Low-cost, fast prototyping method of fabrication of the microreactor devices in soda-lime glass

Abstract

In the context of biochips, the novel concept of a fast prototyping platform with application to surface chemistry studies is presented. Microreactors are fabricated in microscope glass slides, allowing for parallel screening. Both, the use of low-cost soda-lime glass and simplified, yet effective, method of masking render our approach affordable and flexible—which are two essential features of fast prototyping. The fabrication process is based on the deep etching of soda-lime glass slides. A photoresist/Cr bi-layer was used as a mask in order to support the deep etching of glass. A buffered oxide etchant/hydrochloric acid mixture was employed as an etching reagent, and etching conditions, namely composition and temperature, were optimised. Structures with depths exceeding 100 μm and surface roughness below 10 nm were obtained. The application of the devices for oligopeptide immobilisation is demonstrated.

Full text

Page 1

Available online at www.sciencedirect.com
Sensors and Actuators B 128 (2008) 552–559
Low-cost, fast prototyping method of fabrication of
the microreactor devices in soda-lime glass
R. Mazurczyka,1, G. El Khourya,1, V. Dugasb,2, B. Hannesa,1, E. Laurenceaua,1,
M. Cabreraa,1, S. Krawczyka,1, E. Souteyranda,1, J.P. Cloareca,1, Y. Chevolota,∗,1
aLaboratoire d’Electronique, d’Opto´ electronique et Microsyst` emes-LEOM UMR CNRS 5512 Ecole Centrale de Lyon,
36 Avenue Guy de Collongue, 69134 Ecully, France
bROSATech, 36 Avenue Guy de Collongue, 69134 Ecully, France
Received 25 April 2007; received in revised form 29 June 2007; accepted 6 July 2007
Available online 12 July 2007
Abstract
Inthecontextofbiochips,thenovelconceptofafastprototypingplatformwithapplicationtosurfacechemistrystudiesispresented.Microreactors
are fabricated in microscope glass slides, allowing for parallel screening. Both, the use of low-cost soda-lime glass and simplified, yet effective,
method of masking render our approach affordable and flexible—which are two essential features of fast prototyping. The fabrication process
is based on the deep etching of soda-lime glass slides. A photoresist/Cr bi-layer was used as a mask in order to support the deep etching of
glass. A buffered oxide etchant/hydrochloric acid mixture was employed as an etching reagent, and etching conditions, namely composition and
temperature, were optimised. Structures with depths exceeding 100?m and surface roughness below 10nm were obtained. The application of the
devices for oligopeptide immobilisation is demonstrated.
© 2007 Elsevier B.V. All rights reserved.
Keywords: Soda-lime glass; Etching; Microreactor; Surface immobilisation; Prototyping
1. Introduction
Microarraysonsolidsurfaceshavecausedarevolutioninpro-
teomics,allowingforhigh-throughputscreeningofproteininter-
actions.Meanwhileitpromptsthequestionofproteinbehaviour
at the surface of a material. Protein properties may be affected
differently by the surface vicinity. One important difficulty is
the grafting of functional protein layers on “useful” surfaces.
UnlikeDNA,proteinsmustretainanintricatethree-dimensional
structure to maintain their functionality. The integrity of this
molecular structure upon immobilisation depends on many
physico-chemical factors such as surface chemistry (surface
energy, surface charge, spacer length, etc.), surrounding buffer
∗Corresponding author. Tel.: +33 472 18 62 40.
E-mail address: yann.chevolot@ec-lyon.fr (Y. Chevolot).
1Present address: Institut des Nanotechnologies de Lyon-INL UMR CNRS
5270 Ecole Centrale de Lyon, 36 Avenue Guy de Collongue, 69134 Ecully,
France.
2Present address: Biotray, ENS-Lyon 46, All´ ee d’Italie 69364 Lyon Cedex
07, France.
(pH,ionicstrength,etc.).Asaresultoftheformationofanionic
double layer (Helmholtz double layer) at a solid/liquid inter-
face of an immersed material and due to the presence of specific
surface chemical groups, pH and ionic strength are usually dif-
ferent at the surface of a material (Debye length 0.5–1nm) than
in the bulk solution. In this context and due to the lack of a
general model, there is a crucial need of developing tools for
rapid screening of chemical environment surrounding immo-
bilised protein, i.e. immobilisation protocols and procedure for
their use in the presence of an appropriate buffer.
Theuseofmicrowells,onthecontrarytoflatsubstrates,ascer-
tains the confinement of various reactants, independently from
the surface energy.
Soda-lime glass is the material of choice for the fabrica-
tion of biochip/lab-on-a-chip devices due to its optical and
electrical properties, the availability of silanes for its surface
functionalisation, and its low cost. Moreover, recent studies
have demonstrated the possibility of monolithic integration of
microfluidic vessels and detection optics in one, common soda-
lime glass substrate [1,2]. Direct comparison of parameters on
the immobilisation yield or biorecognition efficiency, requires
0925-4005/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.snb.2007.07.033

