Abstract
The photoreduction of carbon dioxide (CO2) into renewable synthetic fuels is an attractive approach for generating alternative energy feedstocks that may compete with and eventually displace fossil fuels. However, it is challenging to accurately trace the products of CO2 photoreduction on account of the poor conversion efficiency of these reactions and the imperceptible introduced carbon contamination. Isotope-tracing experiments have been used to solve this problem, but they frequently yield false-positive results because of improper experimental execution and, in some cases, insufficient rigor. Thus, it is imperative that accurate and effective strategies for evaluating various potential products of CO2 photoreduction are developed for the field. Herein, we experimentally demonstrate that the contemporary approach toward isotope-tracing experiments in CO2 photoreduction is not necessarily rigorous. Several examples of where pitfalls and misunderstandings arise, consequently making isotope product traceability difficult, are demonstrated. Further, we develop and describe standard guidelines for isotope-tracing experiments in CO2 photoreduction reactions and then verify the procedure using some reported photoreduction systems.
Similar content being viewed by others
Introduction
If sunlight can be efficiently harnessed as a primary energy source for the photoreduction of CO2 into synthetic fuels, it will contribute a large flow of carbon-neutral energy in a material format that is compatible with existing fossil fuel infrastructure1,2,3,4,5,6,7,8. At a large enough scale, this technology may enable the control of atmospheric CO2 levels and make fossil fuels obsolete in the coming decades9,10,11,12. However, considerable scientific and technical challenges must be overcome to realize CO2 photoreduction reactions with high selectivity and conversion efficiency13,14,15,16,17,18. After many decades of exploration, most experimental studies focus on optimizing materials and reaction systems19,20,21,22,23,24. Most reported CO2 photoreduction systems are not yet able to deliver the desired products on a practical scale and with long-term operational stability. The low selectivity and CO2 conversion efficiency of most existing reaction systems25 is a significant roadblock in this field because it is challenging to attribute potential products (e.g., CO, alkanes, alcohols, carboxylic acids, and alkenes) solely to CO2 photoreduction processes26,27. Unfortunately, the efficiency of some high-performance reaction systems is later attributed to the decomposition of the carbon contaminants on the photocatalyst or the reaction system28,29,30,31,32.
Increasing awareness of the challenge of attributing products to CO2 photoreduction has made 13C isotope labeling experiments essential33,34,35. Lehn et al. carried out isotope labeling experiments with 13CO2 in 1983, revealing that the reduction product of CO indeed came from CO2 by using gas chromatography-mass spectrometry (GC-MS). However, no standard test method was mentioned in this research36,37. Later, Willner et al. used 13C nuclear magnetic resonance spectroscopy (NMR) to show that H13COO− originates from H13CO3− (the dissolved 13CO2 in solution)38. Over the years, various techniques using NMR spectroscopy and Fourier transform infrared spectroscopy (FT-IR) have been employed in isotope-tracer studies: the approach of 1H-NMR is efficient for liquid product analysis through peak shifts and coupling constants of the 13C-linked hydrogen; FT-IR equipped with a gas cell is effective for gaseous product analysis via the increased path length of a beam by multiple internal reflections39,40. Still, both methods are relatively insensitive to various carbon isotopes. Thus GC-MS-based techniques are still the most reliable and universal strategy for isotope-tracer studies in CO2 photoreduction41,42. For example, numerous organic carbon-containing photocatalysts, such as conjugated polymers, Metal-organic frameworks (MOFs), and Covalent organic framework (COFs), have good activities in CO2 photoreduction43,44,45,46,47,48. Isotope-tracing studies with GC-MS allow researchers to exclude carbon contamination from the decomposition of these catalysts or the materials synthesis process. Current isotope-tracer methods can measure simple samples such as 13CO, H13COOH, 13CH4, 13C2H6, 13C2H4, and 13CH3OH due to the isotope effects induced by different mass-to-charge ratios49,50,51,52. However, once the abovementioned series of molecular substances are generated during CO2 photoreduction, the inherent ionization process of mass detection could not only charge the molecules but also inevitably cause chemical bonds to break in these molecules, making the isotope-tracing via mass detection becomes increasingly difficult as molecular fragments with similar m/z ratios interfere and lead to misidentification and errors in quantification. Even for the emerging synchrotron vacuum ultraviolet photoionization mass spectrometry (SVUV-PIMS) strategy, the existence of interfering factors induced by the long-time sampling is also uncertain53. So the gas chromatograph (GC) is essential to be added to the sampling system to separate molecular species before collecting mass spectra (MS)54,55.
Despite the fact that isotope-tracer experiments are regarded as solid corroborating evidence for CO2 photoreduction, the current protocols are still substandard, causing the literature to be rife with false-positive results56. Meanwhile, the accurate and effective solution of isotope detection for various products in CO2 photoreduction is still lacking, which causes the community to be suspicious of all results. In the present work, we confirmed the pros and cons of the current mass spectrometry protocol as well as the emerging technology of synchrotron vacuum ultraviolet photoionization mass spectrometry (SVUV-PIMS) when they applied to the isotope-tracer experiments in CO2 photoreduction. We also experimentally performed isotope-tracer experiments on various isotope standards and described a standard protocol assessing various potential products of CO2 reduction. In addition, some classic CO2 photoreduction systems reported in the literature were also used to validate our protocols. It is imperative to settle on an appropriate scientific method for isotope-tracer studies to promote trust in the community as we develop and benchmark some materials systems for CO2 photoreduction. This research illustrates the pitfalls and misunderstandings in isotope-tracer studies and also provides examples and references for 13C isotope-tracer methods so we can all obtain solid evidence that reduction products are firmly attributed to CO2 photoreduction reactions.
