Anal. Chem.
2005, 77, 3887-3894
Advances in Polymerase Chain Reaction on
Microfluidic Chips
Michael G. Roper,† Christopher J. Easley,† and James P. Landers*,†,‡
Department of Chemistry, University of Virginia, Charlottesville, Virginia 22904, and Department of Pathology, University of
Virginia Health Science Center, Charlottesville, Virginia 22908
Review Contents
Background
Genome Era
Microfluidic Heating Methods
Microchip PCR Background
Scoring Method for Device Comparison
PCR Reaction Volume
High-Speed PCR
Future Strategies
Potential Improvements
Real-Time PCR
Whole Genome Amplification
Conclusions
Literature Cited
BACKGROUND
Genome Era. The mapping and sequencing of the human
genome represents a monumental accomplishment that will
impact medicine and biology in ways that we are only beginning
to comprehend. The successful sequencing reported in 2001 (1)
ushered in what has been dubbed the “postgenome era”. While
this implies that the genome era is over, it is, in fact, just getting
underway in so many ways. A map of the more than 3 billion bases
and their corresponding 30 000+ genes will not only be critical
in fleshing out the structure and function of these genes but will
leverage the discovery of new genes and a newfound understand-
ing of the potential roles the gene products play in health and
disease. Genetic analytical methods have been, and will continue
to be, critical to such discovery, providing the means to dissect
gene structure and gene function. These methods have almost
invariably involved two general processes: sample preparation and
separation/detection.
Capillary electrophoresis (CE), a means for separating DNA
fragments in a manner that allows for decoding DNA sequence,
has supplanted the tedious, time-consuming methods associated
with sequencing slab gels. As a consequence of this, the role of
CE in accelerating the sequencing of the genome is well-
recognized (2). More recently, there is a palpable interest in the
next paradigm shiftshigh-resolution separations on microfabri-
cated devices that provide comparable resolution to CE separa-
tions but with enhancements in speed, throughput capabilities,
and cost-effectiveness (3). One example is the current interest
from the forensics arena to carry out short tandem repeat typing
* To whom correspondence should be addressed. Phone: (434)-243-8658.
Fax: (434)-924-3048. E-mail: landers@virginia.edu.
† University of Virginia.
‡ University of Virginia Health Science Center.
10.1021/ac050756m CCC: $30.25 © 2005 American Chemical Society
Published on Web 05/19/2005
on high-throughput microdevices as a means of reducing the
3887 backlog of samples awaiting analysis in the United States (4).
3887 Advances in sample preparation, at least those akin to what
3888 microfabricated devices have brought to separations (enhanced
3888 speed and throughput, decreased cost), are more difficult to find.
3888 Sample preparation involves three general processes that, when
3889 carried out sequentially, reduce samples of diverse origin and
3890 character to a form amenable to separation. These processes often
3891 include a cell sorting step (to isolate the cells of interest from a
3891
mixture) but almost universally involve the extraction of DNA from
3891
the sample matrix, followed by DNA amplification via the poly-
3891
3892
merase chain reaction (PCR). The role of the PCR is obvious: to
3892 create adequate copies of select DNA sequences so that these
can be detected by separation. The role of the purification step is
simple: to rid the sample of any species that could potentially
interfere with a successful and robust amplification by the PCR
process. Pre-PCR steps serve the sole function of presenting a
DNA-containing sample that is amenable to optimal amplification,
while post-PCR separation serves to display the product(s) yielded
from the PCR process. From this perspective, one can easily argue
that PCR is the keystone of genetic analysis.
The PCR is an enzyme-mediated process where, in response
to temperature cycling, a single DNA molecule can be rapidly
amplified into many billions of molecules. PCR was introduced in
1985 by Mullis and co-workers (5) where he described the
transitioning of a PCR mixture (bacterial polymerase, dNTPs,
template DNA, and primers) through different temperatures that
facilitated the denaturation of the template DNA (94 °C), the
annealing of primers to complimentary sequences (e.g., 55 °C),
and the extension of complimentary sequence from the primer
(72 °C). Repeated cycling through these temperatures resulted
in an exponential increase in the number of copies of a specific
DNA sequence (in size up to ∼1000 base pairs, bp) relative to
the original number of DNA template copies. The ability to create
millions to billions of copies from a single DNA template has
revolutionized the field of molecular biology, supported by the
fact that this pioneering work earned Mullis a portion of the 1993
Nobel Prize in Chemistry.
Since its conception, the PCR has become one of the main
laboratory tools for scientists in the life sciences. An effective PCR-
based amplification is now an indispensable part of genetic
analysis, whether it involves molecular biological-, diagnostic-,
forensic-, or agricultural-related analysis. PCR is nearing its quarter
century anniversary and plays a key role in the evolution of newer
and more powerful molecular biological techniques; and the
continuous morphing of molecular technology is, indeed, real.
Despite the power of PCR technology for routine diagnostics, there
Analytical Chemistry, Vol. 77, No. 12, June 15, 2005 3887
exists trepidation among some diagnostic laboratories to adopt
new molecular methods. Some speculate that it is the ever-
morphing nature of molecular biological techniques that precludes
adoption by some diagnostic laboratories, primarily because of
concerns that the technology will rapidly become outdated (6).
The conventional format of PCR involves the temperature
cycling of the PCR components in thin-walled, plastic tubes in a
temperature-controlled metal block between the desired temper-
atures to effect the amplification of a specific DNA sequence. The
amplified volumes are typically in the 2-50-µL range with cycling
speeds (the time to traverse through one denature/anneal/extend
temperature cycle) on the order of 20-40 cycles are typically
carried out. Thermal mass is the key to fast thermocycling, and
with conventional PCR instrumentation, this resides in the heating
block used to control solution temperature. If we look to microchip
technology to do for PCR what it has done for electrophoretic
separations, then the volume of PCR mixture thermocycled will
be reduced by 1-2 orders of magnitude, and the speed with which
thermocycling can occur will be increased, not only because the
volumes involved are reduced but because thermal mass should
be as well.
