Embed Size (px)
Citation preview
Apa
PMa
b
c
d
h
�
�
�
�
a
ARRAA
KIMFIS
tr
T
m
0h
Analytica Chimica Acta 803 (2013) 135– 142
Contents lists available at ScienceDirect
Analytica Chimica Acta
j ourna l ho mepage: www.elsev ier .com/ locate /aca
nalytical isotachophoresis of lactate in human serum using dry filmhotoresist microfluidic chips compatible with a commerciallyvailable field-deployable instrument platform
etr Smejkala, Michael C. Breadmorea, Rosanne M. Guijtb, Frantisek Foretc, Fritz Bekd,irek Mackaa,∗
University of Tasmania, ACROSS and School of Chemistry, Hobart, AustraliaUniversity of Tasmania, ACROSS and School of Pharmacy, Hobart, AustraliaInstitute of Analytical Chemistry of the Academy of Sciences of the Czech Republic, v.v.i., Brno, Czech RepublicAgilent Technologies, Waldbronn, Germany
i g h l i g h t s
A dry film resist polymer chipdesigned for analysis of lactate inhuman serum.LIF/LEDIF indirect fluorescencedetection used.Polymer ITP chip compatible witha field-deployable commercial plat-form.Full research flexibility using a com-mercial microfluidic separation plat-form.
g r a p h i c a l a b s t r a c t
r t i c l e i n f o
rticle history:eceived 28 November 2012eceived in revised form 21 January 2013ccepted 22 January 2013vailable online 1 February 2013
a b s t r a c t
A dry film resist (DFR) chip compatible with the Agilent Bioanalyzer 2100 was designed and fabricatedfor use in the analysis of lactate in serum by chip isotachophoresis (ITP). The Agilent Bioanalyzer 2100is a commercially available field deployable analytical instrument originally developed for the elec-trophoretic analysis of DNA, RNA and proteins. The DFR chip was designed for the ITP separation oflactate in human serum within 1 min and was made compatible with the Bioanalyzer after packaging in
eywords:sotachophoresis
icrofluidicsluorescencendirect detection
the plastic caddies normally used for the DNA chips. A 20-fold improvement in sensitivity was obtainedfor the DFR chips in comparison with the standard chips used in earlier work. The limit of detection andlimit of quantification for lactate were 24 �M and 80 �M, respectively. This new approach enables theuse of commercial platforms like the Agilent Bioanalyzer for new applications including the analysis ofsmall molecules.
erum
Abbreviations: R6G, rhodamine 6G; LE, leading electrolyte; TE, terminating elec-rolyte; HPMC, hydroxypropylmethyl cellulose; PC, polycarbonate; DFR, dry filmesist.∗ Corresponding author at: Private Bag 75, School of Chemistry, University of
asmania, Hobart 7001, Australia. Tel.: +61 362266670; fax: +61 362262858.E-mail addresses: [email protected],
[email protected] (M. Macka).
003-2670/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.aca.2013.01.046
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
Miniaturization has been a major focus in the development ofmodern highly sensitive separation techniques, such as liquid chro-matography [1,2] and electrophoresis [3] in the last two decades.While the transition from classical analytical separation systems to
a microfluidic format has been widely adapted in the research, onlyfew microfluidic chip-based platforms appeared on the market ascommercial devices [4–6]. Therefore the scarcity of commerciallyavailable portable or field deployable platforms, which are fully
1 himic
fltaTat
wspar2ad
lclw2cpfcci(aaapabdB
twttdbdsccatpcaTs
cidihtt
rab
36 P. Smejkal et al. / Analytica C
exible for research, has been a major limitation and impedimento the use of portable and field deployable microfluidic chip basednalytical instruments in solving real world analytical problems.his has led to the development of simple microfluidic devices thatre compatible with existing and common laboratory instrumen-ation such as centrifuges and PCR thermocyclers [7].
One of the first commercial microfluidic systems on the marketas the Agilent Bioanalyzer 2100, a platform for chip electrophore-
is with fluorescence detection for analysis of DNA, RNA androteins, and later for pressure driven flow cytometry. The generaldvantage of electrophoresis is that it enables high resolution sepa-ations to be easily achieved in a small instrument. The Bioanalyzer100 is a small footprint desktop instrument, and its dimensionsnd weight make it suitable as a portable and field deployableevice [8].
In our previous research we examined the use of the Bioana-yzer as a general research-flexible microfluidic platform for chipapillary electrophoresis (chip-CE) [9], which opens the Bioana-yzer up for a more diverse range of applications than what it
as originally designed for. Later, Lloyd et al. used the Bioanalyzer100 for the separation of amphetamines labeled with fluores-ein isothiocyanate [10]. While these illustrate the possibility oferforming simple zone electrophoretic separations in this plat-orm, the separation performance was inferior to that obtained inapillaries due to the short channels and low field strengths thatan be applied. To overcome this, we established the applicabil-ty of the Agilent Bioanalyzer 2100 for analytical isotachophoresisITP) with indirect fluorescence detection using the commerciallyvailable Agilent DNA chips and demonstrated this for the sep-ration of carboxylic acids in soft drinks [11]. ITP is an idealpproach for miniaturization because it retains all of the separationower of zone electrophoresis while the self-sharpening bound-ries counter-act diffusion allowing high resolution separations toe obtained at relatively low field strengths. In this work we furtherevelop chip-ITP separations using a range of in-house designedioanalyzer-compatible polymer chips.