Results
Importance of separation for multicomponent samples
GC-MS is the most common instrument for tracing isotopes in CO2 reduction reactions currently. The GC-MS instrument is composed of a GC to separate molecular species based on affinity with a column material and an MS to detect their mass or their mass fragments (Fig. 1a). However, the current method for tracing isotopes in CO2 reduction reactions ignores this. We verified the reliability of the current method. To mimic the products and conditions of a real photoreduction reaction, we directly employed standard gases such as 13CO2, 13CO, and vapor in the CO2 reduction system. The volume ratio of 13CO >5% v/v is much higher than the yield of CO in the usual CO2 photoreduction process. Then these premixed gases were collected from the CO2 reduction system and injected into the GC-MS equipped with a commercial HP-5ms column (see “Methods” for experimental details). As shown in Fig. 1b, a peak at m/z = 29 in the mass spectrogram could be obtained from TIC at an RT of 6.4 min. This peak has a definite value, but it could originate from three sources: (i) it is a molecular ion of 13CO (13CO+, m/z = 29), (ii) a fragment ion of 13CO2 (13CO+, m/z = 29), or (iii) it is a molecular ion of 15N14N (the natural abundance of nitrogen isotope, 15N14N+, m/z = 29). This could be further confirmed via the GC-MS analysis of pure 13CO2, 13CO, and N2 (Supplementary Figs. 1–3). Although the natural abundance of the nitrogen isotope exists in traces, the 13CO2 fragment ion is abundant and has the same characteristics to interfere with the isotope-tracer results for 13CO. Consequently, even if there is no 13CO in the injected mixture (only 13CO2 and vapor were injected into GC-MS), a similar mass spectrogram can be obtained under the same conditions (Fig. 1c), generating a peak at m/z = 29. This peak could be mainly attributed to the fragment ion (13CO+) generated from the dissociation of 13CO2. Even the deactivated fused silica tube (the length is 5 m), without any separation effect, acts as the connector to let all components (13CO2 and vapor) enter the quadrupole together; a similar result can also be obtained in the mixture of 13CO2 and vapor even without a photocatalytic reduction process (Supplementary Fig. 4). Moreover, the selected m/z = 45, 29 and 17 can be detected (Supplementary Fig. 5) when the detection method of GC-MS is set to the selected ion monitoring (SIM) mode (see “Methods” for experimental details). It is always used as evidence of traceability for the products originating from CO2 photoreduction, and signals at m/z = 17, 29, and 45 are assigned to CH4, CO, and CO2 molecular ions, respectively57. However, based on the above analysis, the presence of dissociated 13CO2 (generated fragments of 13CO+, m/z = 29) and vapor (generated fragments of HO+, m/z = 17) throws these conclusions of traceability into doubt.
For such confusing results, we could ascribe them to the detection principle of MS detector for charged species. Namely, the process of charging molecules in an MS detector is not so controllable, and there is some probability that the dissociation of the molecules coincides with ionization. In particular, when the amount of the sample is large, the probabilistic dissociation and dissociative recombination become inevitable, causing molecular ions and fragmented ions with the same characteristics to interfere with each other on the MS detector57. As a result, confusion could occur in the isotope detection for the product of 13CO due to the simultaneous process of ionization and dissociation over the 13CO2 and H2O reactants, 13CO product, and air components (N2, O2, Ar), respectively (Fig. 2a).
To avoid this confusion, suitable chromatographic conditions are a prerequisite to meeting the requirement of sample separation. As shown in Fig. 2b, a GC separator equipped with a suitable chromatographic column can separate mixed gases of CO2 and CO so that they are ideally queued in a single file before entering the MS detector. Thus although both CO2 and CO can generate the same fragment ions in the ion source, their different entry times provide resolution. As a result, two signal peaks appeared at different retention times (RT) in the total ion chromatography (TIC). Their corresponding mass spectrogram (MS1 and MS2 selected at different RT) show both molecular ions and fragment ions, which can be clearly attributed to either CO or CO2. In contrast, chromatography columns with bad separation characteristics allow components of CO2 and CO to enter the ion source simultaneously and generate complex molecular ions and fragment ions. These then enter the quadrupole together, resulting in a single peak on TIC and a corresponding mass spectrogram (MS). This spectrogram is very close to the abovementioned MS1 obtained from the fully separated samples. However, the signal corresponding to CO+ could be attributed to a fragment ion of CO2 or a molecular ion of CO, making it impossible to distinguish the source of CO+. Similarly, it is impossible to attribute the signal from C+ or O+ because both CO2 and CO can generate it.
Apart from the abovementioned method of mass spectrometry for isotope traceability, recently, an emerging technique called the synchrotron vacuum ultraviolet photoionization mass spectrometry (SVUV-PIMS) was also used to assign CO products to CO2 photoreduction reactions58. This analytical technique is useful because it delivers high-intensity, high-resolution, and widely tunable photon energies for photoionization processes. SVUV-PIMS enables selective and sensitive ionization (see “Methods” for experimental details) and can overcome the drawbacks of other ionization techniques and minimize fragmentation interference in MS detection. As shown in Fig. 3a, the m/z = 45 signal gives an ionization threshold of 13.75 eV in the photoionization spectra of pure 13CO2, while pure 13CO exhibits an increased ionization threshold of 14.00 eV in the m/z = 29 signal and the response rate of CO 10 times less than that of CO2. However, the SVUV-PIMS signal at m/z = 29 is also observed at 13.91 eV for pure 13CO2 (Fig. 3b). This feature matches the weak fragmentation of 13CO2 or via impurities introduced during the production of 13CO2 gas, interfering with the detection of 13CO to some extent. Therefore, when the photon energy of SVUV-PIMS was set at 14.50 eV to monitor the isotope product of 13CO, the peak at m/z = 29 can be found in the SVUV-PIMS spectra of pure 13CO2 (Fig. 3c) that is similar to the spectra obtained from the mixture of 13CO and 13CO2 (inset of Fig. 3c). Moreover, the intensity of m/z = 45 signal in pure 13CO2 is strongly associated with the increased repulsion voltage and sampling time (Fig. 3d). Although SVUV-PIMS minimizes fragmentation of the potential analytes, the popularity of this technique and the excessive demand of sample (hundred milliliters) hamper the extensive application of SVUV-PIMS to some extent. Moreover, the existence of interfering factors in this strategy induced by the long-time sampling is just like the SIM model in GC-MS, which cannot be ignored in the isotope-tracer experiment for CO2 photoreduction. Therefore, a complete separation of multiple components in GC is essential to obtain reliable isotope-tracing results for the products of CO2 photoreduction in MS.