Page 2

R. Mazurczyk et al. / Sensors and Actuators B 128 (2008) 552–559
553
having a similar specific surface area for each microreactor on
the slide. Specific surface area depends on surface roughness.
Thus,thispropertyshouldbecarefullycontrolledwhenplanning
the fabrication technology.
Inthiscontext,thereisagrowinginterestfordevelopingdeep
etching technologies of soda-lime glass.
A number of concepts of fabrication of microstructures in
glass already exist. Mechanical processing methods such as
sandblasting or diamond sawing [3–5] feature simplicity, but
they tend to generate a rough surface. Although the success-
ful use of dry etching methods for structuring glasses has been
reported[6–8],relativelylowetchrates(e.g.forsoda-limeglass
significantly less than 1?m/min) represents a major drawback
of this approach—in particular, if structures with depths of
100?m or more, are considered. Moreover, non-volatile chemi-
calspecies(fluoridesofmetals)aregeneratedduringtheprocess
[7] deteriorating the surface’s quality.
For years, wet etching has been used for the fabrication of
microstructures in glass [9–19]. Two principal issues related
to this method are: the appropriate choice of etching reagents
and the design of masking layers. As for the former, HF-based
etchants are typically used. Unfortunately, in the case of soda-
limeglass,theproblemofnonwater-solubledepositsarisesfrom
thepresenceofCaO,MgO,andAl2O3intheglass[15,16].These
depositspreventtheetchingsolutionfromcompleteanduniform
access to the structured substrate, resulting in the rough surface
of a processed microdevice. The addition of HCl to the etching
mixtureisthetypicalwayofovercomingthistechnologicalhur-
dle [15,16]. Various materials, such as photoresist [9,10,13,19],
polysilicon [9,15], Cr/Au multilayers [9,11,15,17,19], anodi-
cally bonded silicon wafer [12] were used as masking layers
in the wet etching processes. As far as the photoresist masking
is concerned, both its chemical resistance and adhesion to the
surface of glass, turn out to be insufficient in a harsh etching
environment. Hence, this approach is limited to the cases when
the etched depths do not exceed 30?m approximately [13].
In our case, a photoresist/Cr bi-layer was used as a mask. Cr
layers feature good adhesion to glass, however pinholes spoil
the integrity of the Cr mask and, as a result of the penetration
of the etching solution, these pinholes cause local etching of the
surface of a device [19], leading to surface defects. Due to this,
the mere Cr film cannot be considered as a mask—its role is
to improve adhesion between a glass substrate and overlaying
films.Inparticular,aphotoresist/Crbi-layerappearstobeworth
mentioning here due to simplicity of its fabrication [18].
Here we report the fabrication technology of micro to pico-
volumetricreactorsinsoda-limeglass,whichallowsforparallel
screening of various protocols of immobilisation or working
buffers (up to a few thousands). In our work, we developed
a straightforward, affordable and flexible manufacturing pro-
cess of such microsystems in low cost soda-lime microscope
slides. The use of photoresist/Cr bi-layer as a mask remark-
ably reduced the number of technological steps needed and still
proved effective in etching features deeper than 100?m. As a
proof-of-concept,wedemonstratetheapplicationofourdevices
for immobilisation and detection of oligopeptides.
2. Experimental
2.1. Substrate preparation and photolithography
Soda-lime microscope slides (Menzel, LR90) were used as
substrates for the fabrication of microreactors. The approximate
chemical composition of this glass (according to the manufac-
turer) is as follows: SiO272.20%, Na2O 14.30%, K2O 1.20%,
CaO 6.40%, MgO 4.30%, Al2O31.20%, Fe2O30.03%, SO3
0.30%.
The process flow chart is presented in Fig. 1. First, the
substrates were cleaned for 10min with Piranha mixture (sul-
phuric acid 95%:hydrogen peroxide 30%, 3:1 volume ratio).
After 20min of rinsing in running DI water and spin drying,
a sputtering (MRC 822 system) method was used in order to
deposit 150nm layer of chromium on both sides of the sub-
strates (step 1). Next, (step 2), the standard photolithography
process, based on the SRJ 5740 photoresist (Shipley), was car-
ried out. At the end of the process, substrates were post-baked
in the furnace at 110◦C for 10min. Openings in the masking
layer were of circular shape, with their diameter ranging from
100?m to 4mm.
SEM images were taken with a Jeol JSM-T200 apparatus.
Depth and roughness were measured with a mechanical pro-
filer (Alfa-step 500 from KLA Tencor), over the distance of
200?m.
Fig. 1. Technology process flow of the device fabrication.

