Internally hollow Cu2O nanoframes with the abundance of {110} facets enhance direct propylene epoxidation
Authors: Yueming Qiu, Yichen Zhang, Ronghui Zhang, Meng Huang, Kok Bing Tan, Guowu Zhan, Gang Fu, Qingbiao Li, Jiale Huang
Journal: Nature Communications (2025)
DOI: https://doi.org/10.1038/s41467-025-63059-0
Affiliations:
- Department of Chemical and Biochemical Engineering, College of Chemistry and Chemical Engineering, Xiamen University, Xiamen, Fujian, China
- College of Chemical Engineering, Integrated Nanocatalysts Institute (INCI), Huaqiao University, Xiamen, China
- State Key Laboratory of Physical Chemistry of Solid Surfaces, Xiamen University, Xiamen, Fujian, China
- College of Ocean Food and Biology Engineering, Fujian Provincial Key Laboratory of Food Microbiology and Enzyme Engineering, Jimei University, Xiamen, Fujian, China
- These authors contributed equally: Yueming Qiu, Yichen Zhang
Abstract
The gas-phase direct epoxidation of propylene (DEP) using molecular oxygen, which has been deemed the ‘dream reaction’ for propylene oxide (PO) production due to its efficiency and environmental benefits, remains highly regarded by researchers. In this contribution, we engineer a series of Cu2O nanocatalysts by employing the ligand-protection/selective facet-etching technique. Among these, the internally hollow Cu2O nanoframes, featuring increased specific surface area and a prevalence of {110} sites, achieve a triple-win in activity, selectivity, and stability, with an optimal PO formation rate of 0.18 mmol g_cat^-1 h^-1 and a selectivity of 83.8% at 175 degrees C. In addition, long-term tests confirm that these internally hollow nanoframes maintain high activity and selectivity for over 300 minutes. Further characterizations, combined with density functional theory calculations, confirm that the unique atomic arrangement of copper and oxygen on the Cu2O {110} facet facilitate the formation of chemically adsorbed oxygen species and propylene oxide as well. We anticipate that the ligand-protection/selective facet-etching approach may serve as a versatile method for fabricating well-defined catalyst architectures.
Introduction
Theselectivecatalyticconversionofpropylene(C H )intopropylene The DEP reaction using Cu-based catalysts, especially Cu O 3 6 2 oxide(PO)isanimportantprocessinthefieldofindustrialcatalysis1–3. nanocrystals,hasbeenthesubjectofintensiveresearch9,11,13–15. Var- POservesasabulkchemicalfeedstockforthemanufactureofversatile ious Cu O nanostructures, such as nanocubes, octahedra, and 2 value-addedchemicals,whichareessentialinsectorsofmicrochipand rhombicdodecahedra,alongwiththeirmodifiedcounterparts,have capitalconstructionfields4–6. Marketresearchdataforecaststhatthe beensynthesized,yieldingsignificantunderstandingoftheinfluence annualdemandof POwillsurpass20milliontonsin20257,8. However, ofcrystallographicorientationoncatalyticoutcomes16,17. Thediffer- conventionalindustrialmethodsfor POproductionareencumbered ent Cu Ofacets,suchas{111},{100}and{110},withdifferentsurface 2 by their cost-ineffectiveness and adverse environmental impact, oxygen coordination structures, are closely associated with the resulting in the generation of multiple by-products that intricately production propensity for acrolein, CO , and PO, respectively9,11. 2 complicate downstream separation processes. To address this chal- However,the POyieldwasnotsatisfactoryduetothelowdensityof lenge,thedirectepoxidationofpropylene(DEP)withmolecularoxy- the active site and the limited specific surface areas on the Cu O 2 genhasemergedasanidealalternativeforthe21stcentury,owingto catalysts. Thesynthesisofhollow Cu Onanoframeshasbeeniden- 2 itspromiseofachieving100%theoreticalatomeconomyandasus- tifiedasapromisingsolutiontoovercometheexistingchallenges, tainable,environmentallyfriendlyprocess8–12. primarily due to their enhanced surface-to-volume ratio and 1Departmentof Chemicaland Biochemical Engineering, Collegeof Chemistryand Chemical Engineering, Xiamen University, Xiamen, Fujian, China.2College of Chemical Engineering, Integrated Nanocatalysts Institute(INCI), Huaqiao University, Xiamen, China.3State Key Laboratoryof Physical Chemistryof Solid Surfaces, Xiamen University, Xiamen, Fujian, China.4Collegeof Ocean Foodand Biology Engineering, Jimei University, Xiamen, Fujian, China.5Theseauthors contributedequally:Yueming Qiu, Yichen Zhang. e-mail:gfu@xmu.edu.cn;kelqb@xmu.edu.cn;cola@xmu.edu.cn Nature Communications|( 2025)1 6:7802 1 ;,:)(0987654321 ;,:)(0987654321 Article https://doi.org/10.1038/s41467-025-63059-0 improvedsubstratetransfercapacity3,18,19. Specificallydesignedhol- efficiencyofthe DEPreaction,owingtothecriticaldependenceof low Cu O nanoframes, with a focus on exposing preferred crystal- propyleneoxide(PO)productionontheexposureofthe{110}facets 2 lographic planes, were expected to demonstrate superior catalytic of Cu O. Therefore, there is an urgent need to develop synthetic 2 performance. For example, Sui et al.20 revealed that hollow Cu O methods specifically aimed at selectively engineering Cu O nano- 2 2 catalysts with {111} edge sites significantly surpassed their solid framestoprioritizetheexposureof{110}facets. counterpartsintheoxidativedegradationofpollutants. Inaddition, In this work, the fabrication of hollow Cu O nanoframes with 2 hollow Cu Onanoframes,characterizedby{111}edgesites,exhibited abundant {110} facets was made possible through an innovative 2 enhanced catalytic performance in photocatalysis reactions as synthesis approach called ligand-protection/selective facet-etching. compared with conventional Cu O cubes21–25. However, the indis- This was achieved by employing lauryl sodium sulfate as a ligand 2 criminateformationofdiverseexposedfacetsdidnotimprovethe protectingagentforpreferentiallybindingtothehigh-surface-energy Fig.1|Microscopiccharacterizationsandschematicillustration. SEMand TEM structuremodelsof(k)c-Cu O-NF,(l)o-Cu O-NF,(m)d-Cu O-NF,(n)s-Cu O-NC 2 2 2 2 imagesof(a)c-Cu2O,(b)c-Cu2O-NF,(c)o-Cu2O,(d)o-Cu2O-NF,(e)d-Cu2O,(f) and(o)s-Cu2O-NFmodels.p Schematicillustrationforthesynthesisprocessof d-Cu O-NF,(g)s-Cu O,(h)s-Cu O-NCand(i,j)s-Cu O-NF. Three-dimensional s-Cu O-NF. 2 2 2 2 2 Nature Communications|( 2025)1 6:7802 2 Article https://doi.org/10.1038/s41467-025-63059-0 Fig.2|Structuralcharacterizations.(a) XRDand(b) FT-IRpatternsofas-prepared Cu Onanocatalysts.c XPSpatternsof Cuofs-Cu O,s-Cu O-NCands-Cu O-NF. 2 2 2 2 Fig.3|Microscopiccharacterizationofs-Cu O-NF.a TEMimagesofthemor- of Cu, O,and Cl. The(e) HRTEMimage(insetisthe Fast Fouriertransformimage), 2 phologychangeofs-Cu O-NFwithdifferentetchingtimes.b HRTEMimagesofthe (f) SAEDpatternsoftheareamarkedwithayellowrectangleandwhitecirclein(d) 2 areamarkedwithayellowrectangleofs-Cu2O-NFafter HCletchingfor15min. TEMimageofs-Cu2O-NF. c High-angleannulardark-fieldimageofs-Cu2O-NFandelementalmappingimages Nature Communications|( 2025)1 6:7802 3 Article https://doi.org/10.1038/s41467-025-63059-0 {110}facets,effectivelyshieldingthemfrom HCletching. Conversely, alsoincreasedduetothecontributionofthe–CH bandat630cm−1in 2 the other Cu O facets, especially the {111} and {100} facets, were theresidual D-glucose34,35. Inaddition,the XPSanalysisof C1s(Sup- 2 selectivelydiminishedby HCletching. Utilizingthistechnique,wehave plementary Fig.3a)revealedonlythepresenceofadventitiouscarbon successfully synthesized spherical Cu O nanoframes (s-Cu O-NF) species(284.8e V)andcarbonatespecies(288.4e V)onthesurfacesof 2 2 characterizedbyaninternallyhollowstructurewiththeabundanceof spherical Cu O nanocatalysts. The XPS analysis of Cu 2p (Fig. 2c) 2 {110} facets. The as-synthesized s-Cu O-NF achieved an impressive manifestedthatthesurfacecopperatomsonvariousspherical Cu O 2 2 selectivityofnearly100%forpropyleneoxide(PO)atarelativelylow were almost the same valence. The in situ DRIFTS (Supplementary operatingtemperatureof150°C,whichwassuperiortothatofother Fig.3b)of COadsorptiondiscernedvibrationalfeaturesat2112,2125 Cu Onanoparticlesinthe DEPreaction. and2172cm−1. Twopeaksat2125and2172cm−1canbeattributedtothe 2 gaseous CO36–39. Another CObandat2112cm−1canbeassignedtothe
