Ff

SD

h

�

�

�

�

�

a

ARAA

KTTSNAS

1

it

0h

Colloids and Surfaces A: Physicochem. Eng. Aspects 412 (2012) 38– 46

Contents lists available at SciVerse ScienceDirect

Colloids and Surfaces A: Physicochemical andEngineering Aspects

journa l h omepa g e: www.elsev ier .com/ locate /co lsur fa

abrication of meso and nanotextured silica surfaces for tunable densities ofunctionalized molecules

ubham Dastidar, Ravi Chahar, Vivekananda Bal, Satyendra Kumar, Siddhartha Panda ∗

epartment of Chemical Engineering, Indian Institute of Technology Kanpur, Kanpur, UP 208 016, India

i g h l i g h t s

We report a method to obtain tun-able meso and nanotextured silicasurfaces.The time of introduction had a strongeffect on the surface particle densi-ties.Surface densities were also governedby the process used for the particlesynthesis.The texturization phenomena werestudied and possible mechanisms areproposed.The textured surfaces were function-alized with antibodies; the effects oftexture were studied.

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 31 May 2012ccepted 11 July 2012vailable online 20 July 2012

eywords:unableexturization

a b s t r a c t

Textured surfaces have applications in different fields including sensing. We report a method to obtaintunable textured surfaces utilizing electrostatic coupling by introducing silanized silica wafer surfacesinto a sol of growing silica meso and nano particles at appropriate times specific to the process. Animportant finding here is that the time of introduction of the wafers into the sol had a strong effect on thesurface particle densities and there exists an optimum time window of introduction for a given processto obtain the maximum surface density. The surface densities were also governed by the process usedfor the particle synthesis, with processes for smaller sized particles yielding higher surface particle den-

ilicaanoparticlentibodyensor

sities. Thus by controlling the particle size and the time of introduction, the surface texture can be tunedfor specific applications. This may be a cheaper alternative in specific cases to expensive (though morecontrolled) techniques. The texturization phenomena were studied and possible mechanisms are pro-posed for the observed experimental results. These textured surfaces were subsequently functionalizedwith biomolecules (antibodies) and the effects of the texture on the surface density of the immobilizedantibodies were studied.

. Introduction

Textured surfaces are utilized for enhanced photon absorptionn photovoltaics [1,2], effective anti-reflective coatings for lightrapping [1], higher data storage [3], improved adhesion and

∗ Corresponding author. Tel.: +91 512 259 6146; fax: +91 512 259 0104.E-mail address: spanda@iitk.ac.in (S. Panda).

927-7757/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.colsurfa.2012.07.012

© 2012 Elsevier B.V. All rights reserved.

friction for tribology [4,5], obtaining superhydrophobic surfaceswith appropriate chemical treatment [6,7], and higher sensitivityin chemical sensing applications [8,9]. Texturing of surfaces bysilica micro and nano particles have been demonstrated by spin-coating [1,4], dip coating [10] and Langmuir–Blodgett techniques

[6,11] using commercially available silica powders by both thesubtractive routes and the additive routes. The subtractive routestypically involve the spin coated silica particles on the surfacewhich act as “natural lithography” sacrificial masks for subsequent

: Phys

reIpwtto

gsoart(mbeiadapltiipwtioAgsaohestw

2

2

2

pmt

ing distance of 10–11 mm. Sizes of the particles from the solutionas well as those on the surface, and the surface densities of the par-

TE

S. Dastidar et al. / Colloids and Surfaces A

eactive ion etch to form a textured surface [12]. However, thisxtra process step adds to the process complexity and the cost.n the additive routes, in processes such as spin coating, thesearticles are physically adsorbed on to the surfaces, which beingeakly bound can be removed by weak shear forces. Also, in both

hese routes which use commercially available silica particles, theunability of the texturing is limited by the size of the particlesbtained from the external source.

Strong electrostatic coupling of protein molecules [3] and nanoold particles [13,14] on to silanized silica surfaces to form texturedurfaces have been reported. Tunable surface densities [13,14] werebtained by creating a one-dimensional gradient of the surfacemine groups and by controlling the solution pH. In this work, weeport a simple and cost effective method to obtain design (andhus tunable) textures of silanized silica surfaces utilizing strongelectrostatic) coupling with silica meso and nano particles. Silica

eso and nano particles, synthesized using variations of the Sto-er technique (hydrolysis of TEOS followed by condensation), werelectrostatically coupled to silanized silica substrates, by introduc-ng them in the sol of the growing particles at appropriate timesnd keeping them immersed also for appropriate times, to obtainesign textured surfaces. To the best of our knowledge, we are notware of any prior reports on the utilization of nucleating/growingarticles (as opposed to grown particles having reached their equi-

ibrium or final sizes) for texturization on to silanized surfaces. Inhis method, tunability of the surface texture is achieved by vary-ng the size and the surface density of the particles. As mentioned,n methods which use commercially available silica powders, thearticle sizes are limited to those obtained from the external sourcehich limits the variability. An advantage of our method is that

he particles are generated in situ and a reasonable (within lim-ts) range of near monodisperse particles can be obtained therebybviating the dependence of external sources for different sizes.lso, there is no need for creating a gradient of the aminosilaneroups on the surface. The time evolutions of the particle sizes weretudied and utilized to obtain tunable surface textures. The mech-nisms of texturization were studied and the appropriate regimesf the introduction of the surfaces into the sols for obtaining theighest surface densities were identified and explanations of thexperimental results are proposed. These textured surfaces wereubsequently functionalized with antibodies [9] and the effects ofhe texture on the surface density of the immobilized antibodiesere studied.