Page 3

554
R. Mazurczyk et al. / Sensors and Actuators B 128 (2008) 552–559
2.2. Etching
Inthestep3(Fig.1),windowsintheCrlayerwereopenedby
wet etching (Merck chromium etchant). As a result, the double,
hard/soft(150nmofCr/8?mofphotoresist),maskinglayerwas
prepared. In this system, the chromium layer served as an adhe-
sion promoter between the glass substrate and the photoresist
film, whereas the layer of photoresist filled the pinholes [11,19]
in the chromium film and provided an additional protection
against harsh environment of the glass etching chemicals.
As a subsequent phase of the fabrication, several wet etch-
ing processes were carried out (step 4). The etching agents
were based either on hydrofluoric acid (HF, 40%) or buffered
oxideetchant(BOE),withadditionsofhydrobromicacid(HBr),
hydrochloric acid (HCl, 37%) or EDTA, all supposed to elimi-
nate the problem of solid precipitates.
Various proportions of the components and process temper-
atures were tested in order to minimise the amount of insoluble
precipitates and possible damages of the masking layers, and to
maximise the etching rate at the same time. Temperature and
mixing rate were controlled using a thermomixer (Eppendorf).
Finally,aftertheetchingstephadbeenfinished,theprotective
layers were removed with the use of chromium etchant (Merck)
and acetone—step 5.
2.3. Silanisation of glass slides
We fabricated micro-reactors of 4mm diameter and 100?m
depth which corresponds to approximately a 1?l volume and
which are convenient to fill up with a micropipette. The slides
typically featured 21 micro-reactors (4mm diameter, 100?m
deep).
The surface of slides was functionalised according to
the protocol developed by Dugas et al. [20]. Briefly,
the slides were piranha washed for 20min, rinsed in DI
water and dried by centrifugation. After heating under
dry nitrogen for 2h at 150◦C, dry pentane was added
at room temperature (RT), followed by 300?l of tert-
butyl-11-(dimethylamino)silylundecanoate. After 2h at RT
under N2, the pentane was removed by evaporation. The
slides were heated at 150◦C overnight, washed in THF
and rinsed in water. The ester was converted into the
acidic function in CH2Cl2/TFA (1/1) for 2h at RT.
Acid groups were activated for amine coupling with N-
hydroxysuccinimide(0.1M)/di(isopropyl)carbodiimide(0.1M)
in dry THF, overnight at RT. The biomolecules were either
directly immobilised on the NHS ester activated slides via their
amine functions.
2.4. Material for oligopeptide immobilisation
The flow chart of the peptide immobilisation is presented in
Fig. 2. The following chemicals were used in this experiment.
Biotinylated dodecapeptide (biot-Gly-His6-Gly2-Lys-Gly2)
was synthesised and purified at LBASB CNRS/University
Claude Bernard Lyon I. Cy3-sterptavidin conjugate solution
(1mg/ml)wassuppliedfromAmershamBiosciencesandbovine
Fig. 2. Flow chart of the peptide immobilisation. (a) Immobilisation of bio-
tynilated peptide. (b) Labelling of immobilised peptide with Cy3 conjugate
streptavidin.
serum albumin (BSA) from Sigma–Aldrich. Ultra-pure water
(18.2M?) was used.
Buffer solutions of the monobasic and dibasic forms of
sodium phosphate (20mM) were prepared at various pH (6.5,
7.4, 8.5, 9.0). One percent Triton X-100 solution was obtained
by diluting 1ml Triton X-100 in 100ml ultra pure water. Block-
ingsolutionwas2%(w/v)BSAin1×D-PBS(pH7.4).Washing
buffer was PBS-T (1× PBS, 0.05% Tween 20).
2.5. Oligopeptide immobilisation
Solutionsofbiotinylateddodecapeptide(2×10−4,10−3and
2×10−3M) were prepared in sodium phosphate buffer, at var-
ious pH (6.5, 7.4, 8.5 and 9.0). These solutions were deposited
onto functionalised reactors, and left to react for 3h at 21◦C
under water vapour saturated atmosphere. The slides were then
rinsedwithultra-purewater,washedwith1%TritonX-100aque-

