Mems Final Paper

8/14/2019 Mems Final Paper

1/25

Macro, Micro, and

NanoLab-on-a-Chip Technology UnderDiffering Flow Regimes

Anthony Salvagno, Brittany Branch, Darin Leonhardt, Martin Donovan

12/17/2008

There are major differences when one compares fluid flow at different scales. Here we take acloser look at many of the various differences of each regime: macroscale (larger than 100 um),

microscale (100 um to 100 nm), and the nanoscale(fewer than 100 nm). We will also present

applications of each scale in order to demonstrate the usefulness of developing Lab-on-a Chip

technologies.


2/25

Introduction

The need for analysis of biological specimens utilizing as little sample as

possible has created an offshoot branch of MEMS (aptly named BioMEMS)

utilizing liquid flow on the very small scale. The research performed in this

field has created many advances in the technology and has allowed for the

creation of laboratory-on-a-chip (LOC) type setups. In the early stages of

development, these devices were constructed to handle chemical analysis on

the order of a lab. This was the creation of the first Micro Total Analysis

Systems (TAS).

The improvementof fluidics technologies allowed researchers to do so much

more on a single chip than just chemical sensing and analysis. Shortly after,

creation and development of pumps, valves, mixers, motors, and anything

else that could be used in a fluidics lab was miniaturized to the microscopic

scale and further enhanced fluidics [1-4]. People began to realize that this

could be applied to more than just chemistry, and looked to the ever

advancing field of genomics and microbiology for new applications.

During typical biological experiments (on the macroscale) large sample sizes

are used on the order of millions or even billions of cells, proteins, amino

acids, etc. On the smaller scales of micrometers and nanometers,

experiments may be carried out where single molecules can be analyzed and

characterized. In the field of genomics, for instance, this is particularly

useful because there is a gap in understanding how molecular interactions

affect gene expression [10].

The need to reduce cost and improve speed and efficiency were initially the

driving forces behind LOC technology. Eventually it became obvious that

higher resolution could be obtained because now one could analyze (down

to) individual molecule behavior. Reducing the need for such sample sizes,

as in the case of micro-/nanofluidic experiments, provides a much more cost

and effort efficient process.


3/25

The design of these tools even allows for high throughput processing, simply

because the entire chip is completely automated. Fluids can be moved,

separated, mixed, screened, etc. all on a single chip without ever needing tocontact human hands. The control gained from this automation is quite

impressive and may range from spatial and temporal characterizations on

the subcellular level [5] or organismal level [6], patterning of molecules and

cells [7], and passive and active cell handling and environment control [8] up

to the cellular level [9].

While most of the previously mentioned research has been done on the

microscopic level, there is as potentially impactful work going on at the

nanoscopic level too. The benefits here are similar to microfluidics in that

reduced cost of reagents, parallelization of experiment, and high resolution

(and sensitive) detection are all available technologies and could even be

more useful on the smaller nanoscale. The smaller structures of the

nanochannels (thus entering a completely different regime of physics) could

also present interesting interactions between the fabricated devices and the

molecules/particles to be studied.

It is pretty clear that these technologies haveits usefulness in the biologicalspectrum of science, but still we have not seen the explosion of technology

that LOC brings. Currently there is a slight gap between the biologists who

wish to employ the use of LOC technologies and the engineers who fabricate

them. This divide is ever narrowing as science becomes more open and

researchers themselves are pursuing interdisciplinary study. This

collaborative effort is rapidly changing the future of study for both fields and

will assuredly enable high-impact research.

Here, we present some of the technologies that have been and are being

developed in the fields of microfluidics and nanofluidics. Each regime is

dictated by its own set of parameters and will be discussed. We will provide

a compare/contrast of both and discuss the limitations that come with each

technology. We will also detail some of the technologies that were previously

(and still are) utilized in a biological setting, the uses for each, and the


4/25

limitations they provide. By discussing the current technologies, we hope to

demonstrate the usefulness of further developing both microfluidics and

nanofluidics to the point the LOC may be pushed to the forefront of current

research.

Background

Investigations into the complex and dynamic pathways governing the

function of organisms require the isolation and purification of specific

molecular populations from the myriad inhabiting the intracellular milieu. In

the biological sciences, a broad spectrum of separation technologies are

employed to both isolate individual components from the diverse andextensive mixtures comprising the cell, and to dissect the genome into

manageable segments for analysis or manipulation. Indeed, the field of

biology in general tends to makes its greatest advancements shortly

following the development of novel and powerful separation and isolation

techniques, e.g. electrophoresis, polymerase chain reaction (PCR), and

microarray technologies. Within the cell, the primary molecules to be

separated and analyzed are the deoxyribonucleic acids (DNAs), ribonucleic

acids (RNAs), and proteins; additional populations include segments of the

lipid bilayersof the plasma membrane, or carbohydrate sugars integral to the

cell signaling pathways that malfunction during pathological conditions.

