Supporting Information Solar Hydrogen Production by Plasmonic Au ...
Supporting Information ______________________________________________________________________
Solar Hydrogen Production by Plasmonic Au-TiO2 Catalysts: Impact of Synthesis Protocol and TiO2 Phase on Charge Transfer Efficiency and H2 Evolution Rates. Jacqueline B. Priebe, Jörg Radnik, Alastair J. J. Lennox, Marga-Martina Pohl, Michael Karnahl,† Dirk Hollmann, Kathleen Grabow, Ursula Bentrup, Henrik Junge, Matthias Beller, and Angelika Brückner * Leibniz-Institut für Katalyse e.V. an der Universität Rostock, Albert-Einstein-Str. 29a, 18059 Rostock, Germany, [email protected]
Table of Content
page
Catalyst preparation “Bro”
2
Figure S1
Radiation spectra of the light source
2
Figure S2
Photocatalytic H2 production curves
3
Table S1
H2 evolutions of selected catalysts with light > 420 nm
4
Figure S3
XPS of Au-TiO2 catalysts in the Ti 2p region
5
In situ particle formation in AuP25-DP
5
Figure S4
In situ UV-vis DR spectra of AuP25-SIM and AuP25DP
6
Figure S5
Au particle size distributions of AuP25-DP after the reaction
6
Figure S6
UV-vis DR spectra of AuP25-DP after the reaction
7
Figure S7
In situ FTIR spectra of AuP25-SIM under UV-vis and vis light
8
Figure S8
FTIR spectra of AuP25-SIM and AuP25-DP
9
Figure S9
In situ EPR spectra of calcined AuP25-DP at T = 90 K
10
Figure S10
XRD patterns of Au-TiO2 samples with different supports
11
Figure S11
Photocatalytic H2 production curves of catalysts with different supports
11
Figure S12
H2 production of Au-TiO2 with different supports and preparation methods 12
Figure S13
EPR spectra of Au-TiO2 with different supports and preparation methods
12
1
Catalyst preparation The anatase/brookite mixture (Bro) was prepared as follows: Titanium powder (10 mmol, Aldrich) was dissolved in 30 % H2O2 (40 ml, Roth) and 25 % NH3 (11.2 ml, VWR) and the mixture was stored for 16 h at 5 °C. Glycolic acid (27 mmol, Aldrich) was added at 25 °C and the solution was slowly heated to 95 °C under continuous stirring. Finally, evaporation of water (after 7 h) gave a yellow titanium glycolate complex. This complex (5 mmol) was dissolved in H2O (10 ml) and 25 % NH3 (10 ml), to adjust pH 10.8. The solution was placed in a hydrothermal Teflon reactor (Parr Instruments, 75 ml) and heated to 260 °C for 24 h. The resulting white TiO2 powder was washed repetitively and dried at 40 °C.
Photocatalytic tests
a)
b)
Figure S1. Radiation spectra recorded by Lumatec on their Superlite 400 for two different internal filter settings: a) UV-vis light (320−500 nm, top) and b) visible light (400−700 nm, bottom). It must be emphasized that the vis-filter does not completely filter out light below 400 nm (marked red). This small UV intensity might also contribute to the visible-light activities as discussed in the main manuscript. In order to determine the contribution of the pure plasmon-induced visible-light activity, an additional external filter (> 420 nm) was applied (see Table S1).
2
Figure S2. Experimental hydrogen production curves of AuP25 samples prepared by diverse Au deposition methods from methanol/water mixtures. The measurements represent one preparation batch of each method. The hydrogen evolution rate was constant over a period of 3 h under UV (320−500 nm) as well as over subsequent 21 h under visible light (400−700 nm).
