Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Jun 17.
Published in final edited form as: Chem Catal. 2021 Mar 2;1(1):106–116. doi: 10.1016/j.checat.2021.02.001

Chemo- and Stereoselective Intermolecular [2+2] Photocycloaddition of Conjugated Dienes using Colloidal Nanocrystal Photocatalysts

Yishu Jiang 1, Muwen Yang 1, Yue Wu 1, Rafael López-Arteaga 1, Cameron R Rogers 1, Emily A Weiss 1,2,*
PMCID: PMC8323757  NIHMSID: NIHMS1676176  PMID: 34337591

SUMMARY

The use of visible-light photosensitizers to power [2+2] photocycloadditions that produce complex tetrasubstituted cyclobutanes is a true success of photochemistry, but the scope of this reaction has been limited to activated α, β-unsaturated carbonyls. This paper describes selective intermolecular homo- and hetero-[2+2] photocycloadditions of terminal and internal aryl conjugated dienes – substrates historically unsuited for this reaction because of their multiple possible reaction pathways and product configurations – through triplet-triplet energy transfer from CdSe nanocrystal photocatalysts, to generate valuable and elusive syn-trans aryl vinylcyclobutanes. The negligible singlet-triplet splitting of nanocrystals’ excited states allows them to drive the [2+2] pathway over the competing [4+2] photoredox pathway, a chemoselectivity not achievable with any known molecular photosensitizer. Reversible tethering of the cyclobutane product to the nanocrystal surface results in near quantitative yield of the syn-trans product. Flat colloidal CdSe nanoplatelets produce cyclobutanes coupled at the terminal alkenes of component dienes with up to 89% regioselectivity.

Graphical Abstract

graphic file with name nihms-1676176-f0001.jpg

INTRODUCTION

Photochemical couplings of alkenes to form cyclic structures, particularly [2+2] photocycloadditions, are a powerful tool for simple construction of molecules that are difficult to generate with thermal reactions1–3. Methods that power [2+2] photocycloadditions with visible light – through absorption of photons by molecular photosensitizers4,5 and subsequent triplet-triplet energy transfer (TTEnT) to alkene substrates – have increased the energy efficiency and broadened the applications of these reactions by eliminating the need for UV light. Molecular and solid state templating strategies have led to enantioselective6–11 and diastereoselective12,13 couplings of certain substrates, and colloidal photocatalysts (PCs) have enabled regio- and diastereoselective couplings14. Despite these successes, the scope of substrates for [2+2] photocycloadditions is still largely confined to activated olefins, enones9–11,15 and α, β-unsaturated carboxylic acid derivatives7,8,16 Aryl-conjugated alkenes, including stilbenes and aryl conjugated dienes (like those shown in Figure 1A), are rarely involved in sensitized [2+2] photocycloadditions17–21 despite the facts that (i) they are common and useful precursors in small-scale and industrial organic synthesis, and (ii) their [2+2] coupling products, vinylcyclobutane derivatives, are skeletons of many bioactive natural products22, including pipercyclobutanamide A and piperarborenine B23, and versatile intermediates for alkene additions, ring expansions, and pericyclic reactions24,25. In fact, there are no examples of high-yield, stereoselective [2+2] photocycloadditions of aryl conjugated dienes with themselves or other olefins, either with direct or photosensitized excitation of the substrate.

Figure 1. Substrates and Photocatalysts.

Figure 1.

Open in a new tab

A) Structures of the aryl conjugated diene substrates, 4-(1,3-butadien-1-yl)-benzoic acid 1 and 4-(4-phenylbuta-1,3-dienyl)benzoic acid 2; the heterocoupling partner, 4-(3-oxobut-1-en-1-yl)benzoic acid 3; and the PCs used in this study. The radii of the CdSe cores ranged from 1.2 nm to 1.4 nm. The thickness of the CdSe NPLs is 1.5 nm and its lateral dimensions are 29 nm × 8 nm. The “cross-bond” structure of 1 indicates a trans:cis ratio of 5:2 in the starting material. B) Scheme for the TTEnT process between CdSe QDs and substrate 1 anchored by its carboxylate. Upon direct excitation of the QD to its singlet-like “bright” exciton, it populates the triplet-like “dark” exciton, from which it transfers energy to the substrate to form its triplet excited state, T1, the direct precursor for the [2+2] cycloaddition reaction.

Why has the design of systems for this photochemical coupling been intractable? The first issue is chemoselectivity. Compared with analogous couplings of aryl-conjugated enones, the TTEnT-triggered [2+2] coupling of aryl conjugated dienes is not competitive with the large number of additional reaction pathways available from their singlet excited states and radical intermediates26–28, in particular, the photo-redox-driven [4+2] cycloaddition17, Figure 2A. As a result, the few known [2+2] diene-olefin photocycloadditions (none of which involve aryl conjugated dienes) are intramolecular couplings18,19,29.

Figure 2. Origin of the Chemoselectivity of [2+2] over [4+2] Photocycloadditions with CdSe Nanocrystals.

Figure 2.