Microfluidic Heating Methods. Temperature cycling on
microfluidic devices can be performed either with contact heating
methods or noncontact heating methods. Here, we define contact
heating methods as having heaters fabricated within the microchip
or in thermal contact with the outside the microchip; with both,
the common denominator is that the thermal mass of the
microfluidic device is in contact with the heating element. Not
surprisingly, noncontact methods are defined as using a heating
method that is remote from the microfluidic device and not in
physical contact with the PCR chamber. While our group has
focused extensively on noncontact heating methods for PCR
amplification to reduce the thermal mass of the components to
be cycled (7) and, hence, achieve faster thermocycling rates, the
majority of papers published during 2003-2005 rely on contact
heating. For this reason, the rest of this review will devote its
attention to contact heating methods.
The benefits of microfluidics have always been the ability to
integrate multiple functionalities on one device, decrease analysis
time, and minimize dilution. Therefore, an ideal integrated genetic
analysis system will accept a complex biological fluid (e.g., blood,
serum, or semen), remove PCR inhibitors, amplify the DNA, and,
finally, separate and detect the amplified products (amplicons) of
interest. Due to the multiple steps in this sequence, a short process
time for each step is ideal, which can be achieved in PCR via fast
cycling times, which favors reduced volumes (and therefore
thermal mass) as mentioned previously. Therefore, this review is
divided into two sections focused on what we believe to be two of
the more significant parameters in chip PCR: the PCR reaction
volume and the speed of amplification.
Chemical Abstract Service was searched from 2003 to 2005
including the keywords “microfluidic AND amplification”, “mi-
crofluidic AND reaction”, “lab-on-a-chip AND amplification”, “lab-
on-a-chip AND reaction”. This review is not meant to be a
comprehensive summary of all literature during this time period,
only a cursory review and summary of notable achievements of
fast PCR and low-volume PCR during this era. Table 1 is a
summary of the literature cited in the text and also lists literature
3888 Analytical Chemistry, Vol. 77, No. 12, June 15, 2005
involving PCR on microfluidic devices that are not mentioned in
the text. We refer the readers to other recent reviews for a further
survey of the literature (3, 8-12).
Microchip PCR Background. It is noteworthy to summarize
the initial developments in microchip PCR before outlining the
recent literature. The first report of translating PCR to a chamber
in a microchip device was by Wilding et al. (12), who used a
silicon/glass hybrid device that held 5-10 µL of reaction mixture
and was heated using an external copper block heater. The same
group then emphasized the need for surface passivation for
microchip PCR, and they determined that silanization of the
microchamber surface provided the best results (13). Later, Burns
and co-workers (14) created an integrated microchip including a
resistive heating region with temperature sensors for DNA
amplification, a sample loading region, and a gel-based separation
region; this work demonstrated the ability to integrate single-
strand displacement assay with other sample analysis steps on
one device. Kopp et al. (15) exploited the geometric freedom of
microchip design by developing the first flow-through PCR
microdevice. By simply using the microchip reservoir as the
reaction chamber, Waters and co-workers (16) performed cell
lysis, PCR amplification, and electrophoretic sizing on a single
microchip device by placing the entire chip into a commercial
thermocycler. The first report of noncontact PCR on microchips
was by our group (7e) using a polyimide substrate and cycling
through selective infrared heating and air cooling. Lagally et al.
(18) developed a device with microfabricated resistance heaters,
resistance temperature sensors, and PCR chambers seamlessly
connected to electrophoretic separation channels, including valves
and hydrophobic vents. These publications have been seminal to
the development of microchip PCR, and as such, a few were
included in Table 1 for the purpose of comparison along with
others (7e, 15, 16, 19-30, 33, 36-42).
Scoring Method for Device Comparison. Throughout this
review, we found it necessary to attempt a normalization of the
various microfluidic platforms for performing PCR to enable a
better comparison between the designs. To perform this normal-
ization, three major facets of PCR amplification were included.
First, we calculated the average experimental rate at which PCR
was occurring from the amplicon size and reaction time. This term
therefore included not only the kinetic rate of nucleotide incor-
poration by Taq polymerase but also the effects of general reaction
conditions such as passivation of the substrate. Second, the volume
of the PCR chamber was used as a measure of the “cost” of
amplification. For instance, a small volume would reduce the
overall cost of the system by allowing for lower amounts of
reagents to be used. Third, the limit of detection of the experiment
was reflected in the score. This last parameter was included
because the reaction rate (as defined in the first facet) will be
dependent on the copy number used. In this way, a device with
a slow reaction rate could still receive a relatively high score if
the limit of detection is good. To incorporate these parameters,
the formula for the score was calculated as follows:
bp
Score )
Vol × t × CN
where bp is the number of base pairs of the amplified fragment,
Vol is the volume of the PCR domain, t is time for amplification,
Table 1. Abbreviated List of the PCR Performed within Microfluidic Devices

amplicon grade (bp

vol amplification LOD or amount length µL-1 s-1
ref (nL) time (s) template DNA of DNA used (bp) copy-1)
Low Volume
19 3 3600 1800-bp cDNA LOD ) 60 copies 294 0.45
20 29 5600 genomic Salmonella LOD ) 6 copies 559 0.57
21 40 >3400 genomic DNA LOD ) 0.4 copies 295
22 200 1200 intact E. coli cells LOD ) 2-3 cells 280 0.58
23 200 420 Topo PCR 2.1 vector 15 ng 127
3.5 × 109 copiesa
24 200 600 genomic DNA LOD ) 10 copies 150 0.12
25 240 1860 mouse DNA snrpn template n/ab
Fast PCR
26 total ) 2500 102 λ-phage DNA LOD ) 2 × 107 copies 500 9.8 × 10-8
at 10 mm/s
27 2000 360 1.02-kbp genomic DNA 2.5 × 106 copies 230
Others
15 240 1030c 106-bp M. tuberculosis DNA 9.6 pg, 8.8 × 107 copiesa 106
16 10 000 90 with 72.9 nL/s 1000-bp N. gonorrheae DNA 108 copies 176
7e 1700 240 λ-DNA 6.9 × 104 copies 500
28 500 600 λ-DNA 5 pg, 9.5 × 104 copiesa 500
29 15 000 1800 3.9-kbp genomic DNA 6 ng, 1.4 × 109 copiesa 295
30 119 000 4380 B. cereus DNA 1.6 µg, 2.9 × 108 copiesa 700
33 280 900 2686-bp M13/pUC19 LOD ) 1 copy 136 0.54
cloning vector
36 10 000 780 280-bp genomic DNA 1.56 × 105 copies 230
37 20 000 5400 intact E. coli cells 1000 E. coli in 1 mL blood 221
1 × 103 copiesa
38 15 000 5940 intact white blood cells 10000 cells, 2 × 105 copiesa 379
39 3500 3390 Vero E6 cells and nasal swabs n/a 438
40 3000 4680 genomic DNA 6 × 104-3 × 105 ng 350
2 × 107-9.8 × 107 copiesa
41 8000 4200 Topo PCR 2.1 vector LOD ∼500 copies 600
42 12 8280 λ-DNA 24 pg, 4.8 × 105 copiesa 199
a Indicates our calculation based on molecular weight of template DNA. b n/a, not available. c Single strand displacement assay.

