# Fileset

[adfm76809-sup-0001-suppmat.pdf](https://mdr.nims.go.jp/filesets/8d0cc285-d581-47ea-b40d-c39adce91809/download)

## Creator

Tai Kobayashi, Ryosuke Nishikubo, Yusuke Tomiyori, Fumitaka Ishiwari, Daisuke Asakura, Eiji Hosono, Miho Kitamura, [Hisao Kiuchi](https://orcid.org/0000-0001-9139-8218), Yoshihisa Harada, Akinori Saeki

## Rights

[Creative Commons BY-NC Attribution-NonCommercial 4.0 International](https://creativecommons.org/licenses/by-nc/4.0/)

## Other metadata

[Water Adsorption‐Induced Color Sensor: Insight Into the Sensing Mechanism and Interfacial Engineering for Improved Responsiveness](https://mdr.nims.go.jp/datasets/4b04ff93-379a-4178-b03f-3be299e9584c)

## Fulltext

Supporting Information    Water Adsorption-Induced Color Sensor: Insight into the Sensing Mechanism and Interfacial Engineering for Improved Responsiveness  Tai Kobayashi, Ryosuke Nishikubo,* Yusuke Tomiyori, Fumitaka Ishiwari,  Daisuke Asakura, Eiji Hosono, Miho Kitamura, Hisao Kiuchi,* Yoshihisa Harada,* Akinori Saeki *  Mr. T. Kobayashi, Dr. R. Nishikubo, Prof. Dr. F. Ishiwari, Prof. Dr. A. Saeki Department of Applied Chemistry, Graduate School of Engineering, The university of Osaka, 2-1 Yamadaoka, Suita, Osaka 565-0871, Japan E-mail: nishikubo@chem.eng.osaka-u.ac.jp (R.N.), saeki@chem.eng.osaka-u.ac.jp (A.S.)  Dr. R. Nishikubo, Prof. Dr. F. Ishiwari, Prof. Dr. A. Saeki Innovative Catalysis Science Division, Institute for Open and Transdisciplinary Research Initiatives (ICS-OTRI), The University of Osaka, 1-1 Yamadaoka, Suita, Osaka 565-0871, Japan.  Prof. Dr. F. Ishiwari Faculty & Graduate School of Urban Environmental Science, Tokyo Metropolitan University, 1-1 Minami-Osawa, Hachioji-shi, Tokyo 192-0397, Japan.  Dr. Y. Tomiyori, Dr. H. Kiuchi, Prof. Dr. Y. Harada Institute for Solid State Physics (ISSP), The University of Tokyo, Kashiwa, Chiba 277-8581, Japan. E-mail: harada@issp.u-tokyo.ac.jp (Y.H.)  Dr. Y. Tomiyori AGC Inc., Yokohama, Kanagawa, 230-0045, Japan.  Dr. H. Kiuchi Research Center for Energy and Environmental Materials (GREEN), National Institute for Materials Science (NIMS), Tsukuba, Ibaraki 305-0044, Japan. KIUCHI.Hisao@nims.go.jp (H.K.)  Dr. D. Asakura, Dr. E. Hosono Research Institute for Energy Efficient Technologies, National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Ibaraki 305-8585, Japan.  Dr. M. Kitamura NanoTerasu Center, National Institutes for Quantum Science and Technology (QST), Sendai, Miyagi 980-8579, Japan    Table S1. Summary of the performance of color-recognizable single cell photodetectors. Our work shows a fast photovoltage-based response and recovery without external bias or impedance analysis. (ref. 4 is not self-powdered photodtector, and requires an impedance analyzer to obtain wavelength-dependent signals) Photoactive material Color recognition mechanism Sensing color region Response time Recovery time Ref. TiO2/ASIS WDPE a) UV to green 2–5 s 10 s 1 TiO2/ASIS RWDPE b) UV to green 10–20 s 7 s 2 Perovskite Difference of light penetration depth Blue to red 0.57–3.92 s c) 10 s 3 Perovskite Ion and charge migration Blue to red 21.3 ms d) - e) 4 Graphdiyne/ WSe2 Charge carrier trapping Blue to red 1 s Few seconds  (< 10 s) 5 Perovskite Difference of light penetration depth UV to green - - 6 TiO2/ASIS WDPE UV to