A new artificial photosynthetic system coupling photovoltaic electrocatalysis with solar heating catalysis

: In this work, we present a novel artificial photosynthetic paradigm with square meter (m 2 ) level scalable production by integrating photovoltaic electrolytic water splitting device and solar heating CO 2 hydrogenation device, successfully achieving the synergy of 1 sun driven 19.4% solar to chemical energy efficiency (STC) for CO production (2.7 times higher than state of the art of large-sized artificial photosynthetic systems) with a low cost (equivalent to 1/7 of reported artificial photosynthetic systems). Furthermore, the outdoor artificial photosynthetic demonstration with 1.268 m 2 of scale exhibits the CO generation amount of 258.4 L per day, the STC of ~15.5% for CO production in winter, which could recover the cost within 833 sunny days of operation by selling CO.


INTRODUCTION
Artificial photosynthesis can convert CO2 and H2O into useful fuels, chemicals (CO [1], CH4 [2], etc.) and O2 under solar irradiation, which is the most important way for carbon neutralization [3−9]. The application of large-scale artificial photosynthesis is of great significance to weaken the global warming, overcome the current energy and environmental crisis [10−12]. More recently, several large-sized artificial photosynthetic systems for CO2 utilization have been reported, e. g., the solar fuel production chain with square meters (m 2 ) scale [13], the photovoltaic electrocatalytic device with ~0.1 m 2 scale [14]. To the best of our knowledge, the highest solar to chemical energy efficiency (STC) of large-sized devices is 7.2% by the photovoltaic electrocatalytic system [14]. However, the material cost for constructing large-sized artificial photosynthetic system is too expensive to practical application, due to the using of noble metal catalysts (e.g., Ir, Pt, Rh, Ru) and the costliness of large-sized components (e.g., membranes, solar reactor) in the devices [15,16]. Therefore, it is one of the holy grails of the entire scientific and technological community to achieve a scalable artificial photosynthetic system with high STC and low cost RESEARCH ARTICLE Novel artificial photosynthesis for CO2 and H2O converted as CO and O2 As the solar driven water splitting produced H2 injected into the TiC/Cu based device loaded with 137 g of Fe SACs, 600 sccm of CO2 was simultaneously put into the TiC/Cu based device, which was controlled by mass flow controller (C50 5SLM). The TiC/Cu based device was irradiated by a xenon lamp (ZSL-4000). For the produced gas, the flow rate was tested by mass flowmeter (C50 5SLM) and the composition was tested by GC 7890A equipped with FID and TCD detectors. The CO rate (δ, mmol h -1 ) was calculated as follows: δ (mmol h -1 ) = (1000×L/24.5) (3) L was the CO flow rate (L h -1 ) and the L irradiated by 0.6, 0.8, 1 sun was 0.946, 5.170, 12.030 L h -1 , respectively. The STC calculation of sunlight driven CO2 conversion as CO The STC efficiency of novel artificial photosynthetic system for converting CO2 into CO was calculated as follows: STC = (ΔH×ε)/(I×ST×3600) (4) ΔH was the reaction Enthalpy change energy (H2O (l) + CO2 (g) → CO (g) + 1/2 O2 (g) + H2O (g), ΔH= 326.9754 kJ mol -1 ), ε (mol) was the CO generation amount per hour detected by a flowmeter, I was the light intensity (kW m -2 ), ST was the total irradiated area. The ε irradiated by 0.6, 0.8, 1 sun was 0.0386, 0.211, 0.491 mol, respectively. Since