Microfluid Nanofluid
DOI 10.1007/s10404-013-1283-9
RESEARCH PAPER
Centrifugal automation of a triglyceride bioassay on a low-cost
hybrid paper-polymer device
Neus Godino • Elizaveta Vereshchagina
Robert Gorkin III • Jens Ducrée
•
Received: 10 July 2013 / Accepted: 22 October 2013
Springer-Verlag Berlin Heidelberg 2013
Abstract We present a novel paper-polymer hybrid
construct for the simple automation of fundamental
microfluidic operations in a lab-on-a-disc platform. The
novel design, we term a paper siphon, consists of chromatographic paper strips embedded along a siphon microchannel. The paper siphon relies on two main
interplaying forces to create unique valving and liquidsampling methods in centrifugal microfluidics. At sufficiently low speeds, the inherent wicking of the paper
overcomes the rotationally induced centrifugal force to
drive liquids towards inwards positions of the disc. At
elevated speeds, the dominant centrifugal force will extract
liquid from the siphon paper strip towards the edge of the
Neus Godino and Elizaveta Vereshchagina have contributed equally
to this work.
N. Godino E. Vereshchagina R. Gorkin III J. Ducrée (&)
Biomedical Diagnostics Institute, National Centre for Sensor
Research, School of Physical Sciences, Dublin City University,
Dublin, Ireland
e-mail: [email protected]
N. Godino
e-mail: [email protected]
Present Address:
N. Godino
Fraunhofer Institute for Biomedical Engineering IBMT,
Branch Potsdam, Potsdam, Germany
Present Address:
E. Vereshchagina
Microsystems Centre of Tyndall National Institute, Cork, Ireland
Present Address:
R. Gorkin III
ARC Centre of Excellence for Electromaterials Science,
Intelligent Polymer Research Institute, University
of Wollongong, Wollongong, Australia
disc. Distinct modes of flow control have been developed
to account for water (reagent) and more viscous plasma
samples. The system functionality is demonstrated by the
automation of sequential sample preparation steps in a
colorimetric triglyceride assay: plasma is metered from a
whole blood sample and incubated with a specific enzymatic mixture, followed by detection of triglyceride levels
through (off-disc) absorbance measurements. The successful quantification of triglycerides and the simple fabrication offer attractive directions for such hybrid devices
in low-cost bioanalysis.
Keywords Lab-on-a-disc Centrifugal
microfluidics Paper microfluidics Point of care
diagnostic Triglyceride assay
1 Introduction
Paper microfluidics has garnered attention for the development of low-cost point-of-care diagnostic devices in the
scientific community (Ballerini et al. 2012; Fu et al. 2011)
as well as for commercial purposes (Fulmer 2012). Besides
the game-changing cost advantage of the material, paper
devices offer further benefits when compared to conventional substrates such as silicon, glass or plastics, such as
an easier pathway towards fabrication (Carrilho et al. 2009;
Martinez et al. 2010) and remarkable capabilities for dry
reagent storage and surface functionalization (Ge et al.
2012; Kwong and Gupta 2012). In the last few years, the
field of paper microfluidics has brought various innovative,
low-cost and rapid engineering solutions to biodiagnostics
(Arnaud 2012; Liana et al. 2012; Shah et al. 2013). Furthermore, due to the wide range of inherent advantages of
using paper like flow control through wicking and inherent
123
Microfluid Nanofluid
biocompatibility, paper microfluidic devices have been
harnessed for applications such as human blood typing (AlTamimi et al. 2012; Jarujamrus et al. 2012), plasma separation and detection of various biomolecular entities
(antibodies, enzymes) (Songjaroen et al. 2012; Yang et al.
2012). Recent publications have improved the control over
the two-dimensional flow for the sequential delivery of
liquid reagents (Fu et al. 2012).
While advantageous as a substrate, using paper alone
puts some constraints on functionality of microfluidic
devices to perform certain laboratory unit operations
(LUOs) such as metering or valving. Paper can pose
challenges for sequential processing as liquids cannot
successively run through the same passage. The field has
certainly progressed, with recent advances using the
inherent filter capabilities of paper for human blood typing
on antibody-treated paper (Al-Tamimi et al. 2012; Jarujamrus et al. 2012) and in the development of microfluidics
paper-based analytical devices for plasma separation and
subsequent detection of proteins in plasma (Songjaroen
et al. 2012; Yang et al. 2012).
In this work, we join the advantages of using paper as a
substrate regarding capillarity, filtering, cost and simple
fabrication to perform novel concepts of fundamental microfluidics functionalities in centrifugal platforms. Centrifugal microfluidics is a mature technology with
established benefits for liquid handling. One of the major
developments in centrifugal microfluidics is standardization of method for blood separation in clinical assays. We
describe for the first time the development of composite
paper-centrifugal microfluidic valving and liquid-sampling
method that enables a novel concept that we term a paper
siphon. When combining the centrifugally actuated lab-ona-disc with the autonomous capillary absorption of paper, it
is possible to develop a fundamental set of LUOs in a
simple, inexpensive and rapid fashion (Godino et al. 2012a,
b; Vereshchagina et al. 2012; Hwang et al. 2011).