ITP has been shown to be a powerful separation technique withhe ability to deal with high salinity and complex samples (serum)ithout difficult and time consuming sample pretreatment a fur-
her benefit for its use in portable instrumentation [12–18]. Theechnique was developed in the 1980s and 1990s and a detailedescription of the theory, applications and instrumentation cane found in a monographs such as by Bocek et al. [19]. After theevelopment of a protocol for the determination of benzoate inoft drinks using the DNA chips, we demonstrated the methodould also be applied to serum samples by changing the injectiononditions using a commercial DNA chip [20]. While this appearsttractive from the point of using an existing commercial glass chip,he isotachopherogram from the commercial DNA chip did not dis-lay fully developed ITP steps, and therefore a new microfluidic ITPhip was in-house designed and fabricated from soda-lime glass,nd demonstrated for separation of lactate in human serum [20].he results although successful were far from practical with longeparation times and poor limits of detection.
Here, we introduce a new design ITP microchip and show thatompared to our previous work we can improve the repeatabil-ty and significantly reduce the analysis time. The new chip wasesigned to allow a vacuum injection and improve the repeatabil-
ty of injection. The channels dimensions were varied to reduce theydrodynamic flow between reservoirs, improve detection sensi-ivity [21–23] and also to allow the use of a higher current, andherefore faster isotachophoretic separation.
Despite superior properties of glass as microfluidic chip mate-ial including optical transparency and chemical stability, Lab on
Chip research has moved to the use of polymer devices mainlyecause of their lower manufacturing costs [24]. The material used
a Acta 803 (2013) 135– 142
for the fabrication of our newly designed microchips compatiblewith the Agilent Bioanalyzer 2100 was dry film photoresist (DFR).The Ordyl SY300 series DFR was previously introduced for fabrica-tion of microfluidics chips at low cost using budget infrastructure,such as a LED light source and office laminator [25] and was usedhere to make microchips compatible with the Agilent Bioanalyzer2100. The DFR ITP chips were used for the determination of lactatein human serum using a Bioanalyzer 2100 with indirect fluores-cence detection and the results were confirmed using a CE method[26].
2. Experimental
2.1. In-house dry-film resist chip fabrication
The ITP chip was designed using CAD freeware Draft Sight (Das-sault Systemes, Velizy Villacoublay, France). The DFR chips werefabricated using the same equipment as described previously byGuijt et al. [25] and a slightly modified procedure as follows. A70 mm × 50 mm × 1 mm piece of polycarbonate (PC) sheet (Poly-tech Plastics, WA, Australia) was used as a support upon which theDFR microchip was constructed. A layer of 30 �m thick DFR (Ordyl330, Elga Europe, Italy) was laminated using an office laminator(Peach 3500, Lamination Systems, Australia) onto the PC substrateand exposed using a UV shark series high-flux LED array (OTHL-0480-UV, Opto Technology, Wheeling, IL, USA) with an outputintensity 0.8 mW m−2. After UV exposure the wafers were bakedon a hot plate (ECHOthermTM MODEL HS40, Torrey Pines Scien-tific, CA, USA) fitted with a 5 mm thick piece of polished aluminiumplate. A second layer of DFR was laminated and exposed using thesame lithography source for 1 min with a chrome on sodium-limeglass mask manufactured by Bandwidth Foundry (NSW, Australia).The channels were developed using BMR developer (Elga Europe)and rinsed in BMR rinse (Elga Europe). The profile and dimen-sions of the channels in the Ordyl DFR were examined by using aWyko NT 9100 optical profiler (Veeco, NY, USA). Chips were cut byusing a high-speed rotary tool (Dremel 400 series, Dremel Europe,Breda, Netherlands) obtained from local hardware store. Holes forinjecting electrolytes into the chip were drilled by using a drillpress (Ryobi EDP 2521L, TTI, Hong Kong, China) and 2 mm drill bitsobtained from a local hardware store and laminated with a finallayer of DRF to seal the microchannels. The finished chips wereglued by 5 min epoxy glue obtained from a local hardware storeinto black PMMA plastic frames (supplied by Agilent Technologies,Waldbronn, Germany).
2.2. Isotachophoresis of human serum
2.2.1. Instrumentation and chemicalsThe Bioanalyzer 2100 (Agilent Technologies, Palo Alto, CA, USA),
the commercial DNA chips, and the in-house Ordyl dry-film ITPchips were used for all ITP experiments.