Furthermore, entire cell populations are often also separated from one

another in tissues, e.g. immune response cells from the bone marrow, or

surfactant secreting cells from the lung. The following provides a brief

review of the more commonly employed separation techniques in the

biological sciences to separate and analyze DNA and proteins. It should be

noted that these separations are merely the means to an end, as once the

desired bio-molecule population has been isolated, further investigations

elucidating their structure, morphology, and function will be performed.

However, as these analytic techniques fall outside the scope of this paper,they will not be mentioned further.

Nucleic Acid Separation


5/25

Analytical techniques take advantage of various molecular properties

including size, shape, density, electrical charge, chemical structure or

combinations thereof, to isolate a small population of identical molecules

from all of the other types of molecules comprising the cell. One of the more

ubiquitous separation techniques found in molecular biology is gel

electrophoresis, which employs a combination of molecule size, shape and

electrical charge to separate a heterogeneous solution of nucleic acids into a

ladder of fragments sorted according to size. Due to their abundance of

phosphate groups comprising the backbone of the double helix, nucleic acids

possess an overall negative charge, and accordingly in an electric field they

will head toward the positive pole. By applying an electric field and allowing

the nucleic acids to migrate through a gel matrix comprised of an inert, jello-

like porous material, smaller-sized molecules will travel a greater distance

than their larger counterparts, and DNA fragments of the same size willcluster together in bands on the gel, hence exploiting the electrical charge of

these molecules to separate them according to size. By comparing the size

fragments of the sample DNA to the known size fragments of a control DNA

strand (referred to as the DNA ladder), the lengths, in base-pairs, of the

sample fragments may be determined. This technique is applicable to both

DNA and RNA molecules, although as RNA molecules are single stranded and

possess a tendency to assume a variety of three-dimensional configurations

via intramolecularhydrogen bonds, RNA is usually treated with a detergent

prior to separation, ensuring that only the length of the molecule, and not its

morphology, will factor into its migration rate through the gel. Typical length

scales over which the gel electrophoresis separations are performed range

from 10 20 cm, although resolution of many similar sized DNA fragments

increases with distance.

Protein Separation

In contrast to nucleic acids, which share an identical helical structure,

making them distinct only by their precise nucleotide (base-pair) sequence,

proteins possess discreet properties that make their separation different

from polypeptide to polypeptide. Protein separation techniques exploit

differences in protein size, morphology, charge and in many instances,

function. The primary method for protein separation is column

chromatography, where protein fractions are passed through long ( > 25 cm)


6/25

glass columns containing modified acrylamide or agarosebeads that separate

the proteins on the basis of different properties.

Ion exchange chromatography is a form of the aforementioned column

chromatography, where proteins are separated from a heterogeneousmixture on the basis of their surface ionic charge. The beads in the column

will be modified with chemical groups such that they will possess either an

overall positive of negative charge. Proteins containing ionicallycharged

amino acids will either be repelled or attracted to the column matrix, and as

each protein typically contains numerous acidic and basic amino acids,

multiple interactions will occur simultaneously within the same protein. On

the basis of the strength of these interactions, protein populations will elute

with distinct temporal profiles.

Another variation on column separation is gel filtration chromatography, also

known as size-exclusion chromatography, which exploits the size and shape

of proteins to separate similar molecular populations. As opposed to

possessing a chemical modification, the bead matrix for gel filtration

chromatography is fashioned to contain a variety of different sized pore

throughout their surface. When the heterogeneous protein mixture is passed

over this porous matrix, smaller sized proteins will enter the pores far more

often than their larger counterparts, thereby impeding their flow through the

column. Consequently, larger proteins will elute faster than smaller ones,again separating populations into distinct temporal profiles. An additional

column chromatography method employed with lesser frequency than the

previous two examples is metal binding chromatography. This capitalizes on

the ability of metal ions such as nickel, to bind with strong affinity to specific

amino acids, such as histidine. However, for this technique to be worthwhile,

the protein of interest would have to contain a significant amount of

histidineamino acids exposed on the protein surface, which is not typically

encountered.

While column separations are relatively simple and inexpensive methods of

protein fractionation, they lack adequate resolution such that a single

passage through one type of column will suffice for producing a

homogeneous population. Typically, a protein solution is passed through a

column multiple times, refining its yield, and then the sample solution (e.g.


7/25

the fraction containing the target protein population) is passed through a

different column, for further refinement. An example of this would be the

separation of a large, positively charged protein. The sample would first be

passed through a size exclusion column a few times, isolating a protein

population within the size range of the target. This mixture would then be

filtered through an ion exchange column containing a negatively charged

matrix. Thus, while a heterogeneous protein mixture may contain many

molecules that are similar in size to the protein of interest, the likelihood that

it contains multiple protein types of similar size and charge density is

minimal. However, multiple passages through multiple columns quickly

become time intensive, and therefore higher resolution separation

techniques are employed frequently for protein separations.