Siphoning is one of the most common features in centrifugal microfluidic platforms that is based on the principle
of hydrostatic equilibrium of liquid levels in the rotationally induced artificial gravity field (Siegrist et al. 2009;
Steigert et al. 2007). In a common implementation, two
chambers are interconnected using reverse-U shaped duct
featuring a crest point above the maximum liquid level in
the upstream chamber. When a fluid meniscus passes this
crest point and protrudes radially outward below the liquid
level, liquid is centrifugally ‘‘pulled’’ through the siphon
channel to the downstream chamber. Priming of the siphon
channel has been implemented by flow propulsion through
capillary action, or through pneumatic and thermal actuation (Garcia-Cordero et al. 2010; Schembri et al. 1995;
Gorkin et al. 2010; Godino et al. 2013; Grumann et al.
2006; Kitsara and Ducrée 2013). In this work, we develop a
123
novel siphoning concept utilizing the inherent imbibition of
paper segments. The use of paper in siphon valving has
several important benefits: (1) simple manufacturing (no
need for additional hydrophilization processes); (2) flow
control by both capillary and centrifugation forces; (3) low
cost; (4) long shelf life; (5) accessibility to surface modification techniques, if necessary; (6) biocompatibility
(Pelton 2009) and (7) storing of dry reagents (Fridley et al.
2012).
Our hybrid lab-on-a-disc merges both, the effectiveness
of centrifugal platforms for blood separation with the
straightforward filtering properties of paper. With this
method, we succeeded to integrate a complete assay protocol including plasma isolation, metering and filtering
simultaneously through a disc-based paper siphon. The
protocol is used to perform a commercial triglyceride assay
as a pilot study. Along with cholesterol, triglyceride levels
constitute an important blood testing parameter. Excessive
triglyceride concentrations may indicate an enhanced risk
of heart attack, stroke, diabetes or other metabolic disorders. Thus, low-cost diagnostic devices enabling frequent,
near-patient monitoring of triglyceride levels are highly
desired. The colorimetric assay involves initial blood separation step with extraction of 10 lL of pure plasma, followed by mixing with an enzymatic reagent and off-disc
colorimetric detection. We demonstrate the integration and
automation of the subsequent assay steps on our hybrid labon-disc utilizing the paper siphon as will be shown
throughout this work.
2 Principle
Flow control in the paper-polymer lab-on-a-disc is governed by the interplay of two primary effects: capillarity
action and the centrifugal field. Capillary flow is driven by
imbibition of the dry (hydrophilic) paper segments. The
capillary flow within dry paper is isotropic and its speed is
primarily impacted by the porosity and hydrophilicity of
the paper. The centrifugal field, on the other hand, always
points in the radial direction and scales linearly with the
distance from the centre of rotation and with the square of
the frequency of rotation and can be dynamically controlled by the speed of a spindle motor. By integrating
paper membranes on a disc and setting a sequence of
designated rotational frequencies, it is possible to control
liquid transport and to establish fluid pathways that are
challenging to achieve with individual stand-alone microfluidic components.
Figure 1 depicts the three different conditions regarding
the balance between centrifugal and capillary force when
using a paper strip in a centrifugal platform. In the schematic, the liquid is brought in contact with the paper
Microfluid Nanofluid
Each case describes a different state of the processing on
hybrid paper-polymer disc. Liquids can be driven towards
the centre (Fig. 1a); the meniscus level can be temporarily
held at particular radii (Fig. 1b); or liquids can be
transferred from pre-wetted material towards the disc edge
(Fig. 1c). Additionally, by establishing a spin protocol that
cycles the spin speed slightly below and above the
threshold frequency, a combination of the techniques can
be used to re-circulate fluids through the paper membrane
(Godino et al. 2012a, b) or, as we will show in the following, enable siphoning and metering.
Based on the method of fluidic control described in case
(iii), the extraction of well-defined liquid volumes from a
paper strip can be performed, a LUO that we designate as
‘‘liquid sampling’’. As depicted in Fig. 1d when a paper
strip is saturated with liquid, microdrops are formed at the
surface. The number and shape of these drops depends on
the porosity of the paper and the specific liquid-to-vapour
interface. In conventional paper microfluidics, the equilibrium of these drops is merely governed by the surface
tension and viscosity of the sample. However, when a
thoroughly soaked paper strip is spun, the centrifugal force
causes an accumulation of liquid at the outer edge of the
strip. Once a sufficient volume of liquid sample has
amassed, the surface tension is overcome and liquid drops
are released from the membrane (Fig. 1e). The appearance
Fig. 1 Schematic of forces controlling the hydrodynamics on the
hybrid paper-polymer lab-on-a-disc platform. The red arrow corresponds to the centrifugal force and the green arrow shows the
capillary force. a When the capillary force is higher than the
centrifugal force the liquid wicks from ‘‘wet’’ areas to ‘‘dry’’ areas
even if these ‘‘dry’’ areas are positioned radially inwards. b When the
capillary force and the centrifugal force are at balance the flow stalls.
c When the centrifugal is higher than the capillary, the liquid is
transported outwards towards the perimeter of the disc. This outward
liquid movement causes a visible change in the colour of the ‘‘wet’’
areas of the paper. d When a paper strip is saturated with liquid, the
formation of microdrops on the paper surface is distinguishable.
e When the paper element saturated with the liquid is subjected to
centrifugal force, the accumulation of microdroplets at the edge is
observed. At a certain droplet size, the centrifugal force overcomes
surface tension between the droplet and the paper surface and the
saturated paper strip starts to leak by issuing droplets (colour figure
online)
creating a ‘‘wet’’ area at the outer end of the paper segment.
