Note: Descriptions are shown in the official language in which they were submitted.
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
1
DEVICES AND METHODS FOR INTERFACING MICROFLUIDIC DEVICES WITH
FLUID HANDLING DEVICES
CROSS REFERENCE TO RELATED APPLICATIONS
This application claims priority from U.S. Provisional Patent Application
Serial No.
60/598,598 filed on August 4, 2004, the contents of which are incorporated
herein by reference
in their entirety.
FIELD OF THE INVENTION
The present invention relates to the field of microfluidic circuits for
chemical,
biological, and biochemical processes or reactions. More specifically, it
discloses devices and
methods for interfacing microfluidic devices with fluid handling devices.
BACKGROUND OF THE INVENTION
In recent years, the pharmaceutical, biotechnology, chemical and related
industries
have increasingly adopted devices containing micro-chambers and channel
structures for
performing various reactions and analyses. These devices, cominonly referred
to as
microfluidic devices, allow a reduction in volume of the reagents and sample
required to
perform an assay. They also enable a large number of reactions without human
intervention,
either in parallel or in serially, in a very predictable and reproducible way.
Microfluidic
devices are therefore promising devices to realize a Micro Total Analysis
System (micro-
TAS), definition that characterizes miniaturized devices that have the
functionality of a
conventional laboratory.
In general, all attempts at micro-TAS devices can be characterized in two
ways:
according to the forces responsible for the fluid transport and according to
the mechanism used
to direct the flow of fluids. The former are referred to as motors. The latter
are referred to as
valves, and constitute logic or analogue actuators, essential for a number of
basic operations
such as volumetric quantitation of fluids, mixing of fluids, connecting a set
of fluid inlets to a
set of fluid outputs, sealing containers (to gas or to liquids passage
according to the
application) in a sufficiently tight manner to allow fluid storage, regulating
the fluid flow
speed. A combination of valves and motors on a microfluidic network,
complemented by input
means to load the devices, and readout means to measure the outcome of the
analysis, make a
micro-TAS possible and useful. With increasing performances and
miniaturization of the
devices, the need for a reliable
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
2
and adaptable interface to the macroscopic world becomes a requirement to
allow users to
exploit the functionality of these systems, both for research and commercial
applications.
It is evident that most reagents today are stocked in formats not specifically
designed
for microfluidics, and these formats are heterogeneous: for example, vials and
tubes in the
diagnostics area and in the academic world, micro-plates in the drugs
discovery industry. The
existence of standards (for example, the Society of Biomolecular screening has
defined an
open standard for micro-plates) has stimulated many years of commercialization
of a large
number of fluid handling tools specifically designed for the common
standardized formats.
The availability of a large installed base of instruments makes the
introduction of products not
compliant to the fluid storage standards difficult, for reasons related to
laboratory space
availability, maintenance, costs and user habits.
Fluid handling devices, also called fluid handlers, dispensing devices,
sainple loading
robots, compound dispensers, dispensing means, pipettors, pipette
workstations, have the
purpose of transferring fluids, and in particular liquids, from fluid storage
to further fluid
storage. The components that take part in a typical fluid handling process can
therefore be
classified into three categories, according to their role in the process: (i)
the source of the
original fluid storage, (ii) the means by which the fluid is transferred, and
(iii) the container in
the fluid storage where the fluid is moved to.
In general terms, an automated dispensing device is not always strictly
needed, since
the dispensing operation could be performed by a human operator equipped with
specific tools,
like pipettors or similar devices. However, all dispensing devices can be
described according to
their overall characteristics, like for example operational speed,
performance, cost,
contamination issues and versatility. The desired requirements of fluid
handling devices are
the highest speed possible (to achieve high productivity, but also to allow to
perform assays in
similar conditions like temperature, reagents activity, etc.), minimal
contamination between
sources and containers, minimal fixed cost and minimal cost per dispensing
operation
(consumables), performances (precision of dosing, range of volumes that can be
dispensed,
footprint, etc.) and versatility (multi-format compatibility, type of
operations performed,
automatic identification of source and container, etc.).
All existing fluid handling devices address or partially solve these
requirements, and
the user choice depends on the specific application and on the laboratory
environment. Being
the environments heterogeneous, the dispensing instruments - exactly as it is
for the fluid
storage means - differ significantly and adopt different technologies:
disposable tips and
suction means, metallic pins immerged in the fluids, aspirating needles and
subsequent rinsing
and cleaning
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
3
operations, pumps and tubing, ejection of droplets by piezoelectric or other
mechanical means.
Also the infrastructure surrounding the dispensing technology and its degree
of automation
differ enormously, going from complex installations for compound libraries
management in the
pharmaceutical industry, to simple hand-held devices.
Microfluidic devices deal with volumes that are typically negligible in the
standard
assay environment, so they usually take part to the process in the form of the
dispensing
devices or in the form of containers; in fact it is improbable to move
microscopic volumes of
fluids into macroscopic containers, since detection methods used in the
subsequent step of an
assay could miss sensitivity, or because the reaction would simply require
larger amounts of
samples. An example of microfluidic dispensing device is a piezoelectric
nozzle. An example
of microfluidic container is a microarray for genetic analysis. It should be
noted, however, that
"inicroscale-to-microscale device" fluid transfers will become very important
as soon as a
larger number of assays will be performed in microfluidic formats; in that
case, microfluidic
devices will take part to the process also as sources.
Centripetal devices are a specific class of microfluidic devices, where the
micro-fluidic
devices are spun around a rotation axis in such a way that the centripetal
acceleration generates
an apparent centrifugal force on the microfluidic device itself, and on any
fluid contained
within the microfluidic device. The centrifugal force acts as a motor, in the
radial but also in
the tangential direction if the angular momentum varies. This force, however,
is applied at the
same time to any material contained in the microfluidic device, including the
fluids that are
contained in the inlets. In most centripetal microfluidic devices, like for
example those
developed by Gyros AB, Tecan AG, Burstein Technologies Inc. for example, micro-
fluidic
devices have the shape of disks, and the rotation axis is perpendicular to the
main faces and
passing through the centre of the disk. The centrifugal force, therefore, is
also parallel to the
surface of the disk: it is evident that non-sealed inputs manufactured on the
surface require a
very specific shape in order to prevent overspill of the fluid out of the
inlet aperture.
A possible geometrical shape for an inlet in these devices is a cone with its
apex cut off
by a plane parallel to its base, also known as frustum, where the inlet
aperture is located on the
top of the truncated cone. When the centrifugal force on the fluid
contained.in the inlet exceeds
the gravity and the surface tension forces, the only usable volume of the
input reservoir is the
fraction of volume characterized by radii which are larger than the largest
radius of the inlet
aperture. This clearly limits the capacity of the inlet to a fraction of the
actual reservoir
volume, and cannot prevent undesired overspill if the fluid is dispensed in
excess to this
fraction (for example,
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
4
because of the limited dispensing accuracy of a fluid dispensing system when
dealing with
small volumes).
In addition, it should be known that various technologies, for example
injection
moulding, put constraints in the geometrical shape of the inlet. In injection
moulding the
replicated device has to be extracted from the mould that determined its
shape, and this
operation becomes impossible if the previously mentioned inlet with a
truncated cone shape is
attached to the mould structure by the top. It should be also noted that for
voluines typical of
microfluidic devices the surface tension value characteristic of most fluids
prevents them to
flow out of the device when the inputs are not vertical, so that the
microfluidic device can be
kept at rest - with any orientation in space (and, for example, horizontally).
This phenomenon
reinains valid when the microfluidic device is subject to small acceleration,
or for those
accelerations having appropriate direction.
The challenge of interfacing microfluidic devices in first instance is the
problem of
loading fluids from a conventional source (e.g. vial, micro-titre plate or an
Eppendorf tube)
into a microfluidic device. This interfacing challenge has been typically
addressed in the past
by the engineering of specific, proprietary dispensing devices customized to a
given
microfluidic device, or the design of a suitable "macroscale-to-microscale
interface". This
interface allows the efficient use of existing infrastructures and fluid
loading facilities by
extending their applicability into the micro-scale world. While this approach
has the advantage
of reducing switching costs by using existing infrastructure, it often liinits
the advantages
consequent to the miniaturization effort (e.g. small reagent consumption,
density of data-points
for a given substrate, etc.).
However, when a macroscopic interface is implemented onto a miniaturized
device it is
common that a large active area is sacrificially dedicated to the inputs and
to the spacing in
between. This input area, being implemented on a device manufactured with
advanced high-
resolution replication technologies, has a significant production cost and
reduces the active
space on a fixed micro-structured master size (typically a disk with 4, 6 or 8
inches diameter).
Unfortunately, there is a significant manufacturing cost increase due to the
presence of inputs
organized according the mentioned interface. In addition, a large disk
diameter should
preferably remain inside the standard micro-plate footprint, to avoid the
problem of disk
manipulation in conventional micro-titre plate handlers or the requirement of
substantial
modifications to the software or to the hardware of existing handling systems.
The same
limitation on the maximum disk diameter is evident when the micro-device has
to be used
inside instruments designed for the micro-plate formats, like for example
fluorescence and
absorbance readers, incubators, imaging devices, centrifuges, shakers, barcode
labellers, etc.
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
An additional limitation of current approaches is that a majority of
microfluidic devices
are designed and manufactured according to a two-dimensional process that
generates pseudo
three-dimensional structures. The two dimensional network is transformed into
a three-
5 dimensional micro-structured layer by means of etching, or sometimes
extrusion, of a substrate
at a depth identical for all components (or for a fraction of them) contained
in the network.
Because of this, most microfluidic networks are substantially planar or made
by multi-layers
with a planar conformation.
