Synthesis of Small Gold Nanorods and Their Subsequent Functionalization with Hairpin Single Stranded DNA

Small gold nanorods have a significantly large absorption/scattering ratio and are especially beneficial in exploiting photothermal effects, for example in photothermal therapy and remote drug release. This work systematically investigates the influence of growth conditions on the size, growth yield, and stability of small gold nanorods. The silver-assisted seed-mediated growth method was optimized to synthesize stable small gold nanorods with a high growth yield (>85%). Further study on the influence of silver ions on the growth facilitates the growth of small gold nanorods with tunable longitudinal surface plasmon resonance from 613 to 912 nm, with average dimensions of 13–25 nm in length and 5–6 nm in diameter. Moreover, the small gold nanorods were successfully functionalized with thiol-modified hairpin oligonucleotides (hpDNA) labeled with Cy5. Fluorescence intensity measurements show an increase in the presence of target DNA and an enhanced signal/background ratio when the longitudinal surface plasmon resonance of small gold nanorods overlaps with the excitation and emission wavelength of Cy5. This coincides with a reduced fluorescence lifetime of Cy5 in the hairpin structure, indicating surface plasmon resonance-enhanced energy transfer to the small gold nanorods. This study may provide insight on the synthesis and functionalization of small gold nanorods in biomedical sensing and therapy.


INTRODUCTION
Gold nanorods have been demonstrated as promising agents for various biomedical applications including cancer diagnosis and treatment. 1−3 The choice of gold nanorods stems from their unique optical properties arising from localized surface plasmon resonance. 4,5 The longitudinal surface plasmon mode of gold nanorods depends on the particle aspect ratio 4−6 and can thus be tuned from the visible to the near infrared (NIR) region. This shape-dependent optoelectronic property manifests itself in various processes such as strong surface-enhanced Raman scattering, two-photon luminescence, surface plasmon enhanced energy transfer, and photothermal effects. 7−13 These combined with low toxicity, water solubility, and biocompatibility, have made gold nanorods a versatile nanomaterial with demonstrated applications in various areas including biological sensing, imaging, drug delivery, and cancer therapy. 14−20 Localized surface plasmon resonance results in enhanced optical absorption and scattering. Gold nanorods with a longitudinal surface plasmon mode in the NIR region that matches the biological window are ideal agents for biomedical sensing, imaging, and photothermal therapy in treating cancer. Functionalized gold nanorods have been demonstrated as useful probes in energy transfer studies for understanding molecular interactions, recognition of analytes, and binding sites in intracellular environments. 21,22 For example, the energy transfer process from the fluorophore (donor) to gold nanorods (acceptor) has been exploited to good effect to understand the molecular interactions between proteins, DNA, mRNA, and fluorescence lifetime-based sensing of bio-analytes. 21 −23 In addition, Wei et al. demonstrated the assembly of fluorophore-labeled hairpin DNA (hpDNA) on the large gold nanorod structure for detecting mRNA in a homogenous solution. 24 hpDNA has been utilized in various biological applications because of its increased sensitivity and specificity for target recognition. 24,25 Small gold nanorods are especially beneficial for applications exploiting the photothermal effect, 26,27 as they have a higher absorption to scattering ratio than that of larger gold nanorods, therefore converting more energy to heat. 20 In addition, small size gold nanorods have a large surface area to volume ratio, high cellular uptake, high clearance rate from the liver, spleen, and the kidney, and ease of subcellular accessibility. 28−36 Furthermore, the efficient fluorescence quenching abilities of small gold nanorods render them good quenchers of background signals in fluorescence detection techniques. 37 Small gold nanorods are therefore ideal agents for photothermal therapy, diagnostic, biosensing, and drug delivery. 5 Most of the relevant works so far have used gold nanorods synthesized via a seed-mediated growth method developed by Murphy's and El-Sayed's groups. 38,39 The gold nanorods thus synthesized are typically over 50 nm in length and 11−15 nm in diameter. 29,31,35,40 Recently, Ali et al. 40 synthesized small gold nanorods (18−25 nm in length and 4−5 nm in diameter) via a seedless growth method by adjusting the pH and concentration of sodium borohydride (NaBH 4 ) and hexadecyltrimethylammonium bromide (CTAB), respectively. Jia et al. 26 synthesized small gold nanorods with a diameter of less than 10 nm via the seed-mediated method by simultaneously varying both the amount of seed and CTAB added to the growth solutions to tune the diameter of the gold nanorods. However, synthesis of stable, small gold nanorods with controlled aspect ratios to achieve surface plasmon resonance at selected wavelengths spanning a large wavelength region remains challenging.