Photocatalytic tests with pure visible light (>420 nm) In order to explore the pure plasmonic Au activity, the residual UV content of the visible-light filter (see Figure S1b) was filtered out for selected samples by use of an additional external filter (420 nm, Schott GG420), which was placed at a distance of 1 cm from the end of the optical fibre. Since the automatic burette described in the experimental section of the main manuscript was not sensitive enough to detect very small H2 evolutions, these experiments were carried out as follows: The catalyst (50 mg) was added to a 10 ml Schlenk tube loaded with a magnetic stirrer bar and closed with a glass piece containing a tap and rubber sealant fitting. The whole system was purged with argon three times. Methanol:water (1:1, 10 ml) was added through the rubber sealed screw-fitted joint and both taps were then closed. A Hg lamp (Lumatec, 4 W, filter 0 (400−700 nm)) equipped with the additional 420 nm filter was positioned with the end of the optical fibre cable 4.5 cm away from the Schlenk tube. The tube was submerged in a beaker of water, from which the side of the tube was positioned 0.5 cm away. The lamp was turned on and the solution was stirred for 6−26 hours before a sample (5 cm3) of the gaseous head space was removed and analysed by gas chromatography. Over time, the catalyst adhered to the inside side of the flask during reaction and thus occasionally this was cleared manually using an external magnetic stirrer bar to agitate the one internally. The results of these experiments are presented in Table S1.
3
Table S1. Catalytic performance of selected Au-TiO2 catalysts for H2 production under photocatalytic conditionsa b
Exp.-No.
Catalyst
Time / h
Integrated intensity / 25 µVs
Area/h
I
d
AuP25-SIM
6.75
54
8
II
AuP25-SIM
26
270
10
III
AuP25-DP12
24
2
0
IV
AuBro-DP
24
5
0
V
AuP25-DP
24
34
1
e
AuP25-DP
0.67
167
250
f
AuP25-DP
24.5
126
5
VIII
g
P25
25.5
1
0
IX
P25
24
28
VI
VII
h
c
Vol H2 / µmol
15
12
1
a
Unless otherwise stated an additional 420 nm filter was used, where the power measured after the filter was 3.05 W. b Area taken from TCD2 B trace of a HP 6890 gas chromatogram. Note: area expected to be specific to the GC. c Only amounts of H2 above the calibration limit (1% H2 calibrated to 147 /25 µVs) have been calculated. Free volume of system was measured as 14.75 ml. d Power measured after filter was 2.95 W e No additional filter was applied. Reaction mixture from experiment V was used and conducted immediately after completion of experiment V. f Reaction mixture from experiment V was used and was conducted immediately after completion of experiment VI. g LOT Oriel Xe lamp used. Power measured after filter was 1.30 W h Average of two experiments
4
XPS in the Ti 2p region
Figure S3. XPS spectra for the AuP25 samples prepared by different deposition procedures in the Ti 2p region. The peaks at 456.0 eV (DP) and 454.9 eV (PD) are artefacts due to partial charging (asterisked).
In situ particle formation in AuP25-DP The in situ UV-vis spectra clearly demonstrate the influence of irradiation in H2O/MeOH on the SPR band of AuP25-DP (Figure S4b). The intensity increased due to the ongoing formation of reduced SPR-active Au0 particles. Moreover, a slight shift of the band maximum to higher wavelength indicates a gradual particle growth. This was not observed for catalyst AuP25-SIM, which already contains such Au0 particles in the as-synthesized state (Figure S4b). This may be one reason for the high activity of AuP25-DP under UV light (Table 1, Figure S2), since the in situ formation of Au0 leads to a high dispersion of equally sized Au particles. Their agglomeration during the synthesis procedure, as observed for AuP25-PD, is inhibited. Moreover, the presence of Au0 particles of sufficient size is obviously required for catalytic activity under pure visible light as well. This is clearly evident from experiments with catalyst AuP25-DP in which initial irradiation (>420 nm, 24 h) gives low quantities of H2 (Table S1). However, after a short period of radiation (400−700 nm) that includes a small contribution of UV light, four times the amount of H2 was generated under the same conditions (>420 nm, 24 h). That small UV contribution from using the 400−700 nm filter, also present under solar radiation, apparently provokes a sufficient Au0 particle growth which enables pure SPR-induced H2 generation. The mean Au particle sizes of AuP25-DP catalysts, which have been removed from the reactor after the photocatalytic test using the 320−500 nm filter for 3 h and the 400−700 nm filter for 21 h, were similar with 7.1 nm and 8.8 nm, respectively, (Figure S5) leading in both cases to an SPR-absorption maximum 545 nm (Figure S6). 5
Figure S4. Changes in the UV-vis spectra of a) AuP25-SIM and b) AuP25-DP (black: as prepared) by addition of H2O/MeOH and subsequent irradiation with UV-vis light (300−700 nm, recorded each minute during irradiation (darkening red). In contrast to the AuP25-DP sample prepared by deposition precipitation, the position of the SPR band maximum of AuP25-SIM did not change indicating no relevant change in particle sizes during the reaction.
