U.S. patent application number 14/270202 was filed with the patent office on 2014-08-28 for apparatus and method for improved optical detection of particles in fluid.
This patent application is currently assigned to The Regents of the University of California. The applicant listed for this patent is The Regents of the University of California. Invention is credited to Chun Hao Randy Chen, Victor Jie Lien, Yu-Hwa Lo.
Application Number | 20140244217 14/270202 |
Document ID | / |
Family ID | 37968681 |
Filed Date | 2014-08-28 |
United States Patent
Application |
20140244217 |
Kind Code |
A1 |
Lo; Yu-Hwa ; et al. |
August 28, 2014 |
APPARATUS AND METHOD FOR IMPROVED OPTICAL DETECTION OF PARTICLES IN
FLUID
Abstract
A number of fluidic-photonic devices for allowing optical
detection, systems employing such devices, and related methods of
operation and fabrication of such devices are disclosed herein. In
at least some embodiments, the devices can serve as flow cytometry
devices and/or employ microfluidic channels. Also, in at least some
embodiments, the devices are fluidic-photonic integrated circuit
(FPIC) devices that employ both fluidic channels and one or more
waveguides capable of receiving and/or delivering light, and that
can be fabricated using polymeric materials. The fluidic-photonic
devices in at least some embodiments are capable of functionality
such as on-chip excitation, time-of-flight measurement, and can
experience enhanced fluorescence detection sensitivity. In at least
some embodiments, the devices employ detection waveguides that are
joined by way of a waveguide demultiplexer. In additional
embodiments, a variety of techniques can be used to process
information received via the waveguides, including an iterative
cross-correlation process.
Inventors: |
Lo; Yu-Hwa; (San Diego,
CA) ; Lien; Victor Jie; (San Diego, CA) ;
Chen; Chun Hao Randy; (Arcadia, CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
The Regents of the University of California |
Oakland |
CA |
US |
|
|
Assignee: |
The Regents of the University of
California
Oakland
CA
|
Family ID: |
37968681 |
Appl. No.: |
14/270202 |
Filed: |
May 5, 2014 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
13605925 |
Sep 6, 2012 |
8717569 |
|
|
14270202 |
|
|
|
|
12091414 |
Oct 28, 2008 |
8270781 |
|
|
PCT/US06/60313 |
Oct 27, 2006 |
|
|
|
13605925 |
|
|
|
|
60731551 |
Oct 28, 2005 |
|
|
|
Current U.S.
Class: |
702/189 |
Current CPC
Class: |
G01N 21/49 20130101;
G01N 15/1429 20130101; G01N 21/6408 20130101; G01N 15/1484
20130101; Y10T 29/49124 20150115; B01L 2400/0487 20130101; G01N
15/1434 20130101; B01L 2200/0668 20130101; B01L 2300/0816 20130101;
B01L 3/502715 20130101; G01N 21/05 20130101; G01N 21/53 20130101;
G01N 2021/0346 20130101; B01L 2400/0424 20130101; B01L 2300/0654
20130101; B01L 3/502707 20130101 |
Class at
Publication: |
702/189 |
International
Class: |
G01N 15/14 20060101
G01N015/14; G01N 21/49 20060101 G01N021/49 |
Goverment Interests
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT
[0002] This invention was made with United States Government
support awarded by the following agencies: Air Force Office of
Scientific Research (AFOSR) Grant No. F49620-02-1-0288. The United
States Government has certain rights in this invention.
Claims
1. A device, comprising: a processor; and a memory comprising
processor executable code, the processor executable code when
executed by the processor configures the device to: receive at
least one signal produced by at least one detection device, the
signal corresponding to an interaction between scattered or
fluorescent light and at least one particle suspended in a fluid
within a fluidic channel, the scattered or fluorescent light having
been produced upon illumination of the fluid and the at least one
particle as the at least one particle flows through the fluidic
channel from first position within the fluidic channel to a second
position within the fluidic channel; perform a calculation based
upon the at least one signal and one or more transit times required
for the at least one particle to proceed between the first and the
second positions within the fluidic channel, wherein the at least
one signal includes a plurality of signals, and wherein the
calculation comprises determining a product of a plurality of
values corresponding respectively to the plurality of time-shifted
signals; and produce information indicative of at least one
characteristic of the at least one particle.
2. The device of claim 1, wherein the plurality of time-shifted
signals are shifted relative to one another by the one or more
transit times.
3. The device of claim 2, wherein the one or more transit times are
obtained prior to obtaining the at least one signal.
4. The device of claim 1, wherein the at least one particle has a
diameter of approximately 1 micrometer.
5. The device of claim 1, wherein the at least one signal comprises
two or more signals that are received from two or more detection
devices configured to detect the scattered or fluorescent light
when the at least one particle is at two or more positions within
the fluidic channel.
6. A device, comprising: a processor; and a memory comprising
processor executable code, the processor executable code when
executed by the processor configures the device to: receive at
least one signal produced by at least one detection device, the
signal corresponding to an interaction between scattered or
fluorescent light and at least one particle suspended in a fluid
within a fluidic channel, the scattered or fluorescent light having
been produced upon illumination of the fluid and the at least one
particle as the at least one particle flows through the fluidic
channel from first position within the fluidic channel to a second
position within the fluidic channel; perform a calculation based
upon the at least one signal and one or more transit times required
for the at least one particle to proceed between the first and the
second positions within the fluidic channel, wherein the
calculation comprises iteratively performing two or more
calculations based upon the at least one signal and the one or more
transit times, wherein each additional iteration of the calculation
results in a respective additional piece of information.
7. The device of claim 6, wherein the processor executable code
when executed by the processor further configures the device to
determine whether a threshold number of successive ones of the
respective additional pieces of information have been determined to
be substantially equal to zero and, if so, cease to perform the
additional iterations.
8. The device of claim 6, wherein the processor executable code
when executed by the processor further configures the device to
calculate a sum of the information and the additional pieces of
information, and output at least one of: the information, the sum,
and derivative information based upon at least one of the
information and the sum.
9. A device, comprising: a processor; and a memory comprising
processor executable code, the processor executable code when
executed by the processor configures the device to: receive at
least one signal produced by at least one detection device, the
signal corresponding to an interaction between scattered or
fluorescent light and at least one particle suspended in a fluid
within a fluidic channel, the scattered or fluorescent light having
been produced upon illumination of the fluid and the at least one
particle as the at least one particle flows through the fluidic
channel from first position within the fluidic channel to a second
position within the fluidic channel; perform a calculation based
upon the at least one signal and one or more transit times required
for the at least one particle to proceed between the first and the
second positions within the fluidic channel, wherein the
calculation comprises computing a product between a time-varying
intensity function of the at least one signal and a time shifted
version of the time-varying intensity function of the at least one
signal.
10. The device of claim 9, wherein the one or more transit times
are obtained prior to obtaining the at least one signal.
11. The device of claim 9, wherein the at least one particle has a
diameter of approximately 1 micrometer.
12. A device, comprising: a processor; and a memory comprising
processor executable code, the processor executable code when
executed by the processor configures the device to: receive at
least one signal produced by at least one detection device, the
signal corresponding to an interaction between scattered or
fluorescent light and at least one particle suspended in a fluid
within a fluidic channel, the scattered or fluorescent light having
been produced upon illumination of the fluid and the at least one
particle as the at least one particle flows through the fluidic
channel from first position within the fluidic channel to a second
position within the fluidic channel; perform a calculation based
upon the at least one signal and one or more transit times required
for the at least one particle to proceed between the first and the
second positions within the fluidic channel, wherein the processor
executable code when executed by the processor further configures
the device to perform the calculation by at least: (a) assuming a
time value for each of the one or more transit times; and (b)
performing a cross-correlation computation on the at least one
signal based on the assumed time value(s); (c) upon determination
that the cross-correlation computation result is not below a
threshold value, assuming a new time value for at least one of the
one or more transit times and performing the cross-correlation
computation based on the new assumed transit time value(s); and (d)
upon determination that the cross-correlation computation result is
below the threshold value, obtaining the information.
13. The device of claim 12, wherein steps (a) through (c) are
carried out iteratively for up to a maximum number of times.
14. The device of claim 12, wherein step (d) further comprises:
producing a count as to the number of times the cross-correlation
computation result has remained below the threshold value; and
obtaining the information if the count is greater than or equal to
a predetermined count, obtaining the information.
15. The device of claim 12, wherein step (d) further comprises
computing a sum of all cross-correlation results obtained in steps
(c) and (d).
16. The device of claim 12, wherein the cross-correlation
computation result is representative of number of particles that
are suspended within the fluid as the fluid and particles suspended
therein flow through the fluidic channel.
17. The device of claim 12, wherein the processor executable code
when executed by the processor further configures the device to
control or monitor an operation of the at least one detection
device.
18. The device of claim 12, wherein the threshold value is zero.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application is a divisional application of and claims
priority to U.S. patent application Ser. No. 13/605,925, filed on
Sep. 6, 2012, which is a divisional application of and claims
priority to U.S. patent application Ser. No. 12/091,414, filed on
Apr. 24, 2008, which is a U.S. National Stage application under 35
U.S.C. .sctn.371 of International Patent Application No.
