Open Access Article
Xi Mao
a,
Abrahan Martinez
c,
Samantha Williams
b,
Shengli Zou
*c and
Jie He
*ab
aPolymer Program, Institute of Materials Science, University of Connecticut, Storrs, Connecticut 06269, USA. E-mail: jie.he@uconn.edu
bDepartment of Chemistry, University of Connecticut, Storrs, Connecticut 06269, USA
cDepartment of Chemistry, University of Central Florida, Orlando, Florida 32816, USA. E-mail: shengli.zou@ucf.edu
First published on 11th June 2026
We report polymer-guided assembly of chiral gold nanorods (AuNRs) in end-to-end (EE) and side-by-side (SS) modes, modulating their chiroptical responses. EE assemblies amplify circular dichroism (CD) response due to the interparticle plasmonic coupling along the continuously helical patterns of c-AuNRs along the chain direction, while SS assemblies reduce CD response as a result of racemic twisting.
AuNRs (100 nm × 18 nm) capped with cetyltrimethylammonium chloride (CTAC) were synthesized using a previously reported method by Murray's group (Fig. S1).32 Surface modification was carried out in a DMF/water (100
:
1, v/v) mixture containing thiol-terminated polystyrene (PS161-SH, Mn = 16.8 kDa, PDI = 1.3) and polystyrene-block-poly(ethylene oxide) (PS37-b-PEO140, Mn = 10 kDa, PDI = 1.2). Typically, ∼20 µL of concentrated AuNRs were injected into 2 mL of DMF solution containing PS161-SH and PS37-b-PEO140. After incubating for 4 h, water was added to a final concentration of 15 vol%. The solution mixture was annealed at 90 °C for 30 min prior to purification by centrifugation. As the PS161-SH concentration (CPS161-SH) decreased from 10 to 0.01 µM, the longitudinal localized surface plasmon resonance (LSPR) peak of AuNRs blue-shifted from 1006 to 880 nm (Fig. S2a and b), which was attributed to the slight deformation of AuNRs (length shortening to ∼90 nm). At a CPS161-SH of 10 µM, AuNRs were uniformly coated by PS with a polymer layer thickness of 16.8 ± 2.6 nm (Fig. S3a and d). Then, the PS thickness slightly decreased to 11.7 ± 2.8 nm as the PS concentration decreased to 1 µM. This decrease is attributed to the change of the chain stretching status of PS161-SH, further leading to surface dewetting as reported previously.31 TEM reveals the transition from core–shell AuNRs@PS to dumbbells (Fig. S2c–j), known as surface dewetting under the ligand-deficient condition of PS161-SH.31 With CHS-PS161 in the range of 0.25 to 0.1 µM, dumbbell-like AuNRs with PS domains capped at the two ends were obtained (Fig. S2g and h). Thermogravimetric analysis (TGA) revealed a grafting density (σ) of 0.058 chains nm−2 for PS161-SH at a CPS161-SH of 0.1 µM (Fig. S4), close to the critical grafting density as reported previously.31,33–35 The PS domain size is 12.8 ± 2.3 nm at a CHS-PS161 of 0.1 µM and the yield of dumbbell-like AuNRs is >90% (Fig. S3b–d). Further decreasing CPS161-SH to 0.05 and 0.01 µM, the domain size became smaller, reaching 8.9 ± 2.1 nm (Fig. S2i and j).
c-AuNRs were synthesized via seed-mediated growth using dumbbell-like AuNRs as seeds in the presence of L-/D-cysteine (L-/D-cys) as chiral inducers (Fig. 1a).29 With PS-capped AuNRs, chiral features grew along the lateral facets. This growth increased the diameter and caused the LSPR band to blue-shift to ∼760 nm (Fig. 1b–e and Fig. S5). As shown in TEM images, c-AuNRs with L-/D-cys, labelled as L-AuNRs and D-AuNRs, showed right-/left-handed helices, and their average diameters increased to 38.6 ± 6.7 nm and 40.9 ± 6.2 nm after 8 growth cycles, respectively (Fig. 1b and e). The L-/D-AuNRs exhibited clear, mirror-symmetric CD spectra, showing a transverse surface plasmon resonance (TSPR) band at 600 nm and an LSPR band at 760 nm (Fig. 1f). L-AuNRs displayed a positive Cotton effect at 756 nm and a negative Cotton effect at 603 nm, while D-AuNRs showed the opposite peaks at the same positions. As a control, AuNRs grown without (WO-) L-/D-cys showed no CD response in the same range. The plasmonic CD originates from chiral surface distortion caused plasmonic harmonics.36 The corresponding g-factor spectra of L-/D-AuNRs also showed symmetric signals with maximum g-factors at TSPR (603 nm) of −7.1 × 10−4 and 3.7 × 10−4 (Fig. S6 and batch 1 in Table S1), respectively. Meanwhile, the diameter and CD intensities of c-AuNRs can be modulated by growth cycles. As the number of growth cycles increased from 4 to 12, the diameter of L-AuNRs increased to 48.3 ± 6.9 nm (Fig. S7a–c). The increased diameter caused a blue-shift of the TSPR band to 540 nm (Fig. S7d); meanwhile, the g-factors gradually increased to −1.9 × 10−3 at ∼530 nm (shown in Fig. S7e and summarized in Table S1).
