U.S. patent application number 11/131145 was filed with the patent office on 2005-09-22 for microfluidic flow manipulation device.
Invention is credited to Johnson, Timothy J., Locascio, Laurie E., Ross, David J..
Application Number | 20050207274 11/131145 |
Document ID | / |
Family ID | 26884357 |
Filed Date | 2005-09-22 |
United States Patent
Application |
20050207274 |
Kind Code |
A1 |
Johnson, Timothy J. ; et
al. |
September 22, 2005 |
Microfluidic flow manipulation device
Abstract
Disclosed is an apparatus and method for the mixing of two
microfluidic channels wherein several wells are oriented diagonally
across the width of a mixing channel. The device effectively mixes
the confluent streams with electrokinetic flow, and to a lesser
degree, with pressure driven flow. The device and method may be
further adapted to split a pair of confluent streams into two or
more streams of equal or non-equal concentrations of reactants.
Further, under electrokinetic flow, the surfaces of said wells may
be specially coated so that the differing electroosmotic mobility
between the surfaces of the wells and the surfaces of the channel
may increase the mixing efficiency. The device and method are
applicable to the steady state mixing as well as the dynamic
application of mixing a plug of reagent with a confluent
stream.
Inventors: |
Johnson, Timothy J.;
(Charlestown, MA) ; Ross, David J.; (Silver
Spring, MD) ; Locascio, Laurie E.; (North Potomac,
MD) |
Correspondence
Address: |
The Law Offices of William W. Cochran, LLC
Suite 230
3555 Stanford Road
Fort Collins
CO
80525
US
|
Family ID: |
26884357 |
Appl. No.: |
11/131145 |
Filed: |
May 16, 2005 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
11131145 |
May 16, 2005 |
|
|
|
10188664 |
Jul 1, 2002 |
|
|
|
6907895 |
|
|
|
|
60323509 |
Sep 19, 2001 |
|
|
|
Current U.S.
Class: |
366/341 |
Current CPC
Class: |
Y10T 137/0318 20150401;
G05D 7/0694 20130101; Y10T 137/87652 20150401; Y10S 366/03
20130101; Y10T 137/2224 20150401 |
Class at
Publication: |
366/341 |
International
Class: |
B01F 013/00 |
Claims
1. A mixer of laminar microfluidic streams propelled by
electrokinetic flow comprising: a first inlet channel; a second
inlet channel; a mixing channel starting at the confluence of said
first inlet channel and said second inlet channel; and a plurality
of substantially straight unconnected wells disposed in said mixing
channel, said wells being obliquely oriented substantially across
the width of said mixing channel.
2. The mixer of claim 1 wherein alternating wells are configured
perpendicular to each other.
3. The mixer of claim 1 wherein said wells are configured parallel
to each other.
4. The mixer of claim 1 wherein at least a portion of the surfaces
of said wells have an electroosmotic mobility that is different
from the electroosmotic mobility of at least a portion of said
mixing channel.
5. A splitter of a substantially laminar microfluidic stream
comprising: a splitting channel coupled to at least two inlet ports
and at least one outlet port in which said substantially laminar
microfluidic stream has an axis of flow; and a plurality of
substantially straight unconnected wells disposed in said splitting
channel, said wells being oriented substantially longitudinally
across the width of said channel and diagonally across said axis of
flow said wells being greater in depth than in width.
6. The splitter of claim 5 wherein alternating wells are configured
prependicular to each other.
7. The splitter of claim 5 wherein said wells are configured
parallel to each other.
8. The splitter of claim 5 wherein said microfluidic streams are
propelled by pressure.
9. The splitter of claim 5 wherein said microfluidic streams are
propelled by electroosmosis.
10. The splitter of claim 5 wherein said microfluidic streams are
propelled by electrokinetics.
11. The splitter of claim 5 wherein at least a portion of the
surfaces of said wells have an electroosmotic mobility that is
different from the electroosmotic mobility of at least a portion of
said splitting channel.
12-18. (canceled)
Description
CROSS REFERENCE TO RELATED APPLICATIONS
[0001] This application is based upon and claims priority to U.S.
provisional application No. 60/323,509 entitled "MICROCHANNEL
DESIGNS FOR MIXING AND SPLITTING MICROFLUIDIC STREAMS UNDER
ELECTROKINETIC OR PRESSURE DRIVEN FLOW" filed Sep. 19, 2001 by
Timothy J Johnson, David J Ross, and Laurie E Locascio, the entire
disclosure is specifically incorporated herein by reference for all
that it discloses and teaches.