PCT/US2006/060313, filed on Oct. 27, 2006, which claims benefit of
priority of U.S. Provisional Patent Application No. 60/731,551,
filed on Oct. 28, 2005. The entire contents of the before-mentioned
patent applications are incorporated by reference as part of the
disclosure of this application.
FIELD OF THE INVENTION
[0003] The present invention relates to systems and methods for
optical detection and, more particularly, to systems and methods
for detecting small objects or particles such as cells or DNA.
BACKGROUND OF THE INVENTION
[0004] Over a period of nearly five decades, flow cytometry has
evolved from a simple technique for counting suspended particles
(e.g., analytes, cells or DNA) in fluid into a highly sophisticated
and versatile technique that is critical to clinical diagnosis and
fundamental biomedical research. Early efforts in the development
of flow cytometry focused upon the attainment of a stable flow
system able to transport particles, without disturbance by any
alien aerosol, to regions of laser beam illumination for optical
interrogation via fluorescence or light scattering. A standard
approach for today's flow cytometers is to create a laminar sheath
flow in a transparent capillary tube to minimize noise due to
fluctuations in position and propagation speed of the particles.
Besides such improvements in controlling particle flow and in flow
cytometry instrumentation generally, significant progress has also
been made with respect to other aspects of flow cytometry, for
example, with respect to the methods of cell preparation, new
fluorescent dyes and new markers of cell properties.
[0005] These technological advances in flow cytometry have made it
possible to use flow cytometers in a variety of areas. For example,
flow cytometers are now used for analyzing white blood cells in
AIDS patients. Also for example, flow cytometers are now employed
in performing cancer diagnosis and stem cell sorting. Indeed, flow
cytometry is now widely recognized as an important clinical and
research tool. However, even though the size of a flow cytometer
has been reduced from a piece of equipment occupying an entire room
to a table top system with ever increasing functionality and
performance, flow cytometers continue to cost between $150K and
$1M, and consequently remain a tool affordable only by major
medical centers and laboratories. Size and price reduction by
orders of magnitude (e.g. 1000 times) are necessary to make flow
cytometers a prevailing diagnosis tool that can be afforded by more
hospitals and medical practitioners around the world.
[0006] One technique that holds promise for miniaturizing flow
cytometers is the use of microfabricated flow cells, enabled by
advances in microfluidics. Integrated microfluidic chips that
perform a variety of functions for chemical analysis and biological
screening have found wide applications in the pharmaceutical
industry and have accelerated the progress of research in
biotechnology. Several research groups have demonstrated the
ability to manipulate cells and micro-particles in microfluidic
devices using the effects of fluidic pressure, dielectrophoresis,
optical trapping, and electro-osmosis. More particularly, the
introduction of microfabricated electrodes in the fluidic channels
of microfluidic devices can facilitate the optical detection of
particles by controlling and manipulating the positions, angles,
and populations of the particles in microfluidic channels via the
dielectrophoretic effect.
[0007] There are several reasons that make these results
particularly relevant to the development of compact flow
cytometers. First, biological cell sizes fit well with the
dimensions of the microfluidic devices that can be easily and
precisely fabricated using microfabrication techniques such as
lithography and molding. Second, microfluidic devices tend to
support laminar flow, making the flow control simpler and fluid
transport highly efficient. Third, micro-scale integration allows
more functionality (e.g. pumps, valves and switches) to be
incorporated into the device. Finally, two-dimensional or even
three-dimensional array structures can be fabricated to enhance the
performance of the system and alleviate the limit of device
throughput.
[0008] Although rapid progress has been made in microfluidics that
is applicable to flow cytometry, the scheme of optical detection
employed in flow cytometry has not experienced similarly important
advances or changes. In particular, while the hardware utilized in
performing optical detection has continued to evolve, resulting in
more advanced lasers, more sensitive detectors, and superior
optical mechanical components, there nevertheless has not been any
paradigm shift in terms of the manner in which optical detection is
performed in flow cytometry. As a result, the expensive and bulky
optical setup currently necessary for fluorescence detection
threatens to become a bottleneck restricting the realization of
compact, low-cost flow cytometers. Additionally, the relatively
high cost of lasers and light detectors for use in flow cytometry
is further exacerbated when one reduces the size of the overall
system.
[0009] For at least these reasons, it would be advantageous if an
improved optical detector, optical detection scheme or optical
detection method could be developed for use in detecting small
objects or particles such as cells or DNA, as could be used for,
among other things, performing flow cytometry and related
techniques. More particularly, it would be advantageous if, in at
least some embodiments, such an improved optical detector/detection
scheme/method could be designed in which smaller (in terms of size
and/or weight) optical components could be employed. Additionally,
it would be advantageous if, in at least some embodiments, simpler,
less expensive components could be employed for the purposes of
generating and/or sensing light.
BRIEF SUMMARY OF THE INVENTION
[0010] The present inventors have recognized the desirability of
achieving improved photonic designs and technologies for detecting
small objects or particles, for use in various applications such as
flow cytometry involving the use of microfluidic components. For
example, the present inventors have recognized the desirability of
providing a cost effective solution to the problem of fluorescence
(and side scattering) detection in flow cytometry and, more
particularly, have recognized the importance of at least one of (a)
integrating optical components with fluidic circuits to reduce the
size and weight of the overall system, and (b) developing
innovative architectures of photonic circuits to achieve desired
levels of sensitivity without the need for expensive components
such as lasers (e.g., main frame lasers) and/or ultra sensitive
detectors (e.g., photomultiplier tubes (PMTs)).
[0011] In accordance with at least some embodiments of the present
invention, to achieve such goals the present inventors propose a
microfluidic-photonic integrated circuit optical interrogation
device that can be utilized as a microfabricated flow cytometer.
The device includes a photonic circuit integrated monolithically
with the microfluidic channels such that the optical interrogation
zones are in the proximity of and well aligned to the optical
waveguides that collect the fluorescence and/or scattering light
signals. The use of such a waveguide approach to replace free-space
optics eliminates the needs for lenses and precision mechanics for
optical alignment, making significant size and weight reduction
possible. The device can be fabricated using a fluidic-photonic
integrated circuit (FPIC) process.
[0012] In at least some such embodiments, multiple waveguides are
employed to form an array waveguide structure so that, along the
direction of flow, a particle (e.g., an analyte, cell or segment of
DNA) will pass a series of waveguide-defined optical interrogation
zones, each producing a signal that is correlated in time and space
to the others. In one example of such an embodiment, an array of
eight parallel waveguides is employed so that the signal produced
by a particle will be detected eight times. At the detection end,
an array of eight detectors can be employed or, alternatively, it
is possible to combine the eight waveguides into a single output
waveguide and use only a single detector (or, also alternatively,
more than one but less than eight detectors can be employed). For a
single detector approach, the signals from the eight waveguides can
be multiplexed in the time domain, with a time delay of the
demultiplexed operation being set equal to the transit time of the
particle as it passes between adjacent waveguides.
[0013] In at least some embodiments, the present invention relates
to a device that includes a fluidic channel capable of conducting a
fluid containing at least one particle, a source of electromagnetic
radiation arranged to provide the electromagnetic radiation into
the fluidic channel to interact with the at least one particle
contained within the fluid as the fluid is conducted by the fluidic
channel, and a first plurality of optical waveguides having
respectively a plurality of ends positioned along the fluidic
channel. The optical waveguides receive at least some of the
electromagnetic radiation after the electromagnetic radiation has
interacted with the at least one particle.
[0014] Additionally, in at least some embodiments, the present
invention relates to a fluidic-photonic integrated circuit (FPIC)
device that includes a microfluidic channel, means for exciting a
material within the microfluidic channel, and a first optical
waveguide for receiving electromagnetic radiation as a result of
the exciting of the material. Information regarding the material is
detected based upon the received electromagnetic radiation.
[0015] Further, in at least some embodiments, the present invention
relates to a method of manufacturing a fluidic-photonic integrated
circuit (FPIC) device. The method includes casting pre-polymer onto
a photo-lithographically patterned mold, thermally-curing the
pre-polymer, and demolding a first piece of thermally-cured polymer
from the mold. The method also includes bonding the first piece to
a second piece of polymer material to form a fluidic channel, and
implementing the fluidic channel in relation to a further structure
capable of receiving and guiding electromagnetic radiation away
from the fluidic channel.
[0016] Additionally, in at least some embodiments, the present
invention relates to a method of obtaining information regarding at
least one particle suspended within a flowing fluid. The method
includes applying incident light to the fluid and to the at least
one particle suspended within the fluid as the fluid flows through
a fluidic channel, and guiding scattered or fluorescent light
resulting from an interaction between the incident light and the at
least one particle by way of a plurality of optical waveguides
extending away from the fluidic channel to at least one detection
device. The method further includes deriving at least one signal at
the at least one detection device based upon the guided, scattered
or fluorescent light, and performing a calculation based upon the
at least one signal resulting in the information, the information
being indicative of at least one characteristic of the at least one
particle.