not all H2 produced from solar driven water splitting was used for CO2 hydrogenation, the irradiation area (β) of solar driven water splitting used for CO2 hydrogenation was calculated as follows: β = M/N×0.2800 m -2 (5) The M was H2 used for CO2 hydrogenation as CO, which was equal to the CO production rate of 0.0386, 0.211, 0.491 mol h -1 , under 0.6, 0.8, 1 sun irradiation, respectively. The N was the H2 production rate of 0.3931, 0.5273, 0.6714 mol h -1 , irradiated by 0.6, 0.8, 1 sun, respectively. Therefore, the β was 0.0276 m -2 , 0.1119 m -2 , 0.2047 m -2 , under 0.6, 0.8, 1 sun irradiation, respectively. And the ST = β+0.0240 m -2 (6) Therefore, the ST was 0.0516, 0.1359, 0.2287 m -2 , under 0.6, 0.8, 1 sun irradiation, respectively. Consequently, the STC was 11.3%, 17.6%, 19.4%, under 0.6, 0.8, 1 sun irradiation, respectively. The EC calculation The 1 sun driven EC of photovoltaic electrocatalytic water splitting in this work and reported photovoltaic electrocatalytic CO2 reduction was calculated as follows: EC = STCE/efficiency (7) The STCE was the solar to hydrogen chemical efficiency (19.04%) under 1 sun irradiation. The efficiency was the electric energy generation efficiency of solar cell (23.5%) under 1 sun irradiation. Therefore, the EC was calculated as 81.0%. Outdoor artificial photosynthetic system The outdoor artificial photosynthetic system consisted of two components. One component was photovoltaic electrolysis system, in which the PERC solar cell (182DCB) with 1.07 m 2 of solar irradiation area was used to power electrolytic reactor with 2.782 m 2 of Ni/Stainless steel mesh divided into 12 independent chambers in series. The mixture of 200 g KOH and 1.8 L deionized water was used as the electrolyte. The other component was solar heating system, in which a solar heating device was provided by Hebei scientist research experimental and equipment trade Co., Ltd. with the size of 4 cm in diameter and 50 cm inlength, eqquipped with a reflector of 50 cm inlength and 36 cm in width. For the production of CO, the catalysts used in solar heater were 400 g CuOx/ZnO/Al2O3. For CO production production in solar heating system, the CO2/H2 ratio was 1.5. It was required to control the flow rate to ensure the H2 consumption exceeds 95%. The data were collected by FID and TCD. The data shown in Figure. 3C were tested on December 20, 2021, in Baoding, China. The STC of outdoor artificial photosynthetic system The STC of the outdoor artificial photosynthetic system for converting CO2 into CO was calculated as follows: STC = (ΔH×ε)/(I×ST×3600×22.4) (8) ΔH was the reaction Enthalpy change energy (H2O (l) + CO2 (g) → CO (g) + 1/2 O2 (g) + H2O (g), ΔH= 326.9754 kJ mol -1 ), ε (L) was the CO generation amount per hour detected by a flowmeter, I was the outdoor solar intensity (kW m -2 ), ST was the total irradiated area of 1.268 m 2 . The cost recovery calculation We assumed that the CO production amount of the outdoor artificial photosynthetic system was 258.4 L/day. Due to the variety of CO prices, the quotation of Chae et al. reported result and North Special Gas Co., Ltd. was adopted [17], which was $6 per m 3 CO. Therefore, the income of outdoor artificial photosynthetic system for CO production was 0.2584* $6=$1.55. To achieve an income of $1291, this system required $1291/$1.55=833 sunny days, which were equivalent to the sunny days in 3.5 years, according to the weather in Baoding of 240 sunny days per year.