BACKGROUND OF THE INVENTION
[0002] a. Field of the Invention
[0003] The present invention pertains generally to microfluidic
flow devices and specifically to mixers and splitters of
microfluidic flow.
[0004] b. Description of the Background
[0005] The application of microfluidic analytical devices to
chemical or biological assays has developed rapidly over the last
decade. Although microfluidic devices have been highly successful,
several performance limitations exist, notably reagent mixing.
[0006] Most mixing devices rely on diffusive mixing, wherein the
natural laminar flow effects and the reagent's inherent diffusion
coefficient cause the reagents to mix. Therefore, the mixing
chamber/channel is usually extended to lengths that will ensure a
completely mixed outlet stream. This approach may be acceptable for
low flowrates, but high flowrates (>1 cm/s) or low analyte
diffusion coefficients (>10.sup.-7 cm/s.sup.2) will require
excessively long mixing channels. The difficulty in rapidly mixing
reagents results from the fact that the system is restricted to the
laminar flow regime (Re<2000) and also because the feature sizes
are too small (typically<100 .mu.m) to incorporate conventional
mixing mechanisms.
[0007] The lack of turbulence in microfluidic systems has led to
device designs that utilize multi-laminate, or flow splitting
techniques to accomplish mixing in channels of shorter length.
These designs split the incoming streams into several narrower
confluent streams to reduce the mixing equilibrium time. Once
mixing is complete, the narrow channels are then brought back
together into a larger main channel for further transport,
processing, and/or detection. The effectiveness of the flow
splitting concept is based on the fact that the equilibrium time
scales quadratically with the width of the channel. For example, if
the width of the channel decreases by two, then the equilibrium
time and the channel length decreased by a factor of four, or 25%
of the original length. However, even a mixing length of 25% may
still be unsuitable for some applications.
[0008] Other techniques for mixing may rely on active mechanical
mixing, such as stirring paddles and the like. For very small
fluidic passages, such devices are extremely fragile and difficult
to manufacture.
[0009] It would therefore be advantageous to provide a device and
method of mixing two confluent microfluidic laminar flows that did
not require an excessively long channel to effectively mix the
flows. Further, it would be advantageous to provide a splitting
mechanism that may be able to split a stream of reagents into two
streams of differing concentrations.
SUMMARY OF THE INVENTION
[0010] The present invention overcomes the disadvantages and
limitations of the prior art by providing a device and method for
effectively mixing two confluent laminar reagents within a very
short stream length. This is accomplished by passing the confluent
laminar flows over a series of narrow wells that are angled across
the width of the channel. The device and method may also be used to
split incoming streams into multiple streams of equal or non-equal
proportions. Additionally, the present invention may be used for
the mixing of plugs of reagents while minimizing axial dispersion
of the reagent plug.
[0011] The present invention may therefore comprise a mixer of
laminar microfluidic streams propelled by electrokinetic flow
comprising: a first inlet channel; a second inlet channel; a mixing
channel starting at the confluence of the first inlet channel and
the second inlet channel; and a plurality of wells disposed in the
mixing channel, the wells being obliquely oriented substantially
across the width of the mixing channel.
[0012] The present invention may further comprise a splitter of a
substantially laminar microfluidic stream comprising: a splitting
channel coupled to at least two inlet ports and at least one outlet
port in which the substantially laminar microfluidic stream has an
axis of flow; and a plurality of wells disposed in the splitting
channel, the wells being oriented substantially longitudinally
across the width of the channel and diagonally across the axis of
flow, the wells being deeper in profile than in width.
[0013] The present invention may further comprise a method of
mixing two confluent laminar flows in microchannels comprising:
providing a first inlet stream and a second inlet stream that meet
at a confluence point to produce a confluent stream; passing the
confluent stream through a mixing channel, the mixing channel
comprising a plurality of wells, the wells being oriented
substantially longitudinally across the width of the mixing channel
and diagonally across the mixing channel, the wells being deeper in
profile than in width; and producing a mixed laminar flow at the
output of the mixing channel.
[0014] The advantages of the present invention are that flows may
be combined and mixed without the conventional long lengths of
diffusive mixing. The device may be further adapted to create two
or more streams of equal or non-equal proportions of reagents. The
device may be adjusted to tune the mix of the flows by adding
various wells at different orientations, depths, and with various
electroosmotic mobility coatings, all of which may have a
substantial effect on the performance of the present invention.
BRIEF DESCRIPTION OF THE DRAWINGS
[0015] In the drawings,
[0016] FIG. 1 is an illustration of an embodiment of the present
invention of a microfluidic mixer.
