Supporting Information Superstructural Raman Nanosensors with Integrated Dual Functions for Ultrasensitive Detection and Tunable Release of Molecules Jing Liu #, Jianhe Guo #, Guowen Meng and Donglei Fan* # These authors contributed equally to this work S1: Experimental Details Fabrication of Au nanorods: large arrays of Au nanorods are fabricated by electrodeposition into nanoporous anodic aluminum oxide (AAO) templates. In brief, firstly, a layer of copper (~ 500 nm in thickness) is thermally evaporated on one side of the AAO template to seal the nanopores and serve as the working electrode in a three-electrode electrodeposition setup. Then, Au nanorods are electrodeposited into the nanopores of the AAO templates at a potential of 0.8V (vs. Ag/AgCl in 3M NaCl) in a cyanide based electrolyte (434 HS RTU, Technic Inc.). Next, the nanorods are released after etching the Cu film and the nanoporous AAO template by a mixture of iron chloride (FeCl 3, 1M) and hydrochloride (HCl, 2M), and sodium hydroxide (NaOH, 2M) solution, respectively. After sonicating and centrifuging in ethanol and deionized (D.I.) water alternatively for two times, the nanorods are redispersed in D.I. water. Creation of Au/silica core-shells: silica shells are synthesized on the surface of Au nanorods conformably via hydrolysis of tetraethyl orthosilicate (TEOS). A suspension of Au nanorods (~10 8 /ml, 1 ml) is mixed with a solution made of 3 ml ethanol, 0.8 ml D.I. water, 100-200 µl ammonia (30 wt%) and 400-800 µl TEOS, and subjected to sonication for one hour at room temperature. After sonication and centrifugation in ethanol and D. I. water alternatively, the obtained Au/silica core-shell nanostructures are redispersed in D.I. water. Assembling and embedding polystyrene (PS) nanospheres in silica: firstly, the outmost surface of the Au/silica core-shells is modified with positive charges by immersion in 1
poly(diallyl-dimethyl-ammonium chloride) (PDDA, 2 wt%) and poly(sodium-p-styrene sulfonate) (PSS, 2 wt%), alternatively. Next, the core-shells are dispersed in the 80 nm PS nanosphere suspension (0.125 wt%) for assembling of nanospheres. Then, another layer of silica is conformably coated on the surface of the nanospheres. Repeating the process of silica coating and nanospheres assembling, PS nanospheres can be arranged in concentric layers in the silica, forming into a 3-D superstructure. Creation of nanoporous superstructure by removal of the embedded PS nanospheres: the embedded PS nanospheres are removed by calcination in air at 550 for 6 hours. Nanoporous silica superstructures are obtained after the calcination. Synthesis of plasmonic Ag nanoparticles: Plasmonic Ag nanoparticles are grown on both the outer surfaces and inner nanocavities of the superstructures by catalytic reduction of silver nitrate (AgNO 3 ). Firstly, the nanoporous superstructures are incubated and stirred in a mixture of AgNO 3 (0.06 M, 500 µl) and ammonia (0.12 M, 250 µl) for sufficient adsorption of Ag ions. Then polyvinylpyrrolidone (PVP, 10 ml of 2.5 10-5 M in ethanol) is added to promote the reduction of AgNO 3 and growth of Ag nanoparticles at 70. After 7-hour reaction, dense Ag nanoparticles are grown on both the outer surfaces and inner cavities of the nanoporous superstructures. Characterization of enhancement of surface areas with fluorescent spectroscopy: the porous silica superstructures are incubated in 10-5 M Rhodamine 6G (R-6G) aqueous solutions for 2 hours before testing. After rinsing in D.I. water, they are dispersed and dried on glass substrates. Measurements of fluorescent intensity is taken by using a 532-nm laser with a spot size of 6 µm and power of 0.5 µw. The exposure time for each test is 5 seconds. SERS measurements: the nanoporous superstructural Raman sensors are incubated in aqueous solutions of probing molecules for 2 hours before rinsing in D.I. water and drying. We tested several probing molecules, including R-6G, 1,2-bis(4-pyridyl) ethylene (BPE), and 2