These characteristics are typical of lithographic processes, which are among
the most
common manufacturing techniques. Lithography requires masks, and each mask
typically
corresponds to a given etching depth on a planar substrate. Many other
manufacturing
processes have similar constraints: for example, laser ablation of a substrate
has a limited
etching depth and the microfluidic network is typically created onto a planar
substrate. Also
devices obtained by lamination, where different sheets are cut and laminated
together, are
essentially bi-dimensional. The same is valid for hot embossing wllere the
microstructures are
obtained by embossing a planar substrate onto a press and to the injection
moulding technique.
Injection moulding is probably the most important mass production technology:
a master is
etched - being in silicon, glass, SU8, peek or other material - and possibly
replicated by
electroplating into a metallic mould insert. The micro-structured insert is
positioned in a cavity
that gives shape to the high temperature polymer injected in the mould, and
since the insert
comes essentially from a lithographic procedure (or a substantially planar
technology) the
microstructures replicated in this way are also distributed on a plane.
A coinmon problem in the production of microfluidic devices consists in the
fact that
inputs typically required to load the fluids in the devices have to be
manufactured with a
method different from the one used in the micro-structuring operation. This
problem comes
from the requirement that micro-fluidic devices have to be loaded from the
outside; therefore
inputs have to reach the external surface of the device. Input manufacturing
often requires
post-processing or a specific manufacturing technology. Examples of these
processes are laser
drilling of the substrate body, mechanical drilling, needle penetration
through soft substrates
and assembly of cover structures containing ports on top of the substrate
containing the
microstructures. Any additional procedure in the manufacturing process,
however, is
undesired since it implies significant manufacturing issues like cost
increase, yield reduction,
production rate decrease, dust contamination failures, relative alignment
problems and process
quality control.
The injection moulding process, in particular, is a common method of
fabricating
plastic devices. As it is known in the art, media storage devices can be
produced cheaply
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
6
because of mass production scale considerations, but also because they have no
passing-
through connections, and all fine resolution structures, the pits where data
is encoded, can be
replicated in a single step of microlithography. As soon as passing-through
structures are
required, the moulds for manufacturing become more coinplicated and the mould
cycle time
becomes longer thereby increasing production cost. For example, passing-
through connections
could require the addition of other mould inserts that should match and
comlect exactly to the
insert replicating the microstructures on the device. A fluidic coiuiection
between parts of a
device formed by two different inserts implies a very critical matching of the
position of the
inserts, and also any possible gap in the connection between the two inserts
will be filled by the
fluid polymer at injection, a phenomenon that can potentially interrupt the
fluidic connection in
the replicated piece.
As it is the case for injection moulding, other production technologies are
challenged
by the requirement of manufacturing effectively and reliably input ports for
microfluidic
devices. As a last example, simple mechanical drilling of input ports is also
critical because of
the creation of dust and polymer residues, which could possibly fill the
capillary entrance and
therefore prevent the f-uture passage of fluids.
The planar structure of the microfluidic network de-facto influences and
determines the
overall geometry of the body structure of a microfluidic device. Being all
microstructures are
on a two-dimensional plane, most substrates are substantially planar
polyhedrons,
characterized by having two faces with a large surface area (substantially
larger than the other
faces) and both faces are substantially parallel to the plane where the
microstructures are
located. These faces are to be the "main faces" of the polyhedron, and all the
remaining faces
are called "small faces."
It is understood that all geometries where the faces, in particular the small
faces, are not
planar can be reconnected to this concept, for example by finite eleinents
segmentation. As an
example, the lateral surface of a disk (a cylinder with a small height)
constitutes a non-planar
surface, but the same surface could be represented by a large number of small
faces with
rectangular shape and therefore it is here considered as the small face of the
disk. In addition,
also extensions of the small faces extruding out of the space confined between
the planes
defined by the main faces, are here considered small faces (or part of small
faces) in all
respects.
Following these considerations, it is apparent that most microfluidic devices
have a
substantially planar structure, meaning with "planar" that the microstructures
are positioned on
one or a plurality of surfaces in space. Hereafter, the microfluidic devices
with a substantially
planar structure are also referred to as "tiles".
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
7
Various attempts to address the problems above are exemplified by patents such
as U.S
Patent No. 6.251,343 by Caliper Life Sciences, Inc., which discloses an
interface technology
where the inputs of the microfluidic circuit are created by means of an
additional cover,
mounted on top of the device, comprising a plurality of apertures. The cover
plate is mated to
the ports of the body structure which is in fluidic cominunication with the
microfluidic
network, and the apertures allow dispensing of fluids and application of
electrical connections
witli the fluids contained herein.
This solution relies on bonding quality of the body structure with the cover,
and has the
advantage that the cover manufacturing does not require the same replication
quality required
in the manufacturing of a microfluidic device - therefore it has a lower cost
(but at the expense
of an additional production step). Moreover, this solution is designed for
electrophoresis
where the input ports are loaded with significant ainount of fluids, in order
to guarantee the
filling of capillaries and to allow electrodes to come into electrical contact
with the fluid in the
capillaries. The use of this interface is much less obvious for those devices
and technologies
requiring very low amount of fluids, for example at the micro-litre or sub-
micro-litre level,
since the collection of minute quantities of fluids at the interface between
the cover and the
chip is more critical, happening across a joint between different parts.
In a further approach, WO 00/78456 by Aclara Biosciences, Inc. describes a
microfluidic device whose interface is planar and manufactured on top of a
microstructure
layer. The interface is designed in such a way to be compliant with the well-
to-well spacing of
a standard 96 or 384 micro-titre plate, which is standard within discovery
labs within the
pharmaceutical industry. Using this approach, one single chip can be loaded
fiom a standard
dispensing system as if it would be one single micro-titre plate. The
operation of loading a
plurality of microfluidic devices therefore becomes the repetition of the
single-device loading
procedure a plurality of times, and the loading time is therefore proportional
to the number of
devices to be loaded.
Another approach is disclosed in WO 02/055197 by Evotec OAI AG. In this
disclosure, a sample carrier is disclosed where micro-reactions happen in
wells equivalent to
the standard micro-titre plate, but characterized by a significant reduction
of assay volumes.
This reduction is made possible by specific devices to prevent evaporation,
that include the
tight sealing of the wells by lidding the device with a hard cover,
complemented by specific
dispensing devices optimized for low volumes dispensing and readout means
designed for this
format. It should be emphasized, however that in order to simplify the loading
operations,
Evotec also commercializes devices that are compatible with the standard
1536/384/96 micro-
titre plate formats.
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
8
This approach substantially emulates the current mechanism of fluid handling
and
containers, by specifically addressing the limitations (evaporation and
dispensing accuracy
among others) by customised approaches. In particular, to fully exploit the
reaction
miniaturization, the dispensing accuracy has to be increased according to the
volume reduction,
and Evotec therefore commercialises custom dispensing devices with increased
performances
to substitute the conventional systems used in the industry, that possibly
constitutes a barrier
for adoption.
These custom devices require dispensing heads substantially different from
conventional pipetting system, going for example from technologies where the
dispensing head
is disposable (plastic tip) to technologies where the dispensing head is not
disposable.
Differently from the domain of inkjet printing, where the fluid contained in a
dispensing head
doesn't change during the head lifetime, here the fluids are continuously
substituted, and they
have very different chemical properties. It is of the uttermost importance to
avoid any possible
contamination, and the use of a non-disposable dispensing head therefore
constitutes a
liinitation requiring cleaning and quality check of the cleaning procedure, if
not head
replacement with a significant operational cost increase.
A further approach is disclosed in WO 01/87475 by TECAN AG. This disclosure
describes the implementation of an interface meant to adapt a centripetal
microfluidic disk to a
conventional robotized fluid handling system. This is achieved by
manufacturing, in the
region internal to the area occupied by the microfluidic structures, 48 input
wells with an
interspacing pitch corresponding to the 384 and 96 well-plate standards for
columns and rows
respectively. Using this approach one-half of a micro-titre plate could be
transferred to a
single microfluidic device, and the device could be loaded with conventional
fluid handling
devices. Unfortunately, the surface occupied by the interface cannot be used
for additional
microstructures since fluids move radially outwards, and therefore fluids in
the inputs could
not reach microstructures at smaller radii than the inputs themselves.
Another approach to interfacing fluid handling devices is shown in US Patent
No.
6,620,625 and 6,149,787 to Caliper Life Sciences, Inc. These disclosures
recognize the need
of a high-throughput interface for microfluidic devices for compound sampling
in drug
discovery screening. The Caliper approach addresses this challenge by means of
capillary
forces generated by immersing a capillary into a liquid (sipping). According
to this interface,
the fluid transfer is achieved by first dipping one end of a capillary,
integral part of the
microfluidic circuit, into the fluid source and subsequent filling of the
capillary. Limitations of
the technology consist in the difficulty of sampling different volumes of
fluids, for example
required when the reagents have different concentrations. Using this
technology large volumes
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
9
are impossible to transfer in one operation since the surface tension forces
would not overcome
the gravitational force. A further limitation of this interfacing technology
consists in the
problem of contamination. The sipping operation implies that residues of the
previously
sipped compound can be possibly transferred to the next well, therefore
damaging the source
integrity.
Yet another approach is disclosed in US Patent No. 6,090,251 to Caliper Life
Sciences,
Inc. This patent discloses a custom micro-structured plate for dispensing
fluids into a
microfluidic device. The interface is designed in order to minimize fluid
losses, and is
optimised for the transfer of ininute quantities of fluids in parallel. While
this solution
improves the throughput of the loading operation, it is essentially limited in
the versatility
since the involved volumes are not arbitrary and depend on the. geometry of
the plate and on
the characteristics of the fluids involved, for example the surface tension
properties.