. Experimental and characterization

.1. Formation of textured silica wafer surfaces

.1.1. Synthesis of monodisperse silica particles

Using variations of the Stober technique [15], near monodis-

erse silica meso and nano particles were formed in a two-stepethod. Different process conditions, shown in Table 1, were used

o obtain particles of different final or equilibrium sizes (475, 300,

able 1xperimental conditions used to synthesize the particles.

Sl. no. Step I

TEOS (ml) Ethanol (ml) DI water (ml) Sonicationtime (min)

1 1 10 – 5

2 1 – 10 5

3 0.5 – 10 5

4 1 10 1 5

icochem. Eng. Aspects 412 (2012) 38– 46 39

135, 80 nm). For example, to form particles of a final size of about475 nm diameter, in the first step, an ethanol (10 ml)–TEOS (1 ml)mixture was sonicated in an ultrasonic bath (1.5 l, 60 W, 33 kHz) for5 min. In the second step, an ethanol (10 ml)–ammonium hydroxide(10 ml) mixture was added to the mixture of step 1, and the entiremixture was sonicated for 50 min (the time used to ensure that theparticles have reached the final or equilibrium size). The additionof different concentrations of deionized (DI) water or replacementof ethanol with DI water in either of the step 1 was also used toobtain particles of different sizes. These processes are subsequentlyreferred to by their final sizes; e.g., the process which results in anaverage final size of 475 nm in the sol is referred to as the “475 nmprocess”, and these particles are referred to as the “475 nm” sizedparticles.

2.1.2. Texturization of silicon dioxide wafers with silica particlesSilicon dioxide was grown thermally on (1 1 1) p-doped sili-

con wafers (Wafer World, USA) using furnace oxidation at 1000 ◦Cfor 1 h. The wafer pieces were cleaned sequentially in boilingde-ionized water and acetone for 5 min each and then soni-cated in an ultrasonic bath for 5 min and 2 min, respectively.Further, these wafers were cleaned using the RCA 1 and RCA2 processes. The wafers were hydrolyzed in Piranha solution(3:1 volume ratio H2SO4:H2O2) (to be used with caution) for 2 hat 130 ◦C and then kept in a vacuum oven for 1.5 h at 120 ◦Cto remove the absorbed water from the surface. The waferswere then silanized for 24 h at ambient temperature in a 3%solution of 3-aminopropyltriethoxysilane (APTES), obtained fromSigma–Aldrich, in toluene. After the completion of the silaniza-tion procedure, the wafers were rinsed with toluene (at 50 ◦C), andthen ultra sonicated for 2 min in ethanol to remove any adsorbedaminosilane from the surface. The silanized silicon dioxide waferpieces were introduced into the sol containing the growing sil-ica particles (see Section 2.1.1) which have not yet reached thefinal size, at different time intervals from the start of step 2 inthe particle synthesis process, and were kept in the sol for differ-ent time periods. The time periods used are mentioned in Section3.2.

2.2. Characterization of particle sizes and surface densities oftextured silica surfaces

The morphologies of the synthesized silica particles and themorphologies of the textured surfaces were characterized by scan-ning electron microscopy (SEM) using JEOL Quanta 200 and FE-SEM.The gold coated samples were characterized at 20 kV, with a work-

ticles on the textured surface were measured from the SEM imagesusing ImageJ software. For analysis of the particle sizes in the solu-tion, drops of the sonicated solution were pipetted on to a doublesided carbon tape on brass stubs prior to gold coating.

Step II Average particlesize (nm)

Ethanol (ml) Ammoniumhydroxide (ml)

Sonication time(min)

10 10 50 475 (±4)10 10 50 300 (±4)10 10 50 135 (±5)30 2.5 80 80 (±3)

4 : Phys

2

D

2d

iwtswFfflu

3

t(dtfd

3

aceStfit(s2tsb

tiwssdtttd1atccideas

0 S. Dastidar et al. / Colloids and Surfaces A

.3. Zeta potential measurements

The zeta potential of the silica particles were measured by aelsa Nano C unit (Beckman Coulter).