BRIEF DESCRIPTION OF THE DRAWINGS
[0017] FIGS. 1(a) and 1(b) respectively provide a schematic, side
perspective view and a top plan view of an exemplary integrated
fluidic-photonic device having a fluidic channel, optical
waveguide, and a dielectrophoretic (DEP) cage, in accordance with
at least some embodiments of the present invention;
[0018] FIGS. 2(a)-(c) illustrate steps of an exemplary process of
manufacturing the device of FIGS. 1(a)-(b) having DEP electrodes
integrated with a microfluidic channel;
[0019] FIGS. 3(a) and (b) respectively show a schematic view and a
side cross-sectional view of an exemplary fluidic-photonic
integrated circuit device that can be employed in an improved
optical detector in accordance with at least some embodiments of
the present invention;
[0020] FIG. 4 shows a side cross-sectional view of an alternate
embodiment of the fluidic-photonic integrated circuit device of
FIGS. 3(a) and (b) in accordance with some other embodiments of the
present invention;
[0021] FIGS. 5(a)-(b) show steps of an exemplary process that can
be employed to fabricate one or more fluidic channels employed in
the devices of FIGS. 3(a)-(b) and 4;
[0022] FIG. 6 shows in schematic form an improved optical detector
in accordance with at least some embodiments of the present
invention, where the detector employs a fluidic-photonic integrated
circuit similar to that of FIGS. 3(a)-(b);
[0023] FIGS. 7(a) and (b) are graphs showing exemplary output
signals from a single output of the waveguide demultiplexer of the
fluidic-photonic integrated circuit of FIGS. 4(a)-(b);
[0024] FIGS. 8(a)-(c) are graphs showing exemplary time variation
of photon counts intensity generated from eight waveguide outputs
of an alternate embodiment of the fluidic-photonic integrated
circuit of FIG. 4 where the data in (a-c) were obtained from
fluorescent beads of decreasing size and fluorescence
intensity;
[0025] FIG. 9 is a schematic illustration of time domain
cross-correlation;
[0026] FIG. 10 is a flow chart illustrating steps of an exemplary
iterative cross-correlation process;
[0027] FIGS. 11(a) and (b), respectively, are graphs showing
exemplary cross-correlated signals obtained using the raw data of
FIGS. 8(b) and (c), respectively; and
[0028] FIGS. 12(a) and (b), respectively, are graphs showing
exemplary signal chains obtained using the raw data of FIGS. 7(a)
and (b), respectively.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
[0029] As described in detail below, the present invention is
intended to encompass a variety of different embodiments of
microfluidic-photonic integrated circuits and similar devices, as
well as systems that implement those integrated circuits and
devices. Such devices can be employed for a variety of purposes
including, for example, to detect the presence of biological
particles such as cells and DNA or other particles, and/or in
various applications such as flow cytometry and other techniques.
Additionally, the present invention is intended to encompass
various methods of operating and manufacturing such integrated
circuits and other devices (and/or systems that implement those
integrated circuits and devices). In at least some embodiments, for
example, microfluidic-photonic-dielectrophoretic integrated
circuits can be fabricated by way of a process involving
micro-molding, polymer bonding, and channel waveguides with
capillary filling. At least some of the circuits described herein
can be considered to represent a new class of circuits particularly
attractive to lab-on-a-chip and biomedical applications.
[0030] Referring to FIGS. 1(a) and 1(b), a schematic, side
perspective view and a top plan view are provided of a first
exemplary microfluidic-photonic integrated circuit device 10 that
is capable of allowing fluorescent excitation and detection to be
performed. Although capable of being employed in various
environments, the device 10 in at least some cases is intended to
perform on-chip optical detection of biological particulate
material or analytes such as single cell(s) and/or small
aggregations of DNA. As shown particularly in FIG. 1(a), the device
10 includes first and second waveguides 12 and 14, respectively,
that are aligned with one another and extend away from each other.
Further as shown, the first and second waveguides 12, 14
respectively extend up to opposite sides of a fluidic channel 16
that extends perpendicularly between the waveguides. Light passing
through the first waveguide 12 along a direction generally
indicated by an arrow 13 passes through a target region 17 of the
fluidic channel 16, where it can interact with particulate material
19 flowing through the target region. Some or all (or possibly
none) of the light, depending upon the light's interaction with the
particulate material, then passes through the second waveguide 14
along a direction generally indicated by an arrow 15.
[0031] Additionally, the device 10 also includes a
dielectrophoretic electric cage 18 formed by four pairs of
diagonally-arranged electrodes 20, 22, 24 and 26, respectively,
where each pair of electrodes includes an upper and a lower
electrode as illustrated in FIG. 1(a). The dielectrophoretic
electric cage 18 is designed to trap and rotate the target
particular material electrically. The trapped object tends to
reside at the position in the fluid where the potential energy is
lowest. If an ac voltage is applied to the four pairs of electrodes
20, 22, 24 and 26 with a phase difference (e.g., with approximately
90-degree phase differences between each adjacent pair of
electrodes), then the trapped object can obtain angular momentum
and spin while it is trapped.
[0032] Referring additionally to FIG. 1(b), the top view of the
device 10 further illustrates the relative arrangement of the
integrated pairs of electrodes 20, 22, 24 and 26, the waveguides
12, 14 and the fluidic channel 16. As shown, each of the electrodes
20, 22, 24 and 26 extends at 45 degree angles relative to its
respective neighboring one of the waveguides 12, 14 and relative to
the fluidic channel 16, so as to form an "X" arrangement
overlapping a "+" arrangement formed by the waveguides and fluidic
channel. Also as shown, the electrodes 20, 22, 24, and 26 protrude
slightly into the target region 17 of the fluidic channel 16. It
should be noted that the fluidic channel 16, waveguides 12, 14 and
electrodes 20, 22, 24 and 26 in the present embodiment are small or
microscopic in size, typically having a smallest feature size
between 5 and 50 micrometers.
[0033] A variety of processes can be implemented in order to
manufacture or otherwise form the device 10 of FIGS. 1(a)-(b).
Referring to FIGS. 2(a)-(c), steps of one exemplary process for
fabricating the device 10 are shown. In particular, a self-aligned
process is employed to form the fluidic channel 16 and waveguides
12, 14 between the four pairs of electrodes 20-26 in a manner such
that the electrodes of each pair are parallel to one another. As
shown in FIG. 2(a), in a first step the lower electrode of each
pair of the electrodes 20-26 are all formed atop a
polydimethylsiloxane (PDMS) substrate 29 (more particularly, two
electrodes from the first and second pairs of electrodes 20 and 22
are shown). Atop the four lower electrodes are then positioned the
waveguides 12, 14, and atop the waveguides is positioned a handle
wafer 31. The target region 17 of the fluidic channel 16 is within
the space formed by the adjacent electrodes and waveguides. In the
present embodiment, the electrodes are made from gold (Au), albeit
other materials can be used in alternate embodiments.
[0034] Once the structures have been assembled as shown in FIG.
2(a), and in particular once the lower electrodes have been formed
by way of that step, the handle wafer 31 is delaminated from atop
the waveguides 12, 14, at a step shown in FIG. 2(b). The removal of
the handle wafer allows for the upper electrodes of each of the
pairs of electrodes to be formed above the waveguides 12, 14. It is
desirable that the upper electrodes and lower electrodes of each
pair of electrodes 20-26 be aligned (e.g., parallel) with each
other, and that the set of four upper electrodes and the set of
four lower electrodes be respectively formed on two different
planes on both (opposite) sides of the fluidic channel. To achieve
such alignment, the addition of the upper electrodes of each pair,
which can be formed within an electrode-patterned substrate can
involve the use of an optical mechanical alignment tool such as a
contact mask aligner or a wafer bonder. FIG. 2(c) shows the upper
electrodes to be assembled atop the waveguides 12, 14, and
additionally shows another PDMS substrate 29 to be positioned atop
the upper electrodes. Thus, the target region 17 is entirely
enclosed within the waveguides 12, 14, the pairs of electrodes
20-26 (only two pairs of which are shown in FIG. 2(c)), and the
PDMS substrates 29, and the target region in particular has four
pairs of electrodes around it forming the dielectrophoretic
electric cage 18.
[0035] Turning to FIGS. 3(a)-(b) and 4, additional embodiments of
exemplary fluidic-photonic integrated circuit (FPIC) devices
differing from that of FIGS. 1(a)-(b) are shown. The devices shown
in FIGS. 3(a)-(b) and 4 in particular are guided wave photonic
circuit devices that are more sophisticated in their operation than
the device of FIGS. 1(a)-(b). Further, as with the device 10 of
FIGS. 1(a)-(b), the devices shown in FIGS. 3(a)-(b) and 4 are
fabricated to include microfluidic channels. Through the use of
such microfluidic channels, the devices of FIGS. 3(a)-(b) and 4 can
take a miniaturized form and are capable of delivering new
functions including, for example, functions relating to the
performance of flow cytometry, which is the workhorse for many
biomedical applications. In particular, the devices allow for flow
cytometry with relatively high signal reliability and sensitivity
to be achieved, notwithstanding possible non-uniformities in the
biological samples (or other sensed material) and/or complex flow
patterns. In alternate embodiments, microfluidic channels need not
be employed.
[0036] FIG. 3(a) in particular shows a schematic diagram of a first
FPIC device 40 that can be employed in an improved optical
detector. FIG. 3(b) additionally shows a side cross-sectional view
of a portion of the device 40 (e.g., in cut-away), while FIG. 4
shows a side cross-sectional view of a portion (e.g., in cut-away)
of an alternate embodiment of the device 40, namely, a device 60.