Conception for constructing novel artificial photosynthetic system
It is well known that the widely studied photovoltaic electrocatalytic systems contain the competition of two main reactions: H2O decomposition and CO2 hydrogenation on one system with CO2 transportation through liquid electrolytes. Although various efficient catalysts have been developed, such as metals [18−20], metal compounds [21−23], molecular complexes [24,25], photovoltaic electrocatalysis still faces two intrinsic shortcomings: one is the complex reaction processes in single catalytic site and the other is the sluggish CO2 supply through gas/liquid transportation [26−28]. Here, a new paradigm of artificial photosynthesis is proposed to separate the two reactions of water splitting (2H2O → 2H2 + O2) [29] and CO2 hydrogenation (CO2 + H2 → CO + H2O) [30−32] in space and time. There are four major advantages in this new system: (1) mature technologies can be selected for both water splitting and CO2 hydrogenation; (2) the integrated system can be easy amplified; (3) the systems for the two reactions can be optimized separately, providing a variety of possibilities for efficiency, cost and products; (4) CO2 supply can be boosted by avoiding the gas transport in liquid elelctrolytes. As shown in Figure 1, this is an integrated system in which the hydrogen generated from photovoltaic water electrolysis [33] is directly injected into the solar heating system for CO2 hydrogenation [34−36]. The CO2 transportation of this system is in gas diffusion mode at a rate of 10 -5 m 2 s -1 [37], 10000 times higher than the rate of CO2 diffused through liquid electrolytes (10 -9 m 2 s -1 ) [38] in conventional photovoltaic electrocatalytic systems [39,40], which could meet the CO2 supply for large-sized artificial photosynthetic systems. For integrating such a new artificial photosynthetic system, the two issues should be solved firstly. One is the matching problem of solar energy utilization in this system, that is, how to scientifically distribute the proportion of solar energy irradiated to the two devices to improve the STC; the second is the quality matching of hydrogen production and hydrogen consumption in the new system. [41−43], which could heat the catalysts to 318 °C under 1 kW m -2 intensity of sunlight (1 sun) irradiation to run CO2 hydrogenation ( Figure S1). This is the key for realizing the new artificial photosynthetic system, because the low solar irradiated temperature of conventional photothermal system (~80 °C, Figure S2) can not drive photothermal CO2 hydrogenation under ambient solar irradiation. As the Fe single-atom catalysts (Fe SACs, Figure S3−S7) were used as catalysts for solar heating CO2 hydrogenation, the system showed a CO generation rate of 21.14 mol m -2 h -1 under 1 sun irradiation, corresponding to 24.1% of solar to chemical energy efficiency (detailed calculation seen in Supplementary Methods, Figure S8). More interestingly, as the 1% O2-polluted H2 was used as feed gas, the efficiency of CO2 hydrogenation had little change ( Figure S8A), evidencing the robustness of solar heating catalytic system. The low requirement of hydrogen purity for solar heating catalysis enables us to simplify the photovoltaic electrocatalysis. Besides using commercialized single crystalline silicon solar cells (23.5% efficiency) as electric energy supply, the membrane was eliminated from the electrocatalytic reactor ( Figure 1) and the cheap nickel-plated stainless-steel mesh (Ni/stainless steel, Figure S9) was used as the electrodes to replace the precious electrocatalysts [18−20]. In the membrane free electrocatalytic reactor, the Ni/stainless steel's electrodes could achieve a current density of 10 mA cm −2 in 1 M KOH electrolyte at only 1.53 V ( Figure  S10). The H2 production rate of this photovoltaic electrolytic system was 2.40 mol m -2 h -1 under 1 sun irradiation ( Figure S11), equivalent to 19.04% solar to hydrogen chemical efficiency (detailed calculation seen in Supplementary Methods). It was calculated that the solar cell's electric energy to chemicals energy efficiency (EC) of this photovoltaic electrocatalytic water splitting system was 81% (detailed calculation seen in Methods). The released H2 contained ~0.8% O2, which also meets the purity requirement of solar heating CO2 hydrogenation. Based on the above experimental results, the photovoltaic electrolytic water splitting device with 2800 cm 2 of solar irradiation area and solar heating CO2 hydrogenation device with 240 cm 2 of solar irradiation area were integrated as a new type of artificial photosynthetic system with more than 3000 cm 2 of solar irradiation area in the laboratory ( Figure 1).

[INSERT FIGURE 1 HERE]
The performance of novel artificial photosynthesis Figure 2A showed that the laboratory system could produce CO with a rate of 38, 210, 491 mmol h -1 under 0.6, 0.8, 1 sun irradiation, respectively. Additionally, Figure 2B identified that this system showed a 100% selectivity for CO2 converted into CO under different intensities of solar irradiation due to the +3 oxidation state of Fe-SACs ( Figure S12) [44]. Figure 2C illustrated that the STC of new artificial photosynthetic system was increased from 11.3%, 17.4% to 19.4% along with the 0.6, 0.8 to 1 sun irradiation (Detailed calculation seen in Methods), which was 2.7 times higher than the best record value of scalable artificial photosynthesis with ~1000 cm 2 of solar irradiation area (7.2%) [14]. The CO2 reduction performance of this system was continuously tested for 6 days. The CO production rate was stable maintained at ~500 mmol h -1 ( Figure 2D) and the Fe SACs kept single atom state ( Figure S13), indicating the excellent stability of new artificial photosynthetic system.