Open in a new tab

A) Product skeletons generated from TTEnT-driven [2+2] cycloadditions and photoredox-driven [4+2] Diels-Alder cycloadditions of conjugated dienes. B) Triplet excited state energies (left), oxidation potentials (middle) and reduction potentials (right) of common PCs and aryl-conjugated alkene and conjugated diene substrates (see Figures S1, S2, 6 and S7), including the three PCs and two substrates from this work (bold). A range of energies corresponding to the range of possible particle radii is shown for spherical nanocrystals. PCs in green, blue, and orange have enough triplet energy to sensitize some or all of the listed substrates via TTEnT, but PCs in blue (green) are also able to oxidize (reduce) these alkenes. Only the nanocrystals (orange) can perform TTEnT without the possibility of an accompanying photoredox process.

The second issue with this coupling is selectivity. The regioselectivity of [2+2] diene-olefin photocycloadditions is low due to the presence of multiple alkene coupling sites per molecule. The diastereoselectivity is also intrinsically poor, so the vinylcyclobutane product is in both syn- and anti- configurations, with both cis- and trans- configurations of the remaining alkenes. There are examples of stereoselective [2+2] photocycloadditions of dienes with olefins in the solid state, but the scope is limited (e.g., no conjugated dienes)30–34.

We and others have previously applied colloidal nanocrystals as PCs for a variety of C-C coupling reactions35–37. In particular, we introduced quasi-spherical CdSe quantum dots (QDs) as stable, recyclable, and stereoselective PCs for TTEnT-triggered [2+2] photocycloadditions of carboxylate-functionalized chalcones to produce syn tetrasubstituted cyclobutanes14,38. Here we demonstrate that CdSe nanocrystals, including QDs, core/shell QDs, and quasi-2D nanoplatelets (NPLs), photocatalytically produce the significantly more elusive aryl-conjugated vinylcyclobutanes with high chemo-, regio-, and stereoselectivity, through TTEnT-driven [2+2] couplings of terminal and internal aryl conjugated dienes, Figure 1.

The use of colloidal CdSe nanocrystals as PCs uniquely enables complete selectivity for the TTEnT-driven [2+2] pathway over the photoredox-driven [4+2] pathway, because the negligible energy splitting between the singlet-like and triplet-like excitons of CdSe nanocrystals (<20 meV)39 allows them to have both (i) sufficiently high triplet energies to access the triplet excited states of aryl-conjugated alkenes and (ii) sufficiently mild oxidation and reduction potentials to eliminate the possibility of a photoredox reaction. Even though it is theoretically possible to construct a molecular PC that satisfies these two conditions for the substrates shown in Figure 1A and related conjugated dienes, there is no known molecular PC/photosensitizer that does, as shown in Figure 2B. We note that we have previously published spectroscopic evidence that nanocrystal-alkene systems similar to those studied here undergo energy transfer from the nanocrystal to the alkene14, and that singlet-singlet Förster Resonance Energy Transfer (FRET) is thermodynamically prohibited. This evidence, plus the tendency of the [2+2] pathway to go through an excited state rather than a radical, all support our claim that the [2+2] reactions are TTEnT-triggered.

The yield of [2+2] coupling by the nanocrystals also benefits from the high photoluminescence quantum yield (PL QY) of core/shell QDs40,41 and NPLs42, which leads to more frequent formation of the triplet excited state of the substrate, and from reversible tethering of substrates to the particle surface, which enhances the yield of TTEnT over the diffusion-controlled TTEnT from molecular PCs. With respect to the selectivity of the [2+2] reaction, we present evidence that reversible tethering of the cyclobutane product to the nanocrystal surface allows us to choose which double bonds of the dienes couple by tuning the radius of curvature of the particle. Finally, we show that nanocrystal-catalyzed reactions of aryl conjugated dienes retain the exceptional syn-diastereoselectivity of QD-catalyzed [2+2] photocycloadditions of the chalcone substrates that we reported previously14 and enforce the syn-trans configuration of the product. In particular, our CdSe NPLs accomplish the first high-yield intermolecular [2+2] photocycloaddition (homo- or heterocoupling) of internal aryl conjugated dienes to form a syn-trans vinylcyclobutane derivative coupled at terminal alkene site as the major product.

RESULTS AND DISCUSSION

The Supplementary Information (“SI”) contains details of the syntheses, and electrochemical and spectroscopic characterization of the substrates and the CdSe nanocrystals, and describes our illumination setup, which includes a white LED and fan-cooled sample holder. In all reactions we present here, the nanocrystals are separable from the organic reaction mixtures through simple centrifugation procedures, and product mixtures were analyzed by high resolution mass spectrometry, 1H NMR, 13C NMR, NOESY, COSY, HMBC and HSQC, see the SI. All CdSe nanocrystals are stable under extended illumination; after a 4-day illumination cycle, we observed no etching, aggregation or other degradation (see Figure S5). We determined product yields as described in the figure captions and the SI.