Page 4

R. Mazurczyk et al. / Sensors and Actuators B 128 (2008) 552–559
555
ous solution for 5min at RT under sonication and then for 1h at
80◦C, washed with ultra-pure water for 5min under sonication,
and finally dried by centrifugation for 2min at 1300rpm.
The slides were incubated for 1h with the blocking solu-
tion (2% BSA/1× PBS) at 21◦C, under water vapour saturated
atmosphere, then washed 3× 3min in PBS-T at room temper-
ature, under sonication, rinsed with ultra-pure water, and dried
by centrifugation for 2min at 1300rpm.
2.6. Labelling of immobilised oligopeptide with
Cy3-streptavidin conjugates
Cy3-streptavidin conjugate solution was diluted 1:100 with
water. Two microliters of diluted solution were deposited into
each reactor and left to incubate overnight at 21◦C in darkness
and under water vapour saturated atmosphere. The slides were
then thoroughly rinsed with water, then washed 3× 3min in
PBS-Tatroomtemperature,undersonication,rinsedwithultra-
pure water, and dried by centrifugation for 2min at 1300rpm.
2.7. Fluorescence scanning and image analysis
Slides were scanned using the GenePix®Personal 4100A
microarray scanner (Axon Instruments/Molecular Devices,
Union City, USA) at a laser excitation wavelength of 532nm,
20?m pixel resolution and a PhotoMultiplier Tube (PMT) gain
of 500. Images were analysed and results were stored and man-
aged in a previously defined database. The complete data set
analysed consisted of the average pixel fluorescence intensities
of 21 spots (4mm diameter). The results obtained correspond to
the average of two independent measurements.
3. Results
3.1. Fabrication
PreliminarytestsrevealedthatonlytheBOE/HClbasedmix-
tureswereviableetchingagents.Ineachoftheothercases,either
theetchrateturnedoutunsatisfactoryand/ortheamountofsolid
debrisand/orthehighsurfaceroughnessdidnotfulfilourneeds.
As we have mentioned, the action of HCl in the etching mix-
ture is supposed to eliminate any insoluble precipitates formed
during the process. Therefore, the overall etching mechanism is
controlled by two complementary phenomena: the etching rate
of silicon dioxide by BOE and the removal of insoluble material
by HCl, allowing HF to reach SiO2sites.
In Fig. 3, we present the etching rate temperature depen-
dencies for etchants featuring BOE/HCl ratios between 1.4 and
3. Up to ca. 40◦C, the total etching rate depends mainly on
BOE contents, whereas for the temperature of 55◦C the etch-
ing rate is governed predominantly by temperature. Indeed, we
observed that the etching rate for the 2.33 and 2.06(BOE/HCl)
ratios is higher (or in the same range, as for the 1.4 ratio) as
compared with the ratio of 3. This behaviour can be explained
in the following way.
In the low temperature range the amount of debris formed
is relatively small, thus it is the efficiency of the SiO2removal
Fig. 3. Dependence of the soda-lime glass etch rate on temperature, for various
BOE/HCl ratios.
(proportional to the BOE contents) that controls the process. As
temperature increases, the elimination of precipitates becomes
the limiting factor. Hence, by increasing the concentration of
HCl,itispossibletoeliminatetheinsolublematerialfaster.Con-
sequently, at 55◦C the etching rate is higher for the HCl richer
solutions, since the process is governed by the rate of debris
removal.
Apart from the etching rate, the monitoring of surface rough-
ness over the whole device surface is a factor of paramount
importance. First, it influences the process of surface function-
alisationbyincreasingtheavailablefunctionalgroups.Second,it
affects the optical performance of the fluorescence-based detec-
tion system. The surface roughness for various temperatures
and ratios of BOE/HCl (1.04, 2.06, 2.33 and 3) is illustrated in
Fig. 4. From these data, the following general conclusions can
be drawn. The high percentage of BOE in the etchant results
in the rough surfaces of the devices. Considering the 3 and
2.33BOE/HCl etching mixtures, the surface roughness reaches
tensofnm,whereasforthe1.04and2.06BOE/HCletchingmix-
tures, it stays in the range of just a few nm—which is clearly
more favourable.
For the 1.04 and 2.06 BOE/HCl etching mixtures, the rough-
ness appears to be an increasing function of temperature and at
the temperatures above ca. 40◦C, it exceeds 10nm. Standard
deviation of surface roughness is high for etching solution con-
Fig. 4. Dependence of the surface roughness on temperature, for various
BOE/HCl ratios.

