Photoelectrochemical Synthesis of Ammonia on the Aerophilic ...

Article

Photoelectrochemical Synthesis of Ammoniaon the Aerophilic-Hydrophilic Heterostructurewith 37.8% Efficiency

Jianyun Zheng, Yanhong Lyu,

Man Qiao, ..., Huaijuan Zhou,

San Ping Jiang, Shuangyin

Wang

[email protected] (Y.L.)

[email protected] (H.Z.)

[email protected] (S.P.J.)

[email protected] (S.W.)

HIGHLIGHTS

Aerophilic-hydrophilic

heterostructure achieved a superb

NRR

The heterostructure enriched N2

concentration at the Au active

sites

The heterostructure manipulated

the proton activity with

suppressed HER

DFT calculation indicated the

decrease in the NRR energy

barrier on the structure

A unique aerophilic-hydrophilic heterostructure composed of Au nanoparticles

highly dispersed in a poly(tetrafluoroethylene) porous framework is fabricated on a

Si-based photocathode for N2-to-NH3 fixation. The amphipathic nature of the

heterostructure is considered to be the origin of the enhanced nitrogen reduction

reaction with efficient conversion efficiency and high production rate.

Zheng et al., Chem 5, 617–633

March 14, 2019 ª 2018 Elsevier Inc.

https://doi.org/10.1016/j.chempr.2018.12.003

Article

Photoelectrochemical Synthesisof Ammonia on the Aerophilic-HydrophilicHeterostructure with 37.8% EfficiencyJianyun Zheng,1,2,5 Yanhong Lyu,1,2,5 Man Qiao,3,5 Ruilun Wang,1 Yangyang Zhou,1 Hao Li,1

Chen Chen,1 Yafei Li,3,* Huaijuan Zhou,4,* San Ping Jiang,2,* and Shuangyin Wang1,6,*

The Bigger Picture

Liquid ammonia can be the key

enabler for being an easily

transported energy storage

carrier, which is highly desirable to

be produced from renewable

energy, such as solar or electricity,

under eco-friendly and mild

conditions. However, innovation

in the photoelectrochemical

devices with high activity,

stability, and selectivity for

nitrogen-to-ammonia fixation has

proven to be very challenging

because nitrogen reduction

SUMMARY

Photoelectrochemical nitrogen reduction reaction can provide a useful source of

ammonia and transportable carrier of hydrogen, but the process is limited by the

photocathodes with poor conversion efficiency and low production rate. Here,

we have designed a unique aerophilic-hydrophilic heterostructured Si-based

photocathode for nitrogen-to-ammonia fixation in an acid electrolyte under

mild conditions, achieving a high ammonia yield rate of �18.9 mg,cm�2,hr�1

and an excellent faradic efficiency of 37.8% at �0.2 V versus a reversible

hydrogen electrode. The heterostructure based on the Au nanoparticles highly

dispersed in poly(tetrafluoroethylene) porous framework enriches nitrogen

molecular concentration at the Au active sites while manipulateing the proton

activity with suppressed hydrogen evolution reactions. DFT calculation indicates

that such heterostructure reduces the energy barrier for the nitrogen reduction

reaction. The aerophilic-hydrophilic heterostructure provides a new insight on

designing efficient and robust photocathodes for nitrogen fixation.

reaction competes with the

hydrogen evolution reaction,

which occurs preferentially on the

photocathode surface at a

comparable thermodynamic

potential. Thus, we have designed

a unique aerophilic-hydrophilic

heterostructured Si-based

photocathode for improving the

energy conversion efficiency. The

aerophilic-hydrophilic

heterostructure provides a new

insight on designing efficient and

robust photocathodes for

nitrogen fixation.

INTRODUCTION

Approximately 200 million tons of ammonia (NH3) is produced in the world annually,

reflecting the vast need in agriculture, pharmaceutical production, and other indus-

trial processes.1 NH3 with 17.6 wt % hydrogen is also being considered as an

emerging easily transported carrier of hydrogen energy and carbon-free solar

energy storage carrier.2 As a result of the high energy consumption and CO2 emis-

sion in the traditional Haber-Bosch process,3 it is highly desirable to develop an

alternative route for nitrogen (N2)-to-NH3 fixation under environmentally friendly

and mild conditions. Various strategies have been explored to promote the N2

reduction reaction (NRR), including biochemical, photocatalytic, electrochemical

and photoelectrochemical (PEC).4–6 Among them, PEC NRR is of considerable inter-

est because it can couple with the merits of the photocatalytic and electrochemical

processes. Such an approach not only allows for the possibility of NRR under mild

conditions such as ambient temperature and atmospheric pressure, but also can

be powered by sun.

Recently, significant effort has been devoted to advancing the prospects of produc-

ing NH3 from N2.7–10 The NRR at room temperature yields a certain amount of NH3

production and increased faradic efficiencies via tuning the catalyst structure,11,12

introducing the dopants and defects,13–15 and manipulating the reaction condi-

tions.16,17 In previous work, Ali et al.18 prepared nanostructured p-type Si with Au

nanoparticles (NPs) and achieved a PEC N2-to-NH3 fixation at 6.0 mg,cm�2,hr�1

Chem 5, 617–633, March 14, 2019 ª 2018 Elsevier Inc. 617

1State Key Laboratory of Chem/Bio-Sensing andChemometrics, College of Chemistry andChemical Engineering, Hunan University,Changsha, Hunan 410082, China

2Western Australian School of Mines: Minerals,Energy and Chemical Engineering and Fuels andEnergy Technology Institute, Curtin University,Perth, WA 6102, Australia

3Jiangsu Collaborative Innovation Centre ofBiomedical Functional Materials, School ofChemistry and Materials Science, Nanjing

without applied bias from atmospheric N2 under 2 sun illumination and 7 atm pres-

sure. Separately, a faradic efficiency of 10% for NRR was recently obtained on a

zeolitic imidazolate framework-71 (ZIF)-coated Ag-Au platform by suppressing the

hydrogen evolution reaction (HER) via hydrophobic ZIF.19 Mukherjee et al.20 re-

ported a metal-organic-framework-derived nitrogen-doped nanoporous carbon

with a remarkable NH3 production rate of 3.4 mmol,cm�2,hr�1 and a faradic effi-

ciency of 10.2% for NRR under room temperature by using alkaline aqueous electro-

lyte. In addition, electrochemical NRR with a notable faradic efficiency of �35% was

acquired with N2 and steam in a molten hydroxide suspension under high-

temperature (200�C) and high-pressure (>25 atm) conditions.3 To achieve a higher

faradic efficiency (�60%), Zhou et al.16 employed ionic liquids with high N2 solubility

as electrolyte to suppress the HER and enhance the conversion of N2 to NH3 on a

nanostructured iron catalyst under ambient conditions. However, because water is

the most common and environmentally friendly solvent and source of protons,

development of a water-based system for NRR is of practical and technological

importance. Nevertheless, most reported NRRs in aqueous solution are still very

unfavorable under mild conditions, retaining the conversion efficiency of �10% or

lower.21 The most significant difficulty behind the low conversion efficiency is the

poor selectivity for NH3 production and the overwhelming competing HER from

water-fed systems across all traditional NRR catalysts.

To settle the competition from HER, one of the general strategies is to employ a gas-

diffusion layer to supply an unobstructed gas-diffusion pathway and impede water

absorption at the electrode surface.19,22,23 Porous hydrophobic materials, such as

poly(tetrafluoroethylene) (PTFE), are commonly used to accelerate the gas-diffusion

process by creating a gas molecular-concentrating effect. But, their insulating nature

would obstruct the electron transport and result in a severe performance decay.

On the other hand, the formation of active proton (*H) is limited because of its

hydrophobic structure, further reducing the NRR performance. Therefore, it is

challenging to simultaneously satisfy the stringent requirements on N2 gas diffusion,

electron transport, and production of *H.

In this work, we designed a highly efficient Si-based photocathode for solar-light-

driven conversion of N2 to NH3, which was composed of Si as the photo absorber,

PTFE porous framework as the gas-diffusion layer, and Au NPs as the active sites.

The highly dispersed and nearly free-stacked Au NPs coated on the surface of the

PTFE framework and the Si formed an intimate connection between the PTFE frame-

work and Si, providing an electric contact. Most critical, the PTFE porous framework

with highly dispersed Au NPs created a functional aerophilic-hydrophilic hetero-

structure, facilitating an enriched N2-gas layer and controlling proton activity under

aqueous media. The optimal photocathode showed an ammonia yield rate of

�18.9 mg,cm�2,hr�1 and a faradic efficiency of 37.8% at �0.2 V versus reversible

hydrogen electrode (RHE) at ambient condition.

4State Key Laboratory of High PerformanceCeramics and Superfine Microstructure,Shanghai Institute of Ceramics, ChineseAcademy of Sciences, Shanghai 200050, China

5These authors contributed equally

6Lead Contact

*Correspondence: [email protected] (Y.L.),[email protected] (H.Z.),[email protected] (S.P.J.),[email protected] (S.W.)

https://doi.org/10.1016/j.chempr.2018.12.003

RESULTS AND DISCUSSION

Synthesis and Characterization of Si-Based Photocathodes

Figure 1A shows the scheme for fabrication of the aerophilic-hydrophilic hetero-

structured Si-based photocathode for PEC NRR. A thin Ti layer on the Si surface

(labeled as TS) affords a passivation function in terms of stability under acid condi-

tions24 and serves as an adhesion layer for growing the Au NPs and PTFE porous

framework.25 PTFE, as one of the most hydrophobic materials,26 was chosen as

the porous framework for allowing the diffusion of N2 and stabilizing the

618 Chem 5, 617–633, March 14, 2019

Figure 1. Fabrication and Structural Characterizations of the Si-Based Photocathodes

(A) Schematic illustration of the fabrication procedure of Au/TS and Au-PTFE/TS.

(B) XRD patterns of Au/TS (orange line) and Au-PTFE/TS (purple line). The standard XRD pattern

for Au (light blue star) is shown at the bottom (JCPDS, No. 04-0784). The inset is an FTIR diagram of

Au/TS (orange line) and Au-PTFE/TS (purple line).

(C) XPS spectra of Au 4f of the photocathodes. The black solid lines are XPS data. The red dashed

lines are the fitting of experimental data for the photocathodes, which can be decomposed into a

superposition of two peaks shown as Au0 and Au1+.