Results
Results adsorbed CO on Cu(I) sites11,36,40–42. After the ligand-protection and Structureanalysisofcatalysts selectivefacet-etchingsteps,thespecificsurfaceareawasexpanded Wet-chemical methods were used to synthesize a number of Cu O andtheabundanceof{110}facetswassignificantlyincreased,leading 2 nanocrystals, which included cubic (c-), octahedral (o-), rhombic to more pronounced and sharper infrared absorption peaks of CO dodecahedral (r-),octadecahedral(d-), and spherical (s-) Cu O. Fur- adsorption. 2 thermore, c-, o-, d- and s-Cu O hollow nanoframes (see the experi- The morphological transformation of s-Cu O-NF during the 2 2 mental section for details) have also been prepared. The scanning first two hours of aging, characterized by the predominance of electron microscopy (SEM) and transmission electron microscopy etching processes over crystal growth, has been illustrated in (TEM)wereusedtocharacterizethemorphologiesofthesynthesized Fig.3. Thespherical Cu Osampleretainedasolidformduringthe 2 Cu Osamples. Thesolid Cu Osampleswerefoundtopossesswell- initial phase of HCl-mediated etching, which showed {110} edges 2 2 defined cubic, octahedral, rhombic dodecahedral, octadecahedral, with a mixture of other facets internally. A representative high- andsphericalshapes,whichwerecharacterizedbyuniformdispersion resolution transmission electron microscopy (HRTEM) image andparticlesize,accordingtothesetechniques. Thefabricated Cu O (Fig. 3b), acquired from the region marked with a yellow rec- 2 structuresareshownin Fig.1and Supplementary Fig.1withdetailed tangle, displayed distinct lattice fringes with spacings of 0.21, SEMand TEMimages. Amongthesestructures,thespherical Cu O(s- 0.24 and 0.30nm, corresponding to the {100}, {111} and {110} 2 Cu O) displays a uniformly spherical shape with an average size of facetsof Cu O,respectively11,43–45. Itcanbeobservedin Fig.3d–f, 2 2 ~475nm. The successful preparation of corresponding Cu O nano- after the appropriate etching time, s-Cu O-NF nanocatalysts 2 2 frames/nanocageswithhollowinteriorsthatwereclearlydistinctfrom exhibited a significant exposure of the {110} facets. HRTEM ima- the solid s-Cu O nanocrystals was achieved through selective facet ges from regions outlined by a yellow rectangle in Fig. 3d and 2 etching. Figure 1h demonstrates that the spherical Cu O hollow Supplementary Fig. 4, where the lattice spacings of 0.30nm and 2 nanocage (s-Cu O-NC) assumes a concave, bowl-like form, with a 0.43nmwereidentified,demonstratedthetypicalcharacteristics 2 marginally reduced average size. Conversely, the internally hollow of the {110} facets (Supplementary Fig. 5). This was further con- Cu Onanoframewithsphericalshape(s-Cu O-NF)(Fig.1i,j)adoptsan firmed by Fast Fourier transform and selected area electron dif- 2 2 intricate, football-like geometry, with a wall thickness estimated at fraction(SAED). However,thestructuraldegradationandcollapse 25–40nm. Subsequently,theligand-protection/selectivefacet-etching occurred as a result of prolonged etching time. Supplementary strategy was employed to eliminate the {111} and {100} facets from Fig. 6 illustrates the connection between crystal phase transfor- Cu Ocatalystsinordertosynthesizespherical Cu Onanoframes(s- mation and morphology of s-Cu O-NF during the HCl etching 2 2 2 Cu O-NF)(Fig.1p). Thisillustrationdemonstratedthatthestructure processes. The pure phase of Cu O was clearly indicated by the 2 2 wasinternallyhollowandthattherewerenumerous{110}sites. indexingofallthe XRDpeaks. Duetotheadsorptionof SDSand XRDwasusedtoconfirmthecrystalphasesofthe Cu Onanoca- its protection for the {111} facet, the XRD peak intensities were 2 talysts thathad already been prepared. As shown in Fig. 2a,all the low. By using the appropriate facet-etching technique, the inner diffractionpeakswerewellindexedaccordingtoapurephaseof Cu O surfaceofs-Cu O-NFwasrevealed,withthe XRDintensityratioof 2 2 (space group: Pn 3m, lattice constant a=0.427nm, JCPDS 05- {110}/{111} reaching almost 0.28, which is a significant improve- 066726,27). The samples did not show any detectable impurities. mentfromthatin JCPDS05-0667(nearly0.07). Furthermore,the According to the ligand-protection/selective facet-etching process, relative peak changes in UV-Vis spectroscopy (Supplementary s-Cu O-NFnotonlyexperiencedacrystalgrowthprocessofupto2h, Fig.7a) showed similar emission spectrabutdifferentintensities 2 butalsoexistedinanetchingstageofupto50minforfacet-etching over various Cu O samples throughout the facet-etching pro- 2 andagingprocess,effectivelyincreasingtheintensityofcharacteristic cesses. These findings indicated that the source’s absorbance diffraction peaks. The above two key steps exposed the inner and changes during the s-Cu O-NF facet-etching process within 2 outersurfacesofs-Cu O-NFwiththeabundanceof{110}facets,origi- 50min were responsible for the gradual shift of sample color 2 nating from an internally hollow structure28. Notably, these nano- from yellow to orange-red, which occurred due to the internal frames showed a significantly increased proportion of {110} facets hollowing of the structure. The sample color became lighter as whencomparedtothesolid Cu Onanocrystals. Thisphenomenonwas theetchingtimeincreased,resultinginstructuraldegradationof 2 particularly noticeable in the case of spherical Cu O nanoframes, thes-Cu O-NF(Supplementary Fig.7b)9,46. Theseobservationscan 2 2 whichhadasharp Cu O{220}peak. The chemicalbonds and inner be explained by three distinct stages in the crystal growth and 2 structureswerefurtheranalyzedusing FT-IR. Asdepictedin Fig.2b,it etching process24,47: (1) the reaction of Cu2+ and OH− led to wasdemonstratedthatall Cu Onanocatalystsshowedtheidentified Cu(OH) precipitation;(2)theadditionof NH OH·HClaccelerated 2 2 2 peakpositionsat628cm–1butabsentat550cm–1. Theformercouldbe the conversion of Cu(OH) precipitates into Cu O nanocrystals 2 2 assignedtothestretchingvibrationsof Cu(I)–O,whiletheabsenceof with trace Cl− incorporating into the crystal lattice9; (3) mor- the latter peak, typically attributed to Cu(II)–O vibrations, further phology control by the ligand-protection/selective facet-etching corroborated the predominant presence of Cu O in these strategyconfirmedthepredominantexposureofthe{110}facets. 2 nanocatalysts29–31. Due to the different synthesis methods and the colorofthe Cu Osamples(Supplementary Fig.2),theabsorptionand 2 transmission of infrared rays were different32,33. In addition, as for c-Cu 2 O,theintensityofthe Cu–Ostretchingvibrationat628cm–1was 2Cu2++OH (cid:1)!CuðOHÞ 2 ðsÞ ð1Þ Nature Communications|( 2025)1 6:7802 4 Article https://doi.org/10.1038/s41467-025-63059-0 Fig.4|Catalyticperformanceofpropyleneepoxidationwith O. Propylene and(d)s-Cu O-NFatdifferenttemperatures. Reactionconditions:CH:O:N =10: 2 2 3 6 2 2 conversionandproductselectivityover(a)as-prepared Cu2Onanocatalystsat 5:85vol.