green 150 ms 3.12 s This work a) Wavelength-dependent photovoltaic effect (WDPE) is proposed to be related to the TiO2 photocatalytic effect under UV irradiation. b) Reversed-WDPE is proposed to be related to carrier trapping under visible light and detrapping by UV. c) Response time is dependent on irradiation wavelength (400 nm: 1.75 s, 500 nm: 0.57 s, 650 nm: 3.92 s). d) An impedance analyzer is required to obtain wavelength-dependent signal. e) Hyphen (-) means that there is no description of response/recovery time in the paper.        Note S1. The thickness of the adsorbed H2O on glass substrate/TiO2 was estimated by analyzing the variation in the O Ka fluorescence intensity. The total O Ka intensity at 550 eV, which is non-resonant region, increased by 4.27% relative to the dry state in Figures S3c,d. This increase is attributed to the fact that the contribution of fluorescence generated from the water layer itself exceeds the attenuation of the signal from the underlying glass substrate/TiO2 caused by the water overlayer. To quantify the water layer thickness (𝑑𝐻2𝑂), we simply employed a three-layer model consisting of the water overlayer (Layer 2), the mp-TiO2 film (Layer 1), and the glass substrate (Layer 0). The theoretical fluorescence intensity 𝐼𝑖  from a layer i with thickness 𝑑𝑖  can be expressed by integrating the X-ray production and attenuation over the depth z: 𝐼𝑖 = ∫ 𝑄𝑖 ∙ exp(−𝜇𝑖∗𝑧)𝑑𝑧𝑑𝑖0 where 𝑄𝑖  represents the effective oxygen fluorescence production factor proportional to the oxygen density and the absorption cross-section at the incident energy (550 eV). 𝜇𝑖∗ represents the effective linear attenuation coefficient considering the experimental geometry, defined as: 𝜇𝑖∗ = 𝜌𝑖 (𝜇𝑖,inc sin 𝜃𝑖𝑛+𝜇𝑖,fluosin 𝜃𝑜𝑢𝑡) Here, 𝜌𝑖  is the effective density of the layer. 𝜇𝑖,inc   and 𝜇𝑖,fluo  denote the mass attenuation coefficients for the incident (550 eV) and fluorescent (525 eV) X-rays, respectively. 𝜃𝑖𝑛 (90°) and 𝜃𝑜𝑢𝑡 (45°) are the incident and take-off angles. The integrated intensities for the water layer ((𝐼𝐻2𝑂), the mp-TiO2 layer (𝐼𝑇𝑖𝑂2), and the glass substrate (𝐼𝑔𝑙𝑎𝑠𝑠 𝑠𝑢𝑏) are derived as: 𝐼𝐻2𝑂 = ∫ 𝑄𝐻2𝑂 ∙ exp(−𝜇𝐻2𝑂∗ 𝑧)𝑑𝑧𝑑𝐻2𝑂0 𝐼𝑇𝑖𝑂2= exp(−𝜇𝐻2𝑂∗ ∙ 𝑑𝐻2𝑂) ∙ ∫ 𝑄𝑇𝑖𝑂2∙ exp(−𝜇𝑇𝑖𝑂2∗ 𝑧)𝑑𝑧𝑑𝑇𝑖𝑂20 𝐼𝑔𝑙𝑎𝑠𝑠 𝑠𝑢𝑏 = exp(−𝜇𝐻2𝑂∗ ∙ 𝑑𝐻2𝑂) ∙ exp(−𝜇𝑇𝑖𝑂2∗ ∙ 𝑑𝑇𝑖𝑂2) ∙ ∫ 𝑄𝑔𝑙𝑎𝑠𝑠 𝑠𝑢𝑏 ∙ exp(−𝜇𝑔𝑙𝑎𝑠𝑠 𝑠𝑢𝑏∗ 𝑧)𝑑𝑧∞0 Here, the parameters used for the calculation are listed in Tables S2 and S3. By fitting the total calculated intensity (𝐼𝑡𝑜𝑡𝑎𝑙 = 𝐼𝐻2𝑂 + 𝐼𝑇𝑖𝑂2+ 𝐼𝑔𝑙𝑎𝑠𝑠 𝑠𝑢𝑏 ) to the experimental intensity ratio, the thickness of the water layer was calculated to be 10.0–12.4 nm.        Table S2. Density, thickness, and porosity of each layer used for the calculation.  PCPDTBT (Layer 4) SbSI:Sb2S3 (Layer 3) H2O (Layer 2) mp-TiO2 (Layer 1) Glass substrate (Layer 0) Density (g cm−3) 1.5 ± 0.1[7] 4.6 ± 0.2 1.0 4.26 2.6 Thickness (nm) 35 ± 10 40 ± 10 - 153 ± 30 ∞ Porosity (%) 0 50 ± 10 0 57a 0 a) Porosity was experimentally obtained from the mass of mp-TiO2 layer (0.10 ± 0.02 mg), thickness (153 ± 30 nm), and substrate size (4 × 0.9 cm2).    Table S3. Mass attenuation coefficient of each element from the X-ray interaction database.