As shown, each of the devices 40, 60 of FIGS. 3(a)-(b) and 4
includes a respective fluidic channel 50, 70 including a respective
vertical section 48, 68. As shown particularly in FIG. 3(a), the
fluidic channel 50 of the device 40 extends between a first fluidic
inlet/outlet 33 and a second fluidic inlet/outlet 34 (although not
shown, the fluidic channel 70 of FIG. 4 also extends between two
such inlets/outlets). Although the sections 48, 68 are shown to be
vertically-oriented, the sections need not be oriented in this
manner and instead could take on other orientations, such as
horizontal orientations.
[0037] In addition to the fluidic channels 50, 70 and associated
vertical sections 48, 68, each of the devices 40, 60 includes a
respective first excitation waveguide 42, 62 and a respective
second excitation waveguide 44, 64. The respective excitation
waveguides of each respective device 40, 60 are aligned with, and
extend in opposite directions from opposite ends of, the respective
vertical section 48, 68 of the respective device. Thus, the
waveguides 42, 44 are each aligned with the vertical section 48,
with the first waveguide 42 extending upward away from the top of
the vertical section and the second waveguide 44 extending downward
from the bottom of the vertical section. Likewise, the waveguides
62, 64 are each aligned with the vertical section 68, with the
first waveguide 62 extending upward away from the top of the
vertical section and the second waveguide 64 extending downward
from the bottom of the vertical section.
[0038] The waveguides 42, 44 of the device 40 and the waveguides
62, 64 of the device 60 each perform a similar function to that
performed by the waveguides 12, 14 of the device 10 of FIGS.
1(a)-(b), namely, to cause excitation light to be directed toward
(and possibly away from) a target region, where the target regions
in these embodiments are the vertical section 48 and the vertical
section 68, respectively. More particularly, the two waveguides 42,
44 and 62, 64 near the ends of the respective vertical sections 48,
68 of the respective fluidic channels 50, 70 deliver optical power
for fluorescent excitation. The excitation light can come from
various directions including from the top or the bottom of the
device (e.g., as shown in FIGS. 3(a)-(b) and 4, similar to a
microscope setup in terms of the direction of light passage through
the microscope lens) as well as, in alternate embodiments, from
other directions. The waveguides 42, 44, 62 and 64 as integrated on
the devices 40, 60 in particular are able to provide convenient
access of the excitation light at chosen wavelengths and in
well-defined excitation directions. The well-defined excitation
directions achieved through use of the waveguides 42, 44, 62 and 64
in particular facilitate the measurements of forward, side, and
back scatterings of light as it encounters the sample particles or
analytes flowing through the vertical sections 48, 68 of the
fluidic channels 50, 70.
[0039] In addition to the respective pairs of excitation waveguides
42, 44 and 62, 64, each of the respective devices 40, 60 also
includes a respective first array of horizontal waveguides 52, 72
and a respective second array of horizontal waveguides 54 and 74,
respectively. The waveguides of the respective first and second
arrays 52 and 54 respectively are arranged oppositely one another
on left and right sides of the vertical section 48, and extend
horizontally in opposite directions away from that vertical
section, while the waveguides of the respective first and second
arrays 62 and 64 respectively are arranged oppositely one another
on left and right sides of the vertical section 68, and extend
horizontally in opposite directions away from that vertical
section. More particularly as shown, in the present embodiments,
each of the arrays 52, 54, 72 and 74 has eight waveguides that
extend parallel to one another horizontally away from the
respective vertical section 48, 68, with each waveguide of each
array being spaced apart from the neighboring waveguide(s) of the
respective array by a predetermined amount of distance (e.g., 100
micrometers between the centers of neighboring waveguides). For
each of the waveguides of the left-side arrays 52 and 72, there is
a corresponding waveguide in the respective right-side array 54 and
74 that is aligned with that left-side array waveguide.
[0040] Each of the waveguides of the arrays 52, 54, 72 and 74 is
capable of functioning as a detection waveguide capable of
conducting/guiding light emanating from a respective one of the
vertical sections 48, 68, and thus serves to allow for optical
interrogation. In the present embodiments, optical detection occurs
by sending excitation light into the vertical sections 48, 68 by
way of one or both of the excitation waveguides 42, 44, 62, 64
associated with the respective vertical section, allowing that
light to interact with and be scattered by the liquid-suspended
sample particle(s) (e.g., cells, DNA or microparticles) flowing
through the respective vertical section, and then sensing the
amounts of scattered light that are received within and transmitted
by the waveguides of the arrays 52, 54, 72 and 74. The light
detected from the waveguides of the arrays 52, 54, 72 and 74 thus
is indicative of the liquid-suspended sample particles (e.g.,
cells, DNA or microparticles) flowing through the respective
fluidic channels 50, 70. As will be described in further detail
below, the detection of light by way of multiple detection
waveguides rather than merely one detection waveguide is
particularly advantageous.
[0041] The dimensions of the fluidic channels 50, 70, the
excitation waveguides 42, 44, 62 and 64 and the detection
waveguides of the arrays 52, 54, 72, and 74 can vary depending upon
the embodiment. In the present embodiments, the cross-sectional
dimensions of the fluidic channels 50, 70 and particularly the
vertical sections 48, 68 of those channels is 50.times.50
.mu.m.sup.2, although other cross-sectional dimensions are also
possible. Additionally, the cross-sectional dimensions of each of
the excitation waveguides 42, 44, 62 and 64 as well as each of the
waveguides of the arrays 52, 54, 72 and 74 in the present
embodiments further are 50.times.50 .mu.m.sup.2, although other
dimensions are also possible. While in at least some embodiments,
including the present embodiments, it is desirable that the
waveguides have substantially the same cross-sectional dimensions
(and shapes) as the corresponding fluidic channels, this need not
be the case. Further, in the present embodiments of FIGS. 3(a)-(b)
and 4, both the excitation waveguides 42, 44, 62, 64 and the
detection waveguides of the arrays 52, 54, 72, 74 are multi-mode
devices with a numerical aperture of 0.3.
[0042] The use of the arrays 52, 54, 72 and 74 of waveguides
results is advantageous in several regards. Because each of the
arrays 52, 54, 72 and 74 has eight parallel spaced-apart
waveguides, each device 40, 60 has a capability of detecting a
particle eight times as it passes through the respective vertical
section 48, 68. More particularly, the use of the eight waveguides
of each of the arrays 52, 54, 72 and 74, by providing eight
detection points for the same target flying by, results in detected
signals that have improved signal-to-noise ratios compared with
conventional optical detection systems and in which randomness
caused by Brownian motions is suppressed. In addition, the use of
these arrays of waveguides allows for the performing of
time-of-flight measurements, to determine the velocity of particles
flying by, as well as allows for multi-label fluorescent detection
with wavelength filters. Compared with single-point detection, the
data that can be obtained using such multiple sampling points can
provide rich information about the properties of a particle (or
particles) under scrutiny, as well as the particle's interplay with
the fluid, and the statistic behaviors under Brownian forces.
[0043] As discussed above, on the left sides of the respective
vertical sections 48, 68 of the respective fluidic channels 50, 70
are positioned respective left-side arrays 52, 72, each of which
has eight detection waveguides with separated outputs. Likewise, on
the right sides of those vertical sections 48, 68 are positioned
respective right-side arrays 54, 74, each of which also has eight
detection waveguides. However, while the right-side array 54 of
FIG. 3(b) has eight separated outputs associated with its
respective eight detection waveguides, the right-side array 74 of
FIG. 4 rather includes an 8.times.1 waveguide combiner or
demultiplexer 76 such that all of the eight waveguides of that
array eventually are merged to form a single output waveguide. That
is, while the array 74 of the device 60 includes eight horizontal
detection waveguides extending away from the vertical section 68,
these waveguides eventually bend toward one another and are
merged/joined with one another as they proceed farther away from
the vertical section.
[0044] FIG. 4 in particular shows how four adjacent pairs of the
eight waveguides of the array 74 bend and merge with one another to
form an array of four waveguides 77. However, it should further be
understood that FIG. 4 only shows the portion of the demultiplexer
76 in which eight waveguides are merged into four waveguides, and
that the demultiplexer additionally involves the merging of those
four waveguides into two waveguides and then subsequently into a
single output waveguide. During operation of the device 60, the
demultiplexer 76 receives time-multiplexed signals from eight
detection zones corresponding to the eight waveguides of the array
74. As described in further detail below in relation to FIG. 7(a)
et seq., through the use of the information obtained from the array
74, the 8.times.1 demultiplexer 76 is able to generate an overall
signal that includes all of the information obtained from the eight
detection waveguides, and to restore the time domain signal
chain.
[0045] Although both the devices 40 and 60 of FIGS. 3(a)-(b) and 4
are capable of achieving enhanced detection sensitivity through the
use of the multiple detection waveguides in their respective
waveguide arrays 52, 54, 72 and 74, the use of the demultiplexer 76
in the device 60 makes it possible to reduce the number of
detectors receiving the light communicated by way of the waveguides
to only one detector (or at least to a number of detectors less
than the number of waveguides that are interfacing the fluidic
channel). Thus, use of the demultiplexer 76 allows for a savings of
the hardware costs associated with having multiple detectors
(albeit at an expense of device throughput).
