Radiative loss-determined circular dichroism of plasmonic nanospirals with bendable stability of chiroptical activity

Abstract

There is a lack of analytical approaches to study chiroptical activity of chiral nanoplasmons. Herein, LC circuit theory is proposed to analyze the chiroptical activity of heterojunction nanospirals, revealing the main contribution to be from radiative loss. Furthermore, chiral nanoplasmonic flexible thin films exhibit excellent mechanical stability of their chiroptical activity, paving the way to develop flexible/wearable optoelectronic devices integrated with chiral nanoplasmonics.

---

Chiral metamaterials have exhibited novel optical properties and substantially promoted various applied research studies into, for instance, broadband circular polarizers in the visible-IR-microwave region,^1,2 integrated photonics using electrical actuation to transfer chiroptical signals,³ irradiative polarization detection and thermal imaging using thermal actuation,⁴ molecular detection,^5,6 and chiral plasmon rulers.^7,8

Plasmonic nanospirals (NSs) have a helical pitch (P) of less than 100 nm and are an important member in the family of chiral metamaterials, due to their size being close to the physical limit. In 2013, Fischer et al. used glancing angle deposition (GLAD) to generate a period array composed of gold NSs with a P of ∼20 nm over a macro-scale area, using liquid nitrogen to dramatically reduce substrate temperature and consequently prohibit surface diffusion of Au adatoms.² Since then plasmonic NSs have attracted increasing research interest, and a number of diverse fabrication methods have been developed.^9–11 The chiroptical activity of plasmonic NSs, characterized by circular dichroism (CD) and optical rotatory dispersion, has been investigated by engineering structural helicity, NS-to-NS spacing and NS materials.^12–16 Meanwhile, numerical simulations have been used to study chiroptical principles.^2,17–21 However, analytical modes, of which there is a lack, are highly desired to increase intuitive understanding of chiroptical principles and to help in the instruction of designing new-generation integrated optoelectronic devices.

Flexible optoelectronics and optoelectronic textiles are a vision of the future of electronic devices,^22,23 and could integrate functions, for example, solar energy harvesting,^24,25 superpower energy storage,^26,27 human-motion monitoring,²⁸ molecular detection in health care²⁹ and food production,³⁰ and point-of-care applications.³¹ There is no doubt that chiral metamaterials could play an essential role in flexible/wearable optoelectronic devices. To our best knowledge, however, there is no report on the study of the mechanical stability of plasmonic NSs’ chiroptical activity, which could fundamentally prevent their applications in future integrated optoelectronic systems.

In this work, we propose an analytical LC circuit mode to simulate and analyze the CD of silver NSs (AgNSs), which is considered to be contributed by radiative and ohmic loss. GLAD is used to generate AgNS : Ti nanorod : AgNS heterojunction NSs (HJNSs), and the length of the Ti nanorod (l_TiNR) is engineered to be in the range of 0–177 nm. The CD of the HJNSs in the UV-visible region (with a wavelength λ of 300–700 nm) is monitored as a function of l_TiNR, which can be analytically simulated by LC circuit theory, and the ratio between the radiative and ohmic loss in the AgNSs is precisely calculated. The analytical calculation reveals that the AgNSs have chiroptical activity that is mainly contributed by radiative loss. Furthermore, the AgNSs are deposited on flexible thin films made of ITO-coated polyethylene terephthalate (i.e., ITO-PET), and the CD is monitored while mechanically bending the ITO-PET films deposited with the AgNSs. The chiroptical activity can be retained after bending forwards 50 times, but the CD amplitude is significantly degraded after bending backwards 50 times. The cause of the chiroptical degradation is studied. Substrate temperature is controlled to be −50 °C to prevent PET melting. The selection of ITO-PET is based on the following concerns. First, PET can work in a wide range of operating temperatures and has good planarization.^32,33 Second, the ITO coating can significantly reduce the surface roughness of PET, so that the AgNSs can be deposited uniformly on the thin film. Third, the ITO coating can effectively improve the thermal conductivity of PET to further prevent the melting of PET.