Hydrochloric acid (32%) was purchased from Merck (San Diego,CA, USA). Sodium pyruvate (99%), sodium lactate (syrup, 60%(w/w)), sodium 3-hydroxybutyrate (99%), nicotinic acid (99%),�-alanine (99%), polyvinilpyrolidone (PVP, 1.3 MDa), hydrox-ypropylmethyl cellulose (HPMC, 3500–5600 cP), rhodamine 6 Gchloride salt (99%), 3,5-dinitrobenzoate and cetyltrimethylammo-nium bromide (CTAB) were all purchased from Sigma–Aldrich (St.Louis, MO, USA). Milli-Q (Merck Millipore, Billerica, MA, USA) waterwas used for all experiments.
2.2.2. Leading electrolyte, terminating electrolyte and samplesConditions for ITP of carboxylic acids in human serum were
adopted from our previous work [20] and slightly modified as
P. Smejkal et al. / Analytica Chimica Acta 803 (2013) 135– 142 137
Fig. 1. Schema of the chip design. (A) Functional scheme with reservoirs marked (LE = leading electrolyte, TE = terminating electrolyte, S = sample, V = reservoir for applyingthe vacuum). (B) Well layout to fit with dimensions (mm) to satisfy compatibility with the Agilent Bioanalyzer 2100. (C) Chip channels design with dimensions (mm, unlesso acuuma
fists4s2whwtw
sHwfn(pb
s
bswst
pcoA(tCspa1(bw5t
therwise stated). At this stage of the chip design, the dimension X of where the vdditional simulations were required to establish it.
ollows. The two leading electrolytes (LE) were used for follow-ng experiments. The first LE was prepared daily by mixing twotock solutions in a ratio of 1:1 (v/v). The first stock solution con-ained 100 �M Rhodamine 6G (R6G) dissolved in water, and theecond stock solution was prepared by dissolving 2%(w/v) PVP in0 mM hydrochloric acid, adjusted to pH 3.3 with �-alanine. Theecond LE was prepared daily by mixing three stock solutions.00 �L of the first stock solution (100 �M R6G dissolved in water)as mixed with 40.7 �L of the second stock solution (98.28 mMydrochloric acid with pH adjusted to pH 3.3 with �-alanine) andith 120 �L of the third stock solution (2% HPMC in water). The
otal volume of LE was adjusted to 400 �L by adding 39.3 �L ofater.
The terminating electrolyte (TE) was prepared daily by mixingtock solutions of 20 �L of 100 mM nicotinic acid with 120 �L of 2%PMC. The volume of TE was adjusted to 400 �L by adding 260 �L ofater. All stock solutions for preparing LE and TE were stored in the
ridge at 6 ◦C. In this work nicotinic acid was used instead of propio-ic acid. Both acids have similar electrophoretic mobility at pH 3.3∼8.5 × 10−9 m2 V−1 s−1 nicotinic acid and ∼8.6 × 10−9 m2 V−1 s−1
ropionic acid [27]). Nicotinic acid is a solid crystalline powderetter suited for preparation of solutions.
The artificial samples were prepared by diluting a 100 mM stockolution of lactate to a required concentration.
Human serum samples were collected from a healthy volunteery staff at Hobart Pathology and were pipetted into plastic vials andtored in freezer at −20 C. Before analysis, 45 �L of defrosted serumas mixed with 40 �L of water, 38.6 �L of 2% HPMC, and 5 �L of
tandard lactate addition (0, 5, 10 and 20 mM lactate correspondingo 0, 0.5, 1 and 2 mM lactate addition in serum sample).
The CE method validated for determination of lactate in serum,ublished by Jager and Tavares [26], was used for the verification ofhip-ITP results obtained with the Bioanalyzer. The CZE was carryut using Agilent Technologies HP3DCE capillary analyzer (Palolto, CA, USA). Separation polyimide uncoated fused silica capillary
75 �m ID × 375 �m OD) was 48.5 cm long with effective separa-ion length 40 cm. The BGE was 5 mM 3,5-dinitrobenzoate, 0.1 mMTAB (pH 3.5). Sample was injected by pressure (35 mbar, 2 s),eparation voltage was −15 kV applied to the injection side, tem-erature was 25 ◦C, indirect UV detection of lactate was carry outt 254 nm. Prior the first use the capillary was preconditioned with
mM NaOH (2 bar, 5 min), with water (2 bar, 5 min) and with BGE2 bar, 30 min). The capillary was flushed with BGE (2 bar, 2 min)
etween runs. Serum samples with standard addition of lactateere prepared by mixing 45 �L of human serum with 5 �L water,�L 10 mM lactate (1 mM standard addition) and 5 �L 20 mM lac-ate (2 mM standard addition).