The problems of regular column chromatography are overcome using affinity

chromatography, commonly known as Immunoprecipitation, where the

surfaces of the beads composing the column matrix are coated with an

antibody specific for the protein of interest. With this technique, only the

desired protein will bind strongly, effectively separating them from the

homogeneous cell extract. This differs from the previous chromatographic

separation methods in that the specific binding targets of the protein must

be known ahead of time, which may not be possible when separating a novel

protein. Additionally, while highly specific, coating the beads with an

antibody can be expensive, and once coated the matrix may only be used for

one protein population.

In addition to the techniques that separate nucleic acids and proteins, there

also exist separation techniques that isolate multiple populations of

molecules to assay in vivoassociation and function. One technique employed

extensively in molecular biology is the co-immunoprecipitationassay, which

is used to investigate which proteins bind to, or are localized with, the

protein of interest in the cell. Based on affinity chromatography, a cellcolony is treated with formaldehyde to crosslink any associated proteins with

each other. The cell extract is then passed through beads coated with

antibodies for a protein, and additional proteins bound to it will also be

retained. Interaction between DNA and proteins are investigated with

chromatin Immunoprecipitation (ChIP). This is identical to co-

immunoprecipitation, except the protein that is targeted is typically a DNA


8/25

binding protein, and therefore has a segment of DNA crosslinkedto it. This

technique is commonly employed to determine where in the genome a

known DNA-binding protein is located.

While the separation techniques used to investigate intracellular structureand function are quite diverse with regards to the molecular properties

exploited to isolate a specific intracellular population, they all share a

common feature in that the length scales required to achieve an adequate

resolution reside in the macro-world, ranging from a few centimeters for

electrophoresis to over half a meter or more for size exclusion

chromatography. Additionally, macro-scale separations are not exclusive to

biology, and many common analytical techniques in the physical sciences,

including gas chromatography, HPLC, and mass spectrometry also require

the analyteto travel significant lengths for effective resolution. Accordingly,

the primary challenge encountered when developing separation techniques

on the micro-scale is overcoming the size requirement, meaning that these

miniaturized separation platforms must achieve comparable resolution to

their macro-scale counterparts while utilizing only a fraction of the distance.

This necessitates the development of novel techniques that exploit the

differences in physicochemical properties between the macro and

microscaleto compensate for the reduction in travel length.

Microfluidics

Flow devices at the micron and nanometer scale provide precise control of

fluids and chemical reactions in the fluid-phase. The high surface to volume

ratio at this small size facilitates rapid heat and mass transfer while taking

advantage of physical phenomena that do not normally have an influence at

the macro domain. An important feature of flow at such small transverse

lengths is that the flow occurs only at low Reynolds numbers due to the

dominant dependence of the fluid viscosity, whereas in larger flow systems

turbulent flow can occur further complicating the effects. With the

development of new fluidic devices there is a need to understand the

fundamental differences of fluid flow at these small transverse lengths. Flow

within a microchannel and nanochannel differ greatly in mechanism and

electrical double layer profiles.


9/25

Pressure driven flow and electro-osmotic driven flow are the two common

mechanisms of pumping liquid in microchannels, each are both beneficial

depending on the application. As the size of the channel decreases into the

nanometer range pressure driven flow is no longer feasible. For instance,

consider a pressure driven flow in a 100 nm channel where the average

velocity is on the order of mm/s. The pressure gradient required to achieve

this flow is as high as 3x109 Pa/m assuming a viscosity of 1x10-3kg/m s [18].

The integration of such a high-pressured pump into a nanoscale device

defeats the purpose of miniaturization, therefore electro-osmotic driven flow

is the primary flow mechanism used in nanoscale devices. Electro-osmotic

driven flow for both micro and nanofluidic devices requires an electrical

double layer at the channel wall. The double layer thickness and profile vary

for the different sized channels.

The first study of electro-kinetic flow in channels focused on the electrical

double layer (EDL) thickness. Two phenomenas occur in electro-kinetic flow,

electrophoresis and electro-osmosis.Electrophoresis describes the motion of

a charged surface submerged in a fluid under the action of an applied

electric field.Electro-osmosis refers to the bulk movement of a liquid past a

stationary solid surface, due to an externally applied electric field. Electro-

osmosis requires the existence of an electrical double layer at the solid-liquid

interface within the channel [19,20]. This charged double layer results from

an attraction between bound surface charges and ions in the passing fluid. It

is described by the Poisson equation:

o

e

=2

where

is the electrical field potential,

e

is the free charge density, and

o

and

are the dielectric constants in the vacuum and medium

respectively. The electrical double layer is illustrated in Figure 1.


10/25

Figure 1: Ionic distribution in an electrical double layer for a channel wall in contact with an aqueous

solution.