and CN is the copy number at the limit of detection. As can be
seen, bp/t represents the average experimental rate of the
reaction, while the volume and copy number at the limit of
detection are also included. Thus, a high score will occur if a large
amplicon is generated from a small-volume, high-speed, low-copy
number assay (an example of an “ideal” high score would be
generation of a 200-bp amplicon from a single copy of template
DNA in a 10-nL reactor in 90 ssall numbers that have been
achieved at least individually in chip-based PCRswould provide
a score of 222). We are aware that this is a simplistic formula;
however, it has helped us to provide a more clear comparison
between these PCR chips (discussed in the text and shown in
Table 1), and we hope that it will also help the readers. All scores
were based on parameters at the limit of detection, and if no limit
of detection was given, no score was assigned. Additionally, the
scoring method assumes similar PCR reagent concentrations
(other than DNA). In the conclusion of this review, we address
issues of “weighting” the various parameters in an applications-
dependent manner.
PCR REACTION VOLUME
With microfabricated devices come the ability to react ex-
tremely small volumes of solutionsPCR is not excluded from this
quest. The smallest PCR reaction volume reported to-date comes
from work in the Quake group (19). PCR in a volume of 3 nL
reduces, by almost an order of magnitude, the volume required
by the second lowest report. This volume of solution is amenable
to upstream and downstream components of a completely inte-
grated genetic analysis system, and their use of hydraulic valves
and pneumatic pumps makes integration facile. An added benefit
of this design was the reduction in the amount of manual pipetting
needed to address an N × N matrix of reaction centers from 3N2
to 2N + 1, thereby reducing the possibility of contamination and
errors. Although this work set the lower limit of reaction volume,
speed of analysis was not the major focus and thermocycling was
performed by placing the entire device on a conventional ther-
mocycler. Thus, even with a dramatically reduced PCR reaction
volume, one that would be amenable to extremely fast thermocy-
cling, 1 h was still required for amplification of a 294-bp fragment
of the human β-actin cDNA fragment. With an LOD of 60 template
copies, the calculated score for this work was 0.45 bp µL-1 s-1
copy-1.
A 29-nL PCR reactor was demonstrated on a poly(cyclic olefin)
microfluidic chip that used paraffin wax valves to contain the PCR
mixture followed by electrophoretic separation of the amplified
product (20). Silver/graphite inks were printed onto the plastic
substrate prior to bonding to be used as both the heaters and
electrophoretic contacts, and a resistive temperature detector
(RTD) was used to sense the temperature during PCR amplifica-
tion. Limit of detection for this device was shown to be six copies
for both purified Escherichia coli and Salmonella DNA per 29-nL
reaction vessel. Heating rates attained were 12 °C/s, and cooling
rates were 2 °C/s. Although it was stated that PCR could be
performed with shorter hold times, 80 min was still required for
amplification of DNA, thus not providing much advantage in speed.
With this favorable LOD and low reaction volume a very high
score could have been obtained with more optimization of reaction
Analytical Chemistry, Vol. 77, No. 12, June 15, 2005 3889
time, but with the 80-min reaction time, the work registered a
score of 0.57 bp µL-1 s-1 copy-1.
Matsubara et al. (21) performed PCR amplification in 40-nL
reaction vessels. Forty nanoliters of a solution containing primers
and a fluorescent probe were first loaded into the vessels and
allowed to dry, and 200 µL of oil was applied to avoid evaporation.
Then, 40 nL of PCR mixture containing human genomic DNA
minus primers was added under the oil to each well. The entire
chip was then placed on a conventional thermocycler (amplifica-
tion time was 1 h) and successful amplification determined by
detecting fluorescence from the probe using a charge-coupled
device. A 0.4 DNA copy number was reported for their limit of
detection, but it is important to note that only 2 out of 60 vessels
had detectable amounts of DNA at this concentration. The use of
oil as a cover for their system is a distinct disadvantage and raises
questions as to the applicability for integrated genetic analysis
systems. Due to the LOD discrepancy, a score was not reported
for this work. Again a similar theme is observed, where the volume
and LOD were particularly low, but the reaction time gave no
improvement over conventional PCR.
