«Electronic Supplementary Material (ESI) for Nanoscale. This journal is © The Royal Society of Chemistry 2016 Supplementary Information Tunable ...»

Electronic Supplementary Material (ESI) for Nanoscale.

This journal is © The Royal Society of Chemistry 2016

Supplementary Information

Tunable Top-Down Fabrication and Functional Surface Coating of

Single-Crystal Titanium Dioxide Nanostructures and Nanoparticles

Seungkyu Ha, Richard Janissen, Yera Ye. Ussembayev, Maarten M. van Oene, Belen Solano

and Nynke H. Dekker*

Department of Bionanoscience, Kavli Institute of Nanoscience, Faculty of Applied Sciences,

Delft University of Technology, Lorentzweg 1, 2628 CJ Delft, The Netherlands.

*E-mail: n.h.dekker@tudelft.nl

This document includes:

Supplementary Methods Orientation of the optic axis in single-crystal rutile TiO2 nanocylinder for OTW experiment.

Considerations for optimal fabrication of the Cr etch mask for single-crystal TiO2 etching.

Surface functionalization procedure of single-crystal TiO2.

Evaluation of single-crystal TiO2 surface functionalization efficiency via fluorescence microscopy.

Preparation of DNA construct for OTW experiments.

Preparation of flow cell for OTW experiments.

Bioconjugation of DNA to single-crystal TiO2 nanocylinders for OTW experiments.

OTW instrumentation and DNA measurements with single-crystal TiO2 nanocylinders.

Supplementary Tables Table S1. Dry etching conditions for single-crystal TiO2 during optimization of the CHF3-RIE process.

Table S2. Dry etching conditions for single-crystal TiO2 nanofabrication.

Table S3. Dimensional analysis of high and low aspect-ratio TiO2 nanocylinders.

Supplementary Figures Fig. S1. Diverse applications of TiO2 at the nanoscale.

Fig. S2. The etch rates and etch selectivities of different mask materials.

Fig. S3. Control of sidewall profiles and etch characteristics in single-crystal TiO2 nanocylinders by variation of O2 flow rate in the CHF3-RIE process.

Fig. S4. Dimensions of fabricated single-crystal TiO2 nanocylinders.

Fig. S5. Quantitative comparison of surface functionalization efficiencies on single-crystal TiO2 for different linker molecules.

Fig. S6. DLS measurements of single-crystal TiO2 nanocylinder aggregation in relation to surface coatings and buffer conditions.

Fig. S7. Optical trap calibration of single-crystal TiO2 nanocylinders.

Supplementary References S1 Supplementary Methods Orientation of the optic axis in single-crystal rutile TiO2 nanocylinder for OTW experiment To control polarization-based rotation of optically trapped nanoparticles in an OTW, birefringent positive uniaxial single-crystals are desirable substrate materials. Single-crystal rutile TiO2 is such a material, and it has an exceptionally high birefringence that is advantageous for effective torque transfer in an OTW. In particular, cylindrically shaped nanoparticles align their long axis with the direction of laser beam propagation, fixing two of the three rotational degrees of freedom (DOF). The remaining rotational DOF is controllable via the polarization of the laser beam provided that the optic axis is perpendicular to the nanocylinder’s long axis (Fig.

5a). To appropriately control the orientation of this optic axis within the nanocylinders, it is necessary to etch into (100)-cut single-crystal rutile TiO2 substrates. A similar approach has been employed for the case of X-cut single-crystal quartz SiO2 substrates.1–5 Considerations for optimal fabrication of the Cr etch mask for single-crystal TiO2 etching We consider the optimal fabrication of the Cr etch mask for the desired size of single-crystal TiO2 nanostructures. The deposited Cr layer should be sufficiently thick for the mask to remain functional until the end of etching process, taking into account the fact that the mask top surface will not be perfectly flat. Also, the overall thickness of the Cr mask is limited by that of the used PMMA layer. The thickness of the PMMA should be 2−3 times larger than that of the Cr mask to facilitate complete lift-off, but it has an upper limit determined by its concentration. In practice, Cr layers thicker than ~150 nm tend to cause more severe deformation in patterned PMMA layers, resulting in higher nonuniformity. This deformation is presumably due to the built-up stress in Cr layers during physical vapor deposition.6 A further consideration in mask fabrication is that mask shapes tend to be more cone-like when thicker Cr layers and/or smaller patterned apertures are used7 (Fig. 1, step 4, inset illustration). Such masks are not suitable for anisotropic etching for vertical sidewall due to more rapid erosion of their thinner edges.