Immediately next to the charged wall of the channel is an immobile layer of

ions that are strongly attracted to the surface. This layer is called the

compact layer which is normally several Angstroms thick. The potential

distribution in this area is known as the zeta potential and is determined by

the geometric restrictions that the wall imposes on the ions, along with the

short-range interactions they have. From the compact layer into the neutral

bulk liquid the net charge density reduces to zero. This region is call the

diffuse layer because ions are less affected by the electrostatic charge of the

wall and therefore are mobile in the liquid. The ion and potential distributionin this region is further described by the Poisson-Boltzmann equation in

which the concentration of ions is predicted by the Boltzmann distribution:

2

=2zeno osinhze

kbT

Where z is the valence of the ion, n is the bulk ionic concentration, e is the

charge of a proton, kb is the Boltzmann constant, and T is the absolute

temperature [20]. Solving this equation with appropriate boundary conditions

results in the EDL potential field for a micro-sized channel illustrated in Figure

2.


11/25

The EDL in nano

sized channels

changes dramatically due to the fact that the size of the channel is

comparable to the thickness of the ELD [21]. In addition, the Boltzmann

distribution is no longer accurate since the EDL now affects the

concentration of ions in the bulk fluid. Figure 3 illustrates the overlapped ELD

resulting from the scaled down channel.

The differences in the EDL of micro and nanochannels have a significant

effect on the velocity profiles within the channel. The velocity within a

microchannel can be predicted by the steady state Navier-Stokes equation

for low Reynolds number of incompressible fluid:

2u=- eE

Where is the viscosity of the liquid, u is the velocity distribution,

e

is the

free charge density, and E is the externally applied field. If we assume the

double layer to be very thin compared to the channel width the Stokes

equation can be solved with

e

defined by the Poisson equation resulting in

the Helmholtz-Smoluchowski equation which predicts a plug flow velocity

profile [22].

ueo=ux=- o E

Conversely, the velocity profile for a nanochannel has a parabolic profile. The

velocity near the wall of the channel is slower than the bulk movement of the

liquid. This is due to the overlapping electrical double layer. COMSOL-FEMLAB

was used to model the velocity profile for both micro and nanochannels for

fluidic devices. Figure 4 shows the modeled plug flow velocity profile for

microfluidic channels.

Figure 2: The electrical double layer potential in a microchannel in contact with an aqueous


12/25

Figure 4: The velocity field measured in m/s in a microchannel.

Figure 5 shows the modeled parabolic velocity profile due to the overlapping

EDL for nanofluidic channels.

Fluid flow through small channels has enabled studies of single molecules

and cells. The benefits of this miniaturization include, but are not limited to,

direct manipulation and separation of proteins and nucleic acid along with

single cell analysis on a single chip [23]. Microfluidic devises are becoming

more prevalent with the emergence of biochemical lab-on-chip systems.

There has been successful use of these devices for cell sampling, cell

trapping and sorting, cell treatment, and cell analysis. Microfluidic-based

biological applications including polymerase chain reaction (PCR), DNA

separation, and DNA sequencing have been implemented [24]. Many

techniques in microfluidics have been used for separation including

magnetic, optical, mechanical, and electrical manipulation [25]. Currently

many of the separation techniques employed are the same techniques for

macro separations. There is an initiative to develop separation techniques

unique to microfluidic devices to improve sensitivity [25].

There has been great success in the development of microfluidic devices forseparation and analysis purposes. The envelope is being pushed further to

extend these systems down to the nanoscale. The nanoscale regime offers

unprecedented control of single cell and molecule transport, in manipulation

and separation as well as detection [25].

Nanofluidics

Figure 3: The electrical double layer potential in a nano-sized channel in contact with a aqueous


13/25

Lab-on-a-chip devices have been increasingly used in the last decade for

bioseparation,

detection, and chemicalsynthesis [26-29]. These devices have received growing attention because of

their speed, efficiency, reduced sample consumption, and detection

multiplexing. Many of these advantages led to commercial miniaturized total

analysis systems [30]. Following the first demonstration of such devices for

amino acid [31] and deoxyribonucleic acid (DNA) analysis [32], for instance,

microfluidic devices have emerged as a separation platform. More recently,

however, nanofluidic devices have been explored to achieve greater

efficiency, and lower sample diffusion than those observed in microscale

systems [33]. In particular, nanofluidic systems can provide enhanced

electro-osmotic flow control, compared to microfluidic systems [34], when

the surface charge on channel walls is manipulated in the presence of

significant double layer overlap [35].

This control scheme, where a third potential is applied to a gate electrode

surrounding the channel walls, is very much analogous to that of the field

effect transistors (FETs) used in complementary metal oxide silicon (CMOS)

technology. The fluidic devices that use this control scheme are hence

termed fluidic FETs [36]. Since the typical size of biomolecules, the Debyescreen length of electrolyte solutions containing the biomolecules, and the

size of nanochannels, can be at nearly the same length scale, it is expected

that the nanofluidic FETs provide enhanced manipulation of biomolecular

flow, leading to pronounced separation of biomolecules, while multiple

nanochannels (>104) in an array may simultaneously serve as an integrated

detector with increased sensitivity.