[0017] FIG. 2A is a white light microscopy image of the embodiment
of FIG. 1.
[0018] FIG. 2B is an image of the fluorescence of Rhodamine B
introduced in the first inlet mixed with the buffer solution
introduced in the second inlet.
[0019] FIG. 2C is a similar image as FIG. 2B, except the flow is
0.81 cm/s.
[0020] FIG. 3 illustrates experimental results of the degree of
mixing of the embodiment of FIG. 2A, wherein an electroosmotic flow
of 0.06 cm/s was achieved.
[0021] FIG. 4 illustrates the same experimental set up as FIG. 2C,
with the electroosmotic flowrate of 0.81 cm/s.
[0022] FIG. 5 illustrates the same experimental set up of FIGS. 3
and 4, with a pressure driven flow.
[0023] FIG. 6 illustrates a second preferred embodiment of the
present invention.
[0024] FIG. 7 is a white light microscopy image of the embodiment
of FIG. 6.
[0025] FIG. 8 illustrates experimental results of the degree of
mixing of two reagents using embodiment of FIG. 7 and an embodiment
similar to embodiment of FIG. 7 but with three wells instead of
four.
[0026] FIG. 9 illustrates the results of the same experimental set
up as FIG. 8 with a higher electroosmotic flowrate of 0.81
cm/s.
[0027] FIG. 10 illustrates an embodiment of a stream splitter
wherein two inlet ports form a confluent stream wherein the fluid
on one half of the channel is split into two streams located on
opposite sides of the channel that then exit through two
outlets.
[0028] FIG. 11 illustrates an embodiment of a four-well mixer that
was analyzed with computational fluid dynamics techniques for
variations in the present invention.
[0029] FIG. 12 illustrates some computational analysis of the flow
patterns for various depths of wells, based on the embodiment shown
in FIG. 11.
[0030] FIG. 13 illustrates some computational results of various
angles of the walls, based on the embodiment of FIG. 11.
[0031] FIG. 14A illustrates a plan view of the flow pattern of an
embodiment of the present invention of a mixer with quantity 4
wells oriented at 15 degrees off of the axis of flow.
[0032] FIG. 14B illustrates a cross sectional view of the flow
pattern of FIG. 14A, as observed from the cross section E-E.
[0033] FIG. 15 illustrates some computational results of changes in
the electroosmotic (EO) mobility of the surfaces of the wells.
[0034] FIG. 16A illustrates a plug of fluid introduced to and
transported by a channel.
[0035] FIG. 16B illustrates an embodiment of the present invention
wherein a plug of fluid is introduced into a channel in which four
wells are disposed.
[0036] FIG. 17 illustrates some results of a computational analysis
of the flow of the embodiments of FIGS. 16A and 16B.
[0037] FIG. 18 illustrates the laser apparatus used to manufacture
the experimental devices referenced in the specification.
DETAILED DESCRIPTION OF THE INVENTION
[0038] FIG. 1 illustrates an embodiment 100 of the present
invention of a microfluidic mixer. Two inlet streams 102 and 104
are combined and mixed in the mixing region 106 to produce a mixed
flow that exhausts out of the outlet 108. The mixing region 106
comprises several wells 110, 112, 114, 116, 118, 120, 122, 124,
126, and 128 that are recessed into the outlet 108.
[0039] In a first embodiment, the channels have a uniform,
trapezoidal cross section, with the width at the top being 72
.mu.m, depth 31 .mu.m, and the width at the bottom being 28 .mu.m.
The wells 110, 112, 114, 116, 118, 120, 122, 124, 126, and 128 have
a depth of 85 .mu.m in the center of the well.
[0040] The wells 110, 112, 114, and 116 are parallel to each other
and uniform in size. The wells are angled approximately 45 degrees
from the axis of the outlet 108. The wells 118, 120, 122, 124, 126,
and 128 are perpendicular to each other and are approximately 45
degrees from the axis of the outlet 108.
[0041] The flow of reagents through the embodiment 100 may be
electrokinetic, electroosmotic, or pressure driven flow. Since the
electroosmotic flow is a wall driven phenomenon based on the
surface charge of the microchannel wall and on the local electric
field, the fluid may enter and follow the contour of the wells,
with the slanted well design used to induce lateral transport
across the channel. In a pressure driven flow, the flow is not a
wall driven phenomenon and therefore the fluid is not forced to
enter the wells like electroosmotic flow. However, the presence of
the wells induces some lateral transport across the channel in
pressure driven flow, although not as effective as with
electroosmotic flow. The presence of the wells under electrokinetic
flow, being the combination of both electroosmotic flow and
electrophoretic flow, will also enhance the mixing.
