Supporting Information Ms. ID: se-2015-00051e Paper-based Sensor for Electrochemical Detection of Silver Nanoparticle Labels by Galvanic Exchange Josephine C. Cunningham, a Molly R. Kogan, a Yi-Ju Tsai, a Long Luo, a Ian Richards, b and Richard M. Crooks a,* a Department of Chemistry, The University of Texas at Austin, 105 E. 24th St., Stop A5300, Austin, TX, 78712 USA b Interactives Executive Excellence LLC, 201 North Weston Lane, Austin, Texas 78733 USA 16 Pages S-1
Table of Contents S-1 Cover page S-2 Table of contents S-3 Electrochemical cell for gold electrodeposition on the NoSlip working electrode S-4 3D printed holder design and dimensions S-4 NoSlip and electrode stencil dimensions S-5 NoSlip electrode fabrication S-6 Co-oxidation of Au and ascorbic acid S-7 Scanning electron micrographs of working electrode S-8 Oxidation of Au over-layer to expose Ag S-8 Protocol for AgNP-biotin conjugation S-9 Protocol for binding AgNP-biotin to MµB-streptavidin S-11 ASVs in smaller potential windows prevent Au re-oxidation S-11 Re-deposition of Ag in between ASVs S-14 Effect of scan rate on Ag ASVs S-15 References S-2
Electrochemical cell for gold electrodeposition on the NoSlip working electrode. Gold was electrodeposited onto the working electrode (WE) of the NoSlip using a recessed electrochemical cell made of polytetrafluoroethylene (PTFE). The cell was equipped with a Pt wire counter electrode (CE), and a saturated Hg/Hg 2 SO 4 reference electrode (RE) from CH Instruments (Figure S-1a). The unassembled NoSlip was aligned in the center of the PTFE cell bottom such that the WE was exposed (Figure S-1b). An acrylic disk (3.2 cm diameter, 0.4 cm thick) was placed behind the NoSlip WE for support and a large binder clip was placed on the side of the PTFE cell and acrylic disk to prevent leaking. The dimensions of the PTFE electrochemical cell are provided in Figure S-1b. Figure S-1. (a) A photograph of the electrochemical cell setup for gold electrodeposition on the NoSlip WE. (b) Dimensions of the PTFE cell in (a) with the carbon WE of the NoSlip exposed through the cell bottom. S-3
3D-printed holder design and dimensions. The 3D-printed polylactic acid (PLA) holder that encases the NoSlip and allows for reproducible magnet alignment is shown in Figure S-2a. Before use, the assembled NoSlip is inserted into the 4.80 cm long side-slit in the holder, and then the magnet is placed directly above the WE (Figure S-2b). Figure S-2. (a) An Autodesk 123D 3D drawing of the 3D-printed PLA holder, including dimensions. (b) A photograph of the 3D-printed holder with the NoSlip and magnet inserted. NoSlip and electrode stencil dimensions. The Adobe Illustrator CS6 design that is used to print wax in a particular design onto chromatography paper is shown to scale in Figure S-3a. The regions outlined in blue were removed by laser cutting along the blue lines. The laser cutter was also used to cut out each individual NoSlip (i.e. around the exterior edges of each device). The electrode stencil was fabricated by laser cutting the electrode design into a plastic transparency (Figure S-3b). The stencil was then aligned with the red wax in Figure S-3a, taped on one side to a flat surface, and then the electrodes were stencil printed. S-4