injection microchannel will connect the separation channel was unknown, and
2.2.3. Vacuum loading of electrolytes into a DRF-ITP chipThe in-house DRF-ITP chip was filled with electrolytes as shown
in Fig. 1A. First the reservoirs were filled with LE, TE and sam-ple (S). Vacuum was then applied to the vacuum reservoir (V)by using a 1 mL plastic syringe with a rubber seal (−300 �L,5 s), which caused the electrolytes from reservoirs to fill theseparation channel in the chip. To prevent hydrodynamic flowand to provide a conductive connection with an ITP system, thereservoir V was then filled with 10 �L of LE. Injection and ITPseparation were visualized using an inverted confocal fluores-cence microscope (Nikon Eclipse Ti-U, Tokyo, Japan) with Semrock390/482/563/640 nm BrightLine quad-band bandpass excitationfilter, Semrock DI01 – R405/488/561/635 nm dichroic filter, andSemrock 446/523/600/677 nm multiband emission filter (bothRochester, NY, USA). The schematic of the injection and a videoobtained by the fluorescence microscope are provided in theSupporting information.
3. Results and discussion
3.1. Design of DRF ITP chips
As explained above the microfabrication method uses DRF tech-nology [25], where the selected thickness of dry film layer used forchip fabrication determines the depth of microchannels – in thiswork DRF layer of 30 �m was used. Other factors that have to betaken into account when designing the chip compatible with theBioanalyzer are the positions of the reservoirs to maintain compat-ibility with array of 16 (4 × 4) platinum electrodes and the positionof LIF detection.
The most important aspect of the ITP chip design is the injectionmethod. ITP requires the injection of two different electrolytes (LE,TE) and sample, and the chip design has to be such as to allow it.Electrokinetic and hydrodynamic injection modes can be both usedfor sample introduction in ITP, although as discussed in our previ-ous paper [20], the electrokinetic injection of serum sample faileddue to the high concentration of chloride in the sample. Unfor-tunately hydrodynamic injection by positive pressure is difficultto realize in a simple, controlled and reproducible manner, andcomputer controlled automatically activated channel valves areusually required [28–30]. Vacuum injection used here is injection
simple and more reproducible approach particularly if the volumeof sample injected can be physically defined by the microchan-nel architecture. Vacuum can be applied to only one reservoir,enabling fluidic control of the flow of different electrolytes drawn
1 himica Acta 803 (2013) 135– 142
ft
oictatysmtt(
ertrl1afibtdPn3nc
otWi
isvt5wdiwts
Fig. 2. Results of numerical simulations with Simul 5.0 for three different designsshowing the position where steady state is reached for those designs and injected
38 P. Smejkal et al. / Analytica C
rom the other reservoirs using hydrodynamic resistance throughhe microchannel design (cross-section and length).
One of the practically important insights gained from our previ-us work with in-house glass ITP chips [20] was that wider channelsncrease the amount of injected sample and decrease the electri-al resistance of the ITP system. This is beneficial in ITP becausehe increased sample injection increases the ITP system sensitivity,nd the decreased electrical resistance in wider channels allowshe separation current to be increased, which decreases the anal-sis time. Here, a separation channel composed of a wide channelection (300 �m width, 7.65 mm length), connected to a narrowicrochannel section (30 �m width, 3.65 mm length) was designed
o enhance the detection sensitivity [22,23]. The transition betweenhe wide and the narrow parts of separation channel was 1 mm longFig. 1C).
A schematic of how the chip is filled is shown in Fig. 1A. First,lectrolytes and sample (10 �L of each) were loaded into the cor-esponding reservoirs (LE, TE, S). The electrolytes were drawn intohe 300 �m wide separation channel by applying vacuum to theeservoir V. To prevent hydrodynamic flow due to the differentevel of electrolytes in the reservoirs, the reservoir V was filled with0 �L of LE after the injection was finished. To minimize undesir-ble hydrodynamic flow between termination of the vacuum andlling the reservoir a relatively narrow (30 �m wide) connectionsetween the reservoirs and the wide separation channel were usedo provide a localized resistance point. To avoid too low a con-uctivity the length of these connections was limited to 100 �m.reviously it was mentioned that the length of the narrow chan-el between the wide separation channel and the LE reservoir was.65 mm long (see Fig. 1C). This length was purposely longer thanecessary to ensure that the detection optics focused in the narrowhannel of the in-house designed chip.
In the design of the ITP chip, restrictions due to the size, locationf reservoirs on the Agilent DNA chip frame (Fig. 1B) and the posi-ion of the LIF detector in the instrument, are shown in (Fig. 1B).
ithin these restrictions, the designed microchannel is depictedn Fig. 1C.