Page 5

556
R. Mazurczyk et al. / Sensors and Actuators B 128 (2008) 552–559
Fig. 5. SEM image of the microreactor well.
taininglowconcentrationofHClmeaningthatroughnesscannot
be controlled with this etchant.
In the case of a 2.06 BOE/HCl etching mixture at 37◦C the
surfaceroughnessandtheetchingratewereindependentofetch-
ing time contrary to what was observed by Iliescu et al. [13]. In
ourcase,theconcentrationofHClishigher,leadingtothebetter
dissolution of the debris.
Finally, taking into account the above results, the BOE/HCl
mixture of 2.06 ratio and the working temperature of 37◦C
were selected as our standard etching conditions. This provided
us with an acceptable etching rate of 3.8±0.1?m/min while
yielding a low surface roughness of approx. 5nm. An exam-
ple of the successful application of these conditions is given in
Fig. 5, where the SEM image of a single micro-reactor (diam-
eter 700?m and depth 100?m) is shown. Differences in depth
of the reactors over the microscope glass slide were less than
3?m. Similarly, average roughness was below 5nm over the
entire device.
For smaller features (diameter of 100?m), mixtures of
BOE/HCl diluted with water were studied. Our objective in
this case was to elaborate the etchants featuring lower etching
rates, thus enabling fabrication of subtle structures with high
precision. In Fig. 6, the exemplary performance of these etch-
ing agents is presented, for the proportions of 1:2:4 and 1:2:2
of BOE (7/1):HCl (37%):H2O and process temperatures rang-
ing from 20◦C up to 35◦C. Etching rates fell in the range of
0.4–1.2?m/min in this case. Importantly, no debris of any kind
was revealed in the reactors and, again, excellent surface qual-
ity (roughness below 5nm) was accomplished. The application
of one of these etchants (proportions 1:2:4, process tempera-
ture 30◦C) is demonstrated in Fig. 7, where the SEM image
of 100?m diameter microreactor is shown. The depth of the
structure is 20.0±0.1?m.
3.2. The influence of etching on fluorescence signal
Despiteourinitialconcerns,itturnedoutthatthebackground
fluorescencesignaloftheprocessedslideswasofthesameorder
Fig. 6. Temperature dependence of the etch rate for BOE:HCl:H2O etchants:
volume ratios of 1:2:2 (circles) and 1:2:4 (squares).
Fig. 7. SEM image of a microreactor, diameter of 100?m, depth of 14.8?m.
ofmagnitudeasfortheordinarymicroscopeglassslides.There-
fore,thefabricationprocessdoesnotimpairtheuseofthedevice
for fluorescent detection of biomolecules (the detection method
most commonly used in microarrays).
The fluorescent signal depends, among others, on the focus-
ing of the excitation laser at the surface of the substrate. Due
to the geometry of the device, the bottom of the microreac-
tor and the slide surface cannot be in focus simultaneously. In
Fig. 8, we present the results of an experiment, where spots
of a 10−2mg/ml of Cy3-streptavidin conjugate solution were
depositedonthesurfaceoftheglassslideandatthebottomofthe
reactor. The fluorescence histogram was integrated to take into
Fig. 8. Influence of the etching process on fluorescence signal.

