig. 2 is a schematic cross-sectional view of the analyzer in Fig. 1, taken
along the line 2-2.
- Fig. 3 is a schematic plan view of an implementation of an assembly that can
be used in Fig. 2.
- Fig. 4 is a schematic cross-sectional view of another implementation of an assembly
that can be used in Fig. 2.
- Fig. 5 is a graph illustrating laterally varying light transmission properties
of a transmission structure in Fig. 4.
- Fig. 6 is a schematic cross-sectional view of another implementation of an assembly
that can be used in Fig. 2.
- Fig. 7 is a graph illustrating the laterally varying light transmission properties
of a transmission structure in Fig. 6.
- Fig. 8 illustrates a technique that produces a transmission structure that can
be used in an assembly as in Fig. 2.
- Fig. 9 illustrates another technique for producing a transmission structure
that can be used in an assembly in Fig. 2.
- Fig. 10 is a flowchart showing general operations that can be performed in producing
an analyzer as in Fig. 1.
- Fig. 11 is a schematic diagram of an alternative implementation of part of an
analyzer on a fluidic structure as in Fig. 1.
- Fig. 12 is a schematic cross-sectional view of the alternative implementation
in Fig. 11, taken along the line 12-12.
- Fig. 13 is a schematic plan view of a portion of an alternative implementation
of an analyzer as in Fig. 1.
- Fig. 14 is a schematic cross-sectional view of an alternative implementation
of the sensing component in Fig. 2.
- Fig. 15 is a schematic cross-sectional view of an alternative implementation
to that of Fig. 14.
- Fig. 16 is a schematic cross-sectional view of another application of the alternative
implementation of Fig. 15.
- Fig. 17 is a schematic block diagram of a system that can control the analyzer
of Fig. 1.
- Fig. 18 is a flow chart showing general operations implementing the detect,
readout, and combine routine of Fig. 17.
In this specification, "light" refers herein to electromagnetic
radiation of any wavelength or frequency; unless otherwise indicated, a specific
value for light wavelength or frequency is that of light propagating through vacuum.
The various exemplary implementations described below address
problems that arise in obtaining information about light. One of those problems
is the difficulty of obtaining spectral information about light emanating from moving
objects or from a channel in a fluidic structure rapidly and without bulky, expensive
equipment. This is extremely important for optical characterization of moving objects
that are hard to capture (or which it is not desired to capture) and that must be
measured while moving. In addition, optical techniques for identifying particles
obtain limited information and are constrained by weak interaction with excitation
Some of the photosensing implementations described herein
employ structures with one or more dimensions smaller than 1 mm, and various techniques
have been proposed for producing such structures. In particular, some techniques
for producing such structures are referred to as "microfabrication." Examples of
microfabrication include various techniques for depositing materials such as growth
of epitaxial material, sputter deposition, evaporation techniques, plating techniques,
spin coating, printing, and other such techniques; techniques for patterning materials,
such as etching or otherwise removing exposed regions of thin films through a photolithographically
patterned resist layer or other patterned layer; techniques for polishing, planarizing,
or otherwise modifying exposed surfaces of materials; and so forth.
Fig. 1 shows schematically some components of analyzer
10 on support structure 12, a fluidic structure. Defined in support structure 12
is serpentine channel 14 through which an object 16 can travel, carried by a fluid
or other appropriate substance. Object 16 can, for example, be a droplet or small
volume of fluid that includes an analyte to be analyzed.
Examples of objects that could occur in implementations
as described below include droplets, small volumes of fluid, single molecules, agglomerated
molecules, molecule clusters, cells, viruses, bacteria, proteins, DNA, microparticles,
nanoparticles, and emulsions. A droplet or small volume of fluid may, for example,
include atoms, molecules, or other parades that emit light spontaneously or in response
to excitation; a particle could be a "fluorescent component" of a droplet, fluorescing
in response to excitation. Or a droplet may include particles that scatter light
incident on the droplet in a way that depends on photon energy, so that the droplet
scatters the incident light correspondingly; in this case, a particle could be a
"scattering component" of a droplet. An analyte (i.e. a chemical species being investigated)
in a droplet can act as a fluorescent, absorbent, or scattering component.
Object 16 enters channel 14 carried by a primary fluid
illustrated by arrow 20, and can enter from a supply reservoir (not shown) and a
sample well (not shown), with its entry into the primary fluid controlled by metering
electrodes 22. Rather than electrical metering, as with electrodes 22, pressure
metering could be used. Other techniques could provide a droplet or other small
object to channel 14 such as capillary forces or electro-osmotic flow pumps.
Analyzer 10 could be implemented with any appropriate number
of channels similar to channel 14, and with each channel receiving analyte samples
from a respective sample well. Furthermore, each of the channels could have a different
combination of components suitable to a specific type of analysis such as fluorescence
spectroscopy, laser-induced fluorescence spectroscopy (LIF), absorption spectroscopy,
excitation spectroscopy, Raman scattering, surface-enhanced Raman scattering (SERS),
far-infrared spectroscopy, etc. The channels could be formed by subdividing a broad
channel into several parallel channels.
Additional fluid to carry object 16 may enter as shown
by arrow 24. The path followed by the fluid in channel 14 can be controlled through
a number of devices. For example, the fluid, together with object 16 if appropriately
positioned, can be purged at two outlets as illustrated by arrows 26 and 28 through
toggling of valves 30 and 32, respectively, each of which is at a bifurcation junction.
Other types of gates could be used, such as electric fields to selectively deflect
objects; charged particles could be deflected by Coulomb force, and polarizable
particles could be deflected by dielectrophoretic force. Fluid can also be purged
at a final outlet from channel 14, illustrated by arrow 34.