Using the schematic shown in Fig. 1C, the variable in the designs the position at which the vacuum injection microchannel inter-ects the main separation channel. This defines the sample injectionolume, as well as the microchannel length available for ITP separa-ion prior to entering the narrow section and detection. Using Simul.0 [31] three designs differing in the position of this intersectionere evaluated to ensure steady state ITP would be reached prior toetection (Fig. 2). The details of conditions used for simulations are
ncluded in Supporting information. The results are shown in Fig. 2ith color coding within the microchannel as follows: the transi-
ion from black to red indicates the position at which a steady stateystem is first reached. Design A, in which the smallest amount of
Fig. 3. Photograph of a finished DRF-ITP chip glued inside the plastic black P
sample amounts. The length of the injection zone is indicated in mm. The changefrom black to red indicates the position in the channel where a steady state isachieved.
sample is introduced into the separation channel, reached steadystate well before the channel narrowing. Design B, in which thelargest amount of sample is introduced into the separation channelreached steady state after the detection point. In the third design,design C, steady state is reached just before the transition fromthe wide to narrow channel. This represents the optimal designwith respect to reaching steady state before the detection point butrelatively close to it therefore maximizing the amount of injectedsample and hence the system sensitivity. Consequently, the thirddesign was chosen for fabrication.
3.2. Chip fabrication
The material used for chip fabrication was Ordyl DFR previouslydemonstrated by Guijt et al. [25] as a suitable material for fastand low-cost microfluidic chip fabrication. Ordyl is a thin plasticfoil and the chips made from this material need a support. Guijtet al. used PMMA as the base layer for Ordyl chips. The smallestthickness of PMMA sheet that that we could obtained was 1.5 mm,however, as the commercial DNA chips are made from glass wafers1 mm thick, the Bioanalyzer’s detection optics could not focus onthe separation channel when using thicker substrates. As we couldnot access 1 mm thick PMMA, polycarbonate (PC) with a thicknessof 1 mm was selected for the fabrication of the microchip. Unfor-
tunately, PC is incompatible with the Ordyl developing and rinsingsolutions (BMR developing and rinsing solutions). During the devel-opment and rinse steps, the exposed side of the PC substrate wascovered with office sticky tape. Another issue associated with usingMMA frame, which positioned the chip for use inside the Bioanalyzer.
P. Smejkal et al. / Analytica Chimica Acta 803 (2013) 135– 142 139
Fig. 4. (A) Isotachopherograms showing a comparison of the DNA and the DRF ITP chips at identical conditions: LE = 20 mM HCl, �-alanine (pH 3.3), 50 �M R6G, 1% PVP;T cuumw n a DRc
P1birofwAD(daeitr
3
tTtToHtewcwi
4ctcIdtD2wf
E = 10 mM nicotinic acid; sample = 2 mM pyruvate, lactate, 3-hydroxybutyrate; vaas used at all unused channels). (B) Isotachopherogram for the same separation o
urrent for the presence of the 3-hydroxybutyrate step.
C as the support was that this material bends when placed on10 ◦C hot plate. The resulting deflection caused an uneven distri-ution of temperature in different parts of the PC sheet, resulting
n irregular dimensions in the microchip channels. This issue wasesolved by placing a glass sheet (200 × 100 × 5 mm) on the topf PC sheet while baking on the hot plate. Six microchips wereabricated each on 50 × 75 mm PC substrate. An optical profileras used to select the chips with the correct channel dimensions.pproximate dimensions of narrow and wide channels after theFR fabrication process were 49/28/27 �m and 330/308/28 �m
widthtop/widthbottom/height), respectively. Reservoir holes wererilled using a 2 mm drill bit and the channels were sealed usingnother layer of Ordyl dry film. The chips were glued with 5 minpoxide glue inside the plastic frames specifically designed for usen the Bioanalyzer 2100 (provided by Agilent Technologies), withhe finished chip shown in Fig. 3. After optimizing the manual fab-ication process, the fabrication yield was approx. 50%.
.3. Performance of DFR ITP chips versus commercial DNA chips
The commercial DNA chips were shown in principle to be ableo be used for the ITP separation of carboxylic acids in serum [20].herefore the first experiments with the new design DFR chips wereo compare their performance with the commercial DNA chips.he sample was injected by applying a vacuum (vacuum injectionn the DNA chip was described in [20]). The LE contained 20 mMCl with �-alanine (pH 3.3), 1% PVP and 50 �M R6G. The TE con-
ained 10 mM nicotinic acid, and the sample was composed 2 mMach of pyruvate, lactate and 3-hydroxybutyrate. ITP separationas performed with a constant current of −0.3 �A. The pullback
urrent applied to all unused channels during the ITP separationas 0.05 �A and the resulting isotachopherograms are compared
n Fig. 4A.From Fig. 4A, the background fluorescence was approximately
times higher for the DRF-ITP chips in comparison to the DNAhips, due to the native fluorescence of the Ordyl DFR at the exci-ation wavelength. While this would increase detection limits inonventional CE, it does not have a significant adverse affect onTP separations, as the length of the step, not the height of a peak,efines the detection limit. From this Fig. 4A, it is very obvious thathe ITP steps are much longer for the DFR ITP chip compared to the
NA chip. The length of the lactate step by ITP on a DNA chip was.2 s and 42.2 s in the DFR ITP chip, an increase of 19.2. While thisas a significant improvement, the isotachopherogram obtainedrom the DFR ITP chip was missing the 3-hydroxybutyrate step.
injection; separation Iconst. = −0.3 �A, pull back Iconst. = 0.05 �A (pull back currentF-ITP chip by using separation Iconst. = −3 �A illustrating the role of the separation
This step reappeared in the isotachopherogram after the separa-tion was performed at a higher constant current of −3 �A (Fig. 4B).The reason why the step of 3-hydroxybutyrate was missing whenthe low constant current (−0.3 �A) was applied, was traced to asmall difference between the separation and pull back currents(0.05 �A). We believe that this difference caused migration of 3-hydroxybutyrate into the unused channels where the pull backcurrent was applied. Another advantage of increasing the separa-tion current in the DFR ITP chip was a considerable shortening ofthe total analysis time from 175 s to 35 s.