Page 6

R. Mazurczyk et al. / Sensors and Actuators B 128 (2008) 552–559
557
account difference in the deposited surface area. As illustrated
in Fig. 8, when focused on the surface of the slides, the signal
emitted from the bottom of a 100?m deep reactor is roughly
three times lower than the one originating from the surface of
the slide.
Consequently,thebackgroundsignalcollectedfrombetween
the microreactors is enhanced relatively to the signal originated
from the bottom of the microreactors. This means that one must
be careful when interpreting the image not to overestimate the
background signal.
3.3. Effect of fabrication on the silanisation-illustrated with
the immobilisation of biotinylated oligopeptide
Thenextstepwastakentoensurethatthefabricationprocess
was compatible with surface silanisation of glass.
Immobilisation of a biotinylated dodecapeptide (biot-Gly-
His6-Gly2-Lys-Gly2) was performed on NHS ester activated
surfaces. The peptide was deposited in the microreactors and its
surface density was evaluated with fluorescent conjugate Cy3-
streptavidin (flow chart, Fig. 2). Immobilisation experiments
were performed with six repetitions for three different concen-
trations(2×10−4,10−3and2×10−3M)ofpeptidesolutionsin
sodiumphosphatebufferatpH6.5.Fig.9showstheaveragesig-
nal for each concentration and the standard deviation measured.
Thestandarddeviationofthefluorescentsignalisapproximately
3%. This is compatible with the deviation typical for standard
microarrays. As explained previously, variations of focal length
resulting from different etching depths, contribute to this devia-
tion. However in our case, the variation of etching depth is near
1%, thus we consider its influence minimal.
From these results, we can also conclude that the microre-
actors’ inside surface has properties similar to the non-etched
surface, thereby allowing grafting of a functional oligopeptide.
BiotinislinkedattheN-terminusoftheoligopeptide.Conse-
quently, the oligopeptide was covalently immobilised onto the
solid support via the amino group of the side chain of its lysine
residue. In order to study the influence of the pH on the stabil-
ity and on the quality of the immobilisation, the oligopeptide
was solubilised in sodium phosphate buffer at various pH (6.5,
7.4, 8.5 and 9.0). The density of immobilised biotinylated pep-
Fig. 9. Reproducibility of peptide immobilisation at different concentrations.
Fig. 10. Effect of pH on biotynilated peptide immobilisation characterised by
indirectlabellingusingCy3streptavidinconjugateontoacid-derivatisedsurface.
tide was indirectly determined with Cy3-streptavidin conjugate.
Indeed, the density of oligopeptide molecules immobilised onto
acid-derivative surface was considered as directly proportional
to the fluorescence signal intensity. Aiming at the reduction of
the non-specific adsorption of Cy3-streptavidin conjugate onto
the solid support after peptide immobilisation, remaining active
sites were capped with BSA solution.
As shown in Figs. 9 and 10, the highest fluorescent signal is
obtainedfortheoligopeptideconcentrationof2×10−3M,with
the spotting buffer at pH 6.5. The signal to background ratio is
above three. Furthermore, at this concentration (2×10−3M)
independently of pH, the fluorescent signal is significantly
higher than the background (i.e. buffer only)—Fig. 10. These
results indicate that the surface density of grafted peptide
depends on the pH of the spotting buffer. Indeed, density of
grafted peptide decreases with increasing pH. As we have
explainedabove,onlytheaminefunctionofthelysinesidechain
is available for immobilisation of the peptide onto the surface.
Taking into account that the pKaof this function is 10.67, it is
worth noting that even though the pH of the spotting buffer was
below or equal to 9, a pH dependency is observed for peptide
immobilisation. However, it is likely that pH at the solid–liquid
interface is different from pH in the bulk volume. The covalent
bond formation between the oligopeptide and the solid surface
may be directed by numerous phenomena such as electrostatic
interactions, diffusion, thermodynamic and steric parameters,
etc. To elucidate mechanisms driving this chemical reaction,
further experiments including more conditions are in progress.
From these data, we can conclude that the highest surface
densityofgraftedoligopeptideisobtainedwithspottingsodium
phosphate buffer at pH 6.5.
4. Conclusions
The results presented herein prove that simple and low cost
technology of fabrication of microreactor arrays in soda-lime
glass has been elaborated. The technology retains surface prop-
erties of the original glass substrate, which makes it possible
to use the existing methods of glass surface functionalisation.
Also, the processing methods employed do not increase the
fluorescence background level of the final device. Indeed, par-