Nile Blue (NB). A 532-nm laser with a spot size of 6 µm and power of 50 µw is used for SERS measurements. The exposure time for each test is 5 seconds. In SERS mapping, a 532- nm laser with a spot size of 1 µm and power of 90 µw is used. 1 The samples were scanned with a step size of 200 nm. The exposure time for each testing point is 1 second. Controlling biomolecule release with tunable rate: we controlled the release dynamics of biomolecules from the nanoporous superstructural Raman sensors by applying an AC electric field and monitored the release every two seconds continuously by Raman spectroscopy. Firstly, the Raman nanosensors are incubated in aqueous solutions of NB (or adenine) for 2 hours. After the adsorption of biomolecules, these nanosensors are dispersed in D.I. water in a PDMS well on patterned microelectrodes. AC electric fields with different voltages and frequencies are applied to control the release of biochemicals. The concentration of biochemicals on the nanosensors is detected continuously. The laser has a spot size of 6 µm, power of 70 µw. The exposure time for each testing point is 2 seconds. Reference: [1] Y. Mei, S. Kiravittaya, M. Benyoucef, D. J. Thurmer, T. Zander, C. Deneke, F. Cavallo, A. Rastelli, O. G. Schmidt, Optical Properties of a Wrinkled Nanomembrane with Embedded Quantum Well, Nano Letters 2007 7 (6), 1676-1679. 3
Figure S1. Close-up SEM view of Ag nanoparticles in a few nanocavities. The narrow junctions are of a few nanometers. S2: Characterization of PS nanospheres, Ag nanoparticles and nanojunctions between Ag nanoparticles Figure S2. Size distribution of (a) Ag nanoparticles and (b) nanojunctions between them. Characterization with ImageJ (software) shows that the average size of Ag nanoparticles and nanojunctions between them on the outmost surface of the nanoporous superstructural Raman sensors are 27.38±6.73 nm and 2.51±1.01 nm, respectively. The histograms are shown in Figure S2. Since electric fields are greatly enhanced in narrow nanojunctions, we only counted nanojunctions less than 2 nm as hotspots. We estimated that the density of hotspots on the surface of nanosensors is ~793/µm 2 with an average diameter of 1.50±0.39 nm. To estimate the hotspots contributed by the Ag nanoparticles inside nanocavities, the density of PS nanospheres decorated on the core-shell nanostructures is studied. We estimated the densities of the two concentric layers of PS nanospheres as 95/µm 2 and 103/µm 2, respectively. Since there are 3-4 Ag nanoparticles as shown by the crosssectional image of the nanocavities, we assumed there are 4 Ag nanoparticles in each nanocavities, positioned at tetrahedral sites. Then, 6 hotspots can be contributed by each nanocavity. In another word, the density of the hotspots in the two concentric layers of nanocavities are 570 and 618 hotspots/ µm 2 respectively. 4
S3: Estimation of SERS Enhancement Factor (EF): Figure S3. (a) Raman spectrum of 1 M BPE in ethanol solution (integration time 10 seconds). SERS spectra on (b) solid and (c) porous Raman nanosensors with saturated adsorption of BPE, respectively (integration time 5 seconds). 1, 2-bis(4-pyridyl) ethylene (BPE), a type of non-resonant molecule, is utilized as the probing molecule to determine EF of plasmonic substrates. Enhancement factor (EF) of a plasmonic nanosensor can be calculated by: / / where I SERS and I RS correspond to the Raman intensity at 1200 cm -1 of BPE molecules obtained from a SERS substrate and directly from a suspension, respectively. N SERS and N RS are the number of molecules detected from the SERS substrate and suspension without a SERS substrate, respectively. The value of I RS is obtained from 1 M BPE ethanol solution excited by a 532-nm laser at 50 µw (integration time 10 seconds). As shown in the Raman spectrum in Figure S3(a), I RS =3087 counts/second were achieved. N RS is given by N RS =V scat C BPE N A, where V scat is the volume of BPE that contributes to the Raman signals, C BPE is the concentration of BPE solution (1 M), N A is Avogadro s number. 5