Analysis strategy for the isotope-labeled products of CO2 photoreduction
Potential products of CO2 photoreduction reactions include CO, CH4, C2H6, C2H4, HCOOH, CH3COOH, CH3OH, and CH3CH2OH. Isotope-tracing experiments are complicated because products are inevitably mixed with the reactants (H2O and CO2) and impurities introduced during sampling (e.g., N2 and O2 via the airtight syringe). In addition, chromatographic columns such as HP-5ms in GC-MS cannot separate these gas components in isotope-tracing experiments. Thus, selecting a suitable column for separating permanent gas components is critical. For example, the HP-Molesieve column (see “Methods” for experimental details), as shown in Fig. 4a, can completely separate O2 (RT at 3.65 min), N2 (RT at 4.37 min), CH4 (RT at 5.25 min), and CO (RT at 7.25 min). After complete separation, it prevents the introduced air components during the injection process from interfering with the detection of products. The obtained mass spectra of O2 and N2 closely match the NIST mass spectral library (Supplementary Figs. 6 and 7), and the m/z of carbon-related molecular ions and fragment ions generated from 13CH4 (13CH4+, m/z = 17; 13CH3+, m/z = 16; 13CH2+, m/z = 15; 13CH+, m/z = 14; and 13C+, m/z = 13) and 13CO (13CO, m/z = 29 and 13C, m/z = 13) have a mass shift effect (M+1) compared with the non-isotope-labeled CH4 (Supplementary Fig. 8) and CO (Supplementary Fig. 9), respectively. Nevertheless, the molecular sieve column will irreversibly adsorb the reactant of CO2; thus, MS analysis of the reactant (CO2) and the products (CO or CH4) cannot be obtained simultaneously with the HP-Molesieve. To confirm the isotopic abundance of CO2, the HP-PLOT/Q is employed (see “Methods” for experimental details). As shown in Fig. 4b, the CH4 (RT at 4.87 min) and CO2 (RT at 6.60 min) elute as separate peaks in the TIC, and the corresponding MS of 13CH4 and 13CO2 (13CO2+, m/z = 45; 13CO+, m/z = 29; and 13C+, m/z = 13) could be obtained, respectively. The peaks exhibit a mass shift effect (M+1) compared to the non-isotope-labeled standards of CH4 (Supplementary Fig. 8) and CO2 (Supplementary Fig. 10). However, the air-derived components (e.g., O2 and N2) and CO (RT at 3.83 min) cannot be separated well under these conditions and may interfere with the analysis of CO isotopes. Interestingly, two kinds of C2 hydrocarbons (C2H6, RT at 8.21 min and C2H4, RT at 6.60 min) could be distinguished under these conditions. Although both 13C2H4 and 13C2H6 exhibit the same highest peak at m/z = 30, 13C2H6 possesses two more characteristic peaks at m/z = 31 and m/z = 32 that can only be generated from the fragment ion of 13C2H5+ and molecular ion of 13C2H6+. The ratio between the peaks that recorded in MS of 13C2H4 (13C2H4+, m/z = 30; 13C2H3+, m/z = 29; 13C2H2+, m/z = 28; 13C2H+, m/z = 27; and 13C2+ m/z = 26) and 13C2H6 (13C2H6+, m/z = 32; 13C2H5+, m/z = 31; 13C2H4+, m/z = 30; 13C2H3+, m/z = 29; 13C2H2+, m/z = 28; 13C2H+, m/z = 27; and 13C2+, m/z = 26) is different. It is still consistent with the ratio of the non-isotope-labeled standards C2H4 (Supplementary Fig. 11) and C2H6 (Supplementary Fig. 12), respectively, which comply with the mass shift effect (M+2) caused by two labeled carbon atoms. In addition, another set of peaks that recorded from m/z = 13 to 16 exist both in the MS of 13C2H4 (13CH2+, m/z = 15; 13CH+, m/z = 14; 13C+, m/z = 13) and 13C2H6 (13CH3+, m/z = 16; 13CH2+, m/z = 15; 13CH+, m/z = 14; 13C+, m/z = 13), which can be attributed to fragments caused by broken carbon–carbon bonds.