A similar seed-mediated growth strategy can be extended to AuNR chains. We prepared achiral AuNR chains by adding AuNRs into a DMF/water mixture containing PS161-SH (1 µM) and PS37-b-PEO140 (0.5 mg mL−1), where the water content was set to 3.8 vol%, close to the critical water concentration of PS (∼4.0 vol% for a 1 µM PS solution in DMF).37 The reduced solvent quality of PS ligands in water/DMF mixtures drove the end-to-end (EE) assembly of AuNRs, leading to the formation of achiral AuNR chains (Fig. 2b and c). The nanorod chain length was 578 ± 220 nm (about 6–10 nanorods per chain). Meanwhile, the LSPR band red-shifted from 913 nm to 1145 nm. Using AuNR chains as seeds, the helices were grown with L-/D-cys, denoted as L-/D-chains. After growth, the LSPR peak of the chains red-shifted to 1235 nm (Fig. 2j). This red-shift is likely due to the formation of epiphysis-like nanostructures on individual nanorods, which further enhances the interparticle plasmonic coupling. TEM and SEM images suggested that L-/D-chains were composed of c-AuNRs with right-/left-handedness (Fig. 2d–g), respectively. Their average diameters (short axis of AuNRs) of L-/D-chains were 27.7 ± 3.5 and 26.7 ± 3.3 nm, respectively. Compared with the g-factor of individual L-AuNRs at the TSPR peak (520 nm) after 4 growth cycles (−2.1 × 10−4, Fig. S7e), the g-factor of L-chains at the TSPR peak (618 nm) increased to −5.5 × 10−4, which is 2.6 times that of the individual L-AuNRs (Fig. 2k and l). To verify the impact of plasmonic coupling on chiroptical properties, chiral AuNR chains were disassembled into discrete c-AuNRs by redispersing in DMF, a good solvent for PS (Fig. 2h and i). The LSPR peak blue-shifted from 1213 nm to 895 nm, confirming the disassembly of L-/D-chains (Fig. 2j). The CD spectra of individual c-AuNRs collected in DMF showed the decrease of CD response after disassembly (Fig. 2k). The g-factor of L-chains is 2 times and the g-factor of D-chains is 1.5 times that of their discrete c-AuNRs (Fig. 2l). The enhanced g-factor originates from the synergistic effects of plasmonic coupling and the continuous helical organization of c-AuNRs. The red-shifted LSPR band indicates the formation of collective plasmon excitations, while the extended chiral geometry promotes their coherent chiroptical response, resulting in amplified CD intensity.23,38–40
Furthermore, the CD response of AuNR chains is dependent on the number of growth cycles. Increasing the growth cycles to 8 or 12 resulted in a further red-shift of the LSPR to 1250 nm, while the transverse peak red-shifted to ∼560 nm (Fig. S8a). TEM images showed a progressive diameter increase to 39 nm after 12 growth cycles (Fig. S8c–j). After 12 growth cycles, the g-factor of L-chains increased to −3.3 × 10−3 (Fig. S8b and S9), close to 5.8 times that of chains with 4 growth cycles and about 1.7 times that of individual NRs (L-AuNRs with 12 growth cycles, Table S1). These helical spikes appeared to be periodic along the chain. Note that the overgrowth on NR chains can result in the interconnection of NRs, i.e., the diameter of AuNRs becomes greater than that of polymers; after 8 growth cycles, the chains cannot be disassembled in DMF.