[0046] The detection sensitivity achieved by the devices 40, 60 is
dependent upon the number of detection waveguides of the arrays 52,
54, 72 and 74. This is true both whether the demultiplexer 76 is
employed or not employed (where the demultiplexer is not employed,
as discussed in further detail below, the multiple signals provided
by the multiple waveguides of each array can be used to perform
cross-correlation operations so as to achieve enhanced
signal-to-noise ratios). Indeed, the detection sensitivity can be
enhanced by increasing the number of waveguides of the array 74
and/or the space-demultiplexed waveguide structure. Nevertheless,
in the present embodiments of FIGS. 3(a)-(b) and 4, with a total of
eight waveguide detection channels/zones, the detection sensitivity
can be enhanced by nearly 1,000 times in comparison with that
afforded by a single channel device. This manifests one advantage
of the FPIC approach over free-space optical set-ups used in
performing conventional flow cytometry because for the latter, the
number of interrogation zones is significantly limited by space and
cost.
[0047] The fluidic-photonic integrated circuit (FPIC) devices 40,
60 in the embodiments of FIGS. 3(a)-(b) and 4 are made entirely (or
substantially) of polymer such as PDMS, and are fabricated by way
of micro-molding and waveguide capillary filling. Details of the
micro-molding process that can be utilized to create the devices
40, 60 of FIGS. 3(a)-(b) and 4 are illustrated in FIG. 5(a)-(b). As
shown in FIG. 5(a), a soft lithography process can be employed to
fabricate microfluidic channels such as the channels 50, 70
discussed above. As shown, a semiconductor or glass wafer 150 is
provided at a step 152, and then at a step 154 a spin-on thick
resist 156 is added to the wafer to form a combined structure 158.
Further, at a step 160, the combined structure is exposed to
ultraviolet light 162 by way of a mask (not shown). Subsequent to
the photo-lithographic patterning step 160, the combined structure
158 is now a modified combined structure 165 having a modified
(patterned) thick resist layer 167. The modified combined structure
165 can also be referred to as a "mold master".
[0048] Subsequent to the step 160, the modified combined structure
165 is then baked by way of a thermo-curing process, for example,
thermo-curing at 65.degree. C. for 4 hours. Then, at a step 166, a
pre-polymer layer 168 is cast and cured upon the modified combined
structure 165. Then, at a step 170, the pre-polymer layer 168 is
peeled from the mold master, thus transferring the pattern to the
pre-polymer layer. In at least some embodiments, the mold master is
made from photo-lithographically patterned SU-8-50 photoresist
(such as that available from MicroChem, Inc. of Newton, Mass.) on a
4'' silicon wafer (such as that available from Silicon Quest
International, Inc. of Santa Clara, Calif.). Also, the pre-polymer
layer 168 cast onto the mold master can be a PDMS layer such as
Gelest OE 41 available from Gelest, Inc. of Morrisville, Pa.
[0049] The steps shown in FIG. 5(a) result in the creation of the
pre-polymer (PDMS) layer 168 that has channel patterns 172
complementary to the patterns of the modified thick resist layer
167. Referring additionally to FIG. 5(b), in order to make an
enclosed microfluidic channel, the pre-polymer layer 168 is in turn
bonded to another pre-polymer (e.g., PDMS) layer 174. Typically,
both of the pre-polymer layer 168 and 174 will have the same
refractive index (e.g., 1.407). A short treatment (e.g., 10
seconds) of high power (e.g., 100 Watts) oxygen plasma (e.g., using
the Technics 500-II Plasma Etcher and Asher System) can be used to
activate the surfaces of the pre-polymer layers 168, 174 to
facilitate permanent bonding of those layers, thus completing the
fabrication of one or more microfluidic channels corresponding to
the patterns 172. In alternate embodiments an ultraviolet/ozone
treatment can be employed instead of the oxygen plasma treatment to
achieve bonding.
[0050] Various similar molding techniques can also be used to
fabricate ridge waveguides with a chosen polymer of an appropriate
refractive index, which can be employed as the waveguides 42, 44,
62, 64 or the waveguides of the waveguide arrays 52, 54, 72 and 74
described above. One such technique that can be employed to make
the waveguides is a channel-waveguide filling process. To make
channel waveguides by way of this technique, some chosen channels
are filled with a polymer of higher refractive index, for example,
Gelest OE 42 PDMS (also available from Gelest Inc.) can be chosen
as the core material. Pre-polymer is introduced into the channels
through the inlets. Pre-polymer can completely fill the channels in
a short period of time, e.g., 20 minutes. Then, upon performing the
same thermo-curing procedure as mentioned before, the core material
is solidified and takes on a desired refractive index (e.g., 1.42).
It should be further noted that, if striations at the end of a
waveguide are caused due to waveguide cutting, these can be removed
by polishing; at the same time, facet polishing does not appear to
be usually necessary since the striations have not appeared to
disturb experimental signal measurements.
[0051] For the purpose of realizing monolithic fluidic-photonic
integrated circuits (FPICs), several integration schemes can be
utilized. For example, for high sensitivity chemical sensors that
require long interaction length, one can employ an integrated
structure having a single-mode ridge waveguide inside or aligned
with the microfluidic channel so that the light wave propagates in
the same direction as the flow. More particularly, to achieve such
an integrated structure, both the waveguide and the microfluidic
channel are formed in the same fabrication process, with the
waveguide being formed by creating passage(s) with walls of a
first, lower index of refraction that are subsequently filled with
a liquid that, upon being solidified by heat or UV treatment, takes
on a second, higher index of refraction. The microfluidic channel
can have the same cross-sectional dimensions as the waveguide, or
the two structures can have different cross-sectional dimensions,
as desired. Also, for applications where on-chip optical processing
or contact-free detection is needed, one can also employ a stacking
structure in which waveguides and fluidic channels are located at
different planes so that the waveguides and fluidic channels can be
routed without crossing one another.
[0052] Still another manner of integrating waveguides and fluidic
channels is a self-forming technique that utilizes a capillary
effect and immiscibility between liquids inside microfluidic
channels to create waveguides that intersect the microfluidic
channels. Such an integration scheme is particularly suitable for
highly-localized fluorescent excitation and detection. For the
self-forming process, a fluidic channel and a waveguide channel (or
multiple waveguide channels) are formed that intersect with one
another. The fluidic channel is then filled up with BSA (protein)
aqueous solution before liquid (pre-polymerized) PDMS is provided
to fill the waveguide channel(s) that intersect the fluidic
channel. Since the liquid PDMS and the BSA solution are immiscible,
the liquid PDMS will not enter the fluidic channels even though
they intersect. After the liquid PDMS is thermally (e.g., at 60
degrees C.) or UV cured to become solid PDMS, the BSA solution is
removed from the fluidic channel, yielding an array of waveguides
in very close proximity to the fluidic channels for efficient light
coupling. In at least some embodiments of FPICs to be used for flow
cytometry, a waveguide/channel integration structure similar to but
simpler than what is used in the self-forming technique can also be
employed.
[0053] Notwithstanding the above discussion, the present invention
is intended to encompass a variety of FPIC devices having features,
or being fabricated by way of techniques, other than those
mentioned above. For example, in alternate embodiments, FPIC
devices can have any number of detection waveguides corresponding
to the arrays of waveguides 52, 54, 72 and 74, including more than
eight or less than eight waveguides in each array (or even only one
waveguide on each side of the fluidic channel). Likewise, although
the waveguides of the arrays 52, 54, 72 and 74 are perpendicular to
the vertical sections 48, 68 of the fluidic channels, in alternate
embodiments, the waveguide(s) can approach the fluidic channels at
oblique or other angles. Curved waveguide surfaces can also be
formed to create light focusing effects to either increase the
numerical aperture of the waveguides or to move the waveguides
further away from the fluidic channels. It should further be noted
that, while in at least some embodiments such as those discussed
above, the FPIC devices are PDMS-based microchips, in alternate
embodiments the FPIC devices can be made from other materials and
via other processes than are used to develop PDMS-based
microchips.
[0054] Referring to FIG. 6, an exemplary flow cytometry system 80
employing a FPIC device 100 similar to the device 40 of FIGS.
3(a)-(b) (e.g., without any demultiplexer) is shown in schematic
form. As shown, the FPIC device 100 includes the microfluidic
channel 50 having the vertical section 48 and the first and second
inlet/outlets 33, 34, as well as the excitation waveguides 42 and
44. However, in this embodiment (contrary to that of FIGS.
3(a)-(b)), only one of the waveguide arrays 52, 54 (namely, the
right-side array 54) is employed and, as a result, signals are only
detected along one side of the microfluidic channel 50. Further as
shown, the first inlet/outlet 33 functions as an inlet and receives
pumped fluid from plastic tubing 81, by which the inlet is coupled
to a pumping unit 82. The second inlet/outlet 34 in turn functions
as an outlet that in the present schematic diagram is shown to be
left open but which typically is coupled to a receptical such as a
waste beaker (e.g., in an experimental set-up) or can be coupled
back to the pumping unit 82. As part of the standard microfluidic
device fabrication and packaging process, appropriate connectors
can be fabricated to connect the plastic tubing 81 to the inlet 33
on the chip without leakage. In at least some embodiments, a
syringe pump can be used as the pumping unit 82 to deliver liquid
samples.