Figure S-3. (a) A scaled drawing of the NoSlip design that is printed onto chromatography paper. (b) The design and dimensions of the electrode stencil used for stencil-printing carbon electrodes onto the NoSlip in (a). NoSlip electrode fabrication. The electrode stencil was aligned over the red wax on Layer 1 (Figure S-3a), and then thickened carbon paste was scraped across the surface of the stencil until all void spaces were filled with ink. The thickened carbon ink was prepared by placing a layer of the commercial paste into a glass petri dish and heating in an oven at 65 C for three 30 min intervals (with stirring between intervals) to remove some solvent. The resulting thickened paste was stored at 4 8 C until needed. After printing, the carbon electrodes were cured by heating the NoSlips at 65 C for 1.0 h. Finally, 4.0 µl of a blue dye solution was dispensed onto Layer 2 of the NoSlip and dried under ambient air, and then copper tape contacts were attached directly to the electrode leads. Metallic Au was electroplated onto the WE as follows. First, HAuCl 4 (400.0 µl of 6.0 mm HAuCl 4 in 0.10 M KNO 3 ) 1 3 was placed in the recessed (PTFE) cell (Figure S-1). Second, the WE potential was held at -0.80 V for 15.0 s to electrodeposit Au. The electrodeposited Au S-5
is not deposited uniformly on the stencil-printed carbon electrode, but rather is present as Au nanoparticles dispersed over the surface (Figure S-4). Finally, the NoSlip was assembled by accordion folding and then placed into the 3D-printed PLA holder with a magnet situated above the WE (Figure S-2). Co-oxidation of Au and ascorbic acid. Figure S-4. Cyclic voltammograms of a 0.10 M BCl solution (black trace) and artificial urine (red trace) in the NoSlip from -0.70 V to 1.00 V vs CQRE at a scan rate of 0.010 V/s. S-6
Scanning electron micrographs of NoSlip working electrode. Figure S-4a is a scanning electron micrograph (SEM) of a carbon WE that was stencil printed onto the NoSlip using the procedure described in the Experimental Section of the main text. An SEM was also collected after electrodepositing gold onto the NoSlip WE (Figure S-5b) for the subsequent galvanic exchange procedure. Each SEM was collected by first cutting out a 2 x 1 inch section around a NoSlip WE, and then attaching the cutout to an SEM stage using double-sided carbon tape. Figure S-5. SEM of (a) a stencil printed carbon WE on paper and (b) electrodeposited Au 0 islands on the carbon WE in (a). S-7
Oxidation of Au over-layer to expose Ag. Figure S-6. Cyclic voltammograms of 0.10 M BCl solution in the NoSlip with (black trace) and without (red trace) Au 0 electrodeposited on the WE (ν = 0.010 V/s). Protocol for AgNP-biotin conjugation. Functionalization of AgNPs with thiol-dna-biotin was performed using a fast ph-assisted method previously reported by Liu and coworkers, 4,5 with slight modifications. Unless stated otherwise, all mixing steps were carried out at 1500 rpm and 24 C on the BioShake iq thermomixer. Briefly, 400.0 µl of 0.75 nm citrate-capped AgNPs and 2.33 µl of a 220.0 µm thiol-dna-biotin solution were combined and mixed for 5.0 min. Next, two aliquots of 100.0 mm citrate-hcl buffer (ph 3.0) were added to the solution (26.5 µl and 27.8 µl, respectively, with 5.0 min of mixing between the first and second additions). The citrate buffer was added to render the solution more acidic while simultaneously increasing the salt concentration. After mixing for 25.0 min, 400.0 µl of 100.0 mm HEPES buffer (ph 7.6) was added to neutralize the ph of the solution. The biotinylated AgNPs (AgNP-thiol-DNA-biotin, referred to hereafter S-8
as AgNP-biotin) were then washed two times by centrifuging (20.0 min at 16,000 g), the supernatant was carefully removed, and then the AgNP-biotin conjugate was resuspended in 400.0 µl of 0.10 M borate solution (ph 7.5) after the first centrifugation and BCl solution (0.10 M borate and 0.10 M NaCl, ph 7.5) after the second centrifugation. Protocol for binding AgNP-biotin to MµB-streptavidin. The biotinylated AgNPs were bound to streptavidin-functionalized magnetic microbeads (MµBs) using our previously reported method. 