.4. Immobilization of antibodies on textured silica surfaces andensity measurement

The non-textured wafers were cleaned sequentially in de-onized water, trichloroethylene, acetone and ethanol. Then they

ere hydrolyzed with Piranha solution (to be handled with cau-ion), and subsequently were silanized with 3% APTES in tolueneolvent. The non-textured and textured wafers were cross-linkedith 2.5% glutaraldehyde in ethanol, and then immobilized with

ITC tagged antibodies (anti-BSA IgG molecules). The relative sur-ace densities of the immobilized antibodies were measured usinguorescence spectroscopy. Details of the materials and processessed are provided elsewhere [9].

. Results and discussion

Monodisperse meso and nano sized silica particles were syn-hesized and the time evolutions of the particle sizes were studieddiscussed in Section 3.1). Silanized silica substrates were intro-uced in the sols containing the growing particles to obtainextured surfaces (discussed in Section 3.2). These surfaces wereunctionalized and the effects of texture on the resulting antibodyensities were studied (discussed in Section 3.3).

.1. Synthesis and time evolution of nano and meso sized particles

Table 1 shows the conditions used to synthesize the particlesnd the average final sizes obtained. It is to be noted that theseonditions were obtained after optimization of the process param-ters. The time evolution of the particle sizes were studied from theEMs of the particles which were taken after different times fromhe start of the second step, and the times required to obtain thenal sizes were determined. For the 475 nm process, the SEMs ofhe particles formed after (a) 1 min, (b) 2 min, (c) 3 min, (d) 5 min,e) 10 min and (f) 50 min are shown in Fig. 1. As seen from the mea-ured average particle sizes obtained from the images, which were70 nm, 360 nm, 410 nm, 440 nm, 470 nm and 475 nm, respectively,here was a rapid initial increase in the particle size followed by alower rate of increase, and by about 50 min the particle growthecame insignificant.

The classical nucleation model [16] proposed three regimes inhe generation of mono-dispersed particles: (i) induction – dur-ng which the reaction slowly generates growth units of the solid

here the concentration in solution builds up until a critical super-aturation is reached; (ii) nucleation – which lasts for a relativehort time duration where a certain number of nuclei are intro-uced into the system and the supersaturation is relieved to belowhe critical value; (iii) growth – where the molecular nutrients addo the particles which grow. Several studies have shown that solu-ion based synthesis of monodisperse nano and micro particleso not follow the classical nucleation process [16 and references4–35 therein]; rather another nonclassical or aggregative nucle-tion process occurs where the classical nucleation is followed byhe assembly of critical sized aggregates. These aggregative pro-esses have been analyzed by adapting the mechanisms of thelassical nucleation-growth processes [17]. From the SEMs shownn Fig. 1, it was seen that within the first minute, it was difficult to

iscern differences between the growth rates of the particles. How-ver, by the second minute, some grew more rapidly than othersnd a bimodal distribution, as observed in other systems [17], waseen. Subsequently, by 10 min, the distribution minimized and an

icochem. Eng. Aspects 412 (2012) 38– 46

average size of about 470 nm was reached. By 50 min, an averagesize of about 475 nm was reached which did not further changewith time. Throughout the evolution, it was observed that the par-ticles were not agglomerated. In cases of intermediate times wherebimodal distributions were observed, the average of the larger sizedparticles was used. We believe that aggregative growth along withthe Ostwald ripening mechanism govern the growth of the particleshere.

Similarly, the time evolution for the 300 nm process and the135 nm process (SEMs not shown here) also showed that by about10 min, the average particle sizes had almost reached the finalvalue. However, for the 80 nm particles the final size was reachedin about 60 min. The time evolutions of the particle sizes for thedifferent processes yielding different final sizes are shown in Fig. 2.The SEMs of the particles of processes resulting in average finalparticle sizes of 300 nm, 135 nm and 80 nm are shown in Fig. 3a, band c, respectively. As with the 475 nm particles, here too sphericalmonodisperse particles were observed.

3.2. Surface texturization

Two terms are defined here: (i) “time of introduction” (TI)and (ii) “removal time” (RT). TI is the time of introduction of thesilanized substrates in the growing sol after the initiation of step2 of the Stober reaction (see Section 2.1). RT is the time from thestart of the step 2 when the substrate is removed from the sol.These time periods are schematically illustrated in Fig. 4. For eachof these processes, different values of TI were used. The choiceof the values depended on the time, from the start of the secondstep of the reaction, of the onset of visual observation of turbid-ity in the growing sol. The appearance of turbidity was abrupt forthe 475 nm, 300 nm and 135 nm processes with onset times beingabout 23–25 s, 30–33 s and 53–56 s, respectively. For the 80 nm pro-cess, the appearance of turbidity was a slow process with a faintappearance around 17 min and progressing to a reasonable turbid-ity at about 25 min. The values of TI used ranged from lower thanto higher than this onset time. Thus, for the 475 nm, 300 nm andthe 135 nm processes, TI values of 15 s, 30 s, 40 s, 50 s, 60 s and 75 swere used; and for the 80 nm case, values of 12 min, 17 min, 25 min,35 min, 45 min and 60 min were used. The minimum time neededto nearly reach the final size as seen in the time evolution stud-ies in the sol was used as the RT for each process. For the 475 nm,300 nm, and 135 nm processes, RT values of 10 min were used (asby this time the average values of the particle sizes were close to, i.e.within the experimental error of, the final values) and for the 80 nmprocess, a value of 80 min (even though it reached the equilibriumsize in about 60 min) was used.