Other reports have described PCR with volumes of 200 nL or
higher. Lagally et al. (22) have set the bar with a portable genetic
system that uses an 20.3 cm × 25.4 cm × 30.5 cm box for housing
all optics and electronics for a microdevice that performs PCR
and electrophoretic separation with a 200-nL PCR chamber. Limit
of detection for their system was 2-3 E. coli cells with amplifica-
tion performed in ∼20 min. With intermediate values of volume
and reaction time combined with a good LOD, this work scored
the highest among those reviewed at 0.58 bp µL-1 s-1 copy-1.
Lee’s group (23) developed a hybrid silicon/poly(methyl meth-
acrylate) chip with platinum thin-film heaters and RTD sensors
fabricated on the Si wafer. The authors reported successful PCR
amplification in 390 s although, due to the 200-nL volume in the
PCR chamber, amplified DNA could not be removed from the
chip for analysis. Therefore, SYBR Green I dye was added to the
reaction chamber and increased fluorescence, compared against
fluorescence prior to PCR, was used as an indirect measure of
PCR success. Guttenberg et al. (24) used a surface acoustic wave
pump to drive a 200-nL droplet of PCR mixture to two temperature
zones produced by thin-film heaters on a planar substrate.
Detection was accomplished by pumping the solution to a
hybridization probe spotted on the same device. Amplification
times for this two-temperature PCR were as low as 10 min
beginning with 10 genomic DNA copies. This work registered a
score of 0.12 bp µL-1 s-1 copy-1. Finally, a low-volume PCR
microdevice was developed by Burns and co-workers (25), with
240-nL reaction chambers isolated from other domains within the
microchip through the use of phase-changing valves. Heaters and
temperature sensors were integrated within the device, and
amplification of mouse DNA snrpn template was achieved in 31
min as determined by off-chip separation.
Small-volume PCR is advantageous in reducing the cost of
reagents used for the amplification (especially costly Taq poly-
merase) but should also allow for more rapid thermocycling and,
as a result, faster temperature transitions and more efficient PCR
amplification. However, as described here, the use of small-volume
PCR chambers does not necessarily coincide with reduced
amplification times if the mode of heating does not exploit the
3890 Analytical Chemistry, Vol. 77, No. 12, June 15, 2005
advantages of low-volume design (see Table 1), which is reflected
by the fact that the scores are an order of magnitude below ideal.
This is certainly the case when thermocycling involves heating
the entire device and not simply the PCR domain. This is
supported by the fact that the fastest thermocycling described
here involved chip-integrated heaters, where only the PCR domain
(and the enclosed PCR mixture) was heated (22-24). A drawback
to integrating heaters and temperature sensors directly into the
device is the complicated and costly method of fabrication.
Although several of these reports did not have small-volume
PCR with high-speed amplification as the ultimate goal of the
project, these systems are much faster compared to conventional
thermocyclers. However, if a complete genetic analysis system
for time-sensitive diagnostics (e.g., biowarfare detection) is
needed, the PCR must be performed with greater amplification
speed.
HIGH-SPEED PCR
To obtain high-speed PCR, reducing the thermal mass of the
system can increase the heating rates during temperature cycling.
However, as noted above, small mass (or volume) does not
necessarily translate to fast PCR thermocycling. The fastest PCR
during 2003spresent was performed with flow-through devices
in which the PCR Mastermix and template DNA are pumped past
preheated domains. Thus, only thermal mass was the solution
and not the microdevice.
The work performed by Hashimoto et al. (26) exemplifies the
fast speed attainable with PCR performed in microfluidic chips.
They utilized a flow-through device consisting of discrete tem-
perature zones produced by electrical resistance heaters. This
work leveraged the original work of Manz and co-workers (16),
who performed 20 cycles of amplification in 90 s using a glass
microdevice. In the Hashimoto et al. work, amplification of a 500-
bp region of λ-phage DNA was performed in 1.7 min while a 997-
bp fragment took 3.2 min, which the authors attributed to the
kinetic limit of nucleotide incorporation by Taq polymerase. Finite
element analysis revealed the steady-state temperature distribution
and the temperature of the PCR reagents in each zone as a
function of velocity. It was found that if the flow was too fast
between the denaturation and renaturation step during amplifica-
tion, cooling to the desired temperature was not attained with the
passive system used. The starting copy number of DNA will affect
the speed at which PCR amplification can be performed, and with
a linear velocity of 10 mm/s, the lowest DNA starting copy to
provide amplified signal in 20 cycles was 1 × 107 copies. The
authors attempted lower concentrations at this velocity; however,
amplified product was undetectable in their system. Even with
the extremely fast cycling, combining such an unfavorable LOD
with the large volume used (2.5 µL) resulted in a score of 9.8 ×
10-8 bp µL-1 s-1 copy-1.
Another flow-through PCR device capable of either PCR or
reverse transcriptase PCR (RT-PCR) developed by Obeid et al.
(27) uses four heating blocks, three for PCR amplification and
the fourth for reverse transcription. The device contained outlet
reservoirs after 20, 25, 30, 35, or 40 cycles to allow the user to
collect samples at different cycle numbers. With 107 starting copies
of DNA in 10 µL of reaction buffer, 30 cycles of amplification could
be performed in 6 min using a flow rate of 21 nL/s. RT-PCR was
performed in starting volumes as low as 700 nL with the lower
limit of starting copy number tested at 6.25 × 106. The ability to
perform multiple PCR or RT-PCR amplifications was also shown.
Our laboratory has pioneered the use of infrared-mediated
heating for performing PCR amplification (7). While the IR
radiation impinges only on the solution in the PCR chamber, the
surrounding glass will heat due to transfer of the energy and
partial absorption of the IR light by the glass. By completely
etching glass away near the PCR zone, IR-mediated PCR (25
cycles) has been performed in 5 min, one of the fastest cycling
times using a non-flow-through system.