The PVP was used in these experiments as electroosmotic flow(EOF) suppressor. However, the method repeatability was testedand obtained RSD was 19.5%. To improve this result a polymer withhigher viscosity the HPMC was tested and RSD improved to 2.5%.The low RSD value for HPMC was found satisfactory and also indica-tive of successfully suppressing the EOF. HPMC was used as EOFsuppressor for all further experiments using DFR chips.
3.4. Influence of separation current on the isotachopherogram
The new microchip design allows the use of higher separa-tion current, which significantly reduces the separation time. Toexplore the capabilities of the DRF ITP chip design, we investigatedthe dependence of the applied separation current on the isota-chophoretic separation. The conditions used for this experimentwere slightly different from those used for the previous experi-ments, with the main difference being in the addition of 0.6% HPMCto the LE, TE and sample. HPMC was added to the TE and sampleto maintain similar viscosity with the LE, facilitating a predictablevacuum injection of electrolytes. Fig. 5A shows a plot of the relation-ship between the time necessary for the migration of chloride (LE)from the separation channel to the point of detection and the sep-aration current. The relationship between the length of the lactatestep and the applied separation current is illustrated in Fig. 5B.
As can be seen from Fig. 5A, as the current is increased, thetime for chloride to reach the detector decreases. For example,increasing the current from −1 to −2 �A, decreased the time from37.3 ± 0.1 s to 19 ± 0.3 s. From Fig. 5B, a corresponding decreasein the step length of the lactate zone, reducing from 6.7 ± 0.15 slong to 3.3 ± 0.05 s can be observed when changing the separationcurrent from −1 to −2 �A. However, at higher currents (4–6 �A)
the length of the lactate step did not become shorter. In order toachieve the higher separation currents, higher voltages are requiredwhich were beyond the 1500 V limit of the Bioanalyzer. Once thehighly conductive chloride zone had exited the microchannel, the
140 P. Smejkal et al. / Analytica Chimica Acta 803 (2013) 135– 142
F the LEa .5 mM− ars ind
vraBcriirmtpipi
3
m
Fuatp
F
ig. 5. The effect of the applied current on (A) the step length of chloride from
lanine + 0.6% HPMC + 50 �M R6G; TE = 5 mM nicotinic acid + 0.6% HPMC; sample = 25 and −6 �A. The pull back current applied to all unused channels was 0.05 �A. B
oltage increase required to maintain the constant separation cur-ent with each successively less conductive ITP zone could not bechieved by the instrument. This is rather a technical issue of theioanalyzer 2100 and not of the new chip design. However a higherurrent (−6 �A) can be used at the start of analysis when the sepa-ation channel is still filled with high conductivity ions and reducedmmediately prior to detection of the step and subsequent analyt-cal zones to keep within the limitations of the Bioanalyzer. Thiseduced current, however, also reduces the speed of the ITP zonesigrating through the detection point thereby increasing the sys-
em detection sensitivity. This approach of switching currents hasreviously been used to reduce the total analysis time of high salin-
ty samples and increase sensitivity in a capillary [13]. Using thisrinciple the analysis of lactate in human serum could be performed
n less than a minute.
.5. Lactate determination: method linearity, LOD and LOQ
To determine the performance parameters of the describedethod in the DFR chips, linearity, LOD and LOQ values were
ig. 6. Isotachopherograms of a human serum obtained from DRF ITP chip in Agilent Bioansed for the separation was −6 �A in the first stage and −2 �A (A) and −1 �A (B) in the secnalysis. The sample was prepared by mixing 45 �L of human serum with 45 �L of water awo different samples. All other conditions were the same as those in Fig. 5. Conditions
ropionic acid, sample = human serum (for more details see [13]).
igure (C) was reprinted with authors permission.
and (B) step length of lactate from the sample. Conditions: LE = 10 mM HCl + �- lactate + 0.6% HPMC; applied separation currents were −0.3, −0.5, −1, −2, −3, −4,icate the confidence interval ( = 0.95) based on three replicates.
determined. For the linearity characterization, aqueous lactatesolutions were tested in the range 0.3125–5 mM (0.3125, 1.25, 2.5,5 mM lactate). The constant separation current used for the exper-iment was −1 �A. The response was linear across the whole range(R2 = 1) and the values of limit of detection (LOD) and limit of quan-tification (LOQ) were calculated from the linear calibration usingequations LOD = 3×(SD/S) and LOQ = 10×(SD/S), where SD and Srepresent standard deviation and slope of linear calibration, respec-tively. LOD and LOQ for lactate by ITP on the DFR chips were 24 �Mand 80 �M, respectively (more details in Supporting information).