Using these values of TI and RT, the surface particle densitiesobtained using the 475 nm, 300 nm, 135 nm and 80 nm processesare shown in Fig. 5a, b, c and d, respectively.

An important observation is that the time of introductionhad a strong effect on the particle surface density. For all cases,there exists an optimum time window of introduction of the sub-strates so as to obtain the maximum surface densities. For the475 nm case, average particle densities of 0.48/�m2, 3.06/�m2,2.43/�m2, 1.72/�m2, 1.24/�m2, 0.90/�m2 were obtained for TIvalues of 15 s, 30 s, 40 s, 50 s, 60 s and 75 s, respectively. For the300 nm case, average particle densities of 0.73/�m2, 3.97/�m2,2.79/�m2, 2.07/�m2, 2.05/�m2, 1.19/�m2 were obtained for TIvalues of 15 s, 30 s, 40 s, 50 s, 60 s and 75 s, respectively. For the135 nm case, average particle densities of 1.46/�m2, 2.23/�m2,2.94/�m2, 4.83/�m2, 2.74/�m2, 2.50/�m2 were obtained for TI

values of 15 s, 30 s, 40 s, 50 s, 60 s and 75 s, respectively. For the80 nm case, average particle densities of 1.12/�m2, 87.70/�m2,235.21/�m2, 160.1/�m2, 119.2/�m2 and 25.5/�m2 were obtainedfor TI values of 12 min, 17 min, 25 min, 35 min, 45 min and 60 min,

S. Dastidar et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 412 (2012) 38– 46 41

F min aT , resp

rmrtcr

ig. 1. SEMs showing the particles at (a) 1 min, (b) 2 min, (c) 3 min, (d) 5 min, (e) 10he average particle sizes are 270 nm, 360 nm, 410 nm, 440 nm, 470 nm and 475 nm

espectively. As can be seen from the above results, the maxi-um surface densities in all four cases were obtained when the

espective TI value was close to the time of the visual onset ofurbidity. Also, smaller the particle size, higher was the parti-le surface density. In addition, with smaller particle sizes, theange of surface densities obtained was higher. Thus by utilizing

nd (f) 50 min after the start of step 2 of the Stober method for the 475 nm process.ectively.

the appropriate process for obtaining a desired final particlesize and the time of introduction, the surface texture can be

tuned for specific applications. This may be a cheaper alternativein specific cases to expensive (though more controlled) tech-niques such as AFM manipulation of surface micro/nano particles[18].

42 S. Dastidar et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 412 (2012) 38– 46

FF3

s(s

Fm

ig. 2. Evolution of the average size of the particles, measured from SEM data fromig. 1, as a function of time after the start of the Stober method for the 475 nm,00 nm, 135 nm and 80 nm processes.

SEMs of the textured surfaces with the highest surface den-ities (conditions as shown in Fig. 5) for the (a) 475 nm process,b) 300 nm process, (c) 135 nm process and (d) 80 nm process arehown in Fig. 6. The bond strengths of these surface immobilized

ig. 3. SEMs of particles of different average final sizes (a) 300 nm (b) 135 nm (c) 80 nmentioned in Table 1.

Fig. 4. Schematic depicting the terms used, “time of introduction” (TI) and “removaltime” (RT).

particles were tested by sonicating the textured wafer pieces inDI water for 1 h. These bond strengths on the silanized surfaceswere sufficient to withstand sonication as seen from SEMs (notshown here) before and after sonication whereas those from thenon-silanized surfaces, which can be considered as “control” sur-

faces, did not withstand the sonication. Also, it is been observedthat the average size of the particles attached to the surfaces weresmaller than those in the solution. The sizes obtained for the 475 nmprocess, 300 nm process, 135 nm process and 80 nm process were

, obtained at the end of step 2 of the synthesis using different process parameters

S. Dastidar et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 412 (2012) 38– 46 43

F 80 nmd

4Tttd

sftsdtootobwiwassufcs

ttIawc

ig. 5. Surface particle densities for the (a) 475 nm (b) 300 nm (c) 135 nm and (d)eviation about the mean.

22 ± 24 nm, 280 ± 10 nm, 120 ± 5 nm and 46 ± 3 nm, respectively.he smaller sizes (compared to the particles growing in the solu-ion) could be due to the geometrical hindrances to the transport ofhe monomers for growth as well as energy considerations. Also, theispersities were higher for the particles growing on the surfaces.