a)
b)
Figure S5. Particle-size distributions obtained by TEM analysis of AuP25-DP after the photocatalytic tests a) using the UV-vis filter (Figure S1a, 320−500 nm) for 3 h and b) using the visible-light filter (Figure S1b, 400−700 nm) for 21 h revealing mean Au diameters of 7.1 and 8.8 nm, respectively.
6
Figure S6. UV-vis absorption spectra of AuP25-DP as-synthesized (black), re-isolated after the catalytic testing with pure visible light for 21 h (green) and with UV-vis light for 3 h (blue) corresponding to the particle-size distributions of Figure S5.
7
ATR-IR measurements Spectra were recorded on a Nicolet Avatar 370 (Thermo Electron) FTIR spectrometer equipped with a MCT detector. A Specac Gateway multi reflection horizontal accessory, coupled to a custom-made 2 ml flow-through cell with quartz window containing a ZnSe crystal on the bottom plate, was used for in situ experiments. The 45° internal reflection element of 72 x 10 x 6 mm enables 6 infrared bounces. All spectra were recorded with 64 scans at 4 cm-1 resolution. Before measurement, the ZnSe crystal was coated with the catalyst. To this end, AuP25-SIM was suspended in distilled water (c = 1.47 g/l) and the suspension was treated for 30 min in an ultrasonic bath. Then, 3 ml of this suspension was spread on the ZnSe crystal and dried in vacuum overnight. For the experiment, a He stream (20 ml/min) saturated with a mixture of Methanol/H2O (1:1) was flushed through the ATR-IR cell. After 10 min Helium flush (20 ml/min) a spectrum was recorded, which was subtracted from the spectra measured under irradiation. Then, the ATR-IR cell was irradiated with a Xe lamp for 30 min without filter (UV-vis light) or for 120 min with visible light by use of an 420 nm cutoff filter (300 W, LOT-Oriel GmbH & Co. KG).
a)
UV-vis
b)
vis
Figure S7. In situ FTIR measurements on AuP25-SIM: Depicted are the difference spectra obtained by subtracting the respective dark spectrum from the spectrum of the sample after 3 h UV-vis irradiation (purple lines: no filter) or from that after 24 h pure visible-light irradiation (green lines: λ > 420 nm). The positive bands demonstrate the increasing concentration of decomposition products over time, while negative bands result from the consumption of methanol. The same band positions were observed for the decomposition under visible light− though with lower rate of increase – compared to the experiment under UV-vis light. This shows that hydrogen is evolved also by the pure plasmonic excitation of the Au electrons.
8
FTIR measurements in transition mode The FTIR measurements were carried out on a Thermo Scientific Nicolet 6700 spectrometer equipped with a heatable and evacuable homemade reaction cell with CaF2 windows connected to a gas-dosing system. The sample powders were pressed into self-supporting wafers with a diameter of 20 mm and a weight of 50 mg. All spectra were recorded at room temperature with a resolution of 4 cm-1 and 64 scans. Before measurement, the samples were pretreated by heating in He up to 100°C for 30 min and then cooled to room temperature.
Figure S8. FTIR spectra of AuP25-DP (blue) and AuP25-SIM (red) a) high wavenumbers (inset: ν(OH) region) and b) low wavenumbers. In contrast to the DP procedure, SIM does not lead to a specific OH-coverage, so that no O2•− signal can be formed by trapping of the positive holes at OH-groups. The spectra of AuP25-SIM furthermore indicate the presence of organic residues probably due to the use of PVA as stabilizer.