The flow of the fluid can be maintained by conventional
propulsion components such as electro-osmotic pumps 40 along the length of channel
14. A wide variety of other propulsion components could be used, including, for
example, gas pressure pumps, positive displacement pumps, micro-peristaltic pumps,
electro-kinetic pumps, piezo pumps, and thermal mode pumps. In addition to maintaining
flow of fluid, propulsion components can also perform system flush and initial fluid
loading functions, with pressure driven techniques. Appropriate circuitry can coordinate
the various pumps and other components to work in a synchronized manner.
Along channel 14 is a series of sensing components, each
of which obtains information about object 16 as it travels within a respective straight
portion of channel 14; the straight portions are separated by 180-degree curved
portions. Coulter counter 50 and Mie scatter sensor 52, for example, are conventional
sensing components, illustratively along parts of one straight portion of channel
14. Coulter counter 50 is an example of an electrically based particle size detector.
Mie scatter sensor 52 is an example of an optical detector that relies on particle-induced
scattering of light entering from the side of channel 14.
Coulter counter 50 can be implemented to size particles
in the 1-10µm range within a continuous liquid stream. The Coulter counter
technique should also work for other particle sizes as long as the inner diameter
of channel 14 in the sensing region is not more than an order of magnitude larger
than the particles being measured.
The series of sensing components also includes optical
(e.g. visible or infrared) absorption sensing component 54, first fluorescence sensing
component 56, second fluorescence sensing component 58, and Raman scatter sensing
component 60. Analyzer 10 could include any other suitable combination of sensing
components, including some that are not connected in series. Additional sensing
components could include conventional optical or electrical trigger elements that
provide a signal indicating when an analyte with properties meeting certain criteria
moves past a position along channel 14. Furthermore, it may be possible to include
sensing components for electrical impedance spectroscopy (EIS) for electronic pathology
rather than sensing differential resistance for bioparticle sizing.
A series of sensing components as in Fig. 1 makes it possible
to obtain spectral information about moving particles or other objects in order
to achieve orthogonal characterization and reliable identification. Characterization
is orthogonal if sensing components obtain information about orthogonal characteristics
of a moving object, such as by photosensing different ranges of photon energies;
sensing components could also be suitable for different intensity ranges. By choosing
suitable materials, it is possible to obtain spectral information for the entire
range from the deep ultraviolet to the far infrared or even for frequencies in the
Analyzer 10 can be designed to perform multi-signal analysis
for a specific application, whether high wavelength resolution or broadband detection
is desired. The technique illustrated in Fig. 1 also takes advantage of the motion
of object 16 with a geometry that enables long integration times without sacrificing
throughput capacity. Highly sensitive optical characterization methods can be used,
such as fluorescence spectroscopy (illustratively in more than one range of photon
energies) and Raman spectroscopy. The use of multi-signal analysis makes it possible
to perform reagentless bioagent identification.
Each of sensing components 54, 56, 58, and 60 includes
a respective one of ICs 64, 66, 68, and 70. Each of these ICs includes a photosensor
array, and the sensing component includes a set of cells of the photosensor array.
The set of cells photosenses photons within a range of photon energies; for example,
the sets of cells in ICs 66 and 68 could photosense different ranges of photon energies
in the visible to ultraviolet range, and, as noted above, the set of cells in IC
70 could photosense in the infrared. Furthermore, more than one IC, such as ICs
66 and 68, could photosense fluorescing photons that are in the same energy range,
but that result from excitation at different wavelengths such as from different
LED or laser light sources. The set of cells for each of sensing components 54,
56, 58 and 60 includes a subset of cells, each of which photosenses in a respective
subrange, and the subranges of at least two of the cells are different from each
Sensing components 56, 58, and 60 can each be implemented
with any suitable excitation or illumination technique to cause emanation of light
from objects. Enhanced light-target interaction is especially important if analyzer
10 is characterizing single particles or low concentrations of biological or chemical
agents. In ICs 66, 68, and 70 are supported on spacers 72, providing a suitable
gap between each IC and the respective portion of channel 14 to avoid interference
with anti-resonant waveguiding.
Suitable anti-resonant waveguide configurations can include
an aerosol in a glass capillary tube or a liquid film between glass slides. The
excitation could be any appropriate electromagnetic radiation.
Additional background suppression of excitation light can
be obtained using a wavelength filtering component as part of the wall of channel
14 or as an additional coating on top of a photosensor array.
Fig. 2 shows schematically a cross-section of analyzer
10 taken along the line 2-2 in Fig. 1. Similar features would be found in first
fluorescence sensing component 56 and, to an extent, in Raman scatter sensing component
As object 16 travels through portion 80 of channel 14 in
the downstream direction indicated by arrow 82, it receives light from an excitation
component, illustratively light source 84 which could be a laser or an LED. Portion
80 can function as an anti-resonant waveguide in response to light from source 84,
or it can function in another way that provides enhanced light-target interaction.
For example, other techniques that provide continuous excitation to a fluorescing
molecule include tracking the molecule in motion with a scanning laser beam; using
a linear array of LEDs to sustain particle excitation along its path; arranging
a collimated beam along the particle path without waveguiding; and providing a Fabry-Perot-style
cavity in which light passes through the medium containing the particle several
Sensing components using anti-resonant waveguide modes
are especially advantageous in combination with fluidic devices because the fluidic
channels themselves can be used as anti-resonant waveguides in various configurations.
Examples of configurations include an aerosol carrying analytes in a capillary,
a liquid film carrying analytes within a channel or between glass slides, etc.
In response to light from source 84, an analyte within
object 16 fluoresces, emitting light with a characteristic spectrum of photon energies.
A portion 86 of the light is emitted toward assembly 87, which includes at least
IC 68 and possibly also other structures. Photons in portion 86 can therefore be
photosensed by cells of a photosensor array on IC 68. Assembly 87 is positioned
so that the photosensor array on IC 68 is close to and parallel to the path of object
16 through portion 80, to increase light collection efficiency.