(D) Top view FESEM image of Au-PTFE/TS. The inset is the magnification image of the specified

area in Au-PTFE/TS.

(E) Cross-sectional FESEM image of Au-PTFE/TS. The inset is the corresponding EDS-line

measurements with the blue arrow.

photocathode during PEC NRR. In addition, Au NPs as the active catalysts for NRR27

were dispersed on the surface of photocathodes (Figure 1A). We investigated the Au

NPs on the TS photocathode (denoted as Au/TS) with hydrophilicity and the one

coated by the hydrophobic PTFE porous framework with suitable thickness and Au

NPs, denoted as Au-PTFE/TS (Figures 1A and S1; details described in the Experi-

mental Procedures) toward PEC NRR. In addition, we also prepared other photo-

cathodes with different levels of thickness of PTFE porous framework (as described

in the Experimental Procedures).

Successful loading of the crystalline Au catalyst onto Au/TS and Au-PTFE/TS was

confirmed by X-ray diffraction (XRD; Figure 1B). Two broad peaks at 38.2 and

Chem 5, 617–633, March 14, 2019 619

44.4� were observed on Au/TS, which can be indexed to (111) and (200) planes of Au

(PDF no. 04-0784, JCPDS), respectively. In comparison to Au/TS, Au-PTFE/TS

showed broader and lower peaks, implying smaller Au NPs in Au-PTFE/TS (Fig-

ure S2). Meanwhile, the crystalline Au NPs were also found in the other samples

with different thicknesses of PTFE (Figure S2). In addition, Fourier transform infrared

spectroscopy (FTIR) was used to further assess the compositions of the samples.

Two peaks at�1,150–1,250 cm�1 confirm the presence of the –CF3 functional group

(inset of Figures 1B and S3),26 indicating the successful coating of PTFE. Subse-

quently, X-ray photoelectron spectroscopy (XPS) spectra were employed for

studying the elemental features and states of the samples (Figures S4 and S5).

The Ti thin layer was oxidized in ambient to form amorphous TiO2 layer (Figure S6),

which can play a role of protection layer to stabilize the PEC operation of the photo-

cathode as demonstrated in our previous work.28 The peaks at�291.8 eV of C 1s and

�688.9 eV of F 1s (Figures S7 and S8) were close to the binding energies of the C–F

and the F–C bonds, in line with FTIR data. In the Au 4f spectra (Figures 1C and S9),

the peaks at�84.1 and�87.8 eV for Au/TS were assigned to metallic Au (Au0). Inter-

estingly, additional peaks at �84.8 and �88.4 eV were displayed in Au-PTFE/TS,

which is associated with Au ion (Au1+).29

To reveal the architecture of the photocathodes, field emission scanning electron

microscopy (FESEM) with energy dispersive spectroscopy (EDS) set at line-scanning

mode was implemented. An obvious 3D porous structure constructed by PTFE

was observed in Au-PTFE/TS (Figure 1D). Instead of coating the Au NPs on the

flat surface (Figure S10), Au NPs with the size range from �2 to �10 nm were highly

dispersed on the surface of the PTFE porous framework and the TS for Au-PTFE/TS

(inset of Figures 1D, S11, and S12). Additionally, the thickness of the PTFE porous

layer in Au-PTFE/TS was �24 nm (Figure 1E), while other photocathodes with thick

PTFE porous frameworks were also provided in Figure S12. Meanwhile, EDS line

profiling indicated the successive construction of Au-PTFE/TS consisting of Au

NPs, PTFE porous framework, and Ti layer on the Si surface. High-resolution

transmission electron microscopy (HRTEM) images of Au-PTFE/TS (Figure S13)

showed a nanoparticle with lattice spacing of �0.23 nm on an amorphous frame-

work, further clarifying the presence of Au NPs with the (111) plane and PTFE frame-

work. Finally, the photocathodes with PTFE porous framework performed with a

lower reflectance in the wavelength range from 380 to 1,200 nm than Au/TS (Fig-

ure S14). According to the indirect allowed transition,30 the band gap of �1.1 eV

derived from Si was observed in all photocathodes.

Physicochemical Properties of Si-Based Photocathodes

Figures 2A and 2D show the droplet shapes of the liquid (0.05 M H2SO4 electro-

lyte with 0.05 M Na2SO3) on the surface of Au/TS and Au-PTFE/TS in the air,

respectively. Au-PTFE/TS showed a hydrophobic surface, and its liquid contact

angle (CAl) is about 125�, whereas Au/TS with CAl of �78� was hydrophilic.

Indeed, the hydrophobic properties were also observed on the surface of the

other photocathodes with different thicknesses of PTFE porous framework (Fig-

ure S15). It should be noted that the photocathode coated with the pure PTFE

porous layer possessed higher CAl than Au-PTFE/TS, which simultaneously

possessed the hydrophobic porous framework (PTFE) and the hydrophilic sites

(Au NPs). These results show the manipulation of the hydrophilic and hydropho-

bic properties of Au NPs catalytic active sites with the incorporation of hydropho-

bic porous PTFE framework. In addition, the hydrophobic performance of

Au-PTFE/TS was stable and durable under illumination and long-time exposure

in ambient conditions (Figure S16).

620 Chem 5, 617–633, March 14, 2019

Figure 2. Physicochemical Properties of the Si-Based Photocathodes

(A and D) The droplet shapes of the liquid (0.05 M H2SO4 electrolyte with 0.05 M Na2SO3) on the

surface of Au/TS (A) and Au-PTFE/TS (D).

(B and E) The shapes of underwater N2 bubble on the surface of Au/TS (B) and Au-PTFE/TS (E).

(C and F) Original (black solid line) and deconvoluted (red dashed line) spectra of interfacial water

at the surface of Au/TS (C) and Au-PTFE/TS (F). The orange, purple, and light-blue lines stand for

the peaks centered at �3,600, �3,400, and �3,200 cm�1, respectively. The insets are schematic of

the orientation of water molecules next to the surface of Au/TS and Au-PTFE/TS (surface and water

molecules not to scale).

(G) AFM topography image (left) and typical I-V curves of marked position (right) for Au-PTFE/TS.

Positions 1�10 in AFM image match in the colored I-V curves.

The aerophilic properties of the photocathodes were assessed by captive-bubble

measurements.31 The surface of Au-PTFE/TS with a gas-bubble contact angle

(CAg) of �88� provided an aerophilic property with a strong interaction toward N2

bubble (Figure 2E), in stark contrast to Au/TS, which showed a high CAg (�111�),implying a relatively weak interaction with the N2 bubble (Figure 2B). Meanwhile,

the change of aerophilic properties was also exhibited in the other photocathodes

with different thicknesses of PTFE porous framework (Figure S17). The transition

from aerophobicity of Au/TS to aerophilicity of Au-PTFE/TS is due to the incorpora-

tion of the PTFE porous framework. The N2-bubble adhesion force of the

Au-PTFE/TS surface was also evaluated; however, it could not be obtained

precisely because the gas bubble would be absorbed on the surface and detach

the bubble capture (Figure S18B),23 further demonstrating the strong gas adhesion.

In comparison, the adhesion force of the Au/TS was about 30.1 mN (Figure S18A),

which is lower than that of Au-PTFE/TS. Consequently, when Au-PTFE/TS was

immersed into the electrolyte, it would be more difficult for the liquid to get into

the PTFE porous framework, and the gas pockets would form on the surface of the

photocathode.

Chem 5, 617–633, March 14, 2019 621

The orientation of interfacial water molecules near the surface of Au/TS and

Au-PTFE/TS was probed by FTIR with attenuated total reflection (ATR) accessory,

as given in Figures 2C and 2F, respectively. To distinguish between the two spectra,

the broad band of vibrational spectrum of water between 3,100 and 3,800 cm�1 (the

OH stretching region), which yielded against the background spectrum of the clean

diamond with the water, was deconvoluted into three peaks centered at �3,200,

3,400, and 3,600 cm�1, which are ascribed to OH stretching modes associated

with the tetrahedral structure of bulk water molecules, hydrogen-bonded OH strad-

dling the interface, and non-hydrogen-bonded OH pointing toward the surface,

respectively.32,33 As reported in the literature,32,33 a hydrophobic material shows

a higher magnitude at �3,600 cm�1 than a hydrophilic material. In Figures 2C and

2F, the intensity of the peak at �3,200 cm�1 related to the OH stretching of bulk wa-

ter was almost identical for the surface of Au/TS and Au-PTFE/TS. In contrast, the

peak at �3,600 cm�1 was higher for Au-PTFE/TS than that for Au/TS by a factor of

�2. Correspondingly, Au-PTFE/TS also displayed a higher peak at �3,400 cm�1,

suggesting that its surface has a larger number of OHs straddling the interface. To

some extent, the intensity of the peaks at �3,400 and �3,600 cm�1 increases with

the increase in the thickness of the PTFE layer (Figure S19). Furthermore, the photo-

cathode with the thicker PTFE layer (Au-PTFE_1/TS; Figure S19C) showed a lower in-

tensity of the peaks at�3,400 and�3,600 cm�1 than Au-PTFE_0.75/TS (Figure S19B)

because more Au NPs were stacked on the surface, in line with XPS data. As a result,

we can imagine that the hydrogen-bonding network of the water molecules next

to the surface of Au-PTFE/TS is destroyed as sketched in the inset of Figure 2F.

One hydrogen-bond vector pointing toward the surface is associated with the

hydrogen bonding between the Au atom and the OH bond of the water molecule,

whereas the remaining three hydrogen-bond vectors point away from the surface.

In contrast, Au atoms and amorphous TiO2 on the surface of Au/TS form hydrogen

bonds with interfacial water molecules, resulting in a hydrophilic hydration structure

(see the inset of Figure 2C).

In order to explore the charge carrier transport, the electrical measurements of the

photocathodes were conducted by conductive atomic force microscopy (c-AFM) in a

vertical configuration. An AFM topographic image showing the typical current-

voltage (I-V) curves of the marked position for Au-PTFE/TS is presented in Figure 2G.

