1493
Electrophoresis 2013, 34, 1493–1509
Petr Smejkal1 Danny Bottenus2 Michael C. Breadmore1 Rosanne M. Guijt3 Cornelius F. Ivory2 ˇ Frantisek Foret4 Mirek Macka1 1 ACROSS
and School of Chemistry, University of Tasmania, Hobart, Australia 2 Voiland School of Chemical Engineering and Bioengineering, Washington State University, Pullman, WA, USA 3 ACROSS and School of Pharmacy, University of Tasmania, Hobart, Australia 4 Institute of Analytical Chemistry of the Academy of Sciences of the Czech Republic, v.v.i., Brno, Czech Republic
Received January 16, 2013 Revised March 6, 2013 Accepted March 7, 2013
Review
Microfluidic isotachophoresis: A review Electromigration methods including CE and ITP are attractive for incorporation in microfluidic devices because they are relatively easily adaptable to miniaturization. After its popularity in the 1970s, ITP has made a comeback in microfluidic format (-ITP, microITP) driven by the advantages of the steady-state boundary, the self-focusing effect, and the ability to aid in preconcentrating analytes in the sample while removing matrix components. In this review, we provide an overview of the developments in the area of -ITP in a context of the historic developments with a focus on recent developments in experimental and computational ITP and discuss possible future trends. The chip–ITP areas and topics discussed in this review and the corresponding sections include: PC simulations and modeling, analytical -ITP, preconcentration ITP, transient ITP, peak mode ITP, gradient elution ITP, and free-flow ITP, while the conclusions provide a critical summary and outlook. The review also contains experimental conditions for -ITP applications to real-world samples from over 50 original journal publications. Keywords: Chip / Isotachophoresis / Microfluidics / Miniaturization DOI 10.1002/elps.201300021
Additional supporting information may be found in the online version of this article at the publisher’s web-site
1 Introduction
Correspondence: Professor Mirek Macka, School of Chemistry, University of Tasmania, Private Bag 75, Hobart 7001, Australia E-mail:
[email protected] Fax: +61-3-62262858
Joule heating in large bore capillaries. The popularity of ITP decreased with the introduction of narrow-bore fused silica capillaries with an inner diameter of tens of micrometers, enabling Jorgenson to introduce CZE in 1981 [3, 4]. The resurgence of ITP in the 1990s was initiated by the need for a powerful preconcentration technique to increase the sensitivity in CZE, which was achieved by coupling ITP and CZE [5], later followed by the instrumentally less complicated transient ITP-CZE [6]. Microfluidic devices for CZE and MEKC separations appeared in the early 1990s with the introduction of the micro-total analytical systems (TAS) and Lab-on-a-Chip concepts [7], with first formal report of ITP on a chip was published in 1998 by Walker et al. [8]. However, the first ITP on a chip–concept-based separation was actually reported in 1975 by Bocek et al. who demonstrated the ITP separation of six carboxylic acids of Krebs cycle in a Perspex device [9] (a schematic overview of the ITP device is shown in Supporting Information Fig. 1). While the methods for fabrication of planar microfluidic chips were strongly limited at that time, the rationale of the advantages of a planar platform is identical to that behind the Lab-on-a-Chip concept introduced 23 years later. Advances in ITP over last two decades have been well monitored and are included in reviews by a number of groups
Abbreviations: FEKS, floating electrokinetic supercharging; FFITP, free-flow ITP; GEITP, gradient elution ITP; LE, leading electrolyte; -ITP, micro-ITP; TAS, micro-total analytical systems; TE, terminating electrolyte; t-ITP, transient ITP
Colour Online: See the article online to view Figs. 1, 3, 4 and 6–8 in colour.
Ninety years ago, Kendall and Crittenden introduced the “Ionic Migration Method,” separating rare earth metals and some simple acids [1]. The “Ionic Migration Method,” is fundamentally identical to the method known as ITP, the name introduced in the 1970s. In the early stages of development of electromigration methods, other separation methods based on ITP principles were introduced before the term ITP was established, such as the “Steady-State-Stacking” step in Disc Electrophoresis in the early 1960s [2]. The period of the greatest popularity of ITP was the 1970s, when this method gained fame in NASA’s third Skylab mission (http://history.nasa.gov/SP-401/ch12.htm; February 18, 2013). ITP could be performed in large bore capillaries (id hundreds of micrometers), which were not suitable for CZE due to excessive Joule heating. The self-sharpening effect between the boundaries of adjacent analyte zones counteracts diffusion and makes ITP resistant to band broadening due to
C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.electrophoresis-journal.com
1494
P. Smejkal et al.
Electrophoresis 2013, 34, 1493–1509
Table 1. Reviews and monographs offering some coverage of -ITP. The coverage is aimed to be exhaustive for the last 3 years (2010–2012) and also capturing primarily significant contributions through the last decade
Year Authors
Title
Pages
Refs Figs Tabs Ref.#
Comments (primary focus)
2013 Petr Smejkal, Danny Bottenus, Michael C. Breadmore, Rosanne M. Guijt, Cornelius F. Ivory, Frantiˇsek Foret, Mirek Macka 2012 Supreet S. Bahga, Juan G. Santiago
Microfluidic isotachophoresis: a review
17
150
14
4
This review
Coupling isotachophoresis and capillary electrophoresis: a review and comparison of methods
24
74
11
0
[10]
2013 Michael C. Breadmore, Alia I., Recent advances in enhancing the sensitivity of Shallan, Heide R. Rabanes, electrophoresis and Daniel Gstoettenmayr, electrochromatography in Aemi S. Abdul Keyon, capillaries and microchips Andras Gaspar, Mohamed (2010–2012) Dawod, Joselito P. Quirino
26
235
11
0
[11]
´ Petr Gebauer, 2012 Zdena Mala, ˇ Petr Bocek
Recent progress in analytical capillary isotachophoresis
10
89
5
0
[12]
2012 G. Garcia-Schwarz , A. Rogacs, S. S. Bahga, J. G. Santiago
On-chip isotachophoresis for separation of ions and purification of nucleic acids
8
18
5
2
[13]
2012 Alexander Stoyanov
IEF-based multidimensional applications in proteomics: toward higher resolution
10
160
1
0
[14]
2012 Yingying Wen, Jinhua Li, Jiping Ma, Lingxin Chen
Recent advances in enrichment 20 techniques for trace analysis in capillary electrophoresis
183
12
3
[15]
2012 Tomas Krizek, Anna Kubickova
Microscale separation methods for enzyme kinetics assays
11
100
3
2
[16]
2012 Fumihiko Kitagawa, Takayuki Kawai, Kenji Sueyoshi, Koji Otsuka
Recent progress of on-line sample preconcentration techniques in microchip electrophoresis Recent developments in CE and CEC of peptides (2009–2011)
9
56
8
0
[17]
Chip ITP areas discussed, sections: PC simulations and modeling, analytical ITP, preconcentration ITP, transient ITP, peak mode ITP, gradient elution ITP, and free-flow ITP Physical principles underlying ITP and the two standard modes of operation: “peak” and “plateau” modes including three movies Developments in the field of stacking, covering all methods from field-amplified sample stacking and large volume sample stacking, through ITP, dynamic pH junction, and sweeping Continuation of a series of regularly published reviews on the topic Physical principles underlying ITP and the two standard modes of operation: “peak” and “plateau” modes including three movies Alternative isoelectrofocusing methods and IEF-related techniques in protein analysis and characterization Enrichment techniques containing sample off-line and on-line preconcentration to enhancing sensitivity in CE for trace analysis Applications of separations in capillary or chip formats: MEKC, LC, GE, IEF, and ITP On-line sample preconcentration techniques to enhance the detection sensitivity in MCE
26
294
10
0
[18]
12 2011 Stacy M. Kenyon, Michelle M. Recent developments in electrophoretic separations on Meighan, Mark A. Hayes microfluidic devices
88
6
0
[19]
´ 2011 Petr Gebauer, Zdena Mala, ˇ Petr Bocek
84
0
0
[20]
´ ˇ 2012 Vaclav Kaˇsicka
Recent progress in analytical capillary isotachophoresis
C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
7
High-performance electroseparation methods in capillary and microchip formats: ZE, ITP, IEF, AE, EKC, and EC Electrophoretic concentration, sample preparation/conditioning, and separation on-chip Continuation of a series of regularly published reviews covering 2009–2010
www.electrophoresis-journal.com
Microfluidics and Miniaturization
Electrophoresis 2013, 34, 1493–1509
1495
Table 1. Continued
Year Authors
Title
Refs Figs Tabs Ref.#
Comments (primary focus)
22
269
12
0
[21]
14
75
6
0
[22]
Capillary electrophoresis
17
227
0
0
[23]
Stacking, covering all methods from field-amplified sample stacking and large volume sample stacking, through ITP, dynamic pH junction, and sweeping Main progress in electrophoresis techniques in order to achieve separation of NPs: CZE, GE, or ITP Anal. Chem. reviews series
Dynamic computer simulations of electrophoresis: a versatile research and teaching tool
29
215
19
0
[24]
Basic principles of electrolyte chemistry for microfluidic electrokinetics. Part I: acid-base equilibria and pH buffers Basic principles of electrolyte chemistry for microfluidic electrokinetics. Part II: coupling between ion mobility, electrolysis, and acid-base equilibria Micro free-flow electrophoresis: theory and applications
17
81
7
3
[25]
16
72
6
4
[26]
12
57
11
0
[27]
2011 Michael C. Breadmore, Mohamed Recent advances in enhancing Dawod, Joselito P. Quirino the sensitivity of electrophoresis and electrochromatography in capillaries and microchips (2008–2010) Electrophoretic methods for the 2011 Angela I. Lopez-Lorente, analysis of nanoparticles Bartolome M. Simonet, Miguel Valcarcel 2010 Nicholas W. Frost, Meng Jing, Michael T. Bowser 2010 Wolfgang Thormann, Michael C. Breadmore, Jitka Caslavska, Richard A. Mosher
2009 Alexandre Persat, Robert D. Chambers, Juan G. Santiago
2009 Alexandre Persat, Robert D. Chambers, Juan G.
2009 Ryan T. Turgeon, Michael T. Bowser
2008 James P. Landers (Ed)
Pages
Handbook of Capillary and 1592 Microchip Electrophoresis and Associated Microtechniques, 3rd edn., CRC Press, Boca Raton 2008
´ Petr Recent progress in capillary ITP 2007 Petr Gebauer, Zdena Mala, ˇ Bocek Miniaturised isotachophoresis 2006 Lin Chen, Jeff E. Prest, Peter R. analysis Fielden, Nicholas J. Goddard, Andreas Manz, Philip J. R. Day
–
–
–
[28]
7
96
0
0
[29]
14
61
7
3
[30]
2004 Christopher J. Evenhuis, Rosanne Determination of inorganic ions M. Guijt, Miroslav Macka, Paul using microfluidic devices R. Haddad
23
96
7
11
[31]
´ Masar, ´ Electrophoretic separations on 2003 Duˇsan Kaniansky, Marian ´ Zˇ uborov ´ ´ ´ Bodor, Maria a, chips with hydrodynamically Robert ¨ ´ Matthias Johnck, ¨ Eva Olveck a, closed separation systems Bernd Stanislawski
20
131
17
0
[32]
Software is available, which simulates all systems, including moving boundary electrophoresis, ZE, ITP, IEF, and EKC, and their combinations Fundamental and applied acid–base equilibrium chemistry useful to microfluidic electrokinetics—a tutorial review The importance of the coupling between electromigration and electrophoresis, acid–base equilibria, and electrochemical reactions—a tutorial review Separations using various modes such as zone electrophoresis, IEF, ITP, and field-step electrophoresis have been demonstrated Several chapters relevant to ITP, especially 5, 10, 11, and 13: “Online Sample Preconcentration for Capillary Electrophoresis” by Dean S. Burgi and Braden C. Giordano Summarizes the progress since 2002 The basic features of microchip-based ITP and its applications to the analysis and pretreatment of ionic compounds and biomolecules Inorganic analysis using microfluidic devices using ITP, CE, and hyphenated ITP-CE, together with a brief account of flow injection analysis ITP-ZE combination in trace analysis applications of the miniaturized systems is discussed in a broader extent
Refs = no. of references, Figs = no. of figures, Tabs = no. of tables, Ref.# = reference number in this review. GE: gel electrophoresis.