6 First, 100.0 µl of 1.11 pm streptavidin-coated MµBs were placed in a microcentrifuge tube, and then the MµBs were washed three times with 50.0 µl of 10.0 mm phosphate buffer (ph 7.4). After the third wash, the MµBs were resuspended in 200.0 µl of the previously synthesized AgNP-biotin solution. The resulting solution was mixed at 24 C for 30 min and then washed three times with 100.0 µl aliquots of BCl solution. Note that unless specified otherwise, all washing steps were carried out by magnetic separation; that is, by holding a magnet against the sidewall of the microcentrifuge tube for 30 s, followed by removal of the supernatant and resuspension in the specified solution. Figure S-7 shows the UV-vis spectra of the AgNP-biotin solution before and after incubation with streptavidin-coated MµBs. Prior to addition of the MµBs, two peaks are present: the one at 400 nm arises from the plasmon excitation of individual AgNPs, and the one centered at 260 nm corresponds to the DNA coating on the AgNPs. After conjugation, the peak at 400 nm in the spectrum of the first supernatant is significantly smaller, indicating successful S-9
attachment of the AgNPs to the MµBs. The peak heights at 400 nm before and after incubation were used to calculate the number of AgNPs bound to the MµBs. Figure S-7. UV-vis spectra of the AgNP-biotin solution before incubation with streptavidin-coated magnetic microbeads (black trace) and the first supernatant after removal of the microbeads (red trace). ASVs in smaller potential windows prevent Au re-oxidation. In the main text we claim that the in Figure 3 the peak is larger in the second ASV than the first because some of the Au 0 over-layer is oxidized between -0.20 and 0.20 V, thereby allowing Ag 0 to be electrochemically accessible. This claim is supported by the following experiment. A NoSlip was injected with 50.0 µl of 33.8 pm MµB-AgNP composite (suspended in 0.10 M BCl solution) and the galvanic exchange electrochemical procedure was followed as described in the Experimental Section of the main text, with one modification: the potential was scanned from -0.70 to -0.20 V for the first and second ASVs instead of -0.70 to 0.20 V. The scan was stopped at -0.20 V, S-10
because as shown in Figure S-7 Au 0 starts to oxidize from the WE around -0.20 V. The results of this experiment are shown in Figure S-8. The first and second ASV peaks are nearly identical in size and shape, presumably due to the Au 0 over-layer not being oxidized. Figure S-8. First- and second-scan ASVs of the MµB-AgNP composite in the NoSlip. The scans started at -0.70 and ended at -0.20 V, and the scan rate was 0.010 V/s. Re-deposition of Ag between ASVs. As discussed in the main text, we expected that the Ag peaks in anodic stripping voltammograms (ASVs) subsequent to the first one would get smaller, because the Ag + will diffuse away from the WE and is therefore not fully re-deposited. However, all Ag ASV peaks after the first one are approximately the same size. This behavior will be further justified in the following control experiments. The reduction potential of Ag + was determined by injecting the MµB-AgNP composite (33.8 pm AgNP labels) into the NoSlip and performing galvanic exchange with one modification: a cyclic S-11
voltammogram was collected instead of the second ASV. As seen in Figure S-9a, reduction of Ag occurs between approximately -0.70 and -0.40 V vs CQRE. This means that during successive scans, Ag 0 can be re-deposited at the beginning of the ASVs. However, at a scan rate of 0.010 V/s, Ag 0 would be deposited for ~30 s, which can be compared to the 200.0 s of deposition in the normal sensing experiment (see main text). To test whether Ag 0 is re-deposited at the beginning of the scans we collected ASVs with different starting potentials (Figure S-9b). The peak area was 6.9 µc, 6.9 µc, and 6.3 µc when the ASV started at -0.70, -0.60, and -0.50 V, respectively. Integrating the peak that resulted from the ASV scan starting at -0.40 V is not possible because the baseline and peak are convoluted. Importantly, the cathodic current observed at the beginning of the ASV scans is present even in the absence of the MµB-AgNP composite (black trace), and therefore we attribute it to oxygen reduction (none of the solutions used in the NoSlips is degassed). In conclusion, re-deposition of Ag 0 is not the primary explanation for the Ag ASV peaks maintaining their size in successive scans, because there is not a significant difference in the Ag ASV peak size when scans are started at different potentials (and hence the total deposition time changes). S-12