Reactive groups on the silica wafer surface, arising from theilanization process, were needed for efficient capture of particlesrom the sol to enable texturization. For a typical condition, usinghe 475 nm process, using a TI value of 30 s, an average particle den-ity of 3.1/�m2 was obtained on a silanized surface, while a particleensity of 0.08/�m2 was obtained on a non-silanized surface. Elec-rostatic interaction between the positively charged amine groupsn the wafer surfaces and the negatively charged hydroxyl groupsf the growing silica particles in the sol are believed to result inhe attachment of the particles on the surface [3,14]. The densityf the particles attached to the surface depends on the quantitativealance of surface charge on both the particles and the silanizedafers [3]. The zeta potentials of the particles formed after hav-

ng reached their final sizes in the sol were measured. The valuesere −12.2 mV (80 nm), −21.5 mV (135 nm), −22.3 mV (300 nm)

nd −60.3 mV (475 nm), respectively, at a pH about 11.5. Largerized particles with more surface charges are more effective thanmaller particles in reducing the net charge on the wafer surfacepon immobilization. Reduction of the net charge on the wafer sur-ace reduces the extent of immobilization of the particles [3]. Thisould explain our observation of higher particle densities of smallerized particles on the surfaces.

We hypothesize that the regime of particle growth occurring inhe solution at the time of introduction of the reactive surface intohe solution has a strong effect on the nature of the texturization.

f the onset of the visually observed turbidity can be considereds an indicator of the progress of nucleation of the particles, thene believe that the introduction during or around regime of parti-

le nucleation leads to a maximum reactivity between the surface

processes obtained at different TI values. The error bars represent one standard

groups ( NH2 on wafer and OH on particle). Use of experimentaltechniques such as small angle scattering [19], thermocalorime-try [20], acoustic spectroscopy, laser diffraction [21] coupled withdeterministic modeling and stochastic simulation routes [21,22]and growth models [23,24] could provide considerable insight intothe mechanisms of particle texturization and temporal evolution inour case. Such analysis is however outside the scope of this presentwork and our current assumption of the visually observed occur-rence of the onset of turbidity as an indicator of the progress ofnucleation stands to be corrected in the future. As mentioned, thereare three regimes of particle growth: induction, nucleation (includ-ing aggregative nucleation) and growth. In our experiments, we didnot measure the time dependent nucleation rates as has been doneby other groups [17,25,26]; instead, we tried to identify the time ofthe onset of nucleation by considering the initiation of the observedturbidity in the solution. As mentioned, these times were about23–25 s, 30–33 s, 53–56 s, and 17–25 min for the 475 nm, 300 nm,135 nm, and the 80 nm processes, respectively. Our results indicatethat the onset of nucleation is earlier for particles with a higher finalparticle size, and this pattern is consistent with the results observedby Shields et al. [17] but is opposite to that observed by Wette et al.[25] for different systems. In our method, the time process win-dow for introduction of the wafers can be varied using differentprocesses; with processes using smaller sized particles providing awider window and also providing a wider range of surface particledensities.

The positively charged NH2 groups (i.e. NH3+) [3,13,14]

on the wafer surface get electrostatically bonded to the nega-tively charged OH (i.e. O−) groups, present both in the solution(referred to as the solution OH) and on the particle surfaces

(referred to as the particle OH). Particle capture by the wafersurface lowers the net wafer surface charge, thereby lowering thecapacity of subsequent capture of the particles on the wafer sur-face. Hence particles with slightly negative zeta potential will have

44 S. Dastidar et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 412 (2012) 38– 46

F th theh

abg

f

wTmaZiazafsEcwNwttta

ig. 6. SEMs of the textured surfaces with the highest surface densities obtained wiighest surface densities for each process are provided in Fig. 5.

higher tendency than those with a higher negative potential toind on to the surface. The fraction of the protonated surface NH2roups (fNH3

+ ) as a function of the solution pH is given by [14]

NH3+ = [1 + 10pH−pK1/2 ]−1 (1)

here pK1/2 is the pH at which half the surface groups are ionized.his fraction is a “reverse S shaped” function of the pH and has ainimal value (i.e. the surface is almost neutral) by a pH of about 9.5

nd higher. Utilizing the technique of chemical force microscopy,hang et al. [27] reported the pK1/2 values of APTES on oxidized sil-con to be about 7.6, a value used by Bhat and Genzer [14] andccepted in this present work as well. However, Bhat and Gen-er [14] obtained a reasonable surface gold nanoparticle density ofbout 70/�m2 even at a solution pH of 11, with the inflexion pointor the particle binding occurring at pH values higher than thateen for the fraction of the surface ionized groups obtained fromq. (1). This was attributed to the higher size of the gold nanoparti-les compared to the size of the NH3

+ groups. Similarly, here tooe obtain particle attachment even at pH values of 11.5 in the sol.ow, let us consider the three regimes of the introduction of theafer in the sol. (i) When the wafers are introduced in the solu-

ion during the induction stage, as there are no particles to bind ono the surface, the solution OH groups neutralize the charge onhe surface. So, by the initiation of the particle nucleation regime,

sufficient electrostatic field may not exist to induce significant

(a) 475 nm, (b) 300 nm, (c) 135 nm and (d) 80 nm processes. The conditions for the

particle immobilization. (ii) When the wafers are introduced to thesolution during the nucleation stage, the smaller sized nucleatedentities, with an expected lower value of negative potential, aremore effective in immobilizing on the positively charged wafer sur-face. (iii) When the wafers are introduced during the growth stage,where the size of the particles on the solution has grown and withit the magnitude of the negative potential, the efficiency of particleimmobilization is reduced. This proposed mechanism is illustratedin a schematic shown in Fig. 7.