%, GHSV=36,000h−1,50.0mgcatalyst.einsitu DRIFTSspectraafter 175°C. Propyleneconversionandproductselectivityover(b)s-Cu2O,(c)s-Cu2O-NC exposings-Cu2O-NFtothemixedflowof C3H6, O2and N2(10:5:85vol.%). the DEPperformancecomparisonofc-,o-andr-Cu Onanocrystals, 2CuðOHÞ +2NH OH!Cu OðsÞ+5H O+N ð2Þ 2 2 2 2 2 2 which exposed {100}, {111}, and {110} facets, respectively. The dis- tributionof DEPreactionproductsdemonstratedthecriticaldepen- Cu OðsÞ+4HCl!2HCu Cl +H O ð3Þ denceof POproductionontheexposuretothe{110}facets. Moreover, 2 2 2 the involvement of{100},{111}, and even higher index facet sites in truncated 26-facet polyhedral Cu O catalysts, as visualized in SEM, 2 The EDXmappinganalysiscanbefoundin Fig.3cand Supple- TEMandsimulatedstructuresimagesin Supplementary Fig.11,was mentary Fig. 8. The structural integrity of the s-Cu O-NF and conducive to the deep oxidation (Supplementary Fig. 12). More 2 s-Cu O-NCnanomaterialswasconfirmedbytheuniformdispersionof importantly, the catalytic performances across various morphologi- 2 Cu and O elements observed in the comprehensive mapping. The callydistinct Cu Ostructures(includingc-,o-,d-,ands-Cu O)andtheir 2 2 hollowconfigurationoftheengineered Cu Onanoframes,asdescri- hollow counterparts, derived via ligand-protection/selective facet- 2 bedearlier,wasadistinctivefeature. Aprofusionofmeso-andmacro- etching strategy,are shown in Fig. 4 and Supplementary Fig. 13. As pores was formed as a result of the combined ligand-protection/ expected, the ligand-protection/selective facet-etching strategy was selective facet-etching strategy and the inherently porous archi- effective and versatile, as evidenced by the enhanced DEP activity tecture. Consequently,therewasasignificantincreaseinthespecific originating from the internally hollow Cu O structures with pre- 2 surfaceareaoftheinternallyhollow Cu Onanoframes,accompanied dominant Cu O{110}facetexposure. Figure4ademonstratesthatthe 2 2 by a concomitant increase in pore volume, surpassing that of con- c-Cu O-NF and s-Cu O-NC/-NF exhibited excellent PO selectivity at 2 2 ventional Cu Onanoparticles(Supplementary Fig.9and Supplemen- 175°C, and the catalytic performance of the above hollow Cu O 2 2 tary Table1). Specifically,thes-Cu O-NFands-Cu O-NCnanomaterials nanoframeswasattributedtothe{110}facetoredge-sitesthatselec- 2 2 exhibited significantly superior S and nitrogen adsorption capa- tively catalyze the propylene epoxidation. Among these, the con- BET citiescomparedtos-Cu O. current presence of hollow structure and exposed {110} facet in 2 s-Cu O-NFexhibitedthehighest POselectivityof83.8%atcomparable 2 DEPwithmolecularoxygen propylene conversions, achieving an optimal PO formation rate of Thereactivityof Cu Onanocatalystsinthe DEPreactionwasevaluated 0.18mmol g –1 h–1. A similar trend of DEP performance can be 2 cat inacontinuous-flowquartzreactor,maintainingagasmixturecom- observed at225°C (Supplementary Fig. 13). Compared with o-Cu O 2 positionof C H /O/N =10:5:85vol.%,pressure(P)of0.1MPa,anda and d-Cu O, the smaller XRD intensity ratio of {110}/{111}, with the 3 6 2 2 2 flow rate of 30.0m Lmin–1. As shown in Supplementary Fig. 10, the residualminor{111}and{100}facetsontheoutersurfaceofo-Cu O-NF 2 facet-controlledselectivityof Cu Ocatalystscanbeobservedthrough and d-Cu O-NF, promoted unwanted over-oxidation pathways, 2 2 Nature Communications|( 2025)1 6:7802 5 Article https://doi.org/10.1038/s41467-025-63059-0 Table1|Oxygencompositionsofs-Cu O,s-Cu O-NCand primaryproductthroughoutthestabilitytest. Incontrast,theactive 2 2 s-Cu 2 O-NF sites on s-Cu 2 O-NC and s-Cu 2 O exhibited lower stability, with PO selectivitydiminishingtonegligiblelevelsafterthetestdurationsur- Catalyst Oα(%)a Oβ(%)a Oγ(%)a Oα+β:Oγ b passed100min(Supplementary Fig.20eandf). Uponfurtherelevating s-Cu O 14.1 52.5 33.4 0.57 2 thereactiontemperatureto200°C,s-Cu O-NFmaintainedstable DEP 2 s-Cu 2 O-NC 5.8 33.0 61.2 1.19 reactivityandconsistent POselectivity,with POpredominatingamong s-Cu 2 O-NF 12.4 34.1 53.5 2.16 reaction products throughout the 300min testing period (Supple- a Thecompositionsweredeterminedby O2-TPDmeasurement. mentary Fig.20c). Post-reactionstructuralintegrityandmorphology b Thecompositionsweredeterminedby XPSanalysis. ofthespents-Cu O-NFwereconfirmedviaextensivecharacterization 2 techniques including XRD, SEM, TEM, HRTEM, and XPS analyses (Supplementary Fig.21). Afterthe DEPreactionat150°Candevenup to200°C,therewerenosignificantchangesinthehollowstructure underscoringthenecessityofpredominantlyexposingthe{110}active and particle size, indicating that the essential {110} facet sites of facets(Supplementary Fig.14). s-Cu O-NFwerestillpreservedwithoutsurfacereconstruction. Allof 2 Interestingly,differentspherical Cu Osamplesdisplayeddrama- these findings accentuated that ligand-protection/selective facet- 2 ticallydifferentcatalyticperformancestowardsthe DEPreaction. The etching, targeting a larger specific surface area and predominant PO selectivity (<10%) was only low for solid Cu O nanocrystals (s- exposure of Cu O {110}, achieves a triple-win: it would not only 2 2 Cu O) at all reaction temperatures (Fig. 4b), while moderate PO enhance the activity and selectivity for the DEP reaction, but also 2 selectivity(40–45%)wasattainedforhollow Cu Onanocages(s-Cu O- strengthencatalystdurability. 2 2 NC)inthetemperaturerangeof150°Cto175°C(Fig.4c). Incompar- ison,theas-obtaineds-Cu O-NF,withthepredominantexposureofthe Structuresensitivityof Cu Ocatalysts 2 2 {110} facets, achieved nearly 100% POselectivity at 150°C(Fig. 4d), Toobtainin-depthknowledgeaboutthestructuresensitivityofdis- whichcanbeintriguingandworthconsideringtheseparationcostfor tinct Cu Onanostructuresinthe DEPreaction,the EIS, Raman, O -TPD, 2 2 theproductsinindustrylevel. Whenthetemperaturewaselevatedto H -TPR and XPS analyses were utilized. In the DEP reaction, it was 2 175°C, 83.8% PO selectivity was still maintained with the C H con- postulatedthatthechemicallyadsorbedoxygenspecies,suchas O–or 3 6 versionof0.14%. Supplementary Fig.15displayedtheapparentacti- O –,wouldpreferentiallyinteractwiththeelectron-rich C=Cbondof 2 vationenergyplotsofs-Cu O,s-Cu O-NC,ands-Cu O-NFdeducedvia propylenetoproduce PO,whilethenucleophiliclatticeoxygenspecies 2 2 2 the Arrheniusequation. Clearly,thecatalystswithhigherreactivityhad (O2-) tends to attack the α C−H bond to yield deep-oxidation lowerapparentactivationenergies,whichindicatedthatligand-pro- products48,49. It is well established that the chemically adsorbed O 2 tection/selectivefacet-etchingtreatmentnotonlygeneratedagreater willbeactivatedonthecatalystsurfacealongwiththecharge-transfer numberofactivesitesbutalsoenhancedintrinsicactivity. phenomenon. Therefore, the surface nature and charge transfer Itwasworthnotingthatthespherical Cu Onanocrystalswithout capacityofcatalystshaveprofoundimplicationsforthepropertiesof 2 etching(s-Cu O-3)containingthedoped Clspecies(asvisualizedin intermediateoxygenspecies,furtherdeterminingthepathwayof DEP 2 SEM,high-angleannulardark-field,andelementalmappingimagesin reactions. The Ramananalysiswascarriedouttoprovideinsightsinto Supplementary Fig.16),displayednegligible DEPreactivityandalmost the surface and local chemical structures of the