A further approach is disclosed in WO 03/035538 by Gyros AB. This disclosure
describes an interface suitable to centripetal systems, where the requirement
of high throughput
dispensing is achieved by dispensing droplets at high repetition rates in a
fixed position, where
at the same time the microfluidic device rotates below the dispenser. This
microfluidic device
presents inputs at constailt radius but at different angular positions. By
synchronization of the
droplet ejection with the disk motion, the drops ai-rive into the right
receptacles present in the
disk. This interface technology optimizes transfer speed and metering accuracy
for small
volumes of fluids, at the price of a loading facility which is custom designed
for this specific
microfluidic device. Unfortunately, a limitation of this dispensing technology
consists in the
contamination of the drop ejecting head, which comes in contact with the fluid
by ineans of a
non-consumable coinponent. To avoid contamination, it has to be accurately
rinsed before
being reused in the next dispensing operation with a different liquid.
Another approach is disclosed in WO 00/78456 to Orchid Biosciences, Inc. This
disclosure is an original implementation of a microfluidic device interface,
since the fluidic
connections are more inspired to the electronic industry than to the
biochemistry traditions.
The chips are connected by fluid-tight sockets to external tubing, and the
liquids flow into the
microfluidic device as consequence of pressure applied to the tubes by
external actuators. The
complexity of the connections makes this solution improbable for high-
throughput liquid
loading, since each chip has to be fully connected to the loading device
before being used.
Tubing contamination is a major challenge, and its systematic replacement
would imply a
significant amount of consumables cost and additional logistics.
A manufacturing method of producing micro-fluidic devices is disclosed in M.
A.
Gretillat et al. (Sensors and actuators A 60 (1997) 219-222). This article
discloses a
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
___ _ 10
manufacturing method for the realization of inputs on a Pyrex microfluidic
device, which is
manufactured according to a inultilayer and multi-substrates structure. The
microfluidic
components, thin capillaries, are manufactured on one layer and communicate
with a second
layer of structures with larger dimensions, the inlets, through connection
holes. The inlets
reach the border of the device, and fluid loading is possible by means of
needles to be inserted
in the bore. In this design, inlets and microstructures sit on two different
layers which are
manufactured with the same technology but independently. The manufacturing of
the overall
device requires structuring of three different planar substrates, one of which
is etched on both
surfaces and shared between the layers, for a total of four different micro-
structuring steps.
SUMMARY OF THE INVENTION
In the current inventive device and method, a plurality of micro-fluidic
devices or tiles
are assembled in a three-dimensional structure while maintaining a two-
dimensional interface
format. This assembly allows fast and efficient loading operations of these
micro-fluidic tiles.
According to the invention, a plurality of tiles can be loaded in parallel as
if they would be a
single conventional micro-titre well plate, and not in sequence as done by
most existing
implementations. In addition, these multi dimensional characteristics of the
inventive
microfluidic tiles can be achieved by loading them by means of conventional
standard liquid
handler devices. The inventive three dimensional assembly can be permanent, or
preferentially
made to allow the detachment of the individual tiles, or a subset of the
tiles, for other
operations including loading, assay processing, readout of the assay, disposal
of the fluids or
partial processing of the assembly.
For the purpose of this disclosure no distinction is made between inputs,
inlets, outlets,
ports, connections, wells, reservoirs and similar words, all referring to the
means by which
fluids can enter, or exit, from the microfluidic network.
According to the invention, ports are not located on the planar faces of the
substrate, as
in prior approaches, but are located on one or a plurality of small faces. In
one illustrative
embodiment the inputs sit in the same plane of the microfluidic structures.
This makes
possible the manufacturing of ports with the same manufacturing technology
used for
replicating the microstructures. Typically, ports will sit in-between,
adjacent or nearby to the
interface cover-substrate or substrate-substrate; this interface is often
present in planar
microfluidic devices, where open-roof structures are created onto a planar
surface and an
additional substrate closes the roof to guarantee fluid tightness. Cover and
substrates can
either have a symmetrical role, for
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
11
example similar dimensions and presence of microstructures in both, but also
could
substantially differ in size, footprint, thickness, dimensions and
manufacturing process.
A class of devices manufactured according to the invention, is the one
consisting of a
sandwich of substrates which are simply connected, and have input ports
accessible from the
outside of the sandwich. A geometrical object is called simply connected if it
consists of one
piece and doesn't have any circle-shaped "holes" or "handles". For instance, a
doughnut (with
hole) is not simply connected, but a ball (even a hollow one) is. A circle is
not simply
connected but a disk and a line are. In a simply connected substrate it is
possible to take a
piece of string and position a first end of the string onto the substrate at
any point. When the
second end of the string is allowed to follow any arbitrary path and it is
connected again with
the first end, the string forms a loop. If it is always possible to detach the
loop from the
substrate without cutting the string or the substrate, the substrate is simply
connected. In other
words, if there is any path that makes it impossible to get the loop of string
out, the substrate is
not simply connected. If no path from any point of entry gets the loop caught
in the substrate,
then it is simply connected.
Advantageously, with respect to the present passing-through solution with
inputs on the
main faces as mentioned, the inveiitive devices and methods allows using a
homogeneous
manufacturing method for ports and inicrostructures, which minimizes
replication costs and
post-processing operations. Many production processes can allow the input
ports on small
faces to be produced at the same time that the microfluidic structures are
produced. This
reduces the cost of production processes and improves related quality control.
According to the invention, production methods such as hot embossing can take
advantage of inputs manufactured on the small faces. The hot embossing
technique relies on
the change of properties of polymers and similar materials, wllich form
substrates according to
the invention, when their temperature is increased. The softening of the
material, aided by
application of pressure on the surface, allows modifying the morphology of the
surface of the
substrate with the purpose of replicating microstructures. In one illustrative
embodiment of the
invention, inlets can be manufactured by means of the same process, without
requiring any
modification to the deep part of the substrate that would be more difficult to
achieve and would
also imply the displacement of large volumes of material, with subsequent
deformation of the
sample. The inputs can therefore be designed directly in the master containing
the microfluidic
structures, so as to replicate the microfluidic components and the inlets in a
single production
step.
In a further illustrative embodiment, production of the inventive tile by
injection
moulding advantageously allows ports on the small faces. In fact, passing-
through inlets
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
12
require the presence of deep structures in the mould, and their design is
critical both in relation
with the connections with the microfluidic structures, as explained earlier,
and in connection to
the filling behaviour of the fluid polymer during injection. The injected flow
in particular
should allow the polymer to reach all empty parts of the cavity, with limited
pressure drop and
temperature decrease, and this becomes more difficult when extruding
structures are present on
the path. Typically, structures with low aspect ratio and positioned on the
outer surface of the
replicated part are preferable, as in the case of side inputs design,
therefore side ports are a
desirable solution for devices replicated by injection moulding.
In a further illustrative embodiment, inputs on the small faces constitute an
advantage
also for the production of silicon microfluidic devices, because there is no
need to penetrate
deeply into the silicon structure. Since silicon is a hard material with
crystalline structure, it is
brittle and difficult to machine with mechanical means. Passing-through inputs
are preferably
generated by chemical etching, which requires a long and aggressive erosion of
the material
that implies particular care in the control of shape and vertical profile of
the inlets. With inputs
on the small faces, the penetration of the process can be limited to the skin
of the substrate,
independently of the ports volume and shape that can be adjusted by the design
of the planar
lithographic masks. The etching process is therefore more reliable and the
time for etching
advantageously reduced.
In a further illustrative embodiment, laser ablation is often used for the
production of
microfluidic devices. In this production method, a laser beam removes desired
material, by
ultra-violet irradiation of a polymer and therefore produces a small pit that
can be moved over
the substrate to design an actual microfluidic structure. With this method,
the realization of a
passing-through input as in prior art approaches would require an unpractical
amount of time,
or additional processing. However, in the case of inputs on the small faces,
ports can be
manufactured on the skin of the substrate.
In the case of traditional main face inputs, which require a thick substrate
or a design
where the liquid containing cavity is larger than the input. The use of a
larger cavity, however,
produces bubbles that prevent an easy filling of the port, which is hardly
ergonomic for
loading. The interface design, according to the invention, allows for a large
variety of input
geometries, both concerning the shape of the opening and the longitudinal
shape of the
reservoir governing the fluid collection. In particular, ports located on the
small faces
according to the invention can be built in two halves, each of which belongs
to different
substrates. The port can be symmetric, for example half on one substrate and
half on the other
substrate of the sandwich, but it could also asymmetric, and for example
completely on one
substrate.
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
_ 13
The shape of the inventive input opening can be in the form of any geometric
shape
including but not limited to a square or hexagonal shaped input. The inventive
input can be
manufactured by hot embossing or injection moulding or by means of joining two
substrates
symmetrically embossed with a rectangular or trapezoidal master. The
longitudinal shape of
the input can be essentially chosen according to the need. It is contemplated
within the scope
of the invention that cones, inverted cones or "expansion chambers" become
feasible, that
would be otherwise very expensive to manufacture in the case of passing-
through ports located
on main faces.
Aiiother adva.ntage of inputs on the small faces according to the invention is
related to
the optical integrity of the main face surface. According to the invention,
the main face of the
microfluidic tile has no additional structures on its outer surface. This
advantage allows
microfluidic structures contained inside the device to be optically accessible
from the outside
through a homogeneous, planar, optical grade substrate surface. This aspect of
the present
invention is particularly relevant for most optical readout means, like for
example microscopes,
confocal imagers, surface plasmon resonance readers, fluorescence readers,
absorbance
readers, light scattering measurement devices, polarization sensitive light
detectors, but also for
devices irradiating the samples or the microfluidic devices with light beams,
for example the
microfluidic device disclosed in the international patent application
W004050242A2, which is
incorporated in its entirety by reference.