Au-PTFE/TS in the positions 4, 5, 6, and 8 shows the current responses. Among

them, positions 4 and 8 in the hole exhibited sensitive and intense current re-

sponses (10 nA), and the currents of positions 5 and 6 on the PTFE porous

framework with Au NPs were less than 3 nA at an applied bias of �10 V. In contrast

to Au-PTFE/TS, Au/TS in all the positions 1–10 produced high currents (10 nA) at low

applied bias (<7 V) (Figure S20). Nevertheless, when the thickness of the PTFE

porous framework was gradually increased, the current response of the sample

was markedly weakened (Figures S21–S23). Finally, there was no current response

in the TS coating the pure PTFE porous framework with a thickness of �24 nm at

applied bias ranging from �10 to 10 V (Figure S24). The results make clear that

the dispersed Au NPs provide the necessary electron transport channels for NRR

as a result of the intimate connection among Au NPs (Figure S25), and the current

response of the Au NPs on the TS surface is stronger and faster than it is on the

PTFE porous framework for the Au-PTFE/TS photocathode.

Photoelectrochemical Properties of Si-Based Photocathodes

Figure 3A schematically illustrates a PEC cell for NRR with the N2 bubbling over the

surface of the photocathode in 0.05 M H2SO4 electrolyte with 0.05 M Na2SO3 under

1 sun illumination. To facilitate the contact between the N2 bubbling and

622 Chem 5, 617–633, March 14, 2019

Figure 3. Photoelectrochemical N2-to-NH3 Fixation of the Si-Based Photocathodes

(A) Schematic diagram of the PEC cell under 1 sun illumination in 0.05 M H2SO4 electrolyte with

0.05 M Na2SO3.

(B) Yield rate of NH3 (column diagrams) and faradic efficiency (point plots) on Au/TS (orange) and

Au-PTFE/TS (purple) at each given potential for 4 hr.

(C) The time dependence of NH3 yield (green balls) and faradic efficiency (brown stars) obtained

from Au-PTFE/TS.

(D) FTIR spectra of the post-electrolysis electrolytes reacted on the Au-PTFE/TS at different

reaction time. Black and red lines are FTIR spectra of the standard ammonium sulfate solution with

the concentration of 0.05 and 0.3 mg,mL�1 in the electrolyte, respectively.

photocathode surface and avoid contamination from ambient conditions, mild stir-

ring was conducted in the NRR process, and the exiting N2 stream was bubbled

through a water-filled vessel to form a liquid seal (Figure S26). As reported in previ-

ous work,18 Na2SO3 was added to the electrolyte to offer alternate electron donors

for scavenging the photogenerated holes and protected the Pt anode. After 24 hr

of electrolysis, the concentration of Pt in the electrolyte was �5.8 3 10�3 mg,mL�1,

evaluated by inductively coupled plasma mass spectrometry (ICP-MS). The yields

of NH3 and hydrazine (N2H4) were measured by the indophenol blue method and

the method of Watt and Chrisp (described in the Experimental Procedures), respec-

tively. The standard solutions of ammonium sulfate ((NH4)2SO4) and N2H4 with

known concentrations in the electrolyte were prepared, and the absorbance at

�650 and 460 nm was used to plot the calibration curves shown in Figures S27

and S28, respectively. No N2H4 was detected in any of the post-electrolysis

electrolytes.

The yield rate of NH3 and the faradic efficiency of Au/TS and Au-PTFE/TS at each

given potential in a 4-hr period are shown in Figure 3B. The yield rate of NH3 and

the faradic efficiency were in the range of 8.6–11.3 mg hr�1 cm�2 and 4.1%–7.4%

on Au/TS, respectively; however, after the incorporation of suitable PTFE porous

framework, both the yield rate of NH3 and the faradic efficiency were increased

markedly on Au-PTFE/TS by nearly 1.5 and 4 times, respectively. Based on the

calculation of the Nernst equation, the standard potential for the reduction of N2

to NH3 is 0.137 V versus RHE under our experimental conditions (see the

Chem 5, 617–633, March 14, 2019 623

Experimental Procedures for the calculations).34 In Figure 3B, the yield rate of NH3

and the faradic efficiency increased as the negative potential increased to �0.2 V

versus RHE, where the maximum yield rate of NH3 and the faradic efficiency were

�18.9 mg hr�1 cm�2 and 37.8% achieved on Au-PTFE/TS for 4 hr, respectively.

Beyond this negative potential, the reduction rate and faradic efficiency decreased,

which is attributed to the strong competitive adsorption of hydrogen species on the

electrode surface at high negative potential. The potential dependence of the yield

rate of NH3 and the faradic efficiency was also observed in the other photocathodes

at different reaction times (Figures S29 and S30). Furthermore, the corresponding

HER performance and hydrogen production on the photocathodes were demon-

strated by linear sweep voltammetric curves and gas chromatography detection

(Figure S31).

Ammonia production and faradic efficiency as a function of time on the Au-PTFE/TS

photocathode at�0.2 V versus RHE is shown in Figure 3C. In comparison to the con-

trol experiments that took place in the dark, replaced N2 with Ar, or removed the

Au co-catalysts (Figure S32), the NH3 production on Au-PTFE/TS was increased pro-

gressively with the increase in reaction time, implying that the NH3 originated from

PEC NRR of Au-PTFE/TS and not from contamination. NH3 production reached a

maximum (�210.5 mg cm�2) on Au-PTFE/TS in a 24-hr period and decayed slightly

to �192.5 mg cm�2 with electrolysis over 48 hr. Nevertheless, the faradic efficiency

decreased gradually with the increased reaction time when the reaction time was

more than 4 hr. This phenomenon can be ascribed to the decrease in the PEC activity

of the photocathode and the increase in NH3 concentration in the electrolyte during

the NRR process. For practical use, evaluating the durability of the photocathodes is

critical. As shown in Figure S33, Au-PTFE/TS demonstrates a good PECNRR stability

for �11.5 hr. Furthermore, FTIR spectrum (Figure S34) and FESEM images with EDS

data (Figure S35) revealed the maintenance of the PTFE porous framework and Au

NPs after the electrolysis for 11.5 hr. However, the loss of the Au-PTFE layer and

the photooxidation of the Si substrate (Figure S36) were found on the surface of

Au-PTFE/TS after 24 hr of electrolysis, which can be the dominating factors in the

degradation of the photocathode in long-term PEC NRR.

To further determine the generation of NH3 unambiguously, FTIR spectra with

ATR was applied to collect the information of post-electrolysis electrolyte with

the identical thickness of liquid film after subtracting the background spectrum

of water. As shown in Figure 3D (red line), two sharp peaks at �2,950 and

2,850 cm�1 arose from the characteristic N-H stretching of NH4+, and the other

peaks at �1,710 and �1,560 cm�1 were attributed to the s(N–H) bending.13,27

After electrolysis for 4 hr, FTIR spectra of the post-electrolysis electrolytes at

�0.2 V versus RHE showed the four peaks reflecting the signature of NH4+. The

intensity of the peaks increased with the increase in reaction time, suggesting

the increased production of NH3 PEC NRR. Moreover, an ammonia-ammonium

ion selective electrode (ISE) was also consolidated for the NH3 production via

PEC NRR. The time and potential dependence of NH3 yield on Au-PTFE/TS under

illumination is in concordance with the results mentioned above, whereas the NH3

production is negligible in the dark (Figures S37–S39). An 15N isotopic labeling

experiment was also performed to qualitatively verify the N source of the NH3

yielded by PEC NRR. In the 1H nuclear magnetic resonance (1H NMR) spectra

(Figure S40), a doublet coupling for 15NH4+ was observed in the electrolyte

when 15N2 was supplied as the feeding gas.2,34 Thus, Au-PTFE/TS with the aero-

philic-hydrophilic heterostructure is very effective for PEC N2-to-NH3 fixation

under ambient conditions.

624 Chem 5, 617–633, March 14, 2019

Calculation and Reaction Mechanism

To further supplement and support the experimental findings, density functional

theory (DFT) calculation was carried out to assess the impact of the aerophilic-

hydrophilic heterostructure on the activation energy barriers and the thermody-

namics of the PEC N2-to-NH3 fixation. Generally, there are two reaction mechanisms

for NRR, namely a dissociative and an associative mechanism. According to the

calculations, the dissociative mechanism is not feasible because the dissociation

of N2 on the Au surface is an extremely high endothermic process.35 On the other

hand, the associative mechanism can be divided into the distal and alternating path-

ways, which is inspired by the proposed mechanism for N2-fixing nitrogenase in

biochemistry.36 Multiple proton-coupled electron transfer (PCET) reactions are

required for the completion of N2 into NH3. For the distal pathway, one N atom of

N2 would be hydrogenated first and then released as NH3, being energetically

more favorable than the alternating pathway that simultaneously produced two

NH3 molecules by hydrogenating two N atoms.37 The free energy profiles at various

states along the distal pathway on Au/TS and Au-PTFE/TS are given in Figure 4A.

The first electron-transfer step for the formation of *NNH species is the rate-

determining step (RDS). The DFT calculation shows that the free energy change

(DG) of N2 hydrogenation to *NNH on Au-PTFE/TS is 2.37 eV, smaller than the

2.52 eV obtained on Au/TS, indicating that the *NNH would be easily generated

on Au-PTFE/TS. The reduced DG of *NNH is most likely related to the unique aero-

philic-hydrophilic heterostructure of Au NPs-PTFE porous framework, leading to the

dramatically increased PEC NRR performance. To gain deeper insights, the plot of

charge-density difference (Figure 4B) was constructed by subtracting the calculated

electronic charge of Au-PTFE/TS from that of individual PTFE and Au/TS. It is

clearly seen that the charge density is redistributed between PTFE and Au/TS with

a considerable charge transfer from PTFE to Au/TS, which is in agreement with the

presence of Au1+ from XPS data (Figure 1). The result suggests that Au-PTFE/TS

can give rise to a stronger binding strength of *NNH species than Au/TS and

consequently shows an excellent catalytic activity. Additionally, the *H specie

(DGH) adsorbed in PTFE-Au and Au is 0.34 and 0.37 eV, respectively. This means

that once the *H intermediate is generated in the Au1+ site, it clings to PTFE frame-

work and will rapidly proceed to the formation of *NNH species through a proton-

coupled electron transfer process (*H + e� + N2 / *NNH). However, in our system,

a portion of naked PTFE without Au is unfavorable for absorbing *H species, which

plays an important role in controlling the rate of HER by limiting the proton

concentrations.