[0055] Also, in the present embodiment, laser excitation light from
a source 84 is coupled by way of multi-mode optical fiber 86 to one
of the two waveguides 42, 44 (in this case the waveguide 44) that
are aligned with and face the respective ends of the vertical
section 48 of the fluidic channel 50. In the present embodiment, to
secure the connection between the optical fiber 86 and the
waveguide 44, a multi-mode fiber is inserted into the waveguide
channel prior to filling of that channel with pre-polymerized PDMS
and solidification of that PDMS (before the core material of the
waveguide is solidified) such that, after PDMS curing and
solidification, an encapsulated fiber-waveguide structure is
created. Such an encapsulated fiber-waveguide structure is capable
of showing low insertion loss (<0.3 dB) and negligible interface
reflection, as well as mechanical robustness. Further as shown in
FIG. 6, the light transmitted by the detection waveguides of the
array 54 away from the fluidic channel 50 is directed toward an
objective lens 88, which in turn provides that light to a CCD
camera 90. A near-field image thus is formed on a CCD camera screen
of the camera 90.
[0056] Depending the upon the embodiment, the camera 90 or another
device in communication with the camera (not shown) can include a
processing devices that receive signals from the camera, allowing
for further processing operations to be performed, some of which
are described in further detail below. The processing can be, for
example, a microprocessor, programmable logic device or integrated
circuit device such as a digital signal processing (DSP) chip or
other processing device. In at least some embodiments, the
processing device can be part of, or assume the role of, a control
device or controller capable not only of processing information but
also capable of generating and controlling the output of (e.g., on
a display or onto a network, such as the internet) signals,
information, or data. In some such embodiments, the controller also
is capable of monitoring and/or controlling other
devices/components of the flow cytometry system 80 such as the
pumping unit, the excitation laser, and/or other
devices/components. Further, in still other embodiments, a
processing device and/or controller can be coupled to receive
information from the detection waveguides via a device other than
the camera 90. It should further be understood that embodiments of
the invention not being employed for the purpose of flow cytometry
also can employ a processing device or controller similar to that
described above. Additionally, it should be understood that the
processing device/controller should be generally understood to
encompass one or more memory devices or computer-readable storage
media capable of governing operation of the processing
device/controller.
[0057] Although not shown in FIG. 6, a demultiplexer such as the
demultiplexer 76 can also be employed as part of (or in conjunction
with) the array of waveguides 54. As mentioned above, the 8.times.1
demultiplexer 76 of FIG. 4 serves to reduce the number of optical
detectors needed to receive the light from the array of waveguides
74. Also, the demultiplexer 76 allows for the restoring of a
time-domain signal chain from noise-masked data. Referring to FIGS.
7(a)-(b), exemplary output of a 8.times.1 demultiplexer such as the
demultiplexer 76 of FIG. 4 in response to detected signals is
provided to illustrate such operation of the demultiplexer. FIG.
7(a) in particular shows data that was obtained using a sample
containing 5 .mu.m fluorescent microbeads, while FIG. 7(b) in
particular shows data that was obtained using a sample containing 1
.mu.m fluorescent microbeads. The data represents the combination
of the signals from 8 separate channels (e.g., the combination of
all of the eight waveguides of the array 74).
[0058] When the detected output signal is weak, the raw data shown
in FIGS. 7(a) and (b) can be largely corrupted by noise due to
stray light and the electronic noise of the CCD camera.
Nevertheless, in spite of the potentially poor signal quality, it
is possible to perform the following algorithm to restore the
signal chain in the time domain, utilizing the property that the
signals detected at the eight sequential detection zones provided
by the eight waveguides of the array 74 are time-correlated.
Specifically,
S(t)=f.sub.1(t)*f.sub.2(t-T)*f.sub.3(t-2T)*f.sub.4(t-3T)*f.sub.5(t-4T)*f-
.sub.6(t-5T)*f.sub.7(t-6T)*f.sub.8(t-7T) (1)
where S(t) is the time-dependent signal and T is the time interval
for a particle to pass through two adjacent waveguide channels
(e.g., to pass from one of the waveguides 74 to a neighboring one
of the waveguides 74). The value of T can be obtained from the
time-of-flight measurement.
[0059] Thus, using Equation (1), the time dependent signal S(t) is
determined based on a concept of time-correlation among the
detection waveguides, as represented by multiple signals
f.sub.i(t). Additionally, it should be noted that, to remove the
effects of high background and baseline drift, it is advisable to
have each signal f.sub.i(t) be passed through a high-pass filter
before performing the operation according to Equation (1). Further,
since all of the various f.sub.i(t) signals represent signal
intensity, negative values will be removed. For instance, should
f.sub.1 happens to show a negative value at a certain time "t",
then the signal S(t) at the time t will become the product of the
remaining 7 terms and its final value will be normalized by the
power of 8/7. In other words, the normalized signal in the above
case at the particular time becomes S.sup.8/7.
[0060] Turning to FIG. 8(a), as mentioned above, the arrays of
eight parallel waveguides such as the array 54 of FIGS. 3(a)-(b)
also can be used to perform time-of-flight measurements. More
particularly, when a bead travels through the interrogation region
(e.g., within the vertical section 48 of FIG. 4), its fluorescence
is detected by each of the eight waveguides of the array 54
sequentially. The output intensity of each of the eight waveguides
is recorded by the CCD camera 90. FIG. 8(a) shows exemplary
intensities of the signals provided by the eight waveguide channels
as functions of time, which were obtained using an experimental
set-up employing 10 .mu.m-diameter fluorescent beads. The
center-to-center difference between intensity peaks (T) is the time
period when a bead particle travels across two adjacent waveguides.
Knowing the distance between the centers of adjacent waveguides of
the array 54, the velocity of the bead particle can be easily
obtained. It should be further noted that, in the case of 10 .mu.m
fluorescent beads, images of high signal-to-noise ratio can be
obtained directly from the output of any single waveguide.
[0061] Referring to FIGS. 8(b)-(c), to demonstrate the ability of
sensitivity enhancement with the array waveguide structure,
exemplary time-of-flight measurements were also performed using
fluorescent beads of smaller diameters than the 10 .mu.m beads that
were the basis of FIG. 8(a), namely, 5 .mu.m (FIG. 8(b)) and 1
.mu.m (FIG. 8(c)). Because these fluorescent micro-beads have
fluorescent dye doped over their entire volumes, the fluorescent
intensity of each respective bead is proportional its volume,
making the fluorescence intensity eight times and one-thousand
times weaker than that of the 10 .mu.m beads. FIGS. 8(b) and 8(c)
respectively demonstrate that the directly-detected signals at the
output of each waveguide channel resulting from the 5 .mu.m beads
and 1 .mu.m beads, respectively, can be obscured and masked by the
noise. Such output from any single channel, analogous to the signal
obtained from a conventional flow cytometer using a low power
excitation source and a low sensitivity detector, cannot produce
any meaningful signal (it is for that reason that conventional flow
cytometry systems require high power lasers and photomultiplier
tubes with photon counting sensitivity, which are expensive and
non-scalable in size).
[0062] FIGS. 8(b)-(c) demonstrate that it can become more difficult
to achieve desired signal-to-noise ratios as the particles to be
sensed become smaller. In these circumstances, the methodology
described above with respect to Equation (1) for restoring a signal
from noisy measurements may be inadequate for achieving output
signals having desired signal-to-noise ratios. In particular,
although Equation (1) provides a method that is mathematically
simple, a more robust method to restore the real signal produced by
each passing particle/analyte may be useful. In accordance with
additional embodiments of the present invention, one such more
robust method for providing output having improved signal-to-noise
ratios is an additional multi-channel detection technique that
involves cross-correlation analysis. This calculation assumes that
signals are correlated to beat between the different waveguides of
a waveguide array such as the array 54, and takes advantage of the
knowledge that the true (light output) signals are time-correlated
while the noise is not.
[0063] The concept of cross-correlation is further illustrated
graphically in FIG. 9. As illustrated, when a given particle passes
through the vertical section 68 of the fluidic channel 50, first
and second signals 151 and 152 will be generated at two of the
neighboring detection waveguides of the array 54. As long as the
signals 151, 152 from the two channels are time correlated, it is
possible to obtain the time delay .tau. between the two signals by
calculating a cross-correlation function R(.tau.) defined in
Equation (2):
R(.tau.)=ff1(t)*f2(t-.tau.)dt (2)
where f1 and f2 are normalized intensity functions of two
individual channels and .tau. is a time domain variable. The
cross-correlation function R(.tau.) (also shown in FIG. 9) that
maximizes .tau. becomes equal to the time-delay between the two
signals.
[0064] The above-described cross-correlation function can be
extended for implementation in relation to an array having an
arbitrary number of detection waveguides. For example, with respect
to the FPIC device 100 in the above-described embodiment of FIG. 6
that has eight waveguide channels, the above analysis can be
extended to calculate an eight-channel cross-correlation, as
follows:
R(.tau.)=f1(t)f2(t-.tau.)f3(t-2.tau.)f4(t-3.tau.)f5(t-4.tau.)f6(t-5.tau.-
)f7(t-6.tau.)f8(t-7.tau.)dt (3)
In comparison with Equation (2) with its two terms, this eight-term
multiplication further serves to amplify the signal and suppress
the noise. It should further be noted that, with respect to
Equations (2) and (3), it is not necessary to assume that every
particle/analyte travel at exactly the same speed as is assumed in
Equation (1), since the application of Equations (2) and (3)
involves the calculating of the travel velocity and time delay of
each passing particle/analyte.