Earlier we had reported aminosilane densities of 2–5/nm2 onsilicon (nontextured and nanotextured) surfaces [28] and in thiswork we obtained densities of about 2.3/nm2 on nontextured silicasurfaces, thereby suggesting that the particle densities obtainedwere not limited by the surface aminosilane densities. There is acompetition between the particle OH groups (which have a higherconcentration) and the solution OH groups to link with the silanesgroups on the wafer surface. Now, while the surface silane densi-ties are about 2.3/nm2 or 2.3 × 106/�m2, we note that the particledensities are about 0.5–250/�m2 for the cases studied here. Thusa small fraction of silanes not neutralized by the solution OH iscapable of particle coupling. Our current hypothesis is supportedby the results of Bhat and Genzer [14] who reported surface par-

ticle densities of about 70/�m2, even at pH values as high as 11,where a high density of OH− existed, and there was a competi-tion between the reactive (negatively charged) groups on the goldparticles and the OH− groups in solution to link with the surface

S. Dastidar et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 412 (2012) 38– 46 45

Fig. 7. A schematic depicting the effect of the time of introduction on the particlestt

abjiios

ohepwafitncadosammlwCcisa

3

tnbiipssi

o

Fig. 8. Fluorescence spectroscopy data showing the fluorescence band area rep-

urface densities with respect to the different regimes of particle evolution. As men-ioned, our current assumption of the visually observed occurrence of the onset ofurbidity as an indicator of the progress of nucleation stands to be corrected.

mines. We would like to mention here that not much work waseen done on the surface attachment of growing particles (the sub-

ect of our work) in solution as opposed to already grown particlesn solution. Further work on this subject (i.e. attachment of grow-ng particles) will elucidate the mechanisms; and as mentioned,ur current hypothesis (based on the current experimental results)tands to be corrected in the future.

As seen from the SEMs in Fig. 6, incomplete surface coveragesf the particles were observed here and such incomplete coveragesave also been observed in other works [3,13] where particles werelectrostatically coupled to the silanized surfaces (utilizing grownarticles as opposed to nucleating/growing particles used in thisork). The surface particle densities depend on the net charge bal-

nce on the surface, and in our case the process conditions could beurther optimized so as to obtain higher surface coverage and thiss a subject of a future work. It needs to be mentioned that whilehere are processes such as spin coating of particles which result inear complete coverage [1], these weakly bound particles are sus-eptible to detachment by weak shear forces, thereby limiting theirpplications. Secondly, our method provides for obtaining tunableensities, which may not possible with the other mentioned meth-ds. While there is a large body of work on the dynamics of colloidalilica particles such as understanding the formation, aggregationnd growth of colloidal silica particles [21–24], mechanisms of for-ation of microporous and mesoporous silicates [29], there is notuch work reported on understanding the mechanisms of the col-

oidal particles binding to functionalized charged surfaces. Furtherork needs to be done to elucidate the phenomena of the binding.ritical understanding of the factors governing the transport of theharged colloidal silica particles to the functionalized charged sil-ca surfaces and factors affecting the binding of the particles to theurfaces are required to further understand the phenomena and is

subject of future work.

.3. Tunable densities of surface immobilized biomolecules

As mentioned, there are several applications of surfaces tex-ured with silica particles, including biosensing. While there is aeed in general to have higher surface densities of the immobilizediomolecules to enhance the capture of target analytes in biosens-

ng applications, these higher densities need not necessarily resultn a higher number of recognition reactions as poorly controlledacking could disturb the biomolecular interactions due to electro-tatic effects and steric hindrances, thus necessitating application

pecific spacing and therefore controlled densities of the surfacemmobilized biomolecules [9,30].

Textured surfaces using the highest surface particle densitiesbtained for the 475 nm, 300 nm, 135 nm and 80 nm processes,

resenting the relative surface antibody densities on the non-textured surface andsurfaces textured with particles of different sizes. The error bars represent onestandard deviation about the mean.

respectively, along with non-textured surfaces (for baseline pur-poses) were immobilized with FITC-tagged antibodies. The fluores-cence signals, which can be taken to represent the relative densitiesof these antibodies, are shown in Fig. 8. As is seen from the figure,the surfaces textured with smaller particle sized process resultsin a higher density of immobilized biomolecules with a sevenfoldincrease in the signal from the non-textured case to the texturedcase with 80 nm sized particles. With simple calculations showingthat the increase in the fluorescence signal is more than proportion-ate to the increase in surface area, the higher biomolecule surfacedensities can be attributed, among other reasons, to lowering ofsteering hindrances [30]. Thus application specific spacing betweenbiomolecules could be realized by utilization of different textures.Further optimization would lead to a wider range of densities andlarger sensitivities and hence improved biosensing performance.