catalysts. Supple- no PO selectivity across all reaction temperatures. Furthermore, as mentary Fig.22arevealedthats-Cu O-NFhadthemostobviouscrystal 2 shownin Supplementary Figs.17and18,theinfluenceofthecontrolled defects,whichwasbeneficialtopromoteelectrontransferfromthe variationof Clsupplyindicatedthatunoptimized Clintroductionled activesitestoadsorbedoxygenspeciesbasedonthestrongdepen- to insufficientcatalyticactivityor structuralcollapse. The PO selec- denceoftransferredelectronsondefectiveness. Furthermore,the EIS tivityremainedthehighestonlywhentheinternalnanoframestructure Nyquistplotsprovidedtheinternalimpedanceofs-Cu Osamplesby 2 waspreserved,withpredominantexposureof{110}facets. Thesynergy comparingtheradiusofthearc,whichrepresentedtheinternalcharge between Clandthe{110}facetsexistsbutdoesnotovershadowthe transfer resistance: the smaller the radius of the arc, the lower the structural factors, which are the primary determinants of catalytic resistance. As shown in Fig. 5a, s-Cu O-NF generated the smallest 2 performance. This emphasizes the irreplaceable role of the ligand- semicircleradiuscomparedwiths-Cu O-NCands-Cu O,accompanied 2 2 protection/selectivefacet-etchingmethod. Byconductingacompre- by the smallest R value (Supplementary Fig. 22b), illustrating that ct hensivecomparativeanalysisofthe DEPperformance(Supplementary s-Cu O-NFpossessedthesmallestinterfacialcharge-transferresistance 2 Fig.19and Supplementary Table2),s-Cu O-NFdemonstratedahigh andthestrongestchargetransportcapacity. Accordingly,the Raman 2 PO selectivity and an unprecedented PO formation rate, out- and EIS results were employed to characterize the charge-transfer performingthoseofothercoinage-metal-basedcatalysts. propertiesofthereactiveoxygenspeciesons-Cu Oanditsderivatives. 2 Theinsitu DRIFTSmeasurementof C H and O adsorptionon Collectively, these results suggest that s-Cu O-NF offers superior 3 6 2 2 s-Cu O-NF atdifferent temperatures corroborated the catalytic per- electron and molecule transfer kinetics, enhancing the electron 2 formance(Fig.4e). Thebandsat1440,1635,2950,2980and3100cm–1 transferfromthesurfacetomolecular O togenerate O–or O –onthe 2 2 wereassignedto CH bending, C=Cstretchingandaseriesofmethyl surface50–52. 2 symmetrical C−H vibrations arising from propylene adsorption on Inaddition, O -TPDwasconductedtogaininsightsintotheoxy- 2 Cu(I)–Osites. Theintensitychangesindicatedapositivecorrelation genspecies. Typically,threetypesofoxygenspeciescanbeidentified betweenpropyleneconversionandreactiontemperature9,11,13. Stability onthe Cu 2 Osurface,namelysurfaceadsorbedoxygen(Oα),chemically tests revealed robust catalytic performance of s-Cu 2 O-NF at 150°C, adsorbedoxygen(Oβ, O–or O 2 –),andlatticeoxygen(Oγ). Asshownin 175°C and 200°C over prolonged periods exceeding 300min. The Fig. 5b, onlyweakdesorptionpeaks,primarilyatthe lattice oxygen s-Cu O-NFcatalyzedthe DEPreactionfor300minwithonlyaslight desorption temperatures, were observed on the s-Cu O, indicating 2 2 decreaseinthe POselectivity(above85.2%)andaslightincreaseof that nearly no adsorbed oxygen species were involved. In contrast, acroleinselectivityat150°C(Supplementary Fig.20a). Asdepictedin s-Cu O-NFexhibited three distinctdesorptionpeakslocatedat243, 2 Supplementary Fig.20b,stabilityevaluationsat175°Cshowedthat PO 331,and404°C. However,fors-Cu 2 O-NC,theweaker Oαintensitybut and acrolein were initially predominant products; however, a slight stronger Oγintensitywasobservedathighertemperatures(361and reductionin POselectivityoccurredovertime,accompaniedbythe 432°C,respectively). Combinedwiththedatasplitgraphof O -TPD 2 emergenceof CO fromdeeperoxidationprocesses. POremainedthe (Supplementary Fig.23)andtherelativecontents(Table1),itcouldbe 2 Nature Communications|( 2025)1 6:7802 6 Article https://doi.org/10.1038/s41467-025-63059-0 Fig.5|Structuresensitivityof Cu2Ocatalystsand DEPprocessstudies.(a) EIS DRIFTSspectraafterexposings-Cu2O-NFtotheflowof C3H6andmixedflowof spectra(usinga1MKOHelectrolytedissolvedindeionizedwater),(b) O2-TPD C3H6, O2and N2(10:5:85vol.%)at(d)150°Cand(e)250°C. patternsand(c) XPSpatternsof Oofs-Cu O,s-Cu O-NCands-Cu O-NF. Insitu 2 2 2 inferred that the excellent PO selectivity via s-Cu O-NF might be adsorbedoxygenspecies(O–or O –). The H -TPRexperiments(Sup- 2 2 2 attributedtotheabundant O–or O –speciesthereon. XPSmeasure- plementary Fig.24)werealsocarriedouttoidentifythesurfaceoxygen 2 ments were further undertaken to investigate the surface oxygen speciesbelow400°C. Theresultsrevealedthatthereductionpeakof speciesofthe Cu Osamples. The O1sspectrapresentedasymmetric s-Cu O-NF shifted to a lower temperature compared to s-Cu O-NC, 2 2 2 peaks, indicating that the presence of different oxygen species on implying greater H consumption. Thus, the conclusion could be 2 Cu Ocatalystscanbeidentifiedashydroxyl(O –,532.8e V),surface drawnthat,incontrasttos-Cu Oands-Cu O-NC,s-Cu O-NFharborsa 2 OH 2 2 2 adsorbed oxygen (Oα+β, 531.5e V), and lattice oxygen (Oβ, 530.0- higher proportion of chemically adsorbed oxygen species that can 530.5e V),respectively53,54. Furthermore,the Oα+βpeakofs-Cu 2 O-NF promote POformation. ands-Cu O-NCshiftedtoahigherbindingenergythanthatofs-Cu O, 2 2 attributedtotheelectrontransferfromthe Cu Osurfaceto O . The Catalyticmechanismstudies 2 2 ratiooftheoxygenspecieshasbeensummarizedin Table1. Impress- Insitu DRIFTSwasalsoutilizedtoinvestigatetheoxidativeprocessof ively,s-Cu O-NF,synthesizedthroughtheligand-protection/selective propylene with O over s-Cu O-NF catalysts. The experiments were 2 2 2 facet-etching strategy, exhibited a higher proportion of chemically conducted under conditions introducing C H and C H +O inthe 3 6 3 6 2 Nature Communications|( 2025)1 6:7802 7 Article https://doi.org/10.1038/s41467-025-63059-0 Fig.6|DFTcalculations. Therelativeenergiesofoxygenspeciesandpossibleintermediatesforthereactionofpropylenewith O*over Cu O{111},{100},and{110} 2 2 surfaces,respectively. range of 150°C to 250°C (0.1MPa). After N purging thoroughly, Computationalstudies 2 s-Cu O-NF was exposed to a C H flow or C H +O mixed flow, Toelucidatetheinfluenceofdifferentmorphologiesof Cu Oonthe 2 3 6 3 6 2 2 resultingintheappearanceofvarious IRfeatures(Fig.5d,eand Sup- DEPreaction,weemployedperiodicdensityfunctionaltheory(DFT) plementary Fig.25). Theassignmentsofthesevibrationalfeatureshave calculations to assess the adsorption characteristics of Cu O {111}, 2 beensummarizedin Supplementary Table S3. Amongthem,thebands Cu O{100}and Cu O{110}(Fig.6and Supplementary Table4). 