Efficient Photocurrent Enhancement from Porphyrin Molecules on

Subscriber access provided by Warwick University Library

Article

Efficient Photocurrent Enhancement from Porphyrin Molecules on Plasmonic Copper Arrays: Beneficial Utilization of Copper Nanoanntenae on Plasmonic Photoelectric Conversion Systems Kosuke Sugawa, Daisuke Yamaguchi, Natsumi Tsunenari, Koji Uchida, Hironobu Tahara, Hideyuki Takeda, Kyo Tokuda, Shota Jin, Yasuyuki Kusaka, Nobuko Fukuda, Hirobumi Ushijima, Tsuyoshi Akiyama, Yasuhiro Watanuki, Nobuyuki Nishimiya, Joe Otsuki, and Sunao Yamada ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b13147 • Publication Date (Web): 05 Dec 2016 Downloaded from http://pubs.acs.org on December 10, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 53

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Efficient Photocurrent Enhancement from Porphyrin Molecules on Plasmonic Copper Arrays: Beneficial Utilization of Copper Nanoanntenae on Plasmonic Photoelectric Conversion Systems

Kosuke Sugawa,†,* Daisuke Yamaguchi,† Natsumi Tsunenari,† Koji Uchida,† Hironobu Tahara,‡ Hideyuki Takeda,† Kyo Tokuda,† Shota Jin,† Yasuyuki Kusaka,§ Nobuko Fukuda,§ Hirobumi Ushijima,§ Tsuyoshi Akiyama, Yasuhiro Watanuki,† Nobuyuki Nishimiya,† Joe Otsuki,† Sunao Yamada †

Department of Materials and Applied Chemistry, College of Science Technology, Nihon University,

Chiyoda, Tokyo 101-8308, Japan ‡Division of Chemistry and Materials Science, Graduate School of Engineering, Nagasaki University, Nagasaki 852-8521, Japan §Flexible Electronics Research Center (FLEC), National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Ibaraki 305-8565, Japan Department of Materials Science, School of Engineering, The University of Shiga Prefecture, Hikone, Shiga 522-8533, Japan Department of Applied Chemistry, Faculty of Engineering, Kyushu University, 744 Motooka, Nishiku, Fukuoka, 819-0395, Japan

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 53

ABSTRACT We have demonstrated the usefulness of Cu light-harvesting plasmonic nanoantennae for the development of inexpensive and efficient artificial organic photoelectric conversion systems. The systems consisted of the stacked structures of layers of porphyrin as a dye molecule, oxidationsuppressing layers, and plasmonic Cu arrayed electrodes. In order to accurately evaluate the effect of Cu nanoantenna on the porphyrin photocurrent, the production of Cu2O by the spontaneous oxidation of the electrode surfaces, which can act as a photoexcited species under the visible light irradiation, was effectively suppressed by inserting the ultrathin linking layers consisting of 16mercaptohexadecanoic acid, titanium oxide, and poly(vinyl alcohol) between the electrode surface and porphyrin molecules. The reflection spectra in an aqueous environment of the arrayed electrodes, which were prepared by thermally-depositing Cu on two-dimensional colloidal crystals of silica with diameters of 160, 260, and 330 nm, showed clear reflection dips at 596, 703, and 762 nm, respectively, which are attributed to the excitation of localized surface plasmon resonance (LSPR). While the first dip lies within the wavelengths where the imaginary part of the Cu dielectric function is moderately large, the latter two dips lie within a region of quite small imaginary part. Consequently, the LSPR excited at the red region provided a particular large enhancement of porphyrin photocurrent at the Qband (ca. 59-fold), as compared with that on a Cu planar electrode. These results strongly suggest that the plasmonic Cu nanoantennae contribute to the substantial improvement of photoelectric conversion efficiency at the wavelengths, where the imaginary part of the dielectric function is small. 2 ACS Paragon Plus Environment

Page 3 of 53

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


KEYWORDS: copper nanostructures, light-harvesting nanoantenna, localized surface plasmon resonance, photocurrent, photoelectric conversion system, porphyrin

3 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 53

1. INTRODUCTION As one of very attractive photofunctional electrodes, nanostructured metal electrodes consisting of some noble metals (i.e. Au and Ag) are very promising for the development of highly efficient artificial organic photoelectric conversion systems driven by visible light,1-17 because they generate stronger local electromagnetic fields than that of the incident light fields in the nano-space around them via an excitation of the localized surface plasmon resonance (LSPR). As a result, these electrodes have been demonstrated to act as “light harvesting plasmonic nanoantennae”, which will be an innovative technique in areas such as artificial photosynthesis,18 photoelectrochemical biosensing,19,20 and logic gates.21,22 Enhancement23,24 as well as decrease25 of light absorption capacity of organic dyes occurred by applying Au (and Ag) nanostructures that excite the LSPR. Furthermore, the exciton−plasmon interaction in hybrids consisting of dye molecules and plasmonic nanomaterials led to a significant enhancement of fluorescence.26 Therefore, the interaction between the dyes and the LSPR of the other plasmonic materials are quite attractive. Indeed, these plasmonic nanostructured electrodes has substantially enhanced the photocurrent of the dye molecules positioned near them by a factor of 2203 due to the enhancement of photoexcitation efficiency.1-17 However, all plasmonic electrodes that have been used in these successful experiments are composed of expensive metals such as Au1-10 and Ag.11-17 Because both Au and Ag nanostructures generate strong local electric fields within the visible wavelength region, they are expected to act as excellent visible light-harvesting nanoantennae. Returning to the fundamental motivation for the development of organic photoelectric conversion 4 ACS Paragon Plus Environment