Various transport and electrokinetic phenomena in nanofluidic channels,

such as electrophoresis, electrostatic control, ion transport, and size and

shape effects, have been intensely studied [37,38]. Complementing these

studies, researchers at UNMs Chemical and Nuclear Engineering Department

have demonstrated an in situ analytical approach to monitor the average

flow speed and localization of molecules in nanochannels during separation

[37]. Their approach relies on an experimental platform where nanofluidic

Figure 5: The velocity field measured in m/s in a nanochannel.


14/25

channels are integrated into an infrared waveguide. The platform enables

multiple internal reflectionFourier transform infrared spectroscopy (MIR-FTIR)

to probe the signature vibrational peaks of molecules flowing through the

channels. The multi-purpose nanofluidic device is made by conventional Si

fabrication technology. The device architecture incorporates a heavily doped

Si gate that surrounds the channels to control the surface charge of

SiO2channel bottom and sidewalls. The device has been shown to

successfully control and monitor the flow of electrolyte solutions containing

fluorescent dye molecules by modulating the surface and to analyze the

interaction between the dye molecules and the internal SiO2channel surface,

using MIR-FTIR. The details of the device architecture will be describes

below.

Two main configurations are used with and without a heavily doped gate that

surrounds the nanochannels. The purpose is to compare the effectiveness of

the flow control, depending on the gate contact resistance. The following

fabrication sequence covers the gate based device, and is shown in Figure 6.

Double-side-polished Si (100) wafers are used as substrates to prevent the

scattering and loss of IR beam intensity during multiple internal reflectionin

the MIR-FTIR analysis system. Each wafer is diced into rectangular pieces,whose length, width, and thickness are 50, 10, and 0.7, respectively. The

rectangular samples are first immersed in a Piranha solution to remove

organics and other contaminants on the surface. Piranha is prepared by

mixing H2SO4 (2M) with H2O2 (30wt%) in 4:1 volumetric ratio. Piranha is a

strong oxidant that forms a thin layer of chemical SiO2on Si. The chemical

oxide is subsequently removed in dilute HF (Fig. 6(a)), for which 48 wt % HF

Figure 6: Schematic flow diagram of fabrication steps for integrating nanochannels into a MIR wa


15/25


16/25

of the nanochannels. Plastic wells are attached

to the through holes to facilitate the injection of

electrolyte solution containing fluorescent dye

molecules (Fig.

8). Fig. 7(c)

shows an SEM

image of the

Pyrex sealed nanochannel array. Note that

since doped Si surrounds only channel bottom

and sidewalls, the gate bias controls surface

charge only on channel bottom and sidewalls

but not on the channel top Pyrex surface. The

width and depth of each channel are

approximately 100 and 400-500 nm,respectively. Therefore, a 3 mm w x 10 mm l

nanochannel section contains well over 8000

nanochannels, and if desired, this number can

be increased to 105over the 1 cm width of the

MIR crystal. The device is referred to as a

nanofluidic waveguide, since the device serves

simultaneously as a separation matrix and as

an analytical tool for MIR-FTIRS. Fig. 8shows conceptually how the

nanofluidic waveguide is operated. For MIR-FTIRS analysis, a focused IR

beam enters one of the beveled edges of the Si MIR crystal and makes

approximately 35 top reflections before the beam exits the opposite end. A

HgCdTedetector collects the IR signal leaving the second beveled edge. Due

to these multiple reflections, the Si MIR crystal is opaque to IR below 1500

cm-1. Note that the width of the channels is much less than the mid-IR

wavelength (2.5 to 15 m). As the channel width increases to the

micrometer level, the IR beam scatters substantially upon each internal

reflection, and very little IR passes through the MIR crystal.

To monitor the electro-osmotic (EO) flow in nanochannels with MIR-FTIR and

with laser-scanning confocal fluorescence microscopy (LS-CFM), the

nanochannels are first completely filled with a desired buffer solution by

capillary force. The LS-CFM technique is considered a standard optical

technique to monitor the flow of fluorescent dye molecules in nanochannels

and also used to provide a baseline comparison with the MIR-FTIR technique.

Figure 7: Cross-sectional SEM images of nanochannels. (a) Nanochanneplasma etching. (b) SiO2 covered nanochannelsafter thermal oxidatio

The channelwidthis less than 100 nm, and the depth is 400 to 500 nm.

Nanochannels sealed with a Pyrex cover after anodic bonding.