In general, the centrifugal force is radially outbound while
capillary action drives liquid to the ‘‘dry’’ areas of siphon
strip. Depending on the balance of forces, the following
scenarios can occur:
1.
2.
3.
At low rotational frequencies, capillarity prevails.
Liquid will be absorbed by the paper and thus move
inbounds and towards dry paper areas of the disc
(Fig. 1a).
If centrifugal and capillary forces balance at a certain
threshold-frequency (Hwang et al. 2011), the liquid
meniscus will remain at rest (Fig. 1b).
Beyond the threshold frequency, the dominant centrifugal force drives the liquid outwards. It is critical to
note that at sufficiently elevated frequencies, a certain
fraction of the liquid can be expelled from the paper
(Godino et al. 2012a, b) (Fig. 1c).
123
Microfluid Nanofluid
of the droplets, the dropping mechanism and the process of
continuous liquid extraction depend on the frequency of
rotation as well as on the intrinsic characteristics of the
liquid and the paper type. We will describe an accurate
metering of water-based and plasma samples through
modulation of rotational frequency.
strips (layer 4) are individually placed in each of the patterned siphon channels. After placing such individual paper
strips in the corresponding siphon channels, layers 5 and 6
are stacked and sealed to complete the disc assembly.
Figure 2b shows a picture of the final devices containing
shaped paper segments in the siphon channel to provide
centrifugo-capillary valving.
3 Experimental methods
3.2 Assay and reagents
3.1 Disc fabrication
For visualization and demonstration of the system fluidics,
various dilutions of food colour ink (Goodall’s, Ireland) in
DI water were used. Once established, a commercial assay
kit (Triglyceride Colorimetric Assay Kit, No. 10010303,
Cayman Chemical, USA) was integrated for the direct
determination of triglycerides. The manufacturer specifically recommended the kit for the detection of triglycerides
in plasma. The triglycerides presented in blood are catalysed by four consecutive enzymatic reactions, starting
with hydrolysis by lipase and giving quinoneimine dye
(brilliant purple colour) as a final product of the fourth
enzymatic reaction. The enzymatic mixture (Lipoprotein
Lipase, Glycerol Kinase, Glycerol Phosphate Oxidase and
Peroxidase) and sodium phosphate buffer are included in
the standard colorimetric Cayman kit No. 10010303.
Human blood samples were collected using ethical
practices from a healthy volunteer into BD Vacutainer
K3EDTA tubes (BD, USA) with pre-stored anti-coagulant.
The blood was used within 24 h after extraction.
Standard CAD software was used for the microfluidic
design of a centrifugo-pneumatic cascade. Figure 2a shows
the multi-layer sketch of the 120-mm-diameter disc. The
hybrid paper-polymer device is composed of polymeric
adhesive and paper substrates all manufactured using rapid
prototyping technologies.
The polymeric substrates, 1.5-mm thick poly(methylmethacrylate) (PMMA) (Radionics, Ireland), corresponding to layers 1, 3 and 6 in Fig. 2a, were patterned using a
CO2 laser ablation system (Zing 16 Laser, Epilog, USA).
Layers 2 and 4 correspond to the 86-lm thick, double-sided
pressure sensitive adhesive layers (PSA, Adhesive
Research, Ireland) used for bonding. Such PSA layers were
contoured using a standard knife plotter (ROBO Pro cutter/
plotter, Graphtec, USA).
To pattern the siphon areas, chromatographic paper
(Whatman type 1, General Electric Healthcare, USA) is
fixed on a layer of the double-sided PSA and then cut using
a standard knife plotter. Utilizing established polymer
lamination techniques, the polymeric and adhesive layers
were consecutively stacked/aligned and irreversibly bonded. Once layers 1, 2 and 3 are assembled, paper-siphon
Fig. 2 a 3D schematic of the disc used for the final application of
detection of triglycerides in blood. All the discs used in this work
consist of a 5-layer stack: three layers of PMMA and two layers of
double-side pressure sensitive adhesive (PSA). The Whatman Chr 1
paper siphons with adhesive on the back are positioned between layer
123
3.3 Instrumental set-up
We performed our experiments on a centrifugal test stand
with features described earlier (Grumann et al. 2005). The
3 and layer 1. b Picture of the assembled disc used for the final
detection. As mentioned, the paper-siphon strips are placed between
the two polymeric layers to allow the transport of plasma between the
blood sedimentation chamber and the final detection chamber
Microfluid Nanofluid
set-up uses a computer-controlled motor (Faulhaber Minimotor SA, Switzerland) to rotate the discs. A sensitive
camera with ls-scale exposure time (Sensicam qe, PCO,
Germany) was mounted on a motorized, 129 zoom lens
(Navitar, USA) for high-resolution imaging of microfluidic
processes on the spinning disc.
Colorimetric measurements were carried out on the
Nanodrop 2000c spectrophotometer (Thermo Fisher Scientific Inc., USA).
4 Results and discussion
The results and discussion section is divided in two sections. First, we present fundamental microfluidic functionalities achieved by integrating paper strips: siphoning
and liquid sampling. Second, as a demonstration of the
advantages of the hybrid device, we show a colorimetric
detection of triglycerides in blood based on a bioassay
consisting of plasma separation from whole blood, plasma
metering, enzymatic reaction and incubation.