[0042] In an experiment with the present embodiment, a confluence
of Rhodamine B in a carbonate buffer was introduced into inlet 102
and carbonate buffer was introduced into inlet 104. The
fluorescence of the Rhodamine B was measured to indicate the degree
of mixing achieved by the embodiment 100.
[0043] The method of manufacture of the embodiment used in the
experiment, as well as the experimental setup and method of
measurements, are given elsewhere in this specification.
[0044] For the experiments, the length of each channel arm was 0.8
cm long, and the dimensions of the channel are 72 .mu.m wide at the
top, 28 .mu.m wide at the bottom, and 31 .mu.m deep. The laser
etched wells that spanned across the entire width of the channel
had a depth of 85 .mu.m in the center of the well relative to the
bottom of the imprinted channel. The intensity measurement was
taken at distance 130 or 443 .mu.m from the beginning of the
confluent region.
[0045] FIGS. 2A-C illustrate the results of the experiments as
detailed elsewhere in this specification. FIG. 2A is a white light
microscopy image of the embodiment of FIG. 1. FIG. 2B is an image
of the fluorescence of Rhodamine B introduced in the first inlet
202 mixed with the buffer solution introduced in the second inlet
204, producing a mixed stream 206. The flowrate is 0.06 cm/s. FIG.
2C is a similar image as FIG. 2B, except the flow is 0.81 cm/s.
[0046] FIG. 3 illustrates the degree of mixing of the embodiment
100, wherein an electroosmotic flow of 0.06 cm/s was achieved. The
horizontal axis is the position across the width of the outlet 108
and the vertical axis is the normalized intensity of the
fluorescence of the Rhodamine B. The curve 302 represents the
results of the measurement taken with no mixing wells present. The
curve 304 represents perfect mixing. Curve 304 is trapezoidal in
shape, following the profile of the trapezoidal outlet 108. The
curve 306 represents the actual experimental results. The details
concerning the experimental procedure and equipment used to perform
all of the experiments referenced in this specification are given
elsewhere in this specification.
[0047] FIG. 4 is a graph that illustrates the same experimental set
up as FIG. 3, with the electroosmotic flowrate of 0.81 cm/s. The
line 402 represents the results of the measurement taken with no
mixing wells present. The line 404 represents perfect mixing. The
line 406 represents the actual experimental results.
[0048] From the results illustrated in FIGS. 2B, 2C, 3, and 4, the
degree of mixing exiting the mixer was 87.2% and 80.5%
respectively, for the flowrates of 0.06 cm/s and 0.81 cm/s. To
achieve the same degree of mixing, theoretical predictions state
that a channel length of 0.2 cm and 2.3 cm for electroosmotic
flowrates would be required if no mixer were present and based on
diffusional mixing, assuming that the diffusion coefficient of the
fluorescent material, Rhodamine B, is 2.8.times.10.sup.-6
cm.sup.2/s. These results indicate that the length of the present
embodiment is 22% and 2% of the length of a comparable diffusive
mixer for the present application.
[0049] FIG. 5 is a graph that illustrates the results of the same
experimental set up of FIGS. 2, 3, and 4, with a pressure driven
flow. The measurements were taken at 443 .mu.m from the beginning
of the confluence region. The curve 502 represents the perfect
mixing results. The curve 504 represents the experimental results
with a pressure driven flow at 0.21 cm/s. The curve 506 represents
the experimental results with a pressure driven flow at 1.25 cm/s.
Since the pressure driven flow is not a wall driven phenomenon and
therefore the fluid is not forced to enter the wells like
electroosmotic flow, the effects of the mixing are not as great
with pressure driven flow as with electroosmotic flow. However, the
presence of the wells does introduce some lateral transport across
the channel. This suggests that a series of wells could be
optimized for mixing under pressure driven flow.
[0050] FIG. 6 illustrates a second embodiment 600 of the present
invention. Two inlet streams 602 and 604 are combined and mixed in
the mixing region 606 to produce a mixed flow that exhausts out of
the outlet 608. The mixing region 606 comprises several wells 610,
612, 614, and 616 that are recessed into outlet 608. The
measurement region 618 is 183 .mu.m from the point of confluence of
the inlets 602 and 604. The shape and dimensions of the channels
and wells are the same as with embodiment 100 of FIG. 1. The wells
610, 612, 614, and 616 are parallel to each other.