Page 7

558
R. Mazurczyk et al. / Sensors and Actuators B 128 (2008) 552–559
allel screening has allowed determination of optimal conditions
for covalent immobilisation of the oligopeptide onto the sur-
face. This technology is also very helpful to study mechanisms
involved in immobilisation process of biomolecules onto solid
surfaces. In future work, our aim is to fabricate devices featur-
ing reactor volumes ranging from nanoliters to picoliters. In this
case, an in-house microprojection machine, capable of supply-
ing such low sample amounts, would be utilised instead of the
pipette. Furthermore, such microreactors on glass slides can be
integratedwithothermicrofluidicanddetectioncomponentsfor
the elaboration of more complex micro-devices.
Acknowledgments
J. Wallach and M.C. Duclos (LBASB CNRS/University
Claude Bernard Lyon) are greatly acknowledged for providing
the synthetic biotinylated dodecapeptide (biot-Gly-His6-Gly2-
Lys-Gly2).
Region Rhˆ one Alpes is greatly acknowledged for financial
support.
Finally,wewouldliketothankDr.ColinMansfieldforcareful
reading of the manuscript and fruitful discussion.
References
[1] J.Vieillard,R.Mazurczyk,C.Morin,B.Hannes,Y.Chevolot,P.L.Desb` ene,
S. Krawczyk, Application of microfluidic chip with integrated optics
for electrophoretic separations of proteins, J. Chromatogr. B 845 (2007)
218–225.
[2] R. Mazurczyk, J. Vieillard, A. Bouchard, B. Hannes, S.K. Krawczyk, A
novelconceptoftheintegratedfluorescencedetectionsystemanditsappli-
cation in a lab-on-a-chip microdevice, Sens. Actuators, B, Chem. 118
(2006) 11–19.
[3] Q.S. Pu, R. Luttge, H.J.G.E. Gardeniers, A. van den Berg, Comparison of
capillary zone electrophoresis performance of powder-blasted and hydro-
gen fluoride-etched microchannels in glass, Electrophoresis 24 (2003)
162–171.
[4] E. Belloy, S. Thurre, E. Walckiers, A. Sayah, M.A.M. Gijs, The introduc-
tion of powder blasting for sensor and microsystem applications, Sens.
Actuators, A, Phys. 84 (2000) 330–337.
[5] L. Szekely, R. Freitag, Fabrication of a versatile microanalytical system
without need for clean room conditions, Anal. Chim. Acta 512 (2004)
39–47.
[6] I.Rodriguez,P.Spicar-Mihalic,C.L.Kuyper,G.S.Fiorini,D.T.Chiu,Rapid
prototypingofglassmicrochannels,Anal.Chim.Acta496(2003)205–215.
[7] T. Ichiki, Y. Sugiyama, R. Taura, T. Koidesawa, Y. Horiike, Plasma appli-
cations for biochip technology, Thin Solid Films 435 (2003) 62–68.
[8] X. Li, T. Abe, M. Esashi, Deep reactive ion etching of Pyrex glass using
SF6plasma, Sens. Actuators, A, Phys. 87 (2001) 139–145.
[9] P.C. Simpson, A.T. Woolley, R.A. Mathies, Microfabrication of 96 sample
capillaryarrayelectrophoresischips,Biomed.Microdevices1(1998)7–26.
[10] M.A. Holden, S. Kumar, E.T. Castellana, A. Beskok, P.S. Cremer, Gener-
ating fixed concentration arrays in a microfluidic device, Sens. Actuators,
B, Chem. 92 (2003) 199–207.
[11] M. Bu, T. Melvin, G.J. Ensell, J.S. Wilkinson, A.G.R. Evans, A new mask-
ingtechnologyfordeepglassetchinganditsmicrofluidicapplication,Sens.
Actuators, A, Phys. 115 (2004) 476–482.
[12] R.Wilke,S.B¨ uttgenbach,Amicromachinedcapillaryelectrophoresischip
with fully integrated electrodes for separation and electrochemical detec-
tion, Biosens. Bioelectron. 19 (2003) 149–153.
[13] A. Grosse, M. Grewe, Henning Fouckhardt, Deep wet etching of fused
silica glass for hollow capillary optical leaky waveguides in microfluidic
devices, J. Micromech. Microeng. 11 (2001) 257–262.
[14] T. Corman, P. Enoksson, G. Stemme, Deep wet etching of borosilicate
glass using an anodically bonded silicon substrate as mask, J. Micromech.
Microeng. 8 (1998) 84–87.
[15] C. Iliescu, J. Jing, F.E.H. Tay, J. Miao, T. Sun, Characterization of masking
layers for deep wet etching of glass in an improved HF/HCl solution, Surf.
Coat. Technol. 198 (2005) 314–318.
[16] M.Stjernstr¨ om,J.Roeraade,Methodforfabricationofmicrofluidicsystems
in glass, J. Micromech. Microeng. 8 (1998) 33–38.
[17] Y. Mourzina, A. Steffen, A. Offenh¨ ausser, The evaporated metal masks
for chemical glass etching for BioMEMS, Microsyst. Technol. 11 (2005)
135–140.
[18] R.N.C.Daykin,S.J.Haswell,Developmentofamicroflowinjectionmani-
foldforthedeterminationoforthophosphate,Anal.Chim.Acta313(1995)
155–159.
[19] C. Iliescu, J. Miao, F.E.H. Tay, Stress control in masking layers for deep
wet micromachining of Pyrex glass, Sens. Actuators, A, Phys. 117 (2005)
286–292.
[20] V. Dugas, G. Depret, Y. Chevalier, X. Nesme,´E. Souteyrand, Immobiliza-
tionofsingle-strandedDNAfragmentstosolidsurfacesandtheirrepeatable
specific hybridization: covalent binding or adsorption? Sens. Actuators, B,
Chem. 101 (2004) 112–121.
Biographies
R. Mazurczyk received his PhD in electrical engineering in 2002 from the
Technical University of Lodz, Lodz, Poland. Between 2003 and 2005 he
was a post-doc fellow at Laboratoire d’Electronique, Opto´ electronique et
Microsyst` emes,EcoleCentraledeLyon,France.Currently,heisaresearchengi-
neer at the same laboratory. His domains of research comprise semiconductor
and microsystems technology, surface and thin films science and technology.
G. El Khoury received her MSc degree in physiology in 2004 from the Claude
Bernard University (Lyon, France). Since 2004, she has been preparing her PhD
thesisinbiotechnologyatEcoleCentraledeLyon(Ecully,France).Herresearch
subject concerns the implementation of protein and peptide microarrays.
V. Dugas received the PhD degree in chemistry and biology from the Ecole
Centrale de Lyon at Ecully (France) in 2001. Between 2001 and 2005, he was
a project manager in RosaTech S.A. (Ecully, France) and his activities were
focusedonthedevelopmentofbiochipsforbiomedicaldiagnosticsapplications.
In November 2005, he joined BioTray S.A.S. (Lyon, France) and develops sci-
entific instruments in the micro-biosystems field. His research interests concern
the surface bio-functionalization and characterization, interfacial phenomena
and microsystems.
B. Hannes obtained his master’s degree in microelectronics and microsys-
tems from the Institut National des Sciences Appliqu´ ees of Lyon (INSA). He
is currently a PhD student at Laboratoire d’Electronique, Opto´ electronique et
Microsyst` emes, Ecole Centrale de Lyon, France. His main interests are new
microfabrication technologies for microsystems in general and the development
and use of microsystems for biomedical applications in particular.
E. Laurenceau obtained her PhD degree in chemistry, speciality biomaterials
in 1995. From 1996 to 2004, she was an assistant professor in chemistry at
Paris XIII-University (France). Since 2004 she has been an assistant professor
in chemistry at Ecole Centrale of Lyon (France). Her fields of interest concern
protein–surface interactions, surface functionalisation, biosensors.
M. Cabrera is a researcher at the Centre National de la Recherche Scientifiqe
at the Institut des Nanotechnologies de Lyon (INL). He received an engineering
degree at the Ecole Sup´ erieure d’Electricit´ e, and his PhD degree in 1986. His
current interests include the development of non-conventional rapid machin-
ing techniques to prepare microfluidics and biosensor devices including micro
contact printing and micro and nano electrical machining.
S. Krawczyk has PhD in physics and technology of electronic materials and
devices (Ecole Polytechnique de Warsaw, 1981), over 20 years of research
experience and 8 years of Industrial R&D and International Marketing expe-
rience. Appointed as University Professor at Ecole Centrale de Lyon (1982), he
introduced scanning photoluminescence as a tool for fast, non-destructive two-