Of two major classes of conjugated dienes, terminal conjugated dienes (compound 1, Figure 1A) are the more commonly used substrates for [2+2] photocycloadditions because they have fewer potential product configurations and competing isomerization pathways; however, the competing photoredox-driven [4+2] photocycloaddition generally dominates the reaction pathway distribution for this type of substrate. Figure 3 shows that, in the homo- and heterocoupling of terminal aryl conjugated dienes, use of CdSe QDs as PCs allows us to (i) selectively drive [2+2] photocycloaddition over [4+2] photocycloaddition in cases where molecular PCs drive both types of reactions, and (ii) selectively form the syn-trans [2+2] cyclobutane product in cases where molecular PCs have no selectively for the syn/anti-cyclobutane nor the trans/cis configuration of the remaining alkene.

Figure 3. Chemo- and Stereoselective Intermolecular [2+2] Photocycloadditions of the Terminal Aryl Conjugated Diene.

Figure 3.

Open in a new tab

Conditions and product distributions for the homocoupling of 4-(1,3-butadien-1-yl)-benzoic acid 1 (A) (trans:cis ratio of 5:2) and its heterocoupling with 4-(3-oxobuten-1-yl)benzoic acid 3 (B), using either CdSe QDs, Ir(ppy)3, or (Ir[dF(CF3)ppy]2(dtbpy))PF6 as PCs. No substrates were directly photoexcited. Each product structure is only drawn once, and is otherwise denoted by its compound number. Compounds with (orange/blue/black) labels are (homocoupled [4+2]/homocoupled [2+2]/heterocoupled [2+2]) products. The yields listed for CdSe QD reactions are isolated yields for the regioisomer and diastereomer drawn. The remainders of the yields (with respect to 1) for the QD-catalyzed reactions are accounted for by starting material 1 (with trans:cis ~ 2:1 in both cases). The yields listed for the molecule-driven reactions are NMR yields; the remainder of the yields (with respect to 1) are accounted for by starting material 1 (trans-isomers and cis-isomers) for all reactions, see Figures S8. In B, the molecular PCs and CdSe QDs also produced [2+2] homocoupled product of 3, see Figure S9.

Figure 3A shows the results of the photodimerization of 4-(1,3-butadien-1-yl)-benzoic acid 1 using either QDs or two common Ir complexes, fac-(Ir(ppy)3) and (Ir[dF(CF3)ppy]2(dtbpy))PF6, as PCs. CdSe QDs generate only syn-trans-[2+2] cyclobutane products (with diastereomeric ratio d.r. > 40:1) with a yield of 94%; the remainder of the yield is accounted for by starting material, 1. In contrast, the molecular Ir PCs generate a mixture of anti-trans and syn-trans major [4+2] products (5 and 6, d.r. = 2:1, 2:1, respectively) and minor [2+2] products (7 and 4, d.r. = 4:1, 2:1, respectively). Figure 3B shows the results of the heterocoupling of 1 with a 3× excess of 4-(3-oxobut-1-en-1-yl)benzoic acid 3 by the same set of PCs. The molar excess of 3 is introduced here to make the heterocoupling of 1 and 3 more competitive with the homocoupling of 1. Again, the only coupled products the QD generates are syn-trans-[2+2] products: 69% heterocoupled cyclobutane 8 (d.r. = 20:1), 18% homocoupled cyclobutane 4 (d.r. > 40:1) (no cis product detected). The remaining yield is accounted for by starting material. The molecular Ir PCs however produce a large yield of [4+2] homoproducts of 1, and only have 6% yield and 11% yield, respectively, of [2+2] heterocoupling. Neither molecular Ir PC enables a selectivity for the syn or anti diastereomer of the cyclobutane or for the cis or trans configuration of the remaining alkene.

The QDs are selective for the [2+2] products over the [4+2] products because their triplet-like excitonic state is sufficiently energetic (ET = 2.55 eV) to form the triplet excited state of substrate 1 (ET = 2.50 eV), but their excited state oxidation and reduction potentials (Epc = −1.57 V, Epa = +0.73 V vs. SCE)43 are not large enough to donate an electron or hole to the substrate (Epc = −2.11 V, Epa = +1.62 V vs. SCE), see Figures S1, S2 and S6. The TTEnT pathway to the [2+2] product is therefore energetically accessible and the photoredox pathway to the [4+2] product is not. In contrast, molecular PCs, Ir(ppy)3 (ET = 2.52 eV, E1/2 III/II = −2.29 V vs. SCE) and (Ir[dF(CF3)ppy]2(dtbpy))PF6 (ET = 2.68 eV, E1/2 III/II = +1.67 V vs. SCE) can access both the TTEnT pathway to form the [2+2] products and the photoredox pathway to form the [4+2] products, see Figure S7.44 Clearly, the availability of the photoredox [4+2] pathway severely limits the chemoselectivity of the desired [2+2] reaction for the molecular PCs, even if that pathway is only available for one of the two reagents present (here 1, see Scheme S2), and even if that reagent is the limiting reagent, as is the case here. We emphasize that the ability of the QD PC to select for the [2+2] pathway when molecular PCs cannot (i) applies to this entire class of aryl conjugated diene substrates (Figure 2A) and all common molecular PCs used for cycloadditions, not just the substrates and PCs in this work, and (ii) is a direct consequence of the near-degeneracy of the singlet-like and triplet-like excitonic states of the QD, which allows the QD to donate all of the absorbed photon energy to the triplet acceptor.