According to the experimental results and theory calculations, the promotion mech-

anism of PEC NRR on Au-PTFE/TS by the aerophilic porous framework and hydro-

philic metal NPs is illustrated in Figure 4C. The Si wafer yields the photogenerated

electron-hole pairs under illumination, and then the electrons reach the photo-

cathode surface to induce the N2-to-NH3 fixation. As a result of the hydrophobic

properties, most possibly the solid-liquid contact happens in the Au NPs depositing

on the tip of PTFE porous framework, where the active protons are obtained even in

low current. At the same time, the overall proton activities, i.e., the adsorption of H+,

at the Au-PTFE/TS are reduced because of the hydrophobic surface. The active

proton transfer occurs from the outside to the inside of the PTFE framework via stir-

ring and N2 bubbling. The close proximity of active protons and N2 molecules to

the Au NPs on the surface of PTFE framework facilitates the intermolecular interac-

tions for hydrogenating N2 and overcoming the RDS of N2-to-NH3 fixation when a

suitable electron energy is dedicated. It is imaged that PTFE porous framework

renders a stable N2 layer under aqueous media resulting in a high-pressure N2

Chem 5, 617–633, March 14, 2019 625

Figure 4. Computational Studies and Proposed Reaction Mechanism

(A) Free energy profile of NRR on Au/TS and Au-PTFE/TS. The inset is the schematic depiction of

distal mechanisms for NRR on Au-PTFE/TS.

(B) The charge density difference of Au-PTFE/TS. The green and pink isosurfaces represent charge

accumulation and depletion, respectively.

(C) Scheme of NRR enhancement by introducing the aerophilic-hydrophilic hierarchical structure

on Si-based photocathode. The proposed reaction route for NRR principally includes the formation

of *H (blue box with dashed line), the hydrogenation of N atom (pink box with dot-dashed line), and

the release of NH3 (brown box with dotted line).

microenvironment on the TS surface. Compared with the suppression of HER,

the release of NH3 can be eventually and favorably performed in the Au NPs coating

the TS surface because of high N2 concentration and electron energy. As a result, the

integrated aerophilic-hydrophilic heterostructure of the Au NPs’ highly dispersed

PTFE porous framework on Si provide a superior catalytic activity and selectivity

toward the PEC reduction of N2 to NH3.

Conclusion

In summary, the concept of forming an aerophilic-hydrophilic heterostructured

Si-based photocathode enables advances in the combination of high conversion

efficiency (37.8%) and high production rate (�18.9 mg hr�1 cm�2) of photoelectro-

chemical N2-to-NH3 fixation. The aerophilic PTFE porous framework provides a

N2-rich concentration on the photocathode surface to tip the reaction balance

toward the synthesis of NH3. At the same time, Au NPs on the PTFE framework

show excellent catalytic activity to efficiently reduce the energy barriers of N2 reduc-

tion. The high photoelectrical-to-chemical power conversion efficiency, when

coupled with the production rate, indicates a promising platform for N2 reduction

applications under mild conditions with renewable solar.

626 Chem 5, 617–633, March 14, 2019

EXPERIMENTAL PROCEDURES

Sample Preparation

A 500-mm-thick, boron-doped, single-side-polished, (100)-oriented, p-type Si wafer

with a resistivity of 1–10 U,cm (Figure S41) was first cleaned sequentially in an ultra-

sonic bath of acetone, ethanol, and distilled (DI) water for 20 min. The clean Si wafer

was immersed in a 5 wt% HF solution for 5 min to remove the native SiO2 layer and

subsequently rinsed with DI water and dried by compressed nitrogen gas. Before

depositing, the Si surface was further cleaned with Ar plasma treatment. A thin Ti

metal layer was deposited on the Si wafer with a magnetron sputtering (MS) appa-

ratus (Chuangshiweina Co. Ltd., MSP-3200) to sputter the planar round metal Ni

target (purity > 99.5 wt %) in pure Ar (99.99%) atmosphere at room temperature

by the power of 100 W for 30 s. Thereafter, the Ti coated Si wafer was transferred

to a vacuum evaporation system (Shenyang Kejing Co. Ltd., GSL-1800X-ZF4) for

fabricating porous poly(tetrafluoroethylene) (PTFE) layer via evaporating PTFE

powder (Aladdin reagent, average grain size of 5 mm) at 110 A for 30 min. Aiming

to explore the effect of PTFE layer on N2 reduction reaction (NRR) performance of

Si-based photocathode, the weight of PTFE powder was 0, 0.25, 0.5, 0.75, and

1.0 g corresponding to Ti-Si (labeled as TS), 0.25 g PTFE/Ti-Si (labeled as

PTFE_0.25/TS), 0.5 g PTFE/Ti-Si (labeled as PTFE/TS), 0.75 g PTFE/Ti-Si (labeled

as PTFE_0.75/TS), and 1.0 g PTFE/Ti-Si (labeled as PTFE_1/TS), respectively.

After evaporation, Au nanoparticles were grown on the surface of the samples by

MS to sputter planar round metal Au targets (purity > 99.9 wt %) at room tempera-

ture by the power of 25 W for 10 s.38 In such conditions, the samples with Au

nanoparticles were simply denoted as Au/TS, Au-PTFE_0.25/TS, Au-PTFE/TS,

Au-PTFE_0.75/TS, and Au-PTFE_1/TS.

The back sides of all samples were first polished and then deposited with an Au layer

of�300 nm thickness and connected to a metal Cu belt, forming an ohmic back con-

tact. Silver paint was applied to affix the Cu belt. After drying, the entire back side

and partial front side of the Si-based electrodes were encapsulated in epoxy, estab-

lishing an exposed active area of �0.1 cm2. Calibrated digital images and ImageJ

were used to determine the geometric area of the exposed electrode surface

defined by epoxy.

Physicochemical Characterization

To investigate the crystalline structure of the samples, a Rigaku diffractometer

(Rigaku Ultima IV) using Cu Ka radiation (l = 0.15406 nm) was applied via 2q X-ray

diffraction (XRD) scans with the grazing angle of 1� at the scan rate of 1�,min�1.

FTIR (Thermo Nicolet is10) at a grazing angle of 80� was used to determine the pres-

ence of PTFE layer on the sample surface. The chemical composition of the sample

was analyzed by X-ray photoelectron spectroscopy (XPS, Thermo escalab 250XI)

with monochromated Al Ka radiation on at a pass energy of 29.4 eV. All binding en-

ergies were referenced to the C1s peak (284.8 eV) arising from adventitious carbon.

The atomic force microscopy (AFM) images presenting both the surface morphology

and microscopic I-V curves were collected by Nanocute SII scanning probe micro-

scopy operated in contact and electric models. To explore the microstructure and

composition of the samples before and after PEC measurements, field emission

scanning electron microscopy (FESEM, Magellan 400) with energy dispersive X-ray

spectroscopy (EDS) was employed to observe the surface and cross section of the

film. High-resolution transmission electron microscopy (HRTEM, Tecnai G2 F20

S�Twin) was employed to further analyze the Au nanoparticles coated on the

PTFE porous framework. The concentration of Pt in the electrolyte solution was

Chem 5, 617–633, March 14, 2019 627

confirmed by inductively coupled plasma mass spectrometry (ICP-MS) with a

Thermo Scientific iCAP-Q instrument.

The optical transmittance characteristics were monitored on a UV-visible (vis)-near-

infrared spectrophotometer (Hitachi, UV-4100) at normal incidence from 350 to

2600 nm with an integrating sphere. The Kubelka-Munk theory is generally used

for the analysis of diffuse reflection (R) spectra to obtain the absorption coefficient

(a) of the samples as follows:

FðRÞ= ð1� RÞ22R

ya; (Equation 1)

where F(R) is the Kubelka-Munk function. The optical energy band gap of the sample

has been estimated with the classical relation of optical absorption

ahn=B�hn� Eg

�m; (Equation 2)

where B, Eg, and hn denote the band tailing parameter, the optical band gap, and

the photon energy, respectively. The value ofm should be taken as 2, a characteristic

value for the indirect allowed transition that dominates over the optical absorption.

To probe the orientation of interfacial water molecules, the spectra of water mole-

cules near the film surface was recorded by a Thermo Nicolet iS 50 FTIR spectrom-

eter in conjunction with an attenuated total reflection accessory (ATR). In the

spectrometer, the ATR was used for analyzing thin water films on the sample surface

and consisted of a diamond hemispherical ATR crystal and a 45� angle of incidence.

Prior to each run, we cleaned the diamond crystal to avoid contamination. For each

measurement, we sandwiched a 10-mL water droplet between the sample and the

diamond crystal by using the built-in pressure applicator in conjunction with a

slip-clutch and a torque screwdriver. Interfacial water spectra were generated by

taking the absorbance spectrum of water against the background spectrum of the

clean diamond with air and water. The technique allowed analysis of a sufficiently

thin film of water (�300 nm)33 and yielded spectra that contained information about

the surface as well as of bulk water.32

Contact Angle Measurements

The wettability of the samples was determined bymeasuring the CA of an electrolyte

droplet on the sample surface. The method of digital video image was used to pro-

cess the sessile droplets by a CA apparatus (Chengde Dingsheng Testing Machine

Co. Ltd, JY-82A) in ambient air at room temperature. A CCD camera with space res-

olution 1,2803 1,024 and color resolution 256 gray levels was applied to capture the

droplet images. A droplet (�5 mL) of 0.05 M H2SO4 electrolyte with 0.05 M Na2SO3

was injected onto the surface with a 1-mL micro-injector. The CA values for each film

before or after solar light irradiation (produced by a Xe lamp with 300 W power and

wavelength range from 250 to 2,000 nm) for 30 min were averaged from five mea-

surements. Furthermore, the nitrogen (N2)-bubble CA with the volume of �1 mL

was performed by the captive bubble method (Dataphysics OCA20) and was

defined as the observed equilibrium CA of electrolyte around the pinned bubbles

on the sample surface, in line with the previous work.23,39

The interaction force between the N2 bubble and photoelectrode interface was as-

sessed with a high-sensitivity micro-electromechanical balance system (Dataphysics

DCAT11, Germany).23,39 A N2 bubble was suspended on a metal ring pre-treated

by hydrophobic fluorine silane under an acid solution (0.05 M H2SO4). The photo-

electrode surface was brought into contact with N2 bubble at a moving rate of

628 Chem 5, 617–633, March 14, 2019

0.02 mm,s�1. Subsequently, when the surfaces left the N2 bubble after contact, the

balance force increased gradually and reached a critical force. Finally, the surfaces

broke away from theN2 bubble, and the cycle of forcemeasurement was completed.