The term "path" is used herein to refer to a substantially
continuous series of positions from which light may emanate (i.e. an "emanation
path") or at which light is incident on a photosensor array (i.e. a "photosensing
path"). A part of a path is referred to herein as a "segment", and segments may
overlap or be included one in another.
A photosensor array is "positioned along" or "along" a
path or a segment if the array is positioned near the path or segment in such a
way that one or more of its photosensors can photosense light emanating from the
path or segment.
Assembly 87 is illustratively supported on spacers 72 to
avoid disturbing anti-resonant waveguiding in portion 80 of channel 14. Spacers
72 are positioned outside portion 80, and, as a result, air gap 88 below assembly
87 prevents disturbance of waveguiding because air has a lower refractive index
than that of the liquid within the waveguide. A thin gap, layer, or film that is
only a few microns thick, e.g. 10 µm, is sufficient to prevent disturbance
of waveguiding if it has a sufficiently low refractive index.
Because object 16 receives excitation continuously throughout
portion 80, fluorescence also occurs continuously along the photosensor array. As
a result, spectral information is collected continuously as object 16 moves through
The structure shown in Fig. 2 could also be used to implement
Raman scatter sensing component 60 in a way that, although not comparable to dedicated
Raman sensors, may provide acceptable performance and resolution with sufficient
spectral range for a given application such as for specific Raman bands of interest.
The output signal could indicate a set of intensity ratios of selected Raman lines
and/or certain narrow intervals of a Raman spectrum rather than a complete Raman
To implement a Raman scatter sensing component as shown
in Fig. 2, it would be necessary that light source 84 and IC 68 meet appropriate
specifications, especially with regard to sensitivity and background light suppression
within analyzer 10. In addition, suitable optical elements would be necessary between
channel 14 and the photosensor array of IC 68 to ensure efficient and suitable light
Exemplary differences between a fluorescence sensing component
and a Raman scatter sensing component would be as follows: A fluorescence sensing
component could include a photosensor array in which cells photosense within a wide
spectral range with rather low resolution, e.g. 400-700 nm with a moderate wavelength
resolution of 2-5 nm. In contrast, a Raman scatter sensing component could include
a photosensor array in which cells photosense within a smaller spectral range close
to the excitation wavelength but with greater resolution, e.g. 800-830 nm with a
resolution of 0.2-0.5 nm or even higher resolution. The sensing range for Raman
scatter sensing must be set in accordance with typical energies of Raman scattered
photons, which are 100 cm-1 to a few 1000 cm-1 wavenumbers
different from the excitation photon energy, where wavenumber k=2&pgr;/&lgr;
in units of 1/cm.
Fig. 2 also illustrates one of the ways in which support
structure 12 could be implemented. Support layer 90 could, for example, be a light-transmissive
glass or silicon substrate. Channel 14 can be defined in a micromolded layer 92
of polydimethylsiloxane (PDMS). In patterning layer 92 and other layers in Fig.
2, the length of channel 14 in which light-target interaction occurs can be chosen
to minimize interference between different analytes.
Techniques for producing a patterned layer of PDMS include,
for example, fabricating a template on glass from SU-8 polymer, and then depositing
PDMS to form a patterned structure within the template. The template can then be
removed. Over layer 92 is plate 94, such as glass and therefore another example
of a light-transmissive structure.
Alternatively, techniques could be used that etch glass
to produce channels. Also, channels could be microfabricated by patterning a layer
of a polymer material such as SU-8 to produce high aspect ratio channel walls. Depending
on the medium that carries analyte through channel 14, parameters of channel 14
can be modified for optimal results.
Other dimensions of the structure shown in Fig. 2 can be
changed to obtain desired results, For instance, the thicknesses of layers 90 and
94 can bear a desired relationship to the height of channel 14, depending on various
constraints, including stability requirements, manufacturing convenience, and, as
noted below, the need to accommodate a desired flow of fluid and objects through
channel 14. Thicknesses of layers 90 and 94 are often greater than or approximately
equal to the height of channel 14. Typical thicknesses range between approximately
100 µm and 2 mm. Channel height ranges from approximately 1 µm or less
up to a few mm.
A specific parameter of channel 14 that can have significant
effects is adhesiveness of the channel wall,
Various techniques can be used to reduce sample loss through
surface adhesion. More specifically, an anti-adhesive coating can be applied to
prevent bioparticles and other analytes from sticking to the walls. Dip-coated polyethylene
glycol (PEG) is a good choice for preventing adhesion of most biomaterials and can
maintain capillary force on aqueous solutions. Other coatings may be viable options
depending on sample properties, material interface chemistry, and operating conditions
and regimes; for example, parylene C or vapor deposited tetraglyme might provide
appropriate coatings. As noted above, anti-resonant waveguiding techniques can employ
a channel with an inner diameter at least up to approximately 1.0 mm and possibly
greater, in which case adhesion is not as great a problem and clogging is unlikely
to occur; nevertheless, anti-adhesion measures may be advantageous to prevent spurious
background signals emitted by material adhering to a channel's inner wall.
Fig. 2 also shows optical component 96 on the side of support
layer 90 opposite PDMS layer 92. Optical component 96 provides an appropriate surface
98 through which light from source 84 can be coupled into the anti-resonant waveguide
within portion 80 of channel 14. Support layer 90 and optical component 96 could
instead be fabricated from a single layer of material by suitable processes.
Fig. 3 is a schematic view of an implementation of assembly
87 in which IC 68 includes photosensor array 100 and also has spacers 72 attached
to it. Photosensor array 100 is illustratively a two-dimensional array, with at
least two rows of cells that include photosensors.
Different rows or other parts of photosensor array 100
can be provided with different coatings or can be otherwise structured so that their
cells photosense different ranges or subranges of photon energies. As a result,
the information obtained from a single IC can provide a detailed analysis of incident
photons over a broad range of photon energies. In addition, reference cells, such
as the cells in row 102, can be used to provide a spatially resolved real-time reference
A feature of array 100 is that it includes one or more
reference cells that are nearby to a subrange cell.