A close-packed array of two-pitch AgNSs was deposited on a non-patterned surface by GLAD. For easy differentiation, two-pitch left-handed AgNSs are denoted 2LH-AgNSs, and their mirror images are denoted 2RH-AgNSs (Fig. 1a). The 2LH/RH-AgNS array exhibits very strong CD signals in the UV-visible region, composed of two CD peaks that are separated at a λ of ∼400 nm and have opposite signs to one another. Although the array is too thick to monitor an extinction spectrum, our previous study revealed that the UV and visible CD peaks are ascribed to the transverse (T) and longitudinal (L) plasmonic mode, respectively. The CD spectrum flips around the zero-CD axis as the helical handedness is switched, illustrating that the AgNSs have chiroptical activity intrinsically stemming from the structural helicity. Inserting a TiNR (with a tilting angle of ∼63° with respect to the substrate normal) in a two-pitch AgNS to separate the two one-pitch AgNSs, which is confirmed by EDS (Fig. S1†), leads to a HJNS that retains the helicity-induced chiroptical activity (Fig. 1b). In the 1LH-AgNS : TiNR : 1LH-AgNS (i.e., 1LH : TiNR : 1LH) HJNSs, elongation of the tilted TiNR to 177 nm causes the two plasmonic modes to reduce in CD amplitude continuously and redshift in a range of 20–40 nm (Fig. 2a).

Fig. 1 CD spectra: (a) 2LH-AgNSs (red line) and 2RH-AgNSs (blue line) with a P of ∼210 nm; (b) 1LH : TiNR : 1LH (red line) and 1RH : TiNR : 1RH HJNSs (blue line) with a P of ∼210 nm and l_TiNR of 106 nm. “T” and “L” denote the transverse and longitudinal plasmonic modes, respectively. Insets: (a and b) schematic diagrams and SEM cross-sectional images of the samples with a scale bar of 200 nm; (b) λ_max is the wavelength at which a CD peak has its maximum amplitude (CD_max).

Fig. 2 (a) CD spectra of the 1LH : TiNR : 1LH HJNSs with a P of ∼210 nm and l_TiNR of 0 (black line), 30 (red line), 106 (green line) and 177 nm (blue line). Plots of CD_max^−0.5 (black squares) and λ_max (red triangles) versus l_TiNR (empty symbols for l_TiNR of 0; solid symbols for l_TiNR > 0): (b) T-mode, (c) L-mode. At l_TiNR > 0, the plots of CD_max^−0.5–l_TiNR are linearly fit (black lines) and extrapolated to l_TiNR = 0.

The TiNR-induced chiroptical weakening is analytically simulated by LC circuit theory (Fig. 3a and b; see ESI S1†), which shows that the plasmonic CD is proportional to the light-induced power loss of the HJNSs.³⁴ Incident light interacts with the HJNSs and excites an oscillating electric current (I) in the HJNSs, which are equivalently composed of five in-series electric components including two identical 1LH-AgNSs, two identical 1LH-AgNS : TiNR contacts and one TiNR. Since the tilted TiNRs don’t have a chiroptical response (Fig. S2†), the CD_max (the maximum CD amplitude of the plasmonic T- and L-modes) of the HJNSs is proportional to the power loss of I²(2R_1LH) (eqn (S1),† where R_1LH is the electrical resistance of the 1LH-AgNS, and “2” denotes the contribution of the two 1LH-AgNSs), at the resonance wavelength λ_max (inset in Fig. 1b) given by

where υ_c is the speed of light, and L_1LH and C_1LH are the electric inductance and capacitance of the 1LH-AgNS, respectively. Since λ_max of the two plasmonic modes varies by less than 10% with l_TiNR (Fig. 2b and c), it can be assumed that L_1LH and C_1LH are independent of l_TiNR. The light-induced power loss, contributing to CD_max, includes radiative loss which is attributed to some of the light energy entering into the AgNSs and being released as re-emitted radiation, and ohmic loss which is ascribed to the rest of the light energy being absorbed by the AgNSs and eventually becoming heat. As a result, the 1LH-AgNS has R_1LH contributed from R_rad,1LH (radiative R_1LH) and R_ohm,1LH (ohmic R_1LH) that are electrically in series (eqn (S5)†). The titled TiNR without chiroptical activity tends to have an indirect effect on the CD_max by altering I (eqn (S2) and (S3)†). The TiNR elongation causes R_TiNR (electric resistance of TiNR) to increase (eqn (S6)†) and thus reduces I, accounting for the decrease of CD_max (Fig. 2a). The AgNS : TiNR contacts may play a non-trivial role in determining I and consequently CD_max, in terms of the contact electric resistance R_c (eqn (S3)†). It is derived from LC circuit theory that increases linearly with l_TiNR (eqn (S7)†), and the interception (a, eqn (S8)†) and slope (b, eqn (S9)†) can be evaluated by using the linear fitting in Fig. 2b and c. The ratio of R_c to R_1LH can be calculated by evaluating a and CD_max,2LH (CD_max of the 2LH-AgNSs, Fig. 1a), according to eqn (S8), (S10) and (S11).† Given l_1LH (the length of 1LH-AgNS) of 675 nm, the ratio of (R_1LH + R_c)/l_1LH to R_TiNR/l_TiNR can be evaluated by eqn (S12).† The ratio of R_rad,1LH to R_ohm,1LH can be evaluated by eqn (S13),† and is equal to that of radiative loss-determined CD_max to ohmic loss-determined CD_max with a given I propagating in the electrically in-series resonance CD circuit.