Here, we have systematically investigated the influence of growth conditions on the size, growth yield, and stability of small gold nanorods. Stable small gold nanorods of improved yield were achieved by optimizing the amount of seeds added to the growth solution (shown in Figure 1). Small nanorods of varying aspect ratios with longitudinal surface plasmon resonance from 613 to 912 nm were successfully synthesized by simply varying the amount of silver nitrate in the growth solution. Specifically, stable short small gold nanorods were synthesized with an average length of 14.2 nm. Moreover, we report for the first time the functionalization of small gold nanorods with thiolated Cy5-labeled single strand hpDNA via the ligand exchange and salt aging process. This demonstrates the possibility of assembling thiolated aptamer and other DNA structures onto small gold nanorods. Therefore, this work should facilitate the development of small gold nanorod-based agents for enhanced performance in targeted analyte sensing, biomolecular labeling, photothermal therapy, and drug delivery.

RESULTS AND DISCUSSION
2.1. Influence of Seeds and CTAB on the Size of Small Gold Nanorods. The silver-assisted seed-mediated growth method was used to synthesize small gold nanorods due to its flexibility and ease of control over particle size and shape. To study the influence of seeds and CTAB on the size of gold nanorods, we adopted the reported silver-assisted seedmediated growth protocol 26 to produce three samples coded as G 9 S 1 , G 8 S 2 , and G 6 S 4 as described in Section 4.2. Figure 2 shows the scanning electron microscopy (SEM) images and extinction spectra of these samples. It was found that the gold nanorods had an average length of less than 17 nm and diameter below 6 nm as shown in Table 1. The distribution of sizes of small gold nanorods for these samples are shown in the Supporting Information ( Figure S1). Both the length and diameter decrease by increasing the amount of seeds added in the growth solution. The decrease in size as the seed amount increases is not surprising as increased nucleation sites results in a reduced amount of gold atoms available to each nanorod. 41,42 Similarly, the aspect ratio decreases as the amount of seeds in the growth solution increases due to the shortening of length 26,43 that leads to a shift in the longitudinal surface plasmon resonance peaks from 776, to 744, to 717 nm for G 9 S 1 , G 8 S 2 , and G 6 S 4 , respectively.
The growth yield of small gold nanorods (number of nanorods/number of total particles) reached 69% in G 9 S 1 as shown in Table 1. It decreases dramatically to 12% in G 6 S 4 as the seed to CTAB ratio in the growth solution increases. This could be due to a reduced amount of CTAB surfactant selectively binding to the sides of each nanorod, resulting in a less anisotropic growth and generation of more nanospheres. Previous reports indicated that more byproducts of nanospheres were formed than gold nanorods as the amount of seed solution in the final growth mixture increased. 26,40,44 Moreover, G 6 S 4 was found not to be stable; Figure 2d shows that the UV−vis extinction spectrum changed from the initial two surface plasmon bands (G 6 S 4 ) to one peak (*G 6 S 4 ) after 24 days of synthesis, indicative of the transformation of gold nanorods to gold nanospheres. This implies the limitation of synthesizing short smaller gold nanorods by merely increasing the seed to growth solution ratio.