Page 8

R. Mazurczyk et al. / Sensors and Actuators B 128 (2008) 552–559
559
dimensional evaluation of compound semiconductor materials and for in-line
controlofentiretechnologicalprocessesduringdevicefabrication.Intheperiod
1995–1998, he conducted several research projects in the field of microsys-
tems and in the period 1998–2000 was visiting scientist at the University of
Tokyo in Japan, where he pioneered exploratory developments of GaN based
microsystems.BacktoFrancein2001,heinitiatednewinterdisciplinaryresearch
programs aimed at the development of Lab-on-a-Chip microfluidic systems for
medical diagnostics applications.
E.SouteyrandreceivedherPhDdegree(1979)fromParisVIUniversityandthe
graduation of “Docteur d’Etat” in Materials Sciences (1985). She joined CNRS
in1981andstartedherresearchactivitiesinelectrochemistry,photovoltaicsand
semiconductors. In 1991, she integrated the Ecole Centrale de Lyon and investi-
gated the field of gas sensors, biosensors and DNA chips. In 2000, she received
theAwardofInnovativeCompagnyCreation,followedtheentrepreneurialtrain-
ingprogramattheLyonSchoolofBusinessandfoundRosatechS.A.,acompany
devotedtothe“Biochipsfordiagnostics”.Currently,sheistheheadofthedepart-
ment of “Biotechnology for Health “at the Lyon Institute of Nanotechnology.
She is the co-author of more than 70 international publications and holds 8
international patents.
J.P. Cloarec obtained PhD degree in material engineering in 1997. He has been
an assistant professor in chemistry at Ecole Centrale de Lyon since 2000. His
fieldsofinterestcoverbiomolecularinteractionsatsolid/liquidinterface,surface
functionalisation, biosensors and microarrays.
Y. Chevolot received the PhD degree in material science from the Ecole Poly-
technique F´ ed´ erale de Lausanne at Lausanne (Switzerland) in 1999. Between
1999 and 2001, he work as assistant of Pr Mathieu at EPFL on bacterial adhe-
sion to PVC endotracheal tubes and was in charge of ToF-SIMS analysis. From
2001 to 2004, he work in the research department of Goemar SA, Saint Malo
(France). Since 2004, he joined the CNRS as a senior scientist. He focuses
on microfabrication, surface chemistry and characterisation, and biochips in
particular glycoarrays.