2 2 at1635and2954cm−1wereindicativeofthe C=Cand CH stretching Initially,weexaminedtheadsorptionofmolecularoxygen(O , 3 2 vibrations, respectively, representing a free C H molecule. Con- indicatingchemicallyadsorbedoxygen)andtheepoxidationproduct 3 6 currently,thepeaksproximateto1440,1473,1540,and2920cm–1were (O+PO). The DFTresultsrevealedthat O *exhibitedstrongadsorp- 2 ascribedto C H adsorbedat Cu(I)sitesorbridging Cu–Ositesof Cu O tiononthe Cu O{111}surfacewithanadsorptionenergyof−1.74e V,in 3 6 2 2 {110}(notedas C H (a) )11,13. Significantly,thevibrationalfea- contrasttothe relativelyweakadsorption on Cu O{100} and Cu O 3 6 Cu&Cu, O 2 2 turesofintermediatesduringthe DEPreactionprocessexhibiteddis- {110}surfaces,whichpresentedadsorptionenergiesof−0.51e Vand tinctvariationswhenamixedflowof C H and O wasintroducedas −0.57e V, respectively. Analysis of Bader charges indicated that the 3 6 2 comparedtothe C H flowat150°C. Theappearanceof IRbandsat adsorbed O *onthesesurfacesacquirednegativechargesof0.58|e|, 3 6 2 2752,2815,and 2856cm−1 arising from C−Hstretching,whichwere 0.10|e|,and0.27|e|,respectively,suggestingtheformationofsuper- split by Fermi resonance, emerged exclusively in the mixture of oxide species. Concerning the epoxidation product, the Cu O {111} 2 C H +O ,signifyingintermediatesinvolvedin POproduction. Inter- surface demonstrated the most substantial adsorption energy 3 6 2 estingly, the same vibrational bands could be observed via PO (−2.89e V), followed by the Cu O {110} (−1.92e V) and Cu O {100} 2 2 adsorption(Supplementary Fig.26)10,49. Inaddition,characteristic IR (−1.46e V) surfaces. These results, however, could not explain why peaksassociatedwithintermediatesforacroleinor CO production, onlythe Cu O{110}exhibitsexcellent DEPperformance. 2 2 including HCOO(a) (1385 and 1560cm−1), C H O(a) (1652cm−1), and Accordingly,weshiftedourattentiontotheadsorptionofpossi- 3 4 CO (manifestingasapairofweak IRbandsat2305and2375cm−1), bleepoxidationintermediates. Asdepictedin Fig.6,theinteraction 2 wereobservedinthepresenceof C H +O butwereabsentwith C H amongthepropylene C=Cbond,theadsorbed O *,andanadjacent 3 6 2 3 6 2 flow11,49,55,56. Theseresultssuggestedthatthereactionprocessoccurred Cu atom can lead to the formation of peroxametallacycle species, onlywhenthecatalystwasexposedtoanexternalsource O at150°C. whichwas usuallypostulated as a critical intermediate inthe epox- 2 Asimilarphenomenoncanbeobservedatareactiontemperatureof idationprocess57–59. DFTcalculationsdemonstratedthattheseperox- 175°C,atwhichthe POformationratebecamemostoptimistic. With ametallacycle species exhibited only marginal adsorption energies, furtherheating,theaforementionedvibrationalfeaturesweresimilar rangingfrom+0.12to−0.17e Vrelativeto O *+C H ,therebydown- 2 3 6 upon chemisorption of C H +O mixed flow or C H flow on playingtheirsignificancein DEP. Alternatively,theinteractionamong 3 6 2 3 6 s-Cu O-NF(Supplementary Fig.27). Basedontheinsitu FT-IRspectra the propylene C=C bond, the adsorbed O *, and a nearby oxygen 2 2 ofpropyleneepoxidation,itcanbeconcludedthatthe DEPreaction atom (rather than Cu atom) can result in relatively stable glycolate catalyzedbys-Cu O-NFwouldfollowthe Langmuir-Hinshelwood(LH) intermediates, with adsorption energies ranging from −1.02 to 2 mechanism at a relatively low operational temperature, utilizing −1.95e V with respect to O *+C H . In principle, these glycolate 2 3 6 surface-adsorbedoxygenratherthanlatticeoxygentointeractwith intermediates are expected to be more stable than the PO inter- C H . In contrast, above 175°C, the Mars-van Krevelen (Mv K) mediatesduetothelowerstrainintheformercase. Intriguingly,our 3 6 mechanismusedlatticeoxygenastheactiveoxygenspecies. Itpre- findingssuggestedthat POwouldbethemoststablespeciesonthe ferentiallyattackedtheαC–Hbondofpropyleneandgraduallytook Cu O{110}surface,whileglycolateintermediatesexhibitedthehighest 2 theinitiativeinthe DEPreaction. Untilreaching250°C,asshownin stabilityonboth Cu O{111}and Cu O{110}surfaces. Thisphenomenon 2 2 Fig.5e,thevibrationalfeaturesof C H chemisorptionweresimilarto arosefromtheuniqueatomicarrangementofcopperandoxygenon 3 6 thoseof C H +O ,exceptthattheirintensitieswerereduced. These the{110}facet,whichprovidedaconfinedenvironmentconduciveto 3 6 2 DRIFTSresultspresentedthats-Cu O-NFcatalyzedthe DEPreaction the formation of a distorted adsorption structure for the glycolate 2 following the Mv K mechanism, and this trend became fully visible, intermediate. Thisassertionwassubstantiatedbytheelongationofthe resultingindeepoxidation. C-O bondlengthon Cu O{110}ascomparedto Cu O{111}and Cu O surf 2 2 2 Nature Communications|( 2025)1 6:7802 8 Article https://doi.org/10.1038/s41467-025-63059-0 {100}, as detailed in Supplementary Table S3. Notably, as shown in Synthesisofspherical Cu Onanoframes(s-Cu O-NF) 2 2 Fig. 5d, e and Supplementary Fig. 27, the IR features belonging to Typically, 91.1m L deionized water, 1.0m L Cu Cl (0.1mol L–1) and 2 glycolateintermediates,suchas C–Ostretching(1180–1210cm–1),and 0.87g SDSweresuccessivelyaddedintoaconicalflask. Theconical the IRbandlocatedat1578cm–1whichwasascribedtotheformationof flaskwasplacedina33°Coilbathwithstirringforthecompletedis- glycolate60–62 can be observed upon adsorption of C H +O mixed solutionof SDSpowder. Afterthat,5.0m LNa OHsolution(5.0mol L–1) 3 6 2 flowatboth150o Cand250°C. Consequently,thesefindingscameto wasadded,anditformedablueprecipitateof Cu(OH) immediately, 2 the conclusion that PO would be predominantly generated on the along with the solution color changing blue. Next, 4.5m L of Cu O{110}surfaceduringthe DEP,whileby-productswouldbemore NH OH·HCl (2.0mol L–1) was quickly injected, and the solution was 2 2 prevalentonthe Cu O{111}and Cu O{110},nicelyaccountingforthe stirredfor20suntilthecolorofthesolutionchangedfromlightblue 2 2 experimentalobservations. to green. After aging for 2h, 9.0m L HCl solution (2.0mol L–1) was To summarize, we have successfully employed a ligand-protec- introduced with stirring for another 20s, and the solution color tion/selective facet-etching strategy to synthesize internally hollow changedfromgreentoyellow. Afteretchingfor50min,theobtained Cu Onanoframes,amplifyingtheexposureof{110}facets. Ourwork product was washed with ethanol and deionized water for several 2 highlightsthedistinctadvantagesof Cu O{110}ingeneratingchemi- timesanddriedundervacuumat70°Cfor12h. 2 cally adsorbed oxygen species (O– or O –), which in turn markedly 2 improved DEPreactivity, POselectivityandcatalyststability,realizing Characterizationmethods atriple-win. Impressively,s-Cu O-NFachieved>99%POselectivityat Thecrystallinestructureidentificationofthe Cu Ocatalystswasana- 2 2 150°C, outperforming any other Cu 2 O nanocatalysts previously lyzedby X-raydiffraction(XRD, Rigaku Ultima IV)with Cu Kαradiation reported. Through comprehensive characterizations, we proposed (λ=1.54Å),whichoperatedatavoltageof35k Vandacurrentof15m A thatatrelativelylowtemperatures,the DEPreactionoccurredthrough withastepsizeof0.02°. The N adsorptionisothermswereobtained 2 the LH mechanismwith adsorbed O species acting asthe oxidant. using anautomaticphysical adsorber (ASAP2460, Micromeritics)at 2 Furthermore, DFT calculations have elucidated that PO becamethe −196°C and the surface area was determined by the Brunauer- dominantproductonlywhenitwasthermodynamicallymorefavor- Emmett-Teller(BET)model. X-rayphotoelectronspectroscopy(XPS, ablecomparedtootherintermediatesforthereactionofpropylene Scientific K-Alpha, Thermo Fisher)analysiswascarriedouttoobtain and O *. thesurfacepropertiesofsamplesunderatubevoltageof15k Vand 2 currentof10m A. Themorphologies,sizes andcrystallographicfea-
Methods