9
Low-temperature EPR measurements (90 K)
a)
b)
Figure S9. a) In situ EPR spectra (black) and simulations (red) of AuP25-DP uncalcined as well as calcined at 200, 400 and 600°C. Depicted are difference spectra (spectra obtained after 15 min irradiation under UV-vis light at T = 90 K minus dark spectra). b) Signal contributions derived by spectra deconvolution through simulation of the EPR signals: O– (g1 = 2.026, g2 = 2.015, g3 = 2.004) corresponding to oxide radicals formed by trapping positive holes at lattice O2–, Ti3+ (anatase: g1 = 1.990, g2 = 1.963; rutile: g1 = 1.979, g2 = 1.954) and C* (g = 2.035), the latter of which may be assigned to surface-bound •O2H radicals (see ref. 34 in the main manuscript), but could not be well resolved. It is clearly seen that calcination at low temperature (200°C) led to inefficient charge separation, as electrons are trapped in TiO2 forming Ti3+ species instead of being transferred to Au, where they cannot be detected by EPR at 90 K. The most active samples AuP25-DP and –DP600 also exhibit the lower amount of bulk Ti3+ species due to effective electron transfer to Au.
10
X-ray diffraction (XRD) powder patterns at ambient conditions were recorded in transmission geometry with Cu Kα radiation (λ = 1.5406 Å) in the 2θ range of 10–70° (step width = 0.25°) on a Stoe STADI P diffractometer, equipped with a linear position sensitive detector (PSD).
Figure S10. XRD patterns of various Au-TiO2 samples prepared by DP method with different support materials.
Figure S11. H2 evolution curves in dependence on the TiO2 phase of Au-TiO2-DP catalysts a) under UV-vis light as well as b) under pure visible light irradiation corresponding to Table 5 in the main manuscript.
11
0
0
0
Figure S12. Catalytic activities in H2 generation from H2O/MeOH mixtures of various Au-TiO2 photocatalysts prepared by use of different methods (black: sol-immobilization SIM, blue: deposition precipitation DP, red: prolonged deposition precipitation DP12) as well as different support materials (P25, Rut and Ana) under UVvis (top) and pure visible light irradiation (bottom).
Figure S13. EPR spectra of various Au-TiO2 photocatalysts (corresponding to Figure S10) prepared by use of different methods (black: sol-immobilization SIM, blue: deposition precipitation DP, red: prolonged deposition precipitation DP12) as well as different support materials (P25, Rut and Ana) under UV-vis irradiation recorded at 290 K.
12
Solar Hydrogen Production by Plasmonic Au-TiO2 Catalysts: Impact of Synthesis Protocol and TiO2 Phase on Charge Transfer Efficiency and H2 Evolution Rates. Jacqueline B. Priebe, Jörg Radnik, Alastair J. J. Lennox, Marga-Martina Pohl, Michael Karnahl,† Dirk Hollmann, Kathleen Grabow, Ursula Bentrup, Henrik Junge, Matthias Beller, and Angelika Brückner * Leibniz-Institut für Katalyse e.V. an der Universität Rostock, Albert-Einstein-Str. 29a, 18059 Rostock, Germany, [email protected]
Table of Content
page
Catalyst preparation “Bro”
2
Figure S1
Radiation spectra of the light source
2
Figure S2
Photocatalytic H2 production curves
3
Table S1
H2 evolutions of selected catalysts with light > 420 nm
4
Figure S3
XPS of Au-TiO2 catalysts in the Ti 2p region
5
In situ particle formation in AuP25-DP
5
Figure S4
In situ UV-vis DR spectra of AuP25-SIM and AuP25DP
6
Figure S5
Au particle size distributions of AuP25-DP after the reaction
6
Figure S6
UV-vis DR spectra of AuP25-DP after the reaction
7
Figure S7
In situ FTIR spectra of AuP25-SIM under UV-vis and vis light
8
Figure S8
FTIR spectra of AuP25-SIM and AuP25-DP
9
Figure S9
In situ EPR spectra of calcined AuP25-DP at T = 90 K
10
Figure S10
XRD patterns of Au-TiO2 samples with different supports
11
Figure S11
Photocatalytic H2 production curves of catalysts with different supports
11
Figure S12
H2 production of Au-TiO2 with different supports and preparation methods 12
Figure S13
EPR spectra of Au-TiO2 with different supports and preparation methods
12
1
Catalyst preparation The anatase/brookite mixture (Bro) was prepared as follows: Titanium powder (10 mmol, Aldrich) was dissolved in 30 % H2O2 (40 ml, Roth) and 25 % NH3 (11.2 ml, VWR) and the mixture was stored for 16 h at 5 °C. Glycolic acid (27 mmol, Aldrich) was added at 25 °C and the solution was slowly heated to 95 °C under continuous stirring. Finally, evaporation of water (after 7 h) gave a yellow titanium glycolate complex. This complex (5 mmol) was dissolved in H2O (10 ml) and 25 % NH3 (10 ml), to adjust pH 10.8. The solution was placed in a hydrothermal Teflon reactor (Parr Instruments, 75 ml) and heated to 260 °C for 24 h. The resulting white TiO2 powder was washed repetitively and dried at 40 °C.