2.2. Influence of Silver Nitrate on the Growth of Small Gold Nanorods. Silver ions play a vital role in the formation of gold nanorods and impact on their aspect ratios and growth yields. Silver ions form complexes with CTAB to regulate the structural evolution of gold seeds to the rod-like shape and improve the nanorod yield. 45,46 To investigate the

ACS Omega
Article influence of silver ions on the yield and morphology of small gold nanorods, we varied the amount of AgNO 3 based on the recipe for sample G 9 S 1 . Specifically, the amount of silver nitrate used in the growth solution was varied from 150 to 20 μL in nine samples, while other reagents were unchanged as detailed in 4.3. Figure 3 shows typical SEM images taken from these nine samples. Average length, diameter, and yield of nanorods derived from SEM images are listed in Table 2. The distribution of the sizes of the synthesized small gold nanorods are shown in the Supporting Information ( Figure S2). It can be seen that the length of the nanorods decreases from about 25 to 14 nm as silver nitrate decreased from 150 to 20 μL. The diameter is less sensitive to the amount of silver nitrate, varying slightly between 5 and 6.5 nm. Silver ions were found to promote the anisotropic growth to a nanorod shape although the specific role of silver ions in the synthesis is still not fully understood 38,39,47 The three mechanisms proposed so far include under-potential deposition to favor growth on longitudinal facets, face-specific capping to block specific facets, and modification of CTAB micelle formation through silver-bromide integrations. 39,48 The dependence of the morphology of small gold nanorods on the amount of silver ions observed in this work is in line with a previous study on conventional (large) gold nanorods. Moreover, high yields of nanorod formation (>85%) were found for all the nine samples. Figure 3j depicts the UV−vis extinction spectra taken from the nine samples. A significant blue-shift in the longitudinal surface plasmon resonance from 912 to 613 nm was observed as the amount of silver nitrate decreases from 150 to 20 μL. A near linear relationship was found between the wavelength of the longitudinal surface plasmon mode (688 to 850 nm) and silver nitrate amount (25 to 80 μL), as depicted in Figure 3k. At a higher silver nitrate concentration (>100 μL), the surface plasmon resonance remains around 900 nm. These findings are consistent with previous reports on the growth of large gold nanorods. 39,41 This could imply that the influence of silver ions on the longitudinal surface plasmon mode of small gold nanorods is similar to conventional gold nanorods. The average lengths of nanorods were also found to fluctuate between 24 and 25 nm. Tong et al. recently found that the control of silver ions on the aspect ratio of conventional gold nanorods occurs during the symmetry breaking period and the length of the final nanorods formed is dependent on the available gold atom concentration. 45 Therefore, the unchanged length at a higher silver nitrate concentration suggests the depletion of the available gold atoms.
2.3. Spectroscopic Study of Small Gold Nanorod-Based Nanoprobes. Functionalization of gold nanoparticles is of importance for their biological and biomedical applications. Recently, we have reported a new RNA nanoprobe based on functionalized conventional GNRs with single strand DNA (ssDNA) in a hairpin structure, and the influence of the hairpin structure and packing density on the performance of nanoprobes. 24 The ssDNA hairpin contains a sequence complementary to the target mRNA. The fluorophores are quenched in the absence of target mRNA. In the presence of the target mRNA, the hairpin opens as a result of RNA−DNA base-paring, and fluorescence is recovered or even enhanced with optimal hairpin design. Here, we employed a similar synthetic approach to functionalize small gold nanorods with ssDNA labeled with Cy5. The fluorescence of Cy5 in the hairpin structure is quenched via energy transfer from Cy5 to the gold nanorod, and the energy transfer is prominent when the surface plasmon resonance of gold nanorods is matched to the emission of the fluorophore. To investigate the influence of surface plasmon resonance on the performance of nanoprobes,  Gold nanorods coated with CTAB are not suitable for direct bio-applications due to their cytotoxicity to biological tissues. 49−51 Therefore, the CTAB surfactant on the surface layer of the small gold nanorods was replaced with mercaptohexanoic acid (MHA) via a round-trip phase transfer ligand exchange. 24,52 Zeta potential measurements showed the change of surface charge from 36.0 ± 1.1 mV (LG 18 S 2 ) and 25.7 ± 2.7 mV (SG 18 S 2 ) prior to ligand exchange, to −20.1 ± 6.7 and −12.9 ± 0.3 mV, respectively, after ligand exchange, indicating the replacement of CTAB by MHA. Figure 4a,b show that the longitudinal surface plasmon resonance bands of LG 18 S 2 and SG 18 S 2 blue-shifted to 839 and 654 nm without significantly broadening after the ligand exchange process, indicating a successful ligand exchange with no evident aggregation. 24 Following ligand exchange, both samples were functionalized with Cy5-labeled hpDNA via a salt aging process as described in Section 4.5. The hpDNA functionalized small gold nanorods were stable with no visible sign of aggregation as shown in the Supporting Information ( Figure S3). The functionality of the small gold nanorod-based nanoprobes was evaluated by hybridizing the probes with complementary DNA (cDNA) in a molar ratio of cDNA to GNR-hpDNA 1000:1. The hybridization of cDNA with hpDNA opens the hairpin structure, changing the separation between the Cy5 and gold surface and switching the fluorophore from its initial quenched dark state to an opened bright state.