We have shown previously that the syn selectivity of the QD-catalyzed homo- and heterocouplings, which is retained here for aryl conjugated diene substrates, is a consequence of the configuration in which pairs of the substrates adsorb to the QD surface45. Substrates (and resulting products) are reversibly anchored to the surfaces of the CdSe nanocrystals46 by carboxylate groups; <5% yield of any coupling product is detected for dienes without carboxylic acid substituents, see Scheme S1. We believe, but have not proven, that, similarly, the trans selectivity of the QD-catalyzed reactions is due to tethering of the vinylcyclobutane product to the QD surface such that its photoisomerization (upon incidental subsequent re-excitation) is inhibited.

Figure 4 shows homo- and heterocoupling of an internal aryl conjugated diene (compound 2, Figure 1A). The use of this class of molecules in [2+2] photocycloadditions has been severely limited because, in addition to the challenges associated with terminal conjugated dienes, internal aryl conjugated dienes participate in very low-yield TTEnT from molecular photosensitizers28 and have numerous potential product configurations26,27. Figure 4 shows that CdSe nanocrystal PCs allow us to (i) perform [2+2] photocycloadditions of an aryl conjugated diene that is unreactive in the presence of molecular PCs, (ii) selectively form syn-trans [2+2] vinylcyclobutane products from that substrate, and (iii) select which of the double bonds couple in the [2+2] photocycloaddition by adjusting the radius of curvature – and therefore the density of vertices and other structural defects – of the surface of the nanocrystal PC.

Figure 4. Chemo- and Stereoselective Intermolecular [2+2] Photocycloadditions of the Internal Aryl Conjugated Diene.

Figure 4.

Open in a new tab

Conditions and product distribution of the homocoupling and heterocoupling of the 4-(4-phenylbuta-1,3-dienyl)benzoic acid 2 with 4-(3-oxobuten-1-yl)benzoic acid 3 photocatalyzed by CdSe NPLs, CdSe/CdS core-shell QDs, CdSe core-only QDs, (Ir[dF(CF3)ppy]2(dtbpy))PF6 and Ir(ppy)3. Each product structure is only drawn once, and is otherwise denoted by its compound number. Compounds with (orange/blue/black) labels are ([4+2]/homocoupled [2+2]/heterocoupled [2+2]) products. The yields listed for the CdSe NPLs for the homocoupling reaction, and for the CdSe/CdS core-shell QDs for the heterocoupling reaction, are isolated yields for the diastereoisomer drawn; the yields listed for the remainder of reactions are NMR yields. The remainders of the yields (with respect to 2) for the CdSe PC reactions are accounted for by starting material 2 (majority trans and minority cis isomers). The remainders of the yields (with respect to 2) for the Ir PC reactions are accounted for by starting material 2 (trans isomer only), see Figures S10 and S11. All reactions with CdSe nanocrystals were illuminated with only >490 nm light so as not to excite the substrate directly. The reactions with molecular PCs were illuminated with the LED without a filter to achieve sufficient excitation of the PCs, but control experiments (see Table S1) show that any incidental direct excitation of the substrates did not lead to measurable product.

Figure 4A shows the results of the homocoupling of 4-(4-phenylbuta-1,3-dienyl)benzoic acid 2 photocatalyzed by three CdSe nanocrystals – CdSe QDs, CdSe/CdS core/shell QDs, and CdSe nanoplatelets (NPLs) (Figure 1A) – plus the two Ir PCs. All three types of CdSe nanocrystals selectively form the trans-[2+2] photocycloaddition products, whereas the (Ir[dF(CF3)ppy]2(dtbpy))PF6 only drives photoredox-mediated coupling of 2 with THF solvent (14 and 15) with a total coupling yield of 22%, and Ir(ppy)3 does not generate any [2+2] or [4+2] or photoredox products. The regioselectivity and diastereoselectivity of the [2+2] coupling by the CdSe nanocrystals improves as we go from core-only QDs (R=1.2 nm) to core-shell QDs (R= 2.1 nm) to quasi-2D nanoplatelets, which are 5 monolayers (1.5 nm) thick and have lateral dimensions of 29 nm × 8 nm. Figure 4B shows the results of the hetereocoupling reactions of 2 with a 3× excess of 3. The molecular PCs are again inactive for homocoupling or heterocoupling of these substrates, while the CdSe nanocrystals selectively produce the syn-trans [2+2] heterocoupled product of 2 and 3 (with all d.r. values >20:1). The yields of the CdSe nanocrystal reactions are limited by the nanocrystals’ preferential sensitization of 3 over 2 (see Figure S4), and the remainders of the yields for these reactions are accounted for by the starting material 2 because of the large amount of the homocoupled product of 3 produced (see Figure S10). These yields would be improved substantially in a reaction where 2 were selectively photosensitized.