These reports represent several of the faster analyses reported
within the past 2 years. Unfortunately, both flow-through devices
used a large amount of template DNA. Though scores were not
assigned to most of these, a similar problem with time and volume
optimization was observed among the fast cycling reports, in which
reaction volumes were frequently in the microliter range. The true
value of high-speed PCR amplification will be realized when low
starting copy numbers can be amplified in a short time in low-
volume chambers.
FUTURE STRATEGIES
While the previous sections have shown small volume and fast
temperature cycling for the PCR, several other noteworthy reports
have focused on unconventional strategies for amplification of
DNA via the PCR or for unconventional means of heating for the
PCR. Many of these methodologies, while still conceptual in
nature, have the potential to impact PCR in the future.
Soper and co-workers (28) have demonstrated an electroki-
netically controlled flow-through PCR microdevice fabricated in
a polycarbonate substrate. Two key features of their method
included the use of a positive polyelectrolyte multilayer, Polybrene,
to coat the channel walls for reversed electroosmotic flow (EOF)
and a low ionic strength PCR buffer to minimize Joule heating.
By using EOF to control flow, cycle numbers could be altered
while one-channel design was maintained. Adequate amounts of
amplified product from a 500-nL PCR reaction could be attained
after 15 cycles (600 s) beginning with 10 ng/µL λ-phage DNA.
One benefit of an electrokinetically controlled PCR flow-through
device is valveless integration of up- and downstream DNA
analysis. However, one must be aware that there is a possibility
of the anionic DNA binding to the cationic Polybrene-coated walls.
Ugaz and co-workers (29) have built upon previous work with
Rayleigh-Bénard convection systems for PCR thermocycling. By
placing an array of 35 µL polycarbonate wells on an aluminum
heating block set at 95 °C and covering the wells with an
aluminum heating block at 61 °C, convection drives the solution
through the various temperature zones, thereby, thermocycling.
To provide more uniform residence times in the temperature
zones, closed-loop devices were developed that were in contact
with heaters and provided the same type of convective thermocy-
cling. Amplification of a 191-bp fragment was observed in the 35-
µL wells in 15 min, and a 295-bp fragment was amplified in 30
min using the 15-µL closed-loop system. Chen et al. (30) have
also demonstrated successful amplification of both a 305- and 700-
bp DNA fragment using a closed-loop system. The authors point
out that most reported closed-loop systems amplify fragments less
than 300 bp. They also note that these convection-based amplifica-
tion devices may be useful on centrifugal microfluidic platforms.
The majority of papers presented in this review have relied
on physical heaters. A completely orthogonal approach was
described by Verpoorte and co-workers (31), who take advantage
of exothermic and endothermic chemical reactions on a microf-
luidic chip to heat and cool microchannels. Evaporation of acetone
from a channel produced a cooling effect down to -5 °C, while
on-chip dilution of concentrated sulfuric acid provided heating to
76 °C. While not demonstrated as functional for PCR amplification,
this clever approach for controlling temperature is simple and does
not require macrocomponents or multiple fabrication steps for
construction of the microfluidic chip.
Finally, an interesting approach to expediting genetic analysis
may be to avoid PCR altogether. This can be achieved with
detection that has single-molecule sensitivity. Park et al. (32) used
a microfluidic chip to mix fluorescently labeled sex-determining
region Y gene with 65-70-nm-diameter silver particles. Confocal
surface-enhanced Raman spectroscopy was used to monitor the
oligonucleotide, and a 10 pM limit of detection was reported.
Although significant technical hurdles exist, it may be feasible to
use this type of system to detect hybridization of a labeled probe
with gene sequences on template DNA, thereby, circumventing
the need for PCR.
POTENTIAL IMPROVEMENTS
Real-Time PCR. Real-time PCR (or quantitative PCR, qPCR)
has the advantage that the PCR amplification process can be
monitored as it occurs, presenting the opportunity to obtain
quantitative PCR informationsthis is not possible with standard
PCR. There has been scant work reported on qPCR in the
literature, which is surprising given the availability of commercial
qPCR kits. In addition, the optical setup used for detection is
simple considering a typical laser-induced fluorescence setup (that
most microfluidic labs use) is all that is needed. Chip-based qPCR
would simplify the chip in many ways because the separation step
could be eliminated. Moreover, the ability to obtain quantitative
information lends itself perfectly to diagnostic applications where
the number of copies of starting template have clinical relevance
(e.g., viral titer).
Whole Genome Amplification. While single-copy PCR has
been demonstrated on-chip (33), having access to a large amount
of genomic DNA would be advantageous with samples where the
number of starting template copies is low. Whole genome
amplification (WGA) provides such a method (34) and is used in
many molecular biological assays associated with the analysis of
human disease, where limited sample/tissue is available. For
example, when a particular focus of interest consists of only a
few hundred cells, the amount of extractable DNA may be
insufficient for genetic analysis. Comprehensive WGA is a chal-
lenge because it requires accurate replication of more than 3
billion bases without the loss or mutation of any particular loci/
alleles. However, unlike the PCR (where specific sequences are
targeted), WGA must represent the entire genome with minimal
amplification bias. Methods for WGA have been developed that
minimize this bias and allow, through conventional means, the
generation of as much as microgram quantities of DNA (35). Since
Analytical Chemistry, Vol. 77, No. 12, June 15, 2005 3891
both thermal and isothermal WGA methods have been developed,
it is not unreasonable to anticipate that this can be accomplished
on-chip with the benefit of fast thermocycling and little dilution.