Fig. 3 LC circuit theory of a 1LH : TiNR : 1LH HJNS: (a) schematic diagram of a HJNS excited by circularly polarized light; (b) equivalent circuit diagram; (c) electric evaluation of the T and L modes. R_rad,1LH and R_ohm,1LH are the electric resistance of the 1LH-AgNS, which are attributed to radiative and ohmic loss, respectively. R_1LH = R_rad,1LH + R_ohm,1LH. R_TiNR and R_c are the electric resistance of the TiNR and AgNS : TiNR contacts, respectively. The 1LH-AgNS resonant system has the electric inductance L_1LH and capacitance C_1LH.

Fig. 3c summaries the LC circuit evaluation in terms of the two plasmonic modes. R_c/R_1LH of the T-mode is evaluated to be 0.045, and that of the L-mode is −0.068. The negative sign results from the fact that the L-mode has (highlighted by the olive dashed rectangle in Fig. 2c). Note that the 2LH-AgNSs are composed of two pitches that grow around one longitudinal axis (i.e., the co-axial structure, Fig. 1a), but the two pitches in the 1LH : TiNR : 1LH HJNSs grow around two longitudinal axes (i.e., the bi-axial structure, Fig. 1b). Since the L-mode is sensitive to the alignment of the longitudinal axes, the difference in the co- and bi-axial structures may account for the negative sign of R_c/R_1LH of the L-mode. Regardless of the sign, the evaluation of R_c/R_1LH reveals that R_c is negligible with respect to R_1LH. The ratio of (R_1LH + R_c)/l_1LH to R_TiNR/l_TiNR of the T-mode is 0.315, and that of the L-mode is 0.399. It is illustrated that the insertion of TiNR increases the electrical resistance R_t of the HJNS, resulting in the decrease of I and CD_max. Given the negligible R_c, eqn (S13)† can be simplified to be

where ρ_Ag and ρ_Ti are the electric resistivity of Ag and Ti, respectively. The evaluation shows that R_rad,1LH/R_ohm,1LH of the T-mode is 7.45, and that of the L-mode is 9.72. It is concluded that the two plasmonic modes have chiroptical activity that is mainly contributed from the radiative loss, and the radiative contribution to the L-mode is slightly higher than that to the T-mode. The resonance of the L-mode occurs in the visible region, and that of the T-mode in the UV region. Compared with visible light, inside the metal UV light has a smaller extinction coefficient and so extends further into the metal.³⁴ This accounts for the larger contribution of radiative loss to the L-mode CD than the T-mode CD. Furthermore, we deposited the 2LH : TiNR HJNSs with l_TiNR of ∼100 nm. LC circuit theory reveals that the two plasmonic modes have CD_{max,1LH : TiNR : 1LH} < CD_{max,2LH : TiNR} < CD_max,2LH (eqn (S15)†), which is in good agreement with the experimental results (Fig. S3†).