When the potential products of CO2 photoreduction are liquids such as alcohols or acids, we used the headspace injection method instead of airtight needles (see “Methods” for experimental details). The temperature of the HP-PLOT/Q column was raised to 120 °C to separate alcohols. The liquids suspected to be methanol or CH3OH (RT at 6.14 min) and ethanol or CH3CH2OH (RT at 12.23 min) eluted after the air components (RT at 2.96 min), CO2 (RT at 3.18 min), and H2O (RT at 3.98 min) in TIC (Fig. 4c). The material at RT = 6.14 min was identified as 13CH3OH based on two sets of peaks: one set at m/z = 29 to 33 and another set at m/z = 13 to 16 attributed to the molecular ions of 13CH3OH+ (m/z = 33) and the fragment ions of 13CH3O+ (m/z = 32), 13CH2O+ (m/z = 31), 13CHO+ (m/z = 30), 13CHO+ (m/z = 29), 13CH3+ (m/z = 16), 13CH2+ (m/z = 15), 13CH+ (m/z = 14), and 13C+ (m/z = 13), respectively. It is important to note that the largest peak in the MS was 13CH3O+ (m/z = 32) due to the easier dissociation of CH3OH. It also exhibits a mass shift effect (M+1) compared to the non-isotope-labeled CH3OH (Supplementary Fig. 13). Despite the fact that 13CH313CH2OH also has its largest fragment ion peak at m/z = 32, the source of this peak is derived from the cleavage of the carbon–carbon bonds, generating hydrogenated 13CH2OH+ instead of 13CH3O+ as was the case for 13CH3OH. Moreover, there is no distinct peak at m/z = 33 in the MS of 13CH313CH2OH that matches 13CH3OH. The MS of the ethanol fraction 13CH313CH2OH also has a set of peaks at m/z from 43 to 48, matching the molecular ions of 13C2H5OH+ (m/z = 48) and fragment ions of 13C2H4O+ (m/z = 47), 13C2H3O+ (m/z = 46), 13C2H2O+ (m/z = 45), 13C2HO+ (m/z = 44), and 13C2O+ (m/z = 43). It exhibited a mass shift effect of (M+2) or (M+1) versus the non-isotope-labeled C2H5OH (Supplementary Fig. 14) and the single-isotope-labeled CH313CH2OH (Supplementary Fig. 15), respectively, due to the two isotope-labeled carbon atoms in one molecule of 13CH313CH2OH (Supplementary Fig. 16). All of these points mentioned above are sufficient enough to match the qualitative requirements for 13CH3OH and 13CH313CH2OH. Likewise, the potential products of carboxylic acids (HCOOH and CH3COOH) could be separated by choosing a column (HP-FFAP) with stronger polarity (see “Methods” for experimental details). As shown in Fig. 4d, separation depends on the different molecular polarities between CH3COOH, HCOOH, and the reactants of H2O and CO2. CH3COOH has a relatively short retention time (RT at 8.07 min) than HCOOH (RT at 10.17 min), which elutes from the column after the air components (RT at 2.62 min) and H2O (RT at 3.15 min). By analyzing the corresponding MS, the non-isotope-labeled CH3COOH (Supplementary Fig. 17) possesses a molecular ion peak at m/z = 60 and two prominent fragment ion peaks at m/z = 43 and m/z = 45, while the corresponding molecular ion and fragment ions peak obtained from isotope-labeled 13CH313COOH shift to m/z = 62, m/z = 45, and m/z = 46, respectively. It reveals a difference in relative proportion compared to non-isotope-labeled CH3COOH, even considering the isotope-induced mass shift effect (M+2). After checking the MS of CH313COOH (Supplementary Fig. 18), a molecular ion peak at m/z = 61 and two fragment ion peaks at m/z = 44 and m/z = 46 were obtained. The two fragment ions of CH313COOH (m/z = 44 and m/z = 46), as well as the corresponding fragment ions of 13CH313COOH (m/z = 45 and m/z = 46) and CH3COOH (m/z = 43 and m/z = 45), are generated by carbon–carbon bond cleavage (COOH+) and dehydroxylation (CH3CO+), respectively (Supplementary Fig. 19). In addition, the isotope-labeled H13COOH possesses H13COOH+ (m/z = 47) molecular ions and fragment ions including 13COOH+ (m/z = 46), 13COO+ (m/z = 45), H13CO+ (m/z = 30), and 13CO+ (m/z = 29), which exhibit a mass shift effect (M+1) compared to the non-isotope-labeled HCOOH (Supplementary Fig. 20).
The four methods mentioned above can identify reactants and products in CO2 photoreduction reactions. However, this process requires multiple setups, and there is no standard strategy to separate and analyze all these products, reactants, and air components simultaneously. Even for the most reported photocatalytic CO2 conversion process of CO2 to CO, a mixture of O2, N2, CO, CH4, CO2, and water vapor cannot be completely separated with one column due to the irreversible adsorption of CO2 in the HP-Molesieve column and poor separation of CO/N2 in the HP-PLOT/Q column. To solve this problem, we designed a parallel connection system (Fig. 5a, see “Methods” for experimental details) containing both HP-Molesieve and HP-PLOT/Q columns. The diameters of the parallel GC columns were optimized to enable the components from different columns to queue into the MS detector separately. As shown in Fig. 5b, O2 (RT at 1.79 min), N2 (RT at 2.20 min), CH4 (RT at 2.48 min), and CO (RT at 4.59 min) elute from HP-Molesieve one by one and generate the corresponding MS spectra (Fig. 5c). It matches the abovementioned standard MS in Fig. 4a, respectively. CO2 and H2O could not be resolved by the HP-Molesieve channel but could be measured in the parallel HP-PLOT/Q channel. They eluted in the HP-PLOT/Q channel as an N2/O2/CO mixture (RT at 10.68 min), CH4 (RT at 11.10 min), CO2 (RT at 12.03 min), and H2O (RT at 16.91 min), which matches the standard MS in Fig. 4b. Based on this parallel strategy, the simulated mixture from reagents to products to air components in the potential photocatalytic CO2 reduction process could be analyzed in one GC-MS experiment.