CONCLUSIONS
Although PCR is a mature technique, its adaptation to the
microchip platform is an evolving technology. It is clear that the
analytical challenges associated with doing efficient, fast, low-
volume PCR in a microfluidic chip have, by no means, been
completely addressed and solved (see Table 1). The evolution
involves trying to test and optimize a number of parameters that
include substrate (one that provides the best characteristics for
both PCR and fabrication), design of the PCR chamber, means of
temperature cycling, and whether the device is stand-alone or
integrated with other on-chip functions. This evolution is apparent
in the recent literature covered here. Ironically, some of the fastest
PCR reactions reported used the largest starting template copy
number and the largest volumes, while several of the lowest
volume PCR chambers thermocycled the slowest. While it may
appear that a random logic dominates the evolution of chip-based
PCR, the volumes thermocycled and the starting template copy
number used depend largely on the type of heating utilized and
the chip substrate. Other parameters, such a flow (for flow-through
systems) and temperature sensing, factor in as well. Consequently,
the morphing that seems apparent will continue until there is
convergence to a few functional systems that, in all likelihood,
will be detemined by the application addressed. Several functional
substrates will be defined, and with reaction volumes that meet
the need of the application. The choice of heating method will be
determined by the speed deemed acceptable for the thermocy-
cling, and where ultrafast thermocycling is required (sub-10 s
cycles), small reaction volumes may be a preprequisite. These
dependencies on the application can be better reflected in the PCR
score as well. A more valuable approach to scoring each device
would be to weight each variable according to its importance. For
example, if access to 1 mL of blood is attainable, but results need
to be expedited; a low weighting for PCR volume and a high
weighting for thermocycling speed could be used allowing for the
ideal chip-PCR method to be used. Also, the device and method
of choice will have to move beyond template DNA such as λ-phage
DNA and PCR vectors, and onto real-world samples. Each of these
strategies for performing PCR have their own merits and down-
falls, and it may be some time before a truly low-cost, low-volume,
high-speed PCR device capable of handling real samples is
attained; however, when these goals are reached, the true benefit
of using a microfluidic device for DNA amplification will be
attained.
ACKNOWLEDGMENT
The authors acknowledge Lindsay Legendre for help with calcula-
tions pertaining to Table 1.
Michael G. Roper is a postdoctoral researcher at the University of
Virginia. He received his B.S. from the University of Texas at Austin in
1998 and attained his Ph.D. from the University of Florida under the
supervision of Professor Robert T. Kennedy in 2003. His research has
centered on using microfluidic chips to measure biomarker protein
concentrations found in serum and measure cellular releasate with fast
temporal resolution.
3892 Analytical Chemistry, Vol. 77, No. 12, June 15, 2005
Chrisopher J. Easley is a graduate student completing his doctoral
degree in analytical chemistry at the University of Virginia. He received
his baccalaureate degree in chemistry from Mississippi State University
in Starkville, MS, in 2002. His current reasearch involves rapid
noncontact microchip PCR integrated with electrophorectic separation for
genetic diagnoses of diseases or pathogen detection.
James P. Landers is a Professor of Chemistry at the University of
Virginia and Associate Professor of Pathology at the University of Virginia
Health Sciences Center. He received his B.S. degree in biochemistry (minor
in biomedicine) at the University of Guelph (Canada in 1984 and his
Ph.D. in biochemistry from the Department of Chemistry and Biochemistry
at the same university in 1988. After a postdoctoral fellowship at the
Banting Institute at the University of Toronto’s School of Medicine, he
served as a Canadian Medical Research Council (MRC) Fellow at the
Mayo Clinic under Thomas Spelsberg. Following his MRC Fellowship,
he was appointed as the Director of the Clinical Capillary Electrophoresis
Facility in Mayo Clinic’s Department of Laboratory Medicine and
Pathology. His research interests include the development of analytical
microchip technology for use in human disease diagnostics. This not only
involves method development for electrophoretic separations on microchips
focused on expediting molecular diagnosis but also includes the develop-
ment of microchip-based sample processing methodologies necessary to
create a totally integrated microchip system fir clinical diagnostics.