One-pitch AgNSs were deposited on ITO-PET, which is transparent in the visible region (Fig. S4a†). Since the UV radiation is blocked by ITO-PET, the flexible AgNS thin films only exhibit the L-mode CD peak that flips around the zero-CD axis as the helical handedness is switched (Fig. S4b†), illuminating that the flexible thin films have the helicity-induced chiroptical response. When the flexible sample was bent forwards (inset in Fig. 4a and b), the helical structures were barely damaged at the macro- (Fig. S5b†), micro- (Fig. 4a) and nano-scale (Fig. 4b). As a result, multiple forward bends hardly deteriorate the L-mode in terms of CD_max (Fig. 5c) and A_CD (the integrated area of a CD peak with respect to the zero-CD axis, Fig. S6†), and causes a slight blueshift of λ_max by less than 3 nm (Fig. 5d). In contrast, backward bending (inset in Fig. 4c) causes the L-mode to have an abrupt decrease of more than 20% in the CD_max and A_CD after the first bending, and then gradually reduce to 60–65% of the original CD_max and A_CD with multiple bending (Fig. 5b). Meanwhile, the backward bending leads to a redshift of the L-mode, which fluctuates in a range of 0–10 nm (Fig. 5d). The chiroptical degradation caused by the backward bending could be attributed to detachment of some of the AgNSs from the flexible substrate (Fig. 4c–g). The backward bending inevitably causes physical collision and repulsion between neighboring AgNSs in the close-packed arrays. Consequently, some AgNSs peel off from the flexible polymer to make the polymer appear to have a lot of stripes where there are no AgNS (Fig. S5c†). It is concluded that the AgNS flexible thin films tend to have very stable chiroptical activity under forward bending, but the chiroptical activity can be mechanically degraded under backward bending. Therefore, backward bending should be avoided.

Fig. 4 Forward (a and b) and backward (c–g) bending of the 1LH-AgNS arrays deposited on ITO-PET. (a, b, e, f and g) SEM top-down images; (c and d) SEM cross-sectional images. Insets: (a) photograph of ITO-PET deposited with 1LH-AgNSs subjected to forward bending; schematic diagram of the forward (b) and backward (c) bending.

Fig. 5 CD spectra of the 1LH-AgNS arrays deposited on ITO-PET, as a function of m (the number of bends): (a) forward bending (+m); (b) backward bending (−m). (c) Plot of (CD_max,m/CD_max,m=0) versus m. (d) Plot of (λ_max,m − λ_max,m=0) versus m, in terms of the forward (blue triangles) and backward (red squares) bending.

Conclusions

The AgNS arrays exhibit strong chiroptical activity (characterized by CD) composed of a plasmonic T-mode in the UV region and L-mode in the visible region, and the two modes have CD sign opposite to one another. The chiroptical activity originates from the AgNS helicity. LC circuit theory shows that the plasmonic CD can be contributed from light-induced radiative and ohmic loss. The AgNS : TiNR : AgNS HJNSs, in which the tilted TiNR is inserted into the middle of the two-pitch AgNS and the TiNR length can be engineered by GLAD, provide a controllable helical model to analytically study the AgNSs’ CD using LC circuit theory. It is revealed that the radiative loss makes main chiroptical contribution to the two modes, and that the L-mode has a greater contribution from the radiative loss than the T-mode. The AgNS : TiNR contacts in the HJNSs have a negligible effect on the HJNSs’ CD. The AgNS arrays deposited on the flexible polymer tend to retain chiroptical activity under forward bending; but the backward bending causes the chiroptical activity to undergo 30–40% degradation, due to the serious detachment of AgNSs from the flexible substrate. Benefiting from the flexible engineering of nanostructures using GLAD, this work provides an analytical model to study the plasmonic CD, opening an alternative door to investigate nanoplasmonic optical activity and to design chiral plasmonic systems. It is also demonstrated that generating flexible/wearable chiral nanoplasmonic systems over a macro-scale area, with excellent mechanical stability of chiroptical activity is feasible.

Experimental section

GLAD of AgNSs onto ITO-PET

In a custom-built physical vapor deposition system (JunSun Tech Co. Ltd., Taiwan) with a high vacuum (10⁻⁷ to 10⁻⁶ Torr), silver pellets (99.99%, Kurt J. Lesker) were evaporated at a rate of 0.3 nm s⁻¹ monitored by a quartz crystal microbalance, using an electron-beam accelerating voltage of 8.0 kV and emission current of 15–25 mA. At a deposition angle (α) of 86° with respect to the substrate normal, Ag was deposited on ITO-PET flexible thin films^35,36 over an area of 1.5 × 1.5 cm². During deposition, the substrate temperature (T_sub) was controlled at approximately −50 °C using an ethanol cooling system. To produce 1RH/1LH-AgNSs with a P of ∼210 nm, the substrate was rotated clockwise/counterclockwise in one circle at a rate (R_r) of 0.11° s⁻¹.