[0065] In view of the above considerations, use of an FPIC
device/system (and especially a microfluidic FPIC) such as the FPIC
device 100 and system 80 of FIG. 6 can be particularly advantageous
when implemented in conjunction with cross-calculation techniques.
Although the cross-correlation calculations according to Equations
(2) and (3) are more computation heavy than those according to
Equation (1), the calculations according to Equations (2) and (3)
alleviate the requirement for keeping all of the particles/analytes
in the streamline of a constant velocity, thus greatly simplifying
the design and processing complexities of fluidic channels.
Further, because the FPIC devices can accommodate essentially any
number of detection waveguide channels without increasing the cost
and complexity of the system substantially, the advantages in
signal quality achievable using the cross-calculation techniques
are fully realizable.
[0066] Additionally, although the cross-correlation calculations
according to Equations (2) and (3) when performed in the manner
described above can be relatively computation intensive, a further
time domain spectroscopy detection process shown by a flow chart
200 in FIG. 10 can reduce the computational intensity of
cross-correlation. In particular, as shown in the flow chart 10,
upon starting the flow chart at a step 202, the process begins at a
step 204 by obtaining the signal information from the waveguides of
the waveguide array (e.g., by way of a CCD camera receiving signals
from the waveguides of the array 54 of FIG. 6). Then, rather than
calculating the time delay .tau. associated with the movement of
particles between neighboring waveguides of the waveguide array, at
a step 206 a value for the time delay is assumed. The time delay
.tau. in particular can be thought of as representing a range of
times centered about a center time that is .tau. (e.g., .tau.=0.5
ms+/-0.5 ms). Subsequently, at a step 208, the cross-correlation
algorithm (e.g., Equation (3) for an eight-waveguide array) is
applied to the signal information and, at a step 210, it is
determined whether the result from performing this operation is
zero (or substantially zero, e.g., negligible).
[0067] If the result is not zero (or substantially below a set
threshold value), then at a step 212 a new value is assumed for the
time delay, and step 208 is re-performed given that new value. The
newly-assumed value of the time delay .tau. is typically an
adjacent, incremental value relative to the previously-assumed
value (e.g., given the first assumed value mentioned above, the
next assumed value would be .tau.=1.5 ms+/-0.5 ms). Given that the
step 212 cycles back to the step 208, the steps 208, 210 and 212
can be repeated iteratively as long as successive cross-correlation
calculations produce non-zero results. However, once the result at
step 210 is determined to be zero (or substantially zero), then it
is further determined at a step 212 whether that has occurred
already for a number of (e.g., N) iterations. If not, then the
process again returns to step 212 at which another time delay is
assumed, and further proceeds to repeat steps 208 and 210. However,
if upon reaching step 212 it is determined that there have already
been N zero results corresponding to N different time delay values,
then the process diverts to a step 214 at which a sum of all of the
different results corresponding to the different assumed time delay
values is calculated, and subsequently to a step 216 at which the
process ends. The step 214 is optional and, in some embodiments is
not performed such that the process diverts directly to step 216
from step 210.
[0068] Due to the use of the assumed values of the time delay
.tau., the cross-correlation computation process of the flow chart
200 can be advantageous in comparison with the previously-described
cross-correlation processes. In particular, by assuming the values
of the time delay, computational effort need not be expended in
determining the actual time delay value. Further, although numerous
(e.g., fifty or more) iterations need to be performed in some
circumstances to reach the criterion of the step 210 at which the
iterations are stopped, this does not take excessive time since,
given an appropriate digital signal processing (DSP) chip/device
that performs time-shifting (e.g., any of several DSP chips
available from Texas Instruments, Inc. of Dallas, Tex. that are
capable of performing several hundred million instructions per
section (MIPS)), the different iterative calculations can be
performed simultaneously or nearly simultaneously in parallel. The
speed at which the calculations are made in particular can surpass
the flow rate of the particles/beads within the fluidic
channel.
[0069] It should also be noted that the result obtained from
applying the cross-correlation algorithm (e.g., Equation (3))
during each iteration at the step 208 is representative of the
amount or number of particles flying by the waveguides of the
waveguide array at a given speed corresponding to the assumed time
delay, and thus the individual results are of individual interest
as being representative of the amount of particles passing through
the fluidic channel at different speeds. This can be valuable in
the context of flow cytometry, particularly where it may be
expected (or of interest) that different cells or different DNA
base pairs travel at different speeds. At the same time, the sum of
all of the individual results calculated optionally at the step 214
also can be of interest, as an indication of the total sample
intensity.
[0070] Referring additionally to FIGS. 11(a) and (b), these FIGS.
provide an illustration of the effectiveness of the above-described
cross-calculation techniques when implemented in relation to the
system 80 of FIG. 6 with its FPIC device 100 having the array 54 of
eight waveguides. More particularly, FIGS. 11(a) and (b)
respectively illustrate exemplary cross-correlation data that can
be obtained by applying Equation (3) to the raw data of FIGS. 8(b)
and 8(c), respectively, with each of the FIGS. 11(a)-(b) displaying
R(.tau.) as a function of .tau.. Clearly the signals have been
completely restored as manifested by pronounced peaks 195 of
R(.tau.) shown in each case. As shown, the maximum of R(.tau.)
occurs at 0.3 sec and 0.08 sec for the two cases involving 5 .mu.m
and 1 .mu.m beads, respectively. These are the durations required
for the respective particles to travel across two neighboring
waveguide channels of the array of waveguides. For a
center-to-center channel spacing of 100 .mu.m, the velocities of
particles in each case are 333 .mu.m/sec and 1250 .mu.m/sec
respectively. Therefore, the eight-channel waveguide array shows
its superb ability of signal enhancement to allow for
time-of-flight measurement on even extremely weak fluorescent
beads. The velocity obtained in this way is a direct measurement of
particle speed and can be used for in-situ calibration of the
fluidic system.
[0071] Turning to FIGS. 12(a)-(b), for particle detection and
sorting, it is further desirable to measure signals in real time as
an intensity signal chain. To generate the signals shown in FIGS.
12(a)-(b), the noise-masked raw data in FIGS. 7(a)-(b) obtained
using the output of the 8.times.1 demultiplexer 76 and pertaining
to the 5 .mu.m and 1 .mu.m bead data, respectively, is processed
through Equation (1), with the value of T obtained from the
previous time-of-flight measurement. Distinctive peaks (e.g.,
groups of spikes) 180 with side-lobes 181 represent passing beads
in real time. From the results in FIGS. 12(a)-(b), it is evident
that the values of the peaks 180 of the 5 .mu.m beads are many
(e.g., 17) orders of magnitude greater than those of 1 .mu.m beads,
which is due to the fact that Equation (1) is a product of eight
correlated signals and therefore significantly magnifies the
difference in signal intensity. These results suggest that the
scheme of multi-channel detection not only improves the
signal-to-noise ratio but also enhances the ability to distinguish
signals of slightly different intensity, which is an important
capability in the context of flow cytometry. (It should be noted
that, in FIGS. 12(a)-(b), intensity can be of any arbitrary unit,
and the dimension "number of frame" corresponds to time with, more
particularly, one frame corresponding to 1/30 second.)
[0072] It should further be noted that, for detection and sorting
in some conventional flow cytometry systems, emission intensity is
used to identify targets of different characteristics. Although
cell sorting by intensity is the predominant and simplest method,
it is not as reliable and accurate as desired when the intensity
difference between normal and targeted samples is small. As shown
in FIGS. 12(a)-(b), the peaks produced by passing beads show
considerable differences in their magnitude even though these beads
belong to the same group within a variation of a few percents in
their size and shape. Variations in the values of the peaks 180 can
be partly attributed to non-uniformity of the beads. These large
intensity variations of these "similar" beads suggest the ability
of detection schemes in accordance with at least some embodiments
of the present invention to distinguish samples of only small
difference, thus making cell detection/sorting by intensity more
reliable and accurate.
[0073] In at least some further embodiments, it is desirable that
compensation be provided to account for variation in the distance
of a given bead passing each waveguide, which otherwise can cause
signal intensity change due to the variation of light coupling
efficiency. In some such embodiments, optical detection from the
opposite side of the fluidic channel can be utilized to eliminate
the effects of positional variation, since any positional variation
of a bead is supposed to produce an anti-correlation signal between
the oppositely-located waveguides. For example, using a FPIC device
similar to the FPIC device 40 having the arrays 52 and 54, signals
provided by the waveguides of the array 52 would compensate for the
signals provided by the waveguides of the array 54. If a bead is
away from the center of the fluidic channel, then the signal
intensity from the waveguides near the bead (for example, the
waveguides of the array 52) increases and the signal intensity from
the waveguides farther from the bead (for example, the waveguides
of the array 54) decreases. This yields a negative correlation
between the output signals from the waveguides on both sides of the
channel. At the same time, intrinsic property variations of the
bead will produce a positive signal correlation. That is, when a
bead has a low fluorescence efficiency, output signals from the
waveguides on both sides are reduced. In addition, it should also
be mentioned that improving the flow channel design such as using
multiple stream laminar flows can also suppress the undesirable
effects of positional variation.