[0051] FIG. 7 is a white light microscopy image of an example of
embodiment 600.
[0052] FIG. 8 is a graph that illustrates the degree of mixing of
two reagents using embodiment 600 and an embodiment similar to
embodiment 600 but with three wells instead of four. The
experimental results of FIG. 8 show results for electroosmotic flow
of 0.06 cm/s taken 183 .mu.m from the point of confluence of the
inlet flows. The horizontal axis is the position across the width
of the outlet stream 608 and the vertical axis is the normalized
intensity of Rhodamine B. The same experimental setup was used for
the results of FIG. 3 as FIG. 8, with the differences being the
configuration of the mixing region and the position of the
measurements.
[0053] The curve 802 represents the mixing profile of the two inlet
streams when no mixing wells are present. The curve 804 represents
the perfect mixing of the two streams. The curve 806 represents the
mixing profile for the electroosmotic flowrate of 0.06 cm/s and a
three well mixer. The curve 808 represents the mixing profile for
the same flowrate and a four well mixer. The three well mixer is
the embodiment 600 with well 616 removed.
[0054] FIG. 9 illustrates the results of the same experimental set
up as FIG. 8 with a higher electroosmotic flowrate of 0.81 cm/s.
The curve 902 represents the mixing profile of the two inlet
streams when no mixing wells are present. The curve 904 represents
the perfect mixing of the two streams. The curve 906 represents the
mixing profile for a three well mixer. The curve 908 represents the
mixing profile for a four well mixer. The three well mixer is the
embodiment 600 with well 616 removed.
[0055] The curve 908 forms two distinct humps, 910 and 912,
indicating that the four well mixer may be able to split the
incoming streams into two streams of similar concentrations. The
number of wells, the shape, dimension, and placement of the wells
may be adapted to provide different dilutions of the incoming
fluid. Such adaptations may depend on the reagents and the
diffusivity constants of the various components of the confluent
streams. As such, the particular result desired, such as splitting
a stream or mixing a pair of confluent streams may be obtained by
adjusting the quantity and position of the various wells.
[0056] FIG. 10 illustrates an embodiment 1000 of a stream splitter
wherein inlet port 1002 and inlet port 1003 form a confluent stream
wherein the fluid that is on one half of the channel is split into
two streams due to the presence of the slanted wells, where the
split streams are located on opposite sides of the channel that
then exit through outlet 1004 and outlet 1006. The flow of the
microfluidic stream passes over three wells 1008, 1010, and 1012 of
similar design and construction as those of other embodiments
described in the present specification.
[0057] In lab-on-a-chip or .mu.-TAS (micro Total Analysis Systems),
the use of a series of wells within a microchannel may greatly
enhance the effectiveness of the entire system, especially when the
system is limited to the laminar flow regime. The present invention
is effective for low flowrates (<1 cm/s) as well as high
flowrates (>1 cm/s). The present invention is further able to
effectively mix flows that are driven electrokinetically,
electroosmotically, or by pressure.
[0058] The present invention may be used to divide or split a
stream into non-equal or equal analyte concentrations. Such an
application may be useful in lab-on-a-chip or .mu.-TAS systems
wherein streams of various concentrations are desired. Several
embodiments of the present invention may be used in series, such
that one stream is separated and split, then separated and split
again, with the end result being several outlet streams with
differing dilutions of the original incoming stream. Such a system
is known as serial dilution.
[0059] FIG. 11 illustrates an embodiment 1100 of a four-well mixer
that was analyzed with computational fluid dynamics techniques for
variations in the present invention. The computational analyses
were designed to correspond to the experimental results shown in
the previous figures. The channel geometry and fluid properties
were selected to closely match those of the experiments. The inlet
1102 contains a buffer fluid with Rhodamine B that is mixed with a
second inlet 1104 that contains only the buffer fluid. The fluid
exits the embodiment 1100 via the outlet 1105. The four wells 1106,
1108, 1110, and 1112 are located at an angle theta 1113 from the
centerline axis of the mixer. The geometry of the embodiment 1100
is similar to the previous embodiments described herein.
Cross-section line A 1114 will be used to illustrate the incoming
streams prior to mixing. Cross-section line B 1116 will be used to
illustrate the mixing of the streams while in the well 1112, the
last of the four wells. Cross-section line C 1118 will be used to
illustrate the mixing of the streams 5 .mu.m past the exit of the
last well. Cross-section line D 1120 will be used to illustrate the
mixing of the streams at a location of 420 .mu.m past the point of
confluence.