V scat is determined by V scat =A obj H obj, where A obj is the size of the laser spot, H obj is the effective height of the detection volume of BPE. The diameter of the laser from a 20 objective is 6 µm, thus A obj =πr 2 =28.26 µm 2. By using the method reported previously (Adv. Mater., 2012, 24, 5457-5463), H obj =56.6 µm was obtained. Therefore, N RS = A obj H obj C BPE N A =28.26 µm 2 56.6 µm 1 mol/l 6.02 10 23 =9.63 10 11 molecules To determine the value of I SERS for both the solid and porous Raman nanosensors, we immersed the nanostructures in 1 mm BPE ethanol solution for 2 hours and rinsed by pure ethanol. Figure S3(b-c) shows the obtained SERS spectra with a 532-nm incident laser (power: 50 µw, integrating time: 5 seconds). We can see that I SERS for the solid and porous Raman nanosensors are 118623 counts/second and 461010 counts/second, respectively. The value of NSERS is estimated by counting the number of BPE molecules existing in the hot spots which make the major contribution to the obtained SERS intensity. Since the laser spot is larger than the length of the nanosensors (5 µm), the effective exposure area on the outer surface of the Raman nanosensors is estimated as (effective factor due to curvature) 850 nm (diameter) 5 µm (length)=1.42 µm 2. Similarly, the effective exposure areas for the inner nanocavities, where the two concentric layers of PS nanospheres locate at, are estimated as 0.53 µm 2 and 0.97 µm 2, respectively. Therefore, the number of effective hotspots contributed by the outer surface and inner nanocavities are 1.42 µm 2 793/µm 2 =1126 hotspots, and 0.53 µm 2 570/µm 2 +0.97 µm 2 618/µm 2 =902 hotspots, respectively. Moreover, we can assume the volume of hotspot is ~3.375 nm 3, which is the average gap size as aforediscussed to the power of three. The volume of BPE molecule is 3 Å 6 Å 10 Å/molecule=180 Å 3 /molecule. Therefore, for solid Raman nanosensors, the value of N SERS can be calculated as: N SERS =1126 hotspots 3.375 nm 3 /(180 Å 3 /molecule) = 21113 molecules, and 6
_ / / 118623/21113 3087/ 9.63 10 1.75 10 For the porous superstructural Raman nanosensors, the N SERS can be calculated as: Then, N SERS = (1126+902) hotspots 3.375 nm 3 /(180 Å 3 /molecule) = 38025 molecules _ / / 461010/38025 3087/ 9.63 10 3.78 10 Figure S4. (a) Raman intensities of Nile Blue (595 cm -1 ) at different concentrations detected from the nanoporous Raman sensors. (b) Linear dependence of Raman intensity on concentrations of Nile Blue from 25 nm to 500 nm. 7
Figure S5. (a) Release rate of NB from porous Raman nanosensors at different AC frequencies. (b) Raman detection of the release of NB molecules with time in an external electric field of 50V, 50-400 khz. Figure S6. Raman detection of dynamic release of adenine molecules from the nanosuperstructures. AC electric field was (a) off in the first 2 minutes, (b) on in the next 2 minutes, (c) and then off again in the last 2 minutes. As shown in Equation (2):, the release dynamics is determined by both the molecule concentration difference from the nanosensors to the bulk solutions ( ) and the release rate ( ). As it is a continuous release process, the values of are different in (a) (b) and (c). The fitted values with errors are shown on top of the respective figures. 8
Figure S7. (a) Schematic of the electrode setup for the rotation of nanosensors. (b) Rotation speed versus frequencies of AC electric fields (14V). (c) Rotation speed versus the square of applied voltages (50 KHz). 9