C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.electrophoresis-journal.com
1496
P. Smejkal et al.
as summarized in Table 1 [10–32]. This paper presents a detailed review of micro-ITP (-ITP). The following text contains a comprehensive overview of ITP techniques with a focus on the underlying principles. Computer simulations of ITP processes are reviewed in Section 2, followed by sections on analytical -ITP, preconcentration -ITP, transient ITP (t-ITP) on-a-chip, peak mode ITP [33], gradient elution -ITP, and free-flow -ITP [34, 35]. This review concludes with tables listing conditions to analyze or concentrate a variety of analytes and samples with a summary of the key experimental conditions extracted from original publications [36–72] provided in Supporting Information Tables 1–3.
2 PC Simulations and modeling Microfluidics and, specifically, Lab-on-a-Chip devices have shown themselves to be particularly amenable to simulation, often providing exquisitely accurate predictions of mass, momentum, and energy transport in both linear and nonlinear electrophoretic processes. This section will focus on simulations of -ITP, including those aimed at improving microfluidic applications by optimizing injection methods, minimizing sample loss, controlling fluid flow throughout the microchannel structure, and elucidating the impact of channel geometry and design on electrophoresis applications. In particular, various several dynamic simulators that model ITP will be discussed. Modern electrophoretic simulations were established in 1983 with the work of Bier et al. [73] who postulated a common basis for the four cardinal electrophoretic techniques: MBE, zone electrophoresis (ZE), IEF, and ITP. A number of recent reviews cover dynamic electrophoretic simulators including Thormann et al. [74], who gave a comprehensive overview of both public and commercial simulators developed over the past 30 years detailing applications, achievements, milestones, and capabilities, including moving boundary electrophoresis, IEF, ITP, ZE, and EKC, describing the important features of each package in table format. The following year, Thormann et al. [24] reviewed case studies, mostly those performed using the dynamic 1D solver, GENTRANS [73, 75–77], including an extensive list of electrophoretic phenomena such as the system peaks that occur in ZE, the gradient stabilization and destabilization processes that occur at the electrodes during IEF, the dynamics of interfaces between adjacent ITP zones, and the electrokinetic migration of neutral analytes with different binding constant that ranged from 0 to 100 000 L/mol with a charged buffer additive during EKC. This review covers simulation software specifically geared toward ITP in simple, linear channels. ITP is the separation and concentration of charged components in an electric field. ITP requires a leading electrolyte (LE) and a terminating electrolyte (TE) that have a higher effective electrophoretic mobility and a slower effective electrophoretic mobility than the sample components, respectively so that the sample components form focused/concentrated zones in order of elec C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Electrophoresis 2013, 34, 1493–1509
trophoretic mobility behind the LE and ahead of the TE. ITP can be difficult to simulate because of the mathematical stiffness that accompanies steep concentration gradients, for example, 10–100 mM over micron distances with high kilovolt per cm electric field strengths (kV/cm). In addition to ITP, we will also review general microfluidic protocols such as injection techniques to minimize sample loss, control of analyte migration through a microchip, preconcentration, and the influence of chip design on electrokinetic phenomenon. The simulation software packages most often used to model microfluidic ITP systems include the 1D solver GENTRANS [73, 75–77] available upon author request (
[email protected]), PeakMaster (only applicable to ZE) and SIMUL [78] both available at http://web.natur.cuni.cz/gas/, the 1D freeware SPRESSO [79] available at http://stanfordspresso.blogspot.com/, as well as commercially available multidimensional software such as CFD-ACE+ (ESI, Huntsville, AL, USA) and COMSOL Multiphysics (COMSOL, Inc., Burlington, MA, USA). Each program uses a different numerical solver; and the evolution of a variety of numerical solvers was recently assessed by Bercovici et al. [79]. In 2011, Mosher et al. [80] compared the dynamic 1D simulators GENTRANS, SIMUL, and SPRESSO based on their performance. Each simulator had its advantages and disadvantages, depending on the simulation problem, so the authors direct you to the aforementioned research article in choosing a 1D simulator for your particular project. The 1D electrophoresis simulation packages have progressed over the years to handle the various electrokinetic phenomena that occur during ITP. For instance, Bercovici et al. [79] included an adaptive grid mechanism in SPRESSO which limited the spatial geometry to regions that included high concentration gradients. This allowed grid points to accumulate in a small zone that moved with the LE interface, producing accurate solutions to ITP simulations while using less computational resources, that is, a 75-fold decrease in computation time compared to the same simulation using a dense, uniform grid. Since the mesh accumulated near the narrow analyte focused zone while dissipating from the broader shallow gradients in the LE and TE, the target zones could be analyzed in a moving reference frame which allowed for smaller working domains, sparser meshing overall, faster solution times, and more accurate solutions. The downside of the adaptive gird is when having multiple boundaries because the grid cannot adapt to all of them in an ideal fashion, thus reducing the performance benefit. The best result in terms of computation time is with a simple ITP boundary system with one step or one spike/peak (depending on the terminology being used). The mathematical equations, which assume that the acid–base reactions are instantaneous, pay particular attention to calculating total concentrations of species rather than individual ionic states; and boundary conditions that allow solution in a moving frame of reference were described in a later paper [81]. SPRESSO now includes an extended Onsager– Fuoss correction to determine effective electrophoretic mobilities, and an extended Debye–H¨uckel formula to calculate and www.electrophoresis-journal.com
Electrophoresis 2013, 34, 1493–1509
correct ionic activities [82] as well as modifying the governing equations to take into account variable cross-sectional area geometries in the context of a 1D geometry [83]. GENTRANS has been modified to handle ITP separations with a complexing agent [84] and a version of SIMUL (SIMUL 5 Complex) can simulate affinity ZE with a complexing agent [78]. SIMUL and SPRESSO can both calculate the pH of solutions made from multiple, general amphoteric compounds, including ionic strength corrections [25]. Persat et al. [25, 26] have also described basic principles of electrolyte chemistry for microfluidic devices including pH calculation, buffer preparation, ionic strength corrections, ion mobility corrections, acid–base equilibria, effects of atmospheric carbon dioxide, and electrode reactions which can be important parameters when simulating ITP systems. These 1D dynamic simulators and their modifications that have been made can provide insight into unexpected electrokinetic phenomena but, by their 1D nature, they do have limitations. 1D dynamic simulators cannot rigorously describe injection methods, sample loss and dispersion during electrokinetic migration past T-junctions, peak spreading due to nonuniform flows, effects, and/or concentration effects due to geometric nonuniformities. 2D and 3D numerical simulations, which are currently available in several commercial packages, offer the advantage of exploring all of these different phenomena and their application in these areas will be addressed in the remainder of this section. 2.1 Injection methods and sample leakage/loss at T-junctions The quantitative analysis of final concentrations relies on the ability to reproducibly inject a precise volume of sample into a microfluidic channel. This is typically accomplished by electrokinetic pumping or pressure-driven flow in a doubleT or dog-leg channel which includes four reservoirs, one separation channel, and two branched channels perpendicular to the separation channel (see Fig. 1). Jacobson et al. [85] described two injection strategies to introduce sample into a glass microfluidic chip which were termed (i) pinched sample loading and (ii) floating sample loading (see Fig. 1A and B). Experimental results indicated that pinched sample loading was superior to floating sample loading in both plug length and temporal stability. However, during pinched injection, sample loss can occur during voltage switching as sample gets pulled back into the side channels resulting in reduced sensitivities. Kurnik et al. [86] used simulation software from CFD Research (Huntsville, AL, USA), consisting of the ACE+, GEOM, and VIEW packages to demonstrate that backt-ITP, where the sample conductivity is much higher than the conductivity of a single background separation buffer, in a T-channel configuration that used “simplest injection” (see Fig. 1C) yielded a fivefold increase in concentration relative to “pinched injection.” CFD-ACE+ software was also used to simulate a sample injection method referred to as floating electrokinetic supercharging (FEKS) [87] which utilizes t-ITP coupled to ZE. C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Microfluidics and Miniaturization
1497
Figure 1. Electrokinetic sample injection and separation potential settings for negatively charged analytes using (A) pinched, (B) floating, and (C) simplest sample loading and separation strategies. The sample injection potential settings are shown in black and the separation potential settings are shown in red. HV is high voltage; BV is back voltage; FV is floating voltage; 0V is ground [86]. Back voltage refers to a small electrical potential applied between the perpendicularly opposed sample introduction channels and the separation cathode to prevent the diffusive “leakage” of sample ions into the separation channel during the separation (a so-called “pull-back” voltage) [129].