The poor activity of the molecular PCs in coupling of internal aryl conjugated alkenes, including stilbenes and 1,2-diarylalkenes, is observed consistently and attributed to the facile cis/trans photoisomerization of substrates28; however, here, the Ir PCs produce very little of the cis-isomer of 2. Given that (i) neither Ir PC triggers photoisomerization of 2 (see Figure S10); (ii) (Ir[dF(CF3)ppy]2(dtbpy))PF6 is able to form addition products of 2 and THF through a photoredox (Michael addition) process, and (iii) the CdSe nanocrystals are able to activate 2 for [2+2] coupling, we attribute the lack of activity of the Ir PCs to a very low yield of TTEnT from those PCs to the substrate. This low yield of TTEnT is confirmed by the lack of quenching of the PL of the Ir PCs in the presence of 2 (see Figure S3). We propose that the colloidal nanocrystals have much higher yields of TTEnT to 2 than the molecular PCs, and therefore are able to sensitize 2 for [2+2] cycloadditions, because TTEnT in the case of molecular PCs is diffusion-controlled whereas the substrate forms a quasi-static complex with the QD. TTEnT efficiency is known to scale with the PL QY47,48, so we further increase the yield of the [2+2] photocycloaddition by moving from a lower-PL QY structure, the core-only QD (QY = 0.1%), to higher-PL QY structures, the core/shell QD (QY = 77%) and NPL (QY = 80%)40,42, see Figure S1B.

The CdSe nanocrystal-catalyzed [2+2] couplings all produce vinylcyclobutanes with exclusively trans diastereoselectivity of the remaining alkenes, despite the fact that substrate 2 is more symmetric and has more potential isomerization pathways than substrate 1. We again attribute this selectivity to inhibition of the isomerization of the cyclobutane product through reversible tethering to the QD surface.

Finally, the couplings catalyzed by the CdSe (core-only) QDs have poor regioselectivity; they produce three regioisomers of the [2+2] product (11, 12, 13) resulting from different combinations of couplings of each of the two double bonds on each substrate, with 11:12:13 = 2.5:1:1. Product 11 is less strained than 12 or 13 when both carboxylates are tethered to a flat surface, but a curved or defective surface could release this strain and thereby better accommodate 12 and 13. This behavior is analogous to the formation of intramolecular defects within self-assembled monolayers at structural defect sites on otherwise flat, crystalline metal surfaces49. The fact that, as shown in Figure 4A, the more strained cyclobutane products also have poorer (syn/anti) diastereoselectivity tells us that indeed those cyclobutanes were not aligned enough to form the π-stacked, energy-minimized conformations that normally dominate the QD-templated [2+2] reactions45. We therefore attempted to simultaneously select for the least strained regioisomer and improve the overall diastereoselectivity of the reaction by increasing the radius of curvature and lowering the defect density of the CdSe particle templating the reaction. This strategy appears to work: on going from spherical core-only QDs to larger-radius core-shell QDs to a planar, low-defect density NPL with similar triplet exciton energy to the QDs (ET = 2.27 eV), we select for structure 11 (11:12:13 = 24:2:1) and improve the overall d.r. to 15:1.

Notably, the reaction accomplished by the NPLs is the first intermolecular [2+2] photocycloaddition (homo- or heterocoupling) of internal aryl conjugated dienes to form a syn-trans vinylcyclobutane derivative as the major product with high yield and stereoselectivity.

CONCLUSION

Quantum-confined semiconductor nanocrystals are remarkable in their ability to selectively access photocatalytic reaction pathways based on their own brand of stereoelectronic control. Here, we have used this control to broaden the scope of high-yield, selective [2+2] photocycloadditions to the production of synthetically and pharmaceutically relevant aryl-conjugated vinylcyclobutane products with near quantitative regio- and diastereoselectivity. We showed that the near-degeneracy of the singlet-like and triplet-like excitonic states of CdSe nanocrystals effectively decouples their triplet energies from their redox potentials, and thereby allows us to selectively access the triplet-initiated [2+2] pathway over the radical-initiated parasitic [4+2] pathway for terminal aryl conjugated dienes. We also showed that, with the high yield of TTEnT available using CdSe nanocrystal photocatalysts, we could perform selective [2+2] homo- and heterocouplings of internal aryl conjugated dienes, with regioselectivity controlled by the radius of curvature of the particle, and high selectivity for a single (syn-trans) configuration of the vinylcyclobutane product enforced by its reversible tethering to the particle surface. In fact, the reaction photocatalyzed by CdSe nanoplatelets is the first intermolecular [2+2] photocycloaddition (homo- or heterocoupling) of internal aryl conjugated dienes to selectively form a syn-trans vinylcyclobutane derivative as the major product.

Our approaches to the challenges posed by aryl conjugated dienes can be generalized to other triplet excited state-initiated photoreactions of dienes and polyenes that must compete with photoredox and photoisomerization pathways and have available a complex set of product configurations. Colloidal nanocrystals are a highly evolved and versatile set of materials, and we have just begun to explore their potential to access new photochemical reaction spaces.