Testing the isotope-tracing protocol using samples generated by well-studied CO2 photoreduction reactions
After optimizing our isotope-tracing methods using multiple standard samples, we tried to verify its reliability using some previously reported photoreduction systems. The homogeneous molecular catalyst FeIII porphyrin complex was used to evaluate CO2 photoreduction (see “Methods” for experimental details) in a closed gas circulation system (Supplementary Fig. 21)59. As shown in Fig. 6a, the obtained TIC of this system is similar to the standard spectra (Fig. 4a) because CH4 and CO are products of this reaction. In MS, the molecular ion peaks, as well as the fragment ion peaks, were assigned to the isotopically labeled 13CH4 and 13CO, respectively, conclusively demonstrating that the products of CH4 and CO originated from the photoreduction of labeled 13CO2 (Supplementary Fig. 22). Bipyridine complexes of ruthenium (Ru(bpy)3Cl2) is also a common homogeneous catalyst for CO2 photoreduction. By comparison with the sample without reaction (Supplementary Fig. 23), we could confirm that the 13CO product comes from the reduction of CO2 but not from the interfering factor of reactants (fragment ions from 13CO2) (see “Methods” for experimental details). We also examined CO2 photoreduction using a heterogeneous reaction system with Cd-doped ZnS as the photocatalyst. Cd-doped ZnS can reduce CO2 into HCOOH under light irradiation and was used to validate our GC-MS isotope-tracing method for liquid-phase products (see “Methods” for experimental details)60. Fig. 6 shows the TIC (Fig. 6c) and MS spectrum (Fig. 6d) collected using the products. The peak at RT = 10 min in the TIC corresponded to three peaks in the MS that match our standard MS for H13COOH (Fig. 4b). These results demonstrate that our method is also capable of tracing liquid products generated by CO2 reduction.
In addition, isotope tracing in all-organic reaction systems is particularly difficult to assess in CO2 photoreduction experiments. We employed our recently reported conjugated polymers (CPs) as a representative photocatalyst to experiment with both gas–solid phases (Supplementary Fig. 24) (see “Methods” for experimental details) and liquid phase (Supplementary Fig. 25) (see “Methods” for experimental details) reactions34. Only an increase in CO could be observed during the liquid phase reaction by comparing the TIC before and after photoreduction (Fig. 7a), and the MS collected at RT = 4.59 min matches the standard MS of 13CO (Fig. 7b), implying that CO originates from CO2 reduction (Supplementary Fig. 26). In contrast, CPs in the gas phase reaction possesses much higher activities and produces additional CH4 compared to the liquid phase reaction. In the corresponding MS (inset of Fig. 6c), CO and CH4 are primarily generated by photothermal effect-induced decomposition on solid CP instead of CO2 photoreduction (Supplementary Fig. 27). This result leaves little doubt that the self-decomposition of catalyst can seriously affect both the qualitative and quantitative analysis of the product when evaluating the CO2 photoreduction in the gas-solid phase of CPs. It also illustrates the importance of accurate isotope-tracing methods to prove the efficiency of CO2 photoreduction in different materials systems. Moreover, it should be emphasized that the isotope experiments and accurate GC-MS tests are not only necessary for the reaction systems using sacrificial agents but also important for the reaction systems without sacrificial agents.
Discussion
In summary, we initially demonstrated the difficulty of accurately ascribing products of CO2 photoreduction in isotope-tracing experiments. Existing methods are known to be rather crude due to the similarity of reactants, products, and the catalyst itself, which negatively affects efforts toward reliable CO2 photoreduction processes. Thus we sought to develop a rigorous strategy to eliminate the false-positive results by providing solid evidence of CO2 photoreduction. We disclosed many often neglected false-positive results in isotope-tracing studies using standard isotopically labeled molecules and the basic principles of GC-MS analysis. Through extensive testing, we systematically presented isotope-tracer experiments methods of standard spectra for various potential products of CO, alkanes, alcohols, carboxylic acids, and alkenes. Then we devised a method to simultaneously analyze these potential products using a parallel connection system using two GC columns. The accuracy of this parallel setup was verified with the standard isotopes described earlier; then, the setup was used to analyze four previously reported systems for CO2 photoreduction. The purpose of this research is to highlight appropriate scientific methods in CO2 photoreduction, help researchers avoid the pitfalls and misunderstandings in isotope-tracing experiments, and propose some standard procedures. More precise procedures will allow researchers to provide better feedback in their efforts to design photocatalysts with high selectivities and conversion efficiencies for CO2 photoreduction. We also provide examples and reference measurements for the 13C isotopes of numerous known reactants and products to help researchers firmly attribute them to CO2 reduction reactions.
Methods
Using HP-5ms as a column for isotope-labeled samples analysis
The 0.5 ml gas samples were collected and injected by gas-tight syringes (the VICI Pressure-Lok Precision Analytical Syringe A-2 Series (050033), 1 ml) and then analyzed by gas chromatography-mass spectrometry (8890-5977B GC-MS instrument, Agilent Technologies, USA) equipped with most used commercial capillary columns (HP-5ms, 5%-Phenyl-methylpolysiloxane, 19091S-433UI-KEY, 30 m × 0.25 mm × 25 μm, Agilent Technologies, USA) in GC-MS. Helium was used as carrier gas. The column was maintained at 150 °C for 25 min, and the flow of the carrier was 0.8 ml l−1. The temperatures of the injector, EI source, and GCITF were set to be 200, 200, and 250 °C, respectively. The selected mass-to-charge ratios of ions were 17, 29, and 45 in SIM mode. Developing a suitable programmed temperature rise process can further shorten the detection time.
Using SVUV-PIMS for isotope-labeled samples analysis
The synchrotron VUV photoionization mass spectroscopy (SVUV-PIMS) experiments were performed on the combustion station of the National Synchrotron Radiation Laboratory (Hefei, China). The test modes are the photoionization energy scan under the determined mass-to-charge ratio and the mass-to-charge ratio scan under the determined photoionization energy, respectively. Synchrotron radiation from the undulator beamline was monochromatized with 200 lines/mm laminar grating (Horiba Jobin Yvon, France), which covered the photon energy from 7.5 to 22 eV with an energy resolving power of 3000 (E/∆E at 10 eV). The average photon flux could reach the magnitude of 1013 photons/s after suppressing the higher-order harmonic radiation by a gas filter filled with noble gas61,62.