The critical force hysteresis to which the N2 bubble was subjected can be regarded

as the adhesive force between the photoelectrode interface and N2 bubble.

Photoelectrochemical Nitrogen Reduction Reaction Measurements

All of the chemical reagents in this study were of analytical grade and were supplied

by Aladdin (USA). A three-electrode sealed cell (that is, Si-based photocathode as

the working electrode, Pt wire as the counter electrode, and Ag/AgCl as the refer-

ence electrode) was implemented for photoelectrochemical (PEC) nitrogen reduc-

tion reaction (NRR) in a PEC 1000 system (PerfectLight Co. Ltd.) with illumination

of solar light (AM 1.5G, 100 mW,cm�2) and a solar simulator (optical fiber source,

FX300). Before each measurement, the solar simulator intensity was calibrated

with a reference silicon solar cell and a readout meter for the solar simulator

irradiance (PerfectLight Co. Ltd., PL-MW 200). 80 mL, 0.05 M aqueous H2SO4 with

0.05 M Na2SO3 was set as the electrolyte. To avoid the loss of Pt, sodium sulfite

was added to the electrolyte used to provide an alternate electron donor to scav-

enge the photogenerated holes and hence protect the Pt anode.18 Additionally,

in H2SO4-Na2SO3 aqueous solution, the product of the overall reaction can be

ammonium sulfate, N2 + 3H2SO3 + 3H2O / 2H2SO4 + (NH4)2SO4, which is a form

of ammonia commonly used as a fertilizer. Photoelectrochemical measurements

were conducted with a CHI 630E electrochemical workstation with illumination at

�25�C. The NRR activity of a photocathode was evaluated with controlled potential

electrolysis and reaction time in the electrolyte. During each test, the electrolyte was

continuously bubbled with N2 at a flow rate of 2 sccm and was agitated with a stirring

bar at a stirring rate of about 300 rpm. Before NRR, N2 was passed through the

electrolyte for 30 min to remove O2. The exiting N2 gas stream was bubbled through

a water-filled collector vessel to construct a sealed reaction. Readings for the Ag/

AgCl electrode were converted to reversible hydrogen electrode (RHE) according

to the following relationship.

EðRHEÞ=EðAg=AgClÞ+ 0:197 V+ 0:0593pH: (Equation 3)

Quantification and Analysis of Ammonia and Hydrazine

The NH3 produced was quantitatively determined by the indophenol blue method40

and an ammonia-ammonium ISE (Bante Instruments, NH3-US). In the indophenol

blue method, 1 mL of the reaction solution was first pipetted from the post-

electrolysis electrolyte. Afterward, the reaction solution was mixed with 1 mL of a

1 M NaOH solution containing salicylic acid and sodium citrate, 1.5 mL of 0.05 M

NaClO, and 0.1 mL of 1 wt % C5FeN6Na2O (sodium nitroferricyanide). The mixture

was gently agitated for 30 s and was then allowed to stand for 2 hr to ensure com-

plete color development. The UV-vis spectrometer was used to measure the

absorbance of the mixture at �650 nm. The concentration-absorbance curves

were calibrated with standard ammonium sulfate solution with a series of concentra-

tions in the H2SO4-Na2SO3 electrolyte. The fitting curve (y = 0.3892x – 0.0009,

R2 = 0.9995) shows good linear relation of absorbance value with NH3 concentration

by three times independent calibrations.

Furthermore, the ISE probe was calibrated with standard ammonia solutions con-

taining the sacrificial agent (at concentrations lower than the electrode manufac-

turer’s salinity limit). To avoid the interference of the residual solution, the ISE probe

was immersed in the deionized water with stirring to achieve the surface potential of

201.2 mV (initial value of surface potential in the deionized water) before and after

Chem 5, 617–633, March 14, 2019 629

testing. The potential-log(concentration) curves were calibrated with standard

ammonium sulfate solution with a series of concentrations in the H2SO4-Na2SO3

electrolyte. The fitting curve (y = �56.857x + 100.774, R2 = 0.9968) shows good

linear relation of potential value with log(concentration) by three times independent

calibrations.

The hydrazine presented in the electrolyte was estimated by themethod ofWatt and

Chrisp.41 Typically, 2 mL of the electrolyte solution was taken out and then mixed

with 2 mL of the coloring solution (4 g of p-dimethylaminobenzaldehyde

dissolved in 20 mL of concentrated sulfuric acid and 200 mL of ethanol). After gently

stirring for 20 min, the absorption spectra of the resulting solution were acquired

with the UV-vis spectrophotometer. The solutions of N2H4 with known concentra-

tions in the H2SO4-Na2SO3 solution were set as calibration standards, and the absor-

bance at �460 nm was used to plot the calibration curve (y = 0.7157x � 0.0079,

R2 = 0.9993).

To further prove the NH3 production, the post-electrolysis electrolytes were

measured by FTIR-ATR. During each testing, the thicknesses of the solutions are

the same. All spectra were presented in transmittance. Moreover, an isotopic label-

ing experiment used 15N2 enriched gas (98 atom% 15N) as the feeding gas to clarify

the source of ammonia. After PEC NRR at �0.2 V versus RHE for 4 hr, 20 mL of the

electrolyte was taken out, and then concentrated to 5 mL by heating at �70�C.Subsequently, 0.9 mL of the resulting solution was taken out and mixed with

0.1 mL D2O containing 100 ppm dimethyl sulfoxide as an internal standard for 1H

nuclear magnetic resonance measurement (Bruker AvanceIII HD500).

The ammonia yield rate (r) and faradic efficiency (FE) were calculated by the

following equations:

rðNH3Þ= ½NH3�3V

t3A; (Equation 4)

FEðNH3Þ= 33 96;4853 ½NH3�3V

MNH33Q

; (Equation 5)

where [NH3] is the measured NH3 concentration, V is the volume of the electrolyte

(80 mL), t is the reaction time, A is the geometric area of photocathode, MNH3 is

the molecular weight, and Q is the total charge passed through the photocathode.

The values of the absorbance at �650 nm, [NH3], A, and Q during PEC NRR pro-

cesses are listed in Tables S1–S3.

The equilibrium potential at 298.15 K under our experimental conditions was calcu-

lated with the Nernst equation under the assumption of 1 atm of N2 and 0.1 mM

NH4OH in the solution.

N2ðgÞ+ 2H2O+ 6H+ + 6e�/2NH4OH�aq�

DG0 = � 33:8 kJ mol�1 (Equation 6)

E0 = � DG0�nF = 0:058 V;

where n = 6 is the number of electrons transferred in the reaction and F is the faraday

constant.

E =E0 � RT

6Fln

½NH4OH�2

½H+ �6!+ 0:059V3pH: (Equation 7)

The equilibrium potential is 0.137 V versus RHE in 0.05 M H2SO4 solution.

630 Chem 5, 617–633, March 14, 2019

Computational Method

All DFT calculations were performed with the plane-wave technique implemented in

Vienna ab initio simulation package (VASP).42,43 The ion-electron interaction was

described according to the projector-augmented plane wave (PAW) approach.44

The generalized gradient approximation (GGA) expressed by Perdew-Burke-

Ernzerhof (PBE) exchange-correlation functional45 and a 420 eV cutoff for the

plane-wave basis set were adopted in all the computations. In this work, we built

three-layer (4O3 3 3) supercell totally 72 Au atoms to model the Au (111) surface,

namely Au/TS. To model Au-supported PTFE framework (Au-PTFE/TS), we placed

five C2F4 repeat-units (C10F22 chains) on top of above mentioned Au (111) slab. In

the calculations, the bottom two layers were fixed at their bulk positions, whereas

the remaining atoms were allowed to relax. The convergence threshold was set as

10�4 eV in energy and 0.04 eV A�1 in force. The Brillouin zone was sampled with a

2 3 5 3 1 Monkhorst-Pack mesh for k-point sampling. The PBE-D3 method was

adopted to describe the van der Waals interactions. The solvent effect on adsor-

bates was simulated with the Poisson-Boltzmann implicit solvation model with a

dielectric constant of 80.46

The free energy (G) of each species is estimated at T = 298 K according to

G=EDFT +EZPE � TS; (Equation 8)

where EDFT, EZPE, and S refer to the DFT total energy, zero-point energy, and en-

tropy, respectively. For adsorbed intermediates, EZPE and S were determined by

vibration frequencies calculations, where all 3N degrees of freedom are treated as

harmonic oscillator approximations with neglecting contributions from the slab.

The values of these molecules were taken from the NIST database (http://cccbdb.

nist.gov/). The contribution of zero-point energy and entropy corrections to G is

provided in Table S4.

SUPPLEMENTAL INFORMATION

Supplemental Information includes 41 figures and 4 tables and can be found with

this article online at https://doi.org/10.1016/j.chempr.2018.12.003.

ACKNOWLEDGMENTS

The authors are grateful to the National Natural Science Foundation of China

(51402100, 21573066, 21825201, and 21805080), the China Postdoctoral Science

Foundation, the Provincial Natural Science Foundation of Hunan (2016JJ1006 and

2016TP1009), the Shanghai Sailing Program (17YF1429800), and the Australian

Research Council (DP180100568 and DP180100731). We acknowledge the facilities

and scientific and technical assistance of the Curtin University Microscopy & Micro-

analysis Facility and theWA X-Ray Surface Analysis Facility. We thank Dr. Jean-Pierre

Veder from the John de Laeter Centre for assistance with XPS measurements. Part of

this research was undertaken with the EM instrumentation (ARC LE140100150) and

XPS instrumentation (ARC LE120100026) at the John de Laeter Centre of Curtin

University.

AUTHOR CONTRIBUTIONS

J.Z. and S.W. conceived the ideas, designed the research, and oversaw the entire

project. J.Z. and Y. Lyu synthesized the catalysts, conducted electrochemical mea-

surements, and analyzed the data. R.W., Y.Z., H.L., and C.C. carried out the quanti-

fication and analysis of NH3 in the electrolyte. M.Q. and Y. Lyu finished the

Chem 5, 617–633, March 14, 2019 631

calculation in theory. J.Z., Y. Lyu, M.Q., H.Z., Y. Li, S.P.J., and S.W. co-wrote the pa-

per and commented on the manuscript.