Figure S-9. (a) Cyclic voltammograms of unconjugated MµBs (i.e., no AgNPs) in 0.10 M BCl solution (black trace) and the MµB-AgNP composite (red trace) in the NoSlip following the galvanic exchange electrochemical procedure (including the second-scan ASV). The voltammograms were initiated at -0.70 V and the potential was reversed at 0.10 V. The scan rate of the cyclic voltammograms was 0.10 V/s. (b) Second-scan ASVs initiated at the starting potentials specified in the legend and concluding at 0.20 V. The scan rate was 0.010 V/s. S-13
Effect of scan rate on Ag ASVs. The following control experiment was performed to support our hypothesis that AgCl (s) forms near the electrode surface and results in reproducible Ag ASVs after the first scan. The MµB-AgNP composite (33.8 pm AgNP labels) was injected into a NoSlip and seven scans were collected at the scan rates indicated in the legend (Figure S-10a). By integrating the ASVs in Figure S-10a we determined the amount of Ag 0 present on each electrode (Figure S-10b). There is only a slight variation in charge with scan rate. The linear correlation between the peak height and the square root of scan rate (Figure S-10c) indicates that the electrochemical reaction is reversible and that the constant presence of the Ag peak results from the limited diffusion of Cl - in the solution. In conclusion, our system involves Cl - diffusing to the electrode surface, Ag undergoing kinetically fast oxidation, and AgCl (s) formation. 7 Similarly, Figure S-10d shows a linear correlation between the baseline current and the scan rate. S-14
Figure S-10. (a) Successive ASVs of the MµB-AgNP composite in the NoSlip from -0.70 V to 0.2 V with a range of scan rates (0.010 to 0.40 V/s). Plots of: (b) charge vs scan rate, (c) peak current vs square root scan rate, and (d) baseline current vs scan rate, all of which were extracted from the ASVs in (a). References (1) Pereira, S. V.; Bertolino, F. A.; Fernandez-Baldo, M. A.; Messina, G. A.; Salinas, E.; Sanz, M. I.; Raba, J. A Microfluidic Device Based on a Screen-Printed Carbon Electrode with Electrodeposited Gold Nanoparticles for the Detection of IgG Anti-Trypanosoma Cruzi Antibodies. Analyst 2011, 136, 4745 4751. (2) Yang, W.; Gerasimov, J. Y.; Lai, R. Y. Folding-Based Electrochemical DNA Sensor Fabricated on a Gold-Plated Screen-Printed Carbon Electrode. Chem. Commun. 2009, 2902 2904. S-15
(3) Cunningham, J. C.; Brenes, N. J.; Crooks, R. M. Paper Electrochemical Device for Detection of DNA and Thrombin by Target-Induced Conformational Switching. Anal. Chem. 2014, 86, 6166 6170. (4) Zhang, X.; Servos, M. R.; Liu, J. Instantaneous and Quantitative Functionalization of Gold Nanoparticles with Thiolated DNA Using a ph-assisted and Surfactant-Free Route. J. Am. Chem. Soc. 2012, 134, 7266 7269. (5) Zhang, X.; Servos, M. R.; Liu, J. Fast ph-assisted Functionalization of Silver Nanoparticles with Monothiolated DNA. Chem. Commun. 2012, 48, 10114 10116. (6) Scida, K.; Cunningham, J. C.; Renault, C.; Richards, I.; Crooks, R. M. Simple, Sensitive, and Quantitative Electrochemical Detection Method for Paper Analytical Devices. Anal. Chem. 2014, 86, 6501 6507. (7) Bard, A. J.; Faulkner, L. R. Electrochemical Methods; 2nd ed.; Wiley: New York, 2001. S-16