Each cell in row 102 photosenses photons throughout a suitable
range, characterized as &lgr;all, to produce a reference for a nearby
cell in row 104. For implementations in which it is advantageous to have signal
strengths of the same order from a cell in row 102 and its paired cell in row 104,
the cell in row 102 must be different from the cells in row 104. For example, it
could have a different sensing area or it could have a gray filter coating different
than a coating over the paired cell in row 104.
Each cell in row 104, on the other hand, photosenses a
respective subrange between &lgr;min and &lgr;max, with
illustrative cell 106 photosensing a subrange centered around &lgr;p.
IC 68 also includes array circuitry as well as peripheral circuitry 110 which perform
various functions relating to readout of photosensed information from array 100.
One advantage of the technique illustrated in Fig. 3 is
that IC 68 provides a compact photosensor array that can be used for various functions
within a system such as analyzer 10. The compactness of IC 68 also allows for an
interactive detection scheme. Subsequent or adjacent ICs within analyzer 10 may
exchange information or trigger events. The combination of analysis results from
several ICs within analyzer 10 may help to obtain orthogonal information and ultimately
enable reliable identification of object 16.
Fig. 4 illustrates another implementation of assembly 87,
showing in greater detail how cells of an array photosense subranges. Assembly 87
as in Fig. 4 can be supported over air gap 88 by spacers 72.
In Fig. 4, a cross-section has been taken through a fragment
150 of a photosensor array, with cells 152 of the fragment 150 shown schematically
in cross-section. Over cells 152 is a transmission structure 160 that receives incident
light 162, such as from an optional Selfoc® or other gradient index (GRIN)
lens array, illustrated by lens array portion 164. Lens array portion 164 can be
designed to receive light from air gap 88 as in Fig. 2 and to provide a parallel
beam to structure 160, increasing spectral resolution.
Transmission structure 160 can be a film with laterally
varying light transmission properties. In the portion of transmission structure
160 shown in Fig.4, wedge-shaped transmissive cavity 170 is enclosed between reflective
films 172 and 174, forming a wedge-shaped Fabry-Perot etalon. Because its thickness
varies as a function of position along the x-axis, transmission structure 160 will
transmit different wavelengths as a function of position along the x-axis.
Transmission structure 160 can be produced with appropriate
coatings on or over a photosensor array. Films 172 and 174 and cavity 170 could
all be produced by exposure to deposition beams in an evaporation chamber; uniform
thicknesses could be produced by appropriate on-axis deposition, while laterally
varying thickness can be produced by appropriate off-axis deposition. Fig. 4 illustratively
shows films 172 and 174 as relatively thick compared to cavity 170, which would
be appropriate for layers of non-metallic material such as SiO2, TiO2,
or Ta2O5. If films 172 and 174 are reflective metal, however,
they could be much thinner.
Specific thicknesses of cavity 170 and films 172 and 174
could be designed from the desired transmitted wavelength &lgr; and the refractive
index n of cavity 170. The thickness of cavity 170 is typically chosen to be &lgr;/(2n)
or an integer multiple thereof, while the thicknesses of Bragg mirror layers within
films 172 and 174 are typically &lgr;/(4n). The number of pairs of such layers
in each of films 172 and 174 can vary between a few (e.g. 2-5) all the way up to
20 or 30, depending on the difference in refractive index between the two materials
used, the desired transmission band width, and the desired stop band reflectivity.
Therefore, in typical implementations, films 172 and 174 are much thicker than cavity
Fig. 5 illustrates the laterally varying light transmission
properties of transmission structure 160. Because its thickness varies as a function
of position along the x-axis, cavity 170 transmits different wavelengths as a function
of position along the x-axis. Wavelengths of photons predominantly transmitted to
nine of cells 152 as in fragment 150 are illustrated by the low reflectivity minima
labeled 1 through 9. The high-transmissivity photon energy range for transmission
structure 160 varies laterally.
Fig. 6 illustrates another implementation of assembly 87.
Assembly 87 includes transmission structure 180, Transmission structure 180 can
be a laterally graded Bragg mirror in which each of layers 182, 184, 186, and 188
is laterally graded. Each of layers 182, 184, 186, and 188 could be produced as
described above for cavity 170.
Fig. 7 illustrates the laterally varying light transmission
properties of transmission structure 180. Transmission structure 180 reflects different
wavelengths as a function of position along the x-axis. Curves 200, 202, 204, and
206 are shown, representing reflectivity of the portion of transmission structure
180 over each of four cells 152 in fragment 150, with curve 200 being for the leftmost
cell of the four in Fig. 6 and curve 206 being for the rightmost cell of the four.
The high-reflectivity photon energy range for transmission structure 180 varies
Fig. 8 illustrates a technique that produces transmission
structure 210 with laterally varying light transmission properties similar to those
illustrated in Figs. 5 and 7 but with lateral variation in each of two dimensions.
Transmission structure 210 is produced on or over cells
152 of photosensor array 150 by using deposition source 212 to provide deposition
beam 214 that can be characterized at any given point on the surface of structure
210 by two angles. One of the two angles results from angular variation of deposition
beam 214 in the x-direction across array 150, while the other results from angular
variation in the y-direction. As a result, the thickness gradient of structure 210
is similarly different in the x- and y-directions. Therefore, cells within each
row extending in one of the two directions will photosense a range of photon energies
similarly to Fig. 7, but the range will be different than the range photosensed
by cells in any other row extending in the same direction.
The technique of Fig. 8 could be implemented in a variety
of ways. For example, during deposition, structure 210 could be formed on a support
structure that is tilted as required, deposition source 212 could be tilted as required,
or both kinds of tilt could be employed.
Fig. 9 illustrates a technique that produces transmission
structure 220 with laterally varying light transmission properties but without variation
in thickness of transmission structure 220. The technique in Fig. 9 can be characterized
as providing laterally varying optical thickness d*n, where d is thickness and n
is index of refraction, but without actual variation in thickness d.