It is contemplated within the scope of the invention to have a microfluidic
device with
side inputs inserted directly or by means of adaptors in a conventional micro-
plate reader,
which is optically accessing the microfluidic reactors from the substantially
flat surface of one
of the main faces, or both. This configuration does not compromise the optical
readout of the
samples, which are optimally accessed through a planar window. Equivalent
optical solutions
having the ports still on the main faces, but displaced from the microfluidic
structures, are less
efficient in terms of manufacturing costs, since the same device would occupy
a larger surface.
In addition to the minimal modification to the production method, side inputs
do not
typically require modifications to materials used in the manufacturing process
of the
microfluidic device, being essentially the same manufacturing process adopted
for the
replication of the microfluidic structures in the device. For example, most
polymers used in
injection moulding, like COC, COP, PC, PMMA, PS, and similar are all suitable
for injection
moulding production of side inputs, and devices with side inputs and different
manufacturing
methods can be made in most of the materials used today like PDMS, glass,
photosensitive
substrates, silicon, metals semiconductors and crystals.
Other advantages of the interface of the present invention become more evident
when
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
- -- - 14
complemented with specific microfluidic technologies, like the one disclosed
in the
international patent application W004050242A2. In this case, the requirement
of accurate
dosing of minute quantities of fluids, which is typically difficult with
conventional dispensing
systems, is achieved by complementing the dispensing device functionality with
precision
metering of the fluids inside the microfluidic device.
The present invention advantageously allows the use of existing dispensing
solutions
designed for the macroscopic world by extending and expanding their use with
microfluidic
devices without the need of additional instruments. For example, from the user
point of view
the metering accuracy of the existing dispensing device is virtually extended
to the
microfluidics, and enhanced for small volumes dispensing. On the other hand,
there is still the
possibility of dispensing large volumes into the microfluidic device, which is
sometimes
necessary for the distribution of buffer liquids. The dynamic range of the
dispensing operation
is therefore increased, and allows more flexible operations with respect to
solutions
specifically designed for microfluidics.
A further advantage of the present invention relates to the loading process,
and in
particular when the performances of the microfluidic devices imply high-
throughput (or high
efficiency) loading. High-throughput loading is a challenging process that
requires
optimization of various methods and device performances, for example the fluid
dispensing
action and the related operations, like tip disposal or needle cleaning for
example, but also the
robotized handling of the microfluidic devices, that determines the time
needed to replace a
device on the fluid handler apparatus with a new unit. These operations, in
particular in drug
discovery, often require the use of automation, not only for reasons of speed,
but also for
reasons of reliability and reproducibility.
The performances of a conventional fluid handling station, therefore, can be
optimized
along various directions: first, by performing more assays on average for
unitary operation of
the fluid handling station, i.e. the loading process. This is typically the
objective of most
microfluidic devices that integrate readout and different degrees of sample
preparation and
metering inside the device itself. Second, by designing the microfluidic
devices and their
interfaces in order to interact only at the beginning or at the end of the
assay process with the
fluid handling system, and allowing the reagents to be stored on the
microfluidic chip for the
duration of the assay, in order not to require external dispensing operations
during the assay
protocol. Third, by reducing the dead-time generated by the replacement of the
microfluidic
devices on the fluid handling system and its related logistics.
According to the invention, parallel loading of a plurality of microfluidic
devices is
performed in a single fluid handling operation. In fact, conventional fluid
handling robots
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
spend a large fraction of time, and also the largest part of consumables cost,
in the operation of
cleaning the dispensing head (or in its replacement), and in the operation of
loading the right
fluid into the dispensing system. Therefore, loading more than one device in
parallel by single
5 or multiple dispensing allows a faster throughput and a reduction in
consumables cost.
It is an object of the invention that a plurality of microfluidic devices are
collectively
organized in a space having a suitable format that presents to. the fluid
handling device a
unitary interface. The interface, possibly compatible with existing standards,
exploits
advantageously the presence of inputs on the small faces of the microfluidic
devices, in order
10 to assemble the tiles in a compact object, hereby referred to as a "brick".
The tiles can be kept
together in the brick by mechanical solutions, like pins, enclosures, slits,
slots, locks, covers,
snap-in elements, spacers, "lego-like" connectors, elastic means but also by
the use of adhesive
layers, magnetic means, or the like.
It is contemplated within the scope of the invention, that the brick can
comprise
15 additional structures, such as a frame, or could be assembled by simply
coimecting the tiles
together in a frameless format. The frame can be designed in order to
reproduce the loading
features of a standard micro-titre plate, but could also be designed in order
to minimize, for
example, dust contamination of the inlets. The fraine can have additional
functional roles, like
tile ejection means or a collapsible structure for tile extraction, or thermal
insulation, heating
and cooling capabilities.
The frame could also be inspired to the structures used in the manipulation of
silicon
wafers in the electronic industry, for example as described in patents
US4248346 and
US5125524, which are incorporated by reference in their entirety, or to the
structures used in
the optical media industry for storage of data. It is contemplated within the
scope of the
invention that the frame can act as a shipping support, protecting the tiles
as if they would be
within a package, or could be simply an alignment mean in order to facilitate
the loading of
liquids with conventional fluid handling devices.
The assembly and disassembly of the brick into its constituent tiles, or the
addition of
one or more tiles to the brick as well as the removal of one or more tiles
from the brick, can.be
achieved in different ways and these operations, individually or collectively,
are here referred
to as packing operations. In some illustrative embodiments the frame could act
as a tile holder,
and the tile position is defined by the frame. In other illustrative
embodiments, the tile position
can be defined by the neighbouring tiles, or by other tiles in the brick. In
some cases
individual tiles could be packed individually, or could be accessible by a
"first-in first-out" or a
"first-in last-out" paclcing approach.
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
16
The tiles can be packed in a brick by means of a "top face" insertion, where
the top face
of the brick is defined as the face constituted by the assembly of tile faces
presenting the
inputs, but also by packing the tiles from the bottom or by one or more of the
lateral faces of
the brick. It is also understood that independently of the presence of the
frame the face of the
brick where liquids enter and exit does not correspond necessarily to the
faces where the tiles
enter or exit for the packing operations.
It is contemplated within the scope of the invention that the brick fluids can
be loaded
at the beginning of the assay, minimizing the time occupation of the fluid
handling system and
at the same time the use of dispensing consumables like tips. The brick can
afterwards be
disassembled into its constituents tiles, which are then processed
independently or in parallel
according to the user needs. The location of inputs on the small faces
according to the
invention allows for a compact, unitary interface for a plurality of devices.
In the brick assembly, the surface occupation is minimized since the main
faces of the
tiles are facing each other, while all inputs remain accessible. If the main
faces are vertical, the
tile occupation is done at the moderate expense of vertical occupation of the
brick.
With particular microfluidic technologies such as disclosed in international
application
W004050242A2, it is possible to exploit additionally the brick geometry since
a large number
of active reactors, and metering elements, can be manufactured on a tile. The
inputs of the
loading interface, therefore, are just entry ports that allow the fluids to
access to a more
complex fluidic logic that allows performing a plurality of assays in
parallel, in a plurality of
conditions.
The functionality of the brick is largely extended with respect to the
functionality of a
micro-titre plate, since the assays can be performed starting from raw
reagents, without the
need of pre-dilution or incubations, and all assays can be performed after the
loading operation,
without the need during the assay protocol of an external dispensing system.
This possibility
allows a significant improvement in throughput and logistics, since the
loading process
becomes the straightforward operation of providing to the brick the reagents
required, and then
the fluid handling instrument can be released for subsequent loading
operations while the tiles
from the loaded brick are being processed.
Another advantage of inputs on the small faces and of the brick design
according to the
invention is related to the intrinsic sensitivity of microfluidic devices to
the presence of dust
particles or residues in the microstructures that could potentially compromise
their
functionality. These particles can enter into the microstructures in various
moments: during the
manufacturing process, when the liquid is inserted in the microfluidic device,
but also when the
air around the inlets contains dust, which enters into the inputs before the
liquid is loaded in the
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
17
device. In the last case, the liquid transports the dust particles inside the
microfluidic device,
and clogging could occur when the size of the particles is similar to the size
of the liquid
passages.
A typical procedure to prevent the deposition of dust particles inside the
inputs consists
in the systematic protection of the inlets by application of seals, films,
covers or similar means.
This procedure is simpler, more effective and more economical when it is
performed on the
side inputs, since the number of sealed inputs per unit surface of the cover
is larger in a tile
with side inputs with respect to a tile with the same inputs on the main
faces, and more tiles
can be protected at the same time by one cover when they are assembled into a
brick.
An additional advantage of the brick concept according to the invention
consists in the
possibility of sealing the brick as a single object, with the purpose to
preserve the reagents
loaded in the tiles from evaporating in the time lapse between the loading
operation and the
actual assay. This is important since the time-lapse between the loading and
the processing
steps does not affect the result of the assay, allowing for an optimal
allocation and scheduling
of the instru.inent and of the other resources involved. Sealing could be
performed on the
complete set of tiles in the brick, or on a partial set of tiles in the brick,
as well as on the
complete set of inputs of a tile in the brick or on a partial set of inputs of
a tile in the brick, or
on any combination of these solutions.