GLAD of HJNSs

The HJNSs were deposited onto sapphire (MTL Hong Kong) and Si wafers (Semiconductor Wafer, Inc.) over an area of 1.5 × 1.5 cm², at α of 86° and T_sub of approximately −25 °C. To generate 1RH : TiNR : 1RH and 1LH : TiNR : 1LH HJNSs, 1RH/1LH-AgNSs, tilted TiNRs and 1RH/1LH-AgNSs were deposited subsequently. To produce the tilted TiNRs, the substrate wasn’t rotated and Ti pellets (99.999%, NEXTECK Technology Ltd.) were deposited at α of 86° and a rate of 0.05 nm s⁻¹ using an emission current of 50–65 mA. Deposition duration (t_d) was controlled to engineer l_TiNR to be in the range of 0–177 nm. The substrate was rotated clockwise/counterclockwise in two circles to produce 2RH/2LH-AgNSs. To deposit 2RH/2LH : TiNR HJNSs, the tilted TiNRs were deposited on the 2RH/2LH-AgNSs.

Spectroscopic measurement

DSM 1000 CD (Olis Inc.) was used to monitor the UV-visible CD spectra of the samples deposited on sapphire and ITO-PET, under circularly polarized incident along the substrate normal. To eliminate linear birefringence, the sample was rotated at 0.19 rpm to monitor CD. Four CD spectra were subsequently recorded and algebraically averaged to obtain a CD spectrum of the sample. To study the mechanical stability of the CD of the samples deposited on ITO-PET, the samples were mechanically bent forwards/backwards as a function of m followed by monitoring the CD. UV-visible absorption spectra (HP UV/Vis spect.) of the samples were recorded under non-polarized incident along the substrate normal.

Structure characterization

The as-deposited substrates were mechanically split, leaving the freshly exposed surfaces for SEM and EDS characterization (Oxford, LEO 1530).

Acknowledgements

The authors thank Dr Daniel W. J. Kwong and Ms Anna O. Y. Chan (Chemistry, HKBU) for their technical support with CD, Mr F. Bai (Physics, HKBU) for his technical support with GLAD, Dr Jack Ng (Physics, HKBU) for his discussions about LC circuit theory and financial support by NSFC/21473149, HKBU8/CRF/11E (GLAD), FRG2/14-15/030.