EXPERIMENTAL PROCEDURES

Resource Availability

Lead Contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Emily Weiss (e-weiss@northwestern.edu).

Materials Availability

All unique reagents generated in this study will be made available on request to the Lead Contact.

Data and Code Availability

There is no dataset and code associated with the paper. Full experimental procedures are provided in the Supplemental Information.

Supplementary Material

1

ACKNOWLEDGMENTS

We thank R. Thomson, R. Qiu, and G. Schatz for helpful discussions, and C. Lin, H. Mao, R. Young and M. Wasielewski for use of their phosphorescence setup. This work was supported by the Air Force Office of Scientific Research through grant nos. FA-9550-17-1-0271 and FA-9550-20-1-0364, and made use of the IMSERC NMR facility at Northwestern University, which has received support from the Soft and Hybrid Nanotechnology Experimental (SHyNE) Resource (NSF ECCS-2025633), NIH 1S10OD012016-01 / 1S10RR019071-01A1, and Northwestern University.

Footnotes

DECLARATION OF INTERESTS

The authors declare no competing interests.

SUPPLEMENTAL INFORMATION

Supplemental Information can be found online.

REFERENCES

  • 1.Hoffmann N (2008). Photochemical Reactions as Key Steps in Organic Synthesis. Chem. Rev 108, 1052–1103. [DOI] [PubMed] [Google Scholar]
  • 2.Bach T, and Hehn JP (2011). Photochemical Reactions as Key Steps in Natural Product Synthesis. Angew. Chemie Int. Ed 50, 1000–1045. [DOI] [PubMed] [Google Scholar]
  • 3.Kärkäs MD, Porco JA, and Stephenson CRJ (2016). Photochemical Approaches to Complex Chemotypes: Applications in Natural Product Synthesis. Chem. Rev 116, 9683–9747. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Prier CK, Rankic DA, and MacMillan DWC (2013). Visible Light Photoredox Catalysis with Transition Metal Complexes: Applications in Organic Synthesis. Chem. Rev 113, 5322–5363. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Romero NA, and Nicewicz DA (2016). Organic Photoredox Catalysis. Chem. Rev 116, 10075–10166. [DOI] [PubMed] [Google Scholar]
  • 6.Zhang P, Wang Y, Bao R, Luo T, Yang Z, and Tang Y (2012). Enantioselective Biomimetic Total Syntheses of Katsumadain and Katsumadain C. Org. Lett 14, 162–165. [DOI] [PubMed] [Google Scholar]
  • 7.Alonso R, and Bach T (2014). A Chiral Thioxanthone as an Organocatalyst for Enantioselective [2+2] Photocycloaddition Reactions Induced by Visible Light. Angew. Chemie Int. Ed 53, 4368–4371. [DOI] [PubMed] [Google Scholar]
  • 8.Maturi MM, and Bach T (2014). Enantioselective Catalysis of the Intermolecular [2+2] Photocycloaddition between 2-Pyridones and Acetylenedicarboxylates. Angew. Chemie Int. Ed 53, 7661–7664. [DOI] [PubMed] [Google Scholar]
  • 9.Blum TR, Miller ZD, Bates DM, Guzei IA, and Yoon TP (2016). Enantioselective photochemistry through Lewis acid–catalyzed triplet energy transfer. Science (80-.). 354, 1391. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Hörmann FM, Chung TS, Rodriguez E, Jakob M, and Bach T (2018). Evidence for Triplet Sensitization in the Visible-Light-Induced [2+2] Photocycloaddition of Eniminium Ions. Angew. Chemie Int. Ed 57, 827–831. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Hu N, Jung H, Zheng Y, Lee J, Zhang L, Ullah Z, Xie X, Harms K, Baik M-H, and Meggers E (2018). Catalytic Asymmetric Dearomatization by Visible-Light-Activated [2+2] Photocycloaddition. Angew. Chemie Int. Ed 57, 6242–6246. [DOI] [PubMed] [Google Scholar]
  • 12.Elacqua E, Kaushik P, Groeneman RH, Sumrak JC, Bučar D-K, and MacGillivray LR (2012). A Supramolecular Protecting Group Strategy Introduced to the Organic Solid State: Enhanced Reactivity through Molecular Pedal Motion. Angew. Chemie Int. Ed 51, 1037–1041. [DOI] [PubMed] [Google Scholar]