[0060] FIG. 12 illustrates some computational analyses of the flow
patterns for various depths of wells, based on the embodiment 1100
shown in FIG. 11, with a constant well angle of 45.degree.. The
results for cross section A 1202 illustrate the two incoming flows
1204 and 1206 prior to mixing. The results for cross section B 1208
illustrate the flow patterns for the flow within the last of the
four wells. The 10 .mu.m depth results 1210 show that very little
of the mixing is occurring in the well. The 50 .mu.m depth results
1212 show that a substantial portion of the mixing is occurring in
the well. The 85 .mu.m depth results 1214 show that a substantial
portion of the mixing is occurring in the well, but that there is
not much increase in the mixing due to the larger depth over the 50
.mu.m results 1212. These results indicate that there is a finite
depth wherein increasing the depth does not increase the degree of
mixing substantially. Further, these results illustrate that the
wells greatly affect the mixing by forcing the fluids to fold over
each other.
[0061] FIG. 13 illustrates the results of various angles of the
wells as represented by the angle theta 1113 of FIG. 11. For all
well angles illustrated, the depth of the wells was held constant
at 50 .mu.m below the bottom of the imprinted channel. The results
of cross section B-B 1302, cross section C-C 1304, and cross
section D-D 1306 are shown in columns. The results for the various
angled wells are shown in rows. Results along row 1308 are for
wells at a right angle or 90 degrees to the axis of flow. Results
along rows 1310, 1312, and 1314 are for wells at 60 degrees, 30
degrees, and 15 degrees to the axis of flow. The results indicate
that a decreased angle of the well achieves a higher degree of
mixing.
[0062] The results along row 1308 for right angle wells show that
there is no lateral transport across the width of the well. As the
angle of the wells is decreased, there is increased lateral
transport to the point where the flow may be folded over on top of
itself more than once. The folding action is an important mechanism
that causes efficient mixing.
[0063] FIG. 14A illustrates a plan view of the flow pattern of an
embodiment 1400 of a mixer with quantity 4 wells oriented at 15
degrees off of the axis of flow, and well depths set to 50 .mu.m
below the bottom of the imprinted channel. FIG. 14B illustrates a
cross sectional view of the flow pattern of FIG. 14A, as observed
from the cross section E-E. The flow lines 1402 and 1404 illustrate
that the fluid may exit the first well 1406 and reenter another
well 1408 and thereby may fold during the passage through the mixer
1400.
[0064] FIG. 15 illustrates the results of changes in the
electroosmotic (EO) mobility of the surfaces of the wells.
Different manufacturing processes may create different EO
mobilities on various surfaces of the channels. For example, of the
manufacturing processes described for the experiments described
elsewhere in this specification, imprinting a channel has been
shown to yield a different EO mobility than the laser ablation
manufacturing method. Further, other methods such as
polyelectrolyte multilayers, surface chemistry modifications, EO
mobility suppression coatings, and other methods may be used
individually or in combination to selectively change the EO
mobility of selective surfaces of the mixer.
[0065] The results of FIG. 15 illustrate the effects of increasing
the EO mobility of the surfaces of the wells with respect to the EO
mobility of the remaining surfaces of the mixer. The results are
for a four well mixer with 45 degree wells at a depth of 50 .mu.m
below the bottom of the imprinted channel. The column 1502
illustrates the results for section B-B, column 1504 illustrates
the results for section C-C, and column 1506 illustrates the
results for section D-D, all of which relate the cross sections
illustrated in FIG. 11.
[0066] For the purposes of this discussion, a ratio of the EO
mobility of the wells divided by the EO mobility of the remainder
of the surfaces will be r.sub.EOM. The row 1508 illustrates the
results when r.sub.EOM is 1.24. Row 1510 illustrates the results
for r.sub.EOM of 2.00 and row 1512 illustrates the results for
r.sub.EOM of 3.00. Row 1508 is illustrative of the approximate
r.sub.EOM of the experimental results described in FIGS. 7, 8, and
9. The results indicate that as the r.sub.EOM is increased, mixing
can be enhanced. In other words, the increase of the EO mobility,
by different manufacturing processes, selectively applied coatings,
or other methods may dramatically increase the performance of a
mixer of the present invention.
[0067] A use for the present invention is the mixing of plugs of
fluid. Applications for such a use may be for lab on chip
applications wherein several samples of fluid may be analyzed in
succession. It would be desirable for the plugs of fluid to be
efficiently mixed, but to minimize the axial dispersion of the
plug.