Initially, ITP was performed from reservoir P3 (cathode— set to ground) to reservoir P4 (anode), while reservoirs P1 and P2 were floating (Fig. 2). As the analytes passed the T-junction, the cathode voltage was switched from reservoir P3 to reservoirs P1 and P2 and resulted in the immediate transition from t-ITP to ZE. Any sample that was initially lost in the T-junction was quickly pushed back into the separation channel and gradually recovered to the sample plug. The simulation demonstrated a 99% recovery past the T-junction during FEKS compared to a ∼40% sample loss when only ITP was performed. In addition, experimental results demonstrated a 10× higher LOD using FEKS compared to conventional pinched injection. In a similar technique, Santiago and co-workers [88] experimentally demonstrated over a 106 -fold increase in concentration for a fluorescent dye using ITP focusing. On a side note, Bahga et al. [89] simulated and experimentally confirmed a clever way to couple ITP to ZE without voltage switching or buffer exchange. Instead, they used bidirectional ITP where a cationic ITP boundary migrates in the opposite direction of the anionic ITP boundary and, when they pass through, the mobility of the terminating ion www.electrophoresis-journal.com
1498
P. Smejkal et al.
Electrophoresis 2013, 34, 1493–1509
to the inertial forces produced by the inclined injection channel geometry. Although the authors did not explicitly report this, it appeared that dispersion after the T-junction was also reduced. 2.2 Dispersion during ITP due to flow effects—EOF and hydrodynamic flow ITP consists of a discontinuous buffer system where the LE is different than from the TE. As such, the EOF will be different in each region. Mismatched EOF can have a detrimental effect on ITP and can result in zone curvature and dispersion. Schonfeld et al. [93] used COMSOL Multiphysics to describe EOF-induced dispersion between both the LE interface and TE interface. When LE occupies most of the channel, zone dispersion forms a ⊃-shape at the interface; and when TE occupies most of the channel, zone dispersion forms a ⊂-shape at the interface. Schonfeld et al. [93] determined that the dispersion vanishes if the mismatch in electroosmotic mobilities of the LE and the TE is exactly balanced by their electrophoretic mobilities such that: Figure 2. Protocol of performing the floating electrokinetic supercharging (FEKS) which combines electrokinetic injection and ITP preconcentration of samples on the cross-microchip. Reproduced with permission from [87].
is changed thereby disrupting the ITP and ZE commences. This results in a much faster transition and less dispersion of the ITP zones as they evolve into ZE than other methods of transitioning from ITP to ZE. Cui and et al. [90] used COMSOL Multiphysics to simulate dispersion and sample loss as proteins migrated past a T-junction during ITP and ZE. Unlike in ZE where protein zones fail to recover due to dispersion at the T-junction, protein/sample zones during ITP do recover due to ITP’s selfsharpening effect. However, this did not address the problem of sample loss. To address sample loss, automated microelectrodes (changing voltages based on the total current passing through the microchannel) during ITP were inserted into the side channels to act as a valve to reduce dispersion and eliminate sample loss of proteins migrating past a T-junction [91]. This automated microelectrode system was numerically simulated in COMSOL Multiphysics and subsequently validated by experimental results. Instead of relying on microelectrodes that can be difficult to fabricate, the geometry of the microchip, in particular, the angle of the branched side channels can also be adjusted to minimize sample loss near T-junctions. Most microchips include side channels that are perpendicular, that is, at 90⬚ angles to the main channel, as is the case for conventional T-chips, which, as previously mentioned, can have significant sample loss. Tsai et al. [92] simulated sample loss during injection by electrokinetic pumping at different angled side channels. They found that for the classical perpendicular T-channel, sample losses of ∼25% occur, whereas only ∼4% sample losses occur at side channels with an angle of 30⬚ due C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
el EOF TE l =1 el EOF LE t
(1)
EOF where elt , ell , EOF TE , and LE are the electrophoretic mobilities of the terminating ion and the leading ion, and the electroosmotic mobilities of the terminator and leader, respectively. In addition, hydrodynamic flows can have a detrimental effect on ITP zones. Hydrodynamic flows are often necessary when attempting stationary counterflow ITP. Liu and Ivory [94] used COMSOL Multiphysics to describe stationary counterflow ITP in a 2D axially symmetric capillary and determined that molecules with a lower diffusivity had more severe dispersion than molecules with a higher diffusivity. Furthermore, dispersion due to flow effects could be reduced by the addition of a monolith or stationary phase to the capillary. In unobstructed capillaries, the flow is parabolic and leads to dispersion in combination with an oppositely migrating ITP zone (see Supporting Information Fig. 2), whereas in capillaries with a stationary phase, the flow is more plug-like and results in less dispersion and higher concentration (Liu, B. and Ivory, C. F., 2013, personal communication). Using this approach they were able to concentrate fluorescein by 34 000-fold over 8 h of continuous operation.
2.3 Preconcentration ITP inherently preconcentrates sample components by orders of magnitude into focused pure zones in decreasing order of electrophoretic mobility behind an LE and in front of a TE. Several techniques have been attempted to increase concentration factors further with ITP, such as longer channels/capillaries, using a hydrodynamic counterflow [95], and cross-sectional area reductions [96]. Increased concentration factors are necessary when trying to detect trace components www.electrophoresis-journal.com
Electrophoresis 2013, 34, 1493–1509
in the presence of high abundant species as is often the case in biological fluids such as urine, serum, saliva, etc., where the difference in analyte concentrations can span several orders of magnitude. Serpentine channels have been used to increase the channel length on a microfluidic platform without increasing the overall microchip dimension. Paschkewitz et al. [97] used the commercially available 2D finite volume method solver, CFDACE, to simulate ITP through a serpentine channel to show that, because of ITP’s self-sharpening effect, the dispersion that destroys ZE experiments in a curved channel [98] can recover after a distance that is dependent on the curve radius and sample mass in a straight section of channel [99]. An opposing hydrodynamic flow, hydraulic or electroosmotic, to electromigration during ITP is another way to increase the effective channel length and thus to increase analyte concentrations. However, as mentioned previously, parabolic flow patterns can lead to significant dispersion [94]. Breadmore [100] used GENTRANS to simulate the use of an electroosmosis counterflow for stationary ITP and found significant improvements in sensitivity. In addition, using electroosmosis as the counterflow can reduce dispersion seen with parabolic counterflows but can be difficult to control in situ (Liu, B. and Ivory, C. F., 2013, personal communication). A more elegant, alternative approach to preconcentration during ITP may be to incorporate cross-sectional area reductions along the axial length of the channel. As long as the ITP zone is in peak mode, a reduction in cross-sectional area will result in a proportional increase in concentration. This preconcentration strategy has been simulated both in a full 3D model using COMSOL Multiphysics [101] and in a diffusion-free 1D model using SPRESSO [83]. Both simulations revealed that decreases in cross-sectional area along the axial length of the channel resulted in increased sensitivities. However, only the 3D case simulation gave an accurate depiction of the sample shape, dispersion, and migration as sample zones moved into and through these cross-sectional area reductions. On the negative side, the 3D simulation did take weeks to complete. Simulating these ITP systems in more than one dimension requires significantly more computer memory and runtime. Using COMSOL Multiphysics for the 2D numerical simulations described in this section can take hours to days, while 3D simulations may require days to weeks or even longer, depending on the speed, memory, and power. Unfortunately, the computational time and resources needed for the reported simulations were not always provided in the cited papers.
3 Analytical ITP In 1998, Walker et al. revisited the early work by Bocek et al. and presented a constant voltage ITP separation of herbicides (paraquat and diquat) on a single-channel chip using Normal Raman Spectroscopy (2 W, 532 nm NdYVO4 excitation laser) for detection [8]. The curved, 40-m wide, 75-m C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Microfluidics and Miniaturization
1499
deep, and 210-mm long channel was fabricated on a glass microscope slide, and then sealed with a 120-m thick glass cover slip. The sample was electrokinetically injected into the chip channel prefilled with LE (sulfuric acid) by applying constant voltage 4 kV for 1–2 min. After rinsing and refilling the sample reservoir with TE (Tris), the separation was run under HV (high voltage) (at 8 kV). In 1999, the team of Prest and Baldock reported the separation of a mixture of sodium and potassium by ITP (constant voltage) on a PDMS chip [102, 103]. The chip was used as a single-channel device despite the double-T injection geometry design. A platinum–iridium wire was integrated between the PDMS layers for single electrode conductivity detection. Sample was mixed with TE in the TE reservoir and the separation was carried out by applying a constant voltage of 1 kV between the reservoirs. The same approach was used for ITP separation of four metal cations (lithium, lanthanum, dysprosium, and ytterbium) in half the time required in a capillary system [104]. In 2002, a new design was introduced for bidirectional ITP [105]. The sample was hydrodynamically injected from the middle of the cross-geometry separation channel in the PMMA chip with detection electrodes of a dual conductivity detector positioned at each end. A hydrodynamic system using pressure and valves was used to introduce all electrolytes into the chip before a constant voltage of 1000 V, was applied to drive the ITP separation. From the injection cross in the center of the channels, anions (Cl− , NO3 − , SO4 2− , F− ) migrated toward the anode and cations (Cs+ , NH4 + , Na+ , Li+ ) migrated toward the cathode. In 2003, Baldock et al. presented the first fully polymeric microfluidic chips [106]. Injection-molded electrodes were made from conductive polymers (8% carbon black-filled polystyrene, 40% carbon fiber-filled nylon, and 40% carbon fiber-filled polystyrene) for voltage application and detection in a Zeonor/polystyrene device. The chips used simple crossgeometry structure fitted with four reservoirs, connected with valves for fluidic control. After filling the separation channel with LE and TE upstream and downstream from the injection cross, respectively, sample was injected between LE and TE by opening valves at sample and waste reservoirs. During ITP analysis, all valves were closed and constant voltage/current was applied between the polymer separation electrodes. In 2003, Prest et al. fabricated a new PMMA microfluidic chip for ITP with hydrodynamic injection controlled by pressure and valves and with conductivity detection (shown in Supporting Information Fig. 3) [42]. Similar to previous reports, the method was faster than capillary ITP, but not sensitive enough for the analysis of industrial waste stream samples. This chip design was used repeatedly [36–41,68,69], and the applications ranging from inorganic ions to amino and organic acids are summarized in Supporting Information Table 1. In 2004, Baldock et al. introduced a pressure-driven system for the injection of variable amounts of sample between LE and TE [107]. The separation channel was connected to a U–shaped sample-waste channel by a 1-mm long channel at a 45⬚ angle. The separation channel was filled with www.electrophoresis-journal.com
1500
P. Smejkal et al.
Electrophoresis 2013, 34, 1493–1509
Figure 3. Cascading -ITP chip with fluorescent detection. (A) PMMA microfluidic chip and geometrical dimensions showing a pair of orthogonal reductions (a width and a depth reduction) in the cross-sectional area along the major separation axis of the channel. The microchip includes a tee channel between the sample reservoir and the anode reservoir in order to control the sample load volume. (B) Representative images of three sequential ITP separations performed in the microfluidic chip after the proteins, labeled cardiac troponin I (blue) and R-phycoerythrin (red), have migrated past the two cross-sectional area reductions and fully concentrated. The protein bands have been aligned horizontally in these images [115].