The fluorescence spectra taken from both samples before and after hybridization are presented in Figure 4c,d. The emission wavelength of the nanoprobes was found centered at 665 nm for both the samples that have a larger spectral overlap with the surface plasmon resonance of SG 18 S 2 -hpDNA and that of LG 18 S 2 -hpDNA as displayed in the Supporting Information ( Figure S4). A significant increase in the fluorescence intensity in the presence of cDNA is apparent in both cases, demonstrating the successful assembly of hpDNA onto small gold nanorods. A larger fluorescence increase (1.8-fold) was observed for SG 18 S 2 -hpDNA where the longitudinal surface plasmon resonance overlaps with the emission of Cy5. On the other hand, a smaller fluorescence increase (1.4-fold) was observed for LG 18 S 2 -hpDNA because the longitudinal surface plasmon resonance is away from the excitation and emission of Cy5.
A three-exponential fluorescence lifetime analysis revealed one short lifetime component τ 1 arising from the detection of scattered excitation, and two lifetime components, τ 2 and τ 3 , from Cy5 as shown in Table 3. The fluorescence decay curves before and after incubation with cDNA are shown in the Supporting Information ( Figure S5). The fluorescence lifetime of the free Cy5 fluorophore has been reported to be approximately 1.8 ns. 53 The sub-nanosecond lifetime, τ 2 , can be attributed to Cy5 near the gold surface in a closed hairpin structure. 11,22 The longer nanosecond lifetime component, τ 3 , arises from Cy5 in an opened hairpin that is relatively unquenched away from the gold surface. It is noted that the average lifetime of Cy5 in SG 18 S 2 -hp prior to hybridization (0.62 ns) is smaller than that in LG 18 S 2 -hp (0.82 ns), indicating surface plasmon-enhanced energy transfer when the longitudinal mode of SG 18 S 2 -hp matches with the emission of Cy5. The increase of the average lifetime for both nanoprobes upon hybridization is consistent with the steadystate measurement, confirming the hybridization of the nanoprobe with the target DNA. This demonstrates the potential for employing small gold nanorods for mRNA detection and imaging. Furthermore, the synthesis and functionalization approach reported here can be used to develop functional small gold nanorods for targeted drug delivery and photothermal therapy.

CONCLUSIONS
We have demonstrated a successful synthesis of stable small gold nanorods of various sizes and aspect ratios with tunable longitudinal surface plasmon resonance from 613 to 912 nm by the silver-assisted seed-mediated growth method. It was found that a decrease in the amount of silver nitrate in the growth solution shortens the length of small gold nanorods. Moreover, small gold nanorods were functionalized with a Cy5-labeled thiol-modified hpDNA. Significant changes in the fluorescence intensities and lifetimes of nanoprobes upon hybridization with cDNA indicate the successful functionalization of small gold nanorods and the capability of small gold nanorod-based probes to detect a potential target including mRNA, microRNA, and ssDNA. Fluorescence intensity measurements show an enhanced signal/background ratio when the longitudinal surface plasmon resonance of small gold nanorods overlaps with the excitation and emission wavelength of Cy5, manifesting the surface plasmon-enhanced energy transfer, and benefiting the optical tunability of gold nanorods. The synthesis and functionalization methods reported in this work should shine light on further developments of functionalized small gold nanorods for applications in biomedical sensing, drug delivery, and photothermal therapy.
4.2. Synthesis of Small Gold Nanorods by Varying Seeds and CTAB Concentration. Small gold nanorods were synthesized via the silver-assisted seed-mediated growth method. 26 To investigate the influence of seeds and CTAB on the growth of small gold nanorods, the growth method was adopted by varying the concentration of seeds in the growth solution. The seed solution was prepared by adding HAuCl 4 (0.01 M, 0.25 mL) to a solution of CTAB (0.1 M, 9.75 mL). Thereafter, a freshly prepared ice-cold NaBH 4 solution (0.01 M, 0.6 mL) was added to the mixture and stirred with a magnetic stirrer for 2−3 min until the seed solution turned a dark brown color. The solution was incubated at room temperature for 2 h for gold seeds' growth before use.