[INSERT FIGURE 2 HERE]
The outdoor artificial photosynthetic demonstration The commercial single crystalline silicon solar cell panel (1.07 m 2 scale), membrane-free electrolytic water splitting device and factory prepared TiC/Cu based solar heating tube were used to build the outdoor artificial photosynthetic system. For maintaining the solar heating system at high temperature all day, a parabolic reflector with 0.198 m 2 of irradiated area ( Figure S14) was applied to concentrate outdoor sunlight on solar heating device ( Figure 3A). A commercial CuOx/ZnO/Al2O3 (SCST-401, Figure S15) was selected as the catalyst for solar heating reverse water-gas-shift reaction (CO2+H2→CO+H2O). In outdoor test, the membrane-free electrolytic water splitting device was driven by the silicon solar cell panel to produce H2, then the H2 and CO2 entered the solar heating system for CO2 hydrogenation ( Figure 3B). The artificial photosynthetic system for CO production was tested in December 20, 2021, with an ambient temperature of 2~13 °C and a solar irradiation intensity of 0.26−0.49 kW m -2 in the daytime in Baoding City of Hebei Province, China. As shown in Figure 3C, the CO generation occurred at 9:00 AM with a production rate of 27.9 L h -1 . After that, the CO generation rate rose to a peak value of 41.4 L h -1 at 12:00 PM and then gradually decreased to 23.6 L h -1 at 16:00 PM. The total amount of CO produced daily was up to 258.4 L. Although the solar intensity and ambient temperature are the lowest in winter, the outdoor system STC for CO production was still in the range of 15% to 15.8% throughout the operating period ( Figure 3D, detailed calculation seen in Methods). [INSERT FIGURE 3 HERE] Table 1 listed the data of new artificial photosynthetic systems and the most advanced large-sized artificial photosynthetic systems. Firstly, the size of the outdoor artificial photosynthetic system was 1.268 m 2 , and all parts can be processed in the factory, showing that the system could realize mass production directly. Secondly, the STCs of lab and outdoor systems for CO2 reduction as CO were 19.4% and 15−15.8% respectively, which was 2.7 times and 4 times higher than that of reported large-sized artificial photosynthetic systems under lab and outdoor conditions, respectively [13,14]. The total cost of outdoor demonstration was calculated as $1018 per m 2 ( Figure S16). Tab. 1 showed that the cost of large-sized artificial photosynthetic devices is too expensive to calculated cost [13,14,45]. Compared with systems that produce mixture of CO and H2 [13,46], the main product of our artificial photosynthetic system is CO. As far as we known, the cheapest cost of small-sized artificial photosynthetic device reported in literatures was ~$7200 per m 2 (Table. 1) [17], which was 7 times higher than our outdoor demonstration. With the ultra-high STC and ultra-low system cost, the system cost recovery time of the outdoor artificial photosynthetic device was calculated by selling product (CO). Referring to the price of CO ($6 per m 3 ) [17], the outdoor system could recover the cost after 833 days of operation, which corresponds to ~3.5 years (detailed calculation seen in Methods). The service life of the components in this outdoor system for CO production was generally more than 10 years, able to profitable by selling CO.