Methods turesofthe Cu Ocatalystswerecharacterizedbyscanningelectron 2 Materials microscopy(SEM, Zeiss, Sigma)andtransmissionelectronmicroscopy Copper chloride dihydrate (Cu Cl ·2H O), sodium hydroxide (TEM, Philips, TECNAIF30)withenergy-dispersive X-rayspectroscopy 2 2 (Na OH), hydroxylamine hydrochloride (NH OH·HCl), oleic acid (EDX)andadouble-tiltsampleholder. The Fouriertransform-infrared 2 (OA),glucose, L-ascorbicacid(AA),laurylsodiumsulfate(SDS), spectroscopy (FT-IR) of Cu O catalysts were obtained by a Nicolet 2 ethanol, cyclohexane and hydrochloric acid (HCl) were pur- 6700FT-IRspectrometerinthefrequencyrange3500–600cm−1. Dif- chased from Sinopharm Chemical Reagent Co., Ltd. Copper fusereflectanceinfrared Fouriertransformspectroscopy(DRIFTS)was sulfatepentahydrate(Cu SO ·5H O)waspurchasedfrom Macklin measured on a Nicolet 6700 FT-IR spectrometer equipped with an 4 2 Reagent company. Sodium potassium tartrate tetrahydrate insitu Fouriertransforminfraredreactioncellandaliquidnitrogen (SPTT),potassiumhydroxide (KOH),palladium chloride(Pd Cl ) cooled mercury-cadmium-telluride detector. The Raman spectra of 2 and copper acetate monohydrate ((CH COO) Cu·H O) were representative catalysts were measured at room temperature 3 2 2 purchased from Aladdin Biochemical Technology Co., Ltd. employing an instrument with a microscope attachment (Thermo Polyvinylpyrrolidone (PVP) was purchased from Beijing Wokai Fischer DXR2Xi). Localstructuresweredeterminedby Ramanspec- Biotechnology Co., Ltd. All chemical reagents were used as troscopy(DXR2Xi, Thermo Fisher)usinga532nmexcitationsource. received without further purification. All aqueous solutions Electrochemical impedance spectroscopy (EIS) measurements were were prepared using deionized water with a resistivity of based on a Biologic VMP3 electrochemical workstation. H 2 18.2MΩcm–1. temperature-programmed reduction (H -TPR) experiments were 2 conductedusingachemisorptionanalyzerequippedwithathermal Synthesisofspherical Cu Onanocrystals(s-Cu O) conductivitydetector(Baidewo Instruments, MFTP-3060). 2 2 s-Cu O was synthesized according to the method improved by the 2 reportedstudyof Zhanetal.90.073g Cu(NO ) ·3H Oand0.60g PVP O temperature-programmeddesorption(O -TPD) 32 2 2 2 were dissolved in 35.0m Ldeionized water with constantstir. Then, characterization 6.0m LNa OHsolution(0.20mol L–1)wasaddeddropwisetotheabove O -TPDexperimentswerecarriedoutina Micromeritics Autochem II 2 solution(~195s)andcontinuedtostirfor10min. Afterthat,afresh ASAP2920apparatuswasusedfor O -TPDexperimentsthatinvolved 2 ascorbicsolution(0.10mol L–1)wasaddeddropwisewithin10.5min, recording the signal and determining the oxygen species and their andthemixturewasadequatelystirredforanother10min. Theorigi- corresponding concentrations on the catalyst surface with a TCD nal product was washed with ethanol for several times, and dried (Baidewo Instruments, MFTP-3060). Typically,thenanocatalystsam- undervacuumat60°Covernight. ple (50.0mg) was purged for an hour at 150o C with argon flow (30m Lmin–1). Aftercoolingdowntoroomtemperature,itwassatu- Synthesisofspherical Cu Onanocages(s-Cu O-NC) ratedwith5%O /Arflow(30m Lmin–1)at50°Cfor2h. Aftertheabove 2 2 2 s-Cu O-NCwas synthesized according to the method reported by Yu steps,thenanocatalystsamplewasheatedfromroomtemperatureto 2 etal.63. Theexperimentalstepsareasfollows:1.40g(CHCOO) Cu·HO 520o Cwithheliumataheatingrateof5°Cmin–1. 3 2 2 wasaccuratelyweighedanddissolvedin35.0m Ldeionizedwater. The above precursor solution was poured into a 50.0m L steel autoclave Catalyticperformanceevaluation coatedwith Teflon. Subsequently,thesteelautoclavereactorwasplaced The DEP catalytic performance of synthesized Cu O catalysts was 2 inanovenfor24hat200°C. Theresultingmixturewascollectedby monitoredinacontinuousflowfixed-bedmicroreactoratatmospheric centrifugationandthoroughlywashedwithdeionizedwaterandetha- pressure with an axial quartz sheathed thermocouple. Typically, a nol. Thebrick-redproductwasdriedundervacuumatroomtempera- 50.0mg Cu Osamplewasequippedinaquartzreactor. Thereactant 2 tureovernight. gas mixture (CH : O : N =10: 5: 85 vol.%) with a flow rate of 3 6 2 2 Nature Communications|( 2025)1 6:7802 9 Article https://doi.org/10.1038/s41467-025-63059-0 30m Lmin−1 was fed into the reactor to start the reaction, and the 8. Khatib, S. J.&Oyama, S. T. Directoxidationofpropylenetopropylene reactiontemperaturewasincreasedfrom125°Cto275°Catarateof oxidewithmolecularoxygen:Areview. Catal. Rev.57,306–344(2015). 5°Cmin−1. Thelinesandvalvesofthereactorandthegaschromato- 9. Zhan, C.etal. Criticalrolesofdoping Clon Cu Onanocrystalsfor 2 graphs were heated to 110°C, preventing the condensation of directepoxidationofpropylenebymolecularoxygen. J. Am. Chem. products64,65. The reactants and product were analyzed via two gas Soc.142,14134–14141(2020). chromatographs(GC)equippedwiththreecolumnsinrealtime. The 10. Li, W.etal. Directpropyleneepoxidationwithmolecularoxygen outlet CO , C H , and O were analyzed by a thermal conductivity overcobalt-containingzeolites. J. Am. Chem. Soc.144,4260–4268 2 3 6 2 detector(TCD,separatedbya Porapakcolumnandamolecularsieve (2022). 5Acolumn). Theoutletpropyleneoxide(PO),acrolein,acetone,pro- 11. Hua, Q.etal. Crystal-plane-controlledselectivityof Cu Ocatalysts 2 pionaldehyde,isopropanolandacetaldehydeproductsweredetected inpropyleneoxidationwithmolecularoxygen. Angew. Chem. Int. byaflameionizationdetector(FID,separatedbyacapillarycolumn). Ed.126,4956–4961(2014). Consideringthedifferencein carbonnumber ofdifferentproducts, 12. Nguyen, V.etal. Directgas-phasephotocatalyticepoxidationof the catalytic activity, selectivity of PO and its formation rate were propylenewithmolecularoxygenbyphotocatalysts. Chem. Eng. J. calculatedasthefollowingequations: 179,285–294(2012). 13. Xiong, W.etal. Finecubic Cu Onanocrystalsashighlyselective 2 X =R ×A ð4Þ catalystforpropyleneepoxidationwithmolecularoxygen. Nat. i i i Commun.12,5921(2021). X X X X 14. Song, Y.&Wang, G. Theoreticalstudyofpropyleneepoxidation 2X X X = X + Products, C2 + Products, C1 ð5Þ over Cu O(111)surface:Activityof O2–, O–,and O –species. J. Phys. i Products, C3 3 3 2 2 Chem. C122,21500–21513(2018). 15. Wang, Q., Zhan, C., Zhou, L., Fu, G.&Xie, Z. Effectsof Cl−on Cu O X 2 X C = i ×100% ð6Þ nanocubesfordirectepoxidationofpropylenebymolecularoxy- C3H6 XC3H6,infeed gen. Catal. Commun.135,105897(2020). 16. Torquato, L. D. M.etal. Relationbetweenthenatureofthesurface S = N i ×P X i ×100% ð7Þ facetsandthereactivityof Cu 2 Onanostructuresanchoredon i 3 X i Ti O 2 NT@PDAelectrodesinthephotoelectrocatalyticconversionof CO tomethanol. Appl. Catal. BEnviron. Energy261,118221 2 (2020). POformationrate= C C3H6 ×F C3 R H × 6 × T P × r m essure×S PO ð8Þ 17. Li, Y.etal. Effectiveperiodateactivationbypeculiar Cu 2 Onano- crystalforantibioticsdegradation:Thecriticalroleofstructureand Where X i represents the concentration of reactants or products, A i underlyingmechanismstudy. Appl. Catal. BEnviron. Energy341, representsconversionofthepeakareaobtainedby GC, R iistherela- 123351(2024). tivecorrectionfactorresultingfromthestandardgases, C C3H6repre- 18. An, Y.etal. Hollowstructuredcopper-loadedself-floatingcatalyst sents C 3 H 6 conversion, X C3H6,infeed represents the concentrations of insulfite-inducedoxidationofarsenic(III)atneutralp H:Kineticsand C 3 H 6 attheinlet, N irepresentsthecarbonnumberineachproduct, S i mechanismsinvestigation. Chem. Eng. J.407,127193(2021). representstheselectivityofvariousproducts. F C3H6representstheflow 19. Dong, Y., Jiang, X., Mo, J., Zhou, Y.&Zhou, J. Hollow Cu Onano- rateof C 3 H 6 ,m catrepresentsthemassofthecatalyst. particlesincarbonmicrospherespreparedfromcellulose- cuprammoniumsolutionasanodematerialsfor Li-ionbatteries. Dataavailability Chem. Eng. J.381,122614(2020). Datawillbemadeavailableonrequest. Sourcedataareprovidedin 20. Sui, Y.etal. Synthesisof Cu Onanoframesandnanocagesby 2 thispaper. selectiveoxidativeetchingatroomtemperature. Angew. Chem. Int. Ed.49,4282–4285(2010).