Page 5 of 53

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


systems, however, major advantage is the low cost of organic molecules. It is therefore essential that the other components of light-harvesting nanostructured electrodes used in these systems should be inexpensive. Thus, the achievement of photocurrent enhancement using “inexpensive plasmonic materials” is an important objective. Cu is significantly less expensive than Au and Ag, and is easily recycled. Furthermore, the LSPR of typical spherical Cu nanoparticles occurs within the visible region.27-30 Despite these attractive properties, a highly efficient organic photoelectric conversion system utilizing plasmonic Cu nanostructured electrodes has never been developed. We suggest that two critical problems need to be solved for the development of Cu-based LSPR electrodes to be successfully developed. First, Cu LSPR, which generally occurs around 570 nm,27-30 suffers from substantial damping, because the imaginary part (2) of the Cu dielectric function ( = 1 + i2 = (n + ik)2, 𝑖 = √−1) is fairly large below ca. 590 nm due to the strong electronic interband transition from the valence band to the Fermi level.31-33 Therefore, efficient utilization of the Cu nanoantenna effect below ca. 590 nm would not be allowed. Second, Cu is easily and rapidly oxidized in the atmosphere, which results in the formation of copper oxide phases on the surface. Cuprous oxide (Cu2O) is a p-type semiconductor with a direct bandgap of ca. 2.0 eV, which is produced primarily by the spontaneous oxidation at room temperature,34 leading to broadening and damping of the Cu LSPR band.35 In addition, Cu2O has been found to act as a photoexcited species under the visible light irradiation, resulting in photocurrent generation and 5 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 53

obscuring the role of organic dyes.36 Therefore, for organic photoelectric conversion systems based on dye molecule/nanostructured Cu electrodes, formation of the interfacial Cu2O would not only hinder the effect of Cu LSPR on the photoelectrochemical responses of the molecules, but would also change their electron transfer mechanisms. Recently, we demonstrated that the LSPR excited on regular plasmonic Cu arrays led to significant enhancement of the fluorescence signals from fluorophores located close to them.31 From this result, it is expected that a plasmonic Cu electrode could act as an excellent light-harvesting nanoantenna in the visible region. Herein, we demonstrate effective improvement of photoelectric conversion efficiency of a porphyrin molecule positioned on a plasmonic Cu electrode surfaces due to the interaction with the Cu LSPR, as compared with that on a Cu planar electrode with mirror surfaces. The achievements in this study are as follows. Effective suppression of the spontaneous oxidation of plasmonic Cu electrodes by protecting the surfaces with oxidation-suppressing ultrathin layers has allowed us to accurately evaluate the effect of the Cu nanoantenna on the molecular photocurrent. Based on this ability, we have succeeded in enhancing up to ca. 59-fold the photocurrent from porphyrin as a dye molecule by tuning the generating wavelength of the Cu LSPR to wavelengths at which the 2 value is comparatively small. This enhancement factor compared well with those obtained utilizing plasmonic Au and Ag electrodes in some previous reports.2,7,16

2. RESULTS AND DISCUSSION 6 ACS Paragon Plus Environment

Page 7 of 53

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2.1. Morphological Characterization of Regular Arrayed Cu Electrodes. The regular Cu arrays (Cu half-shell arrays, denoted hereafter as CuHS(d) where d is the silica particle diameter / nm), which acted as the light-harvesting electrodes in this study (Figure 1(A)), were fabricated by a modification of the methods in our previous report.31 Full details of the experimental protocol are described in EXPERIMENTAL SECTION. Typically, two-dimensional (2D) silica colloidal crystals were deposited on transparent indium-tin-oxide (ITO) electrode surfaces from colloidal butanol solutions of silica particles with diameters of 160 ± 9, 260 ± 12, and 330 ± 10 nm (mean ± SD) by a self-assembly method.37 As shown in Figure S1 (Supporting Information), the extinction spectra of the 2D colloidal crystals with diameters of 260 and 330 nm showed a clear diffraction peak (stop band) at 363 and 421 nm, respectively, which suggests the formation of highquality subwavelength structural periodicity. The crystals of the 160 nm nanospheres showed no clear peaks within the wavelength region scanned; according to Bragg’s law, their diffraction is expected to appear at ca. 180 nm. Therefore, this band may occur beyond the measured spectral region.38 After the chemisorption of 3-mercaptopropyltriethoxysilane (MPTS) onto the crystal surfaces to strengthen the interfacial adhesion between Cu and the silica surfaces, Cu was thermally deposited onto the upper hemispheres of the 2D crystals.39 The CuHS arrays with 50-nm thickness of Cu, which were fabricated in our previous study, did not exhibit electrical conductivity.31 However, we confirmed that by increasing the thickness of depositing Cu to one half of the underlying silica diameters (CuHS with thicknesses of 80, 130, and 165 nm for silica with diameters of 160, 260, and 330 nm, 7 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 53

respectively), the arrays gained sufficient electrical conductivity to enable their use as an electrode. SEM images of the CuHS array electrodes are shown in Figure 1(B). For all of the electrodes, the formation of high-quality CuHS arrays was confirmed by the presence of smooth Cu films on the upper hemispheres of the 2D crystals over the entire surface.