4. Conclusions

Monodisperse silica particles of different final sizes (80–475 nm)were synthesized using different, two-step modified Stober pro-cesses. The time evolution of the particle sizes were studied andthe time required to obtain the final sizes were determined. Webelieve that aggregative growth along with the Ostwald ripeningmechanism governed the growth of the particles here. The onsetof nucleation was earlier in the processes for the larger sized par-ticles. Silanized silica wafers were introduced at appropriate timesinto the sols of the growing silica particles to obtain tunable tex-tured surfaces. Electrostatic forces between the negatively charged

OH groups on the particle surfaces and the positively chargedNH2 groups on the wafer surface are believed to result in electro-

static coupling and the particles withstood sonication. The surfaceparticle densities depended on the process used and the time ofintroduction of the substrate into the sol. Processes for obtain-ing lower final sized particles resulted in higher surface densitiesand this was attributed to the lower negative zeta potential of thelower sized particles as the extent of texturization was governedby the charge balance. The regime of particle growth occurring inthe solution at the time of introduction of the reactive surface intothe sol had a strong effect on the nature of the texturization. Webelieve that the introduction during or around the nucleation leads

to a maximum reactivity between the surface groups. Introductionduring the induction regime results in a low particle immobiliza-tion as the wafer surface gets passivated by the OH groups in thesolution. And introduction during the growth regime also results

4 : Phys

ilasitSiitbtdp

A

feora

A

SCao

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

6 S. Dastidar et al. / Colloids and Surfaces A

n low densities as particles which have grown to larger sizes areess efficient in immobilization. These surfaces were functionalizednd the effects of texture on the resulting antibody densities weretudied. These textured surfaces have several applications includ-ng biosensing. These surfaces were immobilized with fluorescentagged antibodies and these antibody densities were measured.urfaces textured with smaller particle sized processes resultedn higher densities of immobilized biomolecules, with a sevenfoldncrease in the signal from the non-textured case to the tex-ured case with 80 nm sized particles. Application specific spacingetween biomolecules could be realized by utilization of differentextures and further optimization would lead to a wider range ofensities and larger sensitivities and hence improved biosensingerformance.

uthors contribution

SD conducted the experiments on the particle synthesis and sur-ace texturization and analyzed the results, RC and VB conductedxperiments on particle synthesis, SK conducted the experimentsn bio-functionalization, SP guided the research work, analyzed theesults and wrote the paper with inputs from other co-authors. Alluthors have approved the final article.

cknowledgments

This work was supported by funding from the Department ofcience and Technology, Govt. of India. SEM facilities at the Samtelentre for Display Technologies, and the DST Unit for Nanosciencesre acknowledged. The help of Dr. R.S. Anand with the oxidation andf Mr. N. Kumar with some experiments are acknowledged.

eferences

[1] M. Tao, W. Zhou, H. Yang, L. Chen, Surface texturing by solution deposition foromnidirectional antireflection, Appl. Phys. Lett. 91 (2007) 081118/3.

[2] Z. Fan, H. Razavi, J. Do, A. Moriwaki, O. Ergen, Y. Chueh, P.W. Leu, J.C. Ho,T. Takahashi, L.A. Reichertz, S. Neale, K. Yu, M. Wu, J.W. Ager, A. Javey,Three-dimensional nanopillar-array photovoltaics on low-cost and flexiblesubstrates, Nat. Mater. 8 (2009) 648–653.

[3] K. Yamada, S. Yoshii, S. Kumagai, I. Fujiwara, K. Nishio, M. Okuda, N. Mat-sukawa, I. Yamashita, High-density and highly surface selective adsorption ofprotein–nanoparticle complexes by controlling electrostatic interaction, Jpn. J.Appl. Phys. 45 (2006) 4259–4264.

[4] M. Zou, L. Cai, H. Wang, Adhesion and friction studies of a nano-textured surfaceproduced by spin coating of colloidal silica nanoparticle solution, Tribol. Lett.21 (2006) 25–30.

[5] A.K. Geim, S.V. Dubonos, I.V. Grigorieva, K.S. Novoselov, A.A. Zhukov, S. Yu

Shapoval, Microfabricated adhesive mimicking gecko foot-hair, Nat. Mater. 2(2003) 461–463, http://dx.doi.org/10.1038/nmat917.

[6] P.-S. Tsai, Y.-M. Yang, Y.-L. Lee, Fabrication of hydrophobic surfaces by couplingof Langmuir–Blodgett deposition and a self-assembled monolayer, Langmuir22 (2006) 5660–5665.

[

[

icochem. Eng. Aspects 412 (2012) 38– 46

[7] S. Jeong, L. Hu, H.R. Lee, E. Garnett, J.W. Choi, Y. Cui, Fast and scalable printing oflarge area monolayer nanoparticles for nanotexturing applications, Nano Lett.10 (2010) 2989–2994.