According to the invention, sealing consists in the deposition of a film on
top of the
brick input surface, coinposed by the tiles inputs. The sealing film can be a
layer of polymer,
metal or a coinbination of both. The film can be applied by means of
additional pressure
sensitive or heat sensitive adhesives, but also the film itself could present
intrinsic adhesive
properties. Heat sealing is one of the options most compatible with reagents,
and it is used
both for temporary sealing (peelable films that prevent evaporation) or
permanent sealing (long
term storage that guarantees the integrity of the sample, like in drugs
packages). Other
embodiments of sealing options comprise the use of films that can be pierced
by needles or
tips, allowing the passage of fluids during dispeilsing but preventing the
passage of gas after
the fluid dispensing has been performed as disclosed in US Patent No.
US5789251. It is
contemplated within the scope of the invention that the design of the side
inputs could
reproduce one or a plurality of rows (or columns) of a standard micro-titre
plate, so that most
of the existing sealing technologies for micro-titre plates can be used.
It is also contemplated within the scope of the invention that when a brick
has been
sealed, individual tiles can be separated and processed independently if
required by cutting the
film sealing the brick in the direction parallel to the main faces, therefore
with the possibility
of keeping the tile sealed after removal from the brick assembly.
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
18
The sealing of individual tiles becomes more important when complemented with
specific microfluidic technologies, like the one disclosed in the
international application
W004050242A2. With the valving technology described in this international
application, the
liquid contained in the sealed reservoir can be transferred into the
microfluidic structures
without requiring the opening of the seal. Therefore an individual tile, pre-
loaded with
reagents, can be processed directly without requiring the opening of the
sealed reservoirs that
could be therefore permanently sealed. In fact, the reservoir can be put in
fluidic
communication with the microfluidic circuit by the opening of two lines, one
required for the
liquid flow and the second one required for the passage of gas, typically air,
to prevent the
formation of an under pressure in the reservoir that would prevent the
extraction of the liquid.
With this method, tile pre-loading becomes possible and can also be applied to
a subset of the
inputs present in the tile.
Another advantage of assembling tiles into a brick according to the invention
consists
in the possibility of labelling the tiles, either individually, as a block, or
both. It is
contemplated within the scope of the invention that identification of the
brick could follow the
same common practice adopted for micro-plates, and individual tile labels
could be readable by
a user without the need of additional instrumentation for a simple and rapid
sorting of the tiles
in the brick. The same information can be used to know, when the assay is
performed, which
reagents have been loaded into the tile and which assay should be performed
for that specific
tile.
Labelling can be achieved by optical, mechanical, magnetic or radio means, and
the
label readout could require an external instrument, or could also be possible
by simple visual
inspection. Examples of bar-coding implementations are mechanical
modifications of the tiles
or of the brick (punching or removal of tabs), colour of the tiles, graphical
drawings for
ordering the tiles (like for example diagonal lines or texts across many
tiles), application of
adhesive barcode labels, direct printing of labels onto the tiles by inkjet or
thermal methods,
application of substrates with magnetic properties, or insertion of radio
emitters or
transponders.
Optical label information could be encoded in one-dimensional or two-
dimensional
formats, the latter allowing for space savings. Optical barcodes could be
preferentially applied
on the small faces, in such a way that the labels are still accessible and
visible when the tiles
are assembled in the brick format. The optical barcodes could also be
positioned on the same
face where the side inputs are located, but also sideways or in alternative on
the bottom or on
extruding parts.
Another significant advantage of the inventive brick consists in the extremely
compact
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
19
format, where the number of assays per unit volume (or per unit surface) can
be dramatically
increased. This compact format is useful in applications requiring the storage
of compounds
for the pharmaceutical industry, and the mentioned advantages are further
enhanced by the fact
that compounds can be accessed on a brick basis but also by extraction of
individual tiles.
BRIEF DESCRIPTION OF THE DRAWINGS
These and other advantages, objects and features of the invention will be
apparent
through the detailed description of the embodiments and the drawings attached
hereto. It is
also to be understood that both the foregoing general description and the
following detailed
description are exemplary and not restrictive of the scope of the invention.
Figs. 1A, 1B, 1C and 1D depict an embodiment of a rotor tile according to the
invention, where the inlets are on the small side of the tile and the tile can
be designed to fit
into a brick;
Fig. 2 depicts a design for input interfaces according to the invention,
optimized for
injection moulding mass-production;
Fig. 3 illustrates another specific embodiment according to the invention
where the side
inputs can be manufactured so that microfluidic structures and inlets on the
tiles are physically
separated during the production of the substrate;
Fig. 4 depicts a single tile according to the invention that is partially
sealed by
application of a film that prevents the fluid evaporation;
Figs. 5A, 5B and 5C depict a further illustrative embodiment having a format
compatible with 1536 micro-plates, where only 768 of the inputs are actually
implemented;
Fig. 6 depicts a tile and the related brick assembly according to the
invention;
Fig. 7 depicts tips of a multi-head dispensing device, and loading of the
tiles is
performed as with a micro-titre plate;
Fig. 8 depicts a centripetal microfluidic system according to the invention,
where the
microfluidic tiles are subject to the centrifugal force by means of a spindle
device allowing
moving the fluids inside the microstructures;
Fig. 9 illustrates a further illustrative embodiment according to the
invention, where a
plurality of bricks is used in the loading operations with minor modifications
with respect to
the design of a single brick loader;
Fig. 10 illustrates a further illustrative embodiment according to the
invention, where
tiles within a brick are extracted from the bottom of the brick;
Fig. 11 illustrates a fiu-ther illustrative embodiment according to the
invention, where
an automated extraction solution for tile removal is shown; and
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
Fig. 12 illustrates a rotor adapted for receiving tiles according to the
invention.
DETAILED DESCRIPTION OF THE INVENTION
5 The present invention provides microfluidic tiles that are used within
centrifugal rotors
and microsystems and in particular nano-scale or meso-scale microfluidic
platforms as well as
a number of its applications for providing centripetally-motivated fluid
micromanipulation.
For the purpose of illustration, the drawings as well as the description will
generally refer to
centripetal systems. However, the means disclosed in this invention are
equally applicable in
10 microfluidic components relying on other forces to effect fluid transport.
For the purposes of this specification, the term "sample" will be understood
to
encompass any fluid, solution or mixture, either isolated or detected as a
constituent of a more
complex mixture, or synthesized from precursor species.
For the purposes of this specification, the term "in fluid communication" or
"fluidly
15 connected" is intended to define components that are operably
interconnected to allow fluid
flow between components. In illustrative embodiments, the micro-analytical
platform
comprises microfluidic tiles within a rotatable platform, such as a disk, or
experimental micro-
fluidic chips whereby fluid movement on the chip is motivated by centripetal
force upon
rotation of the chip and fluid movement on the experimental chip is motivated
by pumps.
20 For the purposes of this specification, the term "biological sample",
"sample of
interest" or "biological fluid sample" will be understood to mean any
biologically-derived
analytical sample, including but not limited to blood, plasma, serum, lymph,
saliva, tears,
cerebrospinal fluid, urine, sweat, plant and vegetable extracts, semen, or any
cellular or cellular
components of such sample.
For the purposes of this specification, the term "meso-scale", or "nano-scale"
will be
understood to mean any volume, able to contain as fluids, with dimensions
preferably in the
sub-micron to millimetre range.
Representative applications of microfluidic tiles within a centripetal system
(e.g.,
centrifuge) employ rectangular shaped devices, with the rotation axis
positioned outside the
device's footprint. For the purpose of illustration, the drawings, as well as
the description, will
generally refer to such devices. Other shapes other than rectangular shaped
devices are
contemplated within the scope of the invention including but not limited to
elliptical and
circular devices, irregular surfaces and volumes, and devices for which the
rotation axis passes
through the body structure, may be beneficial for specific applications.
Turning to Figs. 1A and 1B a first illustrative embodiment a tile 101
according to the
invention is shown. The tile 101 is a substantially planar object formed from
a first substrate
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
21
102 and a second substrate 106. It is contemplated within the scope of the
invention that the
tile 101 can be also formed from more than two substrates. The substrates 102,
106 can be of
any geometric shape. The substrates 102, 106 contain depressions, voids or
protrusions that
form microfluidic structures when the substrates are bond together. In a first
illustrative
embodiment the substrates 102, 106 have a film layer 110 sandwiched between
them. The film
layer 110 allows for separation of voids within the substrates forming
microfluidic circuits that
can be placed in fluid communication within each other by perforation of the
film layer 110. It
is contemplated within the scope of the invention that the substrates 102, 106
can be joined
within the film layer 110 in between them.
In this first illustrative embodiment the tile 101 is substantially
rectangular structure
having an input end 103, a bottom end 105, a first planar surface 109 and a
second planar
surface 108. The bottom end 105 has an affixing tab 107 allowing for handling
and insertion
of the tile 101 into a holder or fraine. In this illustrative embodiment the
input end 103, which
is also referred to as a small face, has a plurality of input ports 113. The
input ports 113 are in
fluid communication with at least one fluid handling microfluidic circuit 115.
It is
contemplated within the scope of the invention that these microfluidic
circuits 115 may be
composed of a series of valves, chambers, reservoirs, microreactors and
microcapillaries. It is
also contemplated within the scope of the invention that the series of
microreactors and
microcapillaries are in fluid coinmunication with a detection chamber.
The tile 101 has an accessory area 117, which can be used for the purpose of
manufacturing, handles, structural supports, precision spacers, purging
volumes, bonding
areas, identification areas or the like.
The functionality of a specific microfluidic circuit 115 can be configured
within the tile
101 to perform a desired assay upon a selected sample. It is conteinplated
within the scope of
the invention that any microfluidic or fluidic assay known in the art can be
configured within
the tile 101 to achieve a desired functionality. With reference to Fig. 1 C a
fluidic circuit 121 is
shown having a first state having a reagent contained in a first 120 and
second 122 reservoir.