In the upper part of Fig. 9, homogeneous coating 222 is
deposited by deposition source 224, which provides deposition beam 226 uniformly
over the surface of photosensor array 150.
Then, in the lower part of Fig. 9, light source 230 provides
radiation 232 that is scanned across the coating over array 150 to introduce a laterally
varying change of refractive index in resulting transmission structure 220, For
example, source 230 can be an ultraviolet source that provides intensity I with
a constant value along each line parallel to the y-axis (perpendicular to the plane
of Fig. 9), but varying from Imin for lines nearer the y-axis to Imax
for lines farther from the y-axis. As a result, the wavelengths transmitted to cells
in array 150 can vary along the x-axis from &lgr;min to &lgr;max,
as shown. The same pattern of intensity can be concurrently applied by source 230
to each of a number of arrays that are appropriately arranged, allowing batch fabrication
of arrays. Two-dimensional variation could also be obtained.
The techniques illustrated in Figs. 4-9 could be implemented
in various other ways, with different cells of a photosensor array photosensing
slightly different subranges of a range of photon energies. For example, various
production and calibration techniques of transmission structures could be employed,
such as a step-like transmission structure.
Implementations of the techniques in Figs. 1-9 are examples
of a detector that detects photon energies emanating from objects. The detector
includes a fluidic structure with a set of channels defined therein. Within each
channel objects travel while photons emanate from the objects within an application's
range of photon energies. The detector also includes an IC that has a photosensor
array. The array includes, for each channel, a respective set of cells that photosense
objects traveling within the channel. The detector also includes a transmission
structure that transmits photons from objects traveling within each of a subset
of the channels to the channel's respective set of cells. The transmission structure
has a respective series of regions for the channel, and each region transmits photons
from objects within a segment of the channel to a respective subset of the channel's
set of cells. Each region transmits within a respective subrange, and the subranges
of at least two of the regions are different from each other.
In general, the resolution of a technique as in any of
Figs. 4-9 depends heavily on the number of cells in an array, the full width half
maximum (FWHM) of the transmission peak, and the peak shift per cell. The smaller
the FWHM and the peak shift, the better the resolution. On the other hand, the totally
covered spectral width can be enhanced by increasing the FWHM and the peak shift
per cell. Therefore, the technique can be customized to the needs of a specific
application. For example, the use of a Fabry-Perot cavity as in Fig. 4 enables very
high spectral resolution, while a version with multiple cavities and many layers
as in commercially available products will be favorable for applications with low
light intensities in combination with small spectra! resolution such as with fluorescence.
With such an approach, the spectral width of the transmission window and the reflectivity
of the stop band can be optimized separately, which may be advantageous because
the reflectivity of the stop band determines stray light suppression. It would also
be possible to use a single laterally graded distributed Bragg reflector (DBR) mirror
as in Figs. 6 and 7 to obtain a photosensor array with high light sensitivity but
limited wavelength resolution, appropriate for fluorescence or luminescence sensing.
In a version with only one DBR mirror with slightly graded
transmission properties as in Figs. 6-8, integrated over a photodiode array for
example, the photocurrent in each cell is slightly different from its neighbors
depending on the incident light spectrum. If the transmission properties of the
DBR over each cell are known, the original spectrum of incident light can be reconstructed.
The number of cells defines the number of spectral points that can be reconstructed
and therefore determines spectral resolution. The reconstruction works best for
wavelengths where transmission changes drastically from one cell to the next.
A particular advantage of analyzer 10, when implemented
with techniques similar to those of Figs. 3-9, is that spectral information of objects
can be collected step-by-step as the objects move across or along a series of sensing
components, each of which obtains information about a respective range of photon
energies. As a result, highly sensitive optical characterization techniques can
be combined, including multiple range fluorescence spectroscopy and Raman spectroscopy,
as described above in relation to Fig. 1. Each of sensing components 56, 58, and
60 can be thought of as a chip-size spectrometer that includes a photosensor array
together with a laterally varying filter such as a coating. The laterally varying
transmission and reflection properties of the coating over the photosensor array
define a correlation between position and photon energy. Therefore the spatially
dependent signal from the photosensor array contains information about the incident
spectrum. Because of the distributed nature of the spectrometer and the fact that
the incident light traverses the photosensor array in the process of resolving spectral
distribution, sensitivity is improved, making additional optics unnecessary.
In general, high sensitivity is obtained by the above techniques
because the light from the object is received at any given time by only a few cells
with relatively narrow subranges. But by photosensing light emanating from an object
or another optical signal across the entire array, information about a complete
range of photon energies can obtained. This technique therefore allows longer integration
times than conventional techniques but does not sacrifice throughput capacity. Sensitivity
can be adjusted by selecting the size and number of cells assigned to a specific
subrange of photon energies. Simpler optics can be used and no dispersion element
is necessary. Note that in conventional spectrometers, any light that is diffracted
into the 0th, 2nd, and higher orders is wasted.
In experimental implementations, a coating as in Fig. 4
typically transmits approximately 60% of photons in its respective subrange. The
subranges can be chosen with wavelengths that span between 0.01 and tens of nanometers
(nm), depending on the design and gradient of the coating and the cell size of the
photosensor array. Very high light yield can be achieved by using a highly sensitive
photosensor, such as an avalanche photosensor array.
In contrast to transmission structures 160, 180, 210, and
220, any coating or other transmission structure over row 102 in Fig. 3 must function
as a gray filter across the range &lgr;all in order to provide a suitable
reference. It may also be possible to leave row 102 uncoated in some implementations.
Fig. 10 illustrates exemplary operations in producing an
analyzer like analyzer 10 in Fig. 1.