4.1 Active flow control using centrifugal paper
microfluidics
4.1.1 Paper-based siphoning
In our novel paper siphon, capillarity comes purely from
the natural wicking of (dry) paper. To prime the siphon, we
first tested the wicking distance as a function of the spinning frequency. Paper strips were radially aligned on a
polymeric disc. Liquid was loaded from an outer position
relative to the dry paper areas (Fig. 3a). The disc is spun
and the wicking distance is recorded as a function of the
spin speed (Fig. 3b). As expected, when increasing the
rotational frequency, absorption through capillary action is
suppressed (Hwang et al. 2011). At a certain threshold
frequency, absorption and centrifugal forces are balanced
and no liquid movement is observed (Fig. 3b). In the
present design, the threshold frequency is *3,500 rpm.
Beyond this spin speed, liquid motion completely stalls.
For frequencies lower than 3,500 rpm, absorption towards
the centre of rotation can take place.
In this centrifugal absorption paper siphon, a paper strip
defines the siphon channel as shown in Fig. 3c. For siphon
priming, the strip must first be saturated with liquid.
Spinning below the threshold frequency leads to increasing
liquid absorption. The dotted red arrow in Fig. 3c indicates
the pathway for the liquid along the paper-siphon channel
during this phase. Once the liquid reaches the crest point,
both capillary and centrifugal forces act in the same
direction, thus pulling the liquid radially outwards. The
time to complete radially inward wetting between points A
and B in Fig. 3c increases with the frequency of rotation.
The paper siphon operates akin to the capillarity-primed,
hydrophilic open-channel siphon. Its operation is governed
by the geometry (e.g. width of the siphon), shape, thickness
and absorption properties of the paper substrate.
Establishing control over the priming mechanism is the
first step in demonstrating siphoning. To complete an
extraction protocol, liquid samples must be siphoned from
inner to outer chambers located further downstream. Figure 3d shows the full sequence of liquid extraction from a
paper-siphon at 750 rpm in more detail. First, the liquid
wicks the paper siphon (Fig. 3d-1). Once the siphon strip is
completely saturated, we observe the droplet formation on
the paper surface (Fig. 3d-2). Continued centrifugation
causes the further accumulation of liquid at the edge of the
paper to increase the stress on the droplets. At a certain
point, droplets fall off the paper (Fig. 3d-3) and are collected in the outer reservoir machined into the polymer
disc.
It is important to note that the viscosity and surface
tension of the liquid affect the siphoning procedure. More
viscous solutions will be slower to prime the siphon. If
wicking is slowed, the accumulation of liquid at the edge of
the siphon takes longer and dropping at the outer edge is
hardly achieved. Therefore, the extraction protocol is
optimized for each type of liquid sample.
4.1.2 Liquid sampling
Using the extraction protocol for paper-based siphoning,
we show in this section some metering protocols for processing water-based solutions and for human plasma. The
differences in viscosity and surface tension (plasma viscosity ranges between 1.10 and 1.35 mPa s at 37 C)
(Bascurt and Meiselman 2003; Lowe and Barbebel 1988)
require bespoke extraction parameters, i.e. spin frequency
and extraction time.
For an aqueous solution, we use a continuous extraction
mode; at constant frequency, the liquid moves through the
paper siphon, accumulates at the edge of the strip and drops
out to the following collecting chamber. In the case of plasma,
we chose a pumping extraction mode where at a low spinning
frequency the human plasma is absorbed through the paper
siphon. After certain accumulation time, the frequency is
drastically increased to allow expulsion of the plasma from the
paper strip. For both the continuous and pumping mode, the
protocol consists of optimizing the extraction parameters (for
instance, time and frequency) while keeping the volume
constant to a discrete metering step of 2 lL.
4.1.2.1 Continuous mode In this sampling mode, the
same frequency is used for transporting the liquid between
the chambers (Fig. 4a from point A to B) through the paper
123
Microfluid Nanofluid
Fig. 3 a Schematic showing wicking distance and time at a certain
frequency of rotation. In the set-up, the paper is held radially to the
centre of the disc and the liquid volume is in a position radially
outwards with respect to the ‘‘dry’’ areas of the paper. b Results
showing the liquid wicking distance versus time for a range of
frequencies of rotation. c Schematic representation of the siphon
valving principle. Certain amount of liquid is moved from position
A (initial chamber) to position B (collecting chamber). Capillary force
allows the liquid motion from ‘‘wet’’ areas to ‘‘dry’’ areas, before the
siphon crest point, centrifugal and capillary force counteract while
after the crest point both forces push the liquid to the outer edge of the
disc. d Pictures depicting liquid-ink extraction. The pictures correspond to the lowest part of the paper-siphon channel, close to the
position B (collecting chamber). Step 1: To wet the paper, the
frequency of rotation is reduced below the threshold frequency, in this
specific case the disc was spun at 750 rpm. Step 2: Once the paper
strip is fully saturated, continued rotation forms drops on the paper
surface. Step 3: Continuous accumulation of the liquid at the edge of
the siphon strip increasing the size of the drops. The larger mass
means the drops are more affected by centrifugal force. After a certain
period, the centrifugal force overcomes the surface tension and the
process of liquid extraction begins
siphon and for the continuous droplet generation from the
paper siphon. In order to optimize the continuous liquid
sampling, we study the effect of two parameters: the width
of the paper strip and the frequency of rotation. It should be
noted that the outward chamber (B) possesses a 10-lL
capacity where each measuring mark corresponds to 2-lL
increments to enable visual inspection of the metering
while extracting the liquid.