With further reference to Fig. 1D, the fluidic circuit 121 is shown in a
second state after
valving within a valving matrix 123 is actuated. It is envisioned that the
inventive tiles 101 can
having a plurality of fluidic circuits 121 that can perform processes in
different regions, by
actuating the valving matrix 123 as illustrated by the first and second state
of the fluidic circuit
121 as depicted in Figs. 1C and 1D.
As illustrated in Fig. 1C, a method of joining two fluids in given proportions
at a
selected time is shown with respect to a first reagent within the first
reservoir 120 and a second
reagent within the second reservoir 122. According to the invention the first
and second
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
22
reagents are transfer in a desired proportion to a mixing chamber 125. The
desired proportion
of each reagent is delivered to the mixing chamber 125 by actuating the
valving matrix 123 as
depicted in Fig. 1D. These reagents can include but not be limited to the
dilution of a reagent
into a buffer, the occurrence of a chemical reaction with a given ratio of
volumes of reagents,
modification of the pH of a solution by addition of an acid or a base, an
enzymatic assay where
a protein comes into contact with an antibody, or the like.
The fluid handling process starts by the opening of a valve 130 within the
valving
matrix 123, which could of the type described in the patent application
W004050242A2 ('242
application), wlierein the film layer is perforated to actuate a valve. The
teacliings of the '242
application are incorporated herein by reference. It is contemplated within
the scope of the
invention that the valving mechanism could also be of different types known in
the art such as
a mechanical valve or the like. According to the invention the reservoirs 120,
122 are
positioned onto a different plane with respect to connecting capillaries
within the valving
matrix 123, and they are separated by means of the film layer 110 that can be
perforated at a
selected location(s) by irradiation, therefore producing a virtual valve 130
as shown in Fig. 1D.
The opening of valves 130, together with the application of a non-equilibrated
force
onto fluids, allows for the movement of liquids into the mixing chamber 125.
The non-
equilibrated force could be generated by means known in the art. In this first
illustrative
embodiment the non-equilibrated force is achieved by centrifugation so that
the liquids are
subject to a centripetal acceleration directed towards the bottom of the tile
101. According to
the invention the amount of fluids which are transferred to the mixing chamber
125 is
determined by the radial position of valves 130, since only the fluid
contained above the
corresponding valve 130 is allowed to descend into the inixing chamber 125.
The process
could be replicated in a plurality of subsequent layers, giving the
possibility of successive
dilution over various orders of magnitude, mixing two or more type of liquids
together,
incubating fluids for a given amount of time into the reactors, or even
performing a real-time
protocol over the matrix layers.
Turning to Fig. 2, a second illustrative embodiment depicting a microfluidic
tile
according to the invention is shown. The microfluidic tile 210 is comprised of
a first substrate
200 and a second substrate 201. The joining of the two substrates 200, 201
forms the
microfluidic tile 210. The microfluidic tile 210 has a bottom face 202, an
input face 203, a
first planar face (not shown) and a second planar face 207. The input face
203, also known as
the small face, of the microfluidic tile 201, contains a plurality of input
ports 209 in a first
input row 211 and a second input row 212. The input face 203 is extruded
outside the space
confined between the first and second planar faces in order to cause a
plurality of microfluidic
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
23
tiles 210 forming a brick having a desired portal interface.
In this illustrative embodiment, the input face 203 contains input ports 209
that have a
pitch and opening dimensions of a standard 384 well micro-plate format. It is
contemplated
within the scope of the invention that the input ports 209 can be configured
to adapt to any
standard laboratory interface. The microfluidic tile 210 is suited to manual
loading operations,
since it is easier to avoid cross-contamination between the inputs ports 209
and to locate the
desired input port(s) 209 on the microfluidic tile 210. According to the
invention, inputs ports
209 are manufactured symmetrically on the substrates 200 and 201 forming the
microfluidic
tile 210. These substrates 200, 201 are not simply connected, since their
inputs are in fluidic
communication with the microfluidic components present at the contact surface
of substrates
200 and 201, which is also the surface at which substrates 200, 201 are bonded
together.
Turning to Fig. 3, an example of a device manufactured by bonding simply
connected
substrates is shown. A first substrate 301 and a second substrate 303 form a
microfluidic tile
305. Inputs 307 are manufactured as depressions on either substrate 301, 303.
These
depressions are manufactured by microstructuring means. It is contemplated
within the scope
of the invention that the depressions could also be manufactured by
macroscopic means with
limited accuracy, for example by milling.
During the manufacturing step, the inputs 307 are not in fluid communication
with
microfluidic circuits on either on substrate 301 or 303. When the microfluidic
tile 305 is
assembled there is fluidic communication between the microfluidic circuits and
the inputs 307.
When the two substrates 301, 303 are bond together fluidic communication with
the
microfluidic structures is established through the substrates 301, 303.
Similarly, all other
inputs ports 307 can be put in fluidic communication with the microfluidic
circuit of the
microfluidic tile 305.
As shown in Fig. 4, a typical requirement of permanent storage applications,
like the
distribution of a diagnostics assay on a microfluidic device, require reagents
to be stored in
liquid, solid, encapsulated or lyophilized form inside the microfluidic
device. A tile 401
according to the invention having input ports 401 are subsequently sealed by
the use of an
impermeable cover 403. The use of the impermeable cover 403 covering inputs
ports 402 is
done routinely in drugs discovery when using standard micro-plates between the
operation of
loading reagents and the actual assay. The impermeable cover 403 prevents
minute quantities
of fluid from evaporating, with the consequence of changing their
concentration and therefore
modifying the assay conditions.
It is contemplated within the scope of the invention that the impermeable
cover 403 can
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
24
be fabricated from polymeric material, natural rubber, or any material having
the feature of
being inert to liquids used and pierceable for the introduction of liquids,
while maintaining gas
tightness afterwards to prevent evaporation of store reagents. It is fiirther
contemplated within
the scope of the invention that the impermeable cover 403 can be obtained by
application of a
laminated film containing metallic and polymeric layers. The metallic layer
allows a low
permeability to gas and liquids, and the polymeric layer allows for an easy
and effective
sealing of the store reagents within the tile 402.
Turning to Figs. 5A, 5B and 5C, a planar microfluidic tile 501 is produced by
micro-
structuring a facing surface of one, or botll, of a first 503 and second 504
facing substrates.
Inputs ports 505 are manufactured in one of the two facing substrates 503, 504
and are
completely contained inside one or both of the facing substrates 503, 504. The
inputs ports
505 have a length inside the substrates 503, 504 that can be decided
arbitrarily accordingly to
the fluid volumes to be loaded and the pitch between successive input ports
505 can be chosen
accordingly to existing standards and specific integration needs. The nominal
pitch values of
2.25 min, 4.5 mm or 9 mm correspond to the 1536, 384 and 96 wells micro-titre
plate
standards respectively. In this illustrative embodiment, the pitch chosen
corresponds to the
1536 micro-titre plate format, with input ports 505 having a square opening.
The substrate 503, 504 with input ports 505 are simply connected. The input
ports 505
can be generated by the same mould insert required for the generation of the
microstructures
forming the microfluidic circuit, or by a second insert (or mould component)
sitting on the
same side of the microfluidic circuit generating insert. In both cases,
removing the piece from
the mould is possible without the requirement of movable parts.
In a further illustrative embodiment as shown in Fig. 6, a microfluidic tile
601 as
previously depicted in Fig. 5 contains one row of input ports 602, and a
microstructure valving
matrix 603 as described in Fig. 1. The microfluidic tile 601 is comprised of a
first substrate
and a second substrate facing each other and bonded together with a film layer
in between.
In this illustrative embodiment, the microfluidic tile 601 has 48 input ports
602, an.d 16
microfluidic tiles 601 form a brick 607. The brick 607 is kept in place by a
frame 608 in this
illustrative einbodiment. It is contemplated within the scope of the invention
that other
methods of affixing the microfluidic tiles 601 into bricks 607 can be used.
The brick 607 has
an upper surface 609 and a lower surface 610. The upper surface 609 is formed
from a
plurality input ports 602 of the comprising microfluidic tiles 601. The
plurality of input ports
602 forms a format of 1536 input ports in a micro-titre plate in a first
direction, and the input
ports 602 have a pitch of a 384 inputs micro-titre plate in a second
direction. The upper
surface 609 is a high density region of input ports 602, which allows for an
efficient filling of
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
the brick 607 with standard existing multi-head or single-head dispensing
devices, which
typically have a head pitch compatible with 96 and 384 inputs micro-titre
plate formats.
It is contemplated within the scope of the invention that the inventive
apparatus and
5 method allows for the assembling of microfluidic tiles 601 in the form of a
brick 607 in any
standard laboratory format or custom format. Microfluidic tiles 601 within
this illustrative
embodiment are parallel to the long side of the brick 607, but with a
different tile design a
brick could host tiles parallel to the short side of the brick 607, with 32
input ports 602 per
microfluidic tile 601 (1 series of 32 inputs), the brick 607 containing 16
microfluidic tiles 601.
10 The number of inputs 602 per microfluidic tile 601, the number of
microfluidic tiles
601 in a brick 607, and the orientation of the microfluidic tiles 601 can be
changed to achieve
various configurations having a standard laboratory format or a custom format.
The various
configurations are dependent on the microfluidic tile 601 design and on the
application and
strategy to collect the microfluidic tiles 601 into a micro-plate-like format.
The segmentation
15 of microfluidic tiles 601 and the number of input ports 602 on the
microfluidic tile 601 can be
made without requiring changes to the fluid handling device and to the loading
process.