  • 13.Wang J-S, Wu K, Yin C, Li K, Huang Y, Ruan J, Feng X, Hu P, and Su C-Y (2020). Cage-confined photocatalysis for wide-scope unusually selective [2+2] cycloaddition through visible-light triplet sensitization. Nat. Commun 11, 1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Jiang Y, Wang C, Rogers CR, Kodaimati MS, and Weiss EA (2019). Regio- and diastereoselective intermolecular [2+2] cycloadditions photocatalysed by quantum dots. Nat. Chem 11, 1034–1040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Du J, Skubi KL, Schultz DM, and Yoon TP (2014). A Dual-Catalysis Approach to Enantioselective [2+2] Photocycloadditions Using Visible Light. Science (80-.). 344, 392. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Kokubo K, Yamaguchi H, Kawamoto T, and Oshima T (2002). Substituent Effects on the Stereochemistry in the [2+2] Photocycloaddition Reaction of Homobenzoquinone Derivative with Variously Substituted Alkenes and Alkynes. J. Am. Chem. Soc 124, 8912–8921. [DOI] [PubMed] [Google Scholar]
  • 17.Lu Z, and Yoon TP (2012). Visible Light Photocatalysis of [2+2] Styrene Cycloadditions by Energy Transfer. Angew. Chemie Int. Ed 51, 10329–10332. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Hurtley AE, Lu Z, and Yoon TP (2014). [2+2] Cycloaddition of 1,3-Dienes by Visible Light Photocatalysis. Angew. Chemie Int. Ed 53, 8991–8994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Mojr V, Svobodová E, Straková K, Neveselý T, Chudoba J, Dvořáková H, and Cibulka R (2015). Tailoring flavins for visible light photocatalysis: organocatalytic [2+2] cycloadditions mediated by a flavin derivative and visible light. Chem. Commun 51, 12036–12039. [DOI] [PubMed] [Google Scholar]
  • 20.Liu D, Wang H-F, Abrahams BF, and Lang J-P (2014). Single-crystal-to-single-crystal transformation of a two-dimensional coordination polymer through highly selective [2+2] photodimerization of a conjugated dialkene. Chem. Commun 50, 3173–3175. [DOI] [PubMed] [Google Scholar]
  • 21.Wang Y, Wang M-F, Young DJ, Zhu H, Hu F-L, Mi Y, Qin Z, Chen SL, and Lang J-P (2021). Tuning the configuration of the flexible metal–alkene-framework affords pure cycloisomers in solid state photodimerization. Chem. Commun, DOI: 10.1039/D0CC06939F. [DOI] [PubMed] [Google Scholar]
  • 22.Dembitsky VM (2008). Bioactive cyclobutane-containing alkaloids. J. Nat. Med 62, 1–33. [DOI] [PubMed] [Google Scholar]
  • 23.Gutekunst WR, Gianatassio R, and Baran PS (2012). Sequential C-H Arylation and Olefination: Total Synthesis of the Proposed Structure of Pipercyclobutanamide A. Angew. Chemie Int. Ed 51, 7507–7510. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Namyslo JC, and Kaufmann DE (2003). The Application of Cyclobutane Derivatives in Organic Synthesis. Chem. Rev 103, 1485–1538. [DOI] [PubMed] [Google Scholar]
  • 25.Seiser T, Saget T, Tran DN, and Cramer N (2011). Cyclobutanes in Catalysis. Angew. Chemie Int. Ed 50, 7740–7752. [DOI] [PubMed] [Google Scholar]
  • 26.Liu RSH, Turro NJ, and Hammond GS (1965). Mechanisms of Photochemical Reactions in Solution. XXXI. Activation and Deactivation of Conjugated Dienes by Energy Transfer. J. Am. Chem. Soc 87, 3406–3412. [Google Scholar]
  • 27.Hammond GS, Turro NJ, and Liu RSH (1963). Mechanisms of Photochemical Reactions in Solution. XVI. 1 Photosensitized Dimerization of Conjugated Dienes. J. Org. Chem 28, 3297–3303. [Google Scholar]
  • 28.Poplata S, Tröster A, Zou Y-Q, and Bach T (2016). Recent Advances in the Synthesis of Cyclobutanes by Olefin [2+2] Photocycloaddition Reactions. Chem. Rev 116, 9748–9815. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Wender PA, and Correia CRD (1987). Intramolecular photoinduced diene-diene cyaloadditions: a selective method for the synthesis of complex eight-membered rings and polyquinanes. J. Am. Chem. Soc 109, 2523–2525. [Google Scholar]
  • 30.Liu D, Ren Z-G, Li H-X, Lang J-P, Li N-Y, and Abrahams BF (2010). Single-Crystal-to-Single-Crystal Transformations of Two Three-Dimensional Coordination Polymers through Regioselective [2+2] Photodimerization Reactions. Angew. Chemie Int. Ed 49, 4767–4770. [DOI] [PubMed] [Google Scholar]
  • 31.Park I-H, Sasaki K, Quah HS, Lee E, Ohba M, Lee SS, and Vittal JJ (2019). Metal Organo-Polymeric Framework via [2+2] Cycloaddition Reaction: Influence of Hydrogen Bonding on Depolymerization. Cryst. Growth Des 19, 1996–2000. [Google Scholar]
  • 32.Wang M-F, Mi Y, Hu F-L, Niu Z, Yin X-H, Huang Q, Wang H-F, and Lang J-P (2020). Coordination-Driven Stereospecific Control Strategy for Pure Cycloisomers in Solid-State Diene Photocycloaddition. J. Am. Chem. Soc 142, 700–704. [DOI] [PubMed] [Google Scholar]