Figure 4b shows the dependence between the time and the
width of the paper strip to meter 8 lL of diluted, water-based
ink at a fixed spin rate of 750 rpm. The time was recorded in
2-lL metering steps for each width. We concluded that the
time for metering 8 lL decreases with the width of the paper
strip. At constant ratio between centrifugal and capillary
forces, the increase in paper width results in higher flow rates
of sample. Using the wider paper-siphon strips, accumulation of sufficient liquid at the edge of the paper siphon to
trigger the extraction is achieved faster. Moreover, as the
crest of the channel is kept at the same position for the different paper widths, for the wider siphon there are shorter
liquid pathways from chamber A to chamber B, so the edge
of the strip is reached more quickly. The 6-mm wide paper
siphons and strips were used in all experiments. The experimental conditions for extraction were optimized with
respect to these structures.
Figure 4c shows the effect of the frequency of rotation
to meter 2 lL using a 6-mm wide paper siphon. As it is
shown in Fig. 4c, at 1,500 rpm, there is an optimal frequency where the time necessary to meter 2 lL is minimum (42 ± 6 s). The metering process depends on two
counteracting effects regarding of the rotational frequency.
On the one hand, the frequency must be large enough to
allow the collection and release of microdroplets at the
edge of the paper strip. However, the elevated frequencies
123
Microfluid Nanofluid
Fig. 4 a Schematic of the liquid-sampling principle. The red dotted
arrow shows the pathway and the direction of the liquid motion on the
paper siphon. Position A corresponds with the initial chamber where a
liquid is loaded. Position B is the collection chamber. In our assay, the
collection chamber has a 10-lL capacity with independent marks for
2-lL steps. b Time to meter 8-lL in 2-lL steps of DI diluted ink for
several paper siphon widths: 2, 4, 5, 5.5 and 6 mm at 750 rpm.
c Average time for metering 2 lL of DI diluted ink for different
continuous pumping mode for 6 mm wide. d Average time for metering
2 lL of plasma using the discrete pumping mode. The disc is spun
always during 30 s at 375 rpm and suddenly switched to high frequency
for the necessary time to meter 2 lL of plasma (colour figure online)
of rotation restrict the liquid movement through the
inbound part prior to the crest of the paper-siphon channel
where centrifugal and capillary forces counteract. All
‘‘new’’ liquid imbibing the paper experiences the centrifugal force up to the crest point and only then the centrifugal and capillary force start to act in parallel to jointly the
liquid in the radial direction. Thus, there is an optimal
frequency of rotation at which the time necessary to meter
a certain amount of liquid is minimized. This frequency
must be sufficiently high to enable the dropping effect and
low enough to not drastically throttle the wicking process.
The optimal frequency also depends on the geometry and
dimensions of the paper siphon as well as the liquid
characteristics, so it is always necessary to tailor these
parameters for a new assay. For the results showed in
Fig. 4c, there is an optimal frequency around 1,500 rpm.
Moreover, the results showed in Fig. 4c present a proper
trend despite of the considerable large errors at some frequencies probably due to the visual quantification of the
metering. The associated sampling error is estimated to be
around 0.5 lL due to the viability of visualizing marks in
the metering chamber. This should be considered as an
absolute error of the complete metering.
4.1.2.2 Pumping mode When using more viscous liquids,
as the case of human plasma, there is not an optimal
continuous metering frequency. More viscous liquids
require increased frequencies of rotation to initiate the
123
Microfluid Nanofluid
extraction process at the edge of the paper strip, so the
wicking of ‘‘new’’ liquid sample through the siphon paper
is reduced. In other words, the metering process in continuous mode for high viscous liquids is too slow. In this
case, we use a pumping mode that alternates between two
frequencies of rotation. By inducing a series of saturating
(low frequency; wicking) and extracting (high frequency,
dropping) phases, the required sample volume is ‘‘discretely pumped’’ through the paper siphon. So to extract
certain amount of liquid V0, we perform N steps of smaller
amount Vi, for example to meter a 10-lL sample volume
we carry out N = 5 steps of Vi = 2 lL each. Such an
alternating mechanism is necessary as it is not possible to
extract high amount of liquid in a single cycle.
Figure 4d shows the mean time for a standard extraction
step of 2 lL of plasma for different extraction frequencies
(high frequency). The low frequency part of the ‘‘pumping’’
cycle (absorption) was kept constant: 30 s at 375 rpm.
However, for the first 2 lL extraction, it is necessary to spin
the disc at 375 rpm for an extended period (around 5 min) in
order to saturate the whole paper siphon with plasma. In the
following cycle steps, the low frequency (absorption) is kept
at 375 rpm for just 30 s. For extraction, the frequency is
elevated again for a defined interval of time until reaching
2 lL of plasma. Figure 4d shows the average time for four
‘‘pumping’’ cycles of 2 lL metering of plasma each at different, elevated frequencies (extraction frequency). As
Fig. 4d shows, there is a minimum extraction time at
2,250 rpm for metering 2 lL of plasma within 45 s.