LITERATURE CITED
(1) Venter, J. C.; Adams, M. D.; Myers, E. W.; Li, P. W.; Mural, R. J.;
Sutton, G. G.; Smith, H. O.; Yandell, M.; Evans, C. A.; Holt, R. A.;
Gocayne, J. D.; Amanatides, P.; Ballew, R. M.; Huson, D. H.;
Wortman, J. R.; Zhang, Q.; Kodira, C. D.; Zheng, X. H.; Chen, L.;
Skupski, M.; Subramanian, G.; Thomas, P. D.; Zhang, J.; Gabor
Miklos, G. L.; Nelson, C.; Broder, S.; Clark, A. G.; Nadeau, J.;
McKusick, V. A.; Zinder, N.; Levine, A. J.; Roberts, R. J.; Simon,
M.; Slayman, C.; Hunkapiller, M.; Bolanos, R.; Delcher, A.; Dew,
I.; Fasulo, D.; Flanigan, M.; Florea, L.; Halpern, A.; Hannenhalli,
S.; Kravitz, S.; Levy, S.; Mobarry, C.; Reinert, K.; Remington, K.;
Abu-Threideh, J.; Beasley, E.; Biddick, K.; Bonazzi, V.; Brandon,
R.; Cargill, M.; Chandramouliswaran, I.; Charlab, R.; Chaturvedi,
K.; Deng, Z.; Di Francesco, V.; Dunn, P.; Eilbeck, K.; Evangelista,
C.; Gabrielian, A. E.; Gan, W.; Ge, W.; Gong, F.; Gu, Z.; Guan, P.;
Heiman, T. J.; Higgins, M. E.; Ji, R. R.; Ke, Z.; Ketchum, K. A.;
Lai, Z.; Lei, Y.; Li, Z.; Li, J.; Liang, Y.; Lin, X.; Lu, F.; Merkulov,
G. V.; Milshina, N.; Moore, H. M.; Naik, A. K.; Narayan, V. A.;
Neelam, B.; Nusskern, D.; Rusch, D. B.; Salzberg, S.; Shao, W.;
Shue, B.; Sun, J.; Wang, Z.; Wang, A.; Wang, X.; Wang, J.; Wei,
M.; Wides, R.; Xiao, C.; Yan, C.; Yao, A.; Ye, J.; Zhan, M.; Zhang,
W.; Zhang, H.; Zhao, Q.; Zheng, L.; Zhong, F.; Zhong, W.; Zhu,
S.; Zhao, S.; Gilbert, D.; Baumhueter, S.; Spier, G.; Carter, C.;
Cravchik, A.; Woodage, T.; Ali, F.; An, H.; Awe, A.; Baldwin, D.;
Baden, H.; Barnstead, M.; Barrow, I.; Beeson, K.; Busam, D.;
Carver, A.; Center, A.; Cheng, M. L.; Curry, L.; Danaher, S.;
Davenport, L.; Desilets, R.; Dietz, S.; Dodson, K.; Doup, L.;
Ferriera, S.; Garg, N.; Gluecksmann, A.; Hart, B.; Haynes, J.;
Haynes, C.; Heiner, C.; Hladun, S.; Hostin, D.; Houck, J.; Howland,
T.; Ibegwam, C.; Johnson, J.; Kalush, F.; Kline, L.; Koduru, S.;
Love, A.; Mann, F.; May, D.; McCawley, S.; McIntosh, T.;
McMullen, I.; Moy, M.; Moy, L.; Murphy, B.; Nelson, K.;
Pfannkoch, C.; Pratts, E.; Puri, V.; Qureshi, H.; Reardon, M.;
Rodriguez, R.; Rogers, Y. H.; Romblad, D.; Ruhfel, B.; Scott, R.;
Sitter, C.; Smallwood, M.; Stewart, E.; Strong, R.; Suh, E.; Thomas,
R.; Tint, N. N.; Tse, S.; Vech, C.; Wang, G.; Wetter, J.; Williams,
S.; Williams, M.; Windsor, S.; Winn-Deen, E.; Wolfe, K.; Zaveri,
J.; Zaveri, K.; Abril, J. F.; Guigo, R.; Campbell, M. J.; Sjolander,
K. V.; Karlak, B.; Kejariwal, A.; Mi, H.; Lazareva, B.; Hatton, T.;
Narechania, A.; Diemer, K.; Muruganujan, A.; Guo, N.; Sato, S.;
Bafna, V.; Istrail, S.; Lippert, R.; Schwartz, R.; Walenz, B.; Yooseph,
S.; Allen, D.; Basu, A.; Baxendale, J.; Blick, L.; Caminha, M.;
Carnes-Stine, J.; Caulk, P.; Chiang, Y. H.; Coyne, M.; Dahlke, C.;
Mays, A.; Dombroski, M.; Donnelly, M.; Ely, D.; Esparham, S.;
Fosler, C.; Gire, H.; Glanowski, S.; Glasser, K.; Glodek, A.;
Gorokhov, M.; Graham, K.; Gropman, B.; Harris, M.; Heil, J.;
Henderson, S.; Hoover, J.; Jennings, D.; Jordan, C.; Jordan, J.;
Kasha, J.; Kagan, L.; Kraft, C.; Levitsky, A.; Lewis, M.; Liu, X.;
Lopez, J.; Ma, D.; Majoros, W.; McDaniel, J.; Murphy, S.;
Newman, M.; Nguyen, T.; Nguyen, N.; Nodell, M.; Pan, S.; Peck,
J.; Peterson, M.; Rowe, W.; Sanders, R.; Scott, J.; Simpson, M.;
Smith, T.; Sprague, A.; Stockwell, T.; Turner, R.; Venter, E.; Wang,
M.; Wen, M.; Wu, D.; Wu, M.; Xia, A.; Zandieh, A.; Zhu, X. Science
2001, 291, 1304-1351.
(2) Dovichi, N. Science 1999, 285, 1013.
(3) Paegel, B. M.; Blazej, R. G.; Mathies, R. A. Curr. Opin. Biotechnol.
2003, 14, 42-50.
(4) Yeung, S. H. I.; Greenspoon, S. A.; Ban, J. D.; McGuckian, A.;
Crouse, C.; Mathies, R. A. In preparation.
(5) Saiki, R. K.; Scharf, S.; Faloona, F.; Mullis, K. B.; Horn, G. T.;
Erlich, H. A.; Arnheim, N. Science 1985, 230, 1350-1354.
(6) Millar, B. C.; Moore, J. E. Methods Mol. Biol. 2004, 266, 139-
166.
(7) (a) Oda, R. P.; Strausbauch, M. A.; Huhmer, A. F.; Borson, N.;
Jurrens, S. R.; Craighead, J.; Wettstein, P. J.; Eckloff, B.; Kline,
 B.; Landers, J. P. Anal. Chem. 1998, 70, 4361-4368. (b) Huhmer,
A. F. R.; Landers, J. P. Anal. Chem. 2000, 72, 5507-5512. (c)
Giordano, B. C.; Copeland, E. R.; Landers, J. P. Electrophoresis
2001, 22, 334-340. (d) Legendre, L. A.; Bienvenue, J. M.; Roper,
M. G.; Ferrance, J.; Landers, J. P. In preparation. (e) Giordano,
B. C.; Ferrance, J.; Swedberg, S.; Huhmer, A. F.; Landers, J. P.
Anal. Biochem. 2001, 291, 124-132.
(8) Liu, R. H.; Grodzinski, P. A. J. Microlithogr., Microfabr., Microsyst.
2003, 2, 340-455.
(9) Bilitewski, U.; Genrich, M.; Kadow, S.; Mersal, G. Anal. Bioanal.
Chem. 2003, 377, 556-569.
(10) Kricka, L. J.; Wilding, P. Anal. Bioanal. Chem. 2003, 377, 820-
825.
(11) Auroux, P. A.; Koc, Y.; deMello, A.; Manz, A.; Day, P. J. Lab Chip
2004, 4, 534-546.
(12) Kelly, R. T.; Woolley, A. T. Anal. Chem. 2005, 77, 96A-102A.