[8] Element Mass attenuation coefficient (cm2 g-1) Incident X-ray at 550 eV Fluorescent X-ray at 525 eV  H 49.18 57.29  C 10763 12106 N 15558 17327 O 20708 1199 Si 7851 8823 S 11688 13002 Ti 20011 22125 Sb 24024 3416 I 3679 3973     Figure S1. The wavelength dependent behavior of an SbSI:Sb2S3 photovoltaic (PV) device reported by our previous papers.[1,2] Wavelength-dependence of a) JV characteristics and b) time-dependent photovoltage (TDPV) decay. For TDPV measurement, the light intensity was modulated to output the close value of photocurrent. The corresponding short circuit photocurrent density (Jph0, unit: mA cm−2) and the irradiated photon density (PD, unit: photons cm−2 s−1) are displayed in the graph. c) Transient photovoltage decay and d) extracted charge with different delay time obtained by charge extraction by linearly increasing voltage (CELIV) measurement. For this measurement, continuous UV or visible light was irradiated for 30 s, followed by a rapid integration (~5 s) under 532 nm pulse laser irradiation to obtain TPV or CELIV decay. e) The schematic image of operando electron spin resonance (ESR) measurement and f) the obtained ESR signals. Figures a-d and e-f are extracted from ref S1, S2, respectively. 00.10.20.30.40.5-5 0 5 10 15 20 25 30 35Photovoltage (V)Time (s)515 nm(Jph0 = 4.0, PD = 6.5×1016)375 nm(Jph0 = 5.8, PD = 7.1×1016)-15-10-50510150 0.1 0.2 0.3 0.4 0.5Current density (mAcm-2)Voltage (V)00.240.370.490.721PD375 + PD515PD375a) b)time (ms)ΔVoltage(mV)051015202530354045-20 0 20 40 60 80 100τ515 = 31.5 msτ375 = 14.9 ms02468101214160 20 40 60 80 100Extracted charge (µC)Delay time (µs)ーWithout pre-irradiationー UVー Visc) d)DarkWith UVWithout UVTiO2/SbSI:Sb2S3/PCPDTBT318 320 322 324 326 328H (mT)−0.8−0.4−0.20.00.20.40.8ESR signals (l Mnunit) 0.6−0.6Xe lampUCF  (Only for the measurement without UV)ESRFilm samplee) f)  Figure S2. a) A schematic illustration of sample setting with the designed measurement cell and a picture of the measurement cell and connections. b) The schematic illustration of the detailed structure of the cell. “SDD” means silicon drift detector.   Ar with or w/o H2OSDDSoft X-ray excitationSoft X-ray emissionHumidity sensorGas inGas outSiCmembraneFilm sampleRubber sealVacuum Frangea)b)Measurement cell  Figure S3. O K-edge XAS spectra and the corresponding difference spectra ratio of a,b) a glass substrate, c,d) TiO2, e,f) ASIS, g,h) HTM, and i,j) TiO2/ASIS/HTM. The difference spectra ratio (δ1, δ2 and δ3) are calculated by (S2 – S1)/S2, (S3 – S2)/S2 and (S3 – S1)/S2. Water adsorption onto bare glass was negligible (Figure S3a). It should be noted that normal glass substrates were used for a-g, while a quartz substrate was used for i. This is the reason for wider background spectra of a-h (533–550 eV) compared to that of i (536–550 eV). 525 530 535 540 545 550Photon energy (eV)O K-edge― δ1 = (S2 − S1)/S2― δ2 = (S3 − S2)/S2― δ3 = (S3 − S1)/S20812−12−84−4Intensity difference ratio(%)525 530 535 540 545 550Photon energy (eV)― δ1 = (S2 − S1)/S2― δ2 = (S3 − S2)/S2― δ3 = (S3 − S1)/S2GlassHTM0812−12−84−4 Differential (%)Intensity difference ratio(%)525 530 535 540 545 550Photon energy (eV)― δ1 = (S2 − S1)/S2― δ2 = (S3 − S2)/S2― δ3 = (S3 − S1)/S20812−12−84−4Intensity difference ratio(%)525 530 535 540 545 550Photon energy (eV)GlassTiO2― δ1 = (S2 − S1)/S2― δ2 = (S3 − S2)/S2― δ3 = (S3 − S1)/S20812−12−84−4Intensity difference ratio(%)525 530 535 540 545 550Photon energy (eV)GlassIntensity difference ratio(%)0812−12−8― δ1 = (S2 − S1)/S2― δ2 = (S3 − S2)/S2― δ3 = (S3 − S1)/S24−4525 530 535 540 545 550Photon energy (eV)Intensity (a.u.)