To prepare the growth solutions for three samples (G 9 S 1 , G 8 S 2 , and G 6 S 4 ), various volume of 0.1 M CTAB solutions (9,8, and 6 mL) solutions were made. Then, HAuCl 4 (0.01 M, 0.5 mL), AgNO 3 (0.01 M, 0.1 mL), and HCl (1.0 M, 0.2 mL) were sequentially added to each solution, respectively. After thorough mixing, AA (0.1 M, 0.08 mL) was added to partially reduce gold ions (Au 3+ to Au + ). Stirring continued until the mixture turned colorless. 41 Finally, 1, 2, and 4 mL of the seed solution were added to each sample, respectively, and these were then kept at room temperature overnight. The seed solution acts as a template for growth of rod shape in addition to reducing Au + to Au 0 . 39 The chemicals used for preparing G 9 S 1 , G 8 S 2 , and G 6 S 4 are listed in Table 4. After stirring for a few minutes, 1 mL of the gold seeds, produced under same conditions as described in Section 4.2, was added to each solution and kept at room temperature overnight. The samples thus made were labeled as G 9 S 1 -150, G 9 S 1 -135, G 9 S 1 -120, G 9 S 1 -80, G 9 S 1 -65, G 9 S 1 -50, G 9 S 1 -35, G 9 S 1 -25, and G 9 S 1 -20. The chemicals used for preparing the growth solution to investigate the influence of silver nitrate on the size of small gold nanorods are shown in the Supporting Information Table S1.
4.4. Ligand Exchange of Short Small Gold Nanorods. The CTAB surfactant on the surface of as-synthesized gold nanorods is not bio-friendly and prevents the direct assembly of hpDNA to the gold surface, therefore ligand exchange was performed to replace the CTAB surfactant via ligand exchange. 24 We performed ligand exchange for two samples (LG 18 S 2 and SG 18 S 2 ) using round-trip phase transfer ligand exchange protocol previously reported in the literature. 24,52 Specifically, the CTAB surfactant on the surface of small gold nanorods was replaced by MHA. The pellets of ligandexchanged small gold nanorods were suspended in TBE (pH 8.5, 50 μL) buffer.
4.6. Hybridization of hpDNA-GNRs Nanoprobes with cDNA. Excessive complimentary DNA (cDNA) was incubated with hpDNA-GNRs nanoprobes in a ratio of 1000:1 for 2 h at 4°C for each sample prior to measuring the fluorescence intensity and lifetime response of the nanoprobes. 4.7. Characterization of Gold Nanorods and hpDNA GNRs. SEM (FEI Quanta FEG 250) was used for the morphological characterization of the gold nanorods using a 30 kV electron beam and bright field/dark field scanning transmission electron detectors. ImageJ was used for size analysis. Over 200 nanoparticles imaged at different sample

ACS Omega
Article areas were analyzed to obtain the average length and diameter. The zeta potential of the GNRs was measured using a Zetasizer Nano-ZS (Malvern Panalytical, UK). The gold nanorods were centrifuged at 13 000 rpm for 15 min and resuspended in 2 mL of distilled water in a conventional plastic cuvette (1 cm path length). The absorption spectra of gold nanorods were measured with a UV−visible spectrophotometer (Lambda 2, PerkinElmer). A Fluorolog spectrofluorometer (HORIBA Jobin Yvon Ltd., Middlesex, UK) was used to measure the fluorescence spectra at an excitation of 635 nm.
Time-resolved fluorescence measurements were conducted using the time-correlated single-photon counting technique with an IBH FluoroCube fluorescence lifetime system (HORIBA Jobin Yvon IBH Ltd., Glasgow, UK). The samples were excited with a 638 nm pulsed light-emitting diode (NanoLED) source operating at a repetition rate of 1 MHz. Fluorescence decay measurement was taken at the magic angle (54.7°) to nullify anisotropy effects. Data analysis was performed using the DAS6 package. Fluorescence lifetimes were analyzed by fitting the decay curves to the multiexponential decay model given as where τ i are the decay times and B i the associated amplitudes. The fractional contribution of each lifetime component to the steady-state intensity is represented by The average lifetime (τ ̅ ) is calculated as.
The decay curves were fitted to a 3-exponential model to account for the fluorescence decay components arising from Cy5 in hybridized and closed hpDNAs and light scattering (τ 1 ) that was fixed at 0.5 channel.

* S Supporting Information
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsomega.9b01200.
Preparation of growth solution to investigate the influence of silver nitrate on the size of small gold nanorods, histogram of the length and width of small gold nanorods prepared by varying the ratio of growth solution (G) to seed solution, G 6 S 4 , G 8 S 2 , and G 9 S 1 , histogram of the length and width of small nanorods prepared using different amounts of silver nitrate (AgNO 3 , 0.01 M) from 20 to 135 μL, UV−vis absorption spectrum of Cy5-functionalized small gold nanorods with/without cDNA, emission spectrum of Cy5 and the absorption spectra of SG 18 S 2 -CTAB and LG 18 S 2 -CTAB, and fluorescence decay curves of SG 18 S 2 and LG 18 S 2 nanoprobes (PDF)