[0068] FIG. 16A illustrates a plug of fluid 1602 introduced into
the channel 1604. The channel 1604 illustrates a case wherein the
wells of the present invention are not present and represents a
baseline case. The mixed plug 1606 is shown downstream.
[0069] FIG. 16B illustrates an embodiment of the present invention
wherein a plug of fluid 1608 is introduced into a channel 1610 in
which four wells 1612, 1614, 1616, and 1618 are disposed. The mixed
plug 1620 is shown downstream. For FIG. 16B, r.sub.EOM is set to
2.00.
[0070] FIG. 17 illustrates the results of a computational analysis
of the flow of the embodiments of FIGS. 16A and 16B. The curves
1702 and 1704 illustrate the average concentration of the plug as
it passes the outlet of the mixing channel over time. Curve 1702
represents the plug of fluid from FIG. 16A and curve 1704
represents the plug of fluid from FIG. 16B, with r.sub.EOM equal to
2.00.
[0071] The cross section 1706 represents the analysis results for
the point 1708 and cross section 1710 represents the analysis for
the point 1712. Both cross sections are the approximate high point
of the concentration. For cross section 1706, the plug flow with no
wells, the average concentration of the reagent is approximately
28% higher than the cross section 1710. However, the standard
deviation, an approximate measure of the degree of mixing, is
approximately 3.6 times higher for cross section 1706 wherein no
wells were present. The lower standard deviation of the mixture
that passed through the inventive wells indicates that the plug of
reagent was very well mixed. Further, from the curve 1712, the plug
of fluid is still intact, although slightly elongated when compared
to the reagent that was not passed over the inventive wells.
[0072] The present invention is a passive device that greatly
enhances the mixing of reagents under electrokinetic flow, and to a
lesser degree, under pressure driven flow. The present invention
significantly decreases the channel length required for mixing
reagents by placing wells in the flow channel at oblique angles to
the axis of flow. The wells may be of various depths, however, for
a given set of reagents, flowrates, and channel geometries, there
may be an optimum depth of a well wherein an increased depth may
not increase the mixing effectiveness.
[0073] Electroosmotic flow is a surface driven mechanism that may
be enhanced to change the performance of the present invention. For
example, increasing the electroosmotic mobility of selective
surfaces such as the wells has been shown to increase the
effectiveness of the mixer.
[0074] The manufacturing process and equipment used in the
experiments referenced in this specification are herein
defined.
[0075] Reagents and Materials. Laser Grade Rhodamine B was used as
supplied by Acros Organics (Belgium) and dissolved in 20 mM, pH 9.4
carbonate buffer to a final concentration of 0.11 mM Rhodamine B.
The buffer solution was made using deionized water from a Millipore
Milli-Q system (Bedford, Mass.), and was filtered before use with a
syringe filter (pore size 0.22 .mu.m).
[0076] Microchannels were made using polycarbonate sheet (PC;
Lexan, GE Co., Mt. Vernon, Ind.). Poly(ethylene terephthalate
glycol) (PETG; Vivak, DMS Engineering Plastic Products, Sheffield,
Mass.) was used to cover and seal the microchannel substrate. The
glass transition temperature of PC and PETG are approximately
150.degree. C. and 81.degree. C., respectively. Polycarbonate was
chosen as the substrate material because it has a high absorption
cross section to 248 nm light (the wavelength of the excimer
laser), therefore ablated structures have minimal surface roughness
(<5 nm). PETG was chosen to seal the microchannels because its
glass transition temperature is well below that of PC. Therefore,
thermal sealing can be performed at a temperature that does not
cause distortion of the PC microchannel.
[0077] Hot Imprinting Method. Prior to imprinting, the PC substrate
was blown clean with ionized air. Channels were hot imprinted in
the substrate material using a silicon stamp with a
trapezoidal-shaped raised T-channel. The PC was place over the
silicon stamp, the two items were then placed between two aluminum
heating blocks, and then the temperature was raised to 155.degree.
C. Next, the assembly was placed in a hydraulic press and a
pressure of 13.8 MPa (2000 psi) was applied for 1.5 hours. The
imprinted substrate was then removed from the template and allowed
to cool to room temperature. Channel dimensions were measured by
optical profilometry.