References
References 21. Shang, Y.etal. Afaciletop-downetchingtocreatea Cu Ojagged 2 1. Teržan, J., Huš, M., Likozar, B.&Djinović, P. Propyleneepoxidation polyhedroncoveredwithnumerous110edgesand111cornerswith usingmolecularoxygenovercopper-andsilver-basedcatalysts:A enhancedphotocatalyticactivity. Chem. Eur. J.18,14261–14266 review. ACSCatal.10,13415–13436(2020). (2012). 2. Du, M.etal. Ionicliquid-enhancedimmobilizationofbiosynthesized 22. Tang, L.etal. One-potsynthesisofetched Cu Ocubeswith 2 Aunanoparticleson TS-1towardefficientcatalystsforpropylene exposed{110}facetswithenhancedvisible-light-drivenphoto- epoxidation. J. Catal.283,192–201(2011). catalyticactivity. Phys. Chem. Chem. Phys.17,29479–29482 3. Lin, M., Xia, C., Zhu, B., Li, H.&Shu, X. Greenandefficientepox- (2015). idationofpropylenewithhydrogenperoxide(HPPOprocess)cat- 23. Tsai, Y.etal. Fabricationofdiverse Cu Onanoframesthroughface- 2 alyzedbyhollow TS-1zeolite:A1.0kt/apilot-scalestudy. Chem. selectiveetching. J. Phys. Chem. C117,24611–24617(2013). Eng. J.295,370–375(2016). 24. Kuo, C.&Huang, M. Facilesynthesisof Cu Onanocrystalswith 2 4. Wu, J., Wu, Y.&Wang, J. Comparativestudyoncorrosionbehavior systematicshapeevolutionfromcubictooctahedralstructures. J. of Cuand Snunder UVlightilluminationinchloride-containing Phys. Chem. C112,18355–18360(2008). boratebuffersolution. Corros. Sci.186,109471(2021). 25. Qin, C., Campbell, B. M., Shen, M., Zhao, T.&Sadtler, B. Light- 5. Nijhuis, T. A., Makkee, M., Moulijn, J. A.&Weckhuysen, B. M. The driven,facet-selectivetransformationofcuprousoxidemicro- productionofpropeneoxide: Catalyticprocessesandrecent crystalstohollowcoppernanoshells. Chem. Mater.31,8000–8011 developments. Ind. Eng. Chem. Res.45,3447–3459(2006). (2019). 6. Kube, P.etal. Greensynthesisofpropyleneoxidedirectlyfrom 26. Wang, W.etal. Aphotocatalystofcuprousoxideanchored MXene propane. Nat. Commun.13,7504(2022). nanosheetfordramaticenhancementofsynergisticantibacterial 7. Wang, L.etal. Titaniumsilicalite-1zeoliteencapsulating Auparticles ability. Chem. Eng. J.386,124116(2020). asacatalystforvaporphasepropyleneepoxidationwith H/O:a 27. Subhan, F.etal. Ammoniaassistedfunctionalizationofcuprous 2 2 matterof Au–Tisynergicinteraction. J. Mater. Chem. A8,4428–4436 oxidewithinconfinedspacesof SBA-15foradsorptivedesulfuriza- (2020). tion. Chem. Eng. J.339,557–565(2018). Nature Communications|( 2025)1 6:7802 10 Article https://doi.org/10.1038/s41467-025-63059-0 28. Zhang, Y., Deng, B., Zhang, T., Gao, D.&Xu, A.-W. Shapeeffectsof 49. Wang, Y., Ma, W., Wang, D.&Zhong, Q. Studyonthereaction Cu Opolyhedralmicrocrystalsonphotocatalyticactivity. J. Phys. mechanismofthepropyleneoxiderearrangementviainsitu 2 Chem. C114,5073–5079(2010). DRIFTS. Chem. Eng. J.307,1047–1054(2017). 29. Bhaisare, M. L., Khan, M. S., Pandey, S., Gedda, G.&Wu, H.-F. 50. Hu, Y., Wan, X., Tie, S., Lan, S.&Gao, X. Synthesisofanewporous Shape-orientedphotodynamictherapyofcuprousoxide(Cu O) Cu O@Cu (SO )(OH) ·H Ocore-shellthermocatalystbyinsitu 2 2 4 4 62 2 nanocrystalsforcancertreatment. RSCAdv.7,23607–23614(2017). etchingforefficientcatalyticdegradationofdyeunderdarkcon- 30. Guo, Y.etal. Fabricationof Ag–Cu O/reducedgrapheneoxide dition. J. Solid State Chem.298,122107(2021). 2 nanocompositesassurface-enhancedramanscatteringsubstrates 51. Xu, Y.etal. Ultralow-content Pdin-situincorporationmediated forinsitumonitoringofperoxidase-likecatalyticreactionandbio- hierarchicaldefectsincorner-etched Cu Ooctahedraforenhanced 2 sensing. ACSAppl. Mater. Interfaces9,19074–19081(2017). electrocatalyticnitratereductiontoammonia. Appl. Catal. B 31. Wei, B., Yang, N., Pang, F.&Ge, J. Cu O–Cu Ohollownanospheres Environ. Energy306,121094(2022). 2 asaheterogeneouscatalystforsynergeticoxidationof CO. J. Phys. 52. Huo, H.etal. Solar-driven CO -to-ethanolconversionenabledby 2 Chem. C122,19524–19531(2018). continuous CO transportviaasuperhydrophobic Cu Onano 2 2 32. Al-Saeedi, S. I., Al-Senani, G. M., Abd-Elkader, O. H.&Deraz, N. M. fence. Chem. Sci.15,1638–1647(2024). Onepotsynthesis,surfaceandmagneticpropertiesof Cu O/Cu 53. Zhu, Y.etal. Regulating Ce O morphologiesonthecatalyticoxi- 2 2 and Cu O/Cu ONanocomposites. Crystals11,751(2021). dationoftolueneatlowertemperature:Astudyofthe 2 33. Wang, W.etal. Propertiesanalysisofself-luminouscement-based structure–activityrelationship. J. Catal.418,151–162(2023). materialswithdifferentcolorsandtheirvisualcomfortevaluation 54. Wang, Y.etal. Metalsupportinteractionofdefective-rich Cu Oand onpavement. Sol. Energy247,214–227(2022). Auwithenhanced COlow-temperaturecatalyticoxidationand 34. Shoeib, M. A., Abdelsalam, O. E., Khafagi, M. G.&Hammam, R. E. moistureresistance. Adv. Powder Mater.2,100119(2023). Synthesisof Cu Onanocrystallitesandtheiradsorptionandpho- 55. Chen, S.etal. Probingsurfacestructuresof Ce O , Ti O ,and Cu O 2 2 2 2 tocatalysisbehavior. Adv. Powder Technol.23,298–304(2012). nanocrystalswith COand CO Chemisorption. J. Phys. Chem. C 2 35. Ahmed, A., Gajbhiye, N. S.&Joshi, A. G. Lowcost,surfactant-less, 120,21472–21485(2016). onepotsynthesisof Cu Onano-octahedraatroomtemperature. J. 56. Liang, M., Zhu, X.&Ma, W. Thepropyleneoxiderearrangement 2 Solid State Chem.184,2209–2214(2011). catalyzedbythelewisacidsitesof ZSM-5catalystwithcontrollable 36. Teng, Z.etal. COself-sustainedcatalyticcombustionovermorpho- surfaceacidity. Catal. Lett.149,942–949(2019). logicalinversemodel Ce O/Cu Ocatalystsexposing(100),(111) 57. Yang, H., Li, G., Jiang, G., Zhang, Z.