3.6. Determination of lactate in human serum
The main practical aim of this chip design was to determine car-boxylic acids in complex samples, such as human serum. Serum hashigh salinity and is rich in proteins content. Most of the advanced
separation methods will require pretreatment of serum samplesbefore injection. In contrast, ITP is able to handle high salinity sam-ples, which makes ITP an ideal choice for the analysis of these typesof samples. To mix the serum sample with aqueous solution ofalyzer 2100 (A and B) and from the literature (C) [13]. Conditions: constant currentond stage. The pullback current applied to unused channels was 0.05 �A during thend 38.6 �L of 2% HPMC. (A) and (B) were obtained from two DFR ITP chips and fromfor (C): LE = 10 mM HCl + 0.3% polyethylene glycol + �-alanine (pH 3.3), TE = 10 mM
P. Smejkal et al. / Analytica Chimic
Fig. 7. Lactate concentrations determined in three different serum samples quan-tified by CZE with indirect UV detection at 254 nm (yellow) and by ITP withindirect fluorescence detection in the DFR ITP chips using Agilent Bioanalyzer 2100(red). Both methods showed comparable results. Bars indicate confidence intervals(to
Ho
wlcCtpccybrfmh
roTvsIiTpT8omomf
4
s3s
[[
[
[[
[[
[
˛ = 0.95) based on three replicates. Conditions as in Fig. 6A. (For interpretation ofhe references to color in this figure legend, the reader is referred to the web versionf the article.)
PMC was the only pretreatment required for the method devel-ped here.
Lactate is present in serum in millimolar concentration level,hich is perfectly suitable with analytical ITP. The normal range of
actate in serum is approximately from 0.5 to 2.2 mM (the levelsan vary significantly, for example as a result of physical exercise).oncentration of chloride in serum is approximately 100 mM, andherefore the step of lactate was expected at different time to com-are with the aqueous lactate samples. Indeed, the time required forhloride to exit the separation channel was ca. 200 s for a separationurrent of −1 �A. As mentioned already, to speed up the total anal-sis time for the serum, the current switching protocol was usedy initially applying −6 �A to quickly remove the chloride (timeeduced from ca. 200 s to ca. 34 s) with a current of −2 �A usedor quantification, thus the whole separation required less than a
inute. An isotachopherogram for the determination of lactate inuman serum is shown in Fig. 6A.
The large amount of manual handling in the microchips fab-ication process made each chip unique, a factor that would bevercome once the ITP chips would be produced commercially.herefore, in this research study external calibration was incon-enient and lactate was rather quantified in serum samples bytandard addition. The quantitative results obtained from the DFRTP chip were compared with results obtained by traditional cap-llary zone electrophoresis using a method developed by Jager andavares [26]. The results obtained from three different serum sam-les by ITP in the DFR chips and by CZE are summarized in Fig. 7.he total analysis time for lactate in serum by CZE was just under
min. Eightfold improvement of analysis time was achieved withur in-house DFR ITP chips (less than 1 min). One should keep inind, however, that while analysis on the ITP chip was rapid, only
ne sample per chip could be processed and the chips requiredanual conditioning between uses, while the CE method could be
ully automated.
. Conclusions
A new ITP chip designed for determination of lactate inerum could be fabricated easily and at low-cost in Ordyl SY-00 series dry-film photoresist. Importantly, the DFR chips werehown to be fully compatible with the Agilent Bioanalyzer 2100,
[[
[
a Acta 803 (2013) 135– 142 141
whilecombining a full research flexibility based on own in-housechip designs, with the advantage of a commercial compact instru-mental platform with a sensitive LIF detection. Using 300 �m widechannels, twenty times more sample could be injected than it waspossible with the commercial DNA chips available for use withthe Bioanalyzer, which led to 20-fold increase in concentrationsensitivity with DFR-ITP chips. The wider channels also enableda 20-fold increase in the applied separation current when usedwith high conductivity samples. To optimize performance betweensensitivity and time, a current switching method was used wherehigh currents were used to quickly remove high ionic strength ions(chloride) while a lower current was used to decrease the migra-tion speed during the detection – increase the step length on theseparation record and therefore the sensitivity of the method. Thusa fast analysis was obtained without affecting the precision of theanalysis. Using this approach, lactate could be quantified in humanserum, with sample pretreatment restricted to dilution with waterand an addition of HPMC to maintain equal viscosity across all elec-trolytes in the chip. The results for the DFR-ITP chips were in goodagreement with results from the same samples analyzed by CZE ina capillary electrophoresis instrument, with the chip-ITP methodmore than eight times faster than CZE. Future developments wouldfocus on development of ITP chip designs and analysis protocolsthat would allow fully automated analysis and optimally also par-allel analysis of more than one sample at a time.