Conclusion
In this work, a novel artificial photosynthesis paradigm was proposed, in which the silicon solar cells were used to drive the membrane free electrolyzer for photovoltaic electrolytic water splitting as O2 and H2. Then, the generated H2 and CO2 were injected into the solar heating system based on a TiC/Cu based device to carry out efficient sunlight driven CO2 hydrogenation due to the high 1 sun-heating temperature of 318 °C. The photovoltaic electrolytic reactor eliminated the membrane and used the Ni-plated stainless steel mesh as the electrodes to reduce the cost. As the 240 cm 2 of solar heating CO2 hydrogenation device was integrated to 2800 cm 2 of silicon solar cell driven photovoltaic electrolytic water splitting device, the system exhibited a CO2 conversion rate of 491 mmol h -1 , an STC of 19.4%, a selectivity of 100% for CO production, under 1 sun irradiation. Moreover, an outdoor demonstration with 1.268 m 2 of solar irradiation area was constructed, A c c e p t e d which showed a cost of $1018 per m 2 , the gas production of 258.4 L per day, the STC of 15%−15.8% for CO production in winter, under ambient solar irradiation, which could neutralize device cost by selling the product of CO within 833 sunny operation days, revealing the ability for direct scalable application.

Outlook
The new artificial photosynthetic system has huge space for STC improvement and flexible product regulated ability. As the silicon solar cell was replaced by triple-junction solar cells for photovoltaic electrocatalytic water splitting, the calculated STC of new artificial photosynthetic system was as high as 28.9 %(detailed calculation seen in Supplementary Methods) , which was higher than the best STC (19.1%) of triple-junction solar cells driven artificial photosynthesis [47]. Further, this system could convert product from CO to CH4 by changing the solar heating CO2 hydrogenation catalysts as commercial Ni/Al2O3 ( Figure S17, detailed calculation seen in Methods). Therefore, our system could be a core investigating platform for scientists all over the world to realize carbon neutralization, via converting CO2 and H2O into a variety of chemicals (such as methanol, formic acid even C2+ product) by developing different catalysts. We believe this new artificial photosynthetic system will speed up the development and application of artificial photosynthesis.

Characterizations
The overall composition of the prepared samples were studied by the powder X-ray diffraction (XRD), which was performed on a lattice constant perpendicular to the slab were adopted to avoid the artificial interactions between the slab and its periodic images. During the geometry optimization, all of atoms were allowed to relax by using conjugate-gradient algorithms. Moreover, all geometries were fully optimized to reach the convergence until the Hellman-Feynman force on each ion was smaller than 0.02 eV/Å. The convergence criteria for the electronic structure was set to 10 -6 eV per atom. In addition, the DFT+U correction for strong-correlation Fe's 3d electrons was taken into account and the U-J value of 3.29 eV was used [2]. Furthermore, the DFT-D2 method of Grimme was also applied for a better description of weak van der Waals interactions [3]. The free-energy change (ΔG) is calculated as ΔG=ΔE+ΔZPE where ΔE is the total energy directly obtained from DFT calculations and ΔZPE is the change in zero-point energy.

Novel artificial photosynthetic system for CO2 and H2O converted as CH4 and O2
The CH4 production was similar to that of CO production. And the difference was follows: 1. The catalyst was changed from 137 g of Fe SACs as 100 g of Ni/Al2O3.
2. The supply of CO2 was 1/4 of the H2 supplied by the solar driven water splitting.

The STC calculation of novel artificial photosynthetic system using triple-junction solar cells
The STC was calculated as follows: STC= STCH*S1+STCE*S2/(S1+S2) The STCH of solar heating CO2 hydrogenation was 24.1% under 1 sun irradiation and the S1 was the solar irradiation area of solar heating CO2 hydrogenation (0.0240 m 2 ). As Jia et al had reported triple-junction solar cells (37.5% efficiency) driven photovoltaic electrolytic water splitting with a STCE over 30% [4], we assumed that the 1 sun driven STCE was 30% for triplejunction solar cell driven photovoltaic electrocatalytic water splitting in this system.
