Supplementary materials for:
Plasmonic hotspots of dynamically assembled nanoparticles in
nanocapillaries: towards a miRNA profiling platform
Shoupeng Liu,a Yu Yan, a Yunshan Wang, a Satyajyoti Senapati a and Hsueh-Chia Chang*a
Department of Chemical and Biomolecular Engineering,
University of Notre Dame, Notre Dame, Indiana 46556, USA
Materials and methods
Gold particle synthesis and surface function
Gold nanoparticles were prepared by the citrate reduction of Chloroauric acid HAuCl4,as
described in ref 19, and characterized with scanning electron microscope (SEM). Three DNA
oligomers with different functionalities were purchased from IDT.
Oligo A: Fam-CAG AC AG A ACC CAT GGA ATT CAG TTC TCACTGTCT GAA AAA Athiol ….
Oligo B (complementary to middle segment of Oligo A): 5‘-TGA GAA CTG AAT TCC ATG
GGT T-3‘
Oligo C (2 mismatches): 5‘-TGA GAA CTG AAT TCC ATA GGC T-3‘
Oligo B is the DNA mimic of miRNA hsa-miR-146a MIMAT0000449 and oligo C that of
miRNA hsa-miR-146b-5p MIMAT0002809, which is different from oligo B only at two bases.
The FAM fluorescent group on Oligo A has an absorbance Max at 495 nm and emission Max at
520 nm. The pairing segments of Oligo A at the two end form the stem of the hairpin. The
unpaired loop in between is complementary to Oligo B.
For gold particle packing experiments, 20±5 nm gold nanoparticle solution at a
concentration of 10 nM was mixed with 5 μM thiol-functioned dsDNA hybridized with Oligo A
and Oligo B. To remove excess reagents, the mixture was centrifuged for 20 min at 6x1000G.
Supernatant was discarded and the rest was dissolved again in 1XPBS buffer. The centrifugation
process was repeated 5 times to make sure all free DNA were removed. The final gold-particle-
DNA complex was diluted to about 100 pM. It is estimated that about 200 oligomers were
functionalized onto each gold particle surface.
For microRNA biosensing experiments, the gold particles were functioned in the same
manner except that only Oligo A was used instead of dsDNA from hybridized Oligo A and Oligo
B.
Nanopcapillary chip
Commercial glass capillaries (Borosilicate Capillaries with OD1.0 mm and ID 0.78 mm;
Warner Instruments) were pulled using laser-assisted capillary puller (Sutter Instrument Model
P-2000, Laser-based micropipette puller). The capillaries were characterized with SEM and ion
current measurements were carried out with each pulled capillary. Theoretical models(Yan et al,
J. Chem. Phys, 138, 044705 (2013)) for the current and SEM characterization allow us to
develop a pulling protocol for 100 nm tips. Capillaries with currents or images beyond 20% of
the standard are discarded. Nanocapillary was embedded in a polydimethylsiloxane (PDMS)
device which has two reservoirs, a base-side reservoir connected to the original mm-size
capillary and a tip-side reservoir connected to the 100nm capillary tip.
Ion current measurement and Optical imaging
A schematic of our setup for simultaneous optical and ionic current measurements is shown
in Fig. 1(a). PDMS device with embedded nanocapillary was mounted on a microscope
(Olympus IX71). A tungsten lamp with a 480 nm filter was focused by a 60x air objective to
excite fluorescent molecules. Fluorescence is collected via the same objective and reflected by a
dichroic mirror (520 nm wavelength) to a CCD camera (Q imaging, Retiga EXL). We used a
lower frequency than single NP resonance frequency in anticipation of dimers and aggregates.
The maximum intensity corresponds to higher order aggregate modes with lower resonant
frequencies than the single nanoparticle resonant frequency. A pair of Ag/AgCl electrodes are
inserted into the two reservoirs. A source meter ( Keithley 2636A ) is used to apply a voltage and
detect ionic current through the nanocapillary.
In Au NP packing experiments, we first filled the base-side reservoir with 100 pM gold
particles dispersed in 1X PBS buffer and the tip-side reservoir with pure 1X PBS buffer. The
device was then refrigerated at 4 oC for 20 hours to allow diffusion of the gold particles into the
capillary. To pack the particles at the end of capillary, 1V was applied to the electrodes for 10
min.
For miRNA sensing, hairpin probe (Oligo A) functionalized Au NP was placed at the base
side and the chip is refrigerated at 4 oC for 20 hours. Just before the experiments, a 100 μl
solution of oligo B (miRNA target), with a concentration range between 0.1 fM to 10 nM, is
placed in the tip-side reservoir. A negative voltage (-1V or otherwise stated) is then applied to
drive the miRNAs into the capillary (trapping process) to hybridize with the hairpin probes on
the NP in the capillary. A positive voltage (1V unless stated) is then applied immediately to pack
the Au NPs towards the capillary tip.
The process was monitored with fluorescence imaging. Detecting mismatched miRNA
experiment was done in the same manner except Oligo C was used, with two different bases
from Oligo B. The fluorescence signal is taken with illumination after 5 s. If the illumination is
sustained, photo-bleaching is observed after 100 s and the short illumination is used to minimize
photobleaching.
Estimation of number of DNA entering the capillary
The DNA electrophoretic flux can be simply estimated from the measured ionic current
J
C D
carried by the dominant ions: DNA  DNA DNA where I is the current (20nA), C D NA is the
I /e
C D .
concentration, DDNA is the diffusion coefficient of 22 bases DNA with an estimated value
of 2 1010 m2 / s , C and D are the concentration and diffusion coefficient of Na+ and Cl- (we
9
2
use
137mM NaCl for simplication of 1XPBS solution, DNa   1.3 3 1 0 m / s and
DCl   2.0 3 1 09 m 2 / s ). Then the total number of DNA driven electrophoretically into the
capillary in t=10mins can be estimated from the measured current of 20 nA. This estimate yields
I C DNA DDNA
t  0.03 , which means that electrophoresis is not the dominant
a negligible N 
e  C  D
component of the DNA flux.
The other DNA transport mechanism is a transient diffusion process. This transient diffusion
correspond to a relaxing diffusion front with a width corresponding to the diffusion length
DDNA t and the total number of molecules that diffuse into the capillary over the time interval
2p
2p
(DDNAt)3/2 N aCDNA =
DDNA N aCDNAt DDNAt » 52 where N a is the Avogadro's
3
3
Number. We hence estimate about 50 DNA molecules have entered the capillary in 10 minutes.
t=10 min is N =
Fig. S1 The upper frames shows the white light images of the NP assembly dynamics and the
lower frame shows the fluorescence image: (a) at 0 minute, no fluorescence signal is detected
despite the presence of fluorescently tagged molecules. (b) at 1 minute, the NPs begin to
assemble near the tip and produces a bright “hot spot” fluorescence signal; (c) at 15min, the
particle is even more densely packed, but the fluorescence signal begins to fade.
Fig. S2 (a) The fluorescence intensity as a function of the trapping time, at 3 min intervals. (b)
The fluorescence intensity as a function of packing time (for a 10 min trapping time prior to
packing), the concentration of target molecule is 10 pM.
(a)
40
(b)
Counts
30
20
10
0
10
15
20
25
Particle Diameter/nm
30
FIG.S3 (a) SEM image of functioned gold particles shows that most particles are monomers with
a few dimers. (b) Histogram of the gold nanoparticle diameter from the SEM images.
Fig. S4 Photobleaching with continuous exposure. The photobleaching time is about 100 s.