2.2. Oxidation-Suppressing Ability of Ultrathin Linking Layers on Regular Cu Array Electrodes. Metallic Cu tends to be spontaneously oxidized upon exposure to air,34,35,40 resulting in the rapid formation of Cu2O layers. Recently, we have succeeded in effectively suppressing the production of Cu2O by coating the arrayed surfaces with stacked ultrathin self-assembled monolayers (SAMs) of 16mercaptohexadecanoic acid (MHA) and titanium oxide (Ti(O)).31 However, even a small amount of Cu2O formed on the Cu electrode surfaces largely affected the photoelectrochemical response of the fabricated photoelectric conversion systems. To further suppress surface oxidation, the spontaneouslygenerated Cu2O was removed by treating the CuHS arrays with glacial acetic acid,35 followed by covering the surfaces with ultrathin layers of MHA SAMs, Ti(O), poly(vinyl alcohol) (PVA), and Ti(O) by self-assembly and sequential surface sol-gel processes41 (Steps 1 and 2 in Scheme 1, detailed procedures are described in EXPERIMENTAL SECTION). To verify each step in the modification process onto the CuHS(260) surfaces, X-ray photoelectron spectroscopy (XPS) measurements were conducted using a monochromatized Mg K (1253.6 eV) Xray source (Figure 2). In the survey spectra (Figure 2(A)) a Cu 2p doublet (Cu2p1/2 and Cu2p3/2) at 953 8 ACS Paragon Plus Environment

Page 9 of 53

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


and 933 eV, a Cu 3s peak at 122 eV, a Cu 3p peak at 76 eV, and an Auger Cu LMM triplet at 335, 415, and 486 eV were observed. After modification with the MHA SAMs a peak consisting of a spin-orbit doublet (S2p3/2 and S2p1/2), which is attributed to the formation of copper thiolate, was observed around 162 eV in the S2p narrow spectrum as shown in Figure 2(B).42 Detection of this peak indicates the formation of MHA SAMs on the Cu surfaces. Also, the O1s narrow spectrum ((a) in Figure 2(C)) obtained from the same surfaces showed a peak at 532.5 eV assigned to the carbonyl and hydroxyl oxygens of the carboxyl groups, which are present on the outer surfaces of MHA SAMs.43,44 The next modification with the Ti(O) resulted in the appearance of a new O1s peak at 530.4 eV along with the peak from the carboxyl groups ((b) in Figure 2(C)), indicating the formation of amorphous titanium oxide.45 In addition, peaks assigned to Ti2p3/2 (459.0 eV) and Ti2p1/2 (464.6 eV) were detected on the same surface (Figure 2(D)), indicating the formation of a Ti(O) linking layer on the MHA SAMs. Further modification with the PVA resulted in a detectable increase in the area ratio of the O1s peak at the higher binding energy (532.5 eV) to that at the lower binding energy (530.4 eV) from 0.97 to 3.71 ((c) in Figure 2(C)). This result supports the modification of PVA layers onto the Ti(O) surfaces, as the O1s signal of PVA hydroxyl groups appears around 532 eV, which overlaps with the higher binding energy peak.46 Finally, the O1s signal ratio decreased to 1.01 after the second modification with Ti(O), suggesting the formation of Ti(O) layers on the PVA ((d) in Figure 2(C)). The thickness of the Ti(O)/PVA/Ti(O) layers was estimated to be approximately 1.9 nm (see Supporting Information and Figure S2). Since the thickness of MHA SAMs has previously been reported to be between 2.0 and 2.5 9 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 53

nm,47,48 the total thickness of the oxidation-suppressing layers is estimated to be between 3.9 and 4.4 nm. To evaluate the effectiveness of this modification in suppressing the spontaneous surface oxidation of Cu, we monitored the temporal changes of the Cu LMM Auger narrow-band spectra in air. Figure 3(A) shows the spectral changes with time for the CuHS(260) array electrode with (a) and without (i.e. the bare electrode) (b) the suppressing layers. Figure 3(B) shows the time course of the intensity ratio of the peaks from Cu2O (337.3 eV) to that from Cu (334.9 eV) as calculated from the spectra in Figure 3(A). For the bare arrayed electrode, the intensity ratio after treatment with glacial acetic acid increased rapidly from 0.43 to 1.03 within 1 h and then reached a steady-state value. On the other hand, the ratio for the electrode coated with molecular layers increased slowly from 0.40 and reached only 0.66 even after 3 h, which remained at this low value afterward. Previous reports have demonstrated that Cu oxidation can be suppressed by coating Cu surfaces with SAMs of long-chain alkanethiols.42,49 Also, it was recently reported that PVA ultrathin films prepared by spin-coating effectively protect Cu nanoparticles from surface oxidation.50 Although it appears that the peak at 337.3 eV slightly increased after the acetic acid treatment, the photocurrent possibly arising from Cu2O is nearly completely suppressed by the protective layers as described in the next section.