Turning to Fig. 7, the loading operation of a brick 701 with a 96 inputs micro-
plate
parallel dispenser 702 is depicted. The brick 701 in this illustrative
embodiment is formed
from a plurality of tiles 705 having a plurality of input ports 709. The
parallel dispenser 702
20 has 8 heads 712 and performs the loading by columns. In this illustrative
embodiment, the
heads 712 move parallel to the long side of the brick 701, and allows the
dispensing of a
reagent or other selected fluid into the input ports 709 of the tile 705.
Since many assays
consists in the repetition of a protocol to test different targets or
different chemical entities in
parallel, a fraction of the reagents or selected fluids of the assay are in
common, and a fraction
25 of the reagents are varied. Once a reagent is available in one dispenser
head 712, it can
therefore be distributed over different tiles in an efficient manner since the
tiles require small
volumes and the pipette tip is used once for all tiles contained in the brick.
The parallel dispensing device 702 has a typical pitch since most of the
dispensing
heads are larger than the pitch of a 1536 micro-plate to maintain
compatibility with the lower
density formats containing 384 and 96 wells per micro-plate. In this
illustrative embodiment,
the spacing for the inputs is determined by a protruding structure of the
tiles 705 and by the
briclc frame 710. It is contemplated within the scope of the invention that
the tiles 705 can be
kept vertical by a comb-like support.
As shown in Fig. 8, tiles 801 according to the invention after being
disassembled from
a brick 802 in a manual or automated way are positioned on a spindle support
803 at constant
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
26
radius. The tiles 801 can be processed individually or in groups, according to
the throughput
needs. It is conteiuplated within the scope of the invention that it is not
required to position the
tiles 801 at a constant distance from the rotation axis, and that the tiles
801 can be loaded in
multiple rows in order to save space occupation on the spindle support 803.
According to the
invention, it is preferable to have inputs 804 on the edge of the tiles 801
facing the rotation
axis. This positioning is desirable since fluids subject to the centripetal
acceleration will tend
to move radially towards the outer part of the spindle and the input 804 can
be optimally
designed for fluid collection.
Inputs 804 on the main faces are configured to avoid spill-over. When inputs
804 of
the tile 801 are on the small face as previously discussed, an additional
advantage consists in
the removal of bubbles. In fact, atmospheric pressure air has a density lower
than the density
of any liquid. Gas bubbles are also subject to the Archimedes principle. In
the case of air in a
liquid at rest, a bubble can remain inside the liquid if the weight of the
bubble, summed with
the surface tension forces, overcomes the Archimedes force. In a centripetal
device, gravity is
rapidly overcome by spinning.
A bubble in a centripetal device, therefore, can be subject to a strong force
directed
towards the rotation axis and perpendicular to it, whose intensity is equal to
the apparent
weight of the liquid displaced. Inputs 804 should be placed on the faces of
the tile 801 that are
directed towards the rotation axis, since the centripetal force will push the
bubble towards the
liquid/air interface with the result of bubble disappearance. The same
consideration applies to
the case where the fluid loaded by the external dispensing system sits on top
of an air volume,
a phenomenon that typically occurs when the introduction of liquid does not
happen at the very
bottom of the container itself. This phenomenon is typical of small-sized
ports since the fluid
rapidly occludes, by surface tension occurring at the contact region with the
side walls, the
passage of underlying gas towards the opening. The centripetal acceleration in
the side input
configuration previously described will drive the fluid to the "bottom" of the
inlet.
Processing of a brick 802 can be accomplished in different ways, in relation
with the
specific microfluidic technology contained in the tiles 801. An example of
brick 802
processing can therefore be made with reference to the microfluidic technology
disclosed in
the international application W004050242A2, in the specific embodiment where
the valving
technology is used in a centripetal platform. In this embodiment, the tiles
801 can be
processed on a centripetal platform, that spins in order to position the valve
actuator in the
correct position, can move the fluids inside the tiles by centrifugation, and
allows the readout
sensor to detect the outcome of the assay in a localized position.
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
27
As shown in Fig. 8, the platform is similar in many aspects to a centrifuge
rotor hosting
horizontal tiles. The tiles 801 can be transferred from the brick 802 to the
rotor in many ways,
one method shown in the figure as an example. The steps of the process can be
identified in
brick loading, tile extraction, tile positioning, tile processing, tile
unloading. The brick 802 can
be loaded on the instrument with tiles 801 in the horizontal position,
profiting of the fact that
the fluids do not escape from the inlets due to surface tensions (or by means
of seals applied to
the inlet). Vertical translation of the brick 802 in the picture allows
choosing the tile 801 to be
processed: without the need of direct tiles identification, this method allows
a unique
association of the tile 801 being processed with the micro plate colurnn (or
row) that was
loaded with the reagents.
Tile extraction from the brick 802 can be achieved by application of pressure
through
an external actuator, for example pushing the bottom of the tile in the
direction towards the
rotor axis. In another illustrative embodiment, the tile 801 could be grasped
by a clamp, or
specific structures created on the tile 801 (like a pin, a hole, a flap, a
flange, a bayonet, a
magnet, an adhesive layer) could be used as means to establish a link with the
actuator. Tile
positioning can be achieved by moving the extracted tile 801 vertically inside
the rotor slot
specifically designed to host the tile 801.
In a further illustrative embodiment, the rotor could present slots which are
accessible
from the outer part of the rotor, and the tile 801 is locked inside the slot
by an active
mechanism, like a key or an electromechanical actuator, preventing the tile
801 to escape from
the rotor as consequence of spinning.
Tile processing occurs by opening in an active way the valves in the tile 801,
by means
of an optical pickup positioned below the rotor, and the readout of the assay
is performed by
means of the same optical path. It should be noted that in this configuration
an identifying
barcode of the tile 801 could be also positioned on the main face of the tile
801, and read
during the spindle rotation. In fact, even if the barcode is not optically
accessible when the
tiles 801 are grouped in the brick 802, the reading of the barcode while the
tile 801 is
positioned onto the rotor allows performing the unique association of the tile
position in the
brick (in other words, the column or the row identifier of the micro plates)
with the tile
barcode, making unnecessary additional tile identification procedures.
After processing, tile unloading can be achieved by repositioning the tile 801
inside the
briclc frame (in the same position or in a different position) through the
sanie movement path.
As another possibility for tile unloading, the tile 801 could be disposed by
completely lifting
up (or
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
28
down) the brick vertical translator, to a disposing unit that could be similar
to a brick or to a
simple pile of tiles for disposal.
As shown in Fig. 9, it is envisioned that various schemes for brick processing
are not
limited to transferring a tile 901 fiom a brick 902 to a processing instrument
905, but refers
also to the process of moving tiles 901 from a plurality of bricks 902 to an
instrument without
substantial modifications. In one aspect of the invention, bricks 902 are
stacked in a vertical
pile within a brick loader 907, and selected for loading by simple vertical
firanslation of the
brick loader 907.
The brick 901 according to the invention can therefore be designed to allow
for vertical
stacking of inultiple bricks 902, as it is conventionally done in well plates,
but also to stack
bricks 902 which contain horizontal tiles 901, witli the purpose of side
stacking. The stacking
of the tiles 901 according to the invention could be facilitated by mechanical
positioning
means, for example pins, slots, "lego comiections," extruding complementary
structures and
similar, in order to allow both vertical and side stacking of bricks 902. It
is contemplated
within the scope of the invention that the possibility of assembling a
plurality of bricks 902 and
treating them as a single brick 902 is essentially possible in all steps,
including brick loading
with fluids, being this feature essentially connected to the modular concept
of assembly of the
tiles 901.
The number of loading steps is determined by the overall number of different
basic
reagents present in an assay. In typical cheinical screening procedures, a
number N of
chemical compounds is screened versus a number M of targets (for example,
proteins) on the
basis of the result of an assay, that typically comprises a small handful of
reagents (in the
following consideration and for this purpose, neglected), operation also known
under the term
of compounds profiling.
Compound profiling procedures in the drug industry are common, and for example
one
of them consists in the determination of the enzymatic activity of a family of
kinase proteins in
presence of various kinase inhibitors. Kinase profiling has the important goal
to assess the
potency of a potential drug while measuring the side effects of the same
molecule towards
other proteins of the same family but regulating different biological
processes. In the operation
of compound profiling, the number of useful data points is essentially
proportional to N times
M, while the number of loading operations consists in N plus M steps.
If all the steps subsequent to the loading process are automated, there is a
significant
scaling advantage in collecting together microfluidic devices to produce in
one go as many
data points as possible, as done in the present invention, since the loading
steps will only
moderately increase: for example, screening 10 compounds vs. 10 targets
produces 100 data-
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
29
points with essentially 20 loading steps, while screening 100 compounds vs.
100 targets
produces 100 times more data-points, with only a ten-fold increase of the
loading steps (i.e. the
amount of work done by the user). The same argument for integration and
collective interface
is valid for most drug discovery and diagnostics applications, where a panel
of a plurality of
assays is performed on a plurality of biological samples, and we can predict
that the future
evolution of pharmaco-genomics will increase the demand and the utility of
panels meant to
screen the patient compatibility with potential therapeutic agents.
Tiles according to the invention are advantageously provided having a variety
of
composition and surface coatings appropriate for a particular application.
Tile composition
will be a function of structural requirements, manufacturing processes,
reagent compatibility
and chemical resistance properties. In particular, tiles may be made from
inorganic crystalline
or amorphous inaterials, e.g. silicon, silica, quartz, inert metals, or from
organic materials such
as plastics, for example, poly(methylmethacrylate) (PMMA), acetonitrile-
butadiene-styrene
(ABS), polycarbonate, polyethylene, polystyrene, polyolefins, polypropylene
and metallocene.