Surface functionalization procedure of single-crystal TiO2 For the surface functionalization of single-crystal TiO2 substrates, we have compared four different surface linker molecules (Fig. S5†): ETA (Sigma-Aldrich, The Netherlands), GPDMES (Sigma-Aldrich, The Netherlands), APDMES (Sigma-Aldrich, The Netherlands), and BADMSCP (abcr GmbH, Germany). For the binding of ETA linker, we dissolve ETA in anhydrous dimethyl sulfoxide (DMSO) (Sigma-Aldrich, The Netherlands) to a final concentration of 5 M. We use this ETA/DMSO solution to incubate the substrates for 12 h at room temperature, followed by washing the substrates with DI water. For the binding of epoxysilane linker (GPDMES), we incubate the substrates for 15 min at 75 °C using non-diluted GPDMES solution, followed by chloroform washing. For the binding of the APDMES and BADMSCP linkers, we incubate the substrates in ethanol containing either 5% (v/v) of APDMES or BADMSCP. We carry out the silanization reaction for 12 h at 70 °C and then wash with chloroform (or ethanol). Additionally, we perform PEGylation of ETA-coated surfaces with heterobifunctional NHS-PEG-COOH (MW 5,000, LaysanBio, USA).8 We incubate the surfaces with 2 mM PEG dissolved in 100 mM 2-(N-morpholino)ethanesulfonic acid (MES) buffer (pH 4.7, Sigma-Aldrich, The Netherlands) for 1 h at room S2 temperature. Afterwards, we wash the PEGylated surfaces with DI water. For every washing step involved, we wash three times for 15 s each, and then dry the substrates under a N2 stream. The above protocols can also be applied to other oxidized surfaces, e.g. quartz (SiO2) (Fig. S5†), silicon (Si), silicon nitride (Si3N4), and non-noble metals.

Evaluation of single-crystal TiO2 surface functionalization efficiency via fluorescence microscopy We fabricate 25 × 25 µm2 micro-patterns of PMMA on single-crystal rutile TiO2 and quartz SiO2 substrates for quantitative evaluation of the surface linker functionalization efficiencies via fluorescence microscopy. To fabricate the PMMA micro-patterns, we utilize a similar protocol as described for TiO2 nanocylinders (Methods).

The main differences include spin-coating PMMA 950k A11 to achieve a ~1.9 µm-thick layer and altered ebeam settings (a current of 312 nA, a diameter of 300 nm through defocusing the beam, and a dose of 1000 µC/cm2). For the quartz SiO2 substrates (X-cut, University Wafer, USA) alone, we sputter a 30 nm-thick gold (Au) layer (EM ACE600, Leica, The Netherlands) onto the spin-coated PMMA layer to prevent charging. Following ebeam patterning, we remove the Au layer by a wet etchant (TFA, Transene, USA). As described in Methods, we treat the micro-patterned substrates with O2 plasma (Plasma-PREEN I) prior to the functionalization process.

For the evaluation of the surface functionalization efficiencies of the different linkers to single-crystal TiO2 (and SiO2) substrates, we use amino and NHS-ester modified fluorophores (ATTO 647N, ATTO-TEC GmbH, Germany). They are covalently added to the organic functional groups of the surface linkers employed. For the substrates coated with ETA, APDMES, and BADMSCP, we dissolve NHS-ester labeled fluorophores in PBS buffer (pH 8.4, Sigma-Aldrich, The Netherlands) to a final concentration of 10 µM and add to the functionalized surfaces. After the reaction time of 1 h, we wash the substrates three times each with PBS/TWEEN® buffer (pH 7.4) and DI water to remove residual physisorbed molecules, and dry under a N2 stream. For GPDMES-coated and PEGylated surfaces, we add 10 µM of amino-labeled fluorophores to PBS buffer (pH 7.4) and MES/EDC buffer (100 mM MES (pH 4.7) containing 50 mM EDC (1-Ethyl-3-(3-dimethylaminopropyl)carbodiimid, SigmaAldrich, The Netherlands)), respectively, and the fluorophore coupling reactions take place for 1 h. We wash the substrates with PBS/TWEEN® buffer (pH 7.4) and DI water, followed by drying under a N2 stream, as stated before.

We perform the fluorescence measurements of the functionalized surfaces and nanocylinders using an epifluorescence microscope (IX-81, Olympus, The Netherlands) equipped with a Peltier-cooled back-illuminated electron multiplying charge coupled device (EMCCD) camera (IXON, 512 × 512 pixels, Andor, Ireland), in combination with an oil-immersion objective lens (100×/NA1.3, UPLNFLN, Olympus, The Netherlands). The fluorophores are excited using a diode laser (λ = 640 nm, Cell Laser System, Olympus, The Netherlands). For quantitative measurements, we measure the fluorescence intensity (in photon counts per second) of an area of

12.5 × 12.5 µm2 for different square micro-patterns on each sample. We calculate the average intensity − normalized to the area of 1 µm2 − and the corresponding standard devia ons to compare the different linker molecule coverages (Fig. S5†).