DECLARATION OF INTERESTS

The authors declare no competing interests.

Received: September 4, 2018

Revised: October 2, 2018

Accepted: December 3, 2018

Published: January 17, 2019

REFERENCES AND NOTES

1. Service, R.F. (2014). New recipe producesammonia from air, water, and sunlight. Science345, 610.

2. Chen, G.F., Cao, X., Wu, S., Zeng, X., Ding, L.X.,Zhu, M., and Wang, H. (2017). Ammoniaelectrosynthesis with high selectivity underambient conditions via a Li+ incorporationstrategy. J. Am. Chem. Soc. 139, 9771–9774.

3. Licht, S., Cui, B., Wang, B., Li, F.-F., Lau, J., andLiu, S. (2014). Ammonia synthesis by N2 andsteam electrolysis in molten hydroxidesuspensions of nanoscale Fe2O3. Science 345,637–640.

4. Liu, J., Kelley, M.S., Wu, W., Banerjee, A.,Douvalis, A.P., Wu, J., Zhang, Y., Schatz, G.C.,and Kanatzidis, M.G. (2016). Nitrogenase-mimic iron-containing chalcogels forphotochemical reduction of dinitrogen toammonia. Proc. Natl. Acad. Sci. USA 113,5530–5535.

5. Song, Y., Johnson, D., Peng, R., Hensley, D.K.,Bonnesen, P.V., Liang, L., Huang, J., Yang, F.,Zhang, F., Qiao, R., et al. (2018). A physicalcatalyst for the electrolysis of nitrogen toammonia. Sci. Adv. 4, e1700336.

6. Li, C., Wang, T., Zhao, Z.J., Yang, W., Li, J.F., Li,A., Yang, Z., Ozin, G.A., and Gong, J. (2018).Promoted fixation of molecular nitrogen withsurface oxygen vacancies on plasmon-enhanced TiO2 photoelectrodes. Angew.Chem. Int. Ed. 57, 5278–5282.

7. van der Ham, C.J.M., Koper, M.T.M., andHetterscheid, D.G.H. (2014). Challenges inreduction of dinitrogen by proton and electrontransfer. Chem. Soc. Rev. 43, 5183–5191.

8. Medford, A.J., and Hatzell, M.C. (2017).Photon-driven nitrogen fixation: currentprogress, thermodynamic considerations, andfuture outlook. ACS Catal. 7, 2624–2643.

9. Guo, C., Ran, J., Vasileff, A., and Qiao, S.Z.(2018). Rational design of electrocatalysts andphoto(electro)catalysts for nitrogen reductionto ammonia (NH3) under ambient conditions.Energy Environ. Sci. 11, 45–56.

10. Li, J., Li, H., Zhan, G., and Zhang, L. (2017). Solarwater splitting and nitrogen fixation withlayered bismuth oxyhalides. Acc. Chem. Res.50, 112–121.

11. Li, S.J., Bao, D., Shi, M.M., Wulan, B.R., Yan,J.M., and Jiang, Q. (2017). Amorphizing of Aunanoparticles by CeOx–RGO hybrid supporttowards highly efficient electrocatalyst for N2

632 Chem 5, 617–633, March 14, 2019

reduction under ambient conditions. Adv.Mater. 29, 1700001.

12. Zhao, Y., Zhao, Y., Waterhouse, G.I.N., Zheng,L., Cao, X., Teng, F., Wu, L.Z., Tung, C.H.,O’Hare, D., and Zhang, T. (2017). Layered-double-hydroxide nanosheets as efficientvisible-light-driven photocatalysts fordinitrogen fixation. Adv. Mater. 29, 1703828.

13. Li, H., Shang, J., Ai, Z., and Zhang, L. (2015).Efficient visible light nitrogen fixation withBiOBr nanosheets of oxygen vacancies on theexposed {001} facets. J. Am. Chem. Soc. 137,6393–6399.

14. Liu, Y., Su, Y., Quan, X., Fan, X., Chen, S., Yu, H.,Zhao, H., Zhang, Y., and Zhao, J. (2018). Facileammonia synthesis from electrocatalyticN2 reduction under ambient conditions onN-doped porous carbon. ACS Catal. 8,1186–1191.

15. Zhu, D., Zhang, L., Ruther, R.E., and Hamers,R.J. (2013). Photo-illuminated diamond as asolid-state source of solvated electrons inwater for nitrogen reduction. Nat. Mater. 12,836–841.

16. Zhou, F., Azofra, L.M., Ali, M., Kar, M., Simonov,A.N., McDonnell-Worth, C., Sun, C., Zhang, X.,and MacFarlane, D.R. (2017). Electro-synthesisof ammonia from nitrogen at ambienttemperature and pressure in ionic liquids.Energy Environ. Sci. 10, 2516–2520.

17. Liu, C., Sakimoto, K.K., Colon, B.C., Silver, P.A.,and Nocera, D.G. (2017). Ambient nitrogenreduction cycle using a hybrid inorganic-biological system. Proc. Natl. Acad. Sci. USA114, 6450–6455.

18. Ali, M., Zhou, F., Chen, K., Kotzur, C., Xiao, C.,Bourgeois, L., Zhang, X., and MacFarlane, D.R.(2016). Nanostructured photoelectrochemicalsolar cell for nitrogen reduction using plasmon-enhanced black silicon. Nat. Commun. 7,11335.

19. Lee, H.K., Koh, C.S.L., Lee, Y.H., Liu, C., Phang,I.Y., Han, X., Tsung, C.K., and Ling, X.Y. (2018).Favoring the unfavored: selectiveelectrochemical nitrogen fixation using areticular chemistry approach. Sci. Adv. 4,eaar3208.

20. Mukherjee, S., Cullen, D.A., Karakalos, S., Liu,K., Zhang, H., Zhao, S., Xu, H., More, K.L.,Wang, G., and Wu, G. (2018). Metal-organicframework-derived nitrogen-doped highlydisordered carbon for electrochemical

ammonia synthesis using N2 and H2O inalkaline electrolytes. Nano Energy 48, 217–226.

21. Lv, C., Yan, C., Chen, G., Ding, Y., Sun, J., Zhou,Y., and Yu, G. (2018). An amorphous nobel-metal-free electrocatalyst that enablesnitrogen fixation under ambient conditions.Angew. Chem. Int. Ed. 57, 6073–6076.

22. Dinh, C.T., Burdyny, T., Kibria, M.G.,Seifitokaldani, A., Gabardo, C.M., Garcıa deArquer, F.P., Kiani, A., Edwards, J.P., De Luna,P.D., Bushuyev, O.S., et al. (2018). CO2

electroreduction to ethylene via hydroxide-mediated copper catalysis at an abruptinterface. Science 360, 783–787.

23. Lu, Z., Xu,W., Ma, J., Li, Y., Sun, X., and Jiang, L.(2016). Superaerophilic carbon-nanotube-arrayelectrode for high-performance oxygenreduction reaction. Adv. Mater. 28, 7155–7161.

24. Liu, R., Zheng, Z., Spurgeon, J., and Yang, X.(2014). Enhanced photoelectrochemical water-splitting performance of semiconductors bysurface passivation layers. Energy Environ. Sci.7, 2504–2517.

25. Zhao, J., Cai, L., Li, H., Shi, X., and Zheng, X.(2017). Stabilizing silicon photocathodes bysolution-deposited Ni-Fe layered doublehydroxide for efficient hydrogen evolution inalkaline media. ACS Energy Lett. 2, 1939–1946.

26. Peng, C., Chen, Z., and Tiwari, M.K. (2018). All-organic superhydrophobic coatings withmechanochemical robustness and liquidimpalement resistance. Nat. Mater. 17,355–360.

27. Yao, Y., Zhu, S., Wang, H., Li, H., and Shao, M.(2018). A spectroscopic study on the nitrogenelectrochemical reduction reaction on goldand platinum surfaces. J. Am. Chem. Soc. 140,1496–1501.

28. Zheng, J., Lyu, Y., Wang, R., Xie, C., Zhou, H.,Jiang, S.P., and Wang, S. (2018). CrystallineTiO2 protective layer with graded oxygendefects for efficient and stable silicon-basedphotocathode. Nat. Commun. 9, 3572.

29. Bourg, M.C., Badia, A., and Lennox, R.B. (2000).Gold-sulfur bonding in 2D and 3D self-assembled monolayers: XPS characterization.J. Phys. Chem. B 104, 6562–6567.

30. Zheng, J., Lyu, Y., Xie, C., Wang, R., Tao, L., Wu,H., Zhou, H., Jiang, S., and Wang, S. (2018).Defect-enhanced charge separation andtransfer within protection layer/semiconductor

structure of photoanodes. Adv. Mater. 30,e1801773.

31. Lu, Z., Zhu, W., Yu, X., Zhang, H., Li, Y., Sun, X.,Wang, X., Wang, H., Wang, J., Luo, J., et al.(2014). Ultrahigh hydrogen evolutionperformance of under-water‘‘superaerophobic’’ MoS2 nanostructuredelectrodes. Adv. Mater. 26, 2683–2687.

32. Azimi, G., Dhiman, R., Kwon, H., Paxson, A.T.,and Varanasi, K.K. (2013). Hydrophobicity ofrare-earth oxide ceramics. Nat. Mater. 12,315–320.

33. Zheng, J.Y., Bao, S.H., Guo, Y., and Jin, P.(2014). Natural hydrophobicity and reversiblewettability conversion of flat anatase TiO2 thinfilm. ACS Appl. Mater. Interfaces 6, 1351–1355.

34. Wang, J., Yu, L., Hu, L., Chen, G., Xin, H., andFeng, X. (2018). Ambient ammonia synthesis viapalladium-catalyzed electrohydrogenation ofdinitrogen at low overpotential. Nat. Commun.9, 1795.

35. Montoya, J.H., Tsai, C., Vojvodic, A., andNørskov, J.K. (2015). The challenge ofelectrochemical ammonia synthesis: a new

perspective on the role of nitrogen scalingrelations. ChemSusChem 8, 2180–2186.