Using different columns for isotope-labeled samples analysis
The 0.5 ml gas samples were collected and injected by gas-tight syringes (the VICI Pressure-Lok Precision Analytical Syringe A-2 Series (050033), 1 ml), and the 2 ml liquid samples were collected and placed in a headspace sampler; then the samples were analyzed by gas chromatography-mass spectrometry (8890-5977B GC-MS instrument, Agilent Technologies, USA) equipped with commercial capillary columns. The column was maintained at a certain temperature for 15 min, and the flow of the carrier was 0.8 ml l−1. The temperatures of the injector, EI source, and GCITF were set to be 200, 200, and 250 °C, respectively. The mass-to-charge ratio of the mass scanning mode was set from 2 to 70. The GC-MS was operated the post-run after each injection (the temperature of the column oven increased to 300 °C with a rate of 30 °C and then maintained at 300 °C for 10 min). Developing a suitable programmed temperature rise process can further shorten the detection time.
The information of the columns are listed below:
(HP-Molesieve, 5A molesieve, 19091S-MS8, 30 m × 0.32 mm × 25 μm, Agilent Technologies, USA; HP-PLOT/Q, Bonded polystyrene-divinylbenzene, 19091P-QO4, 30 m × 0.32 mm × 20 μm, Agilent Technologies, USA; HP-FFAP, Modified polyethylene glycol, 19091F-413, 30 m × 0.32 mm × 20 μm, Agilent Technologies, USA).
Using a designed parallel connection system as a column for isotope-labeled samples analysis
The 0.5 ml gas samples were collected and injected by gas-tight syringes (the VICI Pressure-Lok Precision Analytical Syringe A-2 Series (050033), 1 ml) and then analyzed by gas chromatography-mass spectrometry (8890-5977B GC-MS instrument, Agilent Technologies, USA) equipped with designed parallel connection system (HP-Molesieve 15 m × 0.53 mm × 20 μm, HP-PLOT/Q 15 m × 0.32 mm × 20 μm, and CP4016 10 m × 0.32 mm, Agilent Technologies, USA) in GC-MS. Helium was used as carrier gas. The column was maintained at 45 °C for 20 min, and the flow of the carrier was 0.8 ml l−1. The temperatures of the injector, EI source, and GCITF were set to be 200, 200, and 250 °C, respectively. The mass-to-charge ratio of the mass scanning mode was set from 2 to 70. The GC-MS was operated the post-run after each injection (the temperature of the column oven increased to 300 °C with a rate of 30 °C and then maintained at 300 °C for 10 min). Developing a suitable programmed temperature rise process can further shorten the detection time.
Data availability
The data that support the plots within this paper and other findings of this study are available from the corresponding author upon reasonable request. Source data are provided with this paper.
References
Lin, H., Luo, S., Zhang, H. & Ye, J. Toward solar-driven carbon recycling. Joule 6, 294–314 (2022).
Wu, H. L., Li, X. B., Tung, C. H. & Wu, L. Z. Semiconductor quantum dots: an emerging candidate for CO2 photoreduction. Adv. Mater. 31, 1900709 (2019).
Li, X., Yu, J., Jaroniec, M. & Chen, X. Cocatalysts for selective photoreduction of CO2 into solar fuels. Chem. Rev. 119, 3962–4179 (2019).
Li, R., Zhang, W. & Zhou, K. Metal-organic‐framework‐based catalysts for photoreduction of CO2. Adv. Mater. 30, 1705512 (2018).
Wang, H.-N. et al. Recent progress and perspectives in heterogeneous photocatalytic CO2 reduction through a solid-gas mode. Coord. Chem. Rev. 438, 213906 (2021).
Jiao, X. et al. Fundamentals and challenges of ultrathin 2D photocatalysts in boosting CO2 photoreduction. Chem. Soc. Rev. 49, 6592–6604 (2020).
Chang, X., Wang, T. & Gong, J. CO2 photo-reduction: insights into CO2 activation and reaction on surfaces of photocatalysts. Energy Environ. Sci. 9, 2177–2196 (2016).
Gong, E. et al. Solar fuels: research and development strategies to accelerate photocatalytic CO2 conversion into hydrocarbon fuels. Energy Environ. Sci. 15, 880–937 (2022).
Keith, D. W. Why capture CO2 from the atmosphere? Science 325, 1654–1655 (2009).
Chen, C., Kotyk, J. F. K. & Sheehan, S. W. Progress toward commercial application of electrochemical carbon dioxide reduction. Chem 4, 2571–2586 (2018).
Lewis, N. S. & Nocera, D. G. Powering the planet: chemical challenges in solar energy utilization. Proc. Natl Acad. Sci. USA 103, 15729–15735 (2006).
Hoffert, M. I. et al. Advanced technology paths to global climate stability: energy for a greenhouse planet. Science 298, 981–987 (2002).
Jiao, X., Zheng, K., Hu, Z., Sun, Y. & Xie, Y. Broad-spectral-response photocatalysts for CO2 reduction. ACS Cent. Sci. 6, 653–660 (2020).
Ozin, G. A. Throwing new light on the reduction of CO2. Adv. Mater. 27, 1957–1963 (2015).
Pougin, A. et al. Au@TiO2 core-shell composites for the photocatalytic reduction of CO2. Chemistry 24, 12416–12425 (2018).