17/25

Since the depth of focus of the LS-CFM technique is limited to 1um, the

optical images are averaged over the entire channel depth. A platinum wire

is then inserted into each well as an electrode. A positive potential is applied

to one well, while grounding the other (Fig. 8). This potential difference

creates a longitudinal electric field (E) along the channels and induces EU

flow. While the EO flow is in motion, fluorescent dye molecules pre-dissolved

in identical buffer solution are injected into the positively biased well. The

gate potential (VG) is then applied to control the surface charge on the

channel walls. The motion of the fluorescent dye molecules is observed

under EO flow and upon modulating VG, using both MIR-FTIR and LS-CFM for

purpose of comparison.

Fig. 8 Experimental setup to monitor flow control and segregation of dye

molecules within nanochannels in response to the gate bias, using MIR-FTIR.

FET flow control experiments are conducted with Rhodamine B (MW = 479),

C28H31ClN2O3, dissolved in a pH 4 buffer solution previously described. After

filling the nanochannels with only buffer, the EO flow is initiated by applying

+6 V (VEO) to one of the wells, while the opposite well is grounded, and the

gate is left floating (Fig.3). In this case, the gate is heavily doped to

minimize contact resistance. Fig. 9shows LS-CFM images of Rhodamine B

under FET control. Note that within the field of view, approximately 5000

channels are present, extending horizontally in Fig. 9, as the dye molecules

traverse farther into the channels. The dimensional uniformity amongst the

channels will have to be improved to achieve a uniform front.


18/25

Fig. 9 LS-CFM images of Rhodamine B in a pH 4 buffer solution under EO flow,

with and without the gate bias. (a)-(b) show the accelerationofEO flow with

negative gate bias. (c)-(d) show the reversed EO flow under positive gate bias.

In fig. 9, the dotted lines mark the

reference point, and the arrows

mark the direction and magnitude

of the dye flow. The observed EO

flow velocity is 2 m s-1. When anegative gate bias (VG = -30 V) is

applied to produce a large

negative z-potential, the flow

velocity is increased to 6 m s-1.

In Fig. 9(a)-(b), RhodamineB

moves from

right to left

at an

accelerated

pace.

Conversely, Fig. 9(c)-(d) shows that Rhodamine B rapidly reverses its flow at

8 m s-1when +30 V is applied to the gate. The flow response to the change

in the gate bias is immediate and repeatable. The observed flow response is

also independent of the position of the dye molecules with respect to the

gate position.

Figure 10: A time series IR absorbance spectra of RhodamineB in D2O taken

FET flow control. The background spectrum is taken after filling the

nanochannelswith D2O buffer solution. The sample spectra are taken eve

min, while Rhodamine B flows into the nanochannels be electroosmosis.


19/25

From the results, the flow control with Rhodamine B shows a limitation when

the flow is reversed (Fig. 9(d)). Most of the Rhodamine B is observed to

reverse its flow direction, marked by the solid arrow, while the left mostoutlineremains at the same position. This observation indicates that

potentially two layers of Rhodamine B exist in the nanochannels because

Rhodamine B is positively charged in the

H range below 6.0, while the surface of

the nanochannels is negatively charged.

Its hypothesized that an inner layer, away

from the channel walls, is controllable,

while an outer layer near the channel

walls in not, due to electrostatic

interaction between the positively

charged dye molecules and negatively

charged channel walls.

The hypothesis is supported by MIR-FTIR

analysis during FET flow control. The IR

background spectrum is recoded after

completely filling thee channels and

inlet/outlet wells with the buffer solutionand injecting Rhodamine B into the

positive well. While the dye molecules

flow into the nanochannels due to EO flow

and respond to the gate bias, the sample

spectra are recorded every 5 min with 2 cm-1 resolution (Fig. 10). Each

spectrum is averaged over 50 scans to achieve good signal-to-noise- ratio,

while maintaining a reasonable time resolution. Fig. 11(a) shows

representative IR absorbance spectra extracted from Fig. 10, taken during EO

for different gate voltages.


20/25

The focus is made on the COO- stretching mode of Rhodamine B at 1590-

1600 cm-1,

which is its

strongest

vibrational

mode [40].

The increase or decrease of the intensity of the COO-peak represents: a

convolution of an increasing amount of Rhodamine B filling the channels

during EO flow and upon applying a negative gate bias; increasing

concentration of Rhodamine B near the channel bottoms and sidewalls as an

electrostatic response to negative gate bias; decreasing amount of

Rhodamine B during reverse flow with a positive gate bias; and decreasing

concentration of Rhodamine B near the channel bottoms and sidewall as an

electrostatic response to positive gate bias.

The device architecture of an array of nanochannels integrated into a MIR

infrared waveguide with FET flow control was described. The device provides

the capability to simultaneously monitor flow control and probe molecule-

wall interactions, using LS-CFM and MIR-FTIRS. Flow control is demonstrated

using Rhodamine B by changing the gate bias to accelerate and/or reverse

the flow in the nanochannels. The implication of this charge-dependent

molecule-wall interaction is that the polarity and magnitude of channel

surface charge significantly influences the mobility of charged molecules in

nanochannels and that this mobility control can be used as an additional

mechanism to separate the molecules. Further investigations of the impact

of molecule-wall interactions, using fluorescent dye molecules of similar

molecular weight with different charges is needed as a function of pH.