2.3. Effect of Oxidation Suppression on Photoelectrochemical Responses from Cu Arrayed Electrode. 10 ACS Paragon Plus Environment

Page 11 of 53

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


To verify suppression of the inadvertent photocurrent, which is attributed to the photoexcitation of Cu2O, we investigated photocurrent generation from the CuHS(260) array electrode with and without (i.e. the bare electrode) the oxidation-suppressing layers. The photocurrent measurements were performed under the same conditions as that of the porphyrin-immobilized electrodes described below. Typically, the experiments were carried out in a degassed aqueous 0.1 M NaClO4 solution containing 5 mM methyl viologen dichloride as an electron acceptor, using a three-electrode photoelectrochemical cell with the array electrode as working electrode. Ag|AgCl (sat. KCl aq.) and platinum wire were used as the reference and the counter electrodes, respectively. Monochromatized light from a Xe lamp was used as a light source, and the generated photocurrents were detected using a potentiostat. Figure 4(A) shows the external quantum efficiencies (EQEs) of the photocurrents as a function of excitation wavelength (E = 0 V vs. Ag|AgCl) from the array electrode with and without the oxidation-suppressing layers. For the bare electrode, a large anodic photocurrent was observed over a range of 400-550 nm in the visible region ((a) in Figure 4(A)). The wavelength dependence of the EQEs agreed well with the spectral shape of the photocurrent generated by the photoexcitation of Cu2O in a previous report,36 which means that the photocurrent from the bare electrodes can be attributed to the photoexcitation of spontaneously-produced Cu2O. It seemed that the photoexcited electrons generated in the Cu2O phases at the outer surface transferred into the adjacent internal Cu phases, which is capable of acting as a good electron acceptor (inset in Figure 4(A)).51 On the other hand, it should be noted that the electrode modified with the oxidation-suppressing layers showed only negligible photocurrent ((b) in Figure 11 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 53

4(A)). Two scenarios are possible to rationalize the disappearance of photoresponses: (i)

The layers, which were inserted at the electrode (containing small amount of the Cu2O phases) / electrolyte solution interfaces, effectively inhibited the photoinduced electron transfer across the interface.

(ii)

The production of Cu2O acted as a photoexcited species was substantially suppressed by covering with the oxidation-suppressing layers.

We note that photocurrent attributed to the electron transfer from the Cu electrode to porphyrin across the interfaces having the ultrathin layers was clearly observed (vide infra). Therefore, the former possibility (i) can be eliminated. Furthermore, as shown in the Cu LMM Auger narrow spectra (Figure 4(B)), the intensity ratio of the peak from Cu2O to that from Cu was almost unchanged at a low level before (0.39, (a)) and after (0.41, (b)) the photocurrent measurement in an aqueous electrolyte solution, strongly supporting the latter possibility (ii). We concluded that while these ultrathin layers effectively suppress the spontaneous air oxidation of Cu potentially leading to inadvertent photocurrent arising from the resulting Cu2O, the ultrathin layers did not hinder the photoinduced electron transfer between the Cu electrode and porphyrin molecules across themselves. Thus, the modification by the layers allows us to accurately evaluate the effect of Cu nanoantennae on the photoelectrochemical properties of nearby dye molecules.

2.4. LSPR-Generating Wavelength of CuHS Array Electrodes. 12 ACS Paragon Plus Environment

Page 13 of 53

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5(A) shows reflectance spectra in air of the CuHS array electrodes that have been modified with 5,10,15,20-tetrakis(4-carboxyphenyl)porphyrin (TCPP) as a dye molecule via surface attachment to the oxidation-suppressing linking layers (denoted as TCPP/CuHS(d), see Step 3 in Scheme 1 and EXPERIMENTAL SECTION). The spectral shapes of each sample electrode before and after the immobilization of TCPP were nearly the same (Figure S3). A significant decrease in reflection (reflection dip), which is observed at 570, 613, and 642 nm for the TCPP/CuHS(160, 260, and 330) electrodes, respectively (Table 1), suggests the generation of strong local electric fields due to the excitation of the Cu LSPR in the interstitial region between two adjacent silica particles (vide infra).31,52,53 The wavelength that corresponds to the Cu LSPR depends on the diameter of the underlying silica particles;31 the Cu LSPR red-shifted with increasing silica diameter. The dip obtained in the TCPP/CuHS(160) is significantly shallower than those observed for the TCPP/CuHS(260 and 330) as shown in their reflection minimum (Figure 5(A) and Table 1). This can be explained by the effect of the electronic interband transitions of Cu as described previously.31-33,54 It is known that the interband transitions of Cu typically occur at wavelengths below ca. 590 nm, which corresponds to the region of large 2 in the dielectric function of Cu, as shown by the purple-shaded area in Figure 6(A). Because the LSPR dip from the TCPP/CuHS(160) electrode expressed at the wavelengths where 2 is large (Table 1, 2.82 from Rakic’s data),55 the LSPR was effectively damped through dephasing of the optical polarization associated with the electron oscillation.31-33 On the other hand, LSPR from the TCPP/CuHS(260 and 330) electrodes generated at wavelengths 13 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 53