Photocatalytic tests
a)
b)
Figure S1. Radiation spectra recorded by Lumatec on their Superlite 400 for two different internal filter settings: a) UV-vis light (320−500 nm, top) and b) visible light (400−700 nm, bottom). It must be emphasized that the vis-filter does not completely filter out light below 400 nm (marked red). This small UV intensity might also contribute to the visible-light activities as discussed in the main manuscript. In order to determine the contribution of the pure plasmon-induced visible-light activity, an additional external filter (> 420 nm) was applied (see Table S1).
2
Figure S2. Experimental hydrogen production curves of AuP25 samples prepared by diverse Au deposition methods from methanol/water mixtures. The measurements represent one preparation batch of each method. The hydrogen evolution rate was constant over a period of 3 h under UV (320−500 nm) as well as over subsequent 21 h under visible light (400−700 nm).
Photocatalytic tests with pure visible light (>420 nm) In order to explore the pure plasmonic Au activity, the residual UV content of the visible-light filter (see Figure S1b) was filtered out for selected samples by use of an additional external filter (420 nm, Schott GG420), which was placed at a distance of 1 cm from the end of the optical fibre. Since the automatic burette described in the experimental section of the main manuscript was not sensitive enough to detect very small H2 evolutions, these experiments were carried out as follows: The catalyst (50 mg) was added to a 10 ml Schlenk tube loaded with a magnetic stirrer bar and closed with a glass piece containing a tap and rubber sealant fitting. The whole system was purged with argon three times. Methanol:water (1:1, 10 ml) was added through the rubber sealed screw-fitted joint and both taps were then closed. A Hg lamp (Lumatec, 4 W, filter 0 (400−700 nm)) equipped with the additional 420 nm filter was positioned with the end of the optical fibre cable 4.5 cm away from the Schlenk tube. The tube was submerged in a beaker of water, from which the side of the tube was positioned 0.5 cm away. The lamp was turned on and the solution was stirred for 6−26 hours before a sample (5 cm3) of the gaseous head space was removed and analysed by gas chromatography. Over time, the catalyst adhered to the inside side of the flask during reaction and thus occasionally this was cleared manually using an external magnetic stirrer bar to agitate the one internally. The results of these experiments are presented in Table S1.
3
Table S1. Catalytic performance of selected Au-TiO2 catalysts for H2 production under photocatalytic conditionsa b
Exp.-No.
Catalyst
Time / h
Integrated intensity / 25 µVs
Area/h
I
d
AuP25-SIM
6.75
54
8
II
AuP25-SIM
26
270
10
III
AuP25-DP12
24
2
0
IV
AuBro-DP
24
5
0
V
AuP25-DP
24
34
1
e
AuP25-DP
0.67
167
250
f
AuP25-DP
24.5
126
5
VIII
g
P25
25.5
1
0
IX
P25
24
28
VI
VII
h
c
Vol H2 / µmol
15
12
1
a
Unless otherwise stated an additional 420 nm filter was used, where the power measured after the filter was 3.05 W. b Area taken from TCD2 B trace of a HP 6890 gas chromatogram. Note: area expected to be specific to the GC. c Only amounts of H2 above the calibration limit (1% H2 calibrated to 147 /25 µVs) have been calculated. Free volume of system was measured as 14.75 ml. d Power measured after filter was 2.95 W e No additional filter was applied. Reaction mixture from experiment V was used and conducted immediately after completion of experiment V. f Reaction mixture from experiment V was used and was conducted immediately after completion of experiment VI. g LOT Oriel Xe lamp used. Power measured after filter was 1.30 W h Average of two experiments
4
XPS in the Ti 2p region
Figure S3. XPS spectra for the AuP25 samples prepared by different deposition procedures in the Ti 2p region. The peaks at 456.0 eV (DP) and 454.9 eV (PD) are artefacts due to partial charging (asterisked).