Dulla, M., Pougin, A. & Strunk, J. Evaluation of the plasmonic effect of Au and Ag on Ti-based photocatalysts in the reduction of CO2 to CH4. J. Energy Chem. 26, 277–283 (2017).
Moustakas, N. G. & Strunk, J. Photocatalytic CO2 reduction on TiO2-based materials under controlled reaction conditions: systematic insights from a literature study. Chemistry 24, 12739–12746 (2018).
Lee, B. H. et al. Electronic interaction between transition metal single-atoms and anatase TiO2 boosts CO2 photoreduction with H2O. Energy Environ. Sci. 15, 601–609 (2022).
Zhang, L., Zhao, Z.-J., Wang, T. & Gong, J. Nano-designed semiconductors for electro-and photoelectro-catalytic conversion of carbon dioxide. Chem. Soc. Rev. 47, 5423–5443 (2018).
Jiang, Z. et al. Filling metal-organic framework mesopores with TiO2 for CO2 photoreduction. Nature 586, 549–554 (2020).
Yu, H. et al. Synergy of ferroelectric polarization and oxygen vacancy to promote CO2 photoreduction. Nat. Commun. 12, 4594 (2021).
Sahm, C. D., Ucoski, G. M., Roy, S. & Reisner, E. Automated and continuous-flow platform to analyze semiconductor-metal complex hybrid systems for photocatalytic CO2 reduction. ACS Catal. 11, 11266–11277 (2021).
Cortes, M. A. L. et al. Formal quantum efficiencies for the photocatalytic reduction of CO2 in a gas phase batch reactor. Catal. Today 326, 75–81 (2019).
Li, K., An, X., Park, K. H., Khraisheh, M. & Tang, J. A critical review of CO2 photoconversion: catalysts and reactors. Catal. Today 224, 3–12 (2014).
Christopher, P., Jin, S., Sivula, K. & Kamat, P. V. Why seeing is not always believing: common pitfall in photocatalysis and electrocatalysis. ACS Energy Lett. 6, 707–709 (2021).
Yang, C. C., Vernimmen, J., Meynen, V., Cool, P. & Mul, G. Mechanistic study of hydrocarbon formation in photocatalytic CO2 reduction over Ti-SBA-15. J. Catal. 284, 1–8 (2011).
Mei, B., Pougin, A. & Strunk, J. Influence of photodeposited gold nanoparticles on the photocatalytic activity of titanate species in the reduction of CO2 to hydrocarbons. J. Catal. 306, 184–189 (2013).
Das, R., Chakraborty, S. & Peter, S. C. Systematic assessment of solvent selection in photocatalytic CO2 reduction. ACS Energy Lett. 6, 3270–3274 (2021).
Zhang, Y. et al. Photocatalytic CO2 reduction: identification and elimination of false-positive results. ACS Energy Lett. 7, 1611–1617 (2022).
Chen, P. et al. Rapid self-decomposition of g-C3N4 during gas-solid photocatalytic CO2 reduction and its effects on performance assessment. ACS Catal. 12, 4560–4570 (2022).
Yang, C. C., Yu, Y. H., van der Linden, B., Wu, J. C. & Mul, G. Artificial photosynthesis over crystalline TiO2-based catalysts: fact or fiction? J. Am. Chem. Soc. 132, 8398–8406 (2010).
Cybula, A., Klein, M. & Zaleska, A. Methane formation over TiO2-based photocatalysts: reaction pathways. Appl. Catal. B Environ. 164, 433–442 (2015).
Xu, F. et al. Unique S-scheme heterojunctions in self-assembled TiO2/CsPbBr3 hybrids for CO2 photoreduction. Nat. Commun. 11, 4613 (2020).
Wang, S. et al. Intermolecular cascaded π-conjugation channels for electron delivery powering CO2 photoreduction. Nat. Commun. 11, 1149 (2020).
Li, Y. et al. Facile top-down strategy for direct metal atomization and coordination achieving a high turnover number in CO2 photoreduction. J. Am. Chem. Soc. 142, 19259–19267 (2020).
Hawecker, J., Lehn, J.-M. & Ziessel, R. Efficient photochemical reduction of CO2 to CO by visible light irradiation of systems containing Re(bipy)(CO)3X or Ru(bipy)32+-Co2+ combinations as homogeneous catalysts. J. Chem. Soc., Chem. Commun. 9, 536–538 (1983).
Lecomte, J.-P., Lehn, J.-M., Parker, D., Guilhem, J. & Pascard, C. A dinuclear rhodium-carbonyl cryptate of a ditopic macrocyclic ligand. J. Chem. Soc., Chem. Commun. 6, 296–298 (1983).
Willner, I. & Mandler, D. Characterization of palladium-. beta.-cyclodextrin colloids as catalysts in the photosensitized reduction of bicarbonate to formate. J. Am. Chem. Soc. 111, 1330–1336 (1989).
Chatterjee, T., Boutin, E. & Robert, M. Manifesto for the routine use of NMR for the liquid product analysis of aqueous CO2 reduction: from comprehensive chemical shift data to formaldehyde quantification in water. Dalton Trans. 49, 4257–4265 (2020).
Kuhl, K. P. et al. Electrocatalytic conversion of carbon dioxide to methane and methanol on transition metal surfaces. J. Am. Chem. Soc. 136, 14107–14113 (2014).
Fu, Z. et al. A stable covalent organic framework for photocatalytic carbon dioxide reduction. Chem. Sci. 11, 543–550 (2020).
Li, J. et al. Self-adaptive dual-metal-site pairs in metal-organic frameworks for selective CO2 photoreduction to CH4. Nat. Catal. 4, 719–729 (2021).
Diercks, C. S., Liu, Y., Cordova, K. E. & Yaghi, O. M. The role of reticular chemistry in the design of CO2 reduction catalysts. Nat. Mater. 17, 301–307 (2018).