36. Deng, J., Iniguez, J.A., and Liu, C. (2018).Electrocatalytic nitrogen reduction at lowtemperature. Joule 2, 846–856.

37. Skulason, E., Bligaard, T., Gudmundsdottir, S.,Studt, F., Rossmeisl, J., Abild-Pedersen, F.,Vegge, T., Jonsson, H., and Norskov, J.K.(2012). A theoretical evaluation of possibletransition metal electro-catalysts for N2reduction. Phys. Chem. Chem. Phys. 14,1235–1245.

38. Zheng, J., Bao, S., Zhang, X., Wu, H., Chen, R.,and Jin, P. (2016). Pd-MgNix nanospheres/black-TiO2 porous films with highly efficienthydrogen production by near-completesuppression of surface recombination. Appl.Catal. B 183, 69–74.

39. He, J., Hu, B., and Zhao, Y. (2016).Superaerophobic electrode with metal@metal-oxide powder catalyst for oxygen evolutionreaction. Adv. Funct. Mater. 26, 5998–6004.

40. Searle, P.L. (1984). The Berthelot or indophenolreaction and its use in the analytical chemistryof nitrogen. A review. Analyst 109, 549–568.

41. Watt, G.W., and Chrisp, J.D. (1952).Spectrophotometric method for determinationof hydrazine. Anal. Chem. 24, 2006–2008.

42. Kresse, G., and Furthmuller, J. (1996). Efficientiterative schemes for ab initio total-energycalculations using a plane-wave basis set. Phys.Rev. B 54, 11169–11186.

43. Kresse, G., and Furthmuller, J. (1996). Efficiencyof ab-initio total energy calculations for metalsand semiconductors using a plane-wave basisset. Comput. Mater. Sci. 6, 15–50.

44. Kresse, G., and Joubert, D. (1999). Fromultrasoft pseudopotentials to the projectoraugmented-wave method. Phys. Rev. B 59,1758–1775.

45. Perdew, J.P., Burke, K., and Ernzerhof, M.(1996). Generalized gradient approximationmade simple. Phys. Rev. Lett. 77,3865–3868.

46. Mathew, K., Sundararaman, R., Letchworth-Weaver, K., Arias, T.A., and Hennig, R.G. (2014).Implicit solvation model for density-functionalstudy of nanocrystal surfaces and reactionpathways. J. Chem. Phys. 140, 084106.

Chem 5, 617–633, March 14, 2019 633

Chem, Volume 5

Supplemental Information

Photoelectrochemical Synthesis

of Ammonia on the Aerophilic-Hydrophilic

Heterostructure with 37.8% Efficiency

Jianyun Zheng, Yanhong Lyu, Man Qiao, Ruilun Wang, Yangyang Zhou, Hao Li, ChenChen, Yafei Li, Huaijuan Zhou, San Ping Jiang, and Shuangyin Wang

Figure S1. Photograph of Au-PTFE/TS with ~0.1 cm2 area.

Figure S2. Crystalline structure of the photocathodes. (a) XRD patterns of flat Si (black),

Au/TS (red), Au-PTFE_0.25/TS (blue), Au-PTFE/TS (green), Au-PTFE_0.75/TS (pink) and

Au-PTFE_1/TS (yellow). (b) The calculated grain size of the photocathodes by using Scherrer

formula and (111) plane of Au.

Figure S3. FTIR diagram of Au/TS (black), Au-PTFE_0.25/TS (red), Au-PTFE/TS (blue),

Au-PTFE_0.75/TS (green) and Au-PTFE_1/TS (pink).

Figure S4. XPS survey spectra of Au/TS (orange), Au-PTFE/TS (purple) and Au-PTFE_1/TS

(green).

Figure S5. Atomic ratio of various elements on the surface of Au/TS, Au-PTFE/TS and

Au-PTFE_1/TS.

Figure S6. XPS spectrum of Ti 2p on the surface of Au/TS.

Figure S7. XPS spectra of C 1s on the surface of Au/TS (bottom), Au-PTFE/TS (middle) and

Au-PTFE_1/TS (top).

Figure S8. XPS spectra of F 1s on the surface of Au-PTFE/TS (bottom) and Au-PTFE_1/TS

(top).

Figure S9. XPS spectrum of Au 4f and the ratio of Au1+ and Au0 on the surface of the samples.

(a) Au 4f spectrum on the surface of Au-PTFE_0.25/TS. (b) Au 4f spectrum on the surface of

Au-PTFE_0.75/TS. (c) Au 4f spectrum on the surface of Au-PTFE_1/TS. (d) The ratio of Au1+

and Au0 on the sample surface.

Figure S10. Microstructural characterizations of Au/TS. (a) Top view FESEM image of Au/TS.

(b) The magnification image of the specified area in Au/TS. (c) Cross-sectional FESEM image

of Au/TS. (d) The corresponding EDS-line measurements with the blue arrow.

Figure S11. Microstructural characterizations of Au-PTFE_0.25/TS. (a) Top view FESEM

image of Au-PTFE_0.25/TS. (b) The magnification image of the specified area in

Au-PTFE_0.25/TS.

Figure S12. Microstructural characterizations of Au-PTFE_1/TS. (a) Top view FESEM image

of Au-PTFE_1/TS. (b) The magnification image of the specified area in Au-PTFE_1/TS. (c)

Cross-sectional FESEM image of Au-PTFE_1/TS. (d) The corresponding EDS-line

measurements with the blue arrow.

Figure S13. Cross-sectional TEM image of Au-PTFE/TS. The inset in TEM image is the

magnification image of the specified area by HRTEM.

Figure S14. Optical properties of the Si-based photocathodes. (a) The measured total

hemispherical optical reflectance of the Si-based photocathodes. (b) The optical absorption

coefficient as a function of the incident photon energy for indirect allowed transition for the

samples. The orange, pink, purple, light blue and red lines denote as Au/TS,

Au-PTFE_0.25/TS, Au-PTFE/TS, Au-PTFE_0.75/TS and Au-PTFE_1/TS, respectively.

Figure S15. Liquid contact angles on the surface of Au-PTFE_0.25/TS, Au-PTFE_0.75/TS,

Au-PTFE_1/TS and PTFE/TS. The insets are corresponding droplet shapes of the liquid (0.05

M H2SO4 electrolyte with 0.05 M Na2SO3) on the surface. The mean value of the liquid contact

angles on the samples were obtained over 3 times, as represented by error bars (purple line).

Figure S16. The droplet shapes of the liquid on the surface of Au-PTFE/TS after 2 h

illumination (a) and storing in ambient for two months (b).

Figure S17. The shapes of underwater N2 bubble on the surface of Au-PTFE_0.25/TS (a) and

Au-PTFE_1/TS (b), respectively.

Figure S18. Nitrogen bubble adhesion force measurements of Au/TS (a) and Au-PTFE/TS (b).

Figure S19. Original (black solid line) and deconvoluted (red dashed line) spectra of interfacial

water at the surface of the photocathodes: (a) Au-PTFE_0.25TS, (b) Au-PTFE_0.75/TS, and (c)

Au-PTFE_1/TS. The orange, purple and light blue lines stand for the peaks centered at ~3600,

~3400 and ~3200 cm-1, respectively.

Figure S20. AFM topography image (left) and tipical I-V curves of marked position (right) for

Au/TS. Positions 1-10 in AFM image match in the colored I-V curves.

Figure S21. AFM topography image (left) and tipical I-V curves of marked position (right) for

Au-PTFE_0.25/TS. Positions 1-10 in AFM image match in the colored I-V curves.

Figure S22. AFM topography image (left) and tipical I-V curves of marked position (right) for

Au-PTFE_0.75/TS. Positions 1-10 in AFM image match in the colored I-V curves.

Figure S23. AFM topography image (left) and tipical I-V curves of marked position (right) for

Au-PTFE_1/TS. Positions 1-10 in AFM image match in the colored I-V curves.

Figure S24. AFM topography image (left) and tipical I-V curves of marked position (right) for

PTFE/TS. Positions 1-10 in AFM image match in the colored I-V curves.

Figure S25. Schematic of directed Au NPs assembly under different thicknesses of PTFE

porous framework: Au/TS (left), Au-PTFE/TS (middle) and Au-PTFE_1/TS (right).

Figure S26. Photograph of the testing equipment (PEC 1000) for photoelectrochemical

nitrogen reduction reaction.

Figure S27. Absolute calibration of the indophenol blue method using ammonium sulfate

solutions of known concentration as standards. (a) UV-vis curves of indophenol assays with

NH4+ ions after incubated for 2 hours. (b) Calibration curve used for estimation of NH3 by NH4

+

ion concentration. The absorbance at ~650 nm was obtained by UV-vis spectrophotometer.

The fitting curve shows good linear relation of absorbance with NH4+ ion concentration (y =

0.3892x – 0.0009, R2 = 0.9995) of three times independent calibration curves. The inset in (b)

shows the chromogenic reaction of indophenol indicator with NH4+ ions.

Figure S28. Watt and Chrisp (para-dimethylamino-benzaldehyde) method for estimating the

hydrazine solutions of known concentration as standards. (a) UV-vis curves of various

hydrazine concentration. (b) Calibration curve used for estimation of the hydrazine

concentration (y = 0.7157x - 0.0079, R2 = 0.9993). The absorbance at ~460 nm was measured

by UV-vis spectrophotometer. The inset in (b) shows the chromogenic reaction of

para-dimethylamino-benzaldehyde indicator with the hydrazine.

Figure S29. Yield rate of NH3 on the Si-based photocathodes. (a) Yield rate of NH3 on the

Si-based photocathodes at the given potentials from 0 to -0.4 V versus RHE for 2 h. (b) Yield

rate of NH3 on the Si-based photocathodes at the given potentials from -0.1 to -0.3 V versus

RHE for 4 h. The black, red, blue, green and pink columns denote as Au/TS,

Au-PTFE_0.25/TS, Au-PTFE/TS, Au-PTFE_0.75/TS and Au-PTFE_1/TS in Figure S29a and

S29b, respectively. (c) Yield rate of NH3 on the Si-based photocathodes at -0.2 V versus RHE

for 11.5 h (black) and 24 h (red).