(13) Wilding, P. J.; Bau, H. H.; Zemel, J. N.; Shaffner, A. M.; Kricka,
L. J. Clin. Chem. 1994, 40, 1815-1818.
(14) Wilding, P.; Shoffner, M. A.; Cheng, J.; Huichia, G.; Kricka, L. J.
Clin. Chem. (Washington, D. C.) 1995, 41, 1367-1368.
(15) Burns, M. A.; Johnson, B. N.; Brahmasandra, S. N.; Handique,
K.; Webster, J. R.; Krishnan, M.; Sammarco, T. S.; Man, P. M.;
Jones, D.; Heldsinger, D.; Mastrangelo, C. H.; Burke, D. T. Science
1998, 282, 484-487.
(16) Kopp, M. U.; Mello, A. J.; Manz, A. Science 1998, 280, 1046-
1048.
(17) Waters, L. C.; Jacobson, S. C.; Kroutchinina, N.; Khandurina, J.;
Foote, R. S.; Ramsey, J. M. Anal. Chem. 1998, 70, 158-162.
(18) Lagally, E. T.; Emrich, C. A.; Mathies, R. A. Lab Chip 2001, 1,
102-107.
(19) Liu, J.; Hansen, C.; Quake, S. R. Anal. Chem. 2003, 75, 4718-
4723.
(20) Koh, C. G.; Tan, W.; Zhao, M. Q.; Ricco, A. J.; Fan, Z. H. Anal.
Chem. 2003, 75, 4591-4598.
(21) Matsubara, Y.; Kerman, K.; Kobayashi, M.; Yamamura, S.; Morita,
Y.; Takamura, Y.; Tamiya, E. Anal. Chem. 2004, 76, 6434-6439.
(22) Lagally, E. T.; Scherer, J. R.; Blazej, R. G.; Toriello, N. M.; Diep,
B. A.; Ramchandani, M.; Sensabaugh, G. F.; Riley, L. W.; Mathies,
R. A. Anal. Chem. 2004, 76, 3162-3170.
(23) Lee, D. S.; Park, S. H.; Yang, H.; Chung, K. H.; Yoon, T. H.; Kim,
S. J.; Kim, K.; Kim, Y. T. Lab Chip 2004, 4, 401-407.
(24) Guttenberg, Z.; Muller, H.; Habermuller, H.; Geisbauer, A.; Pipper,
J.; Felbel, J.; Kielpinski, M.; Scriba, J.; Wixforth, A. Lab Chip
2005, 5, 308-317.
(25) Pal, R.; Yang, M.; Johnson, B. N.; Burke, D. T.; Burns, M. A.
Anal. Chem. 2004, 76, 3740-3748.
(26) Hashimoto, M.; Chen, P. C.; Mitchell, M. W.; Nikitopoulos, D.
E.; Soper, S. A.; Murphy, M. C. Lab Chip 2004, 4, 638-645.
(27) Obeid, P. J.; Christopoulos, T. K.; Crabtree, H. J.; Backhouse, C.
J. Anal. Chem. 2003, 75, 288-295.
(28) Chen, J.; Wabuyele, M.; Chen, H.; Patterson, D.; Hupert, M.;
Shadpour, H.; Nikitopoulos, D.; Soper, S. A. Anal. Chem. 2005,
77, 658-666.
(29) Krishnan, M.; Agrawal, N.; Burns, M. A.; Ugaz, V. M. Anal. Chem.
2004, 76, 6254-6265.
(30) Chen, Z.; Qian, S.; Abrams, W. R.; Malamud, D.; Bau, H. H. Anal.
Chem. 2004, 76, 3707-3715.
(31) Guijt, R. M.; Dodge, A.; van Dedem, G. W.; de Rooij, N. F.;
Verpoorte, E. Lab Chip 2003, 3, 1-4.
(32) Park, T.; Lee, S.; Seong, G. H.; Choo, J.; Lee, E. K.; Kim, Y. S.; Ji,
W. H.; Hwang, S. Y.; Gweon, D. G. Lab Chip 2005, 5, 437-442.
(33) Lagally, E. T.; Medintz, I.; Mathies, R. A. Anal. Chem. 2001, 73,
565-570.
(34) Hughesab, S.; Arnesona, N.; Donea, S.; Squirea, J. J. Prog. Biophys.
Mol. Biol. 2005, 88, 173-189.
(35) Lasken, R. S.; Egholm, M. Trends Biotechnol. 2003, 21, 531-
535.
(36) Obeid, P. J.; Christopoulos, T. K. Anal. Chim. Acta 2003, 494,
1-9.
(37) Liu, R. H.; Yang, J.; Lenigk, R.; Bonanno, J.; Grodzinski, P. Anal.
Chem. 2004, 76, 1824-1831.
(38) Panano, N. J.; Lou, X. J.; Fortina, P.; Kricka, L. J.; Wilding, P.
Biomol. Eng. 2005, 21, 157-162.
(39) Zhou, X.; Liu, D.; Zhong, R.; Dai, Z.; Wu, D.; Wang, H.; Du, Y.;
Xia, Z.; Zhang, L.; Mei, X.; Lin, B. Electrophoresis 2004, 25, 3032-
3039.
(40) Rodriguez, I.; Lesaicherre, M.; Tie, Y.; Zou, Q.; Yu, C.; Singh, J.;
Meng, L. T.; Uppili, S.; Li, S. F.; Gopalakrishnakone, P.; Sel-
vanayagam, Z. E. Electrophoresis 2003, 24, 172-178.
(41) Lee, T. M.; Carles, M. C.; Hsing, I. M. Lab Chip 2003, 3, 100-
105.
(42) Liu, J.; Enzelberger, M.; Quake, S. Electrophoresis 2002, 23,
1531-1536.
AC050756M
Analytical Chemistry, Vol. 77, No. 12, June 15, 2005 3893