where 2 are smaller (2.50 and 2.38, respectively, as shown in Figure 6(A)) showed deeper LSPR dips (small reflection minimum), which means that the efficient LSPR excitation occurred (see Figure 5(A) and Table 1). Furthermore, the LSPR dip from the TCPP/CuHS(330) was moderately deeper than that from the TCPP/CuHS(260). This may also be due to the slight difference in the 2 values at these LSPR wavelengths. To determine the LSPR wavelengths under aqueous conditions, in which photocurrents were measured, we measured the reflectance spectra of the electrodes in an aqueous solution containing 0.1 M NaClO4, corresponding to the photocurrent measurement condition. As shown in Figure 5(B) and Table 1, the LSPR dip positions in the higher refractive index (RI) media (water: n = 1.33) for the TCPP/CuHS(160, 260, and 330) electrodes red-shifted by 26, 90, and 120 nm, compared to those in the air (air: n = 1), resulting in the expression of the dip at 596, 703, and 762 nm, respectively (Table 1). It was therefore found that the LSPR of Cu, like those of Au and Ag which have been utilized as the plasmonic RI sensing materials,56,57 is also sensitive to the change in the surrounding RI.58 Despite the fact that all of the electrodes experienced the same change in RI, magnitudes of the shifts of the LSPR dip positions were different in the order of TCPP/CuHS(330) > TCPP/CuHS(260) > TCPP/CuHS(160). These results can be qualitatively explained by the dependence of the RI sensitivity (SQS) on the wavelength. The SQS within the quasi-static framework is given by 𝑆QS

∆𝜆0 2𝜀1 𝜕𝜀1 = = ( | ) ∆𝑛 𝑛 𝜕𝜆 𝜆0

−1

(1)

where1, n, λ, and λ0, are, respectively, the real part of the dielectric function of the metal, the refractive 14 ACS Paragon Plus Environment

Page 15 of 53

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


index of the medium, the wavelength of the incident light, and the LSPR peak wavelength.59,60 As shown in Figure 6(B), the value of 1 of the Cu at these wavelengths are significantly different in the order of 1(642 nm) = -12.99 < 1(613 nm) = -11.31 < 1(570 nm) = -8.50, while ∂1/∂λ in the wavelength region is nearly constant (∂1/∂λ = -0.0591 to -0.0651). Therefore, the difference in the magnitude of the red-shift is mainly attributed to the difference in 1’s at the respective LSPR wavelength. It is also noted that, although the LSPR dip from the TCPP/CuHS(160) electrode in the solution became somewhat deeper with the red-shift of the LSPR, it was still definitely shallower than those from the TCPP/CuHS(260 and 330) (Figure 5(B), Table 1). This is due to partial damping of the LSPR by the interband transition because 2 at 596 nm is still moderately large (2.60), as shown in Figure 6(A). On the other hand, the TCPP/CuHS(260 and 330) arrayed electrodes exhibited the rather deep dips in the solution helped by the shift to the wavelengths with the quite low 2 values (2.28 and 2.31, respectively), suggesting that the efficient excitation of Cu LSPR is possible without interruption by the interband transition.

2.5. Effect of Damped Cu Nanoantenna on Porphyrin Photocurrent. To evaluate the effects of Cu LSPR on the photoelectrochemical properties of TCPP, we measured wavelength dependences of EQE for the TCPP/CuHS arrayed electrode as a working electrode in the Q-band region, which corresponds to the wavelengths of LSPR excitation of the present CuHS. We 15 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 53

also measured the photocurrent from TCPP immobilized on a planar Cu electrode with a mirror surface (denoted as TCPP/CuP) in the same manner. This acted as a control electrode because it did not generate any Cu LSPR. The detailed conditions for the photocurrent measurements were the same as those described above and the details are given in EXPERIMENTAL SECTION. It is expected that CuHSs accommodate more TCPP molecules than CuP because the surface area of the ideal semispherical arrays is ca. 1.6 times larger than that of flat surface. The average coverage of TCPP on the TCPP/CuHS(160, 260, and 330) were found to be 2.5-3.8 times larger than that on the TCPP/CuP (8.0 × 10-11, 6.6 × 10-11, 5.2 × 10-11, and 2.1 × 10-11 mol/cm2 with regard to the nominal flat area for TCPP/CuHS(160, 260, and 330) and TCPP/CuP, respectively, see EXPERIMENTAL SECTION and Table S1 for details. The standard deviations were below 15 %). These areas correspond to 3.3, 4.0, 5.2, and 7.9 nm2, respectively, per TCPP molecule. Considering that the occupied areas by a closepacked single TCPP molecule flat-lying on a surface are expected to be ~1.5 nm2,61 submonolayer coverages were attained by the present procedure. The difference in the number of TCPP molecules immobilized on different Cu electrodes was corrected in the calculation of photocurrent efficiencies. The wavelength dependences of EQEs obtained from the TCPP/CuHS(160) electrode, which showed the partly damped LSPR dip at the wavelengths where the 2 is moderately large, and TCPP/CuP electrode are shown in Figure 7(A). EQEs before correction with the number of porphyrin are shown in Figure S4. We measured the oxidation-suppressing layers-immobilized CuHS (and CuP) electrodes before the immobilization of TCPP. Only featureless weak cathodic photocurrents over a 16 ACS Paragon Plus Environment