In situ particle formation in AuP25-DP The in situ UV-vis spectra clearly demonstrate the influence of irradiation in H2O/MeOH on the SPR band of AuP25-DP (Figure S4b). The intensity increased due to the ongoing formation of reduced SPR-active Au0 particles. Moreover, a slight shift of the band maximum to higher wavelength indicates a gradual particle growth. This was not observed for catalyst AuP25-SIM, which already contains such Au0 particles in the as-synthesized state (Figure S4b). This may be one reason for the high activity of AuP25-DP under UV light (Table 1, Figure S2), since the in situ formation of Au0 leads to a high dispersion of equally sized Au particles. Their agglomeration during the synthesis procedure, as observed for AuP25-PD, is inhibited. Moreover, the presence of Au0 particles of sufficient size is obviously required for catalytic activity under pure visible light as well. This is clearly evident from experiments with catalyst AuP25-DP in which initial irradiation (>420 nm, 24 h) gives low quantities of H2 (Table S1). However, after a short period of radiation (400−700 nm) that includes a small contribution of UV light, four times the amount of H2 was generated under the same conditions (>420 nm, 24 h). That small UV contribution from using the 400−700 nm filter, also present under solar radiation, apparently provokes a sufficient Au0 particle growth which enables pure SPR-induced H2 generation. The mean Au particle sizes of AuP25-DP catalysts, which have been removed from the reactor after the photocatalytic test using the 320−500 nm filter for 3 h and the 400−700 nm filter for 21 h, were similar with 7.1 nm and 8.8 nm, respectively, (Figure S5) leading in both cases to an SPR-absorption maximum 545 nm (Figure S6). 5
Figure S4. Changes in the UV-vis spectra of a) AuP25-SIM and b) AuP25-DP (black: as prepared) by addition of H2O/MeOH and subsequent irradiation with UV-vis light (300−700 nm, recorded each minute during irradiation (darkening red). In contrast to the AuP25-DP sample prepared by deposition precipitation, the position of the SPR band maximum of AuP25-SIM did not change indicating no relevant change in particle sizes during the reaction.
a)
b)
Figure S5. Particle-size distributions obtained by TEM analysis of AuP25-DP after the photocatalytic tests a) using the UV-vis filter (Figure S1a, 320−500 nm) for 3 h and b) using the visible-light filter (Figure S1b, 400−700 nm) for 21 h revealing mean Au diameters of 7.1 and 8.8 nm, respectively.
6
Figure S6. UV-vis absorption spectra of AuP25-DP as-synthesized (black), re-isolated after the catalytic testing with pure visible light for 21 h (green) and with UV-vis light for 3 h (blue) corresponding to the particle-size distributions of Figure S5.
7
ATR-IR measurements Spectra were recorded on a Nicolet Avatar 370 (Thermo Electron) FTIR spectrometer equipped with a MCT detector. A Specac Gateway multi reflection horizontal accessory, coupled to a custom-made 2 ml flow-through cell with quartz window containing a ZnSe crystal on the bottom plate, was used for in situ experiments. The 45° internal reflection element of 72 x 10 x 6 mm enables 6 infrared bounces. All spectra were recorded with 64 scans at 4 cm-1 resolution. Before measurement, the ZnSe crystal was coated with the catalyst. To this end, AuP25-SIM was suspended in distilled water (c = 1.47 g/l) and the suspension was treated for 30 min in an ultrasonic bath. Then, 3 ml of this suspension was spread on the ZnSe crystal and dried in vacuum overnight. For the experiment, a He stream (20 ml/min) saturated with a mixture of Methanol/H2O (1:1) was flushed through the ATR-IR cell. After 10 min Helium flush (20 ml/min) a spectrum was recorded, which was subtracted from the spectra measured under irradiation. Then, the ATR-IR cell was irradiated with a Xe lamp for 30 min without filter (UV-vis light) or for 120 min with visible light by use of an 420 nm cutoff filter (300 W, LOT-Oriel GmbH & Co. KG).