The operation in box 250 in Fig. 10 produces a fluidic
structure with a channel in which objects can be carried by fluid. For example,
the operation in box 250 could include manufacturing a fluidic structure by positioning
or otherwise producing a structured spacer layer between two quartz slides. The
spacer layer could be a patterned layer of PDMS or could be any other suitable material
or combination of materials, including, for example, Gelfilm® or quartz. The
operation in box 250 could alternatively be implemented in various other ways, such
as by defining a fluidic channel in a quartz slide by glass etching or by molding
PDMS to produce a channel, and by then combining the resulting structure with an
upper quartz slide. Two layers of PDMS could be fabricated on separate substrates
and then one could be flipped over and aligned with the other by chip-on-chip assembly,
A final substrate of glass, PCB, or PDMS or sufficient hardness could be used to
allow direction connection to control and detection measurement circuitry.
The operation in box 252 then attaches fluidic components
to the fluidic structure produced in box 250. The fluidic components attached in
box 252 can be operated to cause and control movement of objects in the channel.
The operation in box 254 attaches components for enhanced
light-target interaction. The operation in box 254 can attach optical component
96 on the side of support layer 90, providing an appropriate surface through which
light can be coupled into a portion of channel 14 that functions as an anti-resonant
waveguide. Similarly, the operation in box 254 can produce spacers 72 to provide
a suitable gap that avoids interference with anti-resonant wave guiding.
The operation in box 260 attaches photosensor arrays with
cells that photosense in different subranges. The operation in box 260 can be implemented
by attaching detector 87. The detector can also include reference cells.
The operation in box 262 can be performed at a different
time, as suggested by the dashed line. It could be performed in box 254, or it could
be done later. Like the detector, each light source can be attached once, after
which it is stationary. In the operation in box 262, one or more light sources are
positioned to produce excitation of objects being carried within the channel. Like
other operations above, the operation in box 282 can also include attaching wires
or other appropriate circuitry to provide signals from a microprocessor or I/O device
to light sources.
The technique of Fig. 10 could be modified. For example,
the operations in boxes 252, 254, and 260 could be combined in any appropriate way
to facilitate attachment of components in a desired sequence. Also, an additional
operation could be performed to align or attach interconnects between ICs, gates,
and other circuitry, such as connectors to a microprocessor or computer, or this
operation could be partially performed in each of boxes 272, 274, 280, and 282.
Fig. 11 shows an alternative arrangement that could be
produced by an implementation of Fig. 10. First and second fluorescence sensing
components 56 and 58 are next to each other in the series of sensing components
along channel 14. In addition, however, they are positioned so that IC 270 can be
attached over both of them. As a result, the photosensor array of IC 270 includes
both cells along channel 14 within component 56 and also cells along channel 14
within component 58.
Fig. 12 is a cross-section along the line 12-12 in Fig.
11, and shows how detector 272, which includes IC 270, can be supported over air
gap 88 by spacers 72. The lateral variation in optical thickness of the transmission
structure may be such that the ranges and subranges photosensed within sensing component
56 are different from those photosensed within sensing component 58; alternatively,
the ranges and subranges could be the same. Spacers 72 can help to reduce cross-talk
between components 56 and 58 because spacers 72 can be shaped and positioned to
act as light-absorbing walls between the two components.
Fig. 13 shows an alternative arrangement in which detector
272 as in Fig. 12 is positioned over a set of parallel channels 274, which could
be produced by producing walls 276 to subdivide a larger channel into subchannels.
An advantage of the technique illustrated in Fig. 13 is that several streams of
objects can be analyzed in parallel in order to increase throughput or specificity
of an analyzer. Laterally varying optical thicknesses of a transmission structure
can be produced so that a different range of photon energies is photosensed in each
of channels 274, or different sub ranges are photosensed in different channels,
or the same ranges and subranges could be photosensed in all channels.
Fig. 14 shows an alternative arrangement that could be
produced by an implementation of sensing component 58 in Fig. 2, with components
similar to those described above in relation to Fig. 2 having the same reference
numerals. As in Fig. 2, p. Portion 80 of channel 14 functions as an anti-resonant
waveguide in response to light from source 84. Assembly 87 is along portion 80,
separated from plate 94 by spacers 72. Upstream from portion 80 (but downstream
from light source 84) can be positioned a series of triggering photodetectors, represented
by photodetector 290 on spacers 292
Within portion 80, fluorescing objects 300, 302, and 304
are being carried through channel 14. As they fluoresce, objects 300, 302, and 304
emanate photons, represented respectively by rays 310, 312, and 314.
Fig. 15 shows an alternative implementation of sensing
component 58 in Fig. 14, with similar components having the same reference numerals.
As in Fig. 14, portion 80 of channel 14 functions as an anti-resonant waveguide
in response to light from a source. Assembly 87 is along portion 80, separated from
plate 94 by gap 88 due to spacers 72. Upstream from portion 80 (but downstream from
light source 84) can be positioned a series of triggering photodetectors or, alternatively,
another type of detector.
Within portion 80, a closely spaced, continuous sequence
of fluorescing objects are being carried through channel 14. Fig. 15 shows a group
of the objects passing through portion 80 along assembly 87, and the group is led
by object 330 after which follow several intermediate objects 332 and, finally,
object 334. As they fluoresce, objects 330, 332, and 334 emanate photons, and the
photons pass through optical component 340 attached to assembly 87. Optical component
340 can be implemented, for example, as a Selfoc® lens array similar to lens
arrays 164 in Figs. 4 and 6, as described above. If optical component 340 is present,
however, array 164 would be omitted.
As a result of component 340 and as indicated by rays 342,
photons emanating from each of objects 330, 332, and 334 are predominantly incident
on a different cell of assembly 87 than photons emanating from nearby objects. Quantities
read out from a photosensor array in detector 87 can be used to obtain information
about objects 300, 302, and 304 even though a continuous sequence of closely spaced
objects is concurrently traveling past the array.