― 1. Dry (S1)― 2. Wet (S2)― 3. Dry (S3)GlassHTMO K-edge0525 530 535 540 545 550Photon energy (eV)Intensity (a.u.)O K-edge― 1. Dry (S1)― 2. Wet (S2)― 3. Dry (S3)0525 530 535 540 545 550Photon energy (eV)Intensity (a.u.)― 1. Dry (S1)― 2. Wet (S2)― 3. Dry (S3)O K-edge0525 530 535 540 545 550Photon energy (eV)Intensity (a.u.)O K-edgeGlassTiO2― 1. Dry (S1)― 2. Wet (S2)― 3. Dry (S3)0525 530 535 540 545 550Photon energy (eV)Intensity (a.u.)O K-edgeGlass― 1. Dry (S1)― 2. Wet (S2)― 3. Dry (S3)0a) b)c) d)e) f)g) h)i) j)   Figure S4. a) O K-edge XAS spectra of a tri-layer obtained by repeated measurement under high humidity, and b) the corresponding difference spectrum. The hydrophilization effect was observed when XAS measurements were repeated under high humidity.       Figure S5. a) O K-edge XAS spectra of an ASIS single layer obtained under controlled humidity, b) the corresponding difference spectra ratio (δ1, δ2, δ3), and c) the corrected difference spectra ratio (δ’) are calculated by δ1 – δ3/2 (= 1 – (S1 + S3)/2S2). In δ’ spectra, the hydrophilization effect is corrected by linear interpolation.    a) b)525 530 535 540 545 550Photon energy (eV)Intensity (a.u.)― Wet 1st― Wet 2nd0δ'(%)Photon energy (eV)0.00.51.01.52.02.53.0525 530 535 540 545 550Photon energy (eV)δ’(%)525 530 535 540 545 5506824100H2O gas― δ’ = δ1 – δ3/2525 530 535 540 545 550Photon energy (eV)― δ1 = (S2 − S1)/S2― δ2 = (S3 − S2)/S2― δ3 = (S3 − S1)/S20812−12−84−4Intensity difference ratio(%)525 530 535 540 545 550Photon energy (eV)Intensity (a.u.)― 1. Dry (S1)― 2. Wet (S2)― 3. Dry (S3)O K-edge0b)a) c)   Figure S6. The difference XAS spectra ratio without correction (δ1) of mono-layer films and a tri-layer film. The peaks attributed to gas and liquid-like H2O are marked with blue and red dots, respectively. The pre-edge attributed to H2O-TiO2 interaction is marked with red circles. The spectra of SbSI, HTM, and a tri-layer are shifted along the vertical axis for easier comparison.      Figure S7. a) Ti L3-edge XAS spectra of the TiO2/ASIS/HTM tri-layer film obtained under controlled humidity. The spectra were normalized to the peak top intensity. b) Sb M-edge XAS spectra of the TiO2/ASIS/HTM tri-layer film. The positions of the Sb M5 and M4 peaks were extracted from the data reported by Ye et al.[9]  05101520525 530 535 540 545 550Photon energy (eV)δ1(%)― TiO2/SbSI/HTM― HTM― SbSI― TiO2a) b)― 1. Wet ― 2. DryPhoton energy (eV)Intensity (a.u.)Ti L3-edge456 457 458 459 460 4600Intensity (a.u.)Sb M5Photon energy (eV)Sb M4525 530 535 540 545 5500 Figure S8. An illustration of the device structure used in this study and a normalized external quantum efficiency (EQE) spectrum corresponding to the spectra exhibited in Figure 3.     