Schoedel, A., Ji, Z. & Yaghi, O. M. The role of metal–organic frameworks in a carbon-neutral energy cycle. Nat. Energy 1, 16034 (2016).
Nguyen, H. L. & Alzamly, A. Covalent organic frameworks as emerging platforms for CO2 photoreduction. ACS Catal. 11, 9809–9824 (2021).
Fang, Z.-B. et al. Boosting interfacial charge-transfer kinetics for efficient overall CO2 photoreduction via rational design of coordination spheres on metal–organic frameworks. J. Am. Chem. Soc. 142, 12515–12523 (2020).
Wang, J.-W. et al. Facile electron delivery from graphene template to ultrathin metal-organic layers for boosting CO2 photoreduction. Nat. Commun. 12, 813 (2021).
Dai, C. & Liu, B. Conjugated polymers for visible-light-driven photocatalysis. Energy Environ. Sci. 13, 24–52 (2020).
Huang, Q. et al. Multielectron transportation of polyoxometalate-grafted metalloporphyrin coordination frameworks for selective CO2-to-CH4 photoconversion. Natl Sci. Rev. 7, 53–63 (2020).
Wang, Y., Godin, R., Durrant, J. R. & Tang, J. Efficient hole trapping in carbon dot/oxygen‐modified carbon nitride heterojunction photocatalysts for enhanced methanol production from CO2 under neutral conditions. Angew. Chem. Int. Ed. 60, 20811–20816 (2021).
Gao, W. et al. Vacancy-defect modulated pathway of photoreduction of CO2 on single atomically thin AgInP2S6 sheets into olefiant gas. Nat. Commun. 12, 4747 (2021).
Ni, B., Jiang, H., Guo, W., Xu, Q. & Min, Y. Tailoring the oxidation state of metallic TiO through Ti3+/Ti2+ regulation for photocatalytic conversion of CO2 to C2H6. Appl. Catal. B Environ. 307, 121141 (2022).
Shao, W. et al. In-plane heterostructure Ag2S-In2S3 atomic layers enable boosted CO2 photoreduction into CH4. Nano Res. 14, 4520–4527 (2021).
Zhang, K. et al. Current dilemma in photocatalytic CO2 reduction: real solar fuel production or false positive outcomings? Carb. Neutrality 1, 10 (2022).
Vestal, M. L. Methods of ion generation. Chem. Rev. 101, 361–376 (2001).
Khan, I. S. et al. An efficient metal-organic framework‐derived nickel catalyst for the light driven methanation of CO2. Angew. Chem. Int. Ed. 60, 26476–26482 (2021).
Adams, N. G., Poterya, V. & Babcock, L. M. Electron molecular ion recombination: product excitation and fragmentation. Mass Spectrom. Rev. 25, 798–828 (2006).
Xu, J. et al. Efficient infrared‐light‐driven CO2 reduction over ultrathin metallic Ni‐doped CoS2 nanosheets. Angew. Chem. Int. Ed. 60, 8705–8709 (2021).
Rao, H., Schmidt, L. C., Bonin, J. & Robert, M. Visible-light-driven methane formation from CO2 with a molecular iron catalyst. Nature 548, 74–77 (2017).
Meng, X. et al. Efficient photocatalytic CO2 reduction in all-inorganic aqueous environment: cooperation between reaction medium and Cd (II) modified colloidal ZnS. Nano Energy 34, 524–532 (2017).
Zhou, Z. et al. The vacuum ultraviolet beamline/endstations at NSRL dedicated to combustion research. J. Synchrotron Radiat. 23, 1035–1045 (2016).
Pan, Y. et al. Online characterization of isomeric/isobaric compounds in the gas phase of mainstream cigarette smoke by synchrotron radiation vaccum ultraviolet photoionization time-of-flight mass spectrometry and photoionization efficiency curve simulation. Anal. Chem. 85, 11993–12001 (2013).
Acknowledgements
This work received financial support from the National Science Foundation of China (Nos. 51902121, 52073110, 22106106, 22071072, 21975090, and 52003213), the World Premier International Research Center Initiative (WPI Initiative) on Materials Nanoarchitectonics (MANA), MEXT (Japan), KAKENHI (18H02065, 20K05453), MEXT, and Photoexcitonix Project in Hokkaido University, Japan, the Fundamental Research Funds for the Central Universities of China (2662022LXPY001, 2662018QD011, 2662018PY052, SZYJY2022012, and 2662019PY023), and the Natural Science Foundation of Hubei Province (2019CFB322).
Author information
Authors and Affiliations
Contributions
S.W. and J.Ye. conceived the idea and designed the experiments; S.W. established methods for GC-MS analysis; S.W. and B.J. designed the parallel connection system; S.W., H.S., and S.Z. conducted GC-MS analysis for isotope-labeled standard samples; F.X., C.L., and Y.P. conducted the SVUV-PIMS measurements; S.W., X.M., and B.J. performed the isotopic experiments over reported systems. S.W., J.H., and B.J. wrote the manuscript; H.L., J.Yu., H.C., and J.Ye. jointly supervised the work and revised the manuscript. All authors contributed to the discussion of the results and the manuscript preparation.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks Jennifer Strunk and the other anonymous reviewers for their contribution to the peer review of this work. Peer reviewer reports are available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Source data
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Wang, S., Jiang, B., Henzie, J. et al. Designing reliable and accurate isotope-tracer experiments for CO2 photoreduction. Nat Commun 14, 2534 (2023). https://doi.org/10.1038/s41467-023-38052-0
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-023-38052-0
This article is cited by
-
Separating fiction from fact for photocatalytic CO2 reduction
Nature Chemistry (2023)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.