LE and TE downstream and upstream from the intersection, respectively, before sample was injected from the U-shaped channel between the LE and TE. As illustrated in Supporting Information Fig. 4, the 45⬚ angle reduced leaking of the sample after separation current (30 A) was applied. The system that was equipped with a conductivity detector showed good repeatability for analysis of a mixture of ten metal ions (Ca2+ , Mg2+ , Mn2+ , Co2+ , Ni2+ , Zn2+ , La3+ , Nd3+ , Gd3+ , Cu2+ ). In 2006, Prest et al. presented a new double-T device micromilled in PMMA for the separation and quantification of Cl− under 100 s with an LOD 2.2 mg/mL and a throughput of 20–30 samples per hour could be achieved [108]. Unlike in CE, where indirect detection (primarily photometric) has been explored extensively [109, 110], in ITP indirect fluorescence detection with a fluorescent counterion first shown by Reijenga in 1984 [111] remained unutilized until relatively recently. The Santiago group has explored systematically indirect fluorescence detection options. Fluorescent spacers were used for the detection of ITP zones of nonfluorescent analytes using a single channel commercially available Caliper chip (NS-95) [112]. The chip was filled with LE from one reservoir and sample was hydrodynamically loaded from the opposite reservoir. After injection, the sample reservoir was replaced with TE. The separation of two amino acids was visualized using three fluorescent spacers (Oregon green carboxylic acid, fluorescein, and bodipy). The isotachophoregram contained three peaks for the three fluorescent dyes with the length of the space between these peaks related to the concentration of Ser and Phe. A fourth tracer (FITC) was added to enable the inclusion of a third analyte, illustrating a limitation of this approach. Bercovici et al. used fluorescently
C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
labeled ampholytes designed for IEF as spacers for the indirect detection of 2-nitrophenol and 2,4,6-trichlorophenol by -ITP [113]. A hand-held ITP instrument (mass 240 g, powered by a laptop through an USB port) was developed for the analysis of explosives by -ITP with indirect fluorescence detection [43]. The instrument used glass microfluidic chips (25 × 15 × 2.2 mm) with simple cross-geometry and fluorescently labeled ampholytes enabled the detection of 2,4,6trinitrophenolate and 2,4,6-trinitrophenol and the herbicide 2,4-dichlorophenoxyacetic acid in a river water. A fundamentally different approach from the same group was presented by Chambers et al. using a fluorescent counterion tracer in LE (rhodamine 6G) or a fluorescent co-ion tracer in the TE (Alexa Fluor 488) [114]. Individual zones can be visualized because the concentration of the fluorescent dye will adjust to the local electrical field in each ITP zone. In 2011, Bottenus et al. presented a chip containing cascading channels with reductions in width and in depth in different parts of the separation channel for an ITP separation at constant voltage with fluorescent detection of R-phycoerythrin and fluorescently labeled human cardiac troponin I [101, 115]. The chip contained reservoirs for LE (cathode), sample, and TE (anode) joined through cascading separation channel decreasing from 1-mm wide and 100-m deep to 100-m wide and 20-m deep toward the cathode. The sample reservoir was connected to the 1-mm wide and 100-m deep section using a channel of the same dimensions. Pressure was used to fill the chip with LE before loading sample dissolved in LE into the sample reservoir, filling the wide channel directing toward the anode reservoir. The
www.electrophoresis-journal.com
Electrophoresis 2013, 34, 1493–1509
Microfluidics and Miniaturization
1501
Figure 4. In-house designed -ITP chips compatible with the Agilent Bioanalyzer. (A) Glass chips contained four different channels, which enabled testing in different ITP conditions [116]. (B) Dry film photoresist chip designed for ITP separation and quantification of carboxylic acids from human serum (LE: leading electrolyte; TE: terminating electrolyte; S: sample; V: reservoir for applying the vacuum) [49]. Dimensions are in millimetres.
sample reservoir was sealed with sticky tape to prevent hydrodynamic flow while replacing the anode reservoir with TE. A 10 000-fold increase in the concentration of troponin and R–phycoerythrin was obtained in these devices (scheme in Fig. 3 and fabrication in Supporting Information Fig. 5). Later, a slightly modified cascade chip was used for detection of human cardiac troponin I in human serum [44]. Minor changes of chip design improved the sensitivity and minimized problems with hydrodynamic flow during sample injection and loading filling of the anodic TE reservoir. In 2012, Smejkal et al. developed an ITP method for indirect fluorescence detection of benzoate in soft drinks using an Agilent glass DNA chip and Agilent Bioanalyzer 2100 with indirect fluorescence detection [48]. The chip was filled with LE by using pressure; then sample and TE were loaded into reservoirs and, in a single step, electrokinetically injected into the separation channel. The ITP separation was started by applying a constant current between reservoirs with TE and LE while electrodes at all other reservoirs were set to 0 A. The method LOD for carboxylic acids was in millimolar range. Benzoate in soft drinks was quantified by standard addition. The same chip was used to analyze lactate in serum by -ITP [116]. Due to the high ionic strength of the sample, the injection of sample between the TE and LE was achieved by applying vacuum rather than electrokinetically. In the same work, a glass chip was designed for -ITP in the Agilent Bioanalyzer 2100 (Fig. 4A). Four different length separation channels (25.7, 34.5, 62.8, and 125.8 mm) were compared and the shortest channel was found to be sufficient to reach steady state. Smejkal et al. further improved the design using numerical modeling and fabricated the Bioanalyzer-compatible microchip in dry film photoresist (Fig. 4B) [49]. The separation channel comprised wide (300 m, injection and separation) and narrow (30 m, detection) channels. The wide channel allowed injection of large amounts of sample, while the low resistivity of the wide channel enabled use of a higher constant separation current and hence accelerated the analysis. The narrow channel was incorporated to elongate the length of the ITP zone and improve quantitation. LE, TE, and sample were injected in a sin C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 5. -ITP-CE chip developed by research group of Kaniansky made in PMMA. (A) The bottom substrate contained channels in a PMMA block (30 × 70 mm). The volume of terminating electrolyte (TE) reservoir was 8.8 L. Chip contained two sample loops SL1 (9.2 L) and SL2 (0.96 L). SC1 is preseparation column and SC2 analytical column with volumes of 1.2 and 1.7 L, respectively. The separation column splits in two at bifurcation point (BF). The volume of leading electrolyte reservoirs (LE1 and LE2) was 8.8 L. D1 and D2 indicate the position of the conductivity detectors. (B) The cover plate contained platinum electrodes for detector 1 (D1a/b) and detector 2 (D2a/b) as well as the HV electrodes to drive the separations (HV). Reproduced with permission from [119].
gle step by applying vacuum to the waste vial. The analysis of lactate from human serum was conducted in less than 1 min, with an LOD for lactate of 24 M [49]. The synergy of the technical strengths and advantages of a ready-to-use commercial field-deployable instrument (Agilent Bioanalyzer), combined with the research flexibility of in-house designed chips fitted within the instrument, has proved to offer practical benefits, and is an exciting prospect with the potential to diminish barriers preventing a wider spread and success of microfluidics [117].
4 Preconcentration ITP The self-focusing effect makes ITP an ideal preconcentration technique, for example, when combined with a CZE separation. A significant number of contribution in ITP-CZE [5] have been made by the group of Kaniansky. The initial work using chips capable of ITP, ITP–ITP, and ITP-ZE was published in 2000 [118], with chips made in PMMA by hot embossing [119]. The initial design was optimized by straightening the channels to prevent trapping of bubbles formed by electrolysis. While both designs were successfully used for ITP separation of organic acids, well-developed ITP zones were obtained in the updated design despite longer analysis times. The device shown in Fig. 5 was applied to a range of samples [45–47, 50–52, 70–72] as summarized in Supporting Information Table 2. In 2002, Wainright et al. used the ACLARA LabCard chips for ITP-ZE separation of 13 eTagTM fluorescent markers www.electrophoresis-journal.com
1502
P. Smejkal et al.
Figure 6. Schematic of the ITP-ZE chip with five reservoirs. Empty chip (A) is filled with LE (B) after all channels are filled with LE (capillary effect, pressure or vacuum), TE is loaded in reservoir 2 (R2) and sample in R1. Sample is injected by applying voltage between R1 and R3 (C). TE is injected by applying voltage between R2 and R1 (D). ITP is initiated by applying voltage between R2 and R5 (E). When the ITP pass the junction with R4 channel, the ZE is started by applying voltage between R4 and R5 (F) [53].