Fig. 16 shows an alternative implementation without distinguishable
objects 330, 332, and 334. In this implementation, fluid flow in channel 14 can
be approximated as a continuous sequence of small volumes traveling in channel 14.
The stream of fluid is divided into imaginary small volumes, each of which can be
analyzed as if it were an object, allowing for continuous monitoring of how distribution
of photon energies emanating from the fluid changes with time, such as due to changing
composition of the fluid. In Fig. 16, a group of the small volumes are passing through
portion 80 along assembly 87, and the group is led by volume 360 after which follow
several intermediate volumes 362 and, finally, volume 364.
Due to interaction with light, particles in volumes 360,
362, and 364 emanate photons. The distribution of photon energies in each volume
depends on concentrations of molecules that are involved in fluorescence and scattering.
As in Fig. 15, the photons pass through optical component 340 attached to assembly
87. Rays 370 indicate that photons emanating from each of volumes 360, 362, and
364 are predominantly incident on a different cell of assembly 87 than photons emanating
from nearby volumes. Quantities read out from a photosensor array in detector 87
can be used to obtain information about concentrations of molecules in volumes 360,
362, and 364 even though fluid is continuously flowing past the array.
Fig. 17 illustrates system 400 that can operate analyzer
10. System 400 illustratively includes central processing unit (CPU) 402 connected
to various components through bus 404.
System 400 also includes external input/output (I/O) component
406 and memory 408, both connected to bus 404. External I/O 406 permits CPU 402
to communicate with devices outside of system 400.
Additional components can be connected to bus 404. IC I/O
410 is a component that permits CPU 402 to communicate with ICs in analyzer 10;
M ICs are illustrated in Fig. 17 by a series extending from IC(O) 412 to IC (M-1)
414. ICs 412 through 414 illustratively include IC(m) 416 with a photosensor array
418, which includes cells that photosense subranges as described above. Similarly,
fluidic device I/O 420 is a component permitting CPU 402 to communicate with various
Memory 408 includes program memory 430. The routines stored
in program memory 430 illustratively include fluid/object movement routine 440 and
detect, readout, and combine routine 442.
CPU 402 executes fluid/object movement routine 440 to communicate
with fluidic devices 422 through 424. CPU 402 can receive signals from sensors,
perform computations to determine what fluidic operations are necessary, and then
provide signals to activate pumps, metering electrodes, gates, and valves to produce
appropriate movement of fluid and of objects carried by fluid in channel 14.
In executing routine 442, CPU 402 can call a subroutine
implemented as shown in Fig. 18, which could instead be within routine 442. The
subroutine in Fig. 18 can be implemented for single objects moving past arrays as
in Fig. 2; for spaced multiple objects moving past arrays as in Fig. 14; for continuous
sequences of objects moving past arrays as in Fig. 15; or for continuous flow of
fluid past arrays as in Fig. 16. Also, the subroutine in Fig. 18 follows a general
strategy of performing a series of readout operations, after which spectral information
is combined and provided, although it would also be possible to provide the information
from each readout operation immediately.
When CPU 402 executes the operation in box 480, it performs
a pre-sensing readout. The information could be obtained from a series of photodetectors
illustrated by photodetector 290 in Fig. 14 or from reference cells in the photosensor
array, such as the cells in row 102 in Fig. 3. It would also be possible to modify
the photosensor array to include trigger cells positioned along channel 14 upstream
from a line of subrange cells, and uncoated so that they provide information about
all photon energies. The operation in box 480 may not be necessary for object-free
implementations as in Fig. 16 if information about fluid speed is available from
another source (e.g. the pump speed). If necessary, additional well-known techniques
for measuring fluid velocity may be used to trigger sensing.
Using the information from box 480, CPU 402 can obtain
information about each object or small volume and determine an appropriate sensing
period for each object or volume, in the operation in box 482. For example, CPU
402 could perform calculations to determine whether one or more objects are present,
the position of each object, and the speed of each object; in object-free implementations
as in Fig. 16, CPU 402 may need only determine the fluid speed. Using this information
and taking into account previously calculated sensing periods for the same objects
or for similar fluid volumes, if any, CPU 402 can also determine an appropriate
sensing period to be used during sensing readout; in general, the sensing period
must provide an integration time shorter than the time necessary for an object or
small volume to pass each subrange cell. Each object or small volume can therefore
have a unique sensing period. Alternatively, CPU 402 could provide signals to adjust
fluid speed to obtain the same result.
CPU 402 can then perform the sensing readout operation,
in box 484. This operation includes providing signals so that photons are photosensed
cumulatively during the sensing period obtained in box 482, and may also include
signals to peripheral circuitry on an IC so that analog quantities photosensed by
subrange cells are adjusted based on analog quantities sensed by paired reference
cells. After adjustment, if any, analog quantities can be converted to digital signals
The photosensed quantities read out in box 484 can also
be digitally adjusted by CPU 402 before being stored for each object or small volume,
in box 490. The digital adjustment can include adjusting quantities photosensed
by subrange cells based on quantities photosensed by paired reference cells. The
position and speed information about each object or small volume from box 482 can
be used to determine which photosensed quantities result from photons emanating
from each object or small volume.
In performing the operations in boxes 482 and 490, CPU
402 can employ data structures stored in memory 408. For example, one data structure
can store each object's or small volume's previously calculated position and speed,
which can then be used in performing subsequent calculations to identify the same
object or small volume. Also, a readout data structure can be employed to hold all
of the adjusted quantity information about each object or small volume. The operation
in box 490 can update the readout data structure each time it obtains additional
information about the same object or small volume. In an implementation as in Fig.
17, the operations in boxes 480, 482, 484, and 490 can be performed separately for
each of ICs 412 through 414. Further, as suggested by the dashed line from box 490
to box 480, the same operations can be performed repeatedly for each of the ICs.
Between consecutive executions of the operations in boxes
480, 482, 484, and 490, each object's optical signal may move only a few cells along
the photosensing path, and some implementations may require that consecutive objects
be sufficiently separated to avoid confusion.