Page 17 of 53

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


scanned wavelength region were obtained for all of these electrodes (Figure S5). These background currents were subtracted from the photocurrents of TCPP-immobilized CuHS (and CuP) electrodes at each wavelength, extracting only the TCPP-derived photocurrent. For both of the electrodes, the photocurrents were observed in the cathodic direction. Also, the spectral shapes of the EQEs of both electrodes were correlated well with the absorption spectrum of the ethanol solution of TCPP (Figure 7(B)). These results suggest that the photoinduced electron transfer occurred from the photoexcited states of TCPP to methylviologen dissolved in the electrolyte solution. The difference from the flowing direction of the photocurrent from the Cu2O is most likely due to the spatial separation between the porphyrin and Cu electrodes. While the Cu2O makes direct contact to the Cu electrode, the TCPP molecules are separated from the Cu surfaces by the oxidation-suppressing layer with the thickness of ca. 4 nm. Therefore, the methylviologen, which is directly accessible to the TCPP molecules, may preferentially act as an acceptor from the excited TCPP. Furthermore, the hybrids consisting of the TCPP and the CuHS(160) without the oxidation-suppressing layer showed only large anodic photocurrent derived from Cu2O and no porphyrin-derived current was detected (Figure S6), which demonstrates the beneficial effect of oxidation-suppressing layer. In order to evaluate the effect of Cu island structures, which were formed on the ITO electrodes during Cu deposition, on the photocurrent, we measured the photocurrent of the hybrids consisting of TCPP and Cu island structures. These electrodes showed significantly weaker photocurrents than TCPP/CuHS electrodes (Figure S7). It is therefore concluded that the photocurrents obtained from TCPP/CuHS were mainly derived from the 17 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 53

photoexcited TCPP molecules immobilized on the regions above the Cu-covered silica spheres. The EQEs from TCPP/CuHS(160) arrayed electrode were moderately enhanced, as compared with those from TCPP/CuP. As described above, submonolayer coverages of TCPP were attained by the present procedure. In addition, as shown in Figure S8, the fluorescence excitation spectra of all the TCPP/CuHS showed a peak at 420 nm, which corresponds to the Soret band. Peaks corresponding to the Q bands also appeared at the same wavelengths for all the electrodes. These result exclude the possibility of formation of extended H- or J-aggregates, which would result in blue-shifted or redshifted absorption, respectively, and suggest that porphyrin molecules are in the same microenvironment in all the electrodes. In order to quantitatively evaluate this result, the enhancement factors (EFs) of the EQEs for TCPP/CuHS against TCPP/CuP were calculated by dividing the EQE values obtained from the former by that from the latter at each wavelength. As shown in Figure 7(C), it was found that the EF increased to ca. 10 at the 670 nm above the strong interband transition region (> 590 nm). As described in the following section, the maximum EFs of the TCPP/CuHS(260) and TCPP/CuHS(330) were significantly higher than that of the TCPP/CuHS(160). This is most likely due to the combined effects of the ineffective overlapping between the LSPR wavelengths and the Q-band and the effective LSPR damping by the large 2. Because it is well known that Cu has a dielectric function which is subject to the interband transitions extending to a lower energy compared to those of Au (see the wavelength dependence of 2 of Cu dielectric function shown in (c) in Figure 7(C)).6264

The photocurrent was effectively enhanced at the longer wavelengths than the center wavelength of 18 ACS Paragon Plus Environment

Page 19 of 53

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


LSPR where the 2 value is smaller as shown in Figure 6(A).

2.6. Effective Enhancement of Porphyrin Photocurrent by Cu Nanoantenna. The wavelength dependences of EQEs obtained from the TCPP/CuHS(260 and 330), which showed the LSPRs in a red region where the 2 is comparatively low, are shown in Figure 8(A) (EQEs before correction with the number of porphyrin molecules are shown in Figure S9). The EFs of these EQEs against those obtained from the TCPP/CuP are also shown in Figure 8(B). The EQEs of both of the electrodes were significantly enhanced. Particularly, the maximum EF of the TCPP/CuHS(260) achieved up to ca. 59 at 670 nm, which was substantially larger than that of the TCPP/CuHS(330) (ca. 20). This difference can be explained by the LSPR generating wavelength. Typically, the LSPR dip from the TCPP/CuHS(260) is generated with the minimum reflectance at 703 nm (Figure 8(B)), which is well-overlapped with the wavelengths corresponding to the TCPP Q-band (500-670 nm). On the other hand, as shown in Figure 8(B), the LSPR dip from the TCPP/CuHS(330) was generated with the minimum reflectance at 762 nm (see Fig 5(B)), which is far longer than the Q-band wavelengths. As a result, the nanoantenna effect of the TCPP/CuHS(330) could not effectively enhance the photoexcitation efficiency of the TCPP. It should also be noted that the EFs from the TCPP/CuHS(260) electrode was much larger than those from the TCPP/CuHS(160) particularly at the region above the strong interband transition, where 2 is comparatively small. These interesting results can be rationalized by referring to Figure 5. Typically, the light-harvesting nanoantenna effect of the 19 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 53