Lab-on-a-Chip Integration

Figure 11: MIR-FTIR analysis of flow control, segregation, and adsorption/deso

of Rhodamine B in pH 4 buffer within nanochannels. The dye molecule is

positively charged at pH 4. (a) A representative set of IR spectra taken dur

flow in response to the gate bias. (b) Changes in the integrated absorbanc

COO- peak in response to the EO flow and to the gate bias.


21/25

Standard microfluidics and

nanofluidics experiments utilize

flow through channels (of their

respective sizes) to conduct an

experiments whether it is for

analysis, characterization, force

spectroscopy, or other. With

most cases, only a few (if even

more than one) tests are

performed. With the advent of the integrated circuit, one can conduct

numerous tests on a single chip. This was the precursor to todays LOC.

Low fluid volume consumption means less waste, lower reagent cost, and asmaller required sample volume for investigation. Small volumes also

provide the integration of much functionality, in that a single chip can

incorporate a bunch of different diagnostics. High surface to volume ratios,

small heat capacities, and short diffusion distances also lead to faster

analysis. These factors also allow for real time data tracking which is

conduciveto enhanced process control. Because of the small stature of these

devices (another benefit of the LOC), experiments can be run parallel for

high throughput analysis. The low fabrication costs of such technology also

allow for mass production and disposable chips.

No technology is without its disadvantages, however. Because these

concepts are just the beginning, they are not fully developed and there is still

much room for improvement. The physical and chemical effects that are

more dominant at such small scales also make certain processes more

complex to fabricate (and operate) than would be the case in conventional

lab equipment. Another problem could arise from the scaling down of

features and detection principles, especially at the nanoscale. While there

are always possibilities to work around this, in some cases the effort may notbe enough and could render a less-than-effective or inefficient product.

Figure 12: Concept of lab-on-a-chip technology pertaining to DNA


22/25

Conclusion

Typical biological experiments rely on assays that tend to rely on the quantity

of components used for more of an average value instead of a specific

characteristic per molecule (particle or whatever is being analyzed). Byperforming tests on the microscale or even the nanoscale, there can be a

more efficient use of the materials needed for a typical experiment. To

induce further persuasion, it can be said that each regime has its own

advantages and each can provide the backdrop for applications and

techniques that have yet to be fathomed.

Microfluidics for biology has its advantages because it is a somewhat more

mature domain of study. Techniques useful for today and in the future have

been well documented, fabrication methods can be low cost, and automation

allows for total control of the environment. The size of channels can becreated to accompany small particles or even large cells for complete

analysis, and the difference in flow effects at this scale make this possible.

There is a little more discrepancy on the nanofluidics side. The concepts and

techniques used here are less developed and slightly harder to control.

Since this realm is on a smaller length scale (nanometers vsmicrometers),

there is a completely different world of effects to take into account. This

makes some of the more traditional approaches difficult to employ. In the

same way, however, this may facilitate other processes far better than can

be achieved on the microscale.

As of now, the lab-on-a-chip is a novel concept in its infancy. While the

current designs have their purpose, there is still a lot to be gained by utilizing

the full potential of the subject. By developing microfluidic and nanofluidic

technologies (and even the more macroscopic techniques), we can rapidly

increase theuse, effectiveness, and efficiency of LOCs. This can only be

attained with further interdisciplinary and collaborative efforts between the

biologists who employ the technology and the engineers who fabricate it.


23/25

References

1. H.T.G. Van Lintel, F.C.M. Vandepol and S. Bouwatra, Sens. Actuators,

1988, 15(2), 153-167.

2. S. Shoji, M. Esashi, and T. Matsuo, Sens. Actuators, 1988, 14(2), 101-

107.

3. N.T. Nguyen and Z. Wu,J. Micromech. Microeng, 2005, 14, R1-R16.

4. D.J. Laser and J.G. Santiago,J. Micromech. Microeng, 2004, 14, R35-

R64.

5. A. Sawaro, S. Takayama, M. Matsuda, A. Miyawaki, Dev Cell, 2002, 3(2),

245-257.

6. E.M. Lucchetta, M.S. Munson, Ismagilov, Lab Chip, 2006, 6, 185-190.

7. A. Folch, M Toner,Annu Rev Biomed Eng, 2000, 2, 227-256.

8. G.M. Walker, D.J. Beebe, Lab Chip, 2002, 2, 131-134.

9. C.S. Chen, M. Mrksich, S. Huang, G.M. Whitesides, D.E. Ingber, Science,

1997, 276, 1425-1428.