Only a small fraction of an application's range of photon
energies is photosensed and stored at a time by the operation in box 490. As the
operations in boxes 480, 482, 484, and 490 are repeated while an object or small
volume travels along the path past the array, more and more spectral intervals are
resolved. When the object or small volume has passed the whole array, its spectral
information can be recomposed from the stored fractions.
Upon completion of information gathering, CPU 402 can perform
the operation in box 492 to provide photosensed quantities. As shown, this operation
can include combining the sensed quantities for each object or small volume so that
spectral information about the object or small volume can be provided, such as in
the form of a profile or other data structure.
The exemplary implementations described above can provide
compact, inexpensive components that generally require no additional mechanical
or optical parts to perform functions such as spectrometry. For example, a portable,
easy-to-use spectrometer could include an analyzer as described above; a portable,
compact unit could, for example, be standard equipment for emergency response teams
anywhere. The results of photosensing can be read out rapidly and in parallel from
a number of ICs, allowing fast data acquisition; as a result, an initial characterization
of an object may be used to determine whether to perform more refined or detailed
analysis of the object, or to determine which of different types of analysis are
performed. A multi-signal approach like this is compatible with reagentless identification;
also, a wide variety of objects can be identified in a wide variety of fluids, such
as various nanoparticles, microorganisms, bioagents, and toxins in various aerosols,
water, blood, food, and other specimens.
Spectrometry measurements have a wide variety of applications,
including, for example, optical instrumentation, telecommunications, fluorescence
devices, process control, optical signal scanning, detection systems for chemical
and biological agents, and so forth. An example of a specific application is an
in-line detector for manufacturing and functionalizing colloidal particles in an
industrial setting. In this application, processes typically are performed in closed
systems and the properties of colloidal particles can be assessed only after all
processing steps are completed. A small detection platform implemented as described
above can be easily built into an on-line detector directly connected to a manufacturing
vessel. As a result, small amounts of particles can be analyzed continuously in
real time to determine size, chemical composition, and surface conditions. This
approach permits instant process adjustments leading to production of materials
with consistent properties from run to run. In-line Coulter counters for instant
size measurements are already commercially available, but compact detectors as described
above can also probe chemical composition using multiple advanced spectroscopic
methods, an approach not previously available.
Components could have other various shapes, dimensions,
or other numerical or qualitative characteristics.
Some of the above exemplary implementations involve specific
materials, such as in fluidic structures, photosensor arrays, and transmission structures,
but a wide variety of materials could be used with layered structures with various
combinations of sublayers. In particular, photosensor arrays for a desired speed,
sensitivity and wavelength range could have any suitable material, such as silicon,
germanium, indium-gailium-arsenide, gallium arsenide, gallium nitride, or lead sulphide,
and could be produced with any appropriate kind of devices, including, for example,
photodiodes, avalanche photodiodes, p-i-n diodes, photoconductors, and so forth,
with any appropriate technique for sensing and reading out information whether based
on CCD, CMOS, or other techniques. Various commercially available detector arrays
have pixel densities as high as ten megapixels, and some high density ICs have become
Similarly, transmission structures could be fabricated
with any appropriate techniques, including thin film technology such as sputtering,
e-beam or thermal evaporation with or without plasma assistance, epitaxial growth,
MBE, MOCVD, and so forth. To produce Bragg mirrors, appropriate pairs of materials
with low absorption coefficients and large difference in refractive indices could
be chosen, bearing in mind the photon energies of interest; exemplary materials
include SiO2/TiO2, SiO2/Ta2O5,
GaAs/AlAs, and GaAs/AlGaAs. Thicknesses of layer in transmission structures may
vary from 30 nm up to a few hundred nanometers.
Some of the above exemplary implementations involve particular
types of transmission structures, but these transmission structures are merely exemplary,
and any transmission structure that has laterally varying optical thickness could
be used. Various other techniques could be used to produce transmission structures
with lateral variation.
Furthermore, various techniques other than transmission
structures could be used to obtain photosensor arrays in which cells sense different
subranges of photon energy.
Some of the above exemplary implementations employ an arrangement
of ICs relative to fluidic structures in which fluid moves and may carry objects,
and a wide variety of such arrangements, with or without fluidic structures, could
be made. The invention could also be implemented with any other suitable type of
photosensor array, including simple light-to-electric signal transducers arranged
as cells of a photosensor array. The techniques described above allow concurrent
photosensing of multiple objects or even photosensing of volumes of fluid that do
not contain distinguishable objects. Different rows of a single two-dimensional
photosensor array on an IC could be differently coated to photosense in different
Some of the above exemplary implementations employing fluidic
structures also employ enhanced light-target interaction to obtain fluorescence.
In general, the techniques described above could also be used for self-emitting
or auto-fluorescing objects such as particles. Furthermore, various types of fluorescence,
photo-luminescence, chemo-fluorescence, inelastic scattering, and so forth could
be employed. The technique of anti-resonant waveguiding, described above, is only
one of many techniques that could be used for enhanced light-target interaction,
and any such excitation technique could be applied continuously or intermittently
along a path. Various parameters could be adjusted to obtain anti-resonant waveguiding,
including the shape of quartz or glass surrounding the channel; a thinner structure
is generally better, with a surface parallel to the channel generally being required.
The exemplary implementation in Fig. 17 employs a CPU,
which could be a microprocessor or any other appropriate component
The above exemplary implementations generally involve production
and use of various components, following particular operations, but different operations
could be performed, the order of the operations could be modified, and additional
operations could be added within the scope of the invention. For example, in implementations
in which a transmission structure is on a separate substrate from a photosensor
array, the transmission structure could be moved relative to the photosensor array
between consecutive sensing operations. Also, readout of adjusted or unadjusted
sensed quantities from an IC could be performed serially or in parallel, and could
be performed cell-by-cell or in a streaming operation.