and for the detection of the activity of surface protease ADAM 17 of live THP-1 cells (human monocytic cells) [53]. The injection of electrolytes is illustrated in Fig. 6. Using a confocal fluorescence microscope, the activity of ADAM-17 was analyzed by -ITP-ZE of a fluorescent peptide that loses its fluorescence once cleaved by the protease. The method sensitivity was ten cells (THP-1) in 10 L of Hanks’ buffer and could be used for direct evaluation of surface protease activity or for indirect determination of a number of cells. Later, the same chip was used under similar conditions for -ITP-ZE of dsDNA fragments (HaeIII DNA digest) [120]. In 2005, Huang et al. described a glass microfluidic chip for on-line ITP gel electrophoresis (GE) of four FITC-labeled SDS-denatured proteins [121]. The chip architecture and the protocols for its use were similar to the design shown in Fig. 6 with the addition of one reservoir for simultaneous vacuum injection of TE and BGE. Mixture of a sample and LE (50 mM Tris, 0.5% SDS, 2% DTT, pH adjusted with HCl to 6.8) was injected electrokinetically between TE (192 mM glycine, 25 mM Tris, pH 8.3) and BGE (100 mM TrisNaH2 PO4 , 0.1% SDS, 10% glycerol, 10% dextran, pH 8.3). The channels were coated with linear polyacrylamide to prevent EOF and protein absorption. In 2006, Liu et al. used a similar chip for Hepatitis B virus genotyping and was used C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Electrophoresis 2013, 34, 1493–1509
for 30 cycles PCR in serum samples, where PCR products were analyzed by an ITP-ZE microfluidic chip, and the results compared favorably with those of conventional method [54]. In 2006, Ma et al. used a quartz chip again of a similar five reservoir geometry for ITP-ZE with UV detection of the flavonoids quercetin and isorhamnetin [122]. TE (50 mM H3 BO3 , 20% CH3 OH, pH 8.2) and BGE (25 mM MOPS, 50 mM Tris, 55 mM H3 BO3 , pH 8.36) were injected simultaneously by applying vacuum before the BGE in the sample reservoir was replaced with sample in LE (10 mM HCl, Tris, 20% CH3 OH, pH 7.2). The sample was electrokinetically injected before starting the ITP-CZE. Jung et al. demonstrated a concentration factor of 2 × 106 by the ITP-ZE separation of Alexa Fluor 488 and bodipy on a borosilicate glass microfluidic chip with a simple cross-geometry without replacing the sample solution in TE reservoir [88, 123]. The LE and sample mixed with TE were loaded into reservoirs at either end of the separation channel. One of the side reservoirs was filled with LE and vacuum was applied to the other one. Then HV was applied between the reservoirs of the main separation channel and ITP was initiated. The ITP plateau zones positions were monitored using a CCD camera mounted on a fluorescent microscope. Later, the HV was applied between the reservoirs containing LE to initiate t-ITP followed with ZE. In 2008, Hirokawa et al. used a Shimadzu MCE 2010 MCE system with UV detection for the ITP-GE separation of 16 DNA fragments ranging from 50 to 800 bp using a simple cross-geometry chip using a method they named FEKS [87]. In the method, electrokinetic injection and ITP preconcentration of samples was performed in a separation channel on a cross-geometry microchip. At these two stages, the side channels crossing the separation channel were electrically floating. After the ITP-focused zones passed the intersection, a potential was applied to the side arms to dissipate the ITP by injection leading ions, changing rapidly from ITP to GE in hydroxymethylcellulose as sieving matrix. Immunoassays are expensive due to the high price of antibodies (especially labeled ones). The use of a preconcentration step like ITP to increase the sensitivity of these assays could therefore significantly reduce the cost of existing immunoassays techniques. ITP was combined with an electrokinetic analyte transport assay by Kawabata et al. [55]. The complex chip design comprised a main separation channel connecting a TE and LE reservoirs with eight side arms: five reservoirs with sample and reagents and three vacuum ports for simultaneous loading of all reagents. After loading, the separation channel was filled with subsequent zones of TE, antibody in TE, antigen (sample) in TE, labeled antibody in LE and LE. When the field was applied, the antibody mixed with the antigen and concentrated by ITP. This first immune reaction product was transported into the second, labeled antibody zone and a sandwich antibody complex was formed and also focused into a narrow zone by the ITP. Once the complex passed the fifth side channel filled with LE, the potentials were switched and the complexes were separated by www.electrophoresis-journal.com
Electrophoresis 2013, 34, 1493–1509
GE in LE before LIF detection of the fluorescently labeled complex. Using a similar method, Park et al. used quartz chips from Caliper Life Sciences, the instrument Caliper42, and a modified version of LabChip 90 for an immunoassay using ITP-GE [56] with a 200-fold increase in sensitivity in comparison with a conventional assays. Four proteins (transferrin, 2-microglobulin, HSA, and Ig G) were detected in a human urine sample [57]. The PMMA device was fabricated for on-chip desalting and analyte preconcentration before CZE separation and UV quantitation of urinary proteins to provide an alternative for time-consuming sample pretreatment procedures and increase the detection sensitivity. Proteins focused by ITP were subsequently transferred into a separation capillary embedded in microchip for CZE separation and UV quantitation. On the chip, a 21-mm long sample zone is electrokinetically loaded between a TE and LE and the proteins are concentrated into a narrow zone while the small ions migrate into a waste reservoir. Through switching of the potentials, this sample zone is transferred into the fused silica capillary where it is separated by CZE before UV detection. Persat et al. used classical ITP to concentrate nucleic acids from a whole blood lysate for subsequent PCR [58]. The borosilicate chip with simple cross-geometry chip used for this work contained 20-m deep and 90-m wide channels (model NS12A, Caliper). To reduce protein absorption, the channels were treated with the silanizing agent, Sigmacote. The LE and TE contained SYBR green to visualize separated DNA and Triton X-100 to further reduce EOF and protein absorption. The chip was first filled with LE, then one of the side reservoirs was filled with the blood lysate and a vacuum was applied to the TE reservoir. This reservoir was then filled with TE and a voltage was applied in the main separation channel between the TE and LE reservoirs. The process of ITP focusing of DNA was observed with an inverted epifluorescent microscope. When the DNA reached the LE reservoir, the ITP was stopped and the content of this reservoir was used for PCR. This approach was simplified and applied to the concentration of bacterial 16srRNA by Rogacs et al. [124]. As illustrated in Fig. 7, a single-channel device was filled with LE before a bacterial lysate in TE was loaded into the sample reservoir. The nucleic acids were extracted from the sample and concentrated by ITP before being processed off-line by qPCR. Electromigration methods offer the principal advantage of open channels as opposing to chromatography, thus enabling separations of particles including whole cells [125]. Phung et al. [126] developed highly sensitive capillary ITP method for microbial analysis using LIF detection after cells were stained with the universal nucleic acid fluorophore SYTO 9. An LE of 50 mM Tris-HCl was used while the cells were diluted in 5 mM Tris HEPES as the terminator. This allowed analysis of Escherichia coli bacteria with an LOD of 14 cells in a sample volume of 100 L, or 1.35 × 102 cell/mL, which is 47 times lower than reported by CE-LIF and 148 times lower than CE-UV with on-line concentration.
C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Microfluidics and Miniaturization
1503
5 Transient ITP t-ITP in capillaries was introduced in 1992 by Foret et al. [6], although the principle of this type of sample preconcentration had been occasionally applied since the late 1980s. The principle of t-ITP involves the establishment of a temporary ITP stage in a ZE separation that is used to concentrate sample components before the ITP is dissipated and the components separated by ZE [127, 128]. In 2003, Kurnik et al. used computer simulations to study the separation of electrokinetically injected fluorescein inside a chip with a double-T crossgeometry for t-ITP-ZE separation using electrokinetic sample injection [86]. This study described the effects of pinch-andpull-back currents and currents while using t-ITP for sample focusing. The predicted improvement in sensitivity using floating electrodes was experimentally confirmed and could be explained by the sample loss when the pinch-and-pull-back method was used. Practical experiments were done with fluorescein, a tyrosine kinase assay using a FITC-labeled peptide substrate, and 12 eTagTM fluorescent reporter molecules. In the same year, Vreeland et al. described the t-ITP-ZE separation of fluorescent eTagTM reporters in a commercial microfluidic PMMA chip with a double-T cross-geometry [129]. In 2003, Xu et al. used the Shimadzu MCE 2010 with a single-channel device for separation of six SDS-denatured proteins followed by UV detection [130]. Prior to the analysis, the channel was filled with LE and sample was loaded into the TE reservoir and electrokinetically injected. The sample was then manually replaced with TE, forming the ITP system. The voltage was switched off and the TE was manually replaced with BGE (same composition as LE without dextran), creating a t-ITP system concluding in GE when the voltages were re-applied. All times to switch reagents were determined experimentally (optimized conditions were sample injection for 15 s and TE injection for 30 s). The same method was applied for separation of a standard DNA ladder consisting of 16 DNA fragments in a range of 50–800 bp [131], and for analyzing DNA fragments obtained after 30 PCR cycles [59]. Jeong et al. [132] described a sensitive method for t-ITPZE of fluorescein and 2.7-dichlorofluorescein in highly saline samples (250 mM NaCl). An LOD of 3 pM was obtained for both analytes using a PDMS treated with neutral fluorocarbon surfactant to suppress the EOF suppressor. The chip was initially filled with BGE containing TAPS as a terminating anion. Three different double-T geometry devices were used, with serpentine injection plugs of 12-, 20-, and 28-mm long, respectively. The high salinity sample was electrokinetically transported through the serpentine connecting the sample with the waste reservoir, judging the injection time by the current as a steady current indicates that steady state has been reached. The t-ITP separation was initiated by applying voltage between reservoirs at either end of the separation channel, each filled with BGE. When the chloride zone, acting as leading ion, dissipates the analytes are separated by CZE. t-ITP coupled with GE was used for the analysis of fluorescently labeled HSA and its mAb immunecomplex [133], as
www.electrophoresis-journal.com
1504
P. Smejkal et al.
Electrophoresis 2013, 34, 1493–1509
Figure 7. Scheme of bacterial RNA extraction and purification from whole human blood using -ITP. Reproduced with permission from [124].
well as for the analysis of immunecomplex of FITC-labeled BSA with specific mouse antibody [134]. In 2008, Lin et al. presented a chip for separation of dsDNA from PCR reaction [135]. While the authors described the method as ITP-GE, this is actually t-ITP-GE because of the transient nature of the ITP step. All the channels in the PMMA chip were 20-m deep and the functionalities can be divided into three parts. First, the “injection” part contained four reservoirs (TE, sample, LE, and vacuum) joined with a 500-m wide and 10-mm long channel. Second, the “focusing” (stacking) part containing a 140-m wide and 20-mm long channel ending at the junction with the separation channel and the first waste reservoir. Third, the separation channel, also 140-m wide and with a length of 25 mm. Prior to the separation, the chip was hydrodynamically filled with LE and the sample and TE reservoirs were filled with sample and TE, respectively. Then negative pressure was applied to the vacuum reservoir, and zones of TE and sample were introduced between zones of LE in the injection channel. The t-ITP separation was initiated by applying a voltage between the LE reservoir and waste reservoir 1. Once steady state had been reached, the HV was switched to the second waste reservoir for the separation by GE after dissipation of the ITP system. A number of groups have used t-ITP-GE for DNA analysis in cross- or double-T geometry devices [60, 61, 136]. In 2009, Liu et al. showed more than a 100-fold improvement in sensitivity as a result of using t-ITP [62]. Technique called gradient elution ITP (GEITP) was introduced in capillary format in 2007 by Shackman et al. [137] and later applied to a microchip by Davis et al. [138]. Because of the small scale of the capillary separation (a few centimeters) it is also included in this review. The instrument consisted of two reservoirs connected by a 3-cm long capillary with a 5 mm detection window for LIF in the center. The larger, electrically grounded reservoir (1.4 mL) contained LE and could be pressurized. The smaller reservoir (110 L) contained a high-voltage electrode and was filed with sample in TE. Once the HV was applied, the ITP separation was initiated in the smaller reservoir because the pressuredriven flow of LE prevented the separated ions zones from entering the capillary. By gradually reducing the hydrodynamic flow, analytes could sequentially enter the capillary once the counterflow was below the effective electrophoretic C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
velocity of the analytes. The principle is schematically illustrated in Supporting Information Fig. 6. The device was used for the separation of fluorescent dyes (fluorescein, 6-carboxyfluorescein), fluorescein-labeled DNA, fluorescently labeled (5-carboxyfluorescein) mixture of amino acids (Asp, Gly, Ser, Val), and for the separation of natively fluorescent proteins (green fluorescent protein, DsRed). The separated zones could be detected as peaks because of the use of nonfluorescent ampholytes to space the plateaus. Lowpicomolar detection limits were reported despite the use of a mercury arc lamp and a low-cost CCD camera. This instrument was also used for chiral separation of amino acids [139]. Later, Mamunooru et al. incorporated a UV detector for the separation of the UV absorbing amino acids tyrosine and tryptophan [140]. This work was followed by the analysis of chromophore labeled amino acids (Asp, Glu) in cerebrospinal fluid [67]. Davis et al. coupled GEITP with gradient elution transient ITP-CZE of six fluorescently labeled amino acids (Asp, Glu, Gly, Ala, Ser, and Val) using an 11 cm capillary with a 6 cm detection window [138]. GEITP was used to selectively allow the analytes zones to enter the capillary, then the pressure and voltages were stopped to replace the TE in the TE reservoir with LE. When the pressure and voltages were resumed, a t-ITP system enabled refocusing of the diffused zones. After dissipation of the ITP, the analytes were separated, allowing femtomolar detection limits. This work was transferred to a chip, but unfortunately the sensitivity of the chip system was not as good as demonstrated in the capillary.