Similar to the continuous pumping mode, two following
opposing effects explain the occurrence of a valley seen
within this frequency range. On the one hand, the frequency must suffice to initiate the liquid extraction (dropping process). However, during this dropping process, the
liquid ‘‘stored’’ in the paper siphon before the crest point is
pushed backwards to the initial chamber. The higher the
extraction frequency, the larger the amount of liquid displaced back to the loading chamber. This backward
movement increases the wicking distance of the front of
liquid before the crest point. In the pumping protocol, the
time to saturate the paper siphon at low frequency is fixed
to 30 s, so the longer the wicking distance the less amount
of liquid wicked during this saturation time. Yet, using
such a pumping protocol, it is possible to establish optimal
conditions that in this case are: 30 s at 375 rpm and 45 s at
2,250 rpm for metering steps of 2 lL of plasma.
4.2 Colorimetric detection of triglycerides in whole
blood
The concept of paper-mediated siphoning and sampling in
hybrid lab-on-a-disc enables the integration of a
123
colorimetric quantification of triglycerides in plasma filtered from a whole blood sample. The centrifugal integration and automation of the triglyceride colorimetric
assay includes the following steps: blood separation, valving, plasma metering, mixing with the enzymatic
reagents, incubation and colorimetric readout. Following
the manufacturer’s specifications, 10 lL of plasma are
mixed with 150 lL of enzymatic reaction and incubated
for 15 min. Finally, the absorbance at 540 nm is measured
by a commercial spectrophotometer (Nanodrop 2000). We
use a disc configuration as the one showed in Fig. 2a
containing a blood separation chamber, a paper-based
siphoning channel and a reaction chamber where the colorimetric reaction takes place.
Figure 5a–d shows the detection sequence. First, a
120-lL volume of undiluted whole blood is loaded to the
separation chamber (Fig. 5a). In order to minimize its initial
direct contact with the paper-siphon strip, the blood is first
loaded by increasing the frequency in 15-rpm steps, up to
375 rpm, and centrifuged until the sedimentation chamber is
half full. Then, the frequency is ramped up to 6,000 rpm
where a pure plasma supernatant forms in less than 2 min
(Fig. 5b). The lower part of the paper-strip siphoning
channel has to stay permanently wet in order to extract a
reproducible amount of plasma in the following sampling
step. Once the pellet has formed (Fig. 5c), we decrease the
frequency to 375 rpm for 5 min during which the plasma
wicks and saturates the paper siphon. The point of saturation
is estimated to be reached once the accumulation of a larger
liquid volume at the outer siphon edge is observed.
As previously mentioned, the plasma metering has to be
carried out using the pumping mode. The 10-lL plasma
volume is sampled in 2-lL increments, following the
protocol as shown in Fig. 4d. So first 375 rpm frequency is
kept during 5 min to ensure the whole paper siphoning
channel is saturated with plasma (Fig. 5d). Afterwards, the
frequency is increased to 2,250 rpm for 45 s for the first
2-lL extraction. Then, four pumping cycles resulting in
extraction of 2 lL each are performed to retrieve the
remaining 8 lL (30 s at 375 rpm and 45 s at 2,250 rpm).
Finally, the 10 lL of plasma is mixed with 150 lL of
enzymatic reagent and incubated over 15 min (Fig. 5e).
Once the colour has developed the samples are removed
and absorbance is measured at 540 nm. Despite the fact
that the detection is carried out off-disc, there is a potential
for integration of a detection unit has been well documented (Steigert et al. 2006, 2007; Czugala et al. 2012;
Grumann et al. 2006).
The triglyceride levels of six different samples of a
female healthy volunteer (age group 25–35 years) were
measured using the protocol described above. To determine
absolute triglyceride content, the samples are compared
Microfluid Nanofluid
Fig. 5 a 120 lL of blood is loaded to the sedimentation chamber. In
order to minimize the direct contact between the blood and the paper
siphon, the blood is slowly loaded until approximately fill half of the
sedimentation chamber (steps of 15 rpm up to 375 rpm). The rest of
the loading occurs at 6,000 rpm, at a frequency high enough to pellet
red blood cells at the bottom of the sedimentation chamber and to
reduce the absorption by the siphon paper. b After 2 min at
6,000 rpm, the blood is separated between the cellular pellet at the
bottom and the liquid phase (plasma). c The frequency is decreased to
375 rpm and the plasma is absorbed through the paper siphon, after
5 min the complete paper siphon is saturated of plasma. d Once the
paper siphon is saturated, the pumping mode protocol is carried out in
order to extract 10 lL of plasma. There is a first switch to 2,250 rpm
for 45 s where the first 2 lL are extracted. Afterwards, there are four
complete cycles of 30 s waiting time (saturation) at 375 rpm and 45 s
at 2,250 rpm where the rest 8 lL are extracted up to the required 10
lL of plasma. e The extracted 10 lL of plasma are mixed with 150 lL
of enzymatic reaction, and after 15 min incubation, a bright purple
colour develops. For colorimetric detection at 540 nm, we extract
2 lL from the disc and use a commercial spectrophotometer
(Nanodrop 2000c). f Measurements of the triglyceride level (open
circle), the six measurements correspond with the same person and
different days. The calibration points (filled square) were measured
loading 100 lL of different concentrations of triglycerides and using
a continuous protocol for 10 lL (1,500 rpm for 210 s). The female
human healthy range is highlighted with a grey rectangle from 35 to
135 mg dL-1. Scale bars 6 mm
with the previously derived calibration curve (Fig. 5f) that
was obtained from seven known dilutions of the triglyceride standard supplied with the kit. 100 lL of each dilution is loaded using the same disc configuration as for the
triglycerides assay (Fig. 2b). However, in this case, for
metering 10 lL through the siphon paper, we use the
continuous mode protocol: five times the estimated time for
metering 2 lL (spinning at 1,500 rpm during 210 s as
shown in Fig. 4c once the paper strip was already wet).