[0078] Laser Ablation Method. A 248 nm excimer laser system
(LMT-4000, Potomac Photonics, Inc., Lanham, Mad.) was used to
ablate microstructures within the pre-formed PC microchannel. The
excimer laser system, FIG. 18, contains a laser light source 1, a
round aperture (200 .mu.m diameter) 2 for delimiting the size and
shape of the beam 3, a focusing lens (10.times. compound) 4, a
visible light source 5, a CCD camera to image the ablation process
6, and a controllable X-Y stage 7 with a vacuum chuck 8 to hold the
substrate 9 in place. Also, a nozzle 10 was present to sweep
nitrogen over the substrate 9 during processing, and a vacuum
nozzle 11 was located on the opposite side of the stage to remove
debris. For the experiments conducted here, the size-delimiting
aperture was chosen such that the ablated features would be smaller
than the dimensions of the channel. Also, the X-Y stage was moved
linearly at a rate of 1 mm/s, and the ablated wells were at a
45.degree. angle relative to the axis of the main channel. The
average power level per pulse was set to 2.04 .mu.J+/-0.14 .mu.J.
The frequency of pulses was set to 200 Hz, with a constant pulse
width of 7 ns. The light after being focused exposed a circular
area of 1.90.times.10.sup.-6 cm.sup.2.
[0079] Measuring Well Depth and Profile. The depth of the ablated
wells was measured by cutting the substrate with a microtome
(Microm HM335 E, Walldorf, Germany) either perpendicular to the
axis of the outlet channel or parallel to the slanted wells. The
substrate was cut so that the edge of the substrate was within a
few microns of the wells. The wells were then imaged and measured
using white light microscopy.
[0080] Microchannel Sealing Procedure. The pre-formed microchannels
were covered and thermally sealed with a flat piece of PETG
(referred to as the `lid` throughout the rest of the text) of
similar dimensions to the PC. Prior to bonding, the lid and the
channel were cleaned with compressed nitrogen gas. The lid was then
placed on top of the channel, and the two pieces were clamped
together between microscope glass slides and bonded by heating in a
circulating air over at 90.0.degree. C.+/-0.5.degree. C. for 13
minutes. It is important to keep the time and temperature as low as
possible in the sealing process to avoid physical alteration of the
microchannel.
[0081] For the electroosmotic flow studies, 3 mm diameter circular
holes in the lid provided access to the channels and served as
fluid reservoirs. For the pressure driven flow studies, 0.8 mm
diameter circular holes in the lid, located at the ends of each
inlet channel, provided access to insert a section of hollow
stainless-steel tubing. A 3 mm diameter hole in the lid at the end
of the outlet channel served as a waste reservoir. For each
experiment, the channel arms were fixed to a length of 8 mm.
[0082] Flow Image Acquisition. Flourescence imaging of the
rhodamine dye was performed using a research fluorescence
microscope equipped with a 10.times. objective, a mercury arc lamp,
a rhodamine filer set, and a video camera (COHU, San Diego,
Calif.). Digital images were acquired using Scion Image.TM.
software and a Scion LG-3 frame grabber (Scion, Inc., Frederick,
Md.). For each experiment, images were captured every {fraction
(1/60)}.sup.th of a second over a duration of 0.67 s, averaged,
then recorded.
[0083] Experimental Set-up. To image the mixing under
electroosmotic flow, the microchannels were initially filled with
the carbonate buffer solution. Then, an equal amount (typically 40
.mu.L) of buffer was placed in one inlet channel reservoir and in
the outlet channel reservoir, while the second inlet reservoir was
filled with the rhodamine-labeled buffer. Platinum electrodes were
then placed in contact with the solution in the reservoirs such
that the two inlet reservoirs were fixed to ground and the
potential was applied to the outlet channel reservoir. The
microchannel was placed beneath the fluorescence microscope
described in the previous section, and images were acquired at
several different applied voltages (0 to -1750V), beginning with
zero applied voltage to verify that there was minimal flow
resulting from hydrostatic pressure. The current through the
microchannel was determined by measuring the voltage drop across a
100 k.OMEGA. resistor (typically less than {fraction (1/1000)} the
resistance of the microchannel) connected to the high voltage
supply in series with the microchannel. For pressure driven flow
studies, a programmable syringe pump (Harvard Apparatus PHD 2000,
Holliston, Mass.) was interfaced to the stainless tubing in the
inlet reservoirs via Teflon tubing.
[0084] The foregoing description of the invention has been
presented for purposes of illustration and description. It is not
intended to be exhaustive or to limit the invention to the precise
form disclosed, and other modifications and variations may be
possible in light of the above teachings. The embodiment was chosen
and described in order to best explain the principles of the
invention and its practical application to thereby enable others
skilled in the art to best utilize the invention in various
embodiments and various modifications as are suited to the
particular use contemplated. It is intended that the appended
claims be construed to include other alternative embodiments of the
invention except insofar as limited by the prior art.
* * * * *