TCPP/CuHS(160) was substantially reduced due to the partial damping of the LSPR, a phenomenon occurring in the purple-shaded region in Figure 5, leading to an inefficient enhancement of the photocurrent of the nearby porphyrin molecules. Therefore, the stronger nanoantenna effect led to greater improvement of photocurrent in the region of the comparatively small 2. We previously reported that the Au nanoantenna effect of the Au half-shell arrayed electrodes enhanced the porphyrin photocurrent up to ca. 15 times in the Q-band region although the immobilization process of the porphyrin was different from that in this study. 4 Therefore, this study suggests that the nanoantenna effect from Cu LSPR compares well with that produced by more expensive Au LSPR for improving the performance of organic photoelectric conversion systems. To further clarify the effect of the Cu LSPR, we estimated the local electric field distributions generated at the CuHS(160, 260 and 330) arrayed electrodes in an aqueous medium under the irradiation of 670 nm incident light using finite difference time domain (FDTD) calculations (Figure 9). The detailed calculation models are shown in Figure 9(A) and the validity of model is discussed in Supporting Information (Figure S10). As shown in Figure 9(B), strong local electromagnetic fields are distributed in the gap region between the neighboring silica particles for all of the CuHS electrodes. Furthermore, the local field intensity was in the order of CuHS(260) > CuHS(330) > CuHS(160). The maximum field strength (|Emax/E0|2, normalized by the incident light power) was calculated to be 3.20×104. It should be noted from Figure 9(C) that the silica diameter dependence of the local field intensity were qualitatively correlated well with that of the EF of the porphyrin photocurrent at 670 20 ACS Paragon Plus Environment

Page 21 of 53

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


nm. These results strongly suggest that the efficient enhancement of the porphyrin photocurrent was induced by the effect of the plasmonic Cu nanoantenna. It is worth mentioning that as only a part of molecules are placed in the strong electric field region and thus are influenced by the strong electric field, the enhancement of the photocurrent by those particular molecules should be significantly higher than that experimentally observed, average enhancement. The local electromagnetic field intensity of Cu LSPR at longer wavelengths is somewhat larger than that of Au LSPR in the FDTD calculation (Figure S11). The development of higher-performance plasmonic photoelectric conversion systems will be achieved by increasing the interaction between the Cu nanoantenna and the TCPP based on the utilization of further useful Cu nanostructured electrodes. Porphyrins are attractive dye molecules for organic photoelectric conversion systems because of wide absorption spectra, favorable excited-state properties, and synthetic availability, etc. and the structure is similar to that of chlorophyll in natural photosynthesis.65,66,67 However, the absorption coefficient of the Q-band (~104 M-1cm-1) that covers a wide visible to near-infrared range is significantly smaller than that of the Soret band (~105 M-1cm-1) in a near-UV range. While it has been already reported that photocurrents from porphyrin are effectively enhanced when immobilized on nanostructures consisting of expensive plasmonic metal species such as Au and Ag,1-17 this study has demonstrated that the LSPR of Cu, which is much less expensive and is capable of acting as a superior light-harvesting nanoantenna in a red range of the spectrum (> 600 nm), was very useful for the development of efficient photoelectric conversion systems using porphyrin dyes. 21 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 53

3. CONCLUSIONS We have demonstrated that the photocurrent from porphyrin placed close to the plasmonic Cu arrayed electrode was enhanced by the effect of the inexpensive plasmonic nanoantennae. In particular, we found that the nanoantenna effect can be maximally utilized at the wavelengths where the value of 2 of the Cu dielectric function is quite low. The efficiencies of photocurrent enhancement due to the nanoantenna effect of the present Cu arrays compared well with those by the expensive Au and Ag nanostructures. Since the imaginary part 2 of the dielectric function of Cu is quite low approximately from 630 nm to a near infrared region, the Cu nanoantennae will be useful for highly-efficient visible and/or near-infrared light-driven organic photoelectric conversion systems using not only organic dyes but a wide range of photofunctional materials such as carbon nanotubes, quantum dots, and so on. Researches along this line is currently underway in our laboratory. Also, recently, in the development of high-performance next generation photovoltaics such as perovskite solar cells and quantum dot solar cells, beneficial use of near-infrared (NIR) light (above ca. 750 nm) energy, which is heavily contained in the solar spectrum, is strongly desired. The Cu nanoantenna effect will be effectively utilized for the development of plasmonic NIR photovoltaics.

4. EXPERIMENTAL SECTION 4.1. Materials. Milli-Q-grade water (resistivity: 18.2 MΩ·cm) was used to prepare all aqueous solutions. NH3 aq. 22 ACS Paragon Plus Environment