Figure S9. a) The photoabsorption spectrum and b) the corresponding Urbach tail analysis of a compact-TiO2/mp-TiO2 film. The Urbach energy (EU) was calculated as 650–680 meV.    0.000.200.400.600.801.001.20300 400 500 600 700EQEWavelength (nm)0.00.20.40.60.81.01.2300 400 500 600AbsorptionWavelength (nm)a) b)6.07.08.09.010.011.012.01.0 2.0 3.0 4.0Ln(α)Photon energy (eV)α = α0exphν - EgEU(           ) Figure S10 SCLC curve with and without UV irradiation. The device structure is FTO/TiO2/ASIS/C60/BCP/Au (electron only device). The measurement was performed before and after temporal UV irradiation (385 nm, 75 mW cm−2, ~30 s). After UV irradiation was finished, J-V scan started immediately. The calculated trap density was 1.32×1017 and 1.53×1017 cm−3 for without and with UV irradiation, respectively. The measurements were performed with 0.02 V step, requiring approximately 9 s per scan. Hence, the actual trap density during UV irradiation should be higher than 1.53×1017 cm−3.    Figure S11. Time-dependent photocurrent (TDPC) decay of the devices a) without and b-e) with hydrophilic layer. The irradiated photon density (PD, unit: photons cm−2 s−1) are displayed in the graph. The relatively low photo-current density of PVA- and PSS-included devices are due to insulating property of PVA and PSS.  0.01 0.1 1Current density (mA cm–2)Voltage (V)10−210−1100101102Without irradiationWith UV irradiation0.29 V0.33 V0.00.51.01.52.0-1 0 1 2 3 4 5 6 7 8 9Photocurrent (mA cm-2)Time (s)515 nm(PD = 3.17×1017)(PD = 1.46×1017)375 nmTiO2/PVA/ASIS/HTM01234-1 0 1 2 3 4 5 6 7 8 9Photocurrent (mA cm-2)Time (s)515 nm(PD = 1.17×1016)(PD = 1.46×1017)375 nmTiO2/ASIS/PVA/HTM0246810-1 0 1 2 3 4 5 6 7 8 9Photocurrent (mA cm-2)Time (s)515 nm(PD = 1.17×1017)(PD = 8.49×1016)375 nmControl0.00.51.01.52.02.5-1 0 1 2 3 4 5 6 7 8 9Photocurrent (mA cm-2)Time (s)515 nm(PD = 3.30×1017)(PD = 8.49×1016)375 nmTiO2/PSS/ASIS/HTM012345-1 0 1 2 3 4 5 6 7 8 9Photocurrent (mA cm-2)Time (s)515 nm(PD = 1.36×1017)(PD = 8.49×1016)375 nmTiO2/ASIS/PSS/HTMa) b) c)d) e) Figure S12. Periodic response of control, PVA- and PSS-incorporated devices upon 0.2s UV irradiation.    Figure S13. a, c) Photovoltage change (ΔV) and b, d) photovoltage change ratio (ΔV/V0) upon simultaneous irradiation of different color light with varied intensities. ΔV and V0 are voltage decrease upon short 375 nm irradiation (0.5±0.1 s) and base photovoltage under constant 530 nm irradiation, respectively. The background color is a fitting by radial basis function (RBF) interpolation.  0.00.10.20.30.40.50.60 50 100 150Photovoltage (V)Time (s)0.00.10.20.30.40 50 100 150Time (s)0.00.10.20.30.40.50 50 100 150Time (s)Control TiO2/ASIS/PVA/HTM TiO2/ASIS/PSS/HTMa) b)c) d)Control ControlPSS-incorporated PSS-incorporatedΔV/V0ΔVΔV/V0ΔVReferences [1] R. Nishikubo, S. Li, A. Saeki, Adv. Funct. Mater. 2022, 32, 2201577/1–9. [2] T. Kobayashi, R. Nishikubo, Y. Chen, K. Marumoto, A. Saeki, Adv. Funct. Mater. 2023, 32, 2311794/1–10. [3] L. Min, Y. Zhou, H. Sun, L. Guo, M. Wang, F. Cao, W. Tian, L. Li, Light Sci. & Appl. 2024, 13, 280/1–11. [4] Y. Zhou, H. Sun, L. Guo, S. Yang, L. Min, M. Wang, F. Cao, L. Li, Adv. Mater. 2025, 37, 2502889/1–9. [5] H. Tang, W. Shi, W. Jiang, G. Wang, M. Tang, Z. Cai, R. Li, S. Wu, G. Zhang, J. Li, Chem. Eng. J. 2025, 517, 164215/1–9. [6] C.-Y. Chang, Y.-L. Chin, C.-Y. Chang, J. Holovsky, Adv. Funct. Mater. 2025, 35, 2422732/1–15. [7] C. S. Sarap, Y. Singh, J. M. Lane, N. Rai, Sci. Rep. 2023, 13, 21587. [8] B. L. Henke, E. M. Gullikson, J. C. Davis, Atomic Data and Nuclear Data Tables 1993, 54, 181–342. [9] M. Ye, T. Xu, G. Li, S. Qiao, Y. Takeda, Y. Saitoh, S.-Y. Zhu, M. Nurmamat, K. Sumida, Y. Ishida, S. Shin, A. Kimura, Phys. Rev. B 2019, 99, 144413.