a)
UV-vis
b)
vis
Figure S7. In situ FTIR measurements on AuP25-SIM: Depicted are the difference spectra obtained by subtracting the respective dark spectrum from the spectrum of the sample after 3 h UV-vis irradiation (purple lines: no filter) or from that after 24 h pure visible-light irradiation (green lines: λ > 420 nm). The positive bands demonstrate the increasing concentration of decomposition products over time, while negative bands result from the consumption of methanol. The same band positions were observed for the decomposition under visible light− though with lower rate of increase – compared to the experiment under UV-vis light. This shows that hydrogen is evolved also by the pure plasmonic excitation of the Au electrons.
8
FTIR measurements in transition mode The FTIR measurements were carried out on a Thermo Scientific Nicolet 6700 spectrometer equipped with a heatable and evacuable homemade reaction cell with CaF2 windows connected to a gas-dosing system. The sample powders were pressed into self-supporting wafers with a diameter of 20 mm and a weight of 50 mg. All spectra were recorded at room temperature with a resolution of 4 cm-1 and 64 scans. Before measurement, the samples were pretreated by heating in He up to 100°C for 30 min and then cooled to room temperature.
Figure S8. FTIR spectra of AuP25-DP (blue) and AuP25-SIM (red) a) high wavenumbers (inset: ν(OH) region) and b) low wavenumbers. In contrast to the DP procedure, SIM does not lead to a specific OH-coverage, so that no O2•− signal can be formed by trapping of the positive holes at OH-groups. The spectra of AuP25-SIM furthermore indicate the presence of organic residues probably due to the use of PVA as stabilizer.
9
Low-temperature EPR measurements (90 K)
a)
b)
Figure S9. a) In situ EPR spectra (black) and simulations (red) of AuP25-DP uncalcined as well as calcined at 200, 400 and 600°C. Depicted are difference spectra (spectra obtained after 15 min irradiation under UV-vis light at T = 90 K minus dark spectra). b) Signal contributions derived by spectra deconvolution through simulation of the EPR signals: O– (g1 = 2.026, g2 = 2.015, g3 = 2.004) corresponding to oxide radicals formed by trapping positive holes at lattice O2–, Ti3+ (anatase: g1 = 1.990, g2 = 1.963; rutile: g1 = 1.979, g2 = 1.954) and C* (g = 2.035), the latter of which may be assigned to surface-bound •O2H radicals (see ref. 34 in the main manuscript), but could not be well resolved. It is clearly seen that calcination at low temperature (200°C) led to inefficient charge separation, as electrons are trapped in TiO2 forming Ti3+ species instead of being transferred to Au, where they cannot be detected by EPR at 90 K. The most active samples AuP25-DP and –DP600 also exhibit the lower amount of bulk Ti3+ species due to effective electron transfer to Au.
10
X-ray diffraction (XRD) powder patterns at ambient conditions were recorded in transmission geometry with Cu Kα radiation (λ = 1.5406 Å) in the 2θ range of 10–70° (step width = 0.25°) on a Stoe STADI P diffractometer, equipped with a linear position sensitive detector (PSD).
Figure S10. XRD patterns of various Au-TiO2 samples prepared by DP method with different support materials.
Figure S11. H2 evolution curves in dependence on the TiO2 phase of Au-TiO2-DP catalysts a) under UV-vis light as well as b) under pure visible light irradiation corresponding to Table 5 in the main manuscript.
11
0
0
0
Figure S12. Catalytic activities in H2 generation from H2O/MeOH mixtures of various Au-TiO2 photocatalysts prepared by use of different methods (black: sol-immobilization SIM, blue: deposition precipitation DP, red: prolonged deposition precipitation DP12) as well as different support materials (P25, Rut and Ana) under UVvis (top) and pure visible light irradiation (bottom).
Figure S13. EPR spectra of various Au-TiO2 photocatalysts (corresponding to Figure S10) prepared by use of different methods (black: sol-immobilization SIM, blue: deposition precipitation DP, red: prolonged deposition precipitation DP12) as well as different support materials (P25, Rut and Ana) under UV-vis irradiation recorded at 290 K.
12