[0074] Another noticeable feature for the signal peaks 180 in FIGS.
12(a)-(b) is the occurrence of the side-lobes 181. In an ideal case
when the background noise is small, high main-lobe-to-side-lobe
ratios should be obtained due to the algorithm of Equation (1).
However, when the signal of each waveguide channel is comparable to
the noise, the nonzero background potentially raises the magnitude
of the side-lobes, which can become problematic particularly when
the bead population increases sufficiently that the signals
produced by neighboring beads interfere with each other through
their side lobes. Nevertheless, it should be noted that the
occurrence of side lobes or multiple peaks for a single passing
bead is the result of using the simplest algorithm, namely, that of
Equation (1). If Equation (3) is instead employed such that a
sliding time scale is used to define the lower and upper limit of
integration time interval, there will be no side lobes and the
signal appears to be a sequence of peaks similar to those in FIGS.
11(a)-(b) in time domain. In that situation, the specific time for
each peak represents the arrival time of the bead to the first
waveguide. To avoid crosstalk, the flow rate should be controlled
so that two beads are not passing the array at any given time (so
as to set a limit of the device throughput for each fluidic
channel, as well as possibly a system throughput, where the system
throughput is the product of the throughput of each fluidic channel
and the total number of channels).
[0075] In addition to the above-described embodiments, the present
invention is intended to encompass a variety of other devices,
systems and techniques/methodologies that include one or more of
the above devices, systems and techniques/methodologies and/or
portions thereof. For example, although not shown in the FIGS., in
at least some additional embodiments, a CCD connected microscope
can be placed on top of the sample device to simultaneously monitor
the events happening in the fluidic channel in order to verify the
counting accuracy. By comparing the monitoring video with the
waveguide channel data obtained during the same period of time, it
is possible to address the correspondence between individual beads
and peaks.
[0076] Although such a methodology involving both the waveguide
signals and monitoring video can be helpful, it is limited in two
respects. First, the error rate determined by this methodology is
only valid when the flow rate is slow, due to the restriction of
the speed of the monitoring CCD. Further, the verified period of
time can be limited, for example, because the image acquisition and
process are executed by a personal computer. One way of achieving
valuable data notwithstanding these restrictions is to incorporate
real-time signal processing and sorting functions with the current
detection architecture, so that targeted particles can be collected
for counting and calibration. For example, to approach such a
solution, it is possible to design and utilize signal processing
circuits that will trigger the sorting mechanism in real-time when
events are detected. Also, for the sorting part, a new sorting
mechanism using acoustic waves can also be introduced. It is
intended that, in at least some embodiments, a flow cytometer with
complete functions of detection, signal processing and sorting will
be integrated in a single chip platform.
[0077] Although not limited to applications relating to flow
cytometry, at least some embodiments of the present invention can
offer significant cost, size, and performance advantages that have
a potential to improve or even revolutionize conventional flow
cytometry techniques. The technology and the architecture design of
FPICs in accordance with at least some embodiments of the present
invention significantly enhance the detection sensitivity through
multi-point detections, hence opening up the possibility of using
low cost light sources (e.g., light-emitting diodes (LEDs) and
lamps) and detectors (e.g., semiconductor avalanche photodiode
detectors (APDs)) to replace mainframe lasers, photomultiplier
tubes (PMTs), and lock-in amplifiers. It also offers new functions
such as measurements of particle velocity, quantum efficiency
fluctuation, signal difference between similar samples, etc. that
could provide new insights in relation to biosensing. Additionally,
the FPIC platform offers a natural path to form array structures
for parallel processing, which makes up for the possible throughput
reduction due to the lower flow rate of microfluidic circuits. As
described above, FPIC devices can be made of polymer materials by
way of simple yet controllable methods (e.g., micro-molding and
capillary channel filling), and the devices can be readily
transferred to semiconductor or silica substrates for integration
with optoelectronic or electronic devices.
[0078] As discussed above, a significant purpose of at least some
embodiments of the FPIC devices of the present invention when
implemented for on-chip flow cytometry applications is to enhance
the sensitivity of fluorescent detection using the architecture of
array waveguides that provides multiple detection zones for objects
traveling through the fluidic channel. In addition, the waveguide
arrays can perform time-of-flight measurement and multi-label
fluorescent detection with wavelength filters. Further, the
monolithic integration of waveguides with microfluidic channels as
described above is only one example of a variety of possible
implementations of various structures on photonic ICs to realize
flow-cytometry-on-a-chip and for other purposes. For example,
fabrication methodologies such as those described above (or similar
to those described above) can be employed to incorporate more
functional optical waveguide devices, such as
multiplexers/demultiplexers, power splitters, filters, polarizers,
etc. to further enhance and expand the detection and analysis
functions or for other purposes.
[0079] Although both the eight detector and single detector designs
described above produce superior sensitivity in comparison with
conventional optical detectors, designs such as that of FIG. 4
having an integrated demultiplexer are preferred in at least some
circumstances, since in such designs only one detector rather than
multiple detectors in the form of a detector array is required.
Also, in embodiments where a semiconductor APD or a PMT is used,
the detector bandwidth (e.g., >1 MHz) is more than sufficient to
support the sampling rate (e.g., eight times of the particle
throughput) for unequivocal detection of each passing particle,
cell, bead, analyte, etc. In at least some embodiments, the use of
an eight-channel waveguide array such as that mentioned above also
allows for other measurements to be made that can yield useful
information. These include, for example, time-of-flight
measurements and timing jitter measurements to monitor Brownian
motion and flow effects. Such information makes it possible to
track the behavior of each individual particle in the fluidic
channel, producing insight into the particle properties and signals
for downstream control.
[0080] As mentioned above, the present invention is not limited to
applications relating to flow cytometry but rather is intended to
encompass a variety of embodiments of devices, systems and
processes that can be utilized in a variety of biomedical,
biochemical, and other sensing applications. Also, while in at
least some embodiments of the present invention, one or more arrays
of eight detection waveguides are arranged along one or more sides
of a fluidic channel, in alternate embodiments, lesser or greater
numbers of waveguides than eight waveguides can be employed
(indeed, in at least some embodiments, only one detection waveguide
is positioned along one or both sides of the fluidic channel). The
waveguides can be oriented in a perpendicular manner relative to
the fluidic channel as discussed above, but in alternate
embodiments can be oriented in any particular manner relative to
the fluidic channel. Also, while the FPIC devices shown in FIGS.
3(a)-(b), 4 and 6 employ waveguide arrays arranged on one or two
opposing sides of the vertical sections 48, 68 of the fluidic
channels 50, 70, in alternate embodiments it would further be
possible employ two sets of waveguides that extended at right
angles relative to one another (with the vertical sections serving
as the vertex), or at other angles relative to one another rather
than only at 180 degree angles relative to one another. Indeed, in
further alternate embodiments, three, four or possibly even more
arrays of waveguides can be positioned extending away from the
vertical sections along three or more planes.
[0081] Additionally, while the detection waveguides can be square
or rectangular in cross-section as shown, in alternate embodiments
the waveguides can take on alternative cross-sectional shapes
(e.g., circular cross-sections). Although the sections 48, 68 of
the waveguides can be vertical as shown in FIGS. 4(a)-(b) and 5,
the sections need not be vertical but rather can be horizontal or
oriented in another manner and, in certain embodiments, can also be
curved. Further, although the excitation waveguides 42, 44, 62 and
64 shown in FIGS. 3(a)-(b) and 4 are aligned with the vertical
sections 48, 68, respectively of the fluidic channels, in alternate
embodiments, any of a variety of other types of light sources
(including simply light bulb(s)) could be utilized to illuminate
the fluid flowing within the channels. In such other embodiments,
it would not be necessary that the light be shined through/along
the lengths of the fluidic channels as shown in FIGS. 3(a)-(b) and
4; indeed, in certain embodiments, the light can be directed toward
the fluidic channels from any direction. Indeed, the present
invention is intended to encompass embodiments of FPIC devices in
which light (or other electromagnetic radiation) is detected via
one or more waveguides or other conductive structures, but in which
excitation light (or other electromagnetic radiation) is delivered
to the fluidic channel by way of any of a variety of structures
(including simply light bulbs), not merely by way of one or more
waveguides.
[0082] Additionally, while the fluidic channels (e.g., the channels
50,70) of the devices described above are microfluidic channels,
the present invention is further intended to encompass other
embodiments of devices and systems that employ combinations of
fluid channels and waveguides and/or electrodes even where the
fluid channels are not "microfluidic channels". For example, the
present invention is intended to encompass devices having fluid
channels having dimensions substantially greater than those
considered as being "microfluidic" channels, e.g., channels having
cross-sectional dimensions of greater than micrometers or
millimeters. Further, it is intended that the present invention
encompass methods of constructing FPIC devices that involve
conventional techniques for manufacturing microfluidic channels,
and then supplement those conventional techniques with additional
steps to integrate photonic components (e.g., waveguides) with
those channels/carriers.
[0083] It is specifically intended that the present invention not
be limited to the embodiments and illustrations contained herein,
but include modified forms of those embodiments including portions
of the embodiments and combinations of elements of different
embodiments as come within the scope of the following claims.
* * * * *