S3 Preparation of DNA construct for OTW experiments We carry out the DNA extension and supercoiling measurements (Fig. 5) on a linear 21.8 kbp DNA that contains biotin and digoxigenin modified nucleotides (biotin-16-dUTP and digoxigenin-11-dUTP, respectively, Roche Diagnostics, The Netherlands) at the opposite extremities (600 bp each). We prepare the DNA by ligating the biotin- and digoxigenin-enriched handles to a 20.6 kbp DNA fragment that is obtained via a NotI/XhoI digestion of Supercos1-lambda 1,2 plasmid (Agilent Technologies, USA). We create the DNA handles by PCR amplification of a 1.2 kbp fragment from pBlueScript II SK+ (Agilent Technologies, USA) using the primers − 5’GACCGAGATAGGGTTGAGTG and 5’-CAGGGTCGGAACAGGAGAGC − in the presence of either bio n-16-dUTP or digoxigenin-11-dUTP. Prior to ligation using T4 DNA ligase (New England Biolabs, UK), the biotin and digoxigenin containing handles are digested with NotI and XhoI, respectively.9 Preparation of flow cell for OTW experiments We perform OTW experiments (Fig. 5) in a custom-made flow cell assembled from two 24 × 60 mm2 borosilicate coverslips (#1.5, ~170 µm thickness, Menzel GmbH, Germany) separated by a single-layer Parafilm® (Sigma-Aldrich, The Netherlands) spacer.1,8,10 We drill two holes of ~1 mm diameter in the top coverslips using a sand blaster, to connect with inlet and outlet tubings. Prior to flow cell assembly, we clean the coverslips using a 4% (v/v) aqueous Hellmanex® III (Hellma GmbH, Germany) solution and then DI water, in both cases by sonication for 20 min at 40 °C. We dry the cleaned coverslips under a N2 stream. For the bottom coverslips, we perform surface functionalization to attach biomolecules. To increase the density of surface hydroxyl groups, which allows for denser, more homogeneous functionalization, we treat the bottom coverslips with O2 plasma (Plasma-PREEN I) for 1 min with O2 flow rate of 3 scfh and RF power of 200 W. We incubate these coverslips in DMSO solution containing 5 M ETA for 12 h at room temperature. Afterwards, we wash the functionalized coverslips thoroughly with DI water and dry them under a N2 stream. Single-layer Parafilm® spacers are prepared by cutting out the desired flow cell channel shape, which is properly aligned to the holes in the top coverslips. Finally, we assemble the flow cells and seal the channels by melting the Parafilm® spacers between the coverslips on a hotplate for 30 s at 90 °C.

Bioconjugation of DNA to single-crystal TiO2 nanocylinders for OTW experiments In an OTW, we are able to carry out the extension and coiling measurements on individual, torsionally constrained DNA molecules.1–5 We tether the DNA molecules to the bottom surface of the flow cell channel via digoxigenin:anti-digoxigenin coupling and to the functionalized single-crystal rutile TiO2 nanocylinders via biotin:streptavidin coupling (Fig. 5a). To do so, we perform three steps. First, we PEGylate the ETA-coated flow cell channel both to covalently attach digoxigenin antibodies and to ensure an effective surface passivation against non-specific physisorption of streptavidin-coated TiO2 nanocylinders. We achieve this by incubating the channel with 2 mM PEG dissolved in 100 mM MES buffer (pH 4.7) for 1 h. After washing with DI water, we incubate for 1 h with 8 µM digoxigenin IgG antibodies (Roche Diagnostics, The Netherlands) dissolved in MES/EDC buffer (pH 4.7). We wash the channel with PBS buffer (pH 7.4) and incubate BlockAid™ (Life S4 Technologies, USA) for 1 h for additional surface passivation. Subsequently, we wash the channel with PBS buffer (pH 7.4). Second, we attach individual DNA molecules via the digoxigenin handles to the digoxigenin antibody-covered flow cell channel by incubating 5 pM of DNA for 1 h. We remove non-specifically adhered DNA molecules by washing the channel with PBS buffer (pH 7.4). Third, we attach the DNA via the biotinylated handles to the streptavidin-coated TiO2 nanocylinders (Methods) by incubating them in the flow cell channel for ~30 min. We remove non-attached nanocylinders by flushing 1:1 diluted (v/v) BlockAid™ in PBS/Triton™ buffer (pH 7.4). We perform all DNA extension and coiling measurements in PBS/Triton™ buffer.

OTW instrumentation and DNA measurements with single-crystal TiO2 nanocylinders For optical trapping of single-crystal TiO2 nanocylinders in our OTW setup (Fig. 5b,c), we expand and collimate a linearly polarized laser (λ = 1064 nm, Compass CW 1064-4000M, Coherent, The Netherlands) beam using a beam expander (4401-181-000-20, LINOS, Germany) to fill properly the input aperture of an objective lens (60×/NA1.2, CFI-PLAN-APO-VC-60XA-WI, Nikon, The Netherlands). The power entering the objective lens is always set to 100 mW during experiments. The objective lens focuses the laser beam into a flow cell, generating an optical trap.