  • 33.Hu F-L, Mi Y, Zhu C, Abrahams BF, Braunstein P, and Lang J-P (2018). Stereoselective Solid-State Synthesis of Substituted Cyclobutanes Assisted by Pseudorotaxane-like MOFs. Angew. Chemie Int. Ed 57, 12696–12701. [DOI] [PubMed] [Google Scholar]
  • 34.Mandal R, Garai M, and Biradha K (2017). Hydrogen-bonded Two-fold Interpenetrated Diamondoid Networks for Solid-State [2 + 2] Polymerizations of Criss-crossed Olefins: Molecular Connections vs Supramolecular Connections. Cryst. Growth Des 17, 5061–5064. [Google Scholar]
  • 35.Caputo JA, Frenette LC, Zhao N, Sowers KL, Krauss TD, and Weix DJ (2017). General and Efficient C-C Bond Forming Photoredox Catalysis with Semiconductor Quantum Dots. J. Am. Chem. Soc 139, 4250–4253. [DOI] [PubMed] [Google Scholar]
  • 36.Zhu X, Lin Y, San Martin J, Sun Y, Zhu D, and Yan Y (2019). Lead halide perovskites for photocatalytic organic synthesis. Nat. Commun 10, 2843. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Zhang Z, Rogers CR, and Weiss EA (2020). Energy Transfer from CdS QDs to a Photogenerated Pd Complex Enhances the Rate and Selectivity of a Pd-Photocatalyzed Heck Reaction. J. Am. Chem. Soc 142, 495–501. [DOI] [PubMed] [Google Scholar]
  • 38.Yuan Y, Zhu H, Hills-Kimball K, Cai T, Shi W, Wei Z, Yang H, Candler Y, Wang P, He J, et al. (2020). Stereoselective C-C Oxidative Coupling Reactions Photocatalyzed by Zwitterionic Ligand Capped CsPbBr3 Perovskite Quantum Dots. Angew. Chemie Int. Ed 59, 22563–22569. [DOI] [PubMed] [Google Scholar]
  • 39.Mongin C, Garakyaraghi S, Razgoniaeva N, Zamkov M, and Castellano FN (2016). Direct observation of triplet energy transfer from semiconductor nanocrystals. Science (80-.). 351, 369–372. [DOI] [PubMed] [Google Scholar]
  • 40.Greytak AB, Allen PM, Liu W, Zhao J, Young ER, Popović Z, Walker BJ, Nocera DG, and Bawendi MG (2012). Alternating layer addition approach to CdSe/CdS core/shell quantum dots with near-unity quantum yield and high on-time fractions. Chem. Sci 3, 2028–2034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Lai R, Sang Y, Zhao Y, and Wu K (2020). Triplet Sensitization and Photon Upconversion Using InP-Based Quantum Dots. J. Am. Chem. Soc [DOI] [PubMed] [Google Scholar]
  • 42.Flamee S, Cirillo M, Abe S, De Nolf K, Gomes R, Aubert T, and Hens Z (2013). Fast, high yield, and high solid loading synthesis of metal selenide nanocrystals. Chem. Mater 25, 2476–2483. [Google Scholar]
  • 43.Inamdar SN, Ingole PP, and Haram SK (2008). Determination of Band Structure Parameters and the Quasi-Particle Gap of CdSe Quantum Dots by Cyclic Voltammetry. ChemPhysChem 9, 2574–2579. [DOI] [PubMed] [Google Scholar]
  • 44.Lin S, Ischay MA, Fry CG, and Yoon TP (2011). Radical Cation Diels–Alder Cycloadditions by Visible Light Photocatalysis. J. Am. Chem. Soc 133, 19350–19353. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Jones LO, Mosquera MA, Jiang Y, Weiss EA, Schatz GC, and Ratner MA (2020). Thermodynamics and Mechanism of a Photocatalyzed Stereoselective [2+2] Cycloaddition on a CdSe Quantum Dot. J. Am. Chem. Soc 142, 15488–15495. [DOI] [PubMed] [Google Scholar]
  • 46.Fritzinger B, Capek RK, Lambert K, Martins JC, and Hens Z (2010). Utilizing Self-Exchange To Address the Binding of Carboxylic Acid Ligands to CdSe Quantum Dots. J. Am. Chem. Soc 132, 10195–10201. [DOI] [PubMed] [Google Scholar]
  • 47.Huang Z, and Lee Tang M (2018). Semiconductor Nanocrystal Light Absorbers for Photon Upconversion. J. Phys. Chem. Lett 9, 6198–6206. [DOI] [PubMed] [Google Scholar]
  • 48.Okumura K, Mase K, Yanai N, and Kimizuka N (2016). Employing Core-Shell Quantum Dots as Triplet Sensitizers for Photon Upconversion. Chem. – A Eur. J 22, 7721–7726. [DOI] [PubMed] [Google Scholar]
  • 49.Vericat C, Vela ME, and Salvarezza RC (2005). Self-assembled monolayers of alkanethiols on Au(111): surface structures, defects and dynamics. Phys. Chem. Chem. Phys 7, 3258–3268. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1676176-supplement-1.pdf (12.6MB, pdf)

Data Availability Statement

There is no dataset and code associated with the paper. Full experimental procedures are provided in the Supplemental Information.

RESOURCES