Page 23 of 53

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(28%), tetraethyl orthosilicate (TEOS), 1-butanol, 2-propanol, and H2O2 (34.5%) were obtained from Kanto Chemical, Japan. Absolute ethanol (99.5 vol %), TCPP, and titanium butoxide (Ti(OBu)4) were obtained from Wako Pure Chemical, Japan. MPTS was obtained from Tokyo Kasei, Japan, and MHA and PVA (M. N. 31000-50000; 87-89% hydrolyzed) were obtained from Sigma-Aldrich. All materials were used without further purification. ITO electrodes (thickness of ITO: ∼100 nm, ≥ 30 Ω) was obtained from Sanyo Vacuum Industries, Japan. 4.2. Fabrication of Cu Regular Arrayed Electrodes. The colloidal solutions of silica particles with diameters of 160 ± 9, 260 ± 12 and 330 ± 10 nm (mean ± SD) were synthesized according to a modified version of previously reported procedure.68 Typically, NH3 aq. (28%) (2.2, 8.2, and 8.4 mL, respectively) was added to absolute ethanol (26.5, 19.5, and 19.5 mL, respectively) and the mixture was stirred for 10 min at room temperature. TEOS (0.7, 0.62, and 0.90 mL, respectively) was quickly added to the solutions, and the mixtures were stirred for 3 h. The resultant colloidal solutions of silica particles were centrifuged at 9000 rpm for 15 min and then redispersed twice in an equivalent amount of ethanol. They were then centrifuged at 9000 rpm for 10 min and redispersed twice in an equivalent amount of 1-butanol. Plasmonic CuHS(d) arrayed electrodes were fabricated using a modified version of previously reported procedures.31 Firstly, the two-dimensional colloidal crystals consisting of silica particles were fabricated on ITO electrodes. The ITO electrodes (2.0 cm × 1.5 cm) were first treated in mixed solution of NH3 aq./H2O2 aq. (1:1 v/v) at 100 °C for 3 h and then washed thoroughly with Milli-Q water to 23 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 53

produce a hydrophilic surface. A small amount of each colloidal 1-butanol solution of the silica particles was added dropwise to water in a Petri dish, resulting in the formation of 2D colloidal crystals of the silica particles.37 After transferring the crystals onto the ITO surfaces, the samples were annealed at 500 °C for 1 h to physically strengthen the colloidal crystals. Next, the Cu thin films were formed on the colloidal crystals. Typically, a mixed solution containing MPTS (1 mL) and Milli-Q water (1 mL) in 2-propanol (40 mL) was refluxed for 1 h at 100 °C.39 After cooling the solution, the substrate containing the colloidal crystals was immersed into the solution for 10 min, followed by thorough washing with ethanol and drying in air at 75 °C to expose thiol groups onto the silica surfaces as a binder between the surfaces and Cu films. Finally, to fabricate the CuHS arrays, Cu (thickness of 80, 130, and 165 nm for silica with diameters of 160, 260, and 330 nm, respectively) was thermally deposited under a high vacuum (6.0 × 10-7 Torr) onto the surface of the colloidal crystals. A planar Cu electrode (CuP) was prepared by thermal deposition of Cu (thickness: 100 nm) onto the MPTS-immobilized ITO electrodes.

4.3. Immobilization of TCPP on CuHS. A monolayer of TCPP was immobilized on the CuHS arrays as follows (Scheme 1).41,69 Typically, the as-prepared CuHS arrayed electrodes were immersed for 1 h in a 10 mM ethanol solution of MHA, which was degassed by bubbling N2 through it, followed by washing with ethanol and drying under a stream of N2. This process provided the formation of MHA SAMs on Cu surfaces via S-metal strong 24 ACS Paragon Plus Environment

Page 25 of 53

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


bonding. The electrode was next immersed into a 0.1 M degassed toluene/ absolute ethanol (1/1 v/v) solution of Ti(OBu)4 for 3 min to achieve chemical bonding between Ti compound (Ti(O)) and the carboxyl groups on the MHA SAMs. After washing with anhydrous ethanol, the substrate was immersed in H2O for 1 min to generate surface hydroxyl groups and then dried under a stream of N2. It was then immersed in a 0.5 wt% aqueous solution of PVA for 10 min, followed by washing with H2O and drying under a stream of N2 to modify the PVA onto the Ti(O) surfaces. After modifying Ti(O) thin layers onto the PVA surfaces again by immersing the electrode into a Ti(OBu)4 solution, the electrode was immersed in degassed ethanol solution of 1 M TCPP for 10 min, washed with ethanol, and then dried under a stream of N2 to produce the Cu surfaces terminated by a layer of TCPP (TCPP/CuHS(d)). As a control electrode, TCPP was immobilized on the CuP electrode in the same manner (TCPP/CuP).

4.4. Quantitation of the Number of TCPP Molecules Immobilized on the CuHS and CuP. The TCPP/CuHS(d) (or CuP) electrodes were immersed in 2 mL of a 0.2 M aqueous solution of NaOH for 10 min to dissolve the surface TCPP into the solution. After this treatment, it was confirmed that nearly all of TCPP molecules were dissolved into the solution by measuring the absorption spectra of the substrates. The absorption spectra of the solution containing TCPP were measured and the coverage of TCPP on each sample was calculated from the absorbance at Soret band (molar absorption coefficient: 304600 M–1 cm–1, max = 415 nm, see Figure S12). 25 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 53

4.5. Measurements. Reflectance spectra were measured using an ALS SEC2000 UV-Vis multichannel spectrometer. Scanning electron microscopy (SEM) observations were carried out using a HITACHI S-4500 microscope. Surface analysis of Cu films on the silica colloidal crystals that were deposited by thermal evaporation, was performed using X-ray photoelectron spectroscopy (XPS) using an ESCA-3400 electron spectrometer (Shimadzu Co., Japan), using a base pressure of

Efficient Photocurrent Enhancement from Porphyrin Molecules on

Recommend Documents