6 Peak mode ITP When present in an electric field between an LE and TE, minority analyte ions from the sample also form concentrated zones. Based on Kohlrausch’ regulating function, the maximum concentration of a sample zone is limited by its electrophoretic mobility and LE concentration. Once this limiting concentration is reached, the analyte zone forms a plateau separating the LE and TE co-ions. For trace analytes, the limiting concentration is not reached, and mixed ITP zones with widths of the interface width between adjacent zones are formed [141, 142]. The term “peak mode ITP” specifically www.electrophoresis-journal.com
Electrophoresis 2013, 34, 1493–1509
describes the ITP technique where analytes are concentrated between two zones to a maximum concentration below their steady state limit. The Santiago group exploited peak mode ITP to concentrate trace amounts of analytes [33]. The two most important differences between peak mode ITP and classical, plateau mode ITP are in detection and quantification of the analytes. In plateau mode ITP, a universal detector is used to detect the stepwise changes between zones, using the step length as a measure of analyte concentration in the initial sample. In peak mode ITP, an analyte-specific detection technique is used to selectively measure a specific analyte. Because the limiting concentration is not reached, the signal will be a peak, not a step, where the peak intensity is a measure of the analyte concentration. In 2009, Khurana et al. presented peak mode -ITP with continuous sample injection (sample present mixed with TE) for the analysis of fluorescently labeled DNA and natively fluorescent proteins (green fluorescent protein and allophycocyanin) [143]. Peaks formed according to the analyte mobilities between nonfluorescent plateau zones of LE, carbonate, carbamate, and TE; the carbonate and carbamate were present as impurities in the TE. A simple, cross-geometry, glass Caliper NS-95A chip was used as a single-channel device. While demonstrating the potential of peak mode ITP, the method was not really applicable for quantification due to the limited separation capacity and the contamination with carbonate and carbamate. Schoch et al. described peak mode -ITP with fluorescence detection for the extraction, isolation, preconcentration, and quantification of micro-RNAs (miRNAs) [63]. miRNA is an 18–24 nucleotide long noncoding oligonucleotide which regulates gene expression via sequence-specific interactions with mediator RNA. Again a Caliper NS-95 chip was used as single-channel device. In preparation for the analysis, the chip was filled with LE (100 mM HCl, 140 mM 6-aminocaproic acid, pH 4) and placed for 10 min at 4⬚C. Subsequently, TE reservoir was filled with LE containing 30% Pluronic F-127 and the separation channel was filled with LE by applying the negative pressure to LE reservoir. After 10 min at ambient temperature, the viscosity of Pluronic increased to form the sieving matrix and a mixture of the RNA sample in TE was loaded in TE reservoir. The separation was started by applying the HV between TE and LE reservoirs and miRNA could be quantified from ∼900 cells in ∼5 L. Persat et al. continued Schoch’s research into quantification of miRNA and introduced selective ITP using three LEs in a more complex Caliper NS260 chip (eight reservoirs) [64]. The sample of miRNA was mixed with TE. Using the channel geometry and negative pressure, the separation channel was filled with subsequent zones of the three different Les. This multielectrolyte system improved specificity and selectivity of the peak mode ITP method for miRNA sequences shorter than 40 nucleotides and was used for quantification of miRNA in cell cultures (HeLa and Hepa1–6). A similar method was used to visualize the separation of miRNA using specific molecular beacons [65]. The molecular beacons were mixed with the three LEs and because the mobility of C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Microfluidics and Miniaturization
1505
Figure 8. Scheme depicting acceleration of nucleic acid hybridization reactions using ITP. Two single-stranded DNA species A (analyte) and B (probe) are focused at a narrow interface between TE and LE in a microchannel. TE and LE are chosen such that their mobility bound all of the nucleic acid mobility. In this model system, species A is mixed with TE, and species B is mixed with LE, thus reaction occurs only at the interface where both species focus. The high concentrations of reactants at the interface lead to a corresponding increase in hybridization reaction rate. The arrow lengths at the top, respectively, denote the relative speed of species A in TE, of the ITP interface, and of species B in LE (LE and TE ions migrate at velocities equal to that of ITP interface). We consider a control volume moving with the interface at a velocity VITP . The control volume extends over a length L, which is significantly larger than the characteristic interface width, ␦, as shown. Reproduced with permission from [144].
miRNA and molecular beacons was slower than the LE but faster than the mobility of TE, the miRNAs and the molecular beacons migrated in a mixed peak zone. Hybridization and threrefore visualization only occurred in LE3 because of the lower urea concentration. This method enabled the specific detection and quantification of miR-122 in liver tissue. In 2011, Bercovici et al. described the use of specially designed molecular beacons for quantification of 16S rRNA from E. coli in human urine samples using a cross-geometry Caliper NS-95 device [66]. Urine samples were centrifuged, cells from sediment were lysed, mixed with a solution of molecular beacon and subsequently with TE (50 mM tricine, 100 mM bis-tris). The sample was loaded onto the chip that was prefilled with LE and E. coli could be detected in bacterial cultures as well as in urine samples in the clinically relevant range (1 × 106 –1 × 108 /mL). A very exciting prospect is in-line hybridization utilizing the focusing properties of ITP. Bercovici et al. [144] used -ITP to control and increase the rate of nucleic acid hybridization reactions in free solution with four orders of magnitude decreased reaction time. The formation of mixed concentrated zones concentrating two different reactants in the same spot facilitates this and this makes it a very unique feature of this approach. The scheme is illustrated in Fig. 8.
7 Free-flow ITP Most electrophoretic methods are applied to discrete samples, but in the free-flow format where an electric field www.electrophoresis-journal.com
1506
P. Smejkal et al.
for the separation is applied perpendicular to the flow of sample and buffer, sample can be introduced continuously [145–147]. Free-flow electrophoretic methods are suitable for miniaturization and a reviews on miniaturized free-flow electrophoretic techniques were published in 2007 [148] and 2009 [27,149]. In free-flow ITP (FFITP), a discontinuous buffer system is pumped through the separation chamber and was first performed on a PDMS chip in 2006 [150]. Collapsing of the separation chamber was prevented using 30 m square supporting posts in the chamber, leaving only 10 m spaces between the posts for FFITP. A schematic drawing of the FFITP device (without supporting posts) is shown in Supporting Information Fig. 7. The three reservoirs at the north side of the chip were filled with TE, sample, and LE. The west and east reservoirs along the separation chamber contained electrodes and were filled with TE and LE, respectively. The end of the separation chamber was connected to the outlet channels for collection of the separated fractions. In this study, however, the outlet channels were joined to apply a vacuum to drive the hydrodynamic flow in the chamber. The separation of fluorescein, eosin G, and acetylsalicylic acid and the separation of myoglobin-FITC, serine, and FITC were documented by using a CCD color camera. Later, Janasek et al. introduced electrostatic induction FFITP of fluorescein in a glass chip [151]. The mask design of the chip was similar to the chip described previously, but the west and east reservoirs containing electrodes were separated from the main chamber by a 146-m thick glass wall. In the mask, the supporting poles were 40 m, but the wet etching reduced the posts to 20 m high bumps in the 30 m high chamber; it is unclear if the posts played any role in preventing sagging of the glass. The technical advance, however, was that by separating the electrodes from the separation chamber, the potential applied to the electrodes could be transferred to the glass walls but the current was zero. ITP zones were formed in the electric field in the separation chamber but in this no current scenario, there was no movement of the zones toward the electrodes. This phenomenon was called electrostatic induction FFITP and was demonstrated for concentrating fluorescein. Prest et al. [152] introduced a design of miniaturized freeflow electrophoresis device formed of polystyrene with carbon fiber loaded polystyrene drive electrodes and produced using injection molding, containing a separation chamber 45-mm long by 31.7-mm wide with a depth of 50 m with nine inlet and nine outlet holes to allow for fraction collection. The operation of the devices was demonstrated by performing separations of dyes and bacterial samples containing the bacterium Erwinia herbicola, a biological pathogen, and showed that fractionation of the output was achieved.
8 Concluding remarks The first report of microfabricated devices for ITP predating the introduction of the TAS concept by 23 years illustrates the significant advantages of miniaturization and microfabri C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Electrophoresis 2013, 34, 1493–1509
cation for ITP and ITP-coupled methods. After the popularity of ITP sharply declined in the 1990s with the rise of CE following the introduction of fused silica capillaries, the flexibility in design offered by microfluidic chips combined with a need for a robust sample concentration/clean-up technique compatible with miniaturization led to renewed interest in analytical ITP in microfluidic chip format (-ITP). Developments in these areas have been underpinned by the use of 1D dynamic simulators such as GENTRANS, SPRESSO, and SIMUL, which have provided researchers with a deep insight into the physics underlying various electrophoretic separations in a relatively short period of time. However, multidimensional software will be needed to explore the multifaceted dynamics of microfluidic protocols that occur in networked and 2D microchips, particularly as the ability to exploit changes in microchannel geometry presents significant possibilities to improve the ITP performance. The review also highlights that the strength of -ITP is in the area of sample preparation/concentration because of its ability to concentrate trace components while reducing matrix components. In this area in particularly, it has been shown to be useful for the analysis of small molecules, including inorganic ions, organic acids, and fluorescent dyes as well as large biomolecules including proteins. Significantly, when combined with the use of cascading channels with decreasing width and depth, concentration factors of up to 10 000 have been demonstrated using this powerful approach. The use of peak mode ITP for both purification of DNA as well as RNA directly from lysed biological samples and to increase rapid hybridization kinetics is very significant development toward the creation of simple, functional, and usable TAS devices. It is abundantly clear from the literature covered here that there are many applications of -ITP in a range of disciplines and areas. Its greatest strength is the ability to reduce sample complexity by both increasing the concentration of trace components and decreasing matrix ions—in effect it is the great equalizer. It is this unique capability that can be performed in open channels that are technically simple to fabricate that can be used at low voltages because the steady-state boundaries counteract diffusion that ensure -ITP will play a significant role in the development of many true sample-in/answer-out TAS approaches. For the analysis of small molecules using plateau mode ITP, one of the principal developments was the introduction of indirect fluorescence detection, enabling the detection and visualization of nonfluorescent analytes. Peak mode ITP is a powerful way to concentrate trace analytes, and has changed the perspectives on detection in ITP by advocating a role for selective detection. Open channels in electromigration separations as opposed to chromatography offer a principal advantage for separations of whole cells. The role of ITP for their preconcentration, focusing, and sorting is likely to be one of very exiting areas of research in the area of -ITP. In this regard also rapid in-line hybridization has a future potential as it can provide -ITP an edge over competing methods. www.electrophoresis-journal.com
Electrophoresis 2013, 34, 1493–1509
This work was supported by the Grant Agency of the Czech Republic (P301/11/2055 and P206/12/G014) and the institutional research plan (UIACH 68081715). MCB would like to thank the Australian Research Council for funding and provision of a QEII Fellowship (DP0984745), and MM would like to acknowledge the Australian Research Council for funding and provision of a Future Fellowship (FT120100559). The authors have declared no conflict of interest.
9 References [1] Kendall, J., Crittenden, E. D., Proc. Natl. Acad. Sci. USA 1923, 9, 75–78. [2] Ornstein, L., Ann New York Acad. Sci. 1964, 121, 321–349. [3] Jorgenson, J. W., Lukacs, K. D., Anal. Chem. 1981, 53, 1298–1302.