According to the manufacturer of the colorimetric kit, the
healthy range regarding concentration of triglycerides in
blood is 40–160 mg dL-1 for males and 35–135 mg dL-1
for females. Figure 5f highlights a grey rectangle spanning
between 35 and 135 mg dL-1. The measurements displayed for our female volunteer are within the healthy
range. The variation between the measured values representing the same volunteer is attributed to the normal,
intra-day fluctuation of the biochemical composition of
blood. Thus, the time of the day when a blood sample was
drawn may affect the reading up to some extent.
5 Summary, conclusions and outlook
In this work, we have combined the capillary imbibition
inherent to paper with the rotationally controlled centrifugal field in order to significantly improve and extend
control and precision of liquid handling in paper and to
develop novel concept of fundamental LUOs in centrifugal
platforms. Among the advantages of the here described
system are the following:
•
Simplicity: The siphon channels are made of simple
paper strips placed on the disc.
123
Microfluid Nanofluid
•
•
•
•
•
Cost: The substrate is paper, which is an extremely
low-cost material and very appropriate for industrialscale manufacturing. Moreover, standard methods for
siphon valving typically involve additional hydrophilization (coating) processes such as plasma treatment or
chemical vapour deposition. For these back-end processes significant costs accrue.
Long-term stable and tuneable hydrophilic properties:
Capillary flow characteristics in paper are robust and
long-term stable. This long-term stability of paper
allows the storage of the microfluidic device over years
without degradation under reasonable storage conditions. Moreover, the capillary properties of the paper
siphon can be tuned just by changing the properties
such as porosity, composition, thickness and imbibition
behaviour of the paper.
Flow control: The integration of paper strips allows bidirectional flow. Compared to conventional, stationary
systems, the integration of paper strips introduces
rotationally controlled bi-directional flow.
Biocompatibility: Paper is a biocompatible material and
widely used in commercial lateral flow assays.
Other functionalities: The paper siphon can provide
filtering tasks, e.g. for the plasma extraction, or to store
dry reagents that may, for instance, be re-suspended
with the sample.
Based on this new siphoning mechanism, we developed a set
of laboratory unit operations such as plasma extraction and
metering to integrate and automate a colorimetric assay for the
detection of triglycerides on an inexpensive lab-on-a-disc
device. We envision paper fluidic elements may become
important components (e.g. valves, filters and storage) of highly
integrated centrifugal systems for processing various bioassays.
Furthermore, we will expand the concept of centrifugal paper
microfluidics towards increasingly complex assay protocols.
Acknowledgments This work has been supported in part by the FP7 ENIAC programme CAJAL4EU, Enterprise Ireland under Grant
No. IR/2010/0002 and the Science Foundation of Ireland (Grant No.
10/CE/B1821).
References
Al-Tamimi M, Shen W, Zeineddine R et al (2012) Validation of
paper-based assay for rapid blood typing. Anal Chem
84:1661–1668. doi:10.1021/ac202948t
Arnaud CH (2012) Paper devices move forward. Chem Eng News
90:1–4. doi:10.1039/c2lc40126f
Ballerini DR, Li X, Shen W (2012) Patterned paper and alternative
materials as substrates for low-cost microfluidic diagnostics.
Microfluid Nanofluidics :769–787. doi:10.1007/s10404-012-0999-2
Bascurt OK, Meiselman HJ (2003) Blood rheology and hemodynamics. Semin Thromb Hemost 29:435–450
Carrilho E, Martinez AW, Whitesides GM (2009) Understanding wax
printing: a simple micropatterning process for paper-based
123
microfluidics. Anal Chem 81:7091–7095. doi:10.1021/
ac901071p
Czugala M, Gorkin R III, Phelan T, Gaughran J, Curto VF, Ducrée J,
Diamond D, Benito-López F (2012) Optical sensing system
based on wireless paired emitter detector diode device and
ionogels for lab-on-a-disc water quality analysis. Lab Chip
12:5069–5078. doi:10.1039/c2lc40781g
Fridley GE, Le HQ, Fu E, Yager P (2012) Controlled release of dry
reagents in porous media for tunable temporal and spatial
distribution upon rehydration. Lab Chip 12:4321–4327. doi:10.
1039/c2lc40785j
Fu E, Liang T, Houghtaling J, Ramachandran S, Ramsey SA, Lutz B
(2011) Enhanced sensitivity of lateral flow tests using a twodimensional paper network format. Anal Chem 83:7941–7946
Fu E, Liang T, Spicar-Mihalic P, Houghtaling J, Ramachandran S,
Yager P (2012) Two-dimensional paper network format that
enables simple multistep assays for use in low-resource settings
in the context of malaria antigen detection. Anal Chem
84:4574–4579. doi:10.1021/ac300689s
Fulmer T (2012) Paper point of care. Sci Bus Exch 5:1–2. doi:10.
1038/scibx.2012.1021
Garcia-Cordero JL, Barrett LM, O’Kennedy R, Ricco AJ (2010)
Microfluidic sedimentation cytometer for milk quality and
bovine
mastitis
monitoring.