10. R. Arjun, A. van Oudenaarden, Cell, 2008, 135(2), 216-226.

11. D. Mijatovic, J.C.T. Eijkel, A. van den Berg, Lab Chip, 2005, 5, 492-500.

12. S. Haeberleand, R. Zengerle, Lab Chip, 2007, 7, 1094-1110.

13. A.L. Paguirigan, D.J. Beebe, Bio Essays, 2008, 30, 811-821.

14. G.M. Whitesides, Nature, 2006, 442, 368-373.

15. P.S. Dittrich, A. Manz, Nature Reviews, 2006, 5, 210-218.

16. H. Becker, Med Device Technology, 2008, 19(3), 21-24.

17. M.A. Burns et al., Science, 1998, 282, 484-487.

18.Jay Taylor, C.L. Ren, Microfluidic Nanofluidic, 2005, 1, 356-363.

19.W.B.J. Zimmerman, Microfluidics: History, Theory and Applications,

(2006) 49-100.20. Dongquig Li, Electrokinetics in Microfluidics, (2004) 7-28.

21.J.C.T Eijkel, A.V.D Berg, Nanofluidics: what is it and what can we expect

from it?, Microfluidic Nanofluidic, 1 (2005) 249-267.


24/25

22.T.S. Hug, N.F. Rooij, U. Staufer, Fabrication and Electro-osmotic Flow

Measurements in Microchannel Flow and Nanofluidics Channels, 2

(2006) 117-124.

23. C. Yi, C.W. Li, S. Ji, M. Yang, Microfluidics Technology for Manipulationand Analysis of Biological Cells, Analytica Chimica Acta, 560 (2006) 1-

23.

24. M. Kumemura, D. Collard, C. Yamahata, N. Sakaki, G. Hashiguchi and H.

Fujita, Single DNA Molecule Isolation and Trapping in a Microfluidic

Device, Chem Phys Chem, 8 (2007) 1875-1880.

25. Abgrall, A-M Gue, Lab-on-chip technologies: making a microfluidic

network and coupling it into a complete microsystem, Journal of

Micromechanics and Microengineering, 17 (2007) 15-49.

26. A. Daridon, M. Sequeira, G. Pennarun-Thomas, H. Dirac, J. P. Krog, P.Gravesen, J. Lichtenberg, D. Diamond, E. Verpoorte and N. F. de Rooij,

Sens. Actuators, B, 2001, 76, 235-243

27. A. van den Berg and T. S. J. Lammerink, Top. Curr. Chem., 1998, 194,

21-49.

28. S. Liu, Y. Shi, W. W. Ja and R. A. Mathies,Anal. Chem., 1999, 71, 566-

573.

29. P. C. Simpson, D. Roach, A. T. Wooley, T. Thorsen, R. Johnston, G. F.

Sensabaugh and R. A. Mathias, Proc. Natl. Acad. Sci. U. S. A., 1998, 95,2256-2261.

30. A. Manz, n. Graber and M. H. Widmer, Sens. Actuators, B, 1990, 1, 244-

248.

31. D. J. Harrison, K. Fluri, K. Seiler, Z. Fan, C. S. Effenhauser and A. Manz,

Science, 1993, 261, 895-897.

32. R. A. Mathies, P. C. Simpson and A. T. Wooley, in DNA Analysis with

Capillary Array Electrophoresis Microplates, Micro Total AnalysisSystems, ed. D. J. Harrison and A. van den Berg, Kluwer academic

Publishers, Banff, Canada, 1998, pp. 1-7.

33. G. Grandi, TrendsBiotechnol., 2001, 19, 181-188

34. R. B. M. Schasfoort, S. Schlautmann, L. Hendrikse and A. van den Berg,

Science, 1999, 286, 942-945.


25/25

35. R. Fan, R. Karnik, M. Yue, D. Li, A. Majumdar and P. Yang, Nano Lett.,

2005, 5, 1633-1637.

36. K. Ghowsi and R. J. Gale,J. Chromatogr., 1991, 559, 95-101.

37.Y. J. Oh, T. C. Gamble, D. Leonhardt, C. H. Chung, S. R. J. Brueck, C. F.

ivory, G. P. Lopez, D. N. Petsev, and S. M. Han, Lab Chip, 2008, 8, 251-

258.

38. R. Karnik, K. Castelino, R. Fan, P. Yang and A. Majumdar, Nano Lett., 5,

1638-1642.

39. K. F. Lei, S. Ahsan, N. Budraa, W. J. Li and J. D. Mai, Sens. Actuators, A,

2004, 114, 340-346

40. L. F. Johnson, G. W. Kammlott and K. A. Ingersoll,Appl. Opt., 1978, 17,

1165-1181

41. M. Pospisil, P. Capkova, H. Weissmannova, Z. Klika, M. Trchova, M. M.

Chmielova and Z. Weiss,J. Mol. Model., 2003, 9, 39-46.

Mems Final Paper

Documents