Acknowledgements
This work was supported by the Grant Agency of the CzechRepublic (P301/11/2055 and P206/12/G014) and the InstitutionalResearch Plan (UIACH 68081715). MCB would like to thank theAustralian Research Council for funding and provision of a QEIIFellowship (DP0984745), and MM would like to acknowledge theAustralian Research Council for funding and provision of a FutureFellowship (FT120100559).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.aca.2013.01.046.
References
[1] N.V. Lavrik, L.T. Taylor, M.J. Sepaniak, Anal. Chim. Acta 694 (2011) 6.[2] J.P. Kutter, J. Chromatogr. A 1221 (2012) 72.[3] S.M. Kenyon, M.M. Meighan, M.A. Hayes, Electrophoresis 32 (2011) 482.[4] D. Mark, S. Haeberle, G. Roth, F. von Stetten, R. Zengerle, Chem. Soc. Rev. 39
(2010) 1153.[5] C.D. Chin, V. Linder, S.K. Sia, Lab Chip 12 (2012) 2118.[6] P. Smejkal, F. Foret, Chem. Listy 106 (2012) 104.[7] D. Mark, F. von Stetten, R. Zengerle, Lab Chip 12 (2012) 2464.[8] M. Ryvolova, J. Preisler, D. Brabazon, M. Macka, Trac-Trends Anal. Chem. 29
(2010) 339.[9] P. Smejkal, A. Szekrenyes, M. Ryvolova, F. Foret, A. Guttman, F. Bek, M. Macka,
Electrophoresis 31 (2010) 3783.10] A. Lloyd, L. Blanes, A. Beavis, C. Roux, P. Doble, Anal. Methods 3 (2011) 1535.11] P. Smejkal, M.C. Breadmore, R.M. Guijt, F. Foret, F. Bek, M. Macka, Electrophore-
sis 33 (2012) 3166.12] T. Verheggen, F. Mikkers, F. Everaerts, F. Oerlemans, C. Debruyn, J. Chromatogr.
182 (1980) 317.13] V. Dolnik, P. Bocek, J. Chromatogr. 225 (1981) 455.14] A.A.G. Lemmens, J.C. Reijenga, F.M. Everaerts, R.T.P. Janssen, J. Hulsman, C.A.M.
Meijers, J. Chromatogr. 320 (1985) 193.15] V. Dolnik, M. Deml, P. Bocek, Electrophoresis 9 (1988) 839.16] I. Valaskova, J. Balazova, E. Havranek, J. Chromatogr. B Biomed. Appl. 674 (1995)
310.17] J. Sadecka, J. Polonsky, J. Chromatogr. A 735 (1996) 403.
18] J. Sadecka, A. Hercegova, J. Polonsky, J. Chromatogr. B 729 (1999) 11.19] M. Deml, P. Gebauer, V. Dolnik, P. Bocek (Eds.), Analytical isotachophoresis,VCH Verlagsgesellschaft, Weinheim, 1988, ISBN 3527264441.20] P. Smejkal, M.C. Breadmore, R.M. Guijt, J. Grym, F. Foret, F. Bek, M. Macka, Anal.
Chim. Acta 755 (2012) 115–120.
1 himic
[[[
[[[[
[
[29] R. Bodor, M. Zuborova, E. Olvecka, V. Madajova, M. Masar, D. Kaniansky, B.Stanislawski, J. Sep. Sci. 24 (2001) 802.
42 P. Smejkal et al. / Analytica C
21] V. Dolnik, M. Deml, P. Bocek, J. Chromatogr. 320 (1985) 89.22] D. Bottenus, T.Z. Jubery, P. Dutta, C.F. Ivory, Electrophoresis 32 (2011) 550.23] D. Bottenus, T.Z. Jubery, Y.X. Ouyang, W.J. Dong, P. Dutta, C.F. Ivory, Lab Chip 11
(2011) 890.
24] H. Becker, L.E. Locascio, Talanta 56 (2002) 267.25] R.M. Guijt, E. Candish, M.C. Breadmore, Electrophoresis 30 (2009) 4219.26] A.V. Jager, M.F.M. Tavares, Electrophoresis 24 (2003) 1208.27] T. Hirokawa, M. Nishino, N. Aoki, Y. Kiso, Y. Sawamoto, T. Yagi, J.I. Akiyama, J.Chromatogr. 271 (1983) D1.
[
[
a Acta 803 (2013) 135– 142
28] R. Bodor, V. Madajova, D. Kaniansky, M. Masar, M. Johnck, B. Stanislawski, J.Chromatogr. A 916 (2001) 155.
30] J.E. Prest, S.J. Baldock, P.R. Fielden, N.J. Goddard, B.J.T. Brown, J. Chromatogr. A990 (2003) 325.
31] http://web.natur.cuni.cz/gas/, 23.11.2012.