On the other hand, c-AuNRs can assemble in a twisted SS fashion.41 The SS self-assembly of c-AuNRs was constructed by using methanol (MeOH) to disrupt the CTAC bilayer and drive the interaction of the lateral surface of c-AuNRs (Fig. 3a). In a water/MeOH mixture (1/9, vol), the LSPR band of c-AuNRs showed a ∼60 nm red shift after 2 h (Fig. 3b), similar to the reported value.29 The CD spectra also displayed a red-shifted LSPR band (Fig. S9). TEM reveals that SS assemblies comprise twisted but elongated c-AuNR clusters because of the steric repulsion among polymer domains at the two ends of c-AuNRs. The size of SS assemblies grew and a red-shift of its longitudinal peak was observed later. Meanwhile, the CD response of SS assemblies decreased compared with individual c-AuNRs. The g-factor of L-AuNR SS assemblies at 829 nm decreased by 68%, while that at 604 nm decreased by 42%. Similarly, the intensity of D-AuNR SS assemblies at 829 nm decreased by 63%, and that at 604 nm decreased by 39% (Fig. 3e). The reduced chiroptical response in the SS assembly, likely attributed to the racemic twisting within individual assemblies, i.e., the coexistence of both left-/right-handed twist in one single assembly.
To gain further insight into chiroptical modulation via assembly modes, we used the discrete dipole approximation (DDA) method to simulate the impact of chiral SS and EE assemblies on the g-factors. The chiral AuNRs (100 nm × 30 nm) used for the simulation have chiral curls with a width of approximately 3 nm and a height of 2 nm. The structural configurations were not intended to exactly reproduce the experimental configuration; rather, they were chosen to ensure the stability of the calculated g-factor and to examine the effects of the SS and EE assemblies on the g-factor in comparison to a single c-AuNR. The UV-vis spectra showed a blue shift for SS assemblies and a red shift for EE assemblies (Fig. 3f) compared with individual c-AuNRs. The g-factor spectra indicated an ∼1.6-fold decrease for SS assemblies and a 0.3-fold enhancement for EE assemblies (Fig. 3g) relative to individual c-AuNR, consistent with experimental results. We note that the periodicity of surface chiral features is of key importance for this enhancement. In the EE assemblies, the rotation of individual c-AuNRs without periodical chiral structures would lead to the loss of the g-factor. Therefore, the enhanced chiroptical response of the EE assemblies is likely due to continuous dissymmetrical electric field distribution along the longitudinal direction.42,43 In contrast, the reduced chiroptical signals observed for the SS assemblies are attributed to racemic twists that cancel the overall chirality (Fig. 3i).41,44 Because the SS oligomers exhibit direct plasmonic coupling of the TSPR mode, stronger coupling is expected to exert a greater influence on their chiroptical signals, as demonstrated by Wang et al.30
To support this hypothesis, we further use BSA, which is negatively charged to induce single-handed SS assemblies of c-AuNRs.20,24 L-/D-AuNRs with diameters of 44.4 ± 7.2 nm and 39.2 ± 5.6 nm, respectively (Fig. S10a and b), were mixed with BSA at a concentration of 1 µM in 1 mM PBS buffer (pH = 6.4). The polymer domains at the two ends of c-AuNRs suppress EE assembly, while the asymmetric chiral surface-charge distribution of BSA electrostatically guides neighboring c-AuNRs into a preferred handed twist. The red shift of the LSPR band of c-AuNRs was about 15 nm after 90 min (Fig. S11b). TEM images suggested that the SS assemblies were composed of two or three twisted c-AuNRs (Fig. S11c and d). Correspondingly, the CD intensities showed a significant enhancement (Fig. S10c and d). Compared with individual c-AuNRs, the g-factor at 660 nm of D-AuNRs increased 12.4 times, while that of L-AuNRs increased 4 times (Fig. S11e). This enhancement is attributed to the formation of right-handed helices in the presence of BSA.
In summary, we demonstrate a chiroptical modulation strategy by controlling the assembly modes of chiral AuNRs: end-to-end (EE) and side-by-side (SS). Our results revealed that the periodic continuous dissymmetrical features in EE assemblies exhibit chiral enhancements, doubling the CD signals. In contrast, SS assemblies decrease the chiral signals compared with individual c-AuNRs, reducing CD intensity to approximately half in the LSPR region and by about 0.6-fold in the TSPR region, due to the formation of racemic twist. BSA-induced right-handed SS assemblies increase CD signals by 4- to 12-fold because of the decrease of racemic twist and racemic mixture. These findings highlight the critical role of assembly modes in the chiroptical properties of chiral NPs, offering a versatile route to achieve chiroptical modulation. The amplified CD response of EE assemblies may enable sensitive chiroptical detection of stereoisomeric drugs through CD signal changes induced by drug–polymer interactions.45
X. Mao: methodology, investigation, formal analysis, writing – original draft, and writing – review & editing; A. Martinez: software, formal analysis, and review & editing; S. Williams: investigation and review & editing; S. Zou: software, supervision, formal analysis, funding acquisition, and writing – review & editing; J. He: conceptualization, supervision, project administration, funding acquisition, and writing – review & editing.
| This journal is © The Royal Society of Chemistry 2026 |