Decoration of Gold Nanoparticles on Surface-Grown Single-Walled Carbon Nanotubes for Detection of Every Nanotube by Surface-Enhanced Raman Spectroscopy Haibin Chu,, Jinyong Wang, Lei Ding, Dongning Yuan, Yan Zhang, Jie Liu, *, Yan Li *, Beijing National Laboratory for Molecular Sciences, State Key Laboratory of Rare Earth Materials Chemistry and Applications, Key Laboratory for the Physics and Chemistry of Nanodevices, College of Chemistry and Molecular Engineering, Peking University, Beijing 100871, China, Department of Chemistry, Duke University, Durham, North Carolina 27708, USA yanli@pku.edu.cn; j.liu@duke.edu Peking University Duke University Supporting Information I. Supporting experimental methods II. Supporting figures for the controlled decoration of Au NPs onto SWCNTs III. Detailed analysis of the origin of the SERS enhancement IV. Different enhancement factors of SWCNTs G-band intensity with different excitation wavelength V. Blinking effect of SERS VI. Reference S1
I. Supporting experimental methods The Raman spectra were collected from either LabRam ARAMIS (Horiba Jobin Yvon) with excitation lasers at 633 nm and 785 nm, or LabRam HR 800 (Horiba Jobin Yvon) with an excitation laser at 633 nm. The mapping of the Raman intensity of SWCNTs was performed on LabRam HR 800 (Horiba Jobin Yvon) with an excitation laser at 633 nm. II. Supporting figures for the controlled decoration of gold nanoparticles onto SWCNTs Figure S1. (a d) AFM topographical and corresponding phase images of the SWCNT samples after gold seed deposition for 3 min (a), 30 min (b), 3 min + 3 min (c), and 10 min + 10 min (d), respectively. From Figure S1, the number densities of gold seeds on SWCNTs are 3 NPs/μmSWCNT, 10 NPs/μmSWCNT, 20 NPs/μmSWCNT, and >30 NPs/μmSWCNT, respectively. It can be concluded that the density of gold seeds increases with the elongation of reaction time and the increase of repetition times during the seed deposition process. S2
Figure S2. (a c) 3D AFM topographical images of the gold/swcnt composites after seeded growth for 5 min (a), 7 min (b), and 60 min (c), respectively. (d) The evolution of the interparticle distance of the gold nanoparticles on SWCNTs depending on the seeded growth time. The interparticle distance of the gold particles can be easily controlled by the reaction time during the seeded growth process as shown in Figure S2. S3
Figure S3. XRD pattern (a), EDS spectrum (b), and XPS spectra (c, d) of the nanoparticle/swcnt composites obtained via seeded growth based on palladium seeds. The peaks marked with asterisks in (a) are from the quartz substrate. All other diffraction peaks are in good agreement with the literature values of gold (JCPDS 04 0784). Both EDS spectrum and XPS spectra give additional pieces of evidence showing that the nanoparticles decorated on SWCNTs are mainly made of gold. S4
Figure S4. AFM topographical images of nanoparticle/swcnt composites after adsorption of PdCl 2 for 40 s (a), 1 min (b), 2 min (c), and 10 min (d), respectively. The density of nanoparticles on SWCNT is 17 NPs/μmSWCNT in (a), 30 NPs/μmSWCNT in (b), and 38 NPs/μmSWCNT in (c), respectively. Too long time resulted in the poor selectivity of nanoparticles on SWCNTs as shown in figure (d). S5
Figure S5. SEM images (a - c) and corresponding AFM topographical images (d - f) of gold/swcnt composites via H 2 reduction and seeded growth after adsorption of PdCl 2 for 1 min (a, d), 2 min (b, e), and 3 min (c, f), respectively. The size distribution of the gold nanoparticles in the samples of Figure S5 is shown in Figure S6. S6
Figure S6. (a - c) Distributions of the gold particle diameter obtained from the AFM images of gold/swcnt composites via H 2 reduction and seeded growth after adsorption of PdCl 2 for 1 min (a), 2 min (b), and 3 min (c), respectively. (d) Vis-NIR absorption spectra of the three gold SWCNT composites on quartz. It can be concluded from Figure S6 that with the increase of PdCl 2 adsorption time, the number density of the final gold nanoparticles on SWCNTs increases while the size of them decreases with more uniform size distribution. Moreover, the size decrease is further proved by the blue-shift in SPR absorption peaks of the composites (Figure S6d). S7
III. Detailed analysis of the origin of the SERS enhancement Presumably the strong SERS enhancement might result from three causes: 2-4 (i) an electromagnetic SERS enhancement due to the SPR of the gold nanoparticles with the optical fields, (ii) a chemical SERS enhancement due to the charge transfer between SWCNTs and gold nanoparticles, (iii) strain-induced Raman intensity enhancement due to the surface pressure exerted on the SWCNTs by the gold nanoparticles. However, the chemical enhancement was supposed to contribute little to the SERS enhancement of semiconducting SWCNTs. 2 Raman spectra taken from the same place on the same SWCNT excited with laser 633 nm before and after gold decoration (Figure S7) showed no obvious shift of RBM peak, though G band downshifted by 6 cm -1 after gold decoration. Besides, no or small shifts were observed for D band (disorder-induced features) and G band (D band overtone). Thus the charge transfer between SWCNTs and gold nanoparticles had small effect on both SERS enhancement and intrinsic characteristics of SWCNTs. 1 No shift of RBM peaks also indicated that the surface pressure of gold nanoparticles on the SWCNTs played a minor role in the SERS in our samples. 3,4 Therefore, it was concluded that the strong SERS enhancement primarily resulted from the electromagnetic SERS enhancement due to the resonance between the optical fields and the electronic excitations in the high density gold nanoparticles. S8
Figure S7. Raman spectra of the same place of the same SWCNT (a) before and (b) after decoration of the gold nanoparticles. The peak of G-band downshifted by 6 cm -1, but RBM peak did not shift after gold nanoparticles decoration. S9
IV. Different enhancement factors of SWCNTs G-band intensity with different excitation wavelength * The data of decorated SWCNTs under the 442 nm laser excitation were obtained after subtracting background photoluminescence intensity. Table S1. Different enhancement factors for the intensity of G + peaks of the SWCNTs with different excitation wavelengths. I Q is the intensity of Raman peak of quartz at around 464 cm -1. S10
V. Blinking effect of SERS Figure S8. Time evolution of Raman spectra of the same place on a gold-decorated SWCNT under 633 nm laser excitation with typical laser energy of ~1.2 mw. The peaks marked with an asterisk in Figure S8 arise from the quartz substrate. Both the signal intensities and frequencies of G band and RBM peaks did not change in 100 s. S11
VI. Reference (1) Scolari, M.; Mews, A.; Fu, N.; Myalitsin, A.; Assmus, T.; Balasubramanian, K.; Burghard, M.; Kern, K. J. Phys. Chem. C 2008, 112, 391-396. (2) Corio, P.; Brown, S. D. M.; Marucci, A.; Pimenta, M. A.; Kneipp, K.; Dresselhaus, G.; Dresselhaus, M. S. Phys. Rev. B 2000, 61, 13202-13211. (3) Venkateswaran, U. D.; Brandsen, E. A.; Schlecht, U.; Rao, A. M.; Richter, E.; Loa, I.; Syassen, K.; Eklund, P. C. Phys. Status Solidi B 2001, 223, 225-236. (4) Yano, T. A.; Inouye, Y.; Kawata, S. Nano Lett. 2006, 6, 1269-1273. S12