Biomed
Microdevices
12:1051–1059. doi:10.1007/s10544-010-9459-5
Ge L, Wang S, Song X, Ge S, Yu J (2012) 3D origami-based
multifunction-integrated immunodevice: low-cost and multiplexed sandwich chemiluminescence immunoassay on microfluidic paper-based analytical device. Lab Chip 12:3150–3158.
doi:10.1039/c2lc40325k
Godino N, Comaskey E, Gorkin R, Ducrée JD (2012a) Centrifugally
enhanced paper microfluidics. In: The 25th International Conference on Micro Electro Mechanical Systems, 29 January–2
February 2012, Paris, France, p 1017–1020
Godino N, Vereshchagina E, Gorkin R, Ducrée J (2012b) Hybrid
paper-polymer lab-on-a-disc for bioassay integration. In: Proceedings of the 16th International Conference on Miniaturized
Systems for Chemistry and Life Sciences (lTAS 2012)
0:1369–1371
Godino N, Gorkin R, Linares AV, Burger R, Ducrée J (2013)
Comprehensive Integration of homogeneous bioassays via
centrifugo-pneumatic cascading. Lab Chip 13:685–694
Gorkin R, Clime L, Madou M, Kido H (2010) Pneumatic pumping in
centrifugal microfluidic platforms. Microfluid Nanofluidics
9:541–549. doi:10.1007/s10404-010-0571-x
Grumann M, Brenner T, Beer C, Zengerle R, Ducrée J (2005)
Visualization of flow patterning in high-speed centrifugal
microfluidics. Rev Sci Instrum 76:025101. doi:10.1063/1.
1834703
Grumann M, Steigert J, Riegger L, Moser I, Enderle B, Urban G,
Zengerle R, Ducrée J (2006) Sensitivity enhancement for
colorimetric glucose assays on whole blood by on-chip beamguidance. Biomed Microdevices 8:209–214. doi:10.1007/
s10544-006-8172-x
Hwang H, Kim S-H, Kim T-H, Park J-K, Cho Y-K (2011) Paper on a
disc: balancing the capillary-driven flow with a centrifugal force.
Lab Chip 11:3404–3406. doi:10.1039/c1lc20445a
Jarujamrus P, Tian J, Li X, Siripinanond A, Shiowatana J, Shen W
(2012) Mechanisms of red blood cells agglutination in antibodytreated paper. Analyst 137:2205–2210. doi:10.1039/c2an15798e
Kitsara M, Ducrée J (2013) Integration of functional materials and
surface modification for polymeric microfluidic systems. J Micromech Microeng 23:033001. doi:10.1088/0960-1317/23/3/
033001
Kwong P, Gupta M (2012) Vapor phase deposition of functional
polymers onto paper-based microfluidic devices for advanced
Microfluid Nanofluid
unit operations. Anal Chem 84:10129–10135. doi:10.1021/
ac302861v
Liana DD, Raguse B, Gooding JJ, Chow E (2012) Recent advances in
paper-based sensors. Sensors 12:11505–11526. doi:10.3390/
s120911505
Lowe GDO, Barbebel JC (1988) Plasma and blood viscosity. Clin
Blood Rheol 1:11–44
Martinez AW, Phillips ST, Whitesides GM, Carrilho E (2010)
Diagnostics for the developing world: microfluidic paper-based
analytical devices. Anal Chem 82:3–10. doi:10.1021/ac9013989
Pelton R (2009) Bioactive paper provides a low-cost platform for
diagnostics. Trends Anal Chem 28:925–942
Schembri CT, Burd TL, Kopf-Sill AR et al (1995) Centrifugation and
capillarity integrated into a multiple analyte whole blood
analyser. J Autom Chem 17:99–104. doi:10.1155/
S1463924695000174
Shah P, Zhu X, Li C (2013) Development of paper-based analytical
kit for point-of-care testing. Expert Rev Mol Diagn 13:83–91
Siegrist J, Gorkin R, Clime L, Roy E, Peytavi R, Kido H, Bergeron T,
Veres T, Madou M (2009) Serial siphon valving for centrifugal
microfluidic platforms. Microfluid Nanofluidics 9:55–63. doi:10.
1007/s10404-009-0523-5
Songjaroen T, Dungchai W, Chailapakul O, Henry CS, Laiwattanapaisal W (2012) Blood separation on microfluidic paper-based
analytical devices. Lab Chip 12:3392–3398. doi:10.1039/
c2lc21299d
Steigert J, Grumann M, Brenner T, Riegger T, Harter J, Zengerle R,
Ducrée J (2006) Fully integrated whole blood testing by realtime absorption measurement on a centrifugal platform. Lab
Chip 6:1040–1044. doi:10.1039/b607051p
Steigert J, Brenner T, Grumann M, Riegger L, Lutz S, Zengerle R,
Ducrée J (2007) Integrated siphon-based metering and sedimentation of whole blood on a hydrophilic lab-on-a-disk. Biomed
Microdevices 9:675–679. doi:10.1007/s10544-007-9076-0
Vereshchagina E, Bourke K, Meehan L, Dixit C, Glade DM, Ducrée J
(2012) Multi-material paper-disc devices for low cost biomedical
diagnostics. In: The proceedings of the 26th International
Conference on Micro Electro Mechanical Systems, 20–24
January, Taipei
Yang X, Forouzan O, Brown TP, Shevkoplyas SS (2012) Integrated
separation of blood plasma from whole blood for microfluidic
paper-based analytical devices. Lab Chip 12:274–280. doi:10.
1039/c1lc20803a
123