Since the H2 consumption (ε) for solar heating CO2 hydrogenation was 0.491 mol h -1 (equal to the rate of CO production), the irradiation area (S2) of triple-junction solar cells was calculated as S2= (ΔH*ε)/(STCE*I*3600) ΔH was the reaction Enthalpy change energy (H2O (l) → H2 (g) + 1/2 O2 (g), ΔH= 285.83 kJ/mol), I was the light intensity (1 kW m -2 ), ε was 0.491 mol h -1 , STCE was 30%. Therefore, the S2 was calculated as 0.1299 m 2 and the STC was calculated as 28.9%.  As shown in Figure S3A, Fe(NO3)3· 9H2O, citric acid, 2-methylimidazole were added into the water to form a homogeneous solution. And the solution was crosslinked as gel by the addition of ammonia. Then, an argon protected annealing process was applied to form the catalyst. Figure S3B presented the typical transmission electron microscopy (TEM) image of the asprepared catalyst. It was clearly observed that the catalyst was grown in a graphene like morphology and no clear Fe nanoparticles existed in the visual field. Powder X-ray diffraction (XRD) pattern showed that the catalyst has only a typical broaden peak assigned to graphene ( Figure S4A) [5]. Raman spectroscopy was further employed to investigate the graphitization degree of the sample ( Figure S4B). Clearly, the ID/IG ratio of Fe SACs is calculated to be 1.12, exhibiting a high degree of defective nature of the graphene in Fe SACs [6]. Atomic force microscopy (AFM) confirmed that the thickness of the sample was 3.9 nm, revealed its ultrathin nature of sample ( Figure S3C). On the other hand, elemental mapping images demonstrated the homogeneous distribution of N, C, Fe, throughout the sample ( Figure S5), confirming the existence of Fe species. To visualize the Fe species, the high angle annular dark-field scanning TEM (HAADF-STEM) was applied and Figure S3D showed that there were a lot of star like bright spots on the surface of sample, which were assigned to Fe single atoms [7−9]. Therefore, wt%. X-ray absorption spectroscope (XAS) and X-ray photoelectron spectroscope (XPS) were uesd to characterize the coordination structure of Fe in Fe SACs. As shown in the Figure S3E, the energy of Fe K-edge in Fe SACs was higher than that of FeO, Fe foil and similar to that of Fe2O3. It indicated that the valence state of Fe in Fe SACs was +3, which was also consistent with the XPS result ( Figure S6A) [10,11]. Fourier transformed EXAFS (FTEXAFS) of Fe SACs showed a peak located at 1.5 Å, corresponding to the Fe-N coordination ( Figure S3F) [12] and FTEXAFS showed no other peaks for Fe SACs, confirming the single atomic state of Fe as well as it identified that the Fe-N coordination number was 4. Since the N 1s spectrum showed that the contents of pyridine nitrogen [13] and pyrrole nitrogen [14] in Fe SACs were almost equal ( Figure S6B), so we constructed a possible model based on these data. As shown in Figure S3G, the Fe SACs were bonded to two pyridine nitrogen, two pyrrole nitrogen atoms in parallel, which was not only a thermal stable structure but also well fitted with the FT-EXAFS curve of Fe SACs ( Figure S3G). The surface area of Fe SACs was measured as 421 m 2 g -1 ( Figure S7), which is able to provide more catalytic sites for CO2 hydrogenation.    We used the Fe (111) surface as the model of Fe nanoparticles. As shown in Figure S12A   The used amount of KOH in the outdoor electrolytic reactor was 200 g with the price of $1.
The Frame of the outdoor electrolytic reactor was provided by Zhengzhou Shike Technology Co., Ltd. with the price of $213 and the size was 70 cm (length)*35 cm (width)*60 cm (height).
The frame was divided into 12 independent chambers in series. The reflector was an aluminum plate was purched from Hebei scientist research experimental and equipment trade Co., Ltd. with the price of $16 and bent by us into a reflector with 55 cm of length and 36 cm of width.
The solar heating device was provided by Hebei scientist research experimental and equipment trade Co., Ltd. with the price of $349, which had the size of 4cm diameter and 50 cm of length.
The catalyst for CO (SCST-401) production was all purchased from Sichuan Shutai Chemical Technology Co., Ltd. with the price of $70 per 500 g. As the catalysts used amount for outdoor CO production was 400 g, coresponding to the cost of $56.