&Hao, Z. Heterogeneous 2 2 and(110)planes. Appl. Catal. BEnviron. Energy339,123119(2023). selectiveoxidationoversupportedmetalcatalysts:Fromnano- 37. Zou, W.etal. Crystal-planeeffectsonsurfaceandcatalyticprop- particlestosingleatoms. Appl. Catal. BEnviron. Energy325, ertiesof Cu2Onanocrystalsfor NOreductionby CO. Appl. Catal. 122384(2023). Gen.505,334–343(2015). 58. Liu, X.etal. Defectstabilizedgoldatomsongrapheneaspotential 38. Bera, P., Cámara, A. L., Hornés, A.&Martínez-Arias, A. Comparative catalystsforethyleneepoxidation:Afirst-principlesinvestigation. insitu DRIFTS-MSstudyof12CO-and13CO-TPRon Cu O/Ce O cat- Catal. Sci. Technol.6,1632–1641(2016). 2 alyst. J. Phys. Chem. C113,10689–10695(2009). 59. Li, Z.etal. Recentadvancementsingraphene-basedsupportsof 39. Hornés, A.etal. CO-TPR-DRIFTS-MSinsitustudyof Cu O/ metalcomplexes/oxidesforepoxidationofalkenes. Chem. Asian J. Ce 1−x Tb x O 2−y (x=0,0.2and0.5)catalysts:Supporteffectsonredox 13,3790–3799(2018). propertiesand COoxidationcatalysis. J. Catal.268,367–375 60. Xu, G.etal. Alkalietchingofporous Pd Co Znnanosheetsfor (2009). boosting C−Cbondcleavageofethyleneglycoloxidation. Small 40. Yang, T.etal. Coordinationtailoringof Cusinglesiteson C N 20,2306341(2024). 3 4 realizesselective CO hydrogenationatlowtemperature. Nat. 61. Amenabar, I.etal. Hyperspectralinfrarednanoimagingoforganic 2 Commun.12,6022(2021). samplesbasedon Fouriertransforminfrarednanospectroscopy. 41. Zuo, H.etal. Hydrogenproductionand CO capturefrom Linz- Nat. Commun.8,14402(2017). 2 Donawitzconvertergasviaachemicalloopingconcept. Chem. 62. Kang, J.&Kang, D. Effectofaminoacid-derivednitrogenand/or Eng. J.477,146870(2023). sulfurdopingonthevisible-light-drivenantimicrobialactivityof 42. Zhang, Z.etal. Structuresensitivityof Cu Oin COoxidationover carbonquantumdots:Acomparativestudy. Chem. Eng. J.420, Ce O -Cu O/Cu Ocatalysts. J. Catal.405,333–345(2022). 129990(2021). 2 2 43. Zhang, S.etal. Sub-3nmultrafine Cu Oforvisiblelightdriven 63. Yu, H., Yu, J., Liu, S.&Mann, S. Template-freehydrothermal 2 nitrogenfixation. Angew. Chem. Int. Ed.60,2554–2560(2021). synthesisof Cu O/Cu Ocompositehollowmicrospheres. Chem. 2 44. Ye, Y.etal. Cellwalldestructionandinternalcascadesynergistic Mater.19,4327–4334(2007). antifungalstrategyforfungalkeratitis. ACSNano16,18729–18745 64. Zhang, X.etal. Activationofmolecularoxygenover Mn-doped (2022). La Cu O perovskitefordirectepoxidationofpropylene. Catal. Sci. 2 4 45. Yao, S.etal. Anchoringultrafine Cu Onanoclusteron PCNfor CO Technol.12,2426–2437(2022). 2 2 photoreductioninwatervaporwithmuchimprovedstability. Appl. 65. Lei, J.etal. Insightintotheeffectofcoppersubstitutiononthe Catal. BEnviron. Energy317,121702(2022). catalyticperformanceof La Co O -basedcatalystsfordirectepox- 3 46. Xiao, M.etal. Quasi-singlecrystallinecuprousoxidewafersvia idationofpropylenewithmolecularoxygen. ACSSustain. Chem. stress-assistedthermaloxidationforoptoelectronicdevices. Adv. Eng.9,794–808(2021). Funct. Mater.32,2110505(2022). 47. Lv, T.-T.etal. Protectionofhighlyactivesiteson Cu Onanocages: Acknowledgements 2 anefficientcrystallinecatalystforammoniumperchlorate Thisresearchwassupportedbythe National Key Researchand Devel- decomposition. Cryst Eng Comm22,8214–8220(2020). opment Programofthe Ministryof Scienceand Technologyof China 48. Wang, Y., Chu, H., Zhu, W.&Zhang, Q. Copper-basedefficient (No.2022YFA1504601),andthe National Natural Science Foundationof catalystsforpropyleneepoxidationbymolecularoxygen. Catal. China(No.22478322,22132004, U24A20489and22038012),andthe Today131,496–504(2008). Natural Science Foundationof Fujian Province(No.2021J01022). Nature Communications|( 2025)1 6:7802 11 Article https://doi.org/10.1038/s41467-025-63059-0 Authorcontributions Reprintsandpermissionsinformationisavailableat G. F., Q. L.,and J. H.conceivedandsupervisedtheresearch. Y. Q., http://www.nature.com/reprints R. Z.,and M. H.performedthesynthesis,mostofthestructural characterizations,and DEPtests. K. T.assistedwiththeexperi- Publisher’snote Springer Natureremainsneutralwithregardtojur- ments. G. F.and Y. Z.performedthe DFTsimulations. G. F., Y. Q., Y. Z., isdictionalclaimsinpublishedmapsandinstitutionalaffiliations. G. Z.,and J. H.preparedandcheckedthepaper. Y. Q.and Y. Z. contributedequallytothiswork. Allauthorsdiscussedtheresults Open Access Thisarticleislicensedundera Creative Commons andcommentedonthepaper. Attribution-Non Commercial-No Derivatives4.0International License, whichpermitsanynon-commercialuse,sharing,distributionand Competinginterests reproductioninanymediumorformat,aslongasyougiveappropriate Theauthorsdeclarenocompetinginterests. credittotheoriginalauthor(s)andthesource,providealinktothe Creative Commonslicence,andindicateifyoumodifiedthelicensed Additionalinformation material. Youdonothavepermissionunderthislicencetoshareadapted Supplementaryinformation Theonlineversioncontains materialderivedfromthisarticleorpartsofit. Theimagesorotherthird supplementarymaterialavailableat partymaterialinthisarticleareincludedinthearticle’s Creative https://doi.org/10.1038/s41467-025-63059-0. Commonslicence,unlessindicatedotherwiseinacreditlinetothe material. Ifmaterialisnotincludedinthearticle’s Creative Commons Correspondenceandrequestsformaterialsshouldbeaddressedto licenceandyourintendeduseisnotpermittedbystatutoryregulationor Gang Fu, Qingbiao Lior Jiale Huang. exceedsthepermitteduse,youwillneedtoobtainpermissiondirectly fromthecopyrightholder. Toviewacopyofthislicence,visithttp:// Peerreviewinformation Nature Communicationsthanks Xinqing Chen creativecommons.org/licenses/by-nc-nd/4.0/. andtheotheranonymousreviewer(s)fortheircontributiontothepeer reviewofthiswork. Apeerreviewfileisavailable. ©The Author(s)2025 Nature Communications|( 2025)1 6:7802 12