Figure S30. The faradaic efficiency of the Si-based photocathodes for photoelectrochemical

N2-to-NH3 fixation. (a) The faradaic efficiency of the Si-based photocathodes at the given

potentials from 0 to -0.4 V versus RHE for 2 h. (b) The faradaic efficiency of the Si-based

photocathodes at the given potentials from -0.1 to -0.3 V versus RHE for 4 h. The black, red,

blue, green and pink columns denote as Au/TS, Au-PTFE_0.25/TS, Au-PTFE/TS,

Au-PTFE_0.75/TS and Au-PTFE_1/TS in Figure S30a and S30b, respectively. (c) The faradaic

efficiency of the Si-based photocathodes at -0.2 V versus RHE for 11.5 h (black) and 24 h

(red).

Figure S31. (a) Linear sweep voltammetric (LSV) curves of Au/TS (black and blue) and

Au-PTFE/TS (red and pink) recorded in an N2 saturated (solid line) and Ar saturated (dash line)

electrolyte under ambient conditions. (b) The variation of photocurrent density of Au-PTFE/TS

at -0.2 V versus RHE in an N2 saturated electrolyte by chopping light illumination. (c) LSV

curves of Au/TS (black), Au-PTFE_0.25/TS (red), Au-PTFE/TS (blue), Au-PTFE_0.75/TS (pink)

and Au-PTFE_1/TS (dark green) in an Ar saturated electrolyte. (d) Hydrogen production of all

the photocathodes via measuring in PEC NRR for 4 h at -0.2 V versus RHE in an N2 saturated

electrolyte.

Figure S32. The time-dependence of NH3 yield obtained from Au-PTFE/TS with N2 as the

feeding gas in the light (black line with squares), Au-PTFE/TS with Ar as the feeding gas in the

light (red line with dots), Au-PTFE/TS with N2 as the feeding gas in the dark (blue line with

regular triangles), and PTFE/TS with N2 as the feeding gas in the light (pink line with inverted

triangles).

Figure S33. Photoelectrochemical nitrogen reduction reaction of the Si-based photocathodes

at -0.2 V versus RHE for 24 h. (a) Chronoamperometry results of the Si-based photocathodes

at -0.2 V versus RHE for 24 h. (b) UV-vis curves of the post-electrolysis electrolyte measured

on various Si-based photocathodes. The orange, purple, light blue, pink and green lines

denote as Au/TS, Au-PTFE/TS, Au-PTFE_0.25/TS, Au-PTFE_0.75/TS and Au-PTFE_1/TS,

respectively.

Figure S34. FTIR diagram of Au-PTFE/TS after 11.5 h electrolysis.

Figure S35. Microstructural characterizations of Au-PTFE/TS after 11.5 h electrolysis. (a) Top

view FESEM image of Au-PTFE/TS. (b) The magnification image of the specified area in

Au-PTFE/TS. (c) Cross-sectional FESEM image of Au-PTFE/TS. (d) The corresponding

EDS-line measurements with the blue arrow.

Figure S36. Microstructural characterizations of Au/TS and Au-PTFE/TS after 24 h electrolysis.

(a) Atomic ratio of various elements on the surface of Au/TS (orange pentagon) and

Au-PTFE/TS (purple sphere). (b) XPS spectra of Au 4f of Au/TS and Au-PTFE/TS. (c) Top view

FESEM image of Au-PTFE/TS. (d) The corresponding EDS-line measurements with the light

blue arrow.

Figure S37. Calibration curve of the ISE electrode against the standard ammonium sulfate

solution with a serious of concentrations in the H2SO4-Na2SO3 electrolyte. The fitting curve (y =

-56.857x + 100.774, R2 = 0.9968) shows good linear relation of potential value with

log(concentration) by three times independent calibrations.

Figure S38. The ISE potentials of the post-electrolysis electrolyte as a function of the reaction

conditions on Au-PTFE/TS. The green sphere denotes at -0.2 V in light, the purple star

denotes at -0.3 V in light, and the black rhombus denotes as at -0.2 V in dark.

Figure S39. The NH3 yield as a function of the reaction conditions on Au-PTFE/TS via Figure

S38. The green sphere denotes at -0.2 V in light, the purple star denotes at -0.3 V in light, and

the black rhombus denotes as at -0.2 V in dark.

Figure S40. 1H NMR spectra of the post-electrolysis electrolytes with 15N2 and 14N2 as the

feeding gas. In the 1H NMR spectra, a doublet coupling for 15NH4+ and a triplet coupling for

14NH4+ were distinguished for the 15N2- and 14N2- saturated electrolytes after electrolysis,

confirming that the NH3 was synthesized from the feeding gas.

Figure S41. Mott-Schottky plot of the planar Si from capacitance measurement as a function

of potential vs RHE under dark conditions. The donor density of the planar Si can be

calculated to be ~5.43 × 1015 cm-3, corresponding to the resistivity of Si wafer (1‒10 Ω·cm)

basically.

Table S1. Some testing results of photoelectrochemical nitrogen reduction reaction on the

Si-based photocathodes for 2 h. Thereinto, A is the geometric area of photocathode, Q the

total charge passed through the photocathode, and α the values of the absorbance at ~650

nm.

Sample Potential versus RHE (V) A (cm2) Q (c) α

Au/TS 0 0.1 0.224 0.00318

Au-PTFE_0.25/TS 0 0.093 0.152 0.00255

Au-PTFE/TS 0 0.099 0.0489 ignorable

Au-PTFE_0.75/TS 0 0.112 0.00326 ignorable

Au-PTFE_1/TS 0 0.117 0.00017 ignorable

Au/TS -0.1 0.097 0.3971 0.007

Au-PTFE_0.25/TS -0.1 0.119 0.2352 0.00726

Au-PTFE/TS -0.1 0.088 0.1001 0.00902

Au-PTFE_0.75/TS -0.1 0.146 0.08072 0.00665

Au-PTFE_1/TS -0.1 0.129 0.00419 ignorable

Au/TS -0.2 0.104 0.768 0.01091

Au-PTFE_0.25/TS -0.2 0.101 0.5595 0.0134

Au-PTFE/TS -0.2 0.121 0.249 0.0209

Au-PTFE_0.75/TS -0.2 0.103 0.147 0.01299

Au-PTFE_1/TS -0.2 0.105 0.01219 ignorable

Au/TS -0.3 0.111 1.037 0.0108

Au-PTFE_0.25/TS -0.3 0.098 0.6916 0.0123

Au-PTFE/TS -0.3 0.135 0.4479 0.0198

Au-PTFE_0.75/TS -0.3 0.123 0.365 0.0112

Au-PTFE_1/TS -0.3 0.073 0.01265 ignorable

Au/TS -0.4 0.122 6.52 0.0096

Au-PTFE_0.25/TS -0.4 0.099 3.56 0.0109

Au-PTFE/TS -0.4 0.128 2.416 0.0166

Au-PTFE_0.75/TS -0.4 0.121 1.845 0.01

Au-PTFE_1/TS -0.4 0.111 0.0419 ignorable

Table S2. Some testing results of photoelectrochemical nitrogen reduction reaction on the

Si-based photocathodes for 4 h. Thereinto, A is the geometric area of photocathode, Q the

total charge passed through the photocathode, and α the values of the absorbance at ~650

nm.

Sample Potential versus RHE (V) A (cm2) Q (c) α

Au/TS -0.1 0.08 0.6361 0.01252

Au-PTFE_0.25/TS -0.1 0.085 0.4043 0.01279

Au-PTFE/TS -0.1 0.096 0.2241 0.02073

Au-PTFE_0.75/TS -0.1 0.119 0.1784 0.01455

Au-PTFE_1/TS -0.1 0.11 0.00752 ignorable

Au/TS -0.2 0.131 1.807 0.02797

Au-PTFE_0.25/TS -0.2 0.107 1.133 0.02836

Au-PTFE/TS -0.2 0.13 0.4431 0.04701

Au-PTFE_0.75/TS -0.2 0.132 0.371 0.03288

Au-PTFE_1/TS -0.2 0.118 0.03281 0.00135

Au/TS -0.3 0.153 2.59 0.02973

Au-PTFE_0.25/TS -0.3 0.089 1.26 0.0233

Au-PTFE/TS -0.3 0.126 0.843 0.0381

Au-PTFE_0.75/TS -0.3 0.096 0.641 0.01937

Au-PTFE_1/TS -0.3 0.11 0.09485 0.00205

Table S3. Some testing results of photoelectrochemical nitrogen reduction reaction on the

Si-based photocathodes at -0.2 V versus RHE. Thereinto, t is the reaction time, A the

geometric area of photocathode, Q the total charge passed through the photocathode, and α

the values of the absorbance at ~650 nm.

Sample t (h) A (cm2) Q (c) α

Au/TS 11.5 0.077 3.86 0.0394

Au-PTFE_0.25/TS 11.5 0.131 3.25 0.00781

Au-PTFE/TS 11.5 0.13 1.18 0.1045

Au-PTFE_0.75/TS 11.5 0.13 1.08 0.0773

Au-PTFE_1/TS 11.5 0.12 0.1003 0.0043

Au/TS 24 0.143 6.51 0.059

Au-PTFE_0.25/TS 24 0.145 5.93 0.1077

Au-PTFE/TS 24 0.122 1.91 0.124

Au-PTFE_0.75/TS 24 0.112 1.83 0.0828

Au-PTFE_1/TS 24 0.144 0.1948 0.00727

Au-PTFE/TS 48 0.15 3.52 0.1396

Table S4. Zero-point energy correction (EZPE), entropy contribution (TS) and the total free

energy correction (G - Eelec) in this study.

Species EZPE (eV) -TS (eV) G - Eelec (eV)

H2 0.27 -0.41 -0.14

N2 0.15 -0.60 -0.45

NH3 0.90 -0.60 0.30

*NNH on Au-PTFE/TS 0.47 -0.16 0.31

*NNH2 on Au-PTFE/TS 0.79 -0.22 0.57

*N on Au-PTFE/TS 0.07 -0.04 0.03

*NH on Au-PTFE/TS 0.39 -0.04 0.35

*NH2 on Au-PTFE/TS 0.70 -0.07 0.63

*NNH on Au/TS 0.38 -0.04 0.34

*NNH2 on Au/TS 0.79 -0.21 0.58

*N on Au/TS 0.08 -0.04 0.04

*NH on Au/TS 0.38 -0.04 0.34

*NH2 on Au/TS 0.70 -0.07 0.63