[8] S. Santra, P. Zhang, K. Wang, R. Tapec, W. Tan, Conjugation of biomoleculeswith luminophore-doped silica nanoparticles for photostable biomarkers, Anal.Chem. 73 (2001) 4988–4993.

[9] S. Kumar, R. Ch, D. Rath, S. Panda, Densities and orientations of antibodies onnano-textured silicon surfaces, Mater. Sci. Eng. C 31 (2011) 370–376.

10] E. Cunningham, C.J. Campbell, A novel self-assembled nanoparticulate film forcovalent attachment of antibodies to plastic, Langmuir 19 (2003) 4509–4511.

11] M. Bardosova, M.E. Pemble, I.M. Povey, R.H. Tredgold, The Langmuir–Blodgettapproach to making colloidal photonic crystals from silica spheres, Adv. Mater.22 (2010) 3104.

12] T. Nakanishi, T. Hiraoka, A. Fujimoto, S. Saito, K. Asakawa, Nano-patterningusing an embedded particle monolayer as an etch mask, Microelectron. Eng.83 (2006) 1503–1508.

13] R.R. Bhat, D.A. Fischer, J. Genzer, Fabricating planar nanoparticle assemblieswith number density gradients, Langmuir 18 (2002) 5640–5643.

14] R.R. Bhat, J. Genzer, Tuning the number density of nanoparticles by multivarianttailoring of attachment points on flat substrates, Nanotechnology 18 (2007)025301 (6 pp.).

15] W. Stober, A. Fink, E. Bohn, Controlled growth of monodisperse silica spheresin the micron size range, J. Colloid Interface Sci. 26 (1968) 62–69.

16] V.K. LaMer, R.H. Dinegar, Theory production and mechanism of formation ofmonodispersed hydrosols, J. Am. Chem. Soc. 72 (1950) 4847–4854.

17] S.P. Shields, V.N. Richards, W.E. Buhro, Nucleation control of size and disper-sity in aggregative nanoparticle growth. A study of the coarsening kinetics ofthiolate-capped gold nanocrystals, Chem. Mater. 22 (2010) 3212–3225.

18] S. Kim, F. Shafiei, D. Ratchford, X. Li, Controlled AFM manipulation of smallnanoparticles and assembly of hybrid nanostructures, Nanotechnology 22(2011) 115301 (6 pp.).

19] J.D. Rimer, O. Trofymluk, R.F. Lobo, A. Navrotsky, D.G. Vlachos, Thermodynamicsof silica nanoparticle self-assembly in basic solutions of monovalent cations, J.Phys. Chem. C 112 (2008) 14754–14761.

20] J.D. Rimer, D.G. Vlachos, R.F. Lobo, Evolution of self-assembled silica-tetrapropylammonium nanoparticles at elevated temperatures, J. Phys. Chem.B 109 (2005) 12762–12771.

21] M. Tourbin, C. Frances, Experimental characterization and population balancemodelling of the dense silica suspensions aggregation process, Chem. Eng. Sci.63 (2008) 5239–5251.

22] M. Ethayaraja, K. Dutta, R. Bandyopadhyaya, Mechanism of nanoparticle for-mation in self-assembled colloidal templates: population balance model andMonte Carlo simulation, J. Phys. Chem. B 110 (2006) 16471–16481.

23] G.H. Bogush, C.F. Zukoski IV, Uniform silica particle precipitation: an aggrega-tive growth model, J. Colloid Interface Sci. 142 (1991) 19–34.

24] T. Matsoukas, E. Gulari, Monomer-addition growth with a slow initiation step:a growth model for silica particles from alkoxides, J. Colloid Interface Sci. 132(1989) 13–21.

25] P. Wette, H.J. Schöpe, T. Palberg, Microscopic investigations of homoge-neous nucleation in charged sphere suspensions, J. Chem. Phys. 123 (2005)174902/11.

26] T. Gong, J. Shen, Z. Hu, M. Marquez, Z. Cheng, Nucleation rate measurementof colloidal crystallization using microfluidic emulsion droplets, Langmuir 23(2007) 2919–2923.

27] H. Zhang, H.-X. He, J. Wang, T. Mu, Z.-F. Liu, Force titration of amino group-terminated self-assembled monolayers using chemical force microscopy, Appl.Phys. A 66 (1998) S269.

28] N. Rathor, S. Panda, Aminosilane densities in nanotextured silicon, Mater. Sci.

Eng. C 29 (2340) (2009).

29] J.D. Rimer, J.M. Fedeyko, D.G. Vlachos, R.F. Lobo, Silica self-assembly synthesisof microporous and mesoporous silicates, Chem. Eur. J. 12 (2006) 2926–2934.

30] B.J. Hong, V. Sunkara, J.W. Park, DNA microarrays on nanoscale-controlled sur-face, Nucl. Acids Res. 33 (2005) (8 pp.).