Microfluidics and Miniaturization
1507
[24] Thormann, W., Breadmore, M. C., Caslavska, J., Mosher, R. A., Electrophoresis 2010, 31, 726–754. [25] Persat, A., Chambers, R. D., Santiago, J. G., Lab. Chip. 2009, 9, 2437–2453. [26] Persat, A., Suss, M. E., Santiago, J. G., Lab. Chip. 2009, 9, 2454–2469. [27] Turgeon, R. T., Bowser, M. T., Anal. Bioanal. Chem. 2009, 394, 187–198. [28] Landers, J. P., Handbook of Capillary and Microchip Electrophoresis and Associated Microtechniques, CRC, Boca Raton, FL 2008. [29] Gebauer, P., Mala, Z., Bocek, P., Electrophoresis 2007, 28, 26–32. [30] Chen, L., Prest, J. E., Fielden, P. R., Goddard, N. J., Manz, A., Day, P. J. R., Lab. Chip. 2006, 6, 474–487. [31] Evenhuis, C. J., Guijt, R. M., Macka, M., Haddad, P. R., Electrophoresis 2004, 25, 3602–3624.
[4] Jorgenson, J. W., Lukacs, K. D., Science 1983, 222, 266–272.
[32] Kaniansky, D., Masar, M., Bodor, R., Zuborova, M., Olvecka, E., Johnck, M., Stanislawski, B., Electrophoresis 2003, 24, 2208–2227.
[5] Kaniansky, D., Marak, J., J. Chromatogr. 1990, 498, 191–204.
[33] Khurana, T. K., Santiago, J. G., Anal. Chem. 2008, 80, 6300–6307.
[6] Foret, F., Szoko, E., Karger, B. L., J. Chromatogr. 1992, 608, 3–12.
[34] Prus´ık, Z., J. Chromatogr. A 1974, 91, 867–872.
[7] Manz, A., Graber, N., Widmer, H. M., Sensor. Actuat. B-Chem. 1990, 1, 244–248. [8] Walker, P. A., Morris, M. D., Burns, M. A., Johnson, B. N., Anal. Chem. 1998, 70, 3766–3769. [9] Bocek, P., Deml, M., Janak, J., J. Chromatogr. 1975, 106, 283–290. [10] Bahga, S. S., Santiago, J. G., Analyst 2013, 138, 735–754. [11] Breadmore, M. C., Shallan, A. I., Rabanes, H. R., Gstoettenmayr, D., Abdul Keyon, A. S., Gaspar, A., Dawod, M., Quirino, J. P., Electrophoresis 2013, 34, 29–54. ´ Z., Gebauer, P., Bocek, ˇ [12] Mala, P., Electrophoresis 2013, 34, 19–28. [13] Garcia-Schwarz, G., Rogacs, A., Bahga, S. S., Santiago, J. G., J. Vis. Exp. 2012, 61, 3890–3897. [14] Stoyanov, A., Electrophoresis 2012, 33, 3281–3290. [15] Wen, Y. Y., Li, J. H., Ma, J. P., Chen, L. X., Electrophoresis 2012, 33, 2933–2952. [16] Krizek, T., Kubickova, A., Anal. Bioanal. Chem. 2012, 403, 2185–2195. [17] Kitagawa, F., Kawai, T., Sueyoshi, K., Otsuka, K., Anal. Sci. 2012, 28, 85–93.
[35] Kasicka, V., Prus´ık, Z., J. Chromatogr. A 1987, 390, 27–37. [36] Prest, J. E., Baldock, S. J., Fielden, P. R., Goddard, N. J., Brown, B. J. T., Analyst 2003, 128, 1131–1136. [37] Prest, J. E., Baldock, S. J., Fielden, P. R., Goddard, N. J., Brown, B. J. T., J. Chromatogr. A 2004, 1051, 221–226. [38] Prest, J. E., Baldock, S. J., Fielden, P. R., Goddard, N. J., Kalimeri, K., Brown, B. J. T., Zgraggen, M., J. Chromatogr. A 2004, 1047, 289–298. [39] Prest, J. E., Baldock, S. J., Fielden, P. R., Goddard, N. J., Brown, B. J. T., Microchim. Acta 2005, 151, 223–230. [40] Prest, J. E., Beardah, M. S., Baldock, S. J., Doyle, S. P., Fielden, P. R., Goddard, N. J., Brown, B. J. T., J. Chromatogr. A 2008, 1195, 157–163. [41] Prest, J. E., Baldock, S. J., Fielden, P. R., Goddard, N. J., Brown, B. J. T., Anal. Bioanal. Chem. 2009, 394, 1299–1305. [42] Prest, J. E., Baldock, S. J., Fielden, P. R., Goddard, N. J., Brown, B. J. T., J. Chromatogr. A 2003, 990, 325–334. [43] Kaigala, G. V., Bercovici, M., Behnam, M., Elliott, D., Santiago, J. G., Backhouse, C. J., Lab. Chip. 2010, 10, 2242–2250.
[18] Kasicka, V., Electrophoresis 2012, 33, 48–73.
[44] Bottenus, D., Hossan, M. R., Ouyang, Y. X., Dong, W. J., Dutta, P., Ivory, C. F., Lab. Chip. 2011, 11, 3793–3801.
[19] Kenyon, S. M., Meighan, M. M., Hayes, M. A., Electrophoresis 2011, 32, 482–493.
[45] Masar, M., Kaniansky, D., Bodor, R., Johnck, M., Stanislawski, B., J. Chromatogr. A 2001, 916, 167–174.
[20] Gebauer, P., Mala, Z., Bocek, P., Electrophoresis 2011, 32, 83–89.
[46] Bodor, R., Madajova, V., Kaniansky, D., Masar, M., Johnck, M., Stanislawski, B., J. Chromatogr. A 2001, 916, 155–165.
[21] Breadmore, M. C., Dawod, M., Quirino, J. P., Electrophoresis 2011, 32, 127–148. [22] Lopez-Lorente, A. I., Simonet, B. M., Valcarcel, M., TracTrend. Anal. Chem. 2011, 30, 58–71.
[47] Bodor, R., Zuborova, M., Olvecka, E., Madajova, V., Masar, M., Kaniansky, D., Stanislawski, B., J. Sep. Sci. 2001, 24, 802–809.
[23] Frost, N. W., Jing, M., Bowser, M. T., Anal. Chem. 2010, 82, 4682–4698.
[48] Smejkal, P., Breadmore, M. C., Guijt, R. M., Foret, F., Bek, F., Macka, M., Electrophoresis 2012, 33, 3166–3172.
C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.electrophoresis-journal.com
1508
P. Smejkal et al.
[49] Smejkal, P., Breadmore, M. C., Guijt, R. M., Foret, F., Bek, F., Macka, M., Anal. Chim. Acta 2013, in press, DOI: 10.1016/j.aca.2013.01.046. [50] Bodor, P., Kaniansky, D., Masar, M., Silleova, K., Stanislawski, B., Electrophoresis 2002, 23, 3630–3637. [51] Masar, M., Dankova, M., Olvecka, E., Stachurova, A., Kaniansky, D., Stanislawski, B., J. Chromatogr. A 2004, 1026, 31–39. [52] Masar, M., Dankova, M., Olvecka, E., Stachurova, A., Kaniansky, D., Stanislawski, B., J. Chromatogr. A 2005, 1084, 101–107. [53] Wainright, A., Williams, S. J., Ciambrone, G., Xue, Q. F., Wei, J., Harris, D., J. Chromatogr. A 2002, 979, 69–80. [54] Liu, D. Y., Shi, M., Huang, H. Q., Long, Z. C., Zhou, X. M., Qin, J. H., Lin, B. C., J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2006, 844, 32–38. [55] Kawabata, T., Wada, H. G., Watanabe, M., Satomura, S., Electrophoresis 2008, 29, 1399–1406. [56] Park, C. C., Kazakova, I., Kawabata, T., Spaid, M., Chien, R. L., Wada, H. G., Satomura, S., Anal. Chem. 2008, 80, 808–814. [57] Wu, R. G., Yeung, W. S. B., Fung, Y. S., Electrophoresis 2011, 32, 3406–3414. [58] Persat, A., Marshall, L. A., Santiago, J. G., Anal. Chem. 2009, 81, 9507–9511. [59] Xu, Z. Q., Nishine, T., Arai, A., Hirokawa, T., Electrophoresis 2004, 25, 3875–3881. [60] Wang, L. H., Liu, D. Y., Chen, H., Zhou, X. M., Electrophoresis 2008, 29, 4976–4983. [61] Liu, D. Y., Ou, Z. Y., Xu, M. F., Wang, L. H., J. Chromatogr. A 2008, 1214, 165–170. [62] Liu, D. Y., Chen, B., Wang, L. H., Zhou, X. M., Electrophoresis 2009, 30, 4300–4305. [63] Schoch, R. B., Ronaghi, M., Santiago, J. G., Lab. Chip. 2009, 9, 2145–2152. [64] Persat, A., Chivukula, R. R., Mendell, J. T., Santiago, J. G., Anal. Chem. 2010, 82, 9631–9635. [65] Persat, A., Santiago, J. G., Anal. Chem. 2011, 83, 2310–2316. [66] Bercovici, M., Kaigala, G. V., Mach, K. E., Han, C. M., Liao, J. C., Santiago, J. G., Anal. Chem. 2011, 83, 4110–4117.
Electrophoresis 2013, 34, 1493–1509
[75] Palusinski, O. A., Graham, A., Mosher, R. A., Bier, M., Saville, D. A., Aiche J. 1986, 32, 215–223. [76] Saville, D. A., Palusinski, O. A., Aiche J. 1986, 32, 207–214. [77] Mosher, R. A., Saville, D. A., The Dynamics of Electrophoresis, VCH, Weinheim, New York 1992. [78] Hruska, V., Jaros, M., Gas, B., Electrophoresis 2006, 27, 984–991. [79] Bercovici, M., Lele, S. K., Santiago, J. G., J. Chromatogr. A 2009, 1216, 1008–1018. [80] Mosher, R. A., Breadmore, M. C., Thormann, W., Electrophoresis 2011, 32, 532–541. [81] Bercovici, M., Lele, S. K., Santiago, J. G., J. Chromatogr. A 2010, 1217, 588–599. [82] Bahga, S. S., Bercovici, M., Santiago, J. G., Electrophoresis 2010, 31, 910–919. [83] Bahga, S. S., Kaigala, G. V., Bercovici, M., Santiago, J. G., Electrophoresis 2011, 32, 563–572. [84] Breadmore, M. C., Kwan, H. Y., Caslavska, J., Thormann, W., Electrophoresis 2012, 33, 958–969. [85] Jacobson, S. C., Hergenroder, R., Koutny, L. B., Warmack, R. J., Ramsey, J. M., Anal. Chem. 1994, 66, 1107–1113. [86] Kurnik, R. T., Boone, T. D., Nguyen, U., Ricco, A. J., Williams, S. J., Lab. Chip. 2003, 3, 86–92. [87] Hirokawa, T., Takayama, Y., Arai, A., Xu, Z. Q., Electrophoresis 2008, 29, 1829–1835. [88] Jung, B., Bharadwaj, R., Santiago, J. G., Anal. Chem. 2006, 78, 2319–2327. [89] Bahga, S. S., Chambers, R. D., Santiago, J. G., Anal. Chem. 2011, 83, 6154–6162. [90] Cui, H. C., Dutta, P., Ivory, C. F., Electrophoresis 2007, 28, 1138–1145. [91] Cui, H. C., Huang, Z., Dutta, P., Ivory, C. F., Anal. Chem. 2007, 79, 1456–1465. [92] Tsai, C. H., Wang, Y. N., Lin, C. F., Yang, R. J., Fu, L. M., Electrophoresis 2006, 27, 4991–4998. [93] Schonfeld, F., Goet, G., Baier, T., Hardt, S., Phys. Fluids 2009, 21, 092002, DOI: 10.1063/1.3222866. [94] Liu, B., Ivory, C. F., J. Sep. Sci. 2013, in press, DOI: 10.1002/jssc.201300066.