These may be used with unmodified or modified surfaces.
Surface properties of these materials may be modified for specific
applications.
Surface modification can be achieved by such methods as known in the art
including by not
limited to silanization, ion implantation and chemical treatment with inert-
gas plasmas. It is
contemplated witliin the scope of the invention that tiles can be made of
composites or
combinations of these materials, for example, tiles manufactured of a
polymeric material
having embedded therein an optically transparent surface comprising for
example a detection
chamber of the tile.
It is further contemplated within the scope of the invention that tiles can be
fabricated
from plastics such as Teflon, polyethylene, polypropylene, methylmethacrylates
and
polycarbonates, among others, due to their ease of moulding, stamping and
milling. It is also
contemplated within the scope of the invention that tiles can be made of
silica, glass, quartz or
inert metal. The tiles having a fluidic circuit within in one illustrative
embodiment caii be built
by joining using known bonding techniques opposing substrates having
complementary
microfluidic circuits etched therein.
Tiles of the invention can be fabricated with injection moulding of optically-
clear or
opaque adjoining substrates or partially clear or opaque substrates. The tiles
can be square,
rectangular or any geometric form with a thickness approximately comprised
between 1 mm
and 10 mm. Optical surfaces within the substrates can be used to provide means
for detection
analysis or other fluidic operations such as laser valving. Layers comprising
materials other
than polycarbonate can also be incorporated into the tiles.
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
The coinposition of the substrates forming the tile depends primarily on the
specific
application and the requirements of chemical compatibility with the reagents
to be used with
the tile. Electrical layers and corresponding components can be incorporated
in tiles requiring
5 electric circuits, such as electrophoresis applications and electrically-
controlled valves.
Control devices, such as integrated circuits, laser diodes, photodiodes and
resistive networks
that can form selective heating areas or flexible logic structures can be
incorporated into
appropriately wired areas of the tile. Reagents that can be stored dry can be
introduced into
appropriate open chambers by spraying into reservoirs using means known in the
art during
10 fabrication of the tiles. Liquid reagents may also be injected into the
appropriate reservoirs,
followed by application of a cover layer comprising a thin plastic film that
may be utilized for
a means of valving within the fluidic circuits within the tile.
The inventive microfluidic tiles may be provided with a multiplicity of
components,
either fabricated directly onto the substrates forming the tile, or placed on
the tile as
15 prefabricated modules. In addition to the integral fluidic coinponents,
certain devices and
elements can be located external to the tile, optimally positioned on a
component of the tile, or
placed in contact with the tile either while rotating within a rotation device
or when at rest with
a brick formation or with a singular tile.
Fluidic components optimally comprising the tiles according to the invention
include
20 but are not limited to detection chambers, reservoirs, valving mechanisms,
detectors, sensors,
temperature control elements, filters, mixing elements, and control systems.
EXAMPLES
The following examples are provided to illustrate the methods and products of
the present invention with particular choices for the several components
described above. As
described above, many variations on these particular examples are possible.
These examples
25 are merely illustrative and not limiting of the present invention.
Example I
A brick 1000 according to the invention is shown in Fig. 10. The brick 1000 is
comprised of a plurality of microfluidic tiles 1001 within a briclc frame
1005. In a first
illustrative embodiment, the tiles 1001 are extracted from the bottom of the
frame 1002, in
30 order to be processed by related devices. The microfluidic tiles 1001 are
accessed by
microfluidic inlets 1003 on the top face of the brick 1000.
This illustrative embodiment allows a human interface that is designed
independently
from the machine interface. Reagents may be loaded in the inlets 1003 at the
top face of the
brick
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
31
1000, either by manual or automated means. The inlets 1003 are arranged in a
conventional
micro-plate format. As microfluidic technologies consume a very limited amount
of reagents,
the reagent volumes are substantially small. It is known in the art that small
volumes of liquids
are subject to rapid evaporation, that may either deplete the liquid or change
the concentration
of the reagents due to evaporation. A solution to this evaporation problem
consists of the
application of an adhesive polymer film (not shown) on top of the top face
after reagent
loading. The adhesive polymer film can be either temporary or permanent, by
the use of
thermal adhesives, pressure sensitive adhesives or similar means to guarantee
the gas tightness,
which prevents liquid evaporation by an increased vapour pressure.
It is contemplated within the scope of the invention that the same sealing
means can be
used with the brick 1000. The brick 1000 is characterized by bottom extraction
as shown in
FIG. 10. Bottoin extraction has the advantage that a film layer (iiot shown),
positioned on the
top face, can be kept in place until a tile 1001 is extracted from the frame
1002 minimizing the
time that liquids are exposed to air, thereby improving the assay quality and
minimizing the
risk of external contamination.
Tiles 1001 and frames 1002 according to the invention are designed in a manner
so
that, during normal laboratory operations, the tiles 1001 do not exit from the
bottom of the
frame 1002. In one illustrative embodiment adhesive fasteners prevent the
tiles 1001 from
slipping out of the frame 1002. In a further illustrative embodiment
mechanical means are
used that are externally actuated in order to release one or a plurality of
tiles 1001 from the
frame 1002. It is contemplated within the scope of the invention that the
removal of tiles could
be achieved by any mechanical means such as a tab, a lever, or the like.
In a further illustrative embodiment, elastic elements either in the tile 1001
or in the
frame 1002 or in both, exerts pressure in a location of the tile 1001 so that
undesired tile
extraction is avoided. The extraction of tiles 1001 can be achieved by
application in the
direction towards the bottom opening 1007 by means of pushing or pulling pins,
pushing or
pulling rods, various types of clamps, grips, friction wheels, rotating gears,
sliding bars or the
like. In particular, elastic elements may be integrated into the frame 1002,
minimizing the
complexity and the cost of the tiles 1001.
Exatizple II
Turning to Fig. 11, an automated extraction solution for tile removal is
shown. In a
first illustrative embodiment, a selected microfluidic tile 1112 in a brick
1102 is chosen for
extraction by a linear movement of a tray 1101. In this illustrative example,
only one brick
1102 is accessed by the extraction device. It should be understood by those
skilled in the art
that this de-assembly
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
32
procedure can be applied, sequentially or in a desired order, to a plurality
of bricks 1102, either
in an instrument or in a production line. This type of automation is an
efficient solution
allowing for a high throughput or unattended production line, ranging from
compound loading,
reagent distribution, protocol execution and experiment readout. The
production line could be
assembled by a rail or belt driven mechanism where bricks 1102, with or
without reagents, are
fed into the slots of a conveyor, and a continuous flow of experiments can be
performed either
serially or in parallel by means of "bifurcations" of the conveyor, tile
extraction, re-
distribution and brick manipulation.
As shown in Fig. 11, microfluidic tiles 1112 are extracted by means of a
gripper 1103,
that grasps the tile 1112 from the bottom. In a first illustrative example, a
purge voluine 1008
on the tile 1112 is configured so that the movement of the gripper 1103,
actuated by solenoids
1104 or stepper motors, pneumatic actuators or the like, exerts pressure on
the tile 1112
allowing for a firm holding of the tile 1112 in the gripper 1103. A
curvilinear rail 1105 is
configured to transport the gripper 1103 along a complex trajectory, by taking
a vertical tile
1114 contained in a brick (not shown ) to an operational position 1106, where
the tile is
horizontal. The fingers 1107, actuated by pneumatic or electrical means like
solenoids 1108 or
electrical motors, open fixation holders 1109 on a spindle 1110, at which
moment the linear
stage 1111 allow the movement of the spindle 1110 onto the tile 1112 in a set
position 1106.
As is shown in Fig. 12, the spindle 1201 has insertion slots 1202 meant to
keep tiles in
a rotor. The tiles are locked radially by holding means 1203, 1204 actuated by
knobs 1205,
1206 by means of fingers 1207, 1208. With further reference to Fig. 11,
fingers 1107 are de-
energized, and the pressure from the gripper 1103 is released from the tile
1106. Turning to
Fig. 12, the tile is kept inside the slot 1202 of the spindle 1201, and the
subsequent movement
of the spindle 1201 drags the tile away from the loading mechanism, which is
therefore ready
for the next operation. Similarly to loading operations, unloading of tiles
(from spindle to a
frame) can be performed in a similar manner by the inverse path. It should be
noted that a tile
could also be sent to another brick, or to an area dedicated to specific
purposes like tile
disposal or tile incubation.
Although examples of brick processing is specific to the instrument and to the
valving
technology disclosed, it should be understood by those skilled in the art that
the same principle
could apply to systems employing passive valving systems, or to valving
systems based on
mechanical and electrical actuators, both in centripetal and non centripetal
environments.
Although the illustrative microfluidic tiles according to the invention are
construction
of a first and second substrate with a film layer in between, it should be
understood by those
skilled in the art that the microfluidic tiles of the invention can be formed
from a plurality of
CA 02576059 2007-02-05
WO 2006/043181 PCT/IB2005/003985
33
substrates. Likewise it should be understood by those skilled in the art that
the substrates can
be assembled with or without film layers in between.
Although the illustrative microfluidic tiles according to the invention are
utilized in
nano or ineso scale embodiments, it should be understood by those skilled in
the art that the
principle disclosed herein can be applied to fluid handling technologies
regardless of scale.
The principles, preferred embodiments and modes of operation of the presently
disclosed have been described in the foregoing specification. The presently
disclosed however,
is not to be construed as limited to the particular embodiments shown, as
these embodiments
are regarded as illustrious rather than restrictive. Moreover, variations and
changes may be
made by those skilled in the art without departing from the spirit and scope
of the instant
disclosure and disclosed herein and recited in the appended claims.