Notes and references

1. J. K. Gansel, M. Thiel, M. S. Rill, M. Decker, K. Bade, V. Saile, G. von Freymann, S. Linden and M. Wegener, Science, 2009, 325, 1513–1515.
2. A. G. Mark, J. G. Gibbs, T. C. Lee and P. Fischer, Nat. Mater., 2013, 12, 802–807.
3. L. Kang, S. Lan, Y. Cui, S. P. Rodrigues, Y. Liu, D. H. Werner and W. Cai, Adv. Mater., 2015, 27, 4377–4383.
4. X. Yin, M. Schaferling, A. K. Michel, A. Tittl, M. Wuttig, T. Taubner and H. Giessen, Nano Lett., 2015, 15, 4255–4260.
5. H. H. Jeong, A. G. Mark, M. Alarcon-Correa, I. Kim, P. Oswald, T. C. Lee and P. Fischer, Nat. Commun., 2016, 7, 11331.
6. E. Hendry, T. Carpy, J. Johnston, M. Popland, R. V. Mikhaylovskiy, A. J. Lapthorn, S. M. Kelly, L. D. Barron, N. Gadegaard and M. Kadodwala, Nat. Nanotechnol., 2010, 5, 783–787.
7. M. Hentschel, M. Schaferling, T. Weiss, N. Liu and H. Giessen, Nano Lett., 2012, 12, 2542–2547.
8. M. Hentschel, M. Schaferling, B. Metzger and H. Giessen, Nano Lett., 2013, 13, 600–606.
9. M. Esposito, V. Tasco, F. Todisco, A. Benedetti, D. Sanvitto and A. Passaseo, Adv. Opt. Mater., 2014, 2, 154–161.
10. B. Frank, X. H. Yin, M. Schaferling, J. Zhao, S. M. Hein, P. V. Braun and H. Giessen, ACS Nano, 2013, 7, 6321–6329.
11. K. Hoflich, R. B. Yang, A. Berger, G. Leuchs and S. Christiansen, Adv. Mater., 2011, 23, 2657–2661.
12. J. G. Gibbs, A. G. Mark, S. Eslami and P. Fischer, Appl. Phys. Lett., 2013, 103, 213101.
13. J. G. Gibbs, A. G. Mark, T. C. Lee, S. Eslami, D. Schamel and P. Fischer, Nanoscale, 2014, 6, 9457–9466.
14. J. M. Caridad, D. McCloskey, J. F. Donegan and V. Krstic, Appl. Phys. Lett., 2014, 105, 233114.
15. F. Bai, J. Deng, M. Yang, J. Fu, J. Ng and Z. Huang, Nanotechnology, 2016, 27, 115703.
16. G. K. Larsen and Y. Zhao, Appl. Phys. Lett., 2014, 105, 071109.
17. Z. Y. Zhang and Y. P. Zhao, J. Appl. Phys., 2008, 104, 013517.
18. J. K. Gansel, M. Wegener, S. Burger and S. Linden, Opt. Express, 2010, 18, 1059–1069.
19. L. Wu, Z. Y. Yang, M. Zhao, P. Zhang, Z. Q. Lu, Y. Yu, S. X. Li and X. H. Yuan, Opt. Express, 2012, 20, 1552–1560.
20. G. K. Larsen, Y. He, J. Wang and Y. Zhao, Adv. Opt. Mater., 2014, 2, 245–249.
21. M. Schafering, X. H. Yin, N. Engheta and H. Giessen, ACS Photonics, 2014, 1, 530–537.
22. D. Zou, Z. Lv, X. Cai and S. Hou, Nano Energy, 2012, 1, 273–281.
23. W. Zeng, L. Shu, Q. Li, S. Chen, F. Wang and X. M. Tao, Adv. Mater., 2014, 26, 5310–5336.
24. O. M. Ramahi, T. S. Almoneef, M. Alshareef and M. S. Boybay, Appl. Phys. Lett., 2012, 101, 173903.
25. A. M. Hawkes, A. R. Katko and S. A. Cummer, Appl. Phys. Lett., 2013, 103, 163901.
26. L. Hu, M. Pasta, F. L. Mantia, L. Cui, S. Jeong, H. D. Deshazer, J. W. Choi, S. M. Han and Y. Cui, Nano Lett., 2010, 10, 708–714.
27. C. Meng, C. Liu, L. Chen, C. Hu and S. Fan, Nano Lett., 2010, 10, 4025–4031.
28. T. Yamada, Y. Hayamizu, Y. Yamamoto, Y. Yomogida, A. Izadi-Najafabadi, D. N. Futaba and K. Hata, Nat. Nanotechnol., 2011, 6, 296–301.
29. K. Mitsubayashi, J. M. Dicks, K. Yokoyama, T. Takeuchi, E. Tamiya and I. Karube, Electroanalysis, 1995, 7, 83–87.
30. P. Si, X.-C. Dong, P. Chen and D.-H. Kim, J. Mater. Chem. B, 2013, 1, 110–115.
31. A. Kaushik, A. Vasudev, S. K. Arya, S. K. Pasha and S. Bhansali, Biosens. Bioelectron., 2014, 53, 499–512.
32. W. A. MacDonald, J. Mater. Chem., 2004, 14, 4.
33. M.-C. Choi, Y. Kim and C.-S. Ha, Prog. Polym. Sci., 2008, 33, 581–630.
34. J. Deng, J. Fu, J. Ng and Z. Huang, Nanoscale, 2016, 8, 4504–4510.
35. K. Zhang, F. R. Zhu, C. H. A. Huan and A. T. S. Wee, Thin Solid Films, 2000, 376, 255–263.
36. X. Z. Wang, G. M. Ng, J. W. Ho, H. L. Tam and F. R. Zhu, IEEE J. Sel. Top. Quantum Electron., 2010, 16, 1685–1689.

---

Footnote

† Electronic supplementary information (ESI) available: LC circuit theory, EDS spectrum of HJNSs, CD of TiNRs, HJNSs and AgNSs, photograph of AgNSs on ITO-PET, and the normalized integrated area of the L-mode CD peak versus m. See DOI: 10.1039/c6ra19145b

---