[67] Vyas, C. A., Mamunooru, M., Shackman, J. G., Chromatographia 2009, 70, 151–156.
[95] Everaerts, F. M., Vacik, J., J. Chromatogr. A 1970, 49, 262–268.
[68] Prest, J. E., Baldock, S. J., Fielden, P. R., Goddard, N. J., Brown, B. J. T., Analyst 2005, 130, 1375–1382.
[96] Dolnik, V., Deml, M., Bocek, P., J. Chromatogr. 1985, 320, 89–97.
[69] Prest, J. E., Fielden, P. R., Anal. Bioanal. Chem. 2005, 382, 1339–1342.
[97] Paschkewitz, J. S., Molho, J. I., Xu, H., Bharadwaj, R., Park, C. C., Electrophoresis 2007, 28, 4561–4571.
[70] Olvecka, E., Masar, M., Kaniansky, D., Johnck, M., Stanislawski, B., Electrophoresis 2001, 22, 3347–3353.
[98] Kasicka, V., Prusik, Z., Gas, B., Stedry, M., Electrophoresis 1995, 16, 2034–2038.
[71] Grass, B., Hergenroder, R., Neyer, A., Siepe, D., J. Separ. Sci. 2002, 25, 135–140.
[99] Harrison, S. L. M., Chemical Engineering, Washington State University, Pullman 2007, p. 69.
[72] Olvecka, E., Kaniansky, D., Pollak, B., Stanislawski, B., Electrophoresis 2004, 25, 3865–3874.
[100] Breadmore, M. C., Electrophoresis 2008, 29, 1082–1091.
[73] Bier, M., Palusinski, O. A., Mosher, R. A., Saville, D. A., Science 1983, 219, 1281–1287. [74] Thormann, W., Caslavska, J., Breadmore, M. C., Mosher, R. A., Electrophoresis 2009, 30, S16–S26. C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[101] Bottenus, D., Jubery, T. Z., Dutta, P., Ivory, C. F., Electrophoresis 2011, 32, 550–562. [102] Prest, J. E., Baldock, S. J., Bektas, N., Fielden, P. R., Brown, B. J. T., J. Chromatogr. A 1999, 836, 59–65.
www.electrophoresis-journal.com
Electrophoresis 2013, 34, 1493–1509
Microfluidics and Miniaturization
1509
[103] Baldock, S. J., Bektas, N., Fielden, P. R., Goddard, N. J., Pickering, L. W., Prest, J. E., Snook, R. D., Brown, B. J. T., Vaireanu, D. I., Isotachophoresis on Planar Polymeric Substrates, Springer, Dordrecht 2000.
[127] Krivankova, L., Gebauer, P., Bocek, P., J. Chromatogr. A 1995, 716, 35–48.
[104] Prest, J. E., Baldock, S. J., Fielden, P. R., Brown, B. J. T., Analyst 2001, 126, 433–437.
[129] Vreeland, W. N., Williams, S. J., Barron, A. E., Sassi, A. P., Anal. Chem. 2003, 75, 3059–3065.
[105] Prest, J. E., Baldock, S. J., Fielden, P. R., Goddard, N. J., Brown, B. J. T., Analyst 2002, 127, 1413–1419.
[130] Xu, Z. Q., Ando, T., Nishine, T., Arai, A., Hirokawa, T., Electrophoresis 2003, 24, 3821–3827.
[106] Baldock, S. J., Fielden, P. R., Goddard, N. J., Prest, J. E., Brown, B. J. T., J. Chromatogr. A 2003, 990, 11–22. [107] Baldock, S. J., Fielden, P. R., Goddard, N. J., Kretschmer, H. R., Prest, J. E., Brown, B. J. T., J. Chromatogr. A 2004, 1042, 181–188.
´ ´ L., Pantu˚ ckov ˇ ´ P., Bocek, ˇ [128] Kˇrivankov a, a, P., J. Chromatogr. A 1999, 838, 55–70.
[131] Xu, Z. Q., Hirokawa, T., Nishine, T., Arai, A., J. Chromatogr. A 2003, 990, 53–61. [132] Jeong, Y. W., Choi, K. W., Kang, M. K., Chun, K. J., Chung, D. S., Sensor. Actuat. B-Chem. 2005, 104, 269–275.
[108] Prest, J. E., Baldock, S. J., Fielden, P. R., Goddard, N. J., Mohr, S., Brown, B. J. T., J. Chromatogr. A 2006, 1119, 183–187.
[133] Mohamadi, M. R., Kaji, N., Tokeshi, M., Baba, Y., Anal. Chem. 2007, 79, 3667–3672.
[109] Macka, M., Johns, C., Doble, P., Haddad, P. R., LC-GC North America 2001, 19, 38–47.
[134] Wang, J., Zhang, Y., Mohamadi, M. R., Kaji, N., Tokeshi, M., Baba, Y., Electrophoresis 2009, 30, 3250–3256.
[110] Macka, M., Johns, C., Doble, P., Haddad, P. R., LC-GC North America 2001, 19, 178–188. [111] Reijenga, J. C., Verheggen, T. P. E. M., Everaerts, F. M., J. Chromatogr. A 1984, 283, 99–111. [112] Khurana, T. K., Santiago, J. G., Anal. Chem. 2008, 80, 279–286. [113] Bercovici, M., Kaigala, G. V., Santiago, J. G., Anal. Chem. 2010, 82, 2134–2138. [114] Chambers, R. D., Santiago, J. G., Anal. Chem. 2009, 81, 3022–3028. [115] Bottenus, D., Jubery, T. Z., Ouyang, Y. X., Dong, W. J., Dutta, P., Ivory, C. F., Lab. Chip. 2011, 11, 890–898. [116] Smejkal, P., Breadmore, M. C., Guijt, R. M., Grym, J., Foret, F., Bek, F., Macka, M., Anal. Chim. Acta 2012, 755, 115–120.
[135] Lin, C. C., Hsu, B. K., Chen, S. H., Electrophoresis 2008, 29, 1228–1236. [136] Nagata, H., Ishikawa, M., Yoshida, Y., Tanaka, Y., Hirano, K., Electrophoresis 2008, 29, 3744–3751. [137] Shackman, J. G., Ross, D., Anal. Chem. 2007, 79, 6641–6649. [138] Davis, N. I., Mamunooru, M., Vyas, C. A., Shackman, J. G., Anal. Chem. 2009, 81, 5452–5459. [139] Danger, G., Ross, D., Electrophoresis 2008, 29, 4036–4044. [140] Mamunooru, M., Jenkins, R. J., Davis, N. I., Shackman, J. G., J. Chromatogr. A 2008, 1202, 203–211. [141] Svoboda, M., Vacik, J., J. Chromatogr. 1976, 119, 539–547.
[117] Mark, D., von Stetten, F., Zengerle, R., Lab. Chip. 2012, 12, 2464–2468.
[142] Gebauer, P., Bocek, P., Electrophoresis 1995, 16, 1999–2007.
[118] Kaniansky, D., Masar, M., Bielcikova, J., Ivanyi, F., Eisenbeiss, F., Stanislawski, B., Grass, B., Neyer, A., Johnck, M., Anal. Chem. 2000, 72, 3596–3604.
[143] Khurana, T. K., Santiago, J. G., Lab. Chip. 2009, 9, 1377–1384. [144] Bercovici, M., Han, C. M., Liao, J. C., Santiago, J. G., Proc. Natl. Acad. Sci. USA 2012, 109, 11127–11132.
[119] Grass, B., Neyer, A., Johnck, M., Siepe, D., Eisenbeiss, F., Weber, G., Hergenroder, R., Sensor. Actuat. B-Chem. 2001, 72, 249–258.
[145] Canut, H., Bauer, J., Weber, G., J. Chromatogr. B 1999, 722, 121–139.
[120] Wainright, A., Nguyen, U. T., Bjornson, T., Boone, T. D., Electrophoresis 2003, 24, 3784–3792.
¨ Analytische [146] Hannig, K., Fresenius’ Zeitschrift fur Chemie 1961, 181, 244–254.
[121] Huang, H. Q., Xu, F., Dai, Z. P., Lin, B. C., Electrophoresis 2005, 26, 2254–2260.
[147] Roman, M. C., Brown, P. R., Anal. Chem. 1994, 66, 86A–94A.
[122] Ma, B., Zhou, X. M., Wang, G., Huang, H. Q., Dai, Z. P., Qin, J. H., Lin, B. C., Electrophoresis 2006, 27, 4904–4909.
[148] Kohlheyer, D., Eijkel, J. C. T., van den Berg, A., Schasfoort, R. B. M., Electrophoresis 2008, 29, 977–993.
[123] Jung, B. G., Zhu, Y. G., Santiago, J. G., Anal. Chem. 2007, 79, 345–349. [124] Rogacs, A., Qu, Y., Santiago, J. G., Anal. Chem. 2012, 84, 5858–5863. [125] Petr, J., Maier, V., TrAC-Trend. Anal. Chem. 2012, 31, 9–22. [126] Phung, S. C., Nai, Y. H., Powell, S. M., Macka, M., Breadmore, M. C., Electrophoresis 2013, 34, 1657–1662.
C 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ˇ cka, ˇ [149] Kasi V., Electrophoresis 2009, 30, S40–S52. [150] Janasek, D., Schilling, M., Franzke, J., Manz, A., Anal. Chem. 2006, 78, 3815–3819. [151] Janasek, D., Schilling, M., Manz, A., Franzke, J., Lab. Chip. 2006, 6, 710–713. [152] Prest, J. E., Baldock, S. J., Fielden, P. R., Goddard, N. J., Goodacre, R., O’Connor, R., Brown, B. J. T